Cell motility is a prerequisite for the creation of new life, and it is required for maintaining the integrity of an organism. Under pathological conditions, “too much” motility may cause premature death. Studies over the past few years have revealed that ion channels are essential for cell motility. This review highlights the importance of K+ channels in regulating cell motility.
Cell motility is involved in many (patho-)physiological processes, and K+ channels are important regulators of mechanisms underlying cell motility.
The ability of cells to move is a fundamental aspect of life. This holds true for unicellular as well as for multicellular organisms. Consequently, some of the mechanisms underlying cell motility were highly conserved throughout evolution. Thus studying the motility of amebae, ciliates (such as Tetrahymena), or seaurchin sperm can give insights into mechanisms employed by mammalian cells. Two different modes of cell motility will be considered in this review: a swimming movement driven by ciliary or flagellar beating and a crawling or ameboid movement, which will be referred to as cell migration hereafter. Both forms are found in mammals, including humans.
In mammals, cell movement by flagellar beats is restricted to sperm cells. These cells employ a flagellum to swim up the fallopian tube in search of an oocyte (26). In contrast, the ameboid type of movement is used by a wide spectrum of cell types and it is the basis for many physiological and pathophysiological processes. It starts early during embryogenesis (8), and in the brain it does not stop until after birth (108). Movement of fibroblasts or epithelial cells during wound healing also involves the ameboid-type of migration (20). Other processes strongly relying on cell migration are the responses of the immune system, angiogenesis (95), and the formation of tumor metastases (103). These examples indicate that altered function of migrating cells can be associated with severe, possibly life-threatening pathophysiological conditions, for instance tumor metastasis.
We will begin this review on the function of K+ channels in cell motility with short overviews of mechanisms underlying ciliary beating and ameboid cell migration (58, 76, 78, 102). These sections will be followed by a brief description of the role of K+ channels in the motility of ciliates and sperm cells and by a discussion of the impact of K+ channels in ameboid cell migration, respectively. Other types of ion channels, aquaporins, or transporters will not be dealt with systematically. We instead refer to recent reviews that describe the role of these transport proteins in cell motility in detail (34, 64, 82, 84, 93).
Swimming Controlled by Ciliary or Flagellar Beating
Cilia and flagella are hair-like, motile organelles projecting from the cell surface. Cilia are shorter than flagella and are present in large numbers on the cell surface. Their function has traditionally been studied in unicellular organisms like Tetrahymena or Paramecium. In mammals, cilia are found among others on the apical membrane of epithelial cells covering the airways, where they are required for mucociliary clearance (55), whereas flagella are only found in spermatozoa. Undulatory beating of the single flagellum propels sperm cells forward.
Cilia and flagella are usually composed of nine peripheral doublet microtubulues with two central microtubules. This typical “9 + 2” axoneme runs longitudinally through the entire organelle (78). Harboring the same 9 + 2 ultrastructure (78), it is not surprising that motile cilia and flagella share the same mechanism of movement, which is driven by dynein motor molecules. The interconnection of peripheral doublets and the central pair of microtubules converts the sliding movements of the peripheral doublets into a bending motion (78). The ciliary/flagellar beat frequency determines the speed of swimming or creeping on a substrate, whereas the direction of movement is determined by the beat direction and the orientation of the cilia/flagella. The frequency and the orientation of beating are tuned by the local Ca2+ concentration inside the cilium/flagellum (59, 67). An increase in the Ca2+ concentration from 100 nM to 1 μM leads to a gradual acceleration of the forward movement because of a proportionally increasing ciliary beating frequency. As soon as the Ca2+ concentration exceeds a threshold value of 1 μM the direction of beating is reversed and the cell moves backward.
In contrast to motile cilia, single primary cilia, which are found on almost all mammalian cells, lack the two central microtubules (9 + 0) and are therefore usually immotile. They are signaling platforms capable of chemo- and mechanosensing (78). Primary cilia gained much attention recently when their involvement in a number of different pathologies including polycystic kidney disease was discovered (78, 106). Primary cilia respond to mechanical stimulation such as bending by inducing an increase of the intracellular Ca2+ concentration. This mechanical response is mediated by TRPP2 channels that are located in the ciliary membrane (71). So far, it is unknown whether ion channels located within the membrane of primary cilia or activated by the mechanosensory function of primary cilia also affect cell motility.
Potassium Channels Modulate the Motility of Ciliates
It has been known for more than 30 years that the membrane potential is crucial in controlling the motility of ciliated or flagellated cells (24). The membrane potential of ciliates can be modified among other means by mechanical (40), chemical (90), or temperature stimuli (91). A sequence of events elicited by mechanical stimuli, such as the indentation of the plasma membrane or cell/cell collisions, is shown in FIGURE 1⇓. The activation of mechanosensitive Ca2+ channels (40) induces a depolarization, thereby activating voltage-gated L-type Ca2+ channels in the ciliary/flagellar membrane (41) and eventually eliciting an increase of the intraciliary Ca2+ concentration. This in turn initiates accelerated forward or even backward movement. In many ciliated cells, the rise of the intraciliary Ca2+ concentration also activates Ca2+-sensitive K+ channels. The ensuing repolarization of the membrane potential marks the termination of the accelerated forward or backward movement (4). The K+ conductance of ciliates is affected by chemoattractants and pheromones and can, via activation or deactivation of voltage-gated Ca2+ channels, induce characteristic motility patterns that range from moving rapidly back and forth to remaining on the spot (90). So far, there are only few reports on the molecular identification of K+ channels in ciliates. Apparently, Paramecium expresses a large number of different K+ channels that are related to the mammalian Eag channels (31). Transgenic expression of two of them, PAK1 and PAK11, led to the reduction of two separate Ca2+-dependent K+ currents and to a modified swimming behavior (46).
K+ Channels in Sperm Cell Motility
In contrast to ciliates, the molecular identity of K+ channels is well defined in sperm cells (5, 15). K+ channels are required for sperm motility (16). One important function of sperm K+ channels is volume regulation. When sperm is ejaculated into the female reproductive tract, it is surrounded by fluids with a 15-20% lower osmolarity than in the distal epididymis. Thus sperm cells have to regulate ion fluxes to maintain their original volume. Different K+ channels such as Kv1.5, Kv7.1, or TASK2 have been implicated in this process (5). Volume regulation has more than a housekeeping function, since human sperm cells fail to move efficiently in the swollen state (104).
Sperm is chemotactically guided to the egg it is about to fertilize. This holds true for both marine organisms such as sea urchins whose sperm is released into water and so-called “internal fertilizers” (e.g., mammals) whose sperm is deposited in the female reproductive tract (26). Sperm chemotaxis is a Ca2+-dependent process that involves several different voltage-gated Ca2+ channels. Ca2+ influx is regulated by the membrane potential, which brings K+ channels into play. For example, chemotactic activation of sea urchin sperm (Arbacia punctulata) is mediated by the initial activation of cGMP-gated K+ channels, which then sets the stage for Ca2+ influx (94). The K+ channel-dependent hyperpolarization allows recovery of voltage-gated Ca2+ channels from inactivation so that these channels can open when the membrane potential depolarizes after the opening of hyperpolarization-activated and cyclic nucleotide-gated (HCN) channels. A cyclic nucleotide-gated K+ channel, Sp-tetraKCNG, was cloned from sea urchin sperm, which is insofar unique because it has a similar membrane topology as voltage-gated Na+ or Ca2+ channels. Sp-tetraKCNG channels are also composed of a single polypeptide with four KCNG domains, each of which has six transmembrane domains (28). So far, the signaling underlying mammalian sperm chemotaxis is not well understood. Hence, the exact role of K+ channels in this process remains to be elucidated in the future (36).
Mechanisms of Ameboid Movement
The process of migration can be subdivided into two basic steps that are repeated over and over again: protrusion of the front and retraction of the rear part. Migrating cells form at their front a fan-like, 300-nm-thin, and organelle-free process, the so-called lamellipodium. The cell body and the uropod are dragged behind. Migrating cells are polarized in the direction of movement (58) since some components of the cellular migration apparatus are restricted to either one of the two opposing cell poles. Thus actin polymerization and depolymerization play particularly important roles in the lamellipodium (44, 102). Several mechanisms underlie polarity of migrating cells and allow persistent movement in a given direction, even in the absence of external cues (19). One of these involves the Na+/H+ exchanger NHE1. It tethers the actin cytoskeleton via ERM proteins to the leading edge of the lamellipodium (18), and by generating a characterisitc pH nanoenvironment within the glycocalyx, NHE1 generates an asymmetric adhesiveness to the extracellular matrix (92, 96). Accordingly, NHE1-deficient cells have a severely impaired ability to migrate persistently in one direction (17, 86, 96). Another “ionic” mechanism of cell polarity is the generation of a gradient of the intracellular Ca2+ concentration ([Ca2+]I) with [Ca2+]I being higher at the rear than at the front of many migrating cells (51). In chemotactically stimulated amebae or neutrophil granulocytes, lipid metabolites generated by phosphatidylinositol-3 kinase (PI3K), PTEN (phosphatase and tensin homolog in chromosome 10), and phospholipase A2 signaling pathways play crucial roles in controlling directed migration and maintaining polarity (11, 54). The regulation of the opposing cellular events at the front and at the rear part of a migrating cell also involves members of the Rho family of GTPases. In particular, RhoA-C, Rac1–3, and Cdc42 have been implicated in coordinating the spatially distinct cytoskeleton reorganizations underlying migration (35, 74). The intracellular network regulating cytoskeletal remodeling during migration is further complicated by signals originating from focal adhesion contacts, where integrins that are linked to the cytoskeleton interact with extracellular matrix molecules (7, 76). As we will see, the function of integrins and thereby cell adhesion and migration can also be modulated by K+ channels. Finally, primary cilia may function as a cellular “GPS” (13) and control directed migration since they are oriented in the direction of movement (1).
K+ Channels are Part of the Cellular Migration Machinery
Many studies have unequivocally shown that K+ channels belonging to different subfamilies are important components of the cellular migration machinery. When these channels are inhibited, migration is impaired. Studies from the last 10–15 years indicated several common themes by which K+ channels can influence the process of cell migration (see FIGURE 2⇓). Effects related to housekeeping functions of K+ channels, such as regulating the cell membrane potential or cell volume, have to be considered as well as nonconductive properties of K+ channels, i.e., effects that appear to be unrelated to ion transport itself.
K+ channels modulate migration by controlling the cell membrane potential
The most prominent housekeeping mechanism is the regulation of the cell membrane potential. In doing so, K+ channels control the driving force for the electrogenic transport of ions across the plasma membrane (FIGURE 2A⇑). Depolarizing the cell membrane potential by elevating the extracellular K+ concentration inhibits migration of transformed epithelial cells (88). The control of the electrical driving force is of particular importance for Ca2+ ions since migration is a Ca2+-dependent process (51, 68, 108). The intracellular Ca2+ concentration serves both as a temporal as well as a spatial regulator of cell migration coordinating the retraction of the rear part of migrating cells (22, 25, 43) with forward protrusion (27). The membrane potential has a dual effect on Ca2+ influx through Ca2+ channels. On one hand, it sets the electrical driving force for all Ca2+ channels, but on the other hand it also controls the gating behavior of voltage-gated Ca2+ channels. Which one of these two effects prevails depends on the expression pattern of Ca2+ channels in the respective cell type. Most peripheral migrating cells usually express voltage-independent Ca2+ channels whose molecular identity, however, is poorly defined. In contrast, voltage-gated Ca2+ channels are frequently required for migration of cells from the central nervous system (47, 84).
The membrane potential also determines the electrochemical driving force for the flux of Na+ and Cl− through their respective channels, and these fluxes have also been shown to play a role in migration (see Ref. 84 and references therein). Many tumor cells express functional, tetrodotoxin-inhibitable voltage-gated Na+ channels that have a low activity at the resting membrane potential of these tumor cells (21, 47, 75).
Galvanotaxis, i.e., migration within an electric field, is another mechanism that links the cell membrane potential and hence K+ channels to migration. It has been known for a long time that external electric fields can determine the direction of cell migration. Such electric fields are for example generated in a developing embryo or between intact and damaged parts of an epithelial layer (52). Similarly, single cells are surrounded by an electric field that may impose a signal onto a neighboring cell. In addition to signaling molecules employed by chemotactic stimuli (107), K+ channels (Kv1.3) and other ion channels such as TRPC1 and voltage-gated Na+ channels were also proposed to act as antennas that measure electrical fields and thereby affect cell migration (38, 57).
The recent discovery of the voltage-sensitive phosphoinositide phosphatase Ci-VSP points to yet another mechanism by which K+ channels could indirectly modify migration (56), although this possibility has not yet been shown experimentally. The electrical signal generated by K+ channel activation (i.e., hyper-polarization) or inhibition (i.e., depolarization) is translated into a modulation of the PtdIns(4,5)P2 (PIP2)-to-PtdIns(3,4,5)P3 (PIP3) ratio. The potential importance of such voltage-dependent phosphatase activity in cell migration is given by the fact that PIP2 and PIP3 are key regulators of actin dynamics (105) and chemotaxis, respectively (11, 54).
K+ channels modulate migration by regulating the cell volume
Most migrating cells undergo dramatic changes of their shape while moving. This is in part due to the fact that the retraction of the rear part of migrating cells does not always keep pace with the protruding front. It is a common observation that the rear part of migrating cells is stuck for some time while the lamellipodium keeps growing until the rear part eventually “catches up” rapidly. Thus one could model the process of cell migration as intermittent local swelling at the front and shrinkage at the rear part. A local swelling of the protruding lamellipodium was shown in EGF-stimulated adenocarcinoma cells (77). The intracellular “valve” restricting volume changes either to the front or to the rear part of migrating cells is represented by the poroelastic nature of the cytosol (10). Local swelling and shrinkage may also become a necessity for migrating cells when they are obliged to squeeze through narrow spaces of the dense fiber meshwork of the extracellular matrix (FIGURE 2B⇑). This was, for instance, shown for glioma cells invading brain slices (53). Being capable of (locally) changing the volume when needed is therefore advantageous for efficient cell migration. In addition to the above mentioned “mechanical” aspects, the cell volume also plays a homeostatic role in migrating cells. The “correct” cell volume is crucial for the integrity of the cytoskeleton. Cell swelling leads to actin depolymerization, whereas cell shrinkage promotes actin polymerization (66, 87). Since K+ channels are central players in cell volume regulation (33), they create an optimal intracellular environment required for the cytoskeletal migration machinery. Notably, the functional interrelation between K+ channels, cell volume, and the cytoskeleton may be mutual. On the one hand, K+ channels control the cell volume and thereby the actin cytoskeleton. But on the other hand, the actin cytoskeleton may also regulate K+ channel activity (9, 98, 101).
Nonconducting functions of K+ channels
Finally, K+ channels are known to interact directly with proteins that are crucial for cell migration such as integrins (2, 3, 12, 45), cortactin (98, 101), and focal adhesion kinase (FAK) (73). The functional interaction between K+ channels and these proteins is mutual, so that, for example, signaling of integrins can be inhibited by K+ channel blockers, and integrin activation leads to increased K+ channel activity (FIGURE 2C⇑). In addition to G-protein- and tyrosine kinase-mediated signaling, conformational coupling was suggested to account for mutual activation (2).
Contribution of Individual K+ Channel Families to Cell Migration
Ca2+-activated K+ channels
Several members of this family, BK (KCa1.1) (39, 100), SK3 (KCa2.3) (70), and in particular IK1 (KCa3.1) channels (14, 23, 48, 79, 85, 89, 97), are involved in cell migration. In most preparations, inhibition of KCa channels slows migration. However, there are also exceptions pointing toward the importance of the cellular context. When KCa3.1 channels are inhibited in secretory colonic epithelial cells, the restitution of the epithelium is accelerated (48). This effect possibly represents a K+ channel-dependent modification of galvanotaxis during wound closure (52). The contribution of different calcium-sensitive K+ channels is cell-type specific. BK (KCa1.1) channels, for example, are expressed in glioma and microglial cells. Yet, they are only needed for migration of glioma cells but not of microglia (79, 100).
The Ca2+ sensitivity of Ca2+-activated K+ channels provides an excellent means to coordinate their activity and thereby downstream effects such as membrane potential or cell volume with other Ca2+-dependent processes such as cytoskeletal dynamics. This coordination is further fine tuned by spatial gradients and temporal fluctuations of [Ca2+]i. Many migrating cells exhibit an oscillating [Ca2+]i, with resting levels being the highest in the rear part of migrating cells (51). Consequently, KCa3.1 channels are apparently only activated at this cell pole (83, 89), although the density of KCa3.1 channel proteins is the highest at the cell front (60). Fluctuating KCa3.1 channel activity is critical for optimal cell migration since continuously activating KCa channels impair migration (39, 87). Being intermittently activated, KCa channels support migration by inducing localized and transient changes of the cell volume (53, 81). In doing so, they act in concert with volume-sensitive Cl− channels whose molecular identity is discussed controversially. Evidence has been provided for the involvement of volume-regulated anion channels (VRAC) (50, 80) or ClC3 channels (53) in cell migration. KCa2.3 channels were proposed to regulate migration by modulating the driving force for Ca2+ influx (70).
Voltage-gated Kv1.3 channels
Kv1.3 channels have a strong impact on integrin function. Kv1.3 channels physically interact with β1-integrins in lymphocytes (45) and melanoma cells (3) and thereby affect cell adhesion and migration (61). The interaction of Kv1.3 channels and integrins becomes closer on cell adhesion, and it is weakened by Kv1.3 channel blockers. Opening of Kv1.3 channels, not necessarily accompanied by K+ flow, results in an activation of β1-integrins (45), i.e., a depolarization of the cell membrane potential is transduced, via Kv1.3 channels, to an activation of β1-integrins. On the other hand, Kv1.3 channels are stimulated upon integrin activation. The reciprocal relationship is taken as an indication of conformational coupling of both proteins since voltage-dependent gating of Kv1.3 channels and adhesion-dependent activation of integrins are both associated with conformational changes (FIGURE 2C⇑). Thus oscillations of the cell membrane potential of migrating cells could lead to intermittent activation of integrins and downstream signaling pathways. At this point, it remains highly speculative whether the Kv1.3-integrin interaction could thereby contribute to other oscillatory, actin-mediated processes such as a periodic force generation of neutrophils (30).
Kv1.3 channels are clustered with TRPC1 channels at the leading edge of the lamellipodium of neutrophils (38). The functional significance of this co-clustering was seen in the detection of electrical fields. Modeling suggests that changes of the electric field across the plasma membrane following Kv1.3 channel blockade or an elevation of the extracellular K+ concentration could also affect neighboring integrins and possibly contribute to the observed changes of integrin activity (62).
Kv10.1 (Eag1) and Kv11.1 (Erg1) channels
These voltage-gated K+ channels are important regulators of tumor cell proliferation and migration. Expression of Eag1 and ERG1 channels is upregulated in many tumor cells and is negatively correlated with patient prognosis (29, 42, 63, 69). ERG1 channels also form complexes with β1-integrins in which both proteins reciprocally activate each other. ERG1 channels are activated upon integrin-mediated adhesion to fibronectin. On the other hand, integrin-signaling, such as the tyrosine phosphorylation of the focal adhesion kinase, depends on ERG1 channel activity (12). Thus ERG1 channels and β1-integrins modulate each other’s function, presumably in a nonconductive way by conformational coupling. Functionally, they are situated at a similar position as Kv1.3 channels and strongly affect migration by controlling cell-matrix interactions and downstream signaling cascades.
Other potassium channels
Kv1.1 and Kv1.5 channels accelerate the restitution of wounded epithelia by hyperpolarizing the cell membrane potential and thereby maintaining the electrochemical driving force for Ca2+ influx (72). Kv1.5 and Kv3.1 channels are part of the cellular migration machinery of human alveolar macrophages (65) and of embryonic acoustico-vestibular neuroblasts (32), respectively. Furthermore, Kv7.1 and KATP channels have been shown to be part of the EGF-signaling cascade leading to wound healing of alveolar epithelial cells (99). Such cross talk between growth factor or chemokine signaling cascades and K+ channels has also been observed in other migrating cell types (14, 37, 79).
For many years, research on the role of K+ channels in cell motility focused on unicellular, ciliated model organisms. Long before the requirement of K+ channel activity for ameboid migration was recognized, the involvement of K+ channels in the movement of ciliates was firmly established. However, studies over the last 10–15 years have revealed that K+ channels and other ion channels and transporters are essential components of the cellular migration machinery. The number of cell types demonstrated to have K+ channel-dependent ameboid migration is steadily growing. Yet, the field is still too young to draw a conclusive picture. Many migrating cells express more than one “migration-relevant” K+ channel. However, so far it is not known to what extent these different K+ channels cooperate with each other and/or with other ion channels and transporters in a given cell type in modulating cell migration. Moreover, it has to be kept in mind that K+ channels involved in cell migration are expressed by a wide variety of migrating cells performing different physiological tasks. One of the challenges for the next years will be the development of animal models in which the therapeutic potential of modulating K+ channel activity in migrating cells under defined pathophysiological conditions can be explored.
Since migration is central to the pathophysiology of several devastating diseases such as cancer or chronic inflammatory disorders, K+ channel blockade represents a new therapeutic strategy for the treatment of these diseases. This concept is well advanced in the case of autoimmune diseases such as multiple sclerosis where Kv1.3 channel blockade is very effective (6). Eag and KCa3.1 channels are potential drug targets for the treatment of cancer. Both channels are upregulated in cancer cells, and both K+ channels are involved in cell migration, which is a critical step in the metastatic cascade. In addition, they also regulate cell proliferation. Thus blocking Eag or KCa3.1 channels in cancer cells would inhibit two central processes of tumor cell physiology. Interestingly, K+ channels are not the only ion channels whose therapeutic potential for the treatment of cancer is currently being investigated. Voltage-gated Na+ channels (57) and chlorotoxin-sensitive Cl− channels are potential drug targets for tumor therapy too. The latter already entered phase II clinical trials for the treatment of glioblastoma (49). In these trials, chlorotoxin served as a carrier to specifically bring antitumor drugs to the gliomas. These examples are a strong motivation to continue research in this exciting, medically highly relevant field.
We apologize to all colleagues working in the field whose studies could not be cited due to space limitations. We greatly acknowledge the contributions of former and present members of our laboratory.
Work from our laboratory was supported by grants from Deutsche Forschungsgemeinschaft, IZKF Münster, and Innovative Medizinische Forschung Münster.
- © 2008 Int. Union Physiol. Sci./Am. Physiol. Soc.