An Alternative to Conventional Immunosuppression: Small-Molecule Inhibitors of K_v1.3 Channels

A key step in the generation of an immune response, either physiological or pathophysiological (e.g., autoimmune), is the recognition of the antigen presented to the lymphocytes followed by their concomitant activation and clonal proliferation. The activation of T lymphocytes through the T-cell antigen receptor (TCR) initiates several downstream signaling pathways leading to the transcription of genes required for lymphocyte proliferation (e.g., interleukin-2 gene). One axis of downstream signaling is the phospholipase C-γ (PLCγ)-1,4,5-inositol trisphosphate (IP₃)-Ca²⁺ pathway (1). IP₃ releases Ca²⁺ from cytosolic Ca²⁺ stores, and this Ca²⁺ along with the Ca²⁺ influx through the plasma membrane generates a sustained elevation of the cytosolic free Ca²⁺ concentration, [Ca²⁺]_i, required for efficient signal transduction. The calcium signal activates the Ca²⁺–calmodulin–dependent phosphatase calcineurin, which then dephosphorylates the transcription factor nuclear factor of activated T cells (NF-AT), thereby enabling it to accumulate in the nucleus and bind to the promoter element of the IL-2 gene (2). Conventional immunosuppressors, such as cyclosporine A and tacrolimus (FK506), interfere with lymphocyte activation by forming drug–immunophilin–calcineurin complexes, which inhibit the phosphatase activity of calcineurin, thereby preventing NF-AT translocation to the nucleus (3). These calcineurin inhibitors are leading compounds in the management of immune reactions during organ transplantation; however, their therapeutic use is associated with significant nephrotoxicity and an increased risk of infections and cancer(3).

The discovery (4) and the characterization of ion channels in T lymphocytes (5) recently opened a new opportunity for pharmacological interference with T cell activation (6, 7). The most abundant, as well as physiologically most important, voltage- gated K⁺ channel in human T lymphocytes, K_v1.3, opens upon membrane depolarization with an activation threshold close to the resting potential of the cells. The Ca²⁺-activated potassium channel of human T cells, K_Ca3.1 (formerly termed IKCa1), is activated solely by the rise of the cytosolic free calcium concentration to over ~200 nM, independently of the membrane potential. The calcium sensor is calmodulin that is permanently associated with the channel. These two types of K⁺ channels have similar structure; they are composed of four individual subunits, each subunit consisting of six transmembrane-spanning helices and the connecting intra- and extracellular loops. The extracellular loops connecting the fifth and sixth transmembrane helices from each subunit construct the extracellular vestibule of the channel, which serves as the receptor site for many blockers (see below), and also form the selectivity filter for K⁺ (8). K_v1.3 and K_Ca3.1 potassium channels have been identified in a variety of lymphocytes, and these channels along with the Ca²⁺ release–activated Ca²⁺ (CRAC) channel, the encoding genes of which have not been found yet (1, 7), regulate the Ca²⁺ influx through the plasma membrane. The membrane potential–independent CRAC channel is activated by the emptying of cytosolic Ca²⁺ stores by IP₃, and the electrical driving force for sustained Ca²⁺ entry is provided by the negative membrane potential maintained by the activity of K_v1.3 and K_Ca3.1 potassium channels (7). From the membrane potential– dependence of the Ca²⁺ signal one could predict that blockers of K⁺ channels can inhibit the proliferation of T cells, and therefore, suppress immune reactions. Accordingly, early studies reported that depolarization of T cells by specific blockers of K_v1.3 was sufficient to inhibit human T cell activation in vitro (9). Other in vivo studies done over the last decade suggested that K_v1.3 blockers caused generalized immunosuppression based on the ability of K_v1.3 blockers [e.g., margatoxin, correolide, and kaliotoxin (Table 1⇓)] to suppress delayed-type hypersensitivity (DTH) in mini-swine (10, 11). However, the cardinal issues in immunosuppression, namely, selective targeting of specific autoreactive responses without compromising essential immune functions and minimizing the side effects of the therapy, remained to be solved for K⁺ channel inhibitors.

Several recent findings have revolutionized the outlook for K⁺ channel inhibitors as potential immunosuppressors. One set of key observations described the specific expression level of K_v1.3 channels in T lymphocyte subsets (5, 12), whereas structure-based drug design has resulted in the production of high-affinity and high-specificity blockers of K_v1.3 channels (13).

All quiescent human T lymphocytes express ~200 to 400 K_v1.3 channels and 8–10 K_Ca3.1 channels per cell, so the membrane potential of these cells is controlled by K_v1.3 channels. Upon activation by mitogens or antigens, the number of K_Ca3.1 and K_v1.3 channels increases differentially depending on the nature of the T cell subset (5). Naïve and central memory T cells (T_CM) (Box 1) transcriptionally increase the expression of K_Ca3.1 channels; the number of channels increases up to ~500 per cell whereas the number of K_v1.3 channels increases only modestly. Thus, these cells acquire a K_Ca3.1^highK_v1.3^low phenotype. On the other hand, antigenic stimulation of effector memory T cells (T_EM) increases the number of K_v1.3 channels dramatically (~1500 / cell) without any increase in the number of K_Ca3.1 channels. The ionchannel phenotype of T_EM cells is therefore K_Ca3.1^lowK_v1.3^high. The difference in the ion-channel phenotype allows the proliferation of these cells to be modulated by K_v1.3- or K_Ca3.1-specific inhibitors. Activation of quiescent cells can primarily be inhibited by K_v1.3 inhibitors, but not by K_Ca3.1 blockers. Because the transcriptional upregulation of K_Ca3.1 channels in naïve and T_CM cells depends somewhat on the Ca²⁺ signal, the augmentation of the number of K_Ca3.1 channels continues even if the initial activation of these cells is inhibited by K_v1.3 channel blockers. Thus, control of the membrane potential is taken over by K_Ca3.1 channels in activated naïve and T_CM cells (both of which are K_Ca3.1^highK_v1.3^low phenotype) and subsequent proliferation of these T cell subsets escapes further K_v1.3 channel inhibition and becomes sensitive to K_Ca3.1 channel blockade. The proliferation of activated T_EM cells, however, continues to be sensitive to K_v1.3 blockade, because these cells express large amounts of K_v1.3 channels. The use of selective K_v1.3 inhibitors therefore allows the inhibition of the proliferation of T_EM cells, with minimal interference with the activation of naïve and T_CM cells.

What is the significance of these findings? Beeton and colleagues report that in experimental autoimmune encephalomyelitis (EAE), which is a rat model for the human disease multiple sclerosis (MS), chronically activated encephalogenic T cells display an unusual channel phenotype. These repeatedly activated myelin basic protein–specific rat T cells (i.e., their TCRs specifically recognize and bind myelin basic protein) express a large number of K_v1.3 channels and a low number of K_Ca3.1 channels (12). Consistent with the channel phenotype, the proliferation of these cells can be inhibited by a K_v1.3 specific blocker, ShK-Dap²² toxin (14) (Table 1⇓), a derivative of the ShK toxin originally isolated from the sea anemone Stichodactyla helianthus. In vivo experiments also demonstrate, that the symptoms of EAE could be greatly ameliorated by the application of the K_v1.3-specific blocker ShKDap²² alone or in combination with a K_Ca3.1 blocker (TRAM-34). Very recently, Chandy and colleagues have found that myelin-reactive T_EM cells isolated from patients suffering from MS selectively acquire the K_Ca3.1^lowK_v1.3^high channel phenotype, whereas T cells specific for control antigens express the naïve and T_CM phenotype (K_Ca3.1^highK_v1.3^low) (5). Moreover, the proliferation of K_v1.3^high T_EM cells is completely and persistently inhibited by ShK toxin (Table 1⇓), without affecting naïve or T_CM lymphocytes. Thus, a K_v1.3-based therapy that suppresses the activation of T_EM cells without significant impairment of the proliferation of naïve and T_CM cells might have use in the management of MS and other T cell–mediated immune diseases, such as type I diabetes mellitus (15), and also in chronic graft rejection and graft-versus-host [disease sustained by chronically activated T_EM cells (16)]. In addition to T lymphocyte subset-specific expression of K_v1.3, the very limited expression of this channel (6, 7) in other tissues make K_v1.3 an excellent target for specific immune suppression (17).

Parallel to the recent progress in understanding the significance of restricted expression of K_v1.3, extensive research has been performed in several laboratories in order to find specific, high-affinity blockers for K_v1.3 channels. Similar to other ion channels’ inhibitors, the blockers of K_v1.3 were isolated from animals, mostly from the venom of scorpions [reviewed in (18)]. These peptide toxins are made up of approximately thirty-five amino acids, and their rigid and conserved three-dimensional structure is provided by three or four disulfide bridges. The binding site of these toxins is generally in the pore of the channel; bound toxins plug the ion conduction pathway resulting in the block of the ionic currents. The most potent natural toxin blockers of K_v1.3 channels are characterized by equilibrium dissociation constants in the low nM and pM range. These include Pi2 toxin from Pandinus imperator [44 pM (19) and Table 1⇓] and many others reviewed recently (6, 17). The most potent blocker of K_v1.3, Shk, has a dissociation constant for K_v1.3 of ~11 pM (14). Based on the molecular models of the K_v1.3 and K_Ca3.1 pore and the crystal structure of KcsA channel (8), highly selective peptide inhibitors of these channels were synthesized recently. Guided by some unique features of the K_v1.3 pore, the structure of the Shk toxin, and the careful analysis of interacting residues between Shk and the pore of K_v1.3, a highly selective K_v1.3 inhibitor was generated by replacing the central Lys²² in the Shk with a shorter, positively charged non-natural amino acid: diaminopropionic acid (14). The resulting peptide blocker, Shk-Dap²², retained high affinity for K_v1.3 (K_d = 23 pM vs 11 pM for the wild-type toxin), but the affinity of the toxin was ~100-fold reduced for other members of the K_v1 family and also for K_Ca3.1.

In addition to peptide blockers of these channels, several distantly related compounds are reported to block K_v1.3 and K_Ca3.1 channels within a range of T-cell activation (17). The binding site and binding stoichiometry of these compounds is very diverse (20, 21), and the blocking mechanism is more complicated than that of the peptide inhibitors. Some blockers bind in the extracellular vestibule of the channel, or at a site located outside the pore region (20), and many of them bind in the inner cavity of the channels below the selectivity filter (21). The best small-molecule inhibitors block K_v1.3 channels were synthesized based on structural information of blocker and receptor [reviewed in (17)]. Although correolide derivatives (11) and cyclohexyl-substituted benzamides (21) were found to block T-cell activation and DTH response in vivo, their therapeutic potential is hampered by their modest selectivity for K_v1.3 over other members of the K_v1.x family channels.

Recently, a new class of small-molecule K_v1.3 inhibitors derived from using 5-methoxypsoralen (5-MOP) (13). 5-MOP is the major K⁺ channel-blocking material in the extracts of Ruta graveolens; this compound is used clinically in the therapy of psoriasis and is reported to improve the functional deficits in MS patients. Earlier attempts, focused on the reduction of phototoxicity of 5-MOP, resulted in the synthesis H37, of a compound which was effective in reducing the proliferation of myelin-reactive encephalogenic rat T cells but had low affinity for K_v1.3 channels (with an EC₅₀ of 6.7 μM) (22).

Based on earlier structure-activity relationship studies, Chandy and colleagues replaced the methyl group at the 5-position of 5-MOP with a series phenylalkyl or cyclohexylalkyl sub- stituents resulting in 12 different compounds (13). Out of these, 5-(4-phenylbutoxy) psoralen (Psora-4) showed the strongest affinity for K_v1.3 channels (EC₅₀ = 2.9 nM) (Table 1⇓). A detailed analysis of the dose-response curve displayed a Hill coefficient of 2, indicating that more than one molecule interacts with a single K_v1.3 channel, resembling the blocking stoichiometry of PAC (a cyclohexyl-substituted benzamide) (21). Similar to other smallmolecule inhibitors of K_v1.3, Psora-4 also binds preferentially to the C-inactivated state of the channel (11), which is a non-conducting state of the channel entered upon prolonged depolarization. However, in contrast to other small-molecule inhibitors, Psora-4 shows a remarkable selectivity for K_v1.3 over other ion channels of the Shaker family (K_v1.x); it blocks K_v1.1, K_v1.2, K_v1.4, and K_v1.7 channels with 16- to 70-fold lower affinity. In addition, Psora-4 was ineffective against the cardiac K_v11.1 (also termed human ether-a-go-go-related, HERG) channel, the K_v3.1 channel found in murine cytotoxic T cells, and three types of Ca²⁺-activated potassium channels (i.e., K_Ca3.1, SKCa3, and BK_Ca), and had low affinity for the Na_v1.2 sodium channel, which is expressed in mammalian unmyelinated and premyelinated axons. This remarkable selectivity profile correlate with excellent results in in vitro lymphocyte proliferation assays. The CD3-specific antibody – stimulated proliferation of human T_EM cells and myelin basic protein–stimulated proliferation of rat memory T cells are inhibited by Psora-4 with an EC₅₀ of 25 and 60 nM, respectively, as these cells have a K_v1.3^high phenotype. On the other hand, Psora-4 was 10-fold less effective in the suppression the proliferation of naïve and T_CM cells, as these cells escape K_v1.3 inhibition by upregulating K_Ca3.1. Thus, this new class of K_v1.3 inhibitors fulfills the demand for selective inhibition of the proliferation of a set lymphocytes; it has the highest potency for K_v1.3 inhibition among the small-molecule inhibitors; and it displays a outstanding selectivity profile. The only disadvantage of Psora-4 is its 7.7 nM affinity for K_v1.5––a channel responsible for the ultrarapid delayed–rectifier current in the human heart. This fact initiated in vivo toxicity tests, which show that at 33mg/kg/day (for five days), subcutaneous injection of Psora-4 is well tolerated in rats. More extensive toxicity studies are required to verify that Psora- 4 is safe for in vivo use as an immunosuppressor. Although the selective inhibition of T_EM cell proliferation, in principle, could be therapeutically effective in the treatment of MS and other immune diseases related to T_EM cells, this possibility has yet to be verified in extensive in vivo animal experiments modeling human diseases.

The central role of K_v1.3 channels in the regulation of T cell activation has been the subject of extensive research in the past. The recent discovery of the K_v1.3^high phenotype of encephalogenic T_EM cells and the generation of high-affinity and high-specificity blockers of K_v1.3, such as Psora-4, open for therapeutic immunosuppression in autoimmune diseases, without impairing the overall cellular immunity. Other recent publications describing the lateral organization of K_v1.3 channels in the plasma membrane of T cells (23) and the recruitment of K_v1.3 channels into the immunological synapse (24), where antigen presenting cells interact with T cells, also point towards the modulation of immune reactions by K_v1.3 channels. These recent pharmacological and cell biological findings will stimulate further research in this field, which may lead to a K_v1.3-based therapy of immune diseases.

• © American Society for Pharmacology and Experimental Theraputics 2004

Rezsö Gáspár, PhD, and Sándor Damjanovich, MD, Member of the Hungarian Academy of Sciences, are Professors and György Panyi, MD, is an Associate Professor (top to bottom) at the Department of Biophysics and Cell Biology, Medical and Health Science Center, University of Debrecen, Hungary. Their research focuses on biophysical characterization of lymphocyte ion channels and cell surface immune receptors.