Ion channels find a pathway for therapeutic success

Drug discovery and development are complex processes that require long-term commitments to bring such efforts to fruition. Traditionally, there is low probability of success in being able to take new pharmacological agents all of the way through regulatory approval for use in man. There are multiple reasons that account for high failure rates of new investigational drugs, such as the manifestation of safety issues during development or lack of efficacy in clinical trials. Because costs of drug development have significantly escalated, partially as a consequence of stricter regulatory guidelines, it is important initially to identify the highest-quality small-molecule chemical drug leads for prosecution by medicinal chemistry, which fulfill the strictest criteria of potency and selectivity against the therapeutic target of interest. Optimization of such leads would eventually allow interrogation of the target in appropriate experimental medicine protocols to achieve early proof-of-concept in the clinic, which should increase probability of success upon further clinical investigation (1). It is remarkable that although ∼15% of currently used drugs target ion channels (2), this family of proteins remains largely unexplored for therapy, despite their significant contribution to a wide variety of physiological processes. Most approved ion channel drugs, such as Ca²⁺ channel blockers, sulfonylureas, antiarrhythmics, antiepileptics, and local anesthetics, were discovered and optimized using animal models of disease; it was only later that their mechanism of action was shown to be at the level of ion channel proteins (3). Despite the large number of academic and pharmaceutical groups working in this area, and significant progress made in understanding ion channel structure and function, molecular-based ion channel drug discovery research has advanced at a much slower pace than would have been anticipated after cloning of the human genome and identification of all of the ion channel families. Major reasons for the paucity of new drugs specifically targeting ion channels are a consequence of few satisfactory pharmacological tools targeting these proteins and incomplete understanding of precise roles of given ion channels in complex biological systems. To break through these barriers, there needs to be a facile way to identify new selective ion channel modulators with well-defined mechanisms of action that can be used to probe the physiological role of these proteins in native systems, and perhaps even to provide medicinal chemistry leads for new therapeutic agents. In PNAS, the article by Su et al. (4) describes one new elegant approach by which to accomplish these goals.

K⁺ channels are a large family of proteins involved in diverse biological processes. Because of their relative simplicity, K⁺ channels have been studied extensively, and members of this family were among the first channels for which high-resolution X-ray structures were solved, allowing detailed understanding of permeation and gating mechanisms that drive their function (5, 6). Although progress has been made in developing the molecular pharmacology of some K⁺ channels (7–10), other members remain unexplored, and in many cases existing tools are not adequate for probing channels in more complex biological systems. One factor that contributes to the current situation is difficulty in establishing robust, high-capacity functional assays with which to interrogate large chemical libraries, not only on the target of interest, but also on related targets to determine specificity of screening hits. A number of distinct approaches, such as electrophysiology, ion flux, membrane potential-based measurements, ligand binding, and monitoring yeast cell growth have been used for investigating K⁺ channels. However, none of these protocols are optimal for ultrahigh-throughput screening and often the investigation requires use of more than one method to validate the findings (Table 1).

Table 1.

Comparison of different ion channel screening technologies

Protocol	Capacity	Cost	Information content	Ease of operation	Scope
LFA	+++	+	++	+++	+++
Auto EP	++	+++	+++	+	+++
Membrane potential	+++	++	+	++	++
Ion flux	+++	++	++	++	+++
Yeast growth	+++	+	+	+++	+
Ligand binding	+++	+	+	+++	+

+, Low; ++, medium; +++, high. Capacity is related to capability of operation in 96-, 384-, or 1,536-well plates. Ease of operation takes into account the necessity of using mammalian cells. EP, electrophysiology.

For example, electrophysiology represents the gold standard technology by which to monitor ion channel activity, but even with the automated platforms that are currently available, throughput is limited and cost of operation is high (11, 12). Membrane potential-based assays are robust and provide higher capacity (13), but they represent an indirect means of assessing channel activity and, in many cases, inhibition of a significant number of channels is required to obtain a membrane potential readout. Therefore, IC₅₀ values determined in these assays can be significantly shifted to lower potency from those determined by direct measurements of channel activity, and such assays can be subject to significant numbers of false-positives that alter membrane potential by other mechanisms if high concentrations of test articles are screened. Although membrane potential-based assays can support structure-activity relationship studies by medicinal chemistry within individual chemical classes, their broad use to identify leads for ion channels in general is limited because test compound concentration must be restricted to prevent large numbers of false-positives. Ion flux assays that use either thallium or a radioisotope of rubidium as K⁺ surrogates can provide a more direct measurement of channel activity, but use of radioactivity is an issue in high-throughput screening facilities, and thallium assays also have limitations, depending on the host cell where the ion channel of interest is expressed (14). For example, Chinese hamster ovary cells, commonly used for heterologous expression of ion channels, display a substantial endogenous thallium flux pathway that significantly attenuates responses mediated by the expressed K⁺ channel. Yeast growth assays may be more appropriate for exploring those K⁺ channels that are not voltage-gated because yeast maintains a large membrane potential, negative inside, that would attenuate transitions from closed to open states of voltage-sensitive channels (15). In addition, the yeast cell wall may represent a barrier that impedes access of compounds to the ion channel target in the plasma membrane. Finally, ligand binding assays provide high capacity with reproducible characteristics, but they are not functional and require additional follow-up to determine the functional activity of any identified hit (16). In addition, these assays may miss other putative channel modulators that bind to sites, which are not identical to or coupled allosterically to the ligand binding site. Given all of the above, there is a need for establishing new ultrahigh-throughput screening assays that are robust, easy to operate, and that can overcome some of the limitations associated with the currently used technologies.

In PNAS, Su et al. (4), present data that illustrate the feasibility of establishing high-throughput functional assays for a variety of K⁺ channels using a novel and imaginative approach, which the authors refer to as the “liposome flux assay” (LFA). In this approach, K⁺ channels representing members of four of five major subfamilies (17) are purified to homogeneity and reconstituted into lipid vesicles in the presence of KCl. Dilution of these vesicles into a NaCl solution creates a diffusion gradient for K⁺ outward from the proteo-liposomes. Addition of the proton ionophore, carbonyl cyanide m-chlorophenylhydrazone (CCCP) initiates K⁺ efflux by allowing influx of protons to counterbalance the charge movement of K⁺. Influx of protons can be quantitated with the use of a fluorescence dye, such as 9-amino-6-chloro-2-methoxyacridine. This fluorescence dye is permeable in its unprotonated form, but once protons bind the dye becomes trapped inside the proteo-liposome, allowing correlation between activity of K⁺ channels and rate of fluorescence quenching. If channels are closed, or in the presence of an inhibitor, K⁺ efflux will not occur and, as a consequence, there will be no change in dye fluorescence. By using appropriate controls, such as the K⁺ ionophore, valinomycin, which promotes channel-independent K⁺ efflux, it is possible to identify false-positives, such as those with autofluorescence. LFA operates with a good signal-to-noise ratio in 384-well plates, and has demonstrated its utility by screening a 300,000-compound library; this led to identification of novel channel inhibitors in the case of G protein-activated inwardly rectifying K⁺ channel member 2 (GIRK2) and TWIK-related arachidonic acid-stimulated K⁺ channel 1 (TRAAK-1) channels, and to activators of GIRK2 and Slo1. In addition, there is the possibility of adapting the assay to ultrahigh-throughput operation in 1,536-well plates, one of the highest-density screening formats commonly used in the field. Importantly, the assay can be configured for finding either inhibitors or activators of a given channel, demonstrating its excellent flexibility, and the data represent direct interaction of a test agent with the channel under investigation.

LFA provides an easy and cost-effective way to evaluate K⁺ channels not only as therapeutic targets, but also as targets for selectivity and cardiac safety (18). In this later case, the hERG channel LFA assay was explored in great detail by determining its sensitivity to 50 known channel inhibitors and 50 compounds recognized to have no effect on this channel. None of the nonactive compounds caused any significant inhibition: <25%, up to concentrations that approached their limit of solubility. On the other hand, all 50 hERG inhibitors demonstrated an excellent correlation between their IC₅₀ values of inhibition in LFA and those reported in the literature from electrophysiological determinations. There was, however, a 10-fold shift in IC₅₀ values between the two assays, with LFA reporting lower affinity, which Su et al. (4) clearly illustrate to be the result of proton influx being rate-limiting until most hERG channels are inhibited. Although LFA might be considered a semiquantitative high-throughput assay, in particular for hERG, its overall characteristics reveal an approach that could be applied more broadly with any K⁺ channel, providing a substantial advantage over other existing techniques that require functional channel expression in heterologous systems, an empirical task that for certain channels is technically difficult. Indeed, a similar LFA assay has been reported for the Ca²⁺ release-activated Ca²⁺ channel after its purification and reconstitution into liposomes (19), suggesting the possibility that LFA might be adapted for investigating other channel types besides K⁺ channels, as well. The prospect of generating significant amounts of proteo-liposomes in a cost-efficient manner and ease of performing LFA suggest that there is a new means for identifying novel pharmacological tools that could be used to interrogate the physiological role of channels in tissues, or that could serve as leads for new ion channel drugs. One predicts that ion channels are on their way to being drugged!