Fundamental to a variety of biological processes, ion channels are an important class of therapeutical targets. Small-conductance calcium-activated potassium (KCa2.x, or SK1, 2, and 3) channels are activated by increased intracellular Ca2+ to induce potassium efflux and regulate membrane potential (Köhler et al., 1996; Bond et al., 2004; Blatz and Magleby, 1986). In this role, SK1, 2, and 3 mediate cellular excitability and have different but overlapping functions in many cell types including neurons, endothelial cells, and cardiomyocytes (Adelman et al., 2012). In particular, the SK2 channel regulates synaptic transmission and plasticity, learning and memory, and cardiac action potentials and thus has attracted attention as a potential target for the treatment of neurological and cardiovascular diseases (Bond et al., 2004; Hammond et al., 2006; Zhang et al., 2008). SK2 activators reduce cellular excitability and are potential therapeutics for alcohol dependence (Hopf et al., 2011), ataxia (Alviña and Khodakhah, 2010), epilepsy (Anderson et al., 2006), and stroke (Allen et al., 2011). Conversely, SK2 inhibitors increase cellular excitability and have been proposed for the treatment of Alzheimer’s disease (Proulx et al., 2015) and atrial fibrillation (Diness et al., 2011).
The cryo-EM structure of the related SK4 (KCa3.1, IK) channel provided the first insights into the architecture and mechanism of Ca2+-dependent gating for the SK channel family (Lee and MacKinnon, 2018). SK4 channels form non-domain swapped tetramers, with each subunit containing six transmembrane helices S1 to S6 (Köhler et al., 1996; Lee and MacKinnon, 2018). S1–S4 form a voltage-sensor like domain where the S4 helices lack the positively charged residues necessary for voltage sensitivity. The S5 and S6 helices form the potassium pore. Within the potassium pore lies the selectivity filter, a structure unique to potassium channels that is required for rapid and selective conductance of K+ ions (Doyle et al., 1998). Following the S6 helices, there are two intracellular helices (HA and HB) that form the binding site for the Ca2+-binding protein calmodulin (CaM), which acts as the Ca2+ sensor to gate SK channels (Xia et al., 1998). The CaM C-lobe is constitutively bound to the HA and HB helices, and upon an increase in intracellular Ca2+, the CaM N-lobe binds to a unique S4–S5 linker, inducing a conformational change in the S6 helices to open the potassium pore and activate the channel (Lee and MacKinnon, 2018).
SK2 activators described to date bind at the interface of the CaM N-lobe and S4–S5 linker and function by stabilizing this interaction (Lee and MacKinnon, 2018; Brown et al., 2020; Shim et al., 2019). On the other hand, known SK2 inhibitors target the extracellular and/or transmembrane regions and were proposed to function by either direct pore block or negative gating modulation (Brown et al., 2020). The binding site for the bee venom toxin apamin, a cyclic 18-residue peptide inhibitor, has been mapped to the extracellular loop regions of SK2 (Nolting et al., 2007; Weatherall et al., 2011; Lamy et al., 2010). Apamin is of historical importance as it was used to elucidate the physiological role of SK2 and apamin inhibition of SK2 increases neuronal excitability and improves learning and memory (Blatz and Magleby, 1986; Messier et al., 1991). Apamin inhibits SK1, 2, and 3 but not 4 and is most potent against SK2 with an IC50 of ~70 pM (Köhler et al., 1996; Kuzmenkov et al., 2022). Functional mutagenesis of apamin identified two arginine residues that are essential for inhibition (Vincent et al., 1975). Mutagenesis experiments on the SK2 channel indicated that residues in the extracellular loops between S3 and S4 (S3–S4 linker) and between S5 and S6 are important for apamin binding and inhibition (Nolting et al., 2007; Weatherall et al., 2011; Lamy et al., 2010). Since the S3–S4 linker is predicted to be distant to the pore, an allosteric mechanism of apamin inhibition rather than a direct pore block has been suggested (Lamy et al., 2010). Attempts to recapitulate the potency and selectivity of apamin with small molecules resulted in the development of a class of inhibitors, such as UCL1684, that are predicted to have an overlapping binding site with apamin (Ishii et al., 1997; Castle et al., 1993; Chen et al., 2000). These small molecules generally carry two positive charges, which may mimic the two arginine residues in apamin that are essential for inhibition (Vincent et al., 1975). Further characterization of the molecular mechanisms of inhibition by apamin and the small molecule pore blockers is required to understand how the interaction between the S3–S4 linker and the essential arginine residues/positive charges block ion conduction in SK2.
Another class of small molecule SK2 inhibitors acts as negative gating modulators by shifting the Ca2+ dependence of activation to higher Ca2+ concentrations (Jenkins et al., 2011; Simó-Vicens et al., 2017). One such inhibitor, AP31969, is currently in clinical trials for the treatment of arrhythmia (Saljic et al., 2024). Mutagenesis experiments with the structurally related inhibitor AP14145 suggest that these compounds bind within the pore directly below the selectivity filter (Simó-Vicens et al., 2017). Similar potencies on SK1, 2, and 3 channels have been reported most likely due to the homology of pore lining residues in the SK family. However, selective SK1 inhibitors were developed that take advantage of a unique residue, Ser293, on S5 (Hougaard et al., 2012). Interestingly, only small modifications to this family of inhibitors are sufficient to switch the activity profile from inhibition to activation of SK1. However, it remains unclear how compounds that bind the transmembrane regions of SK channels affect Ca2+-dependent gating, which is driven by the interaction between CaM and the intracellular domains.
Despite SK2 being a prominent therapeutic target for both neurological and cardiovascular diseases, no structure of human SK2 has been reported to date. Although the structure of rat SK2 was reported while this manuscript was in preparation (Nam et al., 2025). To enable high-resolution cryo-EM studies of human SK2, we designed a chimera (SK2–4) that contains the transmembrane and extracellular domains of human SK2 and intracellular domains of human SK4. The structures of the SK2–4/CaM complexes in the Ca2+-bound and Ca2+-free conformations demonstrate that SK2 and SK4 adopt similar overall architectures and share a similar mechanism for Ca2+-dependent gating. However, unlike SK4, we observed a structured S3–S4 linker that induces a conformational change in the selectivity filter and forms a hydrophobic constriction at the extracellular opening of the SK2 pore. Apamin binds to the extracellular constriction formed by the S3–S4 linker to block potassium efflux. In addition, high-throughput screening and medicinal chemistry optimization efforts yielded a new class of potent SK2 inhibitors that bind to a novel pocket formed by the S5, S6, and pore helices and induce closure of the S6 helices. Structure-guided design efforts enabled switching the activity profile toward activation while retaining the same binding mode. The detailed understanding of two distinct mechanisms of SK2 channel inhibition, extracellular pore block and negative gating modulation, and a new mechanism for channel activation presented here should facilitate the rational design of potent and selective SK2 modulators.