Discovery and electrophysiological characterization of SKF-32802: A novel hERG agonist found through a large-scale structural similarity search.

Despite the importance of the hERG channel in drug discovery and the sizable number of antagonist molecules discovered, only a few hERG agonists have been discovered. Here we report a novel hERG agonist; SKF-32802 and a structural analog of the agonist NS3623, SB-335573. These were discovered through a similarity search of published hERG agonists. SKF-32802 incorporates an amide linker rather than NS3623's urea, resulting in a compound with a different mechanism of action. We find that both compounds decrease the time constant of open channel kinetics, increase the amplitude of the envelope of tails assay, mildly increased the amplitude of the IV curve, bind the hERG channel in either open or closed states, increase the plateau of the voltage dependence of activation and modulate the effects of the hERG antagonist, quinidine. Neither compound affects inactivation nor deactivation kinetics, a property unique among hERG agonists. Additionally, SKF-32802 induces a leftward shift in the voltage dependence of activation. Our structural models show that both compounds make strong bridging interactions with multiple channel subunits and are stabilized by internal hydrogen bonding similar to NS3623, PD-307243 and RPR26024. While SB-335573 binds in a nearly identical fashion as NS3623, SKF-32802 makes an additional hydrogen bond with neighboring threonine 623. In summary, SB-335573 is a type 4 agonist which increases open channel probability while SKF-32802 is a type 3 agonist which induces a leftward shift in the voltage dependence of activation.

Utility of frozen cell lines in medium throughput electrophysiology screening of hERG and NaV1.5 blockade.

INTRODUCTION: The development of drug candidates must take into account that many compounds have off-target activity against voltage-gated ion channels (VGIC) which may prevent their progression to market. Of particular concern are hERG and hNa(V)1.5. Screening against these ion channels is necessary but expensive, partially due to maintenance of constantly cultured cell lines. Here, we show that frozen HEK-293 cells can be maintained indefinitely, reducing variability in cell performance, time and expense of cell culture. METHODS: Cells, constantly cultured or frozen, were assayed on the PatchXpress 7000A using tool compounds. RESULTS: Amitriptyline, quinidine, compound A, fluoxetine and imipramine inhibited hERG with IC(50)s (paired values denote constantly cultured and frozen, respectively) of 4.8±0.4 and 5.1±0.4, 1.4±0.1 and 1.1±0.1, 24.4±2.4 and 21.9±1.8, 2.1±0.4 and 2.1±0.1, 5.2±0.4 and 4.0±0.2µM. Quinidine, flecainide, mexiletine and amitriptyline inhibited hNa(V)1.5 with IC(50)s of 46.6±4.3 and 28.0±2.3, 7.6±0.7 and 6.2±0.5, 153.5±13.0 and 106.0±4.7, 5.5±0.5 and 4.8±0.2µM. Voltage dependences of activation (V(1/2)) for hERG were statistically identical, 0.4±0.8mV and 2.5±0.5mV. In hNa(V)1.5, the V(1/2) of inactivation and activation were statistically identical, -82.7±0.1mV versus -84.9±0.3mV, -47.5±0.3mV versus -45.0±0.6mV. Current density in both conditions in hERG experiments was similar, 47.0±4.1pA versus 42.3±6.0pA/pF. DISCUSSION: hERG and hNa(V)1.5 screens run using frozen cells have statistically identical IC(50)s, voltage dependence of activation, IV relationships and current density to screens using continuously cultured cells. Frozen cells have more constant performance and allow rapid switching between experiments on several cell lines without sacrificing data quality.