Is drug lipophilicity a robust predictor for potency of beta-cell KATP channel block

P. Smith. Life Sciences, University of Nottingham, Nottingham, UNITED KINGDOM

Introduction The ability of sulphonylurea drugs to potently inhibit ATP-sensitive-potassium channels (KATP) have revolutionised our understanding and treatment of hypoinsulinaemic disorders such as type II diabetes. However adverse hyperinsulinaemia with a variety of structurally diverse drugs unrelated to of sulphonylureas is demonstrated to arise from off target inhibition of KATP (1). By a retrospective analysis of published data, I demonstrate that the potency for non-sulphonylurea compounds that inhibit KATP, is dependent on their lipophilicity.

Methods Publications that measured KATP currents with either whole-cell or inside-out patch clamp in pancreatic beta-cells, beta-cell cell-lines or from cells expressing recombinant Kir6.24/SUR14 constructs were examined (1). Log IC50 values recorded in the presence of intracellular Mg2+, except where Mg2+ did not affect drug action, were extracted. Log P values, pH range 7-9, were from DrugBank (2) else they were calculated from quantitative structural activity relationships with Chemicalize (3).

Results With the exception of sulphonylureas, the pIC50 of 23 drugs to block KATP were linearly correlated to their Log P (r = 0.91, p < 0.001). Linear regression revealed a slope coefficient insignificantly different to 1 (0.83 to 1.3, 95% C.I.) and a y intercept of 1.9 (1.3 to 2.6, 95% C.I.); a value indicative of a membrane concentration of 10 mM at the IC50 for any drug. Expression studies with KIR6.2ΔC26, which forms K+ channels in the absence of SUR1, show, where tested, that the channel pore is the predominant binding site for non-sulphonylureas (1). This finding is supported by drugs which block KATP with a potency unaffected by Mg2+ a condition which uncouples the interaction between SUR1 and Kir6.2 (1). Further support for these ideas comes from diazoxide which opens KATPin the presence of intracellular Mg2+ but blocks in its absence (4) with a pIC50 related to its Log P .

Conclusion For non-sulphonylurea drugs that inhibit KATP their pIC50 is linearly related to Log P and is devoid of obvious structural or chemical motifs. Moreover, these data suggest that the action of these drugs occurs by a simple physiochemical disruption of the membrane bilayer environment and suggest that a prospective analysis of Log P for new drugs will help predict adverse hypoglycaemic action.

References (1) Gribble FM et al (2000) Br. J Pharmacol. 13 (2000) 756-760. (2) Drugbank http://www.drugbank.ca/ (3) Chemicalize www.chemicalize.org/ (4) Kozlowski RZ and Ashford ML (1992) Br. J Pharmacol. 107: 34-43.