Molecular Modelling and Experimental Analysis of the ATP Sensitivity of Cloned Smooth Muscle KATP Channels

Abstract

ATP-sensitive K+ (KATP) channels, comprised of pore-forming Kir6.x and regulatory SURx subunits, play important roles in many cellular functions. Vascular (Kir6.1/SUR2B) and non-vascular (Kir6.2/SUR2B) smooth muscle KATP channels regulate the resting membrane potential in smooth muscle cells and modulate the muscle tone. In spite of the fact that Kir6.1 and Kir6.2 share ∼70% sequence identity, their sensitivity to ATP inhibition differs by more than 30-fold with IC50 values of 3.1±1.3mM (n=7) and 101±22μM (n=17), respectively. We investigated the molecular mechanisms that underlie this differential response. All (17) Kir6.2 residues putatively involved in binding of ATP are conserved in Kir6.1 with the exception of K39 and K185. We constructed a homology model of Kir6.2 and carried out computational docking of ATP to determine the conformation of bound ATP in its putative binding site. The model was validated by assessing the binding free-energy difference of ATP for wild-type (Kir6.2/SUR2B) and a range of Kir6.2 ATP-binding site mutants for which the affinity for ATP was previously experimentally determined. The model predicted that substitution of K39 and K185 on Kir6.2 with the equivalent residues on Kir6.1 (K39S/K185R) did not affect the ATP affinity. Patch-clamp experiments confirmed that there was no difference in the ATP sensitivity between these channels. Furthermore, the open probability (Po) of Kir6.1/SUR2B, Kir6.2/SUR2B and Kir6.2-K39S-K185R/SUR2B were identical (0.41±0.05 (n=7), 0.38±0.06 (n=7) and 0.45±0.02 (n=8), respectively), indicating that their different ATP sensitivity is not caused by a change in channel gating. In conclusion, the difference in ATP sensitivity between Kir6.1 and Kir6.2 is not caused by a difference in the ATP binding site meaning other regions of the channel must be involved.This work was supported by BBSRC (BB/H000259/1) and by CASPUR (gpu11-905)