SUR1 L0 and Kir6.2 M0 can Partner in KATP Gating: Evidences from Diabetogenic Mutations

Abstract

Neonatal diabetes (ND) mutations in different domains of either KATP subunit overactivate KATP via different mechanisms. Mutations in the ABC core of SUR1 augment its Mg-nucleotide-dependent stimulation of Kir6.2. Mutations at the putative ATP site of Kir6.2 supposedly compromise the inhibitory nucleotide binding. Mutations at the putative gate reportedly stabilize its open state. Many ND mutations map to the putative L0 helix preceding the SUR1 core and to the M0 (‘slide’) helix of Kir6.2. We discovered that L0 controls spontaneous bursting and introduced a model in which L0 and M0 partner in KATP gating (Babenko & Bryan 2003). The model predicted that ND mutations in either amphipathic helix at the membrane-cytosol interface hyperactivate KATP by altering its intrinsic gating kinetics. Here we tested the idea. We compared the effects of severe ND mutations in the middle of either interface helix on spontaneous single-channel kinetics, ATP-inhibition, and Mg-nucleotide stimulation of KATP. We found that each of these mutations decreases the rate of burst termination while increasing the rate of burst initiation, thereby reducing the availability of the closed state with the lowest Kd for inhibitory ligands. This mechanism attenuates ATP inhibition but not Mg-nucleotide stimulation, thereby hyperactivating KATP in vivo, and uncouples sulfonylurea binding from Kir6.2 closure. Thus, in support of our model, perturbing either interface helix disrupts SUR1/Kir6.2 inhibitory coupling. Affinity photolabeling with I-azidoglibenclamide indicated that none of the ND mutations altered the proximity of L0 to M0. But each of these mutations changed the helical hydrophobic moment, implying that rotation of either helix along its axis disrupts L0/M0 interactions. We propose that the two interface helices have been optimized to interact and slide together to stabilize the long-lived closed state of SUR1/Kir6.2 complex.