Abstract:
1ATP-sensitive K+-channels (KATP channels) are heterooctamers composed of four poreforming subunits Kir (2 subtypes, Kir6.1 and Kir6.2) and four regulatory subunits SUR (2 subtypes, SUR1 and SUR2 with two splice-variants 2A and 2B). In different subunit-composition they are present in various tissues (e.g. Kir6.2/SUR1 in pancreatic beta-cells, Kir6.2/SUR2A in cardiac myocytes, Kir6.1/SUR2B in vascular smooth muscle cells) and regulate important physiological functions. They are also targets of synthetic drugs such as the sulfonylureas (SUs) and glinides (collectively termed insulinotropes) which close those channels and the channel openers (e.g. bimakalim).
In this work four main issues were investigated:
Selectivity of selected insulinotropes in binding to and closing of the pancreatic as compared to the cardiovascular KATP channels.
Effect of coexpression with Kir6.x on the binding affinity of insulinotropes for the pancreatic SUR (SUR1).
Impact of the mutation Y1206S and of coexpression with Kir6.x on the binding properties of the cardiac SUR (SUR2A).
Access of benzopyran-type KATP openers to their binding site on the non-vascular smooth muscle KATP channel Kir6.2/SUR2B.
In order to assess the binding selectivity of insulinotropes, 3H-glibenclamide (3H-GBC) binding competition experiments were conducted in intact HEK cells expressing the pancreatic or the cardiac channel at 37ºC. There was no difference in selectivity between A-ligands (short chain SU glibornuride and nateglinide, a D-phenylalanine derivative) and A+B-ligands (long chain SUs: glibenclamide, glimepiride). Both ligand classes bound with ~19x higher affinity to the pancreatic than to the myocardial channel. The B-ligands, repaglinide, AZ-DF265, meglitinide and UL-DF9 showed either a strongly reduced selectivity for the pancreatic channel (repaglinide, AZ-DF265), or a slight cardioselectivity (meglitinide, UL-DF9). The effects of repaglinide and glibenclamide on the pancreatic, myocardial and vascular KATP channels were investigated in the whole-cell-configuration of the patch-clamp-technique at 37ºC. Repaglinide and glibenclamide were >=30 and >=1000 times more potent in closing the pancreatic than the cardiovascular channels, respectively, and they did not induce complete inhibition of the myocardial channel. The transduction of inhibitor binding into channel closure was easiest for the Kir6.2/SUR1 channel. The selectivity of repaglinide and glibenclamide shown here supports their therapeutical safety. However, these conclusions must be confirmed in appropriate clinical studies.
Investigating the effects of coexpression with Kir6.x on the affinity of insulinotropes binding to SURx we found that this greatly (>100x) enhanced the affinity of repaglinide and AZ-DF265 for SUR1, whereas the effect was small for other SUs and glinides (<5x). This behaviour depended on the piperidino-substituent, and it makes a major contribution to the selectivity of repaglinide and AZ-DF265 for the pancreatic over the cardiovascular KATP channels. In addition, the effect of coexpression was shown to critically depend on the intact Kir6.2 N-terminus.
In the third point, the impact of the mutation Y1206S in SUR2A (i.e. the replacement of Y1206 by S, the corresponding amino acid in SUR1) and of coexpression with Kir6.x on SUR2A was examined. In intact HEK cells expressing the complete Kir6.2/SUR2A channel, the mutation Y1206S increased the affinity of A- and A+B-ligands ~13x but decreased that of the B-ligands meglitinide and UL-DF 3.5- and 7.1-fold, respectively. Coexpression with Kir6.2 increased the affinity of SUR2A for all insulinotropes. The effect was particularly marked for AZ-DF265 und repaglinide, however to a lesser degree than with SUR1. The mutation Y1206S did not affect the binding of KATP channel openers to SUR2A; coexpression with Kir6.2 reduced opener affinity of wild-type and mutant SUR2A only marginally. In addition, the negative allosteric interactions between GBC- and P1075-binding sites were modified by both mutation Y1206S and coexpression with Kir6.2.
Finally, it was investigated whether benzopyran openers approach their binding site on the Kir6.2/SUR2B channel from the extra- or the intracellular side of the cell membrane. To this end, binding and effect of BMSA, a poorly membrane permeant sulfonic acid derivative of bimakalim and of the cell-permeant mother compound were compared in different configurations. Bimakalim exhibited high affinity binding to the channel in both membranes and intact cells (Ki = 61 and 35 nM, respectively) whereas BMSA binding was much weaker in cells (Ki = 186 µM) than in membranes (Ki = 4,3 µM). In inside-out-patches, bimakalim and BMSA opened the Kir6.2/SUR2B channel to a similar degree whereas in the whole-cell configuration, only bimakalim was effective. This behaviour suggests that benzopyran-type KATP channel openers reach their binding site from the cytosol.
Overall, these studies improve our understanding of the structure-response and structure-binding relationships of the KATP channel modulators. The conclusions on the subtype selectivity of repaglinide and GBC are of clinical importance.