Abstract:
KATP channels are gated by intracellular nucleotides, i.e. they are closed by ATP and opened by MgADP. In this thesis work, the effect of nucleotides, channel blockers, and openers on the activity of recombinant cardiovascular KATP channels was investigated using the patch-clamp technique in the whole cell and the i/o patch configurations.
The effects of nucleotides were investigated using the Kir6.2/SUR2A channel. In the absence of Mg2+, inhibition was observed (ATP: C50=12 µM; ADP: IC50 ~ 800 µM). In the presence of Mg2+, the inhibition curve of ATP was slightly shifted rightwards (IC50=30 µM). With MgADP, a biphasic effect was observed: At low concentrations, activation by ~100 % was seen with EC50=3 µM; at high concentrations, inhibition (IC50=600 µM). It is probable that inhibition by ATP and MgATP is due to the same process (i.e. binding to Kir6.2). The activation by nucleoside phosphates requires Mg2+ and relies upon binding of MgADP to SUR. In case of MgADP, the activatory and inhibitory components are well separated on the concentration scale; for MgATP, the components lie close to one another so that only a slight rightward shift of the inhibition curve by Mg2+ is seen. In the individual patches we observed a strong correlation between the magnitude of the activatory effect of MgADP and the run down of the channels; the underlying molecular mechanism remains unknown. The inhibitory action of nucleoside phosphates is independent of run down.
Amongst the KATP channel blockers, examples of three different structural classes were examined, i.e. the sulphonylurea, glibenclamide, the cardioselective sulphonylthiourea, HMR 1883, as well as PNU-96293, which is derived from the channel opening cyanoguanidines like pinacidil. All three blockers close the channel by binding to SUR. Main points of interest were the dependence of block on the patch-clamp configuration (whole cell vs i/o patch), the effect of nucleotides on the observed block, the selectivity for channel subtypes, and, finally, the effect of the mutation of SUR2(Y1206S).
In the whole cell configuration, glibenclamide und HMR1883 blocked Kir6.2/SUR2x-channels completely at saturating concentrations; PNU-96293, however, induced only a partial block and must be considered a partial antagonist. In i/o patches, the three compounds induced only a partial ((50-70 %) block. This shows that channel species occupied by the blocker can have an open probability >0.
The sensitivity of Kir6.2/SUR2 channel block by glibenclamide and HMR1883 to MgADP was investigated. MgADP weakened the block by reducing the potency (but not the efficacy) of the compounds. Regarding the selectivity profile of the different compounds, slight differences were found which depended on the patch-clamp configuration. The mutation Y1206S in SUR2 increased the potency of glibenclamide in blocking the Kir6.2/SUR2 channels ~ 30x (i/o patch, absence of nukleotides); in the whole cell configuration, the effect was smaller. In contrast, the mutation increased the potency of PNU-96293 only 1.7x; this indicated that the binding site of PNU at SUR differs from that of glibenclamide.
Regarding KATP channel openers, three aspects were investigated: (i) the opening of Kir6.2/SUR2B channels by openers from different structural families in the absence of MgATP; (ii) the action of the novel dihydropyridine, A312110; and (iii) the opening effect of the stilbene, DIDS.
The opening of KATP channels by openers is strongly favoured by MgATP. The signalling chain of this MgATP-dependent pathway involves an opener-induced facilitation of the ATPase activity of SUR and the stabilisation of a posthydrolytic conformation of the protein. On the other hand, some openers at high concentrations can open KATP channels also in the absence of MgATP. Here, I have quantified the opening of Kir6.2/SUR2B-channels by P1075, minoxidil sulfate, diazoxide and nicorandil in the absence of Mg2+. The experiments showed that the different structural classes of openers differed in their efficacy and potency, and that at high concentrations P1075, the MgATP-independent pathway of channel opening contributes up to 20 % to the total effect in the presence of MgATP.
The novel dihydropyridine, A312110, proved to be a potent opener of Kir6.2/SUR2A und /SUR2B channels with a potency equal to that of the standard opener P1075 but a slightly reduced efficacy.
The stilbene DIDS is known to block KATP channels (KIR6.2 + SUR1, SUR2A and SUR2B) with µM potency. Here, I show for the first time that DIDS, at nM concentrations, was able to open Kir6.2/SUR2B channels; however, opening was observed in only 65 % of the experiments. The activating effect required the presence of SUR (as Kir6.2D26 channels were not activated); in addition, the presence of MgATP was obligatory. The PI-3-kinase inhibitors LY-294002 and wortmannin as well as oleyl-CoA did not affect the activating effect of DIDS suggesting that the channel opening mechanism of DIDS was independent of PIP2/PIP3.
In summary, the experiments have shown an intriguing dependence of the effects of KATP channel blockers and openers on the configuration used in the patch-clamp experiments. A second major result is the strong dependence of the drug effects on the presence of Mg2+ salts of nucleotides. The work highlights the complexity of the regulation and modulation of KATP channels which act as metabolic sensors of the cell and which are regulated not only by nucleotides but also by the cytoskeleton and by phospholipids like PIP2 and oleoyl-CoA.