Molecular mechanisms of local anaesthetic action on voltage-gated ion channels

Abstract

In this thesis I try to clarify some molecular mechanisms of local anaesthetic action on ion channels. The traditional view is that local anaesthetics eliminate action potentials by a direct block of Na channels. Other mechanisms, however, have been suggested. For instance bupivacaine has been proposed to affect Na channels in myelinated axons indirectly, by making the resting potential less negative. To test this hypothesis we analysed the effects of bupivacaine on voltage-clamped myelinated axons from Xenopus laevis. Contrary to the suggested hypothesis, the leak current and the resting potential were unaffected. Furthermore, the blocking effect on the K current was explained by two population-specific mechanisms.

In order to gain further insights in the molecular mechanisms of local anaesthetic action, we analyzed the effects of bupivacaine on a series of voltage-gated K (Kv) channels and various mutant channels, expressed in Xenopus oocytes. Two phenomenologically different blocking effects were seen; a timedependent block of Kv1 and Kv3 channels, and a time independent block on Kv2.1. Swapping the S6 helix between Kv1.2 and Kv2.1 introduced Kv1.2 features into Kv2.1. The results suggest that bupivacaine blocks Kv channels by an open-state dependent mechanism, and that Kv2.1 deviates from the other channels in allowing a partial closure of the bupivacaine-bound channel. The results also suggest that the binding site in Kv2.1 is located in the internal vestibule and that residues in the descending P-loop and the upper part of S6 are critical for the binding.

The location of the binding sites was further investigated by automated docking and molecular dynamics methods, using homology models of Kv1.5. Two different models were constructed to describe an open channel. They were based on the crystallized bacterial channels KcsA and MthK. The first model corresponds to a PVP-type of bending hinge in the internal helices, while the second corresponds to a Gly-type of bending hinge. The automated docking and molecular dynamics calculations combined with free energy estimations predicted strongest binding to the PVP region. Surprisingly, no binding was predicted for the Gly-bend model. The results support our electrophysiological data showing that Kv1.5 is unable to close when bupivacaine is bound to the channel.

The standard view of local anaesthetic action assumes a preferential binding to channels in inactivated state. Recently, this view has been challenged. We investigated this issue by comparing the effects of bupivacaine on inactivating and non-inactivating Kv channels of similar subtypes, hypothesizing that the bupivacaine affinity is similar for the two types due to the structural similarities. The results can be explained by a simple kinetic scheme, deviating from the standard scheme in assuming local anaesthetic binding to channels exclusively in open state. By using data from Na channel experiments we could in simulations show that the basic scheme could be used to clarify controversial issues about local anaesthetic effects on Na channel inactivation.