Systematic Analyses of structure/function variability of viral K+ channels for the development of synthetic channels

Abstract

Potassium (K+) channels are an important class of ion channels, which serve crucial physiological functions. All known K+ channels have a similar architecture: a central ion-conducting pore, with a high similarity throughout all different forms of life, eukaryotes, archaea, bacteria and even in viruses. The latter turns out to be a very interesting group: K+ channels isolated from viruses are reduced to an absolute minimum, basically representing the pore module of every K+ channel. Despite their minimall size, they still possess many of the essential and characteristic properties of bigger and more complex channels. This makes them an ideal model system to investigate structure/function correlations that determine ion permeation and gating in K+ channels. In this study, the planar lipid bilayer (PLB) method was used to analyze small viral K+ channels on the single channel level. This reduced electrophysiological measuring system allows a quick and easy modification of the experimental conditions such as pH, ion concentration or lipid environment, as well as a straight forward addition of blockers. In the first part, a novel technique for a fast and artifact-free functional analysis is introduced. We show that adding nanodiscs to an in vitro expression system results in a time-saving and contamination-free method for expression and purification of membrane proteins, which can subsequently be used for single channel analyses with different methods. The viral potassium channel KcvNTS, as well as the model bacterial channel KcsA could successfully be reconstituted into PLBs after expression in the presence of nanodiscs. The experiments also show that not the lipid from nanodisc determines channel function, but the lipid composition of the target membrane in which the channel of interest later incorporates into. The data also show that this technique could be used for the functional reconstitution of the synthetic blue light sensitive channel BLINK1. Unlike expression in cells, in PLBs this channel loses its light sensitivity, which demonstrates a shortcoming of the method regarding posttranslational modifications. The second part of the study is dedicated to the systematic analyses of the small viral K+ channel KcvPBCV-1. A subunit consists of 94 amino acids, which contains two transmembrane (TM) helices, a pore loop including the selectivity filter and a short N-terminal helix. Previous studies with yeast complementation assays have shown that leucine at position 94 plays a crucial role in channel gating. In this preceding study, leucine was exchanged to all other proteinogenic amino acids and the degree of yeast complementation was monitored as an indirect parameter for channel activity. Here, KcvPBCV-1 and the 19 mutants KcvPBCV-1 L94X were analyzed on the single channel level, to determine the effect of each mutation on the key functional parameters of channel function: unitary single channel conductance and open probability. The single channel analyses are not compatible with the yeast complementation assays. This means that the latter method is suitable for screening of basal channel function but provides no information on detailed functional features of a channel. The results of the yeast complementation assays are presumably also influenced by secondary factors such as sorting and translation efficiency of the channel protein. Nonetheless, the single channel data demonstrate that the last amino acid of KcvPBCV-1 has a complex impact on channel function and can modulate the open probability as well as the unitary single channel conductance. The main observations are that KcvPBCV-1 L94P and KcvPBCV-1 L94C have both, an increasing 2 impact on the channels open probability and voltage dependency. Another discovery was that introduction of an amino acid with a basic side chain inverts the voltage dependency of the open probability and reduces the unitary single channel conductance of KcvPBCV-1. Since the amino acid histidine can be titrated in the physiological pH range, KcvPBCV-1 and KcvPBCV-1 L94H were examined within a pH window from 4 to 9. It turns out that the wild type channel already exhibits a mild sensitivity toward H+, which is strongly increased by mutation of L94H. This histidine associated effect is described with a simple two-state model, where KcvPBCV-1 L94H can pass from a state of high conductance (Gmax) to a state of low conductance (Gmin) either via an effect which is inherent to the wild type protein or via an effect which is introduced by the mutation to histidine. Fitting the results based on this model shows that the effect after mutation of L94H completely masks the effect that H+ has on the wild type channel. Further, the effect during the transition from deprotonated to protonated histidine (pH 6) can be mimicked, by addition of NiCl2 to the bath solution during measurements of KcvPBCV-1 L94H at low H+ concentrations. Deprotonated histidine is known to coordinate Ni2+. However, the complexity of this H+ dependency does not allow to use this as a sensor system for pH. In the last part of the study two additional small viral K+ channels, KcvGNLD and KcvMT325 were examined in PLBs. So far both channels have only been characterized on a macroscopic level either by patch clamp measurements in HEK293 cells or with two-electrode voltage clamp measurements in oocytes. The two channels differ slightly in their aligned amino acid sequence but are quite similar in respect to conductance and open probability. The open probability is voltage-dependent, decreasing from 100% to ~10% with positive voltages. With increasing K+ concentrations the voltage-dependent decrease of the open probability is shifted towards more positive voltages. Also, KcvGNLD and KcvMT325 both possess a threonine in the selectivity filter directly before the GY/FG motive, compared to most other Kcv channels which possess a valine at this position. Mutation of threonine to valine results in a loss of the voltage dependency in both, KcvGNLD and KcvMT325, despite the similarity of the two amino acids. The results of the different experiments demonstrate that for small viral K+ channels modification of basic channel properties such as an inversion or loss of voltage dependency, alteration of unitary single channel conductance or open probability can easily be achieved by mutation of only a single amino acid. This feature makes viral K+ channels particularly suitable modules for the construction of synthetic channels. With only little effort, great differences in the phenotype of single channel properties can be achieved.