Gating mechanisms of voltage-activated ion channels

Neurological Disorders And Stroke

Investigators

Linked publications & trials

Abstract

The objective of this study was to ascertain whether the transmembrane interface between voltage-sensing and pore domains in the Shaker Kv channel is tightly packed and whether this region serves an important role in coupling movements of the voltage sensors with those in the gate. To explore potential interactions between the voltage-sensing and pore domains in the Shaker Kv channel we mutated each residue in the S5 and S6 helices to Trp, expressed the resulting channels in oocytes, and used voltage clamp recording techniques to study channel activity. Our experiments were undertaken to answer two specific questions for each mutant. First, is the mutation so poorly tolerated that it causes retention of the protein in the endoplasmic reticulum (ER) due to misfolding? Second, if the mutant channel traffics to the plasma membrane and remains functional, does the mutation perturb movement of the voltage sensors between R and A states? A substantial perturbation might be expected in regions where the two domains interact if this interaction changes as the voltage sensors activate. To address this second question we measured gating currents resulting from movement of charged residues as the voltage-sensing domains fluctuate between R and A states. After evaluating perturbations in gating charge movement we compare these results with our previously reported perturbations in opening of the channel to look for domain interactions that might occur during the concerted opening transition. Of the 53 positions mutated we were able to observe functional activity for mutations at a total of 40 positions. To distinguish whether the remaining 13 mutants disrupt function or are ER retained (ERr) we investigated the extent of glycosylation. Twelve of the 13 non-functional mutants are found in only the core glycosylated form, suggesting that they are ERr. Of the 40 mutants that support functional activity, some display only minor perturbations in charge translocation (V1/2 of the Q-V relation is within 10 mV of control), while others display strong perturbations in gating charge movement. When mapped onto the X-ray structure of the Kv1.2 channel, the ERr residues are located towards the extracellular portion of the protein where they are either involved in packing between S5 and S6, or interact with residues in the reentrant pore loop that forms the selectivity filter. None of the ERr mutants project out towards the surrounding S4 helix, demonstrating the tolerance of this region to mutation and suggesting that the interface between domains is rather unconstrained. We also mapped our results from the functional characterization of gating current phenotypes onto the Kv1.2 structure and observed that the mutants producing strong perturbations are not concentrated on the perimeter of the pore domain, but are found throughout both S5 and S6 helices. The strongly perturbing mutants that are located at the interface between the voltage-sensing and pore domains in the Kv1.2 structure are clustered together near the external side of the membrane. Mutations in this region, which we term the external cluster, can be seen to shift the Q-V towards positive voltages, and thus perturb the equilibrium between R and A states of the voltage sensor in favor of the R state. Although channel opening for each of the mutants in the external cluster is also shifted towards positive membrane voltages, the Kv1.2 structure suggests that these residues are not positioned where they might directly influence the dynamics of the S6 gate, indicating that these mutations influence the equilibrium and dynamics of the voltage sensors proper. To explore domain interactions that may be important for the concerted opening transition we compared the effects of mutants on Q-V relations and channel opening . One particularly interesting group of mutants, which we term the internal cluster, display Q-V relations with small to dramatic shifts towards negative voltages with G-V relations that exhibit pronounced shifts in the opposite direction. These mutants exhibit voltage sensor movements between resting and activated states that are weakly to strongly perturbed in favor of activation, whereas the concerted opening transition is strongly perturbed in favor of the closed state, as if a critical link between conformational changes in the voltage sensors and the gate has been altered. None of the residues within the internal cluster are involved in packing between the S5 and S6 helices, where such dramatic effects on the concerted opening transition might in principle be expected. Instead, residues within the internal cluster form a relatively well-defined patch in the structure of Kv1.2, spanning from the middle to the intracellular end of the membrane. The possibility of a direct interaction between this region of S5 with the internal half of S4 is strengthened by the similarity of the phenotype for the internal cluster to that of the ILT triple mutant within the S4 of Shaker (V369I, I372L and S376T), which also has pronounced effects on the concerted opening transition. Remarkably, the internal cluster is positioned to directly interact with the ILT residues in S4 from the adjacent subunit. The physical proximity and phenotypical similarity between those residues in the voltage-sensing and pore domains suggest that the interactions between the two groups of residues most likely underlie the concerted transition leading to the channel opening. Taken together, our results suggest that much of the transmembrane interface between voltage-sensing and pore domains is relatively unconstrained, consistent with the absence of tight packing that is observed in crystal structures of the KvAP and Kv1.2 channels, but also identify a region within the internal half of the protein where domain interactions are important for the concerted opening transition.