The human Kv1.1 channel is palmitoylated, modulating voltage sensing: Identification of a palmitoylation consensus sequence

Abstract

Voltage-dependent K(+) channels rely on precise dynamic protein interactions with surrounding plasma membrane lipids to facilitate complex processes such as voltage sensing and channel gating. Many transmembrane-spanning proteins use palmitoylation to facilitate dynamic membrane interactions. Herein, we demonstrate that the human Kv1.1 ion channel is palmitoylated in the cytosolic portion of the S(2)-S(3) linker domain at residue C243. Through heterologous expression of the human Kv1.1 protein in Sf9 cells, covalent radiolabeling with [(3)H]palmitate, chemical stability studies of the [(3)H]-palmitoylated protein, and site-directed mutagenesis, C243 was identified as the predominant site of palmitoylation. The functional sequelae of palmitoylation were examined by analysis of whole cell currents from WT and mutant channels, which identified a 20-mV leftward shift in the current-voltage relationship when palmitoylation at C243 (but not with other cysteine deletions) is prevented by site-directed mutagenesis, implicating a role for palmitoylated C243 in modulating voltage sensing through protein-membrane interactions. Database searches identified an amino acid palmitoylation consensus motif (ACP/RSKT) that is present in multiple other members of the Shaker subfamily of K(+) channels and in several other unrelated regulatory proteins (e.g., CD36, nitric oxide synthase type 2, and the mannose-6 phosphate receptor) that are known to be palmitoylated by thioester linkages at the predicted consensus site cysteine residue. Collectively, these results (i) identify palmitoylation as a mechanism for K(+) channel interactions with plasma membrane lipids contributing to electric field-induced conformational alterations, and ii) define an amino acid consensus sequence for protein palmitoylation.

Fig. 1.

Palmitoylation of the Kv1.1. ion channel. Sf9 cells infected with recombinant Kv1.1-containing virus or control virus were incubated in the presence of 2 mCi of [³H]palmitic or [³H]myristic acid for 2 h at 22°C. Immunoblot analysis of immunoprecipitated solubilized whole cell lysates by using rabbit polyclonal antisera demonstrated equal amounts of immunoreactive Kv1.1 in Kv1.1-infected Sf9 cells (lanes 2 and 3) whereas the control virus-infected cells (lane 1) did not contain any immunoreactive material. Autoradiographs of immunoprecipitated, radiolabeled samples detected a radioactive doublet only in the Kv1.1-infected, palmitate-labeled sample (lane 3) but not in either the control virus-infected, [³H]palmitate-labeled sample (lane 1) or the Kv1.1-infected, [³H]myristate-labeled sample (lane 2). These data are representative of the results of three independent experiments.

Fig. 2.

Hydroxylamine sensitivity and site-directed mutagenesis of the palmitoylated Kv1.1 channel protein. (A) Immunoprecipitated proteins were treated with buffer alone (-) or buffer containing 1 M hydroxylamine at pH 7 (+) before SDS/PAGE and autoradiography (Upper) or Western blotting analysis (Lower). (B) Comparison of palmitoylation of the Kv1.1 WT (WT-Kv1.1) with the C35A, C36A double mutant metabolically labeled with [³H]palmitate. Sf9 cells expressing either WT protein (WT-Kv1.1) or the C35A, C36A double mutant were metabolically labeled with [³H]palmitate, immunoprecipitated, and subjected to SDS/PAGE and autoradiography (Upper) or Western blotting (Lower). (C) Comparison of the palmitoylation of the Kv1.1 C243A mutant to palmitoylation of the C-terminal Kv1.1 deletion mutant D478-495 was performed after metabolic labeling with [³H]palmitate or [³⁵S]methionine, immunoprecipitation, and autoradiography.

Fig. 3.

Macroscopic current kinetics of the C243A Kv1.1 channel mutant. (A and B) Whole-cell, voltage clamp recordings with depolarizing voltage steps from -40 to +40 mV (10-mV increments) from a holding potential of -80 mV were performed on infected Sf9 cells expressing WT-Kv1.1 (A) or C243A mutant protein (B). (C) Comparisons of WT-Kv1.1 (□) and C243A (▪) current-voltage relationships. Notably, C243A currents demonstrated a leftward shift of 20 mV in the current-voltage relationship in comparison to WT protein. Error bars represent the x̄ ± SE of three independent preparations.

Fig. 4.

Comparison of the macroscopic current kinetics of the Kv1.1 cysteine mutants. Whole-cell voltage-clamp recordings of Sf9 cells expressing each of the cysteine mutants (C35A, C36A; C243A; and D478-495) were compared withWT Kv1.1-expressing Sf9 cells as described in Methods. Peak macroscopic current is reduced 5-fold in the C243A mutant compared withWT-Kv1.1 or the other cysteine mutants. These studies are representative of three separate experiments.