Calcium/Calmodulin-Dependent Protein Kinase II Regulation of the Slow Delayed Rectifier Potassium Current, IKs, During Sustained Beta-Adrenergic Receptor Stimulation

2019-01-02T16:18:59Z (GMT) by Tyler A. Shugg
Background: Sustained elevations in catecholaminergic signaling, mediated primarily through β-adrenergic receptor (β-AR) stimulation, are a hallmark neurohormonal alteration in heart failure (HF) that contribute to pathophysiologic cardiac remodeling. An important pathophysiological change during sustained β-AR stimulation is functional inhibition of the slow delayed rectifier potassium current, IKs, which has been demonstrated to prolong action potential duration (APD) and increase ventricular arrhythmogenesis in HF. Though functional inhibition of IKs has been consistently reproduced in cellular, animal, and limited human studies of HF, the mechanisms that mediate IKs inhibition during HF remain poorly understood.

In addition, HF results in aberrant calcium handling that is known to contribute to the disease. HF has been demonstrated to increase the expression and function of calcium/calmodulin-dependent protein kinase II (CaMKII), a key regulator of calcium homeostasis and excitation-contraction coupling in cardiomyocytes. Enhanced CaMKII signaling has been consistently demonstrated to contribute to increased arrhythmogenesis in a number of cardiac diseases, including HF. CaMKII is a known pathological regulator of many cardiac ion channels resulting in APD prolongation and the development of arrhythmias.

Objective: This investigation aims to assesses the potential for CaMKII regulation of KCNQ1 (pore-forming subunit of IKs) during sustained β-AR stimulation and to characterize the potential functional implications on IKs. Furthermore, this investigation seeks to elucidate the mechanism underlying CaMKII-mediated IKs inhibition during sustained β-AR stimulation.

Methods: Phosphorylation of KCNQ1 was assessed using a tandem liquid chromatography- mass spectrometry/ mass spectrometry (LCMS/MS) approach during sustained β-AR stimulation via treatment with 100 nM isoproterenol (ISO) for 4-24 hours and during co-expression with KCNE1. Whole-cell, voltage-clamp patch clamp electrophysiology experiments were performed in HEK 293 cells transiently co-expressing wild-type (WT) or mutant KCNQ1 (mutations conferring mimics of dephosphorylation and phosphorylation were introduced at phosphorylation sites identified by LCMS/MS) and KCNE1 (auxiliary subunit) during ISO treatment, treatment with CaMKII or protein kinase A (PKA) inhibitors, or during lentiviral δCaMKII overexpression. A robotic peptide synthesizer was used to create fifteen residue peptide fragments on a nitrocellulose membrane corresponding to KCNQ1 intracellular domains and the KCNQ1 residues identified via LCMS/MS; membranes were incubated with activated CaMKII or PKA in the presence of radiolabeled ATP to identify potential sites of phosphorylation. Bimolecular fluorescence complementation (BiFC) experiments were performed in HEK 293 cells to assess the impact of CaMKII-mediated KCNQ1 phosphorylation on the interaction of KCNQ1 and KCNE1 subunits. Protein immunoblot experiments were performed to (1) assess CaMKII activation during ISO treatment and (2) to assess plasma membrane expression of KCNQ1 and KCNE1 subunits with mimics of differential KCNQ1 phosphorylation following a membrane protein biotinylation procedure.

Results: In Aim 1, we investigated the regulation of the KCNQ1 carboxyl terminus during sustained β-AR stimulation and assessed the associated functional implications on IKs. An LCMS/MS approach identified five novel KCNQ1 carboxyl terminal sites that demonstrated basal phosphorylation, with T482 and S484 having enhanced phosphorylation during treatment with 100 nM ISO for 24 hours (p<0.01 at both sites). Using patch clamp electrophysiology, we demonstrated that treatment with 100 nM ISO for 12-24 hours reduced IKs current density (p=0.01) and produced a depolarizing shift in the voltage dependence of activation (p<0.01) relative to vehicle. Mimics of phosphorylation (mutations to aspartic acid; Triple-D KCNQ1) at S457, T482, and S484 in combination, meanwhile, reduced IKs activation current density relative to dephosphorylation (mutations to alanine; Triple-A KCNQ1) mimics (p=0.02) but did not affect the voltage dependence of activation (p=0.66). Functional assessment of these sites individually revealed that phosphorylation mimics at S457 (p=0.02) and S484 (p=0.04), but not at T482 (p=0.53), reduced IKs current density relative to mimics of dephosphorylation. Similarly, the voltage dependence of activation was right-shifted with phosphorylation mimics at S457 (p=0.03) and S484 (p=0.02), but not at T482 (p=0.99), relative to mimics of dephosphorylation.

The focus of Aim 2 was to assess the potential for CaMKII signaling to regulate increased KCNQ1 phosphorylation and reduced IKs function during sustained β-AR stimulation. Peptide fragments corresponding to the KCNQ1 carboxyl terminal sites demonstrating basal phosphorylation via LCMS/MS analysis were synthesized on a nitrocellulose membrane and exposed to activated δCaMKII. Only peptide fragments corresponding to S484 demonstrated CaMKII phosphorylation. Patch clamp experiments demonstrated that CaMKII inhibition via the chemical inhibitor KN-93 (p=0.02) and the peptide inhibitor CN21 (p<0.01) reversed ISO-treatment associated inhibition of IKs activation current density relative to appropriate controls (KN-92 and CN21-Alanine, respectively). Inhibition with KN-93 and CN21 (p<0.01 for both) also reversed ISO-treatment associated right shifts in the voltage dependence of activation relative to appropriate controls. The ability of ISO treatment to activate CaMKII in HEK 293 cells was confirmed via protein immunoblot wherein T287 phosphorylation (CaMKII residue conferring constitutive activity) was increased during ISO treatment (p<0.05). Lentiviral overexpression of δCaMKII inhibited IKs activation current density with WT IKs (p=0.01) but not with Triple-A IKs (p=0.20) relative to lentiviral control. Inhibition of IKs activation current density during δCaMKII overexpression was attenuated with S484A IKs (p=0.04) but not with S457A (p=0.99) relative to WT IKs during δCaMKII overexpression. The voltage dependence of activation was also right-shifted during δCaMKII overexpression relative to lentiviral control (p=0.03). PKA inhibition with the peptide inhibitor PKI did not reverse ISO-treatment associated inhibition of IKs activation current density (p=0.51), and PKA did not phosphorylate peptide fragments corresponding to any of residues identified via LCMS/MS.

Aim 3 investigated the mechanism through which CaMKII-mediated phosphorylation at KCNQ1 S484 inhibits IKs function. To assess whether interaction with KCNE1 affects KCNQ1 phosphorylation, we performed LCMS/MS experiments during expression of KCNQ1 alone and during co-expression with KCNE1. Phosphorylation at S484 was reduced during co-expression with KCNE1 relative to expression of KCNQ1 alone (p<0.01). In addition, mimics of phosphorylation at S484 (S484D) did not affect activation current density (p=0.96) or the voltage dependence of activation (p=0.51) relative to dephosphorylation mimics (S484A). Based on these results, we hypothesized that S484 phosphorylation affected the interaction between KCNQ1 and KCNE1 subunits; accordingly, we assessed the KCNQ1-KCNE1 interaction using BiFC experiments in HEK 293 cells. In accordance with our hypothesis, Venus fluorescent intensity (corresponding to KCNQ1-KCNE1 interaction) was reduced during ISO treatment relative to vehicle (p<0.05) and with S484D KCNQ1 relative to S484A (p<0.01). The role of CaMKII in mediating this disruption of KCNQ1-KCNE1 interaction was demonstrated BiFC experiments that showed co-treatment with ISO and KN-93 attenuated reduced Venus intensity during co-treatment with ISO and KN-92 (p<0.01). These results were corroborated by BiFC experiments with Long QT Syndrome Phenotype 1 (LQT1) mutations that demonstrated that an LQT1 mutation predicted to disrupt CaMKII phosphorylation at S484 (R481I) attenuated reduced Venus intensity during ISO treatment relative to an LQT1 mutations predicted to not affect CaMKII regulation of S484 (S484T; p<0.01). The ability of S484 phosphorylation to affect KCNQ1 and/or KCNE1 trafficking was assessed via protein immunoblot experiments to detect KCNQ1 and KCNE1 following a biotinylation procedure to isolate plasma membrane-bound proteins. Biotinylation experiments demonstrated that KCNQ1 and KCNE1 plasma membrane expression were reduced by ~15% and ~33%, respectively, with S484D KCNQ1 relative to S484A (p<0.05 for both).

Conclusion: CaMKII phosphorylates KCNQ1 S484 during sustained β-AR stimulation to inhibit IKs function. S484 phosphorylation inhibits IKs function by disrupting the interaction between KCNQ1 and KCNE1 subunits and by reducing the plasma membrane expression of KCNQ1 and KCNE1. Pathological regulation of KCNQ1 by CaMKII (and subsequent inhibition of IKs) during sustained β-AR stimulation may contribute to increased arrhythmogenesis during physiologic states of chronically increased catecholaminergic tone, such as during HF.