Partially dominant mutant channel defect corresponding with intermediate LQT2 phenotype

Abstract

Background: The hereditary Long QT Syndrome is a common cardiac disorder where ventricular repolarization is delayed, abnormally prolonging the QTc interval on electrocardiograms. LQTS is linked to various genetic loci, including the KCNH2 (HERG) gene that encodes the α-subunit of the cardiac potassium channel that carries I(Kr). Here, we report and characterize a novel pathologic missense mutation, G816V HERG, in a patient with sudden cardiac death.

Methods: Autopsy-derived tissue sample was used for DNA extraction and sequencing from an unexpected sudden death victim. The G816V HERG mutation was studied using heterologous expression in mammalian cell culture, whole cell patch clamp, confocal immunofluorescence, and immunochemical analyses.

Results: The mutant G816V HERG channel has reduced protein expression and shows a trafficking defective phenotype that is incapable of carrying current when expressed at physiological temperatures. The mutant channel showed reduced cell surface localization compared to wild-type HERG (WT HERG) but the mutant and wild-type subunits are capable of interacting. Expression studies at reduced temperatures enabled partial rescue of the trafficking defect with appearance of potassium currents, albeit with reduced current density and altered voltage-dependent activation. Lastly, we examined a potential role for hypokalemia as a contributory factor to the patient's lethal arrhythmia by possible low-potassium-induced degradation of WT HERG and haplo-insufficiency of G816V HERG.

Conclusion: The G816V mutation in HERG causes a trafficking defect that acts in a partially dominant negative manner. This intermediate severity defect agrees with the mild clinical presentation in other family members harboring the same mutation. Possible hypokalemia in the proband induced WT HERG degradation combined with haplo-insufficiency may have further compromised repolarization reserve and contributed to the lethal arrhythmia.

Figure 1. Family Pedigree and Location of KCNH2-G816V Mutation

A) The proband is indicated by the arrow. Each individual is assigned as male (squares) or female (circles), carrying the KCNH2-G816V mutation (filled black) or those not carrying the mutation (white). Numbers below indicate QTc intervals from one or more ECGs. B) Leads V4 and II of a 12-lead ECG obtained at rest from the proband showing a prolonged QT interval. C) Schematic representation of the HERG protein topology indicating the position of residue 816 within the C-terminal cyclic-nucleotide binding domain (CNBD).

Figure 2. Baseline Protein Expression of G816V HERG compared to WT HERG in HEK 293 cells

A) Western blot shows whole cell lysate of HEK 293 cells transfected with either WT HERG or G816V HERG plasmid DNA alone or in a 50/50 mix at 48 and 72 hours post-transfection. Antibodies used were against the Myc-tag for both types of HERG and against Na⁺/K⁺ ATPase as a loading control. The molecular weight marker appears at left and HERG appears as a doublet of two bands at 135 kDa and 155 kDa (arrows) separated on 7.5% SDS-PAGE. B) Histogram showing summary data of total HERG protein expression levels normalized to Na⁺/K⁺ ATPase. Graph shows densitometry quantification of total HERG where protein expression of total G816V HERG compared to total WT HERG is reduced both at 48 and 72 hours, n=5 with ** p-value < 0.01 and ***p-value < 0.001. C) Histogram showing the distribution of 135 kD and 155 kD HERG bands for wild-type, 50–50 mix and G816V with totals normalized to 1.

Figure 3. Effect of the G816V Mutation on I_Kr and I_Ks Currents

A) Whole cell current traces of HERG channels with voltage protocol below and brackets indicate the portion of the current tracing that is shown. WT HERG and G816V HERG channels were transiently expressed in CHO cells either alone, in a 50% WT/50% G816V mix, or 50% WT/50% empty vector mix. B) Current density at 48 hrs shown as a function of the fraction of WT subunits. Measurements for current density were taken as peak amplitude at the −40 mV tail current. Numbers in parentheses indicate number of cells assayed. The dashed line indicates the expected relationship for a fully dominant-negative mutant that does not carry current in tetrameric assembly. The dotted line indicates the expected relationship for a non dominant-negative mutant in tetrameric assembly. C) Histogram comparing current density of 50% WT/50% empty vector mix and 50% WT/50% G816V HERG. The difference was not significant. D) Normalized current-voltage relationship shown for WT HERG and 50/50 mixed sample. The measurements for the current-voltage curve were taken at the end of the 3 second depolarization pulse. E) Normalized voltage-dependent activation curves. Curves were fitted using a Boltzmann function. V_1/2 for WT HERG = −17.96±2.1 mV and WT/G816V HERG = −23.67±2.68 mV. n = 8–11, the difference was not significant. F) Whole cell current traces at the −120 tail current for IKs alone, IKs with WT HERG and IKs with G816V HERG. Voltage protocol is shown below and brackets indicate portion of the current trace that is shown. G) Histogram shows summary data for tau of deactivation and numbers in parentheses indicated number of cells assayed. Tail currents were fit with a single-exponential function and the difference in tau for deactivation between I_Ks alone and I_Ks with G816V HERG was significant * p-value < 0.05.

Figure 4. Cell Surface Expression of G816V HERG compared to WT HERG

A) Immunoblot of cell surface proteins which were isolated by biotin labeling and pull-down from transiently transfected HEK cells. Samples were separated by 7.5% linear SDS-PAGE. Calnexin was the negative control for surface labeling and cadherin was the positive control for surface labeling in the lower panels. Western blot was performed with anti-HERG, anti-calnexin, and anti-cadherin antibodies. B) Summary data for part A. Densitometry analysis of the surface expression of HERG proteins was quantified as the amount of surface HERG divided by the total cellular HERG and also normalized for streptavidin pull-down of biotinylated protein (normalization calculated using cadherin and calnexin controls). WT HERG is normalized to 1.0. n=4, * p<0.05. C) Confocal immunofluorescence micrographs of HEK 293 cells transfected with either WT HERG or G816V HERG and counter-stained with anti-cadherin antibody to indicate the cell membrane. Pearson correlation of HERG and cadherin colocalization is 0.571±0.025 for WT HERG and 0.332±0.029 for G816V, p<0.001. n=15, scale bar=10 μm, DIC = differential interference contrast. D) Confocal immunofluorescence micrographs of HEK 293 cells transfected with either HERG-WT or HERG-G816V and counter-stained with anti-calnexin antibody to indicate the ER, an intracellular compartment. Pearson correlation of HERG and calnexin colocalization is 0.831±0.026 for WT HERG and 0.916±0.006 for G816V, p<0.01. n=13, scale bar=10 μm.

Figure 5. Interaction of WT HERG and G816V HERG subunits

A) Co-immunoprecipitation and subsequent Western blot from transiently transfected HEK 293 cells shows 3x FLAG-tagged WT HERG detected from a pull-down of myc-tagged G816V HERG in 3 different mixed combinations with 100% WT HERG and 100% G816V HERG as controls, n=3. Lower panel shows the reverse experiment where myc-tagged G816V HERG was detected in a pull-down of 3x FLAG-tagged WT HERG from the same combinations, n=3. B) Immuno-blots of cell surface biotinylation of 3x FLAG-tagged WT HERG and myc-tagged G816V HERG alone or in a 50/50 mix. Proteins were separated by 7.5% SDS-PAGE. Top panel shows a blot with anti-FLAG antibody with cadherin and calnexin as controls. Lower panel shows a blot with anti-Myc antibody with cadherin and calnexin as controls, n=3.

Figure 6. Protein Expression and Electrophysiology of G816V HERG at Reduced Temperature

A) Western blots of whole cell lysate from transiently transfected HEK 293 cells with WT HERG, 50/50 mixed, or G816V HERG at 48, 72 and 96 hours post-transfection. Anti-HERG antibody and anti-Na⁺/K⁺ ATPase antibodies were used. Top panels show the experiment done at 37°C incubation and middle panels show the same experiment done at 30 degrees incubation. Bottom graph shows a histogram of densitometry quantification of percent of 155 kD over total HERG for the two temperatures at the 96 hour time point, n=3. B) Whole cell current traces of HERG channels patched at 30°C with voltage protocol shown below. WT HERG and G816V HERG channels were transiently expressed in CHO cells either alone, or in a 50% WT/50% G816V mix. C) Current density at 72–96 hours and 30°C shown as a function of the amount of WT subunits. Numbers in parentheses indicate number of cells assayed. D) Normalized current-voltage relationship shown for WT HERG and G816V HERG at 30°C. E) Normalized voltage-dependent activation curves. Curves were fitted using a Boltzmann function. V_1/2 for WT HERG = −8.43±3.8 mV and G816V HERG = −22.64±2.8 mV. n = 8–11. The difference was statistically significant, p-value < 0.05.

Figure 7. Low Potassium Induced Degradation of HERG

Western blots showing whole cell lysate of HEK 293 cells stably expressing WT HERG (left) or G816V HERG (right) treated with low-potassium media for 24 hours. The potassium concentration is indicated at the top and the concentration of potassium in regular media is 5.3 mM. Tubulin was used as the loading control, n=4.