Patient-independent human induced pluripotent stem cell model: A new tool for rapid determination of genetic variant pathogenicity in long QT syndrome

Abstract

Background: Commercial genetic testing for long QT syndrome (LQTS) has rapidly expanded, but the inability to accurately predict whether a rare variant is pathogenic has limited its clinical benefit. Novel missense variants are routinely reported as variant of unknown significance (VUS) and cannot be used to screen family members at risk for sudden cardiac death. Better approaches to determine the pathogenicity of VUS are needed.

Objective: The purpose of this study was to rapidly determine the pathogenicity of a CACNA1C variant reported by commercial genetic testing as a VUS using a patient-independent human induced pluripotent stem cell (hiPSC) model.

Methods: Using CRISPR/Cas9 genome editing, CACNA1C-p.N639T was introduced into a previously established hiPSC from an unrelated healthy volunteer, thereby generating a patient-independent hiPSC model. Three independent heterozygous N639T hiPSC lines were generated and differentiated into cardiomyocytes (CM). Electrophysiological properties of N639T hiPSC-CM were compared to those of isogenic and population control hiPSC-CM by measuring the extracellular field potential (EFP) of 96-well hiPSC-CM monolayers and by patch clamp.

Results: Significant EFP prolongation was observed only in optically stimulated but not in spontaneously beating N639T hiPSC-CM. Patch-clamp studies revealed that N639T prolonged the ventricular action potential by slowing voltage-dependent inactivation of Ca_V1.2 currents. Heterologous expression studies confirmed the effect of N639T on Ca_V1.2 inactivation.

Conclusion: The patient-independent hiPSC model enabled rapid generation of functional data to support reclassification of a CACNA1C VUS to likely pathogenic, thereby establishing a novel LQTS type 8 mutation. Furthermore, our results indicate the importance of controlling beating rates to evaluate the functional significance of LQTS VUS in high-throughput hiPSC-CM assays.

Keywords: Action potential; Extracellular field potential; Human induced pluripotent stem cells; L-type Ca current; Long QT syndrome type 8.

Figure 1.. CACNA1C VUS in kindred with prolonged QT.

(A) Timeline of clinical events for proband (P), brother (B), and mother (M). Time-scale relative to proband’s age at the time of event. (B) Family pedigree. Arrow denotes proband, circles denote female, squares denote male; diagonal line denotes decedent. (C) Lead II from 12-lead ECG obtained from proband. QT rate-corrected with Bazett formula (QTc). (D) Protein topology map of Ca_V1.2 encoded by CACNA1C. Hollow circles represent mutation sites previously reported to prolong QT interval. Red arrow indicates localization of N639T. No LQTS mutations have previously been reported in this region.

Figure 2.. Generation of the patient-independent hiPSC-CM model.

Unrelated healthy donor human induced pluripotent stem cells (hiPSC) that had been previously generated and characterized were used as starting material. Using CRISPR/Cas9 gene editing, the c.C1916A variant was introduced into CACNA1C and confirmed by Sanger sequencing. This heterozygous missense mutation results in an amino acid substitution from asparagine to threonine at position 639 (pN639T). hiPSC lines were then differentiated to generate N639T mutant CM and isogenic control CM.

Figure 3.. EFP measurements in spontaneously-beating hiPSC-CM monolayers.

(A) Timeline and experimental reagents for generating human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CM). At day 30, hiPSC-CM were plated on 96-well plates and EFP recorded at hiPSC-CM day 40. (B) Representative EFP waveform of spontaneously-beating isogenic control (CTRL) hiPSC-CM. Starting from isoelectric baseline, measurements were obtained for the depolarization time (DT), repolarization time (RT), and maximum field potential duration (FPDmax, defined as the time-point when the EFP waveform returned to isoelectric baseline). (C) Comparison of time-based parameters between N639T and isogenic CTRL hiPSC-CM. N = 48 (CTRL) and 42 (N639T, L1) wells, **p<0.01 (D) Rate-correction of FPDmax using Bazett, Fredericia, and Framingham formulae. Values are plotted as median, interquartile range (box), and range (error bars). All p-values obtained using Mann-Whitney test.

Figure 4.. EFP and AP measurements in paced hiPSC-CM.

Representative EFP waveforms (A) and median values (B) for isogenic CTRL and CACNA1C-N639T hiPSC-CM optically paced at 1 Hz. Two control lines: isogenic (I, n=34 wells) and population (P, n=41 wells). Three independently generated heterozygote variant lines: N639T L1 (n=75), N639T L2 (n=13), N639T L3 (n=16). No significant difference between median values for isogenic and population controls (p=0.99). No significant difference between variant lines 1 & 2 (p=0.56), 1 & 3 (p=0.99), and 2 & 3 (p=0.91). ***p<0.001 by Mann-Whitney test compared to both isogenic and population control. (C) Representative AP waveforms of single N639T (L1) and isogenic CTRL hiPSC-CM. (D) AP measurements at 10%, 50%, 90% repolarization (APD10, APD50, APD90). CTRL (n=19) & N639T (n=15), **p < 0.01 by Mann-Whitney test. Range indicated by error bars, interquartile values indicated by box.

Figure 5.. Characterization of Ca_V1.2 currents in single hiPSC-CM.

(A) Representative current records in response to a step membrane depolarization using Ba²⁺ as a charge carrier (I_Ba). (B) Current-voltage relationship for isogenic CTRL (n=12) and N639T (L1, n=10). No significant difference between the two lines was detected (p=0.83, Welch’s t test). (C) Steady-state inactivation curves of isogenic CTRL (n=12) and N639T (L1, n=12). No significant difference between the two lines was detected (p=0.66, Welch’s t test). (D) Representative I_Ba records in response to a long step depolarization. The declining phase of the trace was fitted by a single-exponential function to calculate inactivation time constant. (E) Time constant of I_Ba inactivation. (F) %I_Ba remaining after 350-ms depolarization to −10 mV. CTRL (n=10), N639T (n=8). Range indicated by error bars, interquartile values indicated by box, **p<0.01 by Mann-Whitney test.

Figure 6.. Characterization of Ca_V1.2 currents in HEK293 cells.

(A) Representative current records in response to a step membrane depolarization using Ba²⁺ as a charge carrier (I_Ba). (B) Current-voltage relationship for control (CTRL, n=17) and N639T (n=14). Peak current density was significantly smaller in N639 group compared to control group (p<0.01, t-test). (C) Steady-state inactivation of CTRL (n=17) and N639T (n=16). V_0.5=−30.3±1.5 mV for CTRL vs. −24.5±1.2 mV for N639T (p<0.01). k=4.2±0.3 mV for CTRL vs. 5.2±0.6 mV for N639T (p=0.2). (D) Representative I_Ba records in response to a long step depolarization. The declining phase of the trace was fitted by a single-exponential function to calculate inactivation time constant. (E) Average time constant of I_Ba inactivation. (F) %I_Ba remaining after 350-ms depolarization. CTRL (n=17), N639T (n=17). Range indicated by error bars, interquartile values indicated by box, ***p<0.001 by Mann-Whitney test.

Figure 7.. Comparison of time (A) and reagent cost (B) to generate and characterize patient-specific hiPSC, patient-independent hiPSC and heterologous expression models.

Listed are estimated minimum times required by a research laboratory experienced in the three approaches. The costs of reagents are based on actual costs to our laboratory based in the United States. The E8 stem cell culture media was prepared in-house, which resulted in significant savings compared to purchasing it commercially.