Calcium dysregulation underlies phenotypic diversity in LQT2: Insights from induced pluripotent stem cell-derived cardiomyocytes of a KCNH2 p.Y427H family trio

Abstract

BACKGROUND

Congenital long QT syndrome (LQTS) is an inherited arrhythmia linked to a high risk of sudden cardiac death. LQTS type 2 is mainly caused by KCNH2 mutations. Even among carriers of the same pathogenic KCNH2 variant, clinical manifestations and disease severity vary, with the underlying mechanisms remaining unclear.

AIM

To investigate mechanisms underlying variable clinical severity among carriers of the same KCNH2 mutation using patient-specific induced pluripotent stem cell (iPSC)-derived cardiomyocytes (iPSC-CMs).

METHODS

iPSCs were generated from the peripheral blood mononuclear cells of a family trio: A symptomatic proband, an asymptomatic carrier mother, and a mutation-negative father. Using nonintegrating Sendai virus, these iPSCs were then differentiated into cardiomyocytes using a monolayer-based protocol. The functional properties of the iPSC-CMs were assessed using patch-clamp recordings, calcium imaging, and multi-electrode array recordings.

RESULTS

The proband iPSC-CMs exhibited a significant prolongation of action potential duration compared with the mother and father iPSC-CMs. Both the proband and mother iPSC-CMs, carrying the heterozygous KCNH2 p.Y427H variant, demonstrated a significant reduction in the rapid delayed rectifier potassium current, consistent with a loss-of-function mutation. Notably, the proband iPSC-CMs showed an enhanced L-type calcium current and exaggerated calcium transients. These abnormalities were accompanied by increased arrhythmogenicity, manifesting as early afterdepolarizations and triggered arrhythmias. Treatment with nifedipine partially restored repolarization and effectively suppressed all these arrhythmic events.

CONCLUSION

Calcium dysregulation is a critical modifier of phenotypic severity. Patient-specific iPSC-CMs provide a powerful platform for modeling intra-familial phenotypic variability in LQTS type 2, highlighting their potential for individualized risk assessment and therapeutic evaluation.

Key Words: Congenital long QT syndrome; Induced pluripotent stem cell-derived cardiomyocytes; Precision medicine; Calcium dysregulation; Long QT syndrome type 2

Core Tip: Using induced pluripotent stem cell-derived cardiomyocytes, we modeled a family with congenital long QT syndrome type 2 caused by a KCNH2 variant to explain why clinical severity differs among carriers. Although both affected individuals showed a reduced rapid delayed rectifier potassium current, only the symptomatic proband exhibited calcium handling abnormalities, increased L-type calcium current, and spontaneous arrhythmias. Pharmacological correction with nifedipine suppressed these arrhythmias. This work highlights calcium dysregulation as a key modifier of disease severity and supports the use of patient-specific induced pluripotent stem cell-derived cardiomyocytes for personalized risk assessment and therapy selection.

INTRODUCTION

Congenital long QT syndrome (LQTS) is an autosomal dominant ion channel disorder characterized by prolonged QT intervals and abnormal T waves on the electrocardiogram. These repolarization abnormalities increase the risk of life-threatening arrhythmias, particularly torsades de pointes, which can lead to syncope, sudden cardiac arrest, or sudden cardiac death (SCD)[1,2]. Epidemiological data[3-5] show that the prevalence of LQTS is approximately 1 in 2000 and that the annual SCD rate is 0.3%-0.6% in untreated patients. Notably, LQTS is a leading cause of SCD in children and adolescents, imposing a significant physical and psychological burden on patients and their families.

The pathogenic substrate of LQTS is linked to mutations in at least 17 genes, 11 of which have been identified as pathogenic[6,7]. These genes encode voltage-gated ion channels and their regulatory proteins, primarily involved in repolarization reserve currents[8,9], including I_Ks, I_Kr, I_Na, and I_CaL. Among the various subtypes, congenital LQTS type 2 (LQT2), caused by mutations in KCNH2 encoding the Kv11.1 (hERG) potassium channel, accounts for approximately 30% of LQTS cases. However, clinical outcomes can vary widely even among individuals with the same mutation. Some patients exhibit QT/QTc prolongation in utero, while others remain asymptomatic throughout life. In certain cases, the first clinical manifestation may be SCD[10]. This variable expressivity underscores the need for further investigation into the underlying pathogenic mechanisms to enhance risk stratification and enable personalized treatment strategies.

Traditional heterologous systems and animal models have advanced our understanding of ion channel function but cannot fully replicate human cardiomyocyte (CM) characteristics or disease heterogeneity[11]. In contrast, induced pluripotent stem cell-derived CMs (iPSC-CMs) closely resemble the electrophysiological function of normal embryonic CMs, making them an ideal model for LQTS research[12,13]. In LQT1, Moretti et al[14] first showed that iPSC-CMs from KCNQ1 p.R190Q patients had a prolonged action potential duration (APD) and reduced I_Ks, with increased early afterdepolarizations (EADs) that were rescued by β-blockers. In LQT2, Itzhaki et al[15] first found that iPSC-CMs with the KCNH2 p.A614V mutation displayed APD prolongation and triggered arrhythmias (TAs) due to reduced I_Kr, which could be rescued by nifedipine. In LQT3, Ma et al[16] first generated iPSC-CMs with the SCN5A p.V1763M mutation, showing increased I_Na,L and APD prolongation. Moreover, iPSC-CM models also replicate other I_CaL-related LQTS subtypes, such as Timothy syndrome[17] and Triadin knockout syndrome[18]. By uncovering pro-arrhythmic mechanisms related to the interplay of different ion currents and abnormal calcium handling, iPSC-CMs help explain the development of EADs/TAs, which are key pathogenic mechanisms of ventricular arrhythmia.

Here, we investigated a family trio carrying the KCNH2 p.Y427H mutation, consisting of a symptomatic proband with nine cardiac events, an asymptomatic mutation-positive mother, and a mutation-negative father. To recapitulate the intrafamilial genotype-phenotype discordance observed clinically, we generated iPSC-CMs from this family trio. Comprehensive functional profiling, including patch-clamp recordings, calcium fluorescence imaging, and multi-electrode array (MEA) analyses, was performed to elucidate the cellular mechanisms underlying phenotypic variability.

MATERIALS AND METHODS

Patient recruitment and ethics

A Chinese family clinically diagnosed with an autosomal dominant LQT2 phenotype, carrying the KCNH2 p.Y427H mutation (Figure 1A), was enrolled in the study, with a follow-up period of over 20 years. The female proband first presented with syncope at age 14, with a 12-lead electrocardiogram showing a maximum QT/QTc interval of 640 milliseconds/618 milliseconds. She was treated according to guideline recommendations[19-21], including propranolol and mexiletine, achieving a minimum QT/QTc interval of 500 milliseconds/487 milliseconds during follow-up. Despite optimal treatment, she experienced several episodes of syncope, likely due to medication noncompliance or other unidentified causes. She subsequently underwent implantation of a subcutaneous implantable cardioverter defibrillator for SCD prevention. Post-implantation, the subcutaneous implantable cardioverter defibrillator appropriately delivered shocks, terminating three episodes of malignant arrhythmias, including torsades de pointes and ventricular fibrillation. In total, the proband experienced nine recurrent cardiac events. In contrast, the proband’s mother, carrying the same mutation, had a QT/QTc interval of 560 milliseconds/583 milliseconds on the 12-lead electrocardiogram but did not experience any syncope. The proband’s father, who did not carry the mutation, had a normal QT/QTc interval of 386 milliseconds/434 milliseconds and no syncope (Figure 1B). Two other family members, presumed to be LQT2, suffered SCD at age 17 (pedigree screening is shown in Supplementary Figure 1 and Supplementary Table 1). Genetic analysis revealed the KCNH2 p.Y427H mutation, classified as a likely pathogenic variant according to the 2015 American College of Medical Genetics and Genomics guidelines[22]. The mutation leads to a substitution of tyrosine with histidine at position 427 in the S1-topological domain-S2 of Kv11.1 (hERG) protein (Figure 1C). iPSCs were derived from the peripheral blood mononuclear cells (PBMCs) of the proband and her parents to study the variable expression of LQT2. The study was approved by the Institutional Review Board of Beijing Tsinghua Changgung Hospital (Approval No. 24472-4-01).

Figure 1 Clinical and genetic characterization of a family trio. A: Pedigree of the family trio with congenital long QT syndrome type 2 (LQT2). The proband is indicated by an arrow. The genotypes and clinical phenotypes are annotated; B: 12-lead electrocardiogram of the family trio and treatment of the proband. QTc duration shortening following combination therapy with propranolol and mexiletine. The subcutaneous implantable cardioverter defibrillator tracings confirm appropriate shocks that terminate torsades de pointes and ventricular fibrillation; C: Sanger sequencing confirms heterozygosity for the KCNH2 p.Y427H variant in the proband and her mother, whereas the father does not carry this variant, consistent with findings from their induced pluripotent stem cells. The topology of the KCNH2-encoded Kv11.1 (hERG) channel shows the p.Y427H substitution, where tyrosine at position 427 is replaced by histidine in the S1-S2 topological domain. Tyr: Tyrosine; His: Histidine; CE: Cardiac event.

Genetic sequencing and variant validation

All candidate and known pathogenic variants were initially screened using next-generation sequencing and subsequently confirmed by Sanger sequencing. The panel included 89 genes associated with genetic arrhythmias and cardiomyopathies, such as ABCB1, ABCC9, AKAP9, ALG10, ANK2, BAG3, CACNA1C, CACNA2D1, CACNB2, CALM1, CALM2, CALM3, CASQ2, CAV3, CAVIN1, CDH2, CELF4, CERKL, CTNNA3, CYP2B6, CYP2C19, CYP2C9, CYP2D6, DBH, DES, DNAJC19, DPP6, DSC2, DSG2, DSP, FLNC, GATA5, GATA6, GJA1, GJA5, GNAI2, GPD1 L, HADHA, HCN4, JUP, KCNA5, KCND2, KCND3, KCNE1, KCNE2, KCNE3, KCNE5, KCNH2, KCNJ2, KCNJ5, KCNJ8, KCNQ1, LDB3, LMNA, MYH6, MYH7, MYL4, NKX2-5, NOS1AP, NPPA, NRG3, NUBPL, NUP155, PKP2, PLN, PPA2, PRKAG2, RYR2, SALL4, SCN10A, SCN1B, SCN2B, SCN3B, SCN4B, SCN5A, SDHAF3, SLC22A23, SLCO3A1, SNTA1, TBX5, TECRL, TGFB3, TMEM43, TNNI3, TNNI3K, TNNT2, TRDN, TRPM4, and TTN. Pathogenicity was assessed according to the 2015 American College of Medical Genetics and Genomics guidelines[22]. Variants identified in other genes were classified as benign based on allele frequencies in the general population ≥ 1%. Accordingly, mutations were confirmed by Sanger sequencing after reprogramming PBMCs into iPSCs.

Generation and characterization of iPSCs

Isolation and activation of PBMCs: PBMCs were isolated from peripheral blood using Ficoll-Paque density gradient centrifugation. The cells were resuspended in PBMC activation medium (Cellapy, #3408025, China) and cultured at 37 °C in a humidified incubator with 5% CO₂. On day 7, 1 × 10⁵ PBMCs were collected for reprogramming.

Sendai virus reprogramming: Reprogramming was initiated on day 0 using the Reproeasy iPSCs Reprogramming Kit (Cellapy, #5002002, China) with Sendai virus vectors carrying KOS (MOI = 5), hc-Myc (MOI = 5), and hKlf4 (MOI = 3) for 24 hours. From day 8, emerging colonies were cultured and the medium was switched from Reproeasy medium (Cellapy, #5003050, China) to PGM1 medium (Cellapy, #1007500, China). From day 14, iPSC colonies with characteristic morphology were manually picked and expanded.

Karyotype analysis: Genomic stability was evaluated by karyotyping iPSCs at passage ≥ 10 using G-banding. Cells were treated with colcemid, and slides were stained with Giemsa. For each cell line, 20 metaphase spreads were examined to detect numerical and large structural abnormalities.

Short tandem repeat profiling: The PBMC-derived iPSCs were verified for source accuracy, purity, and consistency with the original sample by short tandem repeat (STR) profiling using the Microread Genomic DNA Kit and Microreader 21 ID System. Polymerase chain reaction products were analyzed with an ABI 3730xl Genetic Analyzer (Applied Biosystems, CA, United States) and interpreted using GeneMapper ID-X software (Applied Biosystems, CA, United States).

Immunofluorescence staining for pluripotency markers: Pluripotency was assessed by immunofluorescence staining using primary antibodies against OCT4 (Santa Cruz, #SC-9081, TX, United States), NANOG (Santa Cruz, #SC-293121, TX, United States), and SOX2 (Santa Cruz, #SC-365823, TX, United States), all diluted 1:100. After incubation, the cells were stained with Alexa Fluor 488-conjugated secondary antibody (Thermo Fisher, #A28175; 1:200, MA, United States). Nuclei were counterstained with DAPI (Solarbio, #C0060, China). Images were acquired using a confocal laser scanning microscope (LSM 980, Zeiss, Germany).

Teratoma formation assay: The tri-lineage differentiation potential of the iPSCs was demonstrated using a teratoma formation assay. Approximately 1 × 10⁷ iPSCs were subcutaneously injected into 4-week-old male NOD/SCID mice (Vital River, China). After 8-12 weeks, tumors formed at the injection site. Mice were euthanized, and tumors were excised, fixed in 4% paraformaldehyde (Sigma, #P6148, MO, United States), and processed for paraffin embedding.

Differentiation and characterization of CMs

CM differentiation: Cardiac differentiation of human iPSCs was induced using the CardioEasy kit (Cellapy, #2004500, China). iPSCs (1 × 10⁶) were seeded into 12-well plates pre-coated with PSCeasy coating solution (Cellapy, #CA3003100, China) and cultured in PGM1 medium (Cellapy, #1007500, China) until 80%-90% confluency. On day 0, the medium was replaced with CardioEasy Medium 1 (Cellapy, #2004500-1, China). On day 3, the medium was switched to CardioEasy Medium 2 (Cellapy, #2004500-2, China), and on day 5 to CardioEasy Medium 3 (Cellapy, #2004500-3, China). From days 7 to 15, iPSC-CMs began spontaneous beating. Metabolic purification was performed using CardioEasy medium (Cellapy, #2005100, China) for 2 days. After purification, cells were maintained in CardioEasy medium (Cellapy, #2015500, China). Experiments began after 25-30 days of differentiation.

Immunofluorescence staining for CM markers: CM markers were assessed by immunofluorescence staining using primary antibodies against cardiac troponin T (Protein Tech, #15513-1-AP, IL, United States) and α-actinin (Sigma, #A7811, MO, United States), both diluted 1:100. After incubation, the cells were stained with Alexa Fluor 594 (Thermo Fisher, #A-11012, MA, United States) and Alexa Fluor 488 (Southern Biotech, #6411-30, AL, United States), both diluted 1:200, and the nuclei were counterstained with DAPI (Solarbio, #C0060, China). Images were acquired using a confocal laser scanning microscope (LSM 980, Zeiss, Germany).

Whole-cell patch-clamp recording

Whole-cell patch-clamp recordings, including current clamp and voltage clamp techniques, were performed on iPSC-CMs to evaluate the action potentials (APs) and ionic currents. Recordings were made at 36 ± 2 °C using a patch-clamp system (HEKA, Germany). After achieving a high-resistance seal, the whole-cell configuration was established, and series resistance was compensated.

The extracellular and intracellular solutions for the AP recordings are provided below. The external solution (in mmol/L) contained 137 NaCl, 4 KCl, 10 HEPES, 10 D-glucose, 1 MgCl₂, and 1.8 CaCl₂; pH 7.4 (NaOH). The intracellular solution (in mmol/L) contained 135 KCl, 2 MgCl₂, 1 EGTA, 10 HEPES, 4 Mg-ATP, and 0.3 Na-GTP; pH 7.4 (KOH). For the I_Kr (Kv11.1) measurement, the external solution was identical to the AP solution, with the intracellular solution (in mmol/L) containing 55 KCL, 10 HEPES, 5 EGTA, 100 K-gluconate, and 5 Mg-ATP; pH 7.2 (KOH). For the I_CaL (Cav1.2) measurement, the extracellular solution (in mmol/L) contained 140 TEA-Cl, 4 KCl, 10 HEPES, 5 D-glucose, 1 MgCl₂, and 10 CaCl₂; pH 7.4 (TEA-OH). The intracellular solution (in mmol/L) contained 110 CsCl, 5 HEPES, 10 EGTA, 4 Na₂-ATP, 4.5 Mg-ATP, and 1 CaCl₂; pH 7.2 (CsOH).

For the current-clamp recordings, the APs were recorded under current-clamp mode with the membrane potential held at -80 mV. Stimulated APs were paced at 0.5 Hz and 1 Hz with a current strength of 1.2 times the threshold for 10 milliseconds. The parameters measured included resting membrane potential, AP amplitude, and AP duration at 50% and 90% repolarization (APD₅₀, APD₉₀). Spontaneous activity was monitored for EADs and TAs. AP with an obvious plateau phase and an APD₉₀/APD₅₀ ratio ≤ 1.3 were classified as ventricular-like AP for analysis.

For the voltage-clamp recordings for I_Kr, a holding potential of -80 mV and depolarization to -40 mV, followed by voltage steps from -40 mV to +40 mV in 20 mV increments for 3 seconds were carried out. Tail currents were elicited by a repolarization phase to -50 mV for 2 seconds. I_Kr was defined as the E-4031-sensitive current and calculated by subtraction before and after the application of 1 μmol/L E-4031 (MedChemExpress, #HY-15551, NJ, United States). For I_CaL, activation kinetics were assessed by 300 milliseconds depolarizing steps from -60 mV to +80 mV in 10 mV increments. The I_CaL inactivation kinetics were assessed followed by repolarization to -60 mV for 2 seconds and depolarization to +10 mV for 300 milliseconds. The inactivation curve was fitted to a Boltzmann function to determine the half-inactivation voltage and slope factor.

Pharmacological agents were dissolved in the AP external solution with a final dimethyl sulfoxide concentration below 0.1%. Solutions were perfused using a perfusion system (Mappinglab PVG-08, United Kingdom). After baseline stabilization, cells were perfused with the drug solution for at least 5 minutes before data collection. The effect of 300 nmol/L nifedipine (MedChemExpress, #HY-B0284, NJ, United States) on paced and spontaneous rhythms of AP was evaluated, and EADs/TAs were recorded before and after drug application.

Calcium imaging

Calcium transients were recorded using the Mappinglab Calcium Imaging System (OMS-PCIE-2022, United Kingdom). iPSC-CMs were incubated with 2 μmol/L Rhod-2 AM (Abcam, #AB142780, MA, United States) and 0.02% Pluronic F-127 (Invitrogen, #P3000MP, CA, United States) for 30 minutes at 37 °C, followed by a 30-minute de-esterification period in Tyrode’s solution containing 1.8 mmol/L Ca²⁺. Stimulation at 0.5 Hz with a current strength of 1.2 times the threshold for 30 milliseconds was delivered using a stimulation device. Regions of interest were manually selected, and the fluorescence intensity (ΔF/F₀) was calculated automatically using OMAPScope software (version 5.9.41, United Kingdom). Quantitative parameters extracted from calcium transients included calcium transient duration (CaTD), upstroke time, decay time (from peak to baseline), and transient amplitude.

MEA recording

Field potential recordings were performed using the CardioExcyte Control 96 system (Nanion, Germany). iPSC-CMs were seeded at 50000 cells per well onto fibronectin-coated CardioExcyte 96-well plates to form a 200 μL system per well. Cells were cultured at 37 °C with 5% CO₂, and the medium was changed every day for 3 days to ensure restoration of cellular beating and signal stability. Nifedipine was dissolved in dimethyl sulfoxide and applied at a final concentration of 300 nmol/L. iPSC-CMs were incubated with nifedipine at 300 nmol/L for 5 minutes, and the MEA parameters were recorded 5 minutes after drug application. The field potential duration (FPD) was corrected using the Fridericia formula [corrected field potential duration (cFPD) = FPD/RR^0.33]. Spontaneous EADs/TAs were recorded before and after drug application.

Statistical analysis

Quantitative data are presented as the mean ± SEM. Normality was assessed with the Shapiro-Wilk test; nonparametric tests were applied when data were non-normal or sample sizes were small. For normally distributed data, 1-way or 2-way analysis of variance with Tukey’s post hoc test was used. For non-normal data, the Kruskal-Wallis test with Dunn’s post hoc test was applied. Paired t-tests were used for comparisons before and after drug treatment. Categorical data are presented as percentages (%) and compared using the χ² test or Fisher’s exact test as appropriate. P < 0.05 was considered statistically significant. ^aP < 0.05, ^bP < 0.05, and ^cP < 0.05 in the comparisons of proband vs father, mother vs father, and proband vs mother iPSC-CMs, respectively. ^dP < 0.05 in the comparisons of pre- vs post-drug administration. Data processing was done in Excel (version 2021, United States); graphs were generated in Origin (version 2021, United States), and statistical analyses were performed in GraphPad Prism (version 10.1.2, United States). Some components of the graphical abstract were created using Figdraw (version 2.0, China).

RESULTS

Generation and characterization of iPSC-CMs from a family trio

PBMCs from the proband and her parents were reprogrammed into iPSCs using a Sendai virus system (Figure 2A). Typical colonies emerged between days 8 and 13, and they were manually expanded and selected between days 14 and 30. All iPSC lines exhibited normal karyotypes, matched STR profiles, and expressed pluripotency markers (e.g., NANOG, OCT4, SOX2). Trilineage differentiation confirmed the capacity for differentiation into all three germ layers. Cardiac differentiation was achieved through Wnt signaling modulation (Figure 2B), and beating CMs were observed from days 7 to 15; the CMs showed consistent morphology across all lines. Immunostaining for the cardiac markers cardiac troponin T and α-actinin confirmed CM identity. Interestingly, during the late maturation phase of simultaneous multicellular culture in 12-well plates, proband iPSC-CMs exhibited prominent disordered contractions, whereas iPSC-CMs from the mother and father showed stable and rhythmic beating without such abnormalities.

Figure 2 Generation and characterization of family trio induced pluripotent stem cell-derived cardiomyocytes. A: Reprogramming process: Representative bright-field images of peripheral blood mononuclear cells and 10^th-passage induced pluripotent stem cells (iPSCs). G-banding karyotyping and short tandem repeat profiling confirm genomic stability. Immunofluorescence staining shows robust expression of pluripotency markers (NANOG, OCT4, SOX2) in green, with a DAPI counterstain in blue. Teratoma formation experiments in immunodeficient mice demonstrate iPSCs differentiation into all three germ layers: Endoderm (respiratory epithelium), mesoderm (adipocytes), and ectoderm (neuronal cells), confirming pluripotency. Scale bars: 100 μm; B: Differentiation process: Morphological progression during cardiac differentiation. Representative images show early, middle, and late stages of iPSC-derived cardiomyocytes formation, as well as single-cell and cell layer morphology. Immunostaining confirms cardiomyocyte identity and sarcomeric structure with cardiac troponin T in red, α-actinin in green, and DAPI in blue. Scale bars: 100 μm (overview), 10 μm (immunostaining). PBMCs: Peripheral blood mononuclear cells; iPSCs: Induced pluripotent stem cells; STR: Short tandem repeat; iPSC-CMs: Induced pluripotent stem cell-derived cardiomyocytes; cTNT: Cardiac troponin T.

Functional properties of iPSC-CMs from a family trio

iPSC-CMs from the KCNH2 p.Y427H family trio exhibited distinct electrophysiological phenotypes. Under a 0.5 Hz paced rhythm (Figure 3A), the proband iPSC-CMs showed a significantly prolonged APD₉₀ (746.15 ± 26.28 milliseconds) and APD₅₀ (637.83 ± 23.31 milliseconds) compared with those of the mother iPSC-CMs (APD₉₀: 485.70 ± 9.38 milliseconds; APD₅₀: 398.48 ± 9.96 milliseconds; both P < 0.05) and father iPSC-CMs (APD₉₀: 306.89 ± 5.96 milliseconds; APD₅₀: 258.13 ± 5.00 milliseconds; both P < 0.05). Similar differences were observed at 1 Hz. As the frequency increased, the APD was shortened. The APD differences in the iPSC-CMs from the family trio were consistent with the clinical QTc phenotype (Supplementary Table 2).

Figure 3 Action potentials and rapid delayed rectifier potassium current characteristics of family trio induced pluripotent stem cell-derived cardiomyocytes. A: Representative action potentials under 0.5-Hz pacing from father induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) (a1), proband iPSC-CMs (a2), and mother iPSC-CMs (a3). Panel shows resting membrane potential and action potential amplitude (a4), while panels present statistical analysis of APD₉₀ and APD₅₀ (a5 and a6). Data represent n = 15 from 3 independent differentiations; B: Representative total potassium currents from father iPSC-CMs (b1), proband iPSC-CMs (b2), and mother iPSC-CMs (b3). Panel shows the I_Kr current, obtained by subtracting the current before and after E4031 treatment (b4). Panel represents the difference in I_Kr tail current density (b5 and b6). Data represent n = 6 from 3 independent differentiations. iPSC-CMs: Induced pluripotent stem cell-derived cardiomyocytes; APA: Action potential amplitude; RMP: Resting membrane potential.

Given the presence of the likely pathogenic KCNH2 p.Y427H mutation, we initially assessed the repolarization reserve current I_Kr (Figure 3B). The I_Kr tail current density was significantly higher in the father iPSC-CMs (1.58 ± 0.11 pA/pF) compared with both the proband (0.34 ± 0.07 pA/pF; P < 0.05) and mother iPSC-CMs (0.36 ± 0.07 pA/pF; P < 0.05) at a membrane potential of +40 mV. No difference was observed between the proband and mother groups. These results indicate a reduced I_Kr current density in the affected individuals, supporting a loss-of-function effect of the KCNH2 p.Y427H mutation.

Moreover, as I_CaL is another key repolarization reserve current contributing to the plateau phase and APD prolongation, we subsequently evaluated I_CaL (Figure 4A). The proband iPSC-CMs showed a significantly greater I_CaL current density (-18.16 ± 1.64 pA/pF) than the father iPSC-CMs (-10.23 ± 1.29 pA/pF; P < 0.05) and the mother iPSC-CMs (-10.19 ± 1.74 pA/pF; P < 0.05) at a membrane potential of +10 mV. The half-inactivation voltage was less negative in the proband iPSC-CMs (-23.58 ± 1.08 mV) compared with the mother iPSC-CMs (-29.59 ± 1.26 mV; P < 0.05) and father iPSC-CMs (-34.80 ± 1.65 mV; P < 0.05). The slope factor was also smaller in the proband iPSC-CMs (4.60 ± 0.76) than in the mother (7.60 ± 0.51; P < 0.05) and father iPSC-CMs (8.02 ± 0.44; P < 0.05). These findings indicate enhanced Ca²⁺ influx and a broader calcium window current in the proband iPSC-CMs.

Figure 4 Calcium handling and field potential characteristics of family trio induced pluripotent stem cell-derived cardiomyocytes. A: Representative I_CaL from the father induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) (a1), proband iPSC-CMs (a2), and mother iPSC-CMs (a3). Panel shows the I-V curve of I_CaL (a4), panel presents the inactivation curve of I_CaL (a5), and panel shows the statistical data of I_CaL density at a membrane potential of +10 mV (a6). Data represent n = 7 from 3 independent differentiations; B: Representative calcium transient waveforms from the three groups (b1). Statistical analysis of the calcium transient peak amplitude (b2). Statistical analysis of the calcium transient at 90% decay (CaTD₉₀) (b3). Data represent n = 15 regions of interest from 3 independent differentiations; C: Representative field potential waveforms from the three groups (c1). Statistical analysis of the field potential duration and corrected field potential duration (c2 and c3). Data represent n = 15 from 3 independent differentiations. iPSC-CMs: Induced pluripotent stem cell-derived cardiomyocytes; CaT: Calcium transient; CaTD: Calcium transient duration; FP: Field potential; FPD: Field potential duration; cFPD: Corrected field potential duration.

Therefore, we next performed Ca²⁺ transient analysis to assess calcium handling (Figure 4B). The proband iPSC-CMs exhibited a significantly prolonged CaT upstroke time (469.72 ± 11.13 milliseconds) compared with the mother iPSC-CMs (284.63 ± 9.56 milliseconds; P < 0.05) and father iPSC-CMs (228.66 ± 17.63 milliseconds; P < 0.05). CaTD₉₀ was also markedly prolonged in the proband iPSC-CMs (1072.4 ± 23.81 milliseconds) vs the mother iPSC-CMs (909.16 ± 25.55 milliseconds; P < 0.05) and father iPSC-CMs (688.00 ± 27.44 milliseconds; P < 0.05), consistent with the CaTD₅₀ trends. In addition, the CaT amplitude (ΔF/F₀) was significantly higher in the proband iPSC-CMs (0.85 ± 0.04) compared with that in mother iPSC-CMs (0.67 ± 0.02; P < 0.05) and father iPSC-CMs (0.66 ± 0.04; P < 0.05). These abnormalities suggest impaired calcium handling, calcium overload, and increased arrhythmia susceptibility in the proband iPSC-CMs (Supplementary Table 3).

Furthermore, we employed MEA to confirm the repolarization abnormalities of the family trio iPSC-CMs at the multicellular level (Figure 4C). MEA recordings showed a significantly prolonged FPD (830.60 ± 21.21 milliseconds) and cFPD (286.72 ± 8.09 milliseconds) in the proband iPSC-CMs compared with those in the mother iPSC-CMs (FPD: 618.85 ± 7.80 milliseconds; cFPD: 207.42 ± 3.64 milliseconds, both P < 0.05) and father iPSC-CMs (FPD: 277.15 ± 5.20 milliseconds; cFPD: 92.66 ± 2.24 milliseconds, both P < 0.05) (Supplementary Table 4).

Arrhythmia susceptibility of family trio iPSC-CMs

We first investigated arrhythmia susceptibility at the single-cell level (Figure 5A). Whole-cell patch-clamp recordings showed a significantly higher EAD incidence in the proband iPSC-CMs at 0.5 Hz compared with the father iPSC-CMs [7/15 (46.7%) vs 0/15 (0%), P < 0.05], and also exhibited a trend towards a higher incidence compared with the mother iPSC-CMs, although the difference was not statistically significant [7/15 (46.7%) vs 2/15 (13.3%), P = 0.11]. Under spontaneous rhythm, EADs were more frequent in the proband iPSC-CMs than in the mother iPSC-CMs [7/15 (46.7%) vs 1/15 (6.7%), P < 0.05] and father iPSC-CMs [7/15 (46.7%) vs 0/15 (0%), P < 0.05], and ventricular tachycardia-like TAs were only observed in 20.0% of the proband iPSC-CMs.

Figure 5 Arrhythmogenic behavior of family trio induced pluripotent stem cell-derived cardiomyocytes. A: Representative early afterdepolarizations under 0.5-Hz pacing and spontaneous rhythms (a1). Ventricular tachycardia-like triggered arrhythmias under spontaneous rhythms (a2). Restoration of a relatively regular rhythm after instantaneous administration of 300 nmol/L nifedipine (a3); B: Representative early afterdepolarizations and triggered arrhythmias observed under spontaneous rhythms in multi-electrode array recordings (b1). Restoration of a relatively regular rhythm after instantaneous administration of 300 nmol/L nifedipine (b2); C: Representative and statistical analysis of action potential duration shortening by nifedipine under 0.5-Hz pacing in patch-clamp recordings (c1). Representative and statistical analysis of corrected field potential duration shortening by nifedipine under spontaneous rhythms in multi-electrode array recordings (c2). Data represent n = 15 and induced pluripotent stem cell-derived cardiomyocytes from 3 independent differentiations. EADs: Early afterdepolarizations; VT: Ventricular tachycardia; TAs: Triggered arrhythmias; APD: Action potential duration; cFPD: Corrected field potential duration.

We then validated arrhythmia susceptibility at the multicellular level (Figure 5B). Similarly, MEA recordings further confirmed that EADs and TAs were more frequent in the proband iPSC-CMs than in the mother iPSC-CMs [7/15 (46.7%) vs 1/15 (6.7%), P < 0.05] and father iPSC-CMs [7/15 (46.7%) vs 0/15 (0%), P < 0.05], with no significant difference between the mother and father iPSC-CMs (Supplementary Table 5).

Finally, rescue experiments with 300 nmol/L nifedipine significantly reduced the APD₉₀ in the proband iPSC-CMs from 746.15 ± 26.28 milliseconds to 615.63 ± 10.09 milliseconds (P < 0.05), with a 17.5% reduction. Consistently, the MEA recordings showed a decrease in cFPD from 286.72 ± 8.09 milliseconds to 222.72 ± 7.24 milliseconds (P < 0.05), corresponding to a 21.6% reduction. Nifedipine also effectively suppressed all EADs and ventricular tachycardia-like TAs in the proband iPSC-CMs (Figure 5C).

DISCUSSION

We employed individual-specific iPSC-CMs to model the intra-familial phenotypic variability associated with the KCNH₂ p.Y427H variant, which was functionally validated in this study as a loss-of-function mutation with potential pathogenicity. Despite carrying the same heterozygous mutation, the symptomatic proband displayed marked prolongation of APD and cFPD, enhanced L-type calcium current, amplified CaT, and a high incidence of EADs and TAs, both at the single-cell and multicellular levels, features that were less prominent in her asymptomatic mutation-positive mother and absent in her mutation-negative father. These findings underscore the power of iPSC-CMs to replicate patient-specific disease severity.

Although LQTS was once considered a “one gene, one disease” model, its genetic architecture is actually highly complex. Increasing evidence suggests that multiple genes or single nucleotide polymorphisms (SNPs) can collectively drive disease development[23,24]. In about 4%-8% of LQTS cases, the combined effects of two mutations within the same gene (such as compound heterozygotes) or mutations in two independent genes (digenic heterozygosity) result in more severe phenotypes[25], helping to explain the variability observed in LQTS phenotypes. For example, common nonsynonymous coding variants within the same gene as the primary mutation can act as modifying factors. Crotti et al[26] have reported a family with a KCNH2 p.A1116V mutation in an LQT2 patient. The proband had sudden cardiac arrest, while most family members were asymptomatic. The proband also carried the KCNH2 p.K897T SNP, which reduced the I_Kr current and worsened the LQT2 phenotype. Similarly, Nof et al[27] have described a family where the mother carried the heterozygous nonsense mutation KCNH2 p.P926AfsX14 with a prolonged QTc but no symptoms. The father, with a normal QTc duration, carried the KCNH2 p.K897T SNP. Two offspring inherited both mutations, leading to miscarriage and sudden infant death. In vitro experiments confirmed that the SNP worsened Kv11.1 dysfunction, causing LQTS-related arrhythmias and infant death. Admittedly, these studies enhance our understanding of phenotypic variation; however, they predominantly utilize heterologous cells, which lack CM characteristics, thus missing the complex interactions of repolarization reserve currents and ions in the full genetic context of real patients.

Since Yamanaka’s development of iPSC technology[13], it has been rapidly applied to the study of LQT2. For example, Itzhaki et al[15] demonstrated the potential of patient-specific iPSCs to model LQT2 caused by KCNH2 p.A614V. The iPSC-CMs showed a prolonged APD due to a reduced I_Kr current, a hallmark of LQTS, and exhibited arrhythmias such as EADs/TAs. The study also tested the effects of several antiarrhythmic drugs, demonstrating the potential of iPSC-CMs for drug testing and personalized medicine in inherited cardiac disorders. In addition, Brandão et al[28] explored how mutations in the KCNH2 gene lead to varying disease manifestations based on the mutation location. CRISPR-Cas9 was used to introduce mutations into iPSC-CMs. The research found that pore region mutations (KCNH2 p.A561T) resulted in prolonged repolarization and a higher risk of arrhythmias compared with tail region mutations (KCNH2 p.N996I), highlighting the potential of iPSC-CMs for assessing mutation severity. However, other genetic variants in individuals that may affect disease severity have been overlooked in studies focusing on single-gene mutations. Shah et al[29] found variable LQT2 (KCNH2 p.L552S) phenotypes within a family using patient-specific iPSC-CMs, showing that symptomatic cells had impaired calcium handling and increased arrhythmogenicity, with an elevated CaT half-width, CaT₉₀, and upstroke time at baseline. Similarly, Chai et al[30] reported that LQT2 patients with the KCNH2 p.R752W mutation showed similar I_Kr deficiency, along with the modifier gene REM2 p.G96A (SNP), which led to enhanced I_CaL, resulting in a more severe phenotype in symptomatic carriers. Consistent with previous studies demonstrating that iPSC-CMs can recapitulate key electrophysiological features of LQT2, including a reduced I_Kr, prolonged APD, and increased susceptibility to EADs and TAs. In our study, the key findings include significant differences in electrophysiological and calcium signaling characteristics among iPSC-CMs derived from the family trio. First, both at the single-cell level (patch clamp) and at the cellular layer level (MEA), the proband iPSC-CMs exhibited the most severe repolarization abnormalities, including APD prolongation and cFPD prolongation. Second, the proband and mother, who carry the KCNH2 p.Y427H pathogenic mutation, showed a reduced I_Kr current density, suggesting that the KCNH2 p.Y427H mutation results in a loss-of-function variant. Additionally, both at the single-cell level (patch clamp) and at the cellular layer level (MEA), the proband iPSC-CMs displayed a significantly higher incidence of EADs and TAs, indicating substantial electrophysiological instability. Unexpectedly, we found that the proband iPSC-CMs exhibited significantly enhanced I_CaL and greater CaT compared with those from the parents. Further, we speculate that dysregulated calcium handling acts as a key “second-hit” modifier in LQT2 pathogenesis. Specifically, exaggerated I_CaL and calcium overload in the proband iPSC-CMs contribute to repolarization instability and the formation of EADs and TAs. Therefore, we evaluated the effects of the L-type calcium channel blocker nifedipine, which effectively suppressed all EADs and TAs, further supporting I_CaL as a key driver of arrhythmogenic events and highlighting its potential as a therapeutic target.

Notably, unlike heterologous expression systems or transgenic animal models, human iPSCs retain the full genetic information of individuals, providing a significant advantage in studying phenotypic differentiation, especially in detecting abnormal repolarization reserve currents (I_CaL) and CaT[31]. In our study, we validated the “second-hit” hypothesis, where the primary pathogenic mechanism of reduced I_Kr, combined with the modifier factors of enhanced I_CaL or calcium dysregulation, leads to a more severe phenotype. However, whether this dysregulation is secondary to APD prolongation or is driven by primary calcium-handling modifier genes (SNPs) remains unclear. Given that the proband inherited approximately 50% of their genetic material from each parent, we hypothesize that the proband may harbor deleterious de novo calcium-handling SNPs or those inherited from her father. This hypothesis warrants further investigation and validation.

This study provides valuable insights into phenotypic variability in LQT2; however, several limitations should be acknowledged. First, the reliance on a single family trio, while enabling a well-controlled within-family comparison, limits the generalizability of the findings. Future studies including additional pedigrees or sporadic LQT2 cases will be important to determine whether calcium dysregulation represents a common modifier across diverse KCNH2 genotypes. Second, although iPSC-CMs faithfully recapitulated phenotypic heterogeneity between carriers, their relative immaturity may influence calcium handling properties. The incorporation of maturation strategies, such as three-dimensional organoid systems, engineered heart tissues, or comparison with more mature adult-like iPSC-CMs, may improve translational relevance. Finally, further investigation using genome-wide association studies or whole exome sequencing in larger cohorts may identify additional calcium-handling SNPs that modulate the variable expressivity of LQT2. Moreover, gene editing technologies, such as CRISPR/Cas9 or base editors, could offer valuable insights into the pathogenicity of these variants and help clarify underlying mechanisms.

CONCLUSION

In summary, our findings reveal that calcium dysregulation serves as a pivotal modifier of arrhythmic susceptibility in LQT2. Patient-specific iPSC-CMs faithfully captured the intra-familial phenotypic diversity and enabled mechanistic and pharmacological interrogation (Figure 6). This platform may provide a powerful tool for individualized risk prediction and therapeutic evaluation in LQT2.

Figure 6 Graphical abstract. PBMCs: Peripheral blood mononuclear cells; iPSCs: Induced pluripotent stem cells; iPSC-CMs: Induced pluripotent stem cell-derived cardiomyocytes; MEA: Multi-electrode array; APD: Action potential duration; cFPD: Corrected field potential duration; CaTD: Calcium transient duration; EADs: Early afterdepolarizations; VT: Ventricular tachycardia; TAs: Triggered arrhythmias; SICD: Subcutaneous implantable cardioverter defibrillator.

ACKNOWLEDGEMENTS

We gratefully acknowledge the support of the Cardiovascular Center, Beijing Tsinghua Changgung Hospital. We also thank our colleagues and technical staff for their contributions to the study.