Kim, Maengjo

Background: Long QT syndrome (LQTS) stems from cardiac action potential duration (APD) prolonging pathogenic variants in KCNQ1 (LQT1), KCNH2 (LQT2), or SCN5A (LQT3). Sodium/glucose co-transporter 2 inhibitor (SGLT2i), a class of drugs used to treat type II diabetes and all forms of heart failure, reduces cardiac late sodium channel current. Objective: To test the ability of three commercially available SGLT2i to reduce APD in patient-specific induced pluripotent stem cell-derived cardiomyocyte (iPSC-CM) models of LQT1, LQT2, and LQT3. Method: Patient-specific iPSC-CMs and a CRISPR/Cas9 gene-edited variant corrected, isogenic control (IC) iPSC-CMs were generated. The variant-corrected IC was used to determine the curative APD in order to establish the degree of APD attenuation. The SCN5A-P1332L (LQT3) and IC iPSC-CMs were tested with 10 μM of either canagliflozin (CANA), dapagliflozin (DAPA), or empagliflozin (EMPA). An ATXII (5μM)-induced APD prolongation iPSC-CM model, mimicking drug-induced LQT3, was also treated. The APD was measured 30 minutes, 1 hour, and 2 hours after treatment using local extracellular action potential recordings on multielectrode array. The efficacy of these SGLT2i were also tested in KCNQ1-V254M (LQT1) and KCNH2-G604S (LQT2) iPSC-CMs. Results: The APD90 was prolonged in SCN5A-P1332L iPSC-CMs compared to its IC (469 ± 4 ms vs 305 ± 2 ms, p<0.0001). While 10 μM of SGLT2i had no APD shortening effect on the KCNQ1-V254M or KCNH2-G604S iPSC-CMs, all three SGLT2i compounds shortened the APD90 of the SCN5A-P1332L iPSC-CMs after 1 hour of treatment from 463±5 ms to 432±6 ms (EMPA, 18.8% attenuation, p=0.0002), 459±8 ms to 430±5 ms (CANA, 17.5%, p=0.004), and 453±6 ms to 434±4 ms (DAPA, 11.4%, p=0.01). Additionally, IC iPSC-CMs treated with 5 μM of ATXII induced significant APD prolongation from 318±9 ms to 379±12 ms (p=0.0003) after 30 min incubation. However, co-incubation of ATXII and SGLT2i abolished the APD prolonging effect of ATXII. Conclusions: Therapeutically inhibiting SGLT2 effectively shortens the cardiomyocyte APD in human heart cell models of the LQT3, but not LQT1 or LQT2. This pre-clinical data compels further assessment of SGLT2i as a novel therapy for patients with congenital LQT3.

Background: Drug-induced QT prolongation (DI-QTP) is a clinical entity in which administration of a HERG/I Kr blocker such as dofetilide prolongs the cardiac action potential duration (APD) at the cellular level and the QT interval clinically, and increases the risk for a potentially lethal, QT-triggered ventricular arrhythmia. There may be a role for inhibition of I NaL to counter DI-QTP. Recently, we have shown that inhibition of serum and glucocorticoid regulated kinase-1 (SGK1) reduces the APD90 in induced pluripotent stem cell derived cardiomyocytes (iPSC-CM) from patients with LQT3 and ameliorates increases in I NaL . Objective: To test the efficacy of novel SGK1 inhibitors (SGK-I) in re-engineered cardiomyocyte models of dofetilide-induced APD prolongation. Methods: Normal iPSC-CMs were treated with dofetilide [5 nM] to produce a DI-QTP iPSC-CM model. The SGK1-I’s therapeutic efficacy for shortening the dofetilide-induced APD90 prolongation was compared to mexiletine. The APD90 values were recorded 4 hours after treatment using FluoVolt, a voltage-sensitive fluorescent dye. Results: The APD90 was significantly prolonged in normal iPSC-CMs treated with dofetilide (673 ± 8 ms with dofetilide vs 436 ± 4 ms with DMSO, p<0.0001). While 10 μM mexiletine shortened the average APD90 of dofetilide-treated normal iPSC-CMs from 673 ± 4 ms to 563 ± 8 ms (46% attenuation, p<0.0001), 30 nM of SGK-I shortened the APD90 from 673 ± 4 ms to 478 ± 10 ms (82% attenuation, p<0.0001). Conclusions: Therapeutically inhibiting SGK1 effectively shortens the cardiomyocyte APD in a human heart cell model of DI-QTP. The novel SGK1 inhibitor normalized the pathological APD prolongation almost fully (> 80%) in the iPSC-CM model treated with dofetilide. This pre-clinical data supports further development of this therapeutic strategy to counter and neutralize the threat of DI-QTP.

Background: Long QT syndrome (LQTS) often stems from pathogenic variants in KCNQ1 (LQT1), KCNH2 (LQT2), or SCN5A (LQT3) and is characterized cellularly by prolongation of the cardiac action potential duration (APD). We have shown that inhibition of serum and glucocorticoid regulated kinase-1 (SGK1) reduces I NaL and is proposed as a novel therapeutic strategy for LQT3. Objective: To test the efficacy of a potent and selective SGK1 inhibitor (SGK1-I) in human cardiomyocyte models of LQT1, LQT2, and LQT3. Methods: Induced pluripotent stem cell derived cardiomyocytes (iPSC-CMs) were generated from patients with either LQT1, LQT2, or LQT3. The mexiletine (MEX)-sensitive SCN5A-P1332L iPSC-CMs were tested initially. A CRISPR/Cas9 P1332L variant-corrected isogenic control (IC) was used as a control. The SGK1-I’s therapeutic efficacy was compared to MEX and to a commercially available SGK1 inhibitor (EMD638683). The novel SGK-I was then tested for APD shortening in SCN5A-R1623Q, KCNH2-G604S and KCNQ1-V254M iPSC-CMs. The APD90 values were recorded 4 hours after treatment using FluoVolt. Results: The APD90 was significantly prolonged in SCN5A-P1332L iPSC-CMs compared to its IC (646 ± 7 ms vs 482 ± 23 ms, p<0.0001). MEX shortened the average APD90 to 560 ± 7 ms (52% attenuation). Interestingly, SGK1-I significantly shortened the APD to 518 ± 5 ms (78% attenuation) but did not further shorten the APD in the IC. SGK1-I also shortened the APD90 of the SCN5A-R1623Q iPSC-CMs from 753 ± 8 ms to 475 ± 19 ms compared to 558 ± 19 ms with MEX. The SGK1-I shortened the APD90 in KCNH2-G604S (666 ± 10ms to 574 ± 18ms for SGK1-I versus 538 ± 15 ms after MEX). EMD had limited APD shortening effects in SCN5A and KCNH2 iPSC-CMs. Interestingly, while neither MEX nor EMD reduced the APD90 in the KCNQ1-V254M iPSC-CMs, the novel SGK1-I reduced the APD90 from 544 ± 10 ms to 475 ± 11ms (p=0.0004). Conclusions: Therapeutically inhibiting SGK1 effectively shortens the cardiomyocyte APD in human heart cell models of the 3 major LQTS genotypes. The novel SGK1-I normalized the pathological APD prolongation almost fully (> 70%) in the patient-derived SCN5A-P1332L iPSC-CM model. These pre-clinical data support further development of SGK1-I as a novel, first-in-class therapy for patients with congenital LQTS.

Introduction: Gain-of-function mutations in SCN5A- encoded cardiac sodium channel (Na V 1.5) cause type 3 long QT syndrome (LQT3). Patients with SCN5A variants localizing to the local anesthetic binding domain might not be sensitive to lidocaine or mexiletine treatment. Hypothesis: Phenytoin, a sodium channel blocker used to treat epilepsy, may be a novel LQT3-directed therapy for a variant involving a critical residue of the Nav1.5’s local anesthetic binding domain. Methods: SCN5A-F1760C was identified in a now deceased infant with severe LQT3 (QTc of 680-1200 ms) who was refractory to left cardiac sympathetic denervation and pharmacological treatments with either lidocaine or mexiletine. SCN5A-F1760C was engineered by site directed mutagenesis and expressed in TSA201 cells. Patient-specific pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) from the SCN5A-F1760C-positive infant and gene-edited/variant-corrected isogenic control (IC) iPSC-CMs were generated. Na V 1.5 currents were recorded in TSA201 cells transfected with SCN5A-WT/ F1760C and action potentials (APs) from iPSC-CMs were recorded before and after treatment with 5μM Phenytoin using the whole cell patch-clamp technique. Results: SCN5A-F1760C did not change the NaV1.5 current density. However, the V1/2 of inactivation was right-shifted significantly by +14.1 mV from -86.3±0.9 mV (WT) to -72.2±0.7 mV (F1760C, p<0.0001) resulting in a marked increase in window current. F1760C increased sodium late current 2-fold from 0.18±0.04% in WT to 0.49±0.07% in F1760C (p=0.0005). Baseline APD to 90% repolarization (APD90) was prolonged significantly in F1760C-derived iPSC-CMs (641±40 ms) compared to IC iPSC-CMs (412±22 ms, p<0.0001). Treatment with 5μM phenytoin normalized the APD90 in F1760C-derived iPSC-CMs to 423+34 ms (p<0.0001), values approximating the ‘gene/variant-edited’ cure. Conclusions: Na V 1.5 gain-of-function caused by a novel SCN5A-F1760C variant results in marked APD90 prolongation which were rescued by phenytoin indicating the antiepileptic drug, phenytoin, may be a mutation-specific, LQT3-directed therapy for those LQT3-causative variants that disrupt the local anesthetic/lidocaine-binding domain.

Mutations in the γ2-subunit of AMPK (encoded by PRKAG2 gene) cause glycogen storage cardiomyopathy. We generated a mouse model with cardiac specific deletion of γ2-AMPK (cKO) to define its function in the heart. The cKO mice were born normal and remained indistinguishable from control mice under baseline conditions but developed greater cardiac hypertrophy and dysfunction after transverse aortic constriction (TAC). The total AMPK activity was normal in cKO hearts and it increased after TAC to the same extent as that of the controls. However, the γ2-AMPK activity was abolished in the cKO and the overall AMPK activity was maintained by a compensatory increase in the γ1-AMPK activity. The lack of γ2-AMPK selectively reduced hexokinase II in the cytosol resulting in decreases in glycogen content and NADPH levels. Knock-down of γ2-AMPK subunit in neonatal rat ventricle cardiomyocytes resulted in increased oxidative stress and exacerbated response to phenylephrine and H 2 O 2 , both were rescued by HKII overexpression. Taken together, these data identify isoform-specific functions of γ2-AMPK in regulating glucose and glycogen metabolism that are critical for cardiac response to stress. The results also indicate that the specific activation of γ2-AMPK is responsible for the glycogen storage cardiomyopathy associated with PRKAG2 mutations.

A de novo missense variant in Rag GTPase protein C (RagCS75Y) was recently identified in a syndromic dilated cardiomyopathy (DCM) patient. However, its pathogenicity and the related therapeutic strategy remain unclear. We generated a zebrafish RragcS56Y (corresponding to human RagCS75Y) knock-in (KI) line via TALEN technology. The KI fish manifested cardiomyopathy-like phenotypes and poor survival. Overexpression of RagCS75Y via adenovirus infection also led to increased cell size and fetal gene reprogramming in neonatal rat ventricle cardiomyocytes (NRVCMs), indicating a conserved mechanism. Further characterization identified aberrant mammalian target of rapamycin complex 1 (mTORC1) and transcription factor EB (TFEB) signaling, as well as metabolic abnormalities including dysregulated autophagy. However, mTOR inhibition failed to ameliorate cardiac phenotypes in the RagCS75Y cardiomyopathy models, concomitant with a failure to promote TFEB nuclear translocation. This observation was at least partially explained by increased and mTOR-independent physical interaction between RagCS75Y and TFEB in the cytosol. Importantly, TFEB overexpression resulted in more nuclear TFEB and rescued cardiomyopathy phenotypes. These findings suggest that S75Y is a pathogenic gain-of-function mutation in RagC that leads to cardiomyopathy. A primary pathological step of RagCS75Y cardiomyopathy is defective mTOR–TFEB signaling, which can be corrected by TFEB overexpression, but not mTOR inhibition.

Background: Prior epidemiological studies demonstrated that the p.D85N-Potassium voltage-gated channel subfamily E member 1 (KCNE1) common variant reduces repolarization reserve and predisposes to drug-induced QT prolongation/torsades de pointes. We sought to develop a cellular model for drug-induced long QT syndrome using a patient-specific induced pluripotent stem cell-derived cardiomyocyte (iPSC-CM). Methods: p.D85N-KCNE1 iPSCs were generated from a 23-year-old female with an exaggerated heart rate-corrected QT interval response to metoclopramide (ΔQTc of 160 ms). Clustered regularly interspaced short palindromic repeats-associated 9 technology was used to generate gene-corrected isogenic iPSCs. Field potential duration and action potential duration (APD) were measured from iPSC-CMs. Results: At baseline, p.D85N-KCNE1 iPSC-CMs displayed significantly longer field potential duration (281±15 ms, n=13 versus 223±8.6 ms, n=14, P <0.01) and action potential duration at 90% repolarization (APD90; 579±22 ms, n=24 versus 465±33 ms, n=26, P <0.01) than isogenic-control iPSC-CMs. Dofetilide at a concentration of 2 nM increased significantly field potential duration (379±20 ms, n=13, P <0.01) and APD90 (666±11 ms, n=46, P <0.01) in p.D85N-KCNE1 iPSC-CMs but not in isogenic-control. The effect of dofetilide on APD90 (616±54 ms, n=7 versus 526±54 ms, n=10, P <0.05) was confirmed by Patch-clamp. Interestingly, treatment of p.D85N-KCNE1 iPSC-CMs with estrogen at a concentration of 1 nM exaggerated further dofetilide-induced APD90 prolongation (696±9 ms, n=81, P <0.01) and caused more early afterdepolarizations (11.7%) compared with isogenic control (APD90: 618±8 ms, n=115 and early afterdepolarizations: 2.6%, P <0.05). Conclusions: This iPSC-CM study provides further evidence that the p.D85N-KCNE1 common variant in combination with environmental factors such as QT prolonging drugs and female sex is proarrhythmic.

RXRA is an endothelial- and early stage–predominant therapeutic target for treating anthracycline-induced cardiotoxicity.

Rationale: We recently reported the association of S75Y, a point mutation in the Rag GTPase protein C ( RRAGC ) gene, with pediatric DCM. Because RRAGC directly binds and regulates mTOR and the S75Y mutation results in abnormally activated mTOR, these in vitro data prompts that the mTOR pathway could be harnessed for therapeutic benefits. Methods: We aimed to integrate in vitro neonatal rat ventricle cardiomyocytes (NRVCM) and in vivo zebrafish system to elucidate underling mechanism of RRAGC -S75Y-based cardiomyopathy, and to seek targeted therapeutic strategies. Results: Overexpression of RRAGC with S75Y mutation, but not RRAGC WT, significantly increased cell size of NRVCM, suggesting the establishment of a cellular model of RRAGC - S75Y-based cardiomyopathy. At the molecular level, overexpression of RRAGC S75Y mutant increased activity of mammalian target of rapamycin (mTOR) and decreased activity of transcription factor EB (tfeb), a direct downstream target of mTOR. Intriguingly, treatment with rapamycin, a mTOR inhibitor, did not affect cellular hypertrophy, whereas adenovirus-mediated overexpression of TFEB was able to rescue cellular hypertrophy induced by RRAGC S75Y mutation. Consistently, treatment with trehalose, an mTOR-independent tfeb activator, rescued cellular hypertrophy induced by RRAGC -S75Y. To confirm the in vitro results, we generated a knock-in (KI) zebrafish line harboring the S75Y corresponding point mutation in rragc via the TALEN (Transcription Activator-Like Effector Nucleases) technology. The cardiac function of rragc KI fish was determined by high frequency echocardiography, and the KI fish showed a significant decrease in fractional shortening compared with control littermates. Conclusion: Both cellular and in vivo models of RRAGC -S75Y-based cardiomyopathy have been generated. Studies in the cellular model suggested that tfeb, but not mTOR, is a therapeutic target for this particular type of cardiomyopathy. We will report our progress in validating this conclusion in our in vivo model, and will discuss potential mechanisms.

Titin-truncating variants (TTNtvs) are the major cause of dilated cardiomyopathy (DCM); however, allelic heterogeneity (TTNtvs in different exons) results in variable phenotypes, and remains a major hurdle for disease diagnosis and therapy. Here, we generated a panel of ttn mutants in zebrafish. Four single deletion mutants in ttn.2 or ttn.1 resulted in four phenotypes and three double ttn.2/ttn.1 mutants exhibited more severe phenotypes in somites. Protein analysis identified ttnxu071 as a near-null mutant and the other six mutants as hypomorphic alleles. Studies of ttnxu071 uncovered a function of titin in guiding the assembly of nascent myofibrils from premyofibrils. By contrast, sarcomeres were assembled in the hypomorphic ttn mutants but either became susceptible to biomechanical stresses such as contraction or degenerated during development. Further genetic studies indicated that the exon usage hypothesis, but not the toxic peptide or the Cronos hypothesis, could account for these exon-dependent effects. In conclusion, we modeled TTNtv allelic heterogeneity during development and paved the way for future studies to decipher allelic heterogeneity in adult DCM.