Introduction of the thesis

(1)

Cover Page

The handle http://hdl.handle.net/1887/56253 holds various files of this Leiden University dissertation

Author: Kosmidis, Georgios

Title: Coming of age of human stem cell derived cardiomyocytes : towards functional maturation of human pluripotent stem cell derived cardiomyocytes and their use in understanding inherited arrhythmia syndromes

Date: 2017-10-11

(2)

Coming of age of human stem cell derived cardiomyocytes

Towards functional maturation of human pluripotent stem cell derived cardiomyocytes and their use in understanding inherited arrhythmia syndromes

(3)

Colophon

Coming of age of human stem cell derived cardiomyocytes. Towards functional maturation of human pluripotent stem cell derived cardiomyocytes and their use in understanding inherited arrhythmia syndromes.

PhD thesis

This thesis was prepared at the department of Anatomy & Embryology of the Leiden University Medical Center, Leiden, The Netherlands

All rights reserved. No part of this book may be reproduced or transmitted, in any form or by any means, without permission in writing of the author. The copyright of the articles that have been published has been transferred to the respective journals.

The research described in this thesis as well as its publication was supported by a grant of the Rembrandt Institute of Cardiovascular Science.

Financial support by the Dutch Heart Foundation for the publication of this thesis is gratefully acknowledged.

ISBN: 978-94-6295695-7

Cover: Immunostaining of a network of human pluripotent derived cardiomyo-

(4)

Coming of age of human stem cell derived cardiomyocytes

Towards functional maturation of human pluripotent stem cell derived cardiomyocytes and their use in understanding inherited arrhythmia syndromes

Proefschrift ter verkrijging van

de graad van Doctor aan de Universiteit Leiden, op gezag van Rector Magnificus prof.mr. C.J.J.M. Stolker,

volgens besluit van het College voor Promoties te verdedigen op 11 oktober 2017 klokke 16.15 uur

door Georgios Kosmidis geboren te Alexandroupolis

in 1982

(5)

Promotor

Prof.dr. C.L. Mummery Co-Promotoren

Dr. M. Bellin Dr. S. Casini

Promotiecommissie Prof.dr. M-J. Goumans Prof.dr. D.E. Atsma

Prof.dr. C.R. Bezzina (Academisch Medisch Centrum, NL) Prof.dr. R. Passier (Universiteit Twente, NL)

(6)

To my parents Heidi and Stavros and in memory of my grandfather Giorgos

Στους Γονείς μου Χάϊδω και Σταύρο και στην μνήμη του παππού μου Γιώργου

(7)

“You want to know the Secret, so did I, low in the dust I sought it and on high Sought it in awful flight from star to star, the sultan’s watchman of the starry sky”.

Omar Khayyám, Rubáiyát

(8)

Contents

Chapter 1. Introduction of the thesis

Chapter 2. Immaturity of human stem-cell derived cardiomyocytes in culture: fatal flaw or soluble problem?

Chapter 3. PGC-1α and reactive oxygen species regulate hESC-derived cardiomyocyte function

Chapter 4. Altered calcium handling and increased contraction force in human embryonic stem cell derived cardiomyocytes following short term dexamethasone exposure

Chapter 5. The combined presence of T3, IGF-1 and Dexamethasone in the cell culture of hPSC-CMs markedly improves the electrophysiological and force of contraction aspects of their functional state

Chapter 6. Readthrough promoting drugs gentamicin and PTC124 fail to rescue Nav1.5 function of hiPSC-derived cardiomyocytes carrying nonsense mutations in the sodium channel gene SCN5A

Chapter 7. Genetic correction of the W156X mutation in a patient- specific hiPSC line carrying compound heterozygous mutations in the SCN5A gene

Chapter 8. Establishing optimal culture conditions for

electrophysiological characterization of human pluripotent stem cell-derived cardiomyocytes

Chapter 9. General Discussion Summary/Samenvatting Acknowledgements Curriculum Vitae List of publications

(9)

(10)

Chapter 1 Introduction of the thesis

(11)

(12)

Inherited cardiac arrhythmias

Interest in the genetic causes of cardiac arrhythmic disorders has recently increased with the emergence of new technologies to investigate underlying mechanisms.

Heart disease as a cause for sudden death became prevalent in the western world during the 1960’s with its primary manifestation, coronary heart disease, resulting from coronary atherosclerosis [1]. Through primary and secondary prevention, the number of sudden deaths associated with heart disease have been declining, although mortality and morbidity from cardiovascular related diseases as a whole still accounts for 30% of global deaths [2] with estimates expecting this to rise in the coming decade. A large group of these patients has not had myocardial infarction or ischemia but still display the pathological features of cardiac arrhythmias. Many of these have turned out to have underlying genetic causes. In 1990, groundbreaking work by Seidman [3] led to the discovery that a mutation in the β-myosin heavy chain gene (MHY7) was linked to hypertrophic cardiomyopathy. A number of studies followed that shed light on the genetic causes of non-ischemic cardiomyopathies (hypertrophic, dilated and arrhythmogenic) and ion channelopathies (Long QT syndrome, Brugada syndrome, catecholaminergic polymorphic ventricular tachycardia and short QT syndrome). Table 1 shows the genes known to be associated with these disorders. Through combined efforts in genetics and clinical electrophysiology we now know many of the underlying mechanisms of inherited cardiac disorders and several treatments developed to alleviate the manifestations of atrial and ventricular fibrillation as well as tachy- cardias. However, these approaches, which include invasive procedures such as catheter ablation or defibrillator implantation, and treatments with beta-blockers or ion (Na⁺, Ca⁺²) channel blockers, only address the symptoms but not the causes of cardiac diseases. Arrhythmias arise because of erratic electrical behavior of small groups of cells in the heart tissue which emit aberrant signals. These can also have non-genetic origins (for instance from anti-cancer compounds) and are of growing concern in drug induced arrhythmias [4] as they cannot yet be prevented.

To devise “mechanism-modulating therapies” [5] model systems are required that provide ways to obtain new insight into mechanisms underlying specific types of arrhythmia and which can also serve as robust drug screening platforms to develop innovative therapies against cardiac diseases.

(13)

Table 1: Genes associated with inherited cardiac arrhythmias

Disease Genes

Brugada syndrome SCN5A, SCN1B, SCN2B, SCN3B, SCN10A,

GPD1L, CACNA1C, CACNB2, CACNA2D1, KCNE3, KCNE5, KCND3, KCNJ8, HCN4, MOG1, SLMAP, TRPM4, HEY2 Long-QT syndrome

(LQT1-LQT15)

KCNQ1, KCNH2, SCN5A, ANK2, KCNE1, KCNE2, KCNJ2, CACNA1C, CAV3, SCN4B, AKAP9, SNTA1, KCNJ5, CALM1, CALM2 Short-QT syndrome

(SQT1-SQT6)

KCNH2, KCNQ1, KCNJ2, CACNA1C, CACNB2, CACNA2D1

Catecholaminergic polymorphic ventricular tachycardia (CPVT1-2)

RyR2, CASQ2

Hypertrophic cardiomyopathy (HCM) MHY7, MYL2, MYL3, TNNT2, TNNI3, TTN, TPM1, MYBPC3, MYOZ2, PRKAG2, PLN ACTC1, ACTN2, CSRP3, CALR3, JPH2, NEXN, TCAP, VCL

Dilated cardiomyopathy (DCM) DMD, ABCC9, ACTC1, ACTN2, ANKRD1, BAG3, CRYAB, CSRP3, DES, DSG2, EYA4, LAMA4, LDB3, LMNA, MYBPC3, MYH6, MYH7, MYPN, PLN, PSEN1, PSEN2, RBM20, SCN5A, SGCD, TAZ, TCAP, TMPO, TNNC1, TNNI3, TNNT2, TPM1, TTN, VCL

Arrhythmogenic right ventricular cardiomyopathy (ARVC)

PKP2, DSC2, DSG2, DSP, JUP, RyR2, TGFB3, TMEM43

Human induced pluripotent stem cell derived cardiomyocytes as a model for cardiac diseases. Insights gained from the hiPSC-CM model.

Various in vitro and in vivo systems are currently used to model cardiac diseases.

Primary adult human cardiomyocytes isolated directly from the heart tissue of patients is one option but these can only be collected during invasive heart sur- gery. Furthermore, since (i) atrial cells from the auricle are most usually accessible to the surgeon and not other cardiac subtypes, (ii) human primary cardiomyocytes survive poorly in culture after isolation and (iii) attempts to immortalize primary human cardiomyocytes to establish cardiomyocyte cell lines have been unsuccessful to date, this approach is generally not considered feasible. Alterna- tively, cardiomyocytes from small mammals such as mouse or rat are more readily available but the phenotype of murine cardiomyocytes differs significantly from humans [6]. The mouse heart beats almost 10 times faster than the human heart.

(14)

such as CACNA1C, KCNQ1 and KCNH2 between the two species and variation in the accessory proteins of their corresponding ion currents. These differences result in distinct action potential (AP) profiles in patch clamp electrophysiology, mouse cardiomyocytes having an AP duration (APD) that can be 10 times shorter than in human cardiomyocytes [7]. Since many mutations exert their arrhythmic effects during this long repolarization phase of the action potential in humans, arrhythmia studies in mice often yield ambiguous results [8] that render relevance to human pathology questionable.

Human induced pluripotent stem cells (hiPSC) are increasingly considered as alternative disease models. After their initial generation by Shinya Yamanaka [9] through reprogramming of somatic cells using integrating retroviral vectors expressing the 4 transcription factors OCT3/4, SOX2, KLF4 and c-MYC, much progress has been made so that it is now possible to generate hiPSC lines using vectors that do not integrate in the host genome [10]. Using methods already developed in human embryonic stem cells, Zhang et al soon showed they could differentiate to cardiomyocytes [11]. Later improvements resulted in greater cardiomyocyte yield in serum-free, chemically defined culture conditions [12].

Whether cardiomyocyte differentiation is in monolayer (2-dimensional or 2D) culture or as aggregates (called embryoid bodies) (3-dimensional or 3D), the basic principles are the same: first hiPSC are induced to form mesoderm by bone morphogenetic protein (BMP) and activating Wnt signaling and then the mesoderm cells are patterned towards the cardiac lineage by inhibiting the Wnt signaling pathway (Fig.1). In most differentiation protocols cardiomyocytes start beating spontaneously at day 7-12 of differentiation. Using these procedures to derive cardiomyocytes from patient hiPSC (hiPSC-CMs), it has been possible to recapitulate the disease phenotype of numerous arrhythmic disorders in vitro (Fig.2). The first study describing this was published by Carvajal-Vergara et al [14]

in 2010. The study modelled the cardiomyopathy of LEOPARD syndrome. This was closely followed by a model of the channelopathy LQT1 syndrome [15].

(15)

Figure 1: Timeline of cardiac differentiation from hiPSC.

The distinct cell populations in the cardiac lineage are shown with their molecular markers at each stage of the differentiation. Growth factors and inhibitors required for each stage are also indicated (figure adapted from Jeziorowskaa et al [13]).

Figure 2:

A. Examples of typical AP profiles (from left to right): i) ventricular cardiomyocytes from the adult heart, ii) wildtype (WT) hPSC-CMs (RMP more depolarized, lower upstroke velocity), iii) hPSC- CMs from patients (APD may be shorter, upstroke velocity lower depending on the genotype), iv) hPSC-CMs showing early after depolarizations (EADs) or delayed after depolarizations (DADs) (sometimes in response to drugs). B. Ion currents and the corresponding ion channel genes in adult ventricular CMs and ventricular-like hiPSC-CMs. Each graph depicts the current during the course of the AP either in adult CMs or WT hiPSC-CMs (corresponding to AP traces in 2A). Some currents, such as IK1 and INa, are reduced in ventricular-like hPSC-CMs while others, such as If,, are aberrant

(16)

Numerous studies that followed have demonstrated the ability of hiPSC-CMs to serve as model systems for cardiac diseases such as LQT1 [16,17], LQT2 [18,19,20,21,22,23,24,25], LQT3 [23,26,27,28,29], Timothy syndrome (LQT8) [30], Catecholaminergic polymorphic ventricular tachycardia (CPVT) [31,32,33,34,35,36,37,38,39], Hypertrophic cardiomyopathy (HCM) [40,41,42,43,44,45], Dilated cardiomyopathy (DCM) [46,47,48,49,50,51,52,53], Arrhythmogenic right ventricular cardiomyopathy (ARVC) [54,55,56,57], Pompe disease [58,59], Friedreich’s ataxia [60,61] and many more. Table 2 provides an overview of the diseases modelled using hiPSC-CMs to date, with a short descrip- tion on the phenotype recapitulated in vitro, and approaches used to rescue the phenotype. Most phenotypic characterization of hiPSC-CMs has been carried out using measures of electrical behaviour, either with classical patch-clamp measurements of AP characteristics and ion currents or microelectrode arrays of external fieldpotential [62]. This is one of the primary approaches used for research carried out in this thesis.

Whilst first use of hiPSC-CMs as cardiac disease models was met with scepticism [63], their potential for providing insights into disease mechanisms and therapeutic approaches was soon revealed. Ma et al [17] presented evidence that a small novel molecule (ML277) that activates the IKs current restores function in LQT1 patient hiPSC-CMs. Itzhaki et al [18] also used novel compounds (pinacidil, ranolazine) to reverse the arrhythmic effects of KCNH2 mutations in LQT2 and demonstrated a role for the hiPSC-CMs model in dosage optimization of these compounds (since high-suboptimal doses would lead to pathological shortening of the QT interval).

Modelling the same disease, Matsa et al [19] tried new experimental potassium enhancers such as nicorandil and PD118057 while showing that the combined treatment with isoprenaline caused early after depolarizations, reversed by another compound, nadolol. In another study the same group [22] presented the possibility of a novel clinical approach using mutation specific siRNAs to target KCNH2 mutations. The study of Bellin et al [21] revealed that the pathogenesis of the N996I KCNH2 mutation is due to a trafficking defect of the protein from the cytoplasm towards the surface of the cell. Also using hiPSC-CMs Spencer et al [23] showed that KCNH2 mutations can have an impact on the cytosolic calcium handling of cardiomyocytes carrying these mutations. Mehta et al [24]

demonstrated that the A561V mutation in KCNH2 results in a trafficking defect of the HERG channel and proposed an alternative therapeutic strategy using the ALLN compound (N-[N-(N-acetyl-l-leucyl)-l-leucyl]-l-norleucine), impacting on the localization of the protein. In an impressive study modelling LQT3 with hiPSC-CMs, Terrenoire et al [27] managed to recapitulate the effects of clinical

(17)

diac diseases modelled with hiPSC-CMs. ardiac disease Gene/mutation or induced conditionDisease phenotype recapitulated Drug rescue (or other)/testReference (first author)

Year T ome T1KCNQ1-R190QAP prolongation, trafficking defect of channel, low IKS densityPropranololMoretti et al** 2010 T1KCNQ1-P631fs/33Prolonged FPD upon drug treatmentPropranololEgashira et al2012 T1KCNQ1-c.922- 1,032del (exon 7)Reduced KCNQ1 protein expression, AP prolongation, low IKS density ML277Ma et al2015 T2KCNH2-A614VAP prolongation, low IKr density, EADs, triggered arrythmiasPinacidil, nifedipine ranolazineItzhaki et al2011 T2 KCNH2-A561TProlonged FPD and APD, drug-induced EADsPropranolol, nicorandil, nadololMatsa et al2011 T2KCNH2-R176WAP prolongation, low IKr density, increased sensitivity to arrhythmogenic drugsErythromycin, sotalol, cisaprideLahti et al2012 T2KCNH2-N996IAP prolongation, low IKr density, trafficking defect of channelGenetic correction of mutationBellin et al***2013 T2 KCNH2-G1681AImpaired glycosylation and channel transport, EADsMutation-specific siRNAsMatsa et al2014 T2KCNH2-A422TAP and calcium transient prolongation,Nifedipine, E-4031, cisaprideSpencer et al2014 T2KCNH2-A561VLow IKr density, prolonged APD, reduced localization of glycosylated hERG, abnormal trafficking of hERG

ALLNMehta et al 2014

(18)

Cardiac disease modelled with hiPSC-CMs Gene/mutation or induced conditionDisease phenotype recapitulated Drug rescue (or other)/testReference (first author)

Year LQT2KCNH2-A561PLow IKr density, prolonged APD trafficking defect of hERG-Jouni et al2015 LQT3SCN5A-1795insDLow INa density, reduced AP upstroke velocity, longer APD-Davis et al2012 LQT3SCN5A-F1473C +polymorphism in KCNH2-K897T

Enhanced late INaMexiletineTerrenoire et al2013 LQT3SCN5A-V1763MAP prolongation, enhanced late INa, altered INa kinetics MexiletineMa et al2013 LQT3SCN5A-N406KAP and calcium transient prolongation,Nifedipine, E-4031, cisaprideSpencer et al2014 LQT3SCN5A-R1644HAP prolongation, altered INa kinetics, EADsMexiletine, ranolazine, bisoprolol, phenytoinMalan et al2016 LQT8-Timothy syndromeCACNA1C-G406RIrregular contraction and calcium transient, AP prolongationRoscovitineYazawa et al2011 Jervell and Lange- Nielsen syndromeKCNQ1-c.478- 2A>T and KCNQ1- c.1781G>A AP prolongation, increased AP amplitude, defective IKs, abolished KCNQ1 protein expression (c.478- 2A>T) and (c.1781G>A) trafficking defect

Propranolol, NS1643 Zhang et al2014 Brugada syndrome BrS type 1SCN5A-R1638X, SCN5A-W156XReduced AP upstroke velocities, low INa densityAtaluren (PTC124), gentamycinKosmidis et al2016

(19)

ardiac disease Gene/mutation or induced conditionDisease phenotype recapitulated Drug rescue (or other)/testReference (first author)

Year RyR2-F2483IDADs upon drug stimulus, longer calcium transient and higher amplitude-Fatima et al2011 RyR2-S406LElevated diastolic calcium, reduced SR calcium content, DADsDantroleneJung et al2012 RyR2-M4109RDADs, calcium transient irregularitiesFlecainide, thapsigarginItzhaki et al2012 RyR2-P2328SDADs, EADs, reduced SR calcium content, abnormal calcium transient-Kujala et al2012 RyR2-Q2311NDADs, irregular calcium transientsKN-93Di Pasquale et al2013 RyR2-F2483IAbnormal calcium signaling, lower calcium stores, higher CICR gain-Zhang et al 2013 RyR2-R420QSarcomere abnormalities, increased diastolic calcium upon isoproteranol-Novak et al2015 RyR2-various missense mutations Calcium transient abnormalitiesDantrolenePenttinen et al2015 CASQ2-D307H (homozygous)DADs, increase in calcium transient amplitude, enlarged SR cisternae and reduced number of caveolae

-Novak et al2012 CASQ2-D307H (homozygous)Sarcomere abnormalities, increased diastolic calcium upon isoproteranol-Novak et al2015

(20)

Year Cardiomyo -pathies LEOPARD syndromePTPN11-T468MLarger size CMs, better sarcomeric organization, localization of NFATC4 in the nucleus, increase in RAS/MAPK phosphorylation

-Carvajal- Vergara et al*

2010 HCMMYH7-R663HLarger size CMs, irregular calcium transient, contractile arrhythmia, elevated intracellular calcium Propranolol, verapamil, nifedipine, diltiazem, lidocaine, mexiletine, ranolazine.

Lan et al2013 HCMMYH7-R442GDisorganized sarcomeres, APD prolongation, elevated diastolic calcium transient, less SR calcium storage, lower decay of calcium transient

Metoprolol, verapamil, pinacidilHan et al2014 HCMMYBPC3-Q1004delLarger cell surface, myofibrillar disarrays at baseline and hypertrophic conditionsETA‐b, ETB‐b Tanaka et al2014 HCMMYBPC3-c.2373dupG Decrease in cMyBP-C protein, decreased traction stress-Birket et al2015 HCMBRAF-T599R3D cardiac tissue model with increased size, twitch force, increased contraction and relaxation rates -Cashman et al2016 HCMMYBPC3-Q1061X and TPM1-D175NDifferences in cell size, DADs, EADs, irregularities in calcium handling-Ojala et al2016 DCMTNNT2-R173WDecrease in force of contraction, abnormal distribution α-actinin, reduced beating rate, abnormal calcium transient Overexpression of SERCA2A metoprolol, norepinephrine

Sun et al2012

(21)

Year LMNA-R225XMicronucleation, increased apoptosis upon electrical stimulus U0126 and selumetinibSiu et al2012 arth ome)TAZ-c.517delG and TAZ-c.328T>CAbnormal cardiolipin processing, ratio of monolysocardiolipin-to-cardiolipin more than 0.3, decreased mitochondrial function and morphology, irregular sarcomeres, lower twitch stress of engineered cardiac tissues

Genetic correction, bromoenol lactone, linoleic acid (suppression of ROS), arginine plus cysteine

Wang et al2014 PLN-R14delDysregulation of calcium transient, abnormal distribution of PLN in cytoplasmGenetic correction and knock down of endogenous PLN

Karakikes et al2015 TTN-S14450fsX4Altered sarcomere assembly, dysregulation of the serum response signaling pathway, contractile dysfunction

AON-mediated exon skippingGramlich et al2015 DMD-deletion over exons 45–52dystrophin deficiency, higher diastolic calcium levels, mitochondrial damage, cell apoptosisPoloxamer 188Lin et al2015 TTN- W976R, TTN- A22352fs, TTN- P22582fs

Sarcomere abnormalities, impaired contraction force, inotropic and chronotropic deficits upon stress, truncated titin protein

-Hinson et al2015 dio- opathyDMD-deletion over exons 4–43 No of expression of dystrophin Introduction of human artificial chromosome Zatti et al2014 ophin- diomyo-pathy

DMD-deletion over exons 3 to 5 and 47 to 48 Reduction in endothelial nitric oxide synthase and neuronal nitric oxide synthase, increase in ROSNicorandilAfzal et al2016

(22)

Year ARVCPKP2-L614PLess PKP2 on the cell surface, larger cardiomyocytes, darker lipid droplets-Ma et al2013 ARVCPKP2-(c.2484C>T) (homozygous)Abnormal plakoglobin nuclear translocation, decreased β-catenin activity, pronounced lipogenesis and cell apoptosis, lower diastolic calcium

-Kim et al*****2013 ARVCPKP2-A324fs335XReduced densities of PKP2, plakoglobin, and connexin-43, prolonged FPD, distorted desmosomes, cell apoptosis

6-bromoindirubin-3’- oximeCaspi et al2013 ARVCPKP2-c.2484C>T PKP2-c.2013delCIncreased lipogenesis and apoptosis upon PPARγ activationGW9662, NAC, ascorbic acidWen et al2015 Pompe diseaseGAA-(not published)High glycogen levels, ultrastructural abnormalities, lower OCR and ECARRecombinant GAA, 3-methyladenine, L-carnitine

Huang et al2011 Pompe diseaseGAA-G2237A and GAA-exon 18 delGlycogen β-particles in lysosomes, hypoglycosylated lysosomal membrane proteins, deficiency in N-linked glycan synthesis

-Raval et al2015 Hypoplastic left heart syndrome(unknown genetic cause)Reduced calcium transient decay, ryanodine receptor dysfunction-Jiang et al2014 Hypoplastic left heart syndromeNOTCH1-P1964L- P1256L compound mutations

Poor myofibril organization-Theis et al2015 Friedreich’s ataxia FXN-GAA triplet codon expansion in the first intron

mitochondria with cristae abnormalities-Hick et al2013

(23)

Year eich’s ataxiaFXN-GAA triplet codon expansion in the first intron

Increased ROS, reduced calcium transient decay, mitochondrial stress under Fe+2 stress idebenone, deferiproneLee et al2016 y diseaseGLA-W162X and GLA-R220XGL-3 accumulation in the lysosomesSAR402671 Itier et al2014 iral myocarditisB3 strain of coxsackievirusAbnormal calcium transient, sarcomere abnormalitiesIFNβ1, ribavirin, fluoxetine, PDTCSharma et al****2014 iabetic diomyopathy Diabetic-like culture conditions (insulin signaling) Sarcomere abnormalities, myofilament disarray, altered calcium transients, lipid accumulation, peroxidation

28 compounds (from a chemical library test)Drawnel et al2014 anon diseaseLAMP2-c.129–130 insAT and LAMP2-c.64+1 G>A

Impaired autophagic flux, increased cell size and apoptosis, abnormal calcium transients, high oxidative stress

N-acetylcysteine Hashem et al2015 ction potential; APD: Action potential duration; FPD: Field potential duration; EAD: Early after depolarization; DAD: Delayed after depolarization; coplasmic reticulum; IKr: Rapid delayed rectifier potassium current; INa: sodium current; IKs: Slow delayed rectifier potassium current; CICR: calcium elease mechanism; CMs: Cardiomyocytes; ROS: Reactive oxygen species; AON: Antisense oligonucleotide; DMD: Duchenne muscular ophy; OCR; oxygen consumption rate; ECAR: extracellular acidification rate irst study using hiPSC-CMs as a cardiac disease model. irst study using hiPSC-CMs to model channelopathies. irst study using hiPSC-CMs in combination with genetic editing. irst study with hiPSC-CMs involving a viral pathogenic form of cardiomyopathy. irst study to induce an “adult-like” hiPSC-CM function, which allowed the phenotype of the disease to be revealed.

(24)

treatment (adjusting the pacing rate combined with mexiletine, a sodium channel inhibitor) in hiPSC-CMs from the same patient who showed positive responses.

In yet another study of the LQT3 syndrome [29], an alternative therapy with phenytoin was proposed to patients after their hiPSC-CMs indicated that this would be efficacious. This was noteworthy because the common treatment with mexiletine is not effective in all LQT3 patients. Evidence was provided by Zhang et al [64] that drug induced IKr activation (NS1643) could counteract the aberrant function of KCNQ1 in Jervell and Lange-Nielsen syndrome patient hiPSC- CMs. Using CPVT patient hiPSC-CMs, Jung et al [32] proposed Dantrolene as a compound that restores calcium spark characteristics and rescues arrhythmic events. In fact in another study by Penttinen et al [38] the same compound was

“benchmarked” in CPVT patients and hiPSC-CMs derived from them, showing anti-arrhythmic effects in 4 out of 6 patients tested. The findings of Di Pasquale et al [35] confirmed the potential of the KN-93 anti-arrhythmic drug, since their data on hiPSC-CMs validated those of a RyR2 knock-in CPVT mouse model.

Wang et al [48] compared three potential treatments using hiPSC-CMs while modeling Barth syndrome and found that only one (linoleic acid) was effective in correcting the metabolic phenotype of Barth hiPSC-CMs. In a DCM model, data by Gramlich et al [50] supported an antisense oligonucleotide (AON) therapeutic strategy (already in clinical trials for Duchene muscular dystrophy) in hiPSC-CMs carrying a titin frameshift mutation. Functional and RNA-Seq data of Hinson et al [52] also from hiPSC-CMs with titin mutations, showed that truncations in the A-band domain of titin cause DCM while truncations in the I-band are better tolerated. In the last phase of Duchenne muscular dystrophy (DMD) patients suffer from DCM. Lin et al [51] revealed that a mitochondrial-mediated signalling network (DIABLO, XIAP, CASP3) might be the cause of enhanced cardiomyocyte death in hiPSC-CMs from DMD patients and that the membrane sealant P188 suppressed the observed apoptosis. Zatti et al [53] tested the promis- ing gene delivery tool of human artificial chromosomes in hiPSC-CMs derived from DMD patients and confirmed that missing dystrophin isoforms could be expressed, thereby restoring dystrophin protein levels and subcellular localization. Using hiPSC-CMs as a model Afzal et al [65] presented a new compound, nicorandil, that has the potential to protect against stress induced cardiac injury in dystrophin deficient cardiomyopathy. While modelling ARVC with hiPSC-CMs, Caspi et al [56] corroborated the findings from a mouse ARVC model in which lipid accumulation was prevented by activating the canonical Wnt pathway with 6-bromoindirubin-3’-oxime. Using hiPSC-CMs, Huang et al [58] discovered a set of genes that could be used as molecular markers for evaluating the efficacy of drug treatment in Pompe disease. The genetic cause of Hypoplastic left heart

(25)

syndrome (HLHS) is not known. The potential of hiPSC-CMs as a model to investigate the genetic mechanism of the disease was capitalized by Theis et al [66], implicating mutations in the NOTCH1 gene as causative. By modelling Friedreich’s ataxia (FRDA) using hiPSC-CMs Lee et al [61] identified deferi- prone, the mitochondrial permeable iron chelator, as an effective compound to treat FRDA related cardiomyopathy. Itier et al [67] used the substrate reduction therapeutic approach by inhibiting glucosylceramide synthase in hiPSC-CMs derived from Fabry patients and observed a reduction of GL-3 in lysosomes (GL-3 accumulation is the hallmark of the disease). hiPSC-CMs have also been used to study cardiac diseases resulting from viral infections. In a novel study by Sharma et al [68], hiPSC-CMs were infected with the B3 strain of the coxsackievirus and various anti-viral approaches were tested. Focusing on one particular approach using IFNβ1, they deduced a mechanism of viral clearance pathways. Drawnel et al [69] applied a high-throughput screening approach using hiPSC-CMs monitoring α-actinin staining, nuclear area and BNP (brain natriuretic peptide) production in combination with a chemical library and identified 28 compounds that prevented the development of diabetic cardiomyopathy in vitro in a dose dependent manner. An additional therapeutic approach was suggested by Hashem et al [70] in which antioxidants such as NAC (N-acetylcysteine) could be used to ameliorate the effects of Danon disease in patient derived hiPSC-CMs.

The potential of hiPSC-CMs as a model system for cardiac disorders has thus been clearly demonstrated in the past few years as new diseases become the focus of hiPSC-CM research and new insights into mechanisms of diseases accrue.

However, some cardiac diseases cannot be accurately modelled with most current culture conditions since hiPSC-CMs are functionally immature, resembling early fetal cardiomyocytes in morphology and function. Much present research thus now focuses on identifying methods to promote hiPSC-CM “maturation” after differentiation. Some studies have already recapitulated a more “adult” cardiomyocyte pathology. For instance in order to reveal the adult disease phenotype of ARVC in hiPSC-CMs, Kim et al [55] added an adipogenic cocktail (insulin, dexamethasone and 3-isobutyl-1-methilxanthine-IBMX) to the differentiation medium, which yielded hiPSC-CMs with adult-like metabolism. Only then could they model the metabolic aspect of the ARVC disease. Wen et al [57] pursued a similar approach with the same disease but in a different model, Birket et al [43]

used a medium containing thyroid hormone (T3), IGF-1 and dexamethasone to reveal contractile defects in hiPSC-CMs carrying mutations in the MYBPC3 gene which result in HCM. Drawnel et al [69] also induced a more mature phenotype

(26)

The cardiac sodium channel and its role in cardiac disease. Using hiPSC-CMs to study Nav1.5 related arrhythmias.

One of the first genes in which a mutation was shown to cause a cardiac arrhythmia is the SCN5A [71], which encodes the main component of the cardiac sodium channel complex. The voltage-gated sodium channel (Nav1.5) is responsible for the fast inward sodium current INa, which is the element behind the rapid depolarization phase (phase 0, Fig.2A) of the AP in cardiomyocytes, generating the upstroke of the AP, which in turn initiates and propagates synchronous cardiac rhythm and contraction. The pore forming alpha subunit of the sodium channel encoded by the SCN5A gene consists of four homologous domains (DI–DIV) (Fig.3).

Each of these domains is composed of 6 segments connected by intracellular and extracellular loops. Segments S5 and S6 line the inner pore while segment S4 serves as a voltage sensor. Initial membrane depolarization by a leak of Na⁺ and Ca⁺² from neighbouring cells via gap junctions allows the pore to open (activation of the channel). The rapid inflow of Na⁺ into the cell causes sudden membrane depolarization (phase 0 of AP). After this, the fast inactivation gate (IFM motif) blocks the pore of the channel from the intracellular side, not allowing Na⁺ ions to pass through (inactivation of the channel). Eventually, membrane repolarization achieved by other ion channels causes closure of the pore while displacing the inactivation gate back to its original position (closed state of the channel). The channel is now recovered from inactivation and ready for the next AP cycle [72].

More than twenty proteins have been shown to interact with the α-subunit of the cardiac sodium channel [73] (Fig.3), each regulating different aspects such as the level of its expression, trafficking and anchoring of the channel to cellular membranes, post- translational protein modifications, altering its biophysical properties and controlling its activity [73,74]. There are however many protein interactions for which the precise function has not yet been determined.

The SCN5A is primarily expressed in cardiomyocytes and mutations in this gene can affect the structure, activity and expression of the cardiac sodium channel.

Various cardiac disorders have been linked to mutations in the SCN5A resulting in electrical disturbances in the heart or/and structural abnormalities (Fig.4).

After the initial discovery linking a particular mutation of the SCN5A to LQT3 [71] , ground breaking work from the same group identified the genetic causes of other types of Long QT syndromes in genes encoding

(27)

Figure 3: Structure of the α-subunit of the cardiac sodium channel and its interactions with other regulatory proteins.

Some proteins have been shown to associate with the channel via co-immunoprecipitation, however their precise interaction sites are not known. (Figure from Shy et al [73])

(28)

Figure 4: The structure of the pore of the cardiac sodium channel and the cardiac diseases caused by mutations in SCN5A. Arrhythmias arising solely from electrical disturbances are shown in red boxeswhile cardiomyopathies arising from structural anomaliesare shown in blue (figure adapted from Zaklyazminskaya et al [75]). potassium currents. To date more than 240 SCN5A mutations (stated in ClinVar as pathogenic or likely pathogenic) have been identified as the genetic cause underlying various cardiac arrhythmias. 2-5% of those mutations are linked with diseases occurring from structural changes in the myocardium while the rest are attributed to arrhythmias arising from electrical disorders in the heart (Fig.4). The high prevalence of such diseases in the population (~1:2000) combined with their high mortality rate after initial diagnosis, render genes such as the SCN5A as prime candidates for drug-related research in order to alleviate cardiac dysfunctions of genetic origin. As shown in Table 1, the hiPSC-CM model has already been used for such purposes with considerable success. Six studies to date (including our own) have generated patient-specific hiPSC lines carrying mutations in the SCN5A that were linked to LQT3 and Brugada syndromes. The disease phenotype was recapitulated in all studies and new molecular insight was gained in most of them [23,27,28,29,76].

Figure 4: The structure of the pore of the cardiac sodium channel and the cardiac diseases caused by mutations in SCN5a. Arrhythmias arising solely from electrical disturbances are shown in red boxes while cardiomyopathies arising from structural anomalies are shown in blue (figure adapted from Zaklyazminskaya et al [75]).

(29)

potassium currents. To date more than 240 SCN5A mutations (stated in ClinVar as pathogenic or likely pathogenic) have been identified as the genetic cause underlying various cardiac arrhythmias. 2-5% of those mutations are linked with diseases occurring from structural changes in the myocardium while the rest are attributed to arrhythmias arising from electrical disorders in the heart (Fig.4).

The high prevalence of such diseases in the population (~1:2000) combined with their high mortality rate after initial diagnosis, render genes such as the SCN5A as prime candidates for drug-related research in order to alleviate cardiac dysfunctions of genetic origin. As shown in Table 2, the hiPSC-CM model has already been used for such purposes with considerable success. Six studies to date (including our own) have generated patient-specific hiPSC lines carrying mutations in the SCN5A that were linked to LQT3 and Brugada syndromes. The disease phenotype was recapitulated in all studies and new molecular insight was gained in most of them [23,27,28,29,76].

The use of readthrough promoting compounds to suppress non-sense mutations and their potential in disease treatment.

A subset of mutations causing some genetic diseases are non-sense mutations, that result in the premature termination of protein translation. In eukaryotic organisms the normal process of terminating protein translation is highly conserved. It begins when one of the three stop codons (UAA, UGA, or UAG) enters the ribosomal A site upon which release factor I (eRF1) recognizes and binds to it. The eRF1 then interacts with eRF3 release factor leading to conformational changes that expose the GGQ motif of eRF1 to the peptidyl transferase center of the ribosome [77]. Subsequently, hydrolysis of the ester bond of the peptidyl-tRNA takes place and the polypeptide chain is released. A competition for codon binding is always occurring between aminoacyl-tRNAs and eRF1 even during the elongation phase of protein translation. Occasionally an amino acid will be incorporated in place of the stop, leading to the synthesis of an extended protein that ends at the next stop codon present in the same reading frame. This naturally occurring phenomenon is the reason why the translation termination process is not 100% efficient, having a low frequency (~0.1%) of “readthrough” taking place, a process also designated as stop-codon suppression [78]. Strategies to increase this chance of readthrough not at natural termination sites but at premature termination sites (PTCs) are of great clinical interest. Non-sense mutations are the cause of such sites, arising from single nucleotide changes within a gene that result in an in-frame PTC and account for ~11.2% of all the mutations associated with human inherited

(30)

taking place at PTCs is not completely understood, several compounds have been identified that induce readthrough at these sites while disregarding the normal stop codons. The first such compound (G418) belonging to an antibiotic fam- ily known as aminoglycosides, was first described in an 1996 study describing a potential treatment for cystic fibrosis patients that carried stop codon mutations in the CFTR [80]. Ever since then numerous studies have investigated the therapeutic potential of other compounds in PTC-mediated diseases such as Duchenne muscular dystrophy, Rett syndrome, Hemophilia, Spinal muscular atrophy, Hurler syndrome and many more [81]. The most prominent of these compounds include other aminoglycosides such as Gentamicin and Neomycin, aminoglycoside derivatives such as NB54 and TC007, another group of antibiotics named Macrolides (tylosin) and various other non-aminoglycoside compounds such as Negamycin, acetylamino benzoic acid, Clitocine, RTC13 and RTC14 [82]. A novel synthetic compound designated PTC124 (Ataluren) is now in the spotlight, being used in in vitro studies as well as clinical trials [83] due to its minimal toxic side-effects to human patients.

Cardiac arrhythmias may also be candidates for readthrough strategies. Olson et al demonstrated full-length protein production and functional restoration of the ultra-rapid delayed rectifier potassium current (IKur) when HEK293 cells overexpressing KCNA5 with non-sense mutation, were treated with Gentamicin for 48h [84]. In a similar approach, Yao et al observed increased IKr current densities and full-length protein production (by Western Blot) following a 24h treatment of Gentamicin and G418 in HEK293 cells overexpressing numerous stop codon mutation variants of the KCNH2 [85]. A stop codon mutation in the SCN5A gene itself was also overexpressed in HEK cells and again after Gentamicin and G418 treatment, Nav1.5 protein production was restored and the activation kinetics of the sodium current showed significant improvement [86]. The previous studies as well as the vast majority of all in vitro studies concerning readthrough promoting compounds have used overexpression systems that have a poor fidelity as a model system compared to the one of hiPSC-CMs. Although we were the first to examine the readthrough promoting capacity of Gentamicin and PTC124 in patient derived hiPSC-CMs, it is certain that more studies will soon follow using the hiPSC-CM platform.

Aim and outline of the thesis

Measurements of electrical phenotype are essential for exploiting hiPSC-CMs as models of many cardiac diseases. My focus in this thesis has been the analysis of

(31)

these electrical phenotypes in cardiac arrhythmia models using hiPSC-CMs but also in advancing and optimizing the value of the model itself.

The overall scope of this thesis is thus two-fold: first, to explore the possibilities of improving the hiPSC-CM disease model as such in terms of its functional output and second, to use patient-derived hiPSC-CMs to model sodium channel related cardiac arrhythmias arising from stop codon mutations in the SCN5A gene and investigate approaches to correct the adverse effect of these mutations. In Chapter 2, we reviewed the problems associated with the immature functional phenotype of human pluripotent stem cell derived cardiomyocytes (hPSC-CMs) and described the numerous experimental strategies proposed to address the issue. In Chapter 3 we demonstrated that controlling reactive oxygen species (ROS) levels is of key importance in leading hPSC-CMs towards a more mature state. Although knock- down of the master regulator gene PGC-1α decreases mitochondria numbers and their activity, the subsequent low levels of ROS have a beneficial impact in AP and calcium transient properties. Short term glucocorticoid treatment, often given to premature neonates, was investigated for its effects on the functional phenotype of hPSC-CMs in Chapter 4. We found that the force of contraction and calcium transient decay were most notably affected while the AP properties and L-type Ca⁺² current density are not. In Chapter 5, however, we observed that both the force of contraction and the AP phenotype were enhanced using T3, IGF-1 and dexamethasone in combination. Chapter 6 describes our efforts to recapitulate the disease phenotype of patients carrying SCN5A non-sense mutations in hiPSC- CMs. Most importantly, we tried to reverse their abnormal electrophysiological status by exploiting the readthrough promoting capabilities of gentamicin and PTC124. In Chapter 7 we used the genetic editing technique of CRISPR-Cas9 in combination with patient-derived hiPSC-CMs carrying compound mutations in the SCN5A. By doing so, we sought to “dissect” the detected disease phenotype and tried to evaluate the impact of each mutation on its severity. Chapter 8 provides a summary of our efforts in optimizing the electrophysiological output of hPSC-CMs by modifying culture conditions. Finally, in Chapter 9 we discuss the outcomes of all previous experimental chapters and present an outline of the future perspectives in the fields of hPSC-CM maturation and disease modelling.

(32)

References

[1] J.E. Dalen, J.S. Alpert, R.J. Goldberg, R.S. Weinstein, The epidemic of the 20(th) century: coronary heart disease, Am J Med 127 (2014) 807-812.

[2] D. Mozaffarian, E.J. Benjamin, A.S. Go, D.K. Arnett, M.J. Blaha, M. Cushman, S. de Ferranti, J.P.

Despres, H.J. Fullerton, V.J. Howard, M.D. Huffman, S.E. Judd, B.M. Kissela, D.T. Lackland, J.H.

Lichtman, L.D. Lisabeth, S. Liu, R.H. Mackey, D.B. Matchar, D.K. McGuire, E.R. Mohler, 3rd, C.S.

Moy, P. Muntner, M.E. Mussolino, K. Nasir, R.W. Neumar, G. Nichol, L. Palaniappan, D.K. Pandey, M.J. Reeves, C.J. Rodriguez, P.D. Sorlie, J. Stein, A. Towfighi, T.N. Turan, S.S. Virani, J.Z. Willey, D. Woo, R.W. Yeh, M.B. Turner, Heart disease and stroke statistics--2015 update: a report from the American Heart Association, Circulation 131 (2015) e29-322.

[3] A.A. Geisterfer-Lowrance, S. Kass, G. Tanigawa, H.P. Vosberg, W. McKenna, C.E. Seidman, J.G. Seid- man, A molecular basis for familial hypertrophic cardiomyopathy: a beta cardiac myosin heavy chain gene missense mutation, Cell 62 (1990) 999-1006.

[4] R. Madonna, C. Cadeddu, M. Deidda, P. Spallarossa, C. Zito, G. Mercuro, Modelling chemotherapy- induced cardiotoxicity by human pluripotent stem cells, Curr Drug Targets (2016).

[5] J.E. Saffitz, D. Corradi, The electrical heart: 25 years of discovery in cardiac electrophysiology, arrhythmias and sudden death, Cardiovasc Pathol 25 (2016) 149-157.

[6] J.M. Nerbonne, Studying cardiac arrhythmias in the mouse--a reasonable model for probing mechanisms?, Trends Cardiovasc Med 14 (2004) 83-93.

[7] G. Hasenfuss, Animal models of human cardiovascular disease, heart failure and hypertrophy, Cardio- vasc Res 39 (1998) 60-76.

[8] R.C. Ahrens-Nicklas, D.J. Christini, Anthropomorphizing the mouse cardiac action potential via a novel dynamic clamp method, Biophys J 97 (2009) 2684-2692.

[9] K. Takahashi, K. Tanabe, M. Ohnuki, M. Narita, T. Ichisaka, K. Tomoda, S. Yamanaka, Induction of pluripotent stem cells from adult human fibroblasts by defined factors, Cell 131 (2007) 861-872.

[10] K. Nishimura, M. Sano, M. Ohtaka, B. Furuta, Y. Umemura, Y. Nakajima, Y. Ikehara, T. Kobayashi, H. Segawa, S. Takayasu, H. Sato, K. Motomura, E. Uchida, T. Kanayasu-Toyoda, M. Asashima, H.

Nakauchi, T. Yamaguchi, M. Nakanishi, Development of defective and persistent Sendai virus vector:

a unique gene delivery/expression system ideal for cell reprogramming, J Biol Chem 286 (2011) 4760- 4771.

[11] J. Zhang, G.F. Wilson, A.G. Soerens, C.H. Koonce, J. Yu, S.P. Palecek, J.A. Thomson, T.J. Kamp, Functional cardiomyocytes derived from human induced pluripotent stem cells, Circ Res 104 (2009) e30-41.

[12] P.W. Burridge, A. Holmstrom, J.C. Wu, Chemically Defined Culture and Cardiomyocyte Differentia- tion of Human Pluripotent Stem Cells, Curr Protoc Hum Genet 87 (2015) 21.23.21-15.

[13] D. Jeziorowska, A. Korniat, J.E. Salem, K. Fish, J.S. Hulot, Generating patient-specific induced pluripotent stem cells-derived cardiomyocytes for the treatment of cardiac diseases, Expert Opin Biol Ther 15 (2015) 1399-1409.