Establishing a new human hypertrophic cardiomyopathy-specific model using human embryonic stem cells

Journal Pre-proof Establishing a new human hypertrophic cardiomyopathy-specific model using human embryonic stem cells Huanhuan Cai, Bin Li, Aobing Bai, Jie Huang, Yongkun Zhan, Ning Sun, Qianqian Liang, Chen Xu PII:

S0014-4827(19)30619-6

DOI:

https://doi.org/10.1016/j.yexcr.2019.111736

Reference:

YEXCR 111736

To appear in:

Experimental Cell Research

Received Date: 20 May 2019 Revised Date:

21 August 2019

Accepted Date: 16 November 2019

Please cite this article as: H. Cai, B. Li, A. Bai, J. Huang, Y. Zhan, N. Sun, Q. Liang, C. Xu, Establishing a new human hypertrophic cardiomyopathy-specific model using human embryonic stem cells, Experimental Cell Research (2019), doi: https://doi.org/10.1016/j.yexcr.2019.111736. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc.

1

Establishing A New Human Hypertrophic Cardiomyopathy-specific

2

Model using Human Embryonic Stem Cells

3

Huanhuan Cai1#, Bin Li1#, Aobing Bai1#, Jie Huang1, Yongkun Zhan2, Ning Sun1,

4

Qianqian Liang1*, Chen Xu1*

5

6

Author Affiliations

7

1.

Sciences, Fudan University, Shanghai 200032, China

8 9

Department of Physiology and Pathophysiology, School of Basic Medical

2.

Children's Hospital, Fudan University, Shanghai 201102, China

10

*Correspondence: Qianqian Liang ([email protected]) and Chen Xu

11

([email protected])

12

#Equal contributors

1

1

Abstract

2

Symptom of ventricular hypertrophy caused by cardiac troponin T (TNNT2)

3

mutations is mild, while patients often showed high incidence of sudden cardiac death.

4

The 92th arginine to glutamine mutation (R92Q) of cTnT was one of the mutant

5

hotspots in hypertrophic cardiomyopathy (HCM). However, there are no such human

6

disease models yet. To solve this problem, we generated TNNT2 R92Q mutant hESC

7

cell lines (heterozygote or homozygote) using TALEN mediated homologous

8

recombination in this study. After directed cardiac differentiation, we found a relative

9

larger

cell

size

in

both

heterozygous

and

homozygous

TNNT2

R92Q

10

hESC-cardiomyocytes. Expression of atrial natriuretic peptide (ANP), brain

11

natriuretic peptide (BNP), and sarcoplasmic reticulum Ca2+-ATPase2a (SERCA2a)

12

were downregulated, while myocyte specific enhancer factor 2c (MEF2c) and the

13

ratio of beta myosin to alpha myosin heavy chain (MYH7/MYH6) were increased in

14

heterozygous

15

cardiomyocytes exhibited efficient responses to heart-related pharmaceutical agents .

16

We also found TNNT2 R92Q heterozygous mutant cardiomyocytes showed increased

17

calcium sensitivity and contractility. Further, engineered heart tissues (EHTs) prepared

18

by combining rat decellularized heart extracellular matrices with heterozygous R92Q

19

mutant cardiomyocytes showed similar drug responses as to HCM patients and

20

increased sensitivity to caspofungin-induced cardiotoxicity. Using RNA-sequencing

21

of TNNT2 R92Q heterozygous mutant cardiomyocytes, we found dysregulation of

22

calcium might participated in the early development of hypertrophy. Our

TNNT2

R92Q

hESC-cardiomyocytes.

2

TNNT2

R92Q

mutant

1

hESC-derived TNNT2 R92Q mutant cardiomyocytes and EHTs are good in vitro

2

human disease models for future disease studies and drug screening.

3

Keywords

4

TALEN, human embryonic stem cells, hypertrophic cardiomyopathy, cardiac troponin

5

T, R92Q mutation

6

1. Introduction

7

Hypertrophic cardiomyopathy (HCM) is often caused by genetic mutations in

8

contractile sarcomeric proteins [1]. Phenotypes of HCM include cardiac hypertrophy,

9

narrowed ventricular chamber, progressive heart failure and arrhythmia [2]. At least

10

20 genes have been identified related to HCM. Among these genes, myosin accounted

11

for a large proportion, while cardiac troponin T (TNNT2) takes up about 5% of the

12

proportion [3, 4]. HCM caused by TNNT2 mutations often showed high incidence of

13

sudden cardiac death even if there is no obvious phenotype [5]. The specific

14

pathogenic mechanism why a single gene mutation cause complex disease-related

15

phenotypes is still unclear.

16

As there are few reliable disease models of human HCM, the research progress is

17

relatively slow. Physiological characteristics and molecular mechanisms of HCM

18

transgenic animals (mice, rats and pigs) were different from human beings [6], and

19

can not mimic the related phenotypes of human diseases completely. In addition, it is

20

very difficult to obtain HCM heart tissues from patients. To facilitate the study of 3

1

pathogenic mechanisms of HCM, it is urgent to develop new human disease models.

2

Recently, development of human pluripotent stem cells (hPSCs) and genomic

3

editing technologies facilitated construction of hPSCs carrying specific HCM

4

mutations [7, 8]. Through TALEN and homogeneous recombination (HR), targeted

5

point mutations could be obtained quickly and efficiently. After cardiac differentiation,

6

we can obtain a large number of cardiomyocytes carrying the relative mutations.

7

These mutant cardiomyocytes can add up to the inadequacy of animal models and

8

provide a good platform for the study of pathogenic mechanism and high throughput

9

drug screening.

10

A missense mutation of cardiac troponin T (cTnT) resulted in an amino acid

11

exchange of glutamine to arginine at residue 92 (R92Q), which is known to cause

12

HCM in humans. As one of the most clinically malignant HCM mutations, individuals

13

with this mutation often showed high incidence of early sudden cardiac death in the

14

absence of overt heart hypertrophy [9]. In this study, we constructed TNNT2 R92Q

15

mutant human embryonic stem cells (hESCs) using TALEN mediated homologous

16

recombination. After cardiac differentiation, the mutant cardiomyocytes showed

17

similar phenotypes as HCM. Combined with decellularized rat heart matrices, we can

18

also construct human HCM heart tissues. CTnT R92Q mutant cardiomyocytes and the

19

engineered mutant heart tissues could serve as new disease models for the study of

20

HCM pathologic mechanisms and individualized treatment of HCM. They can also be

21

used in high throughput drug screening, further accelerate the development of new 4

1

drugs for HCM.

2

2. Materials and methods

3

2.1 TALEN efficiency and donor vector construction

4

PCR amplified genomic sequences flanking the TNNT2 R92 locus from human 293T

5

cells were transfected with the TALEN R92 L1/R1 plasmid and AAVS1 plasmids

6

were served as the positive control. According to the sequencing results, the height of

7

the scrambled peak indicated cutting efficiencies was ~25% at the targeted sites and

8

this pair of plasmids can be applied in subsequent gene editing. For the R92Q

9

mutation donor vector, the length of the 5' and the 3' homology arm is almost 1000 bp

10

each and point mutation in 5' homology arm was fused by overlap PCR. Puromycin

11

cassette and 3' homology arm were amplified by high fidelity PCR and verified by

12

sequencing. 5' homology arm, puromycin cassette, 3' homology arm and Hind

13

digested pUC19 linear plasmid were fused by Gibson Assembling according to

14

manufacturer’s instructions (NEB).

15

2.2 Culture of human ESCs

16

The human ESC line H7 and H9 used in this study was obtained from WiCell

17

Research Institute under specific Material Transfer Agreement. All human stem cells

18

researches followed the ISSCR Guidelines for the Conduct of Human Embryonic

19

Stem Cell Research. Human ESCs were maintained on 6-well or 12-well plate coated

20

with growth factor reduced Matrigel (Corning) using mTeSR1 medium (Stem cell 5

1

Technology). Cultures should be 70 - 90% confluent at the time of passage and were

2

disassociated into small aggregates or single cells with accutase (Sigma).

3

2.3 Construction of the TNNT2 R92Q mutant human ESC lines

4

We next transfected 50000 single H7 or H9 hESCs with the TALEN pair (L1+R1) and

5

homologous recombination donor vector using electroporation. 0.2 mg/ml puromycin

6

was used to screening for positive clones. Then the PGK-puromycin cassette was

7

cutted by the Cre recombinase after the second round of electric transformation. As

8

there were no homozygotes, we repeated the first and second rounds of

9

electroporation to get homozygotes. To prevent the disruption of TNNT2 gene, the

10

PGK-PURO cassette was inserted in the adjacent intron downstream of the R92Q

11

mutation and can be digested by Cre recombinase.

12

2.4 Teratoma formation assay

13

1×106 wild type or mutant hESCs were suspended in 20 µl Matrigel and injected into

14

the oxters of immunodeficiency mice (NOD/SCID) respectively (8 weeks old). Six

15

weeks later, teratomas were dissected and fixed with 4% paraformaldehyde and

16

hematoxylin-eosin staining (HE staining) were used to detected whether there were

17

ectoderm, mesoderm and endoderm layers.

18

2.5 Cardiac differentiation

19

The wild type and mutant hESCs were initially cultured in mTeSR1 medium on

20

Matrigel-coated plates until they were ~80% confluent, medium was changed to 6

1

RPMI/B-27 without insulin, consisting of RPMI 1640 (Corning) and B-27 minus

2

insulin (Life technologies). On day 0-2, medium was supplemented with 8 µM

3

CHIR-99021 (Selleck) in RPMI/B-27 without insulin. The medium was changed on

4

day 3 to RPMI/B27 without insulin for 24 h, followed by RPMI/B-27 without insulin

5

supplemented with 5 µM Wnt inhibitor IWR-1(Sigma-Aldrich) for 2 days, then

6

RPMI/B-27 without insulin for another 2 days. Finally and every 2 day thereafter,

7

medium was changed to RPMI/B-27 medium. Cultures were maintained in a 37 , 5%

8

CO2 environment. Contracting cells should be observed from day 7 post

9

differentiation.

10

2.6 Quantitative real-time PCR

11

All qPCR primers were designed according to the targeted gene sequences of homo

12

sapiens. Trizol reagent (Invitrogen) were used to extract total RNA and the qualities of

13

total RNA were measured by OD260/OD280 ratios. The cDNA synthesis was through

14

reverse transcription reaction according to manufacturer’s instructions (TOYOBO).

15

The PCR should be performed in a reaction system containing 10 µL of 2×SYBR

16

Green (Tiangen), 1 µL of RT reaction mix, 0.5 µL of each primer (10 µmol/L) and 8.5

17

µL ddH2O. Reactions were performed on a CFX96™ Real-Time System instrument

18

(BIO-RAD, USA). The primer sequences were listed in Additional file 4 and all data

19

was analyzed using the 2-∆∆Ct method.

20

2.7 Immunofluorescent staining

21

Cultures were washed with cold 1 × PBS three times (5 min/time), and treated with 7

1

4% paraformaldehyde for 15 min. They were again washed with cold 1 × PBS three

2

times (5 min/time) and treated with 5% Triton X-100 in Tris buffer solution (TBS, 50

3

mmol/L Tris/HCl pH 7.4, 150 mmol/L NaCl) for 20 min at room temperature (RT).

4

After treated with normal goat serum for 30 min at RT, the primary antibodies (1:200)

5

were added and incubated for 2 h at RT or 4°C overnight. Fluorescent secondary

6

antibodies (1:300) were added and the cells were incubated for 1 hour at RT, washed

7

three times for 5 min each with PBST (1% Tween in PBS buffer). The images were

8

analyzed using ImageJ software.

9

2.8 Generation of engineering heart tissues (EHTs)

10

Engineering heart tissue was carried out referring to the protocol described by Qingjie

11

Wang et al [16]. After 6 hours of decellularized perfusion, the decellularized heart

12

ECM was cut into pieces of 2 mm×2 mm size using a surgical scissor in a sterile

13

biosafety cabinet. Individual ECM pieces were put into wells of 96-well plates with

14

the endocardial side facing up and hESCs-derived cardiomyocytes were seeded onto

15

the ECM sheets at 104 cells/mm2. The mixtures were cultured in DMEM

16

supplemented with 10% FBS and the medium changed every 2 days.

17

2.9 The Microelectrode Array (MEA)

18

The electrophysiological properties of cardiomyocytes or engineered cardiac patches

19

were examined using the microelectrode array (MEA) data acquisition system

20

MEA-2100 (Multi Channel Systems). Contracting cardiomyocytes or engineered

21

human cardiac patches were plated on Matrigel coated MEA probes of appropriative 8

1

6-well plate. Local activation time at each of the 60 electrodes was determined and

2

used for the generation of color-coded activation maps as reported [19]. Data analysis

3

was performed using the SPIKE 2 software (CED, UK).

4

2.10 Ca2+ transient

5

Dissociated hESCs-derived cardiomyocytes were seeded in Matrigel-coated 35mm

6

dishes with glass bottom (Nalge Nunc International) and were loaded with 5 mM

7

Fluo-4AM for 15 min at 37°C. Cells were then washed three times with Tyrode’s

8

solution. Ca2+ imaging was conducted with a confocal microscope (Carl Zeiss, LSM

9

510 Meta) with a 63× lens (numerical aperture = 1.4) using Zen software.

10

Spontaneous Ca2+ transients were acquired at room temperature with line scan mode

11

at a sampling rate of 1.92 ms/line.

12

2.11 Relative contractility measurement

13

Representative images showed the relative contractility measured by Nikon edge

14

detection system. Dissociated hESCs-derived cardiomyocytes were seeded in

15

Matrigel-coated 35 mm dish with glass bottom (Nalge Nunc International). After two

16

day’s culture, single cardiomyocyte could beat spontaneously and be used for edge

17

detection. Voltage of Y axes indirectly reflected the relative contractility of beating

18

cardiomyocytes and the number of peaks in 30 seconds indicated the relative

19

contracting rate indirectly.

20

2.12 RNA-Sequencing and analysis 9

1

Total RNA was extracted from heterozygous R92Q cardiomyocytes using the TRIzol

2

Reagent according to the according to manufacturer’s instruction (Life Technologies)

3

and cDNA libraries were prepared using the NGS Multiplex Oligos for Illumina

4

(xCellBioTech Co., Ltd.). Totally, mRNA separated from the total RNA were

5

fragmented and reversed to cDNA, the products were then end-repaired and barcoded

6

with multiplex adapters. Before quantified by Qubit (Invitrogen), AmpureXP beads

7

were used to purify PCR amplified libraries and then run on an Illumina Hiseq

8

(Illumina). OmicsBean (www.omicsbean.cn) was used to screen the differentially

9

expressed genes. Significantly affected genes were considered those with a fold

10

change > 2 and a FDR (false discovery rate) < 0.05. Enrichment clustering of

11

up-regulated genes was performed with Metascape (http://metascape.org/) using

12

custom analysis.

13

2.13 Statistical analysis

14

All experimental data was expressed as means ± SD. Two-tailed Student's t-test was

15

used to compare differences between two groups. Statistical differences among more

16

than two groups were analyzed with one-way ANOVA tests. Significant differences

17

were determined when P value was less than 0.05. *P<0.05, **P<0.01 and

18

***P<0.001.

19

3. Results

20

3.1 The construction of TNNT2 R92Q mutant hESCs

10

1

To investigate the pathogenic mechanism of TNNT2 R92Q point mutations in humans,

2

TALEN and homogeneous recombination technologies were used. We designed six

3

TALEN pairs (three left TALENs × two right TALENs) targeting downstream of the

4

TAG stop codon on exon 10 of the TNNT2 gene. The L1+ R1 TALEN pair showed a

5

~25% cutting efficiency confirmed by DNA sequencing (Additional file 1), and thus

6

was used in the subsequent experiments. To eliminate the influences of genetic

7

background to the mutation induced phenotype, we used two human embryonic stem

8

cell lines (H7 and H9 hESCs) in this study. Single hESCs were electrotransfered with

9

the TALEN pair (L1+R1) and homologous recombination donor in the appropriate

10

proportion (3:1). After five days of puromycin screening, positive single clones with

11

successful homologous recombination were selected. We next transfected these

12

positive clones with plasmids expressing the Cre recombinase and removed the

13

PGK-puromycin selection cassette (Fig. 1A). After sequencing verification,

14

heterozygotes were preferentially obtained and then we transfected the heterozygotes

15

with correspondent vectors to get the homozygotes. Final DNA sequencing of the

16

genomic TNNT2 locus confirmed the correct insertion of targeted R92Q mutation

17

(Fig. 1B and Additional file 5 ).

18

To examine whether our L1+R1 TALEN pair had off-target effects in other locus,

19

we

amplified

the

top

10

20

(http://www.jstacs.de/index.php/TALENoffer) by genomic PCR and performed

21

Sanger’s sequencing. No off-target editing events were found. (Additional file 2 and 11

potential

off-target

sites

1

3 ).

2

3.2 Examination of pluripotency of hESCs carrying TNNT2 R92Q mutation after

3

genomic manipulation

4

The morphology and pluripotency of embryonic stem cells were especially

5

important in their cardiac differentiation process. To investigate whether insertion of

6

the R92Q mutation affected the maintenance and differentiation of hESCs, we used

7

QPCR and immunofluorescence staining to detect the pluripotency related gene

8

expressions. Both heterozygous and homozygous TNNT2 R92Q hESCs exhibited

9

similar morphology (protuberant and clonal morphology without spontaneous

10

differentiation, high ratio of nuclei and cytoplasm, round or ellipse-edged smooth

11

clone) as wild type hESCs and expressed pluripotent markers such as OCT4 (red),

12

SSEA4 (green) (Fig. 2A). Furthermore, teratoma formation assay showed TNNT2

13

R92Q hESCs formed teratoma containing three germ layers, including gut epithelium

14

(endoderm), cartilage (mesoderm) and sebaceous gland (ectoderm), in vivo (Fig. 2B).

15

Meanwhile, high levels of Nanog, Sox2, Oct4, and Klf4 expressed in all the three cell

16

lines (Fig. 2C). All the results indicated that our human embryonic stem cells with

17

heterozygous or homozygous R92Q mutation maintained pluripotency and

18

proliferation capability.

19

3.3 R92Q mutant hESCs-derived cardiomyocytes showed larger cell size and

20

abnormal myofilaments organization

21

Wild type and R92Q mutant hESCs were differentiated into cardiomyocytes 12

1

according to a monolayer-based cardiac differentiation protocol as previously

2

described [10] (Fig. 3A). Spontaneously contracting cells emerged on day 8 post

3

induction

4

hESC-cardiomyocytes. As cTnT is a structural protein in thin filament, mutations in

5

TNNT2 mostly result in the change of myofilament. One month after cardiac

6

differentiation, we found altered myofilament structures and a relative larger cell size

7

in hetero- R92Q cardiomyocytes compared with wild type (WT) cells (Fig. 3B-C and

8

Additional file 6A-B). Meanwhile, homozygous R92Q mutant cardiomyocytes could

9

not beat. In terms of the myofilament arrangement several dotted structures were

10

observed. Highly destructive structure of myofilaments in homozygotes may result in

11

the disability of contraction.

12

3.4

13

phenotypes of HCM

14

Myosin heavy chain 7 (MYH7) and myosin heavy chain 6 (MYH6) are the major

15

proteins comprising thick filament in cardiomyocytes and play a pivotal role in

16

cardiomyocytes contraction. Mutations in MYH7 have been shown to associate with

17

inherited and hypertrophic cardiomyopathy, while mutations in MYH6 associated

18

with late-onset hypertrophic cardiomyopathy. Myocyte-specific enhancer factor 2c

19

(MEF2c) is involved in cardiac morphogenesis and myogenesis and vascular

20

development [11]. We also detected the expression of well-characterized markers of

21

hypertrophy, such as MYH7/MHY6 and MEF2c[12], which also increased in

of

differentiation

Heterozygous

TNNT2

except

R92Q

in

homozygous

hESC-cardiomyocytes

13

TNNT2

showed

R92Q

partial

1

heterozygous R92Q cardiomyocytes compared to wild type cardiomyocytes (Fig. 3D

2

and E).

3

In cardiac muscle cells, atrial natriuretic peptide (ANP) is a peptide hormone which

4

is synthesized and secreted in the walls of the atria while brain natriuretic peptide

5

(BNP) is secreted in the ventricles. ANP and BNP have similar biological effects and

6

same receptors. Our R92Q mutant cardiomyocytes model showed decreased

7

expression of ANP and BNP (Fig. 3E). Sarcoplasmic reticulum Ca2+-ATPase2a

8

(SERCA2a) was found in the membrane of the sarcoplasmic reticulum (SR) and

9

played an important role in Ca2+ cycle by removing cytosolic Ca2+ from the

10

cardiomyocytes and pumping it back into the SR during diastole[13]. In heterozygous

11

and homozygous R92Q cardiomyocytes, SERCA2a showed decreased expression (Fig.

12

3E), which may increase the risk of arrhythmias in patients.

13

3.5 Heterozygous R92Q hESC-cardiomyocytes exhibited efficient responses to

14

heart-related pharmaceutical agents and increased contractility.

15

Heterozygous cardiomyocytes exhibited spontaneous beatings in vitro and thus could

16

be used in testing cardiovascular disease related pharmaceutical reagents. We next

17

studied

18

cardiomyocytes through multiple electrode array (MEA) system. As shown in Figure

19

4A, when treated with epinephrine (0.1 mg/ml), an adrenergic receptor agonist, both

20

wild type and heterozygous cardiomyocytes showed a positive chronotropic response,

21

which were partially reversed by the β-adrenergic receptor blocker metoprolol (0.25

the

electrophysiological

properties

14

of

spontaneously

contracting

1

mg/ml). Compared to wild type (WT) cells, heterozygous cardiomyocytes (Hetero)

2

beat faster and showed arrhythmia after epinephrine treatment and the adrenergic

3

(Epn) sensitivity was relatively increased (Fig. 4 A and C).

4

To investigate the influence of cTnT-R92Q mutation on cardiac contractility, we

5

next measured the relative contraction force of dissociated single beating

6

cardiomyocytes.

7

cardiomyocytes showed significantly increased contraction forces (Fig. 4 B, 4D and

8

Additional file 6C-D). Similar to MEA results, single cardiomyocyte also showed

9

increased beating frequency (Fig. 4B and Additional file 6C). These results indicated

10

that heterozygous R92Q mutation increased the contraction force and beating rate of

11

hESCs-derived cardiomyocytes, which correlated with the hypertrophic phenotypes.

12

3.6 Heterozygous R92Q hESC-cardiomyocytes showed altered Ca2+ handling

13

properties

14

Previous studies showed calcium handling abnormalities were associated with cardiac

15

hypertrophy and arrhythmias [14, 15]. To investigate the Ca2+ handling behaviors of

16

heterozygous R92Q hESC-cardiomyocytes, we next analyzed the spontaneous

17

calcium transients in wild type and heterozygous R92Q cardiomyocytes using

18

fluorescent Ca2+ dye Cal-520 acetoxymethyl ester (AM) (Fig. 5A-B and Additional

19

file 6E-F). Compared with wild type (WT) cardiomyocytes, heterozygous R92Q

20

cardiomyocytes (Hetero) exhibited shorter peak to peak time and transient duration 50

21

(Fig. 5C, 5E and Additional file 6G). The increased decay time implied dysfunction of

Compared

to

WT

cardiomyocytes,

15

heterozygous

R92Q

1

SR (Fig. 5G and Additional file 6I) in heterozygous R92Q cardiomyocytes. However,

2

transient amplitude and irregular Ca2+ transients (Fig. 5D, 5H and Additional file 6J)

3

showed no significant differences compared to wild type cardiomyocytes. Taken

4

together, TNNT2 R92Q mutant hESC-cardiomyocytes exhibited Ca2+ dysregulation,

5

which may play a crucial role in the development of HCM.

6

3.7 Engineering heart tissues (EHTs) using heterozygous R92Q cardiomyocytes

7

Engineered heart tissues constructed using the mutant hESC-cardiomyocytes may be

8

of great value in studying disease mechanisms in the future. We have shown that

9

engineered heart tissues (EHTs) made from human pluripotent stem cell-derived

10

cardiomyocytes exhibited spontaneous contractions in vitro and could be applied in

11

testing different pharmaceutical reagents. We then next set out to construct disease

12

EHTs using mutant R92Q cardiomyocytes. Decellularization of rat hearts were

13

prepared (Fig. 6A) as previously described[16]. The decellularized heart extracellular

14

matrix (ECM) was cut into pieces with desired shape and size (2×2 mm) using a

15

surgical scissor under sterile condition, and hESC-cardiomyocytes were seeded onto

16

the ECM at 104 cells/mm2 (Fig. 6B). We next examined electrophysiological

17

properties of the engineered heart tissues (EHTs) using multielectrode array (MEA)

18

system. EHTs seeded on the MEA probes contracted spontaneously after 2 days (Fig.

19

6C). Similar to multicellular MEA results, both wild type and heterozygous R92Q

20

EHTs led to a positive chronotropic response, which were partially reversed by the

21

β-adrenergic receptor blocker metoprolol (Mtl). Compared to wild type (WT), 16

1

heterozygous EHTs beated faster (Fig. 6D) and the adrenergic (Epn) sensitivity

2

relatively increased. Taken altogether, these results demonstrated that R92Q mutant

3

EHTs showed similar drug responses as to HCM patients, which could be used for

4

further disease mechanism study and cardiac drug screening in vitro.

5

3.8 Evaluation of drug-induced cardiotoxicity in wild type and R92Q mutant

6

cardiomyocytes

7

As patients with preexisting heart diseases such as HCM or DCM are particularly

8

sensitive to cardiotoxic drugs and showed fatal arrhythmias, we attempted to test

9

whether hESC-cardiomyocytes carrying R92Q mutation could model the increased

10

susceptibility of genetic cardiac disorders to drug-induced cardiotoxicity. Caspofungin,

11

a clinical antifungal drug, often causes heart toxicity especially when heavy dose or

12

vein given medicines was used. We used caspofungin to test the response of wild type

13

and R92Q mutant cardiomyocytes to drug-induced cardiotoxicity. First, we used wild

14

type hESC-cardiomyocytes to test the dose causing cardiotoxicity. Cells were treated

15

with a range of doses (0 µg/mL–80 µg/mL) of caspofungin, and changes in the

16

arrangement of myofilament (stained with cTnT) was used to evaluate cardiotoxic

17

response (Fig. 7A). The percentage of disorganized cells in wild type

18

hESC-cardiomyocytes increased after caspofungin treatment and started to show

19

damages at concentrations of 20 µg/mL and obvious cardiotoxicity (

20

shown at concentrations of 40 µg/mL (Fig. 7 A and B). Compared with wild type, we

21

found almost all R92Q mutant cardiomyocytes exhibited severe cardiotoxicity with 20 17

50%) was

1

µg/mL caspofungin (Fig. 7C). These results indicated that hESCs-cardiomyocytes

2

carrying HCM R92Q mutation are more sensitive to caspofungin-induced

3

cardiotoxicity and can recapitulate clinical susceptibility of HCM patients to

4

drug-induced cardiotoxicity.

5

3.9 Dysregulation of calcium might participated in the development of R92Q

6

mutation induced HCM

7

To analyze the mechanism of R92Q troponin T mutation induced HCM phenotype, ,

8

we next performed the whole transcriptomic RNA-sequencing of cardiomyocytes

9

derived from the wild type and R92Q hESCs and compared their gene expression

10

profiles at one month post cardiac differentiation. Since homozygous R92Q

11

cardiomyocytes do not beat at all and genes changed might show developmental

12

defect, we used heterozygous R92Q cardiomyocytes in this study. Metascape

13

enrichment was used to analysis the genes up-regulated in R92Q mutant

14

cardiomyocytes, where bars are discretely colored to encode p-values (Fig. 8A). The

15

intra-cluster and intercluster similarities of enriched terms were showed in the

16

network (Fig. 8B). Results showed that genes involved in calcium channels

17

expression and calcium signaling were mainly enriched in the R92Q mutant

18

cardiomyocytes (Fig. 8 A and C). Genes related to heart and muscle development also

19

increased in R92Q mutant cardiomyocytes (Fig. 8 A and D). Calcium is critical to

20

maintaining the heart-beat, intracellular Ca2+ needed to bind troponin C and induce the

21

actin-myosin crossbridge cycling. As a part of troponin complex, R92Q mutation on 18

1

troponin T affects Ca2+ binding to myofilaments, altered intracellular Ca2+ buffering

2

and abnormal calcium handling might drive deleterious cellular remodeling and

3

participated in the early development of hypertrophy.

4

4. Discussion

5

Mutations of sarcomere protein genes were the molecular basis of most familial HCM

6

and mutations in TNNT2 gene had been confirmed to account for about 5% of HCM.

7

A variety of TNNT2 mutations lead to HCM, such as I79N [17]

8

[19] A104V [20] F110I [21] ∆160E [22] E163R [23] S179F [24] and K273E [25].

9

Interestingly, TNNT2 mutations of HCM patients were mostly missense mutation,

10

these changes in individual amino acids usually caused ventricular hypertrophy,

11

myocardial disorder and fibrosis [26]. Phenotypes caused by the TNNT2 mutations

12

were not obvious, but patients often had high incidence of sudden cardiac death [27].

13

In view of this, it was urgent to elucidate the pathogenesis of TNNT2 gene mutations,

14

especially missense mutations, and to find corresponding therapeutic targets for drugs

15

treatment.

R92Q [18]

R95H

16

TNNT2 gene encode cardiac troponin T (cTnT), which combined with cardiac

17

troponin C (cTnC) and cardiac troponin I (cTnI) to form a complex contracted

18

synergistically with tropomyosin. They played a vital role in the process of the cardiac

19

excitation- contraction coupling [28, 29, 30]. Changes in the structure or function of

20

cTnT could influenced Ca2+ cycle and contraction of cardiomyocytes [31, 32, 33]. The

21

92th arginine to glutamine mutation (R92Q) of cTnT was one of the HCM mutant 19

1

hotspots. CTnT R92Q mutation often leads to severe phenotype of HCM during

2

adolescence and early adulthood and the probability of sudden cardiac death is high

3

[34, 35]. In order to illuminate the pathogenesis of TNNT2 R92Q mutation in

4

hypertrophic cardiomyopathy, we obtained R92Q mutant hESC-cardiomyocytes

5

utilizing TALEN mediated homogenous recombination and cardiac differentiation of

6

hESCs.

7

Through 2D cardiac differentiation, we found heterozygous R92Q mutant

8

cardiomyocytes were capable of independent contracting while homozygous ones

9

cannot. Even though all the three cell lines could express cardiac specific marker

10

cTnT,

homozygous

R92Q

11

myofilament disorganization which may lead to juvenile lethality [36]. Both

12

heterozygous and homozygous R92Q cardiomyocytes exhibited changed sarcomere

13

organization and increased myocardial cell size. In addition, heterozygous R92Q

14

cardiomyocytes

15

(MYH7/MYH6, MEF2c) compared to wild type. The above results were similar to

16

HCM patient's phenotype. Though the expression of ANP, BNP and SERCA2a

17

decreased, they were similar to HCM mouse model with R92Q mutation in previous

18

reports [37, 38].

showed

mutant

upregulated

hESC-cardiomyocytes

expression

of

exhibited

hypertrophic

severe

markers

19

Heterozygous R92Q cardiomyocytes beat faster and showed relatively increased

20

adrenergic (Epn) sensitivity and contractility. Calcium transient results exhibited Ca2+

21

dysregulation, RNA-sequencing also showed primarily changed expression of genes 20

1

related to calcium signaling and heart development. Human cardiomyocytes carrying

2

specific single mutations derived from hESCs were in the most early developmental

3

stage and could show effects of genetic mutations in the initial phases of disease

4

development. Our results demonstrated that dysregulation of calcium most possibly be

5

the core factors participated in the early development of TNNT2 R92Q induced

6

hypertrophy,

7

cardiomyocytes showed similar response to electrophysiological drugs as to HCM

8

patients and these R92Q cardiomyocytes could also model the increased susceptibility

9

of genetic cardiac disorders to drug-induced cardiotoxicity.

arrhythmia

and

remodeling.

EHTs

constructed

using

R92Q

10

In summary, our results indicated that TNNT2 R92Q mutation caused changes in

11

the expression of cardiac genes, abnormal heart rhythms and Ca2+ cycle, elevated

12

contractility, and cardiac hypertrophy in hESCs-derived cardiomyocytes. We believe

13

that these hESC-derived TNNT2 R92Q mutant cardiomyocytes and EHTs could

14

complement to animal models and serve as good in vitro human disease models.

15

Future studies can apply these models in the study of HCM pathogenic mechanism,

16

high throughput drug screening to identify potential therapeutic agents, and drug

17

toxicity screening.

18

19

Competing interests

20

The authors declare that they have no competing interests.

21

1

Ethics approval

2

All human stem cells research followed the ISSCR Guidelines for the Conduct of Human

3

Embryonic Stem Cell Research. The human ESC line H7 and H9 used in this study was

4

obtained from WiCell Research Institute under specific Material Transfer Agreement.

5

Funding

6

This work was supported by the National Natural Science Foundation of China

7

(NSFC No.31571527, N.S.; No.31501098, Q.L.; No.81500241, C.X.), the

8

Postdoctoral Science Foundation (No.KLH1322109, B.L.), the National Key R&D

9

Program of China (2018YFC2000202), and the Science and Technology Commission

10

of Shanghai Municipality (No. 17XD1400300, No.17JC1400200).

11

Acknowledgements

12

We thank Hui Yang and Qingjie Wang for ECM fabrication and rat surgery, Dr

13

Yongkun Zhan for discussions on the modeling methods and Dr. Ying Chen for

14

assistance in Ca2+ transient and relative contractility measurements.

15

16

17

18

19 22

1

References:

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

1

Maron BJ, Ommen SR, Semsarian C, et al. Hypertrophic cardiomyopathy: present and future,

with translation into contemporary cardiovascular medicine. J AM COLL CARDIOL 2014;64:83-99. 2

Lan F, Lee AS, Liang P, et al. Abnormal calcium handling properties underlie familial

hypertrophic cardiomyopathy pathology in patient-specific induced pluripotent stem cells. CELL STEM CELL 2013;12:101-13. 3

Wilde AA, Behr ER. Genetic testing for inherited cardiac disease. NAT REV CARDIOL

2013;10:571-83. 4

Semsarian C, Ingles J, Maron MS, et al. New Perspectives on the Prevalence of Hypertrophic

Cardiomyopathy. J AM COLL CARDIOL 2015;65:1249-54. 5

Anan R, Shono H, Kisanuki A, et al. Patients with familial hypertrophic cardiomyopathy caused

by a Phe110Ile missense mutation in the cardiac troponin T gene have variable cardiac morphologies and a favorable prognosis. CIRCULATION 1998;98:391-7. 6

Montag

J,

Petersen

B,

Flögel

AK,

et

al.

Successful

knock-in

of

Hypertrophic

Cardiomyopathy-mutation R723G into the MYH7 gene mimics HCM pathology in pigs. SCI REP-UK 2018;8:1-12. 7

Xie C, Zhang Y, Song L, et al. Genome editing with CRISPR/Cas9 in postnatal mice corrects

PRKAG2 cardiac syndrome. CELL RES 2016;26:1099-111. 8

Maruyama T, Dougan SK, Truttmann MC, et al. Increasing the efficiency of precise genome

editing with CRISPR-Cas9 by inhibition of nonhomologous end joining. NAT BIOTECHNOL 2015;33:538-42. 9

Watkins H, McKenna WJ, Thierfelder L, et al. Mutations in the genes for cardiac troponin T and

alpha-tropomyosin in hypertrophic cardiomyopathy. N Engl J Med 1995;332:1058-64. 10

Burridge PW, Li YF, Matsa E, et al. Human induced pluripotent stem cell

derived

cardiomyocytes recapitulate the predilection of breast cancer patients to doxorubicin-induced cardiotoxicity. NAT MED 2016;22:547-56. 11

Konno T, Chen D, Wang L, et al. Heterogeneous myocyte enhancer factor-2 (Mef2) activation in

myocytes predicts focal scarring in hypertrophic cardiomyopathy. Proceedings of the National Academy of Sciences 2010;107:18097-102. 12

Rohini A, Agrawal N, Koyani CN, et al. Molecular targets and regulators of cardiac hypertrophy.

PHARMACOL RES 2010;61:269-80. 13

Lopez-Crisosto C, Pennanen C, Vasquez-Trincado C, et al. Sarcoplasmic reticulum

mitochondria

communication

in

cardiovascular

pathophysiology.

NAT

REV

CARDIOL

2017;14:342-60. 14

Yanaga F, Morimoto S, Ohtsuki I. Ca2+ sensitization and potentiation of the maximum level of

myofibrillar ATPase activity caused by mutations of troponin T found in familial hypertrophic cardiomyopathy. J BIOL CHEM 1999;274:8806-12. 15

Knollmann BC. Familial Hypertrophic Cardiomyopathy-Linked Mutant Troponin T Causes

Stress-Induced Ventricular Tachycardia and Ca2+-Dependent Action Potential Remodeling. CIRC RES 2003;92:428-36.

23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

16

Wang Q, Yang H, Bai A, et al. Functional engineered human cardiac patches prepared from

nature's platform improve heart function after acute myocardial infarction. BIOMATERIALS 2016;105:52-65. 17

Wang L, Kim K, Parikh S, et al. Hypertrophic cardiomyopathy-linked mutation in troponin T

causes myofibrillar disarray and pro-arrhythmic action potential changes in human iPSC cardiomyocytes. J MOL CELL CARDIOL 2018;114:320-7. 18

Marian AJ, Zhao G, Seta Y, et al. Expression of a mutant (Arg92Gln) human cardiac troponin T,

known to cause hypertrophic cardiomyopathy, impairs adult cardiac myocyte contractility. CIRC RES 1997;81:76-85. 19

Mickelson AV, Chandra M. Hypertrophic cardiomyopathy mutation in cardiac troponin T (R95H)

attenuates length-dependent activation in guinea pig cardiac muscle fibers. AM J PHYSIOL-HEART C 2017;313:H1180-9. 20

Nakajima-Taniguchi C, Matsui H, Fujio Y, et al. Novel missense mutation in cardiac troponin T

gene found in Japanese patient with hypertrophic cardiomyopathy. J MOL CELL CARDIOL 1997;29:839-43. 21

Anan R, Shono H, Kisanuki A, et al. Patients with familial hypertrophic cardiomyopathy caused

by a Phe110Ile missense mutation in the cardiac troponin T gene have variable cardiac morphologies and a 22

favorable prognosis. CIRCULATION 1998;98:391-7.

Harada K, Takahashi-Yanaga F, Minakami R, et al. Functional consequences of the deletion

mutation deltaGlu160 in human cardiac troponin T. J BIOCHEM 2000;127:263-8. 23

Ferrantini C, Coppini R, Pioner JM, et al. Pathogenesis of Hypertrophic Cardiomyopathy is

Mutation Rather Than Disease Specific: A Comparison of the Cardiac Troponin T E163R and R92Q Mouse Models. J AM HEART ASSOC 2017;6:e5407. 24

Ho CY, Lever HM, DeSanctis R, et al. Homozygous mutation in cardiac troponin T: implications

for hypertrophic cardiomyopathy. CIRCULATION 2000;102:1950-5. 25

Fujino N, Shimizu M, Ino H, et al. A novel mutation Lys273Glu in the cardiac troponin T gene

shows high degree of penetrance and transition from hypertrophic to dilated cardiomyopathy. AM J CARDIOL 2002;89:29-33. 26

Watkins H, McKenna WJ, Thierfelder L, et al. Mutations in the genes for cardiac troponin T and

alpha-tropomyosin in hypertrophic cardiomyopathy. N Engl J Med 1995;332:1058-64. 27

Maass AH. Hypertrophy, Fibrosis, and Sudden Cardiac Death in Response to Pathological Stimuli

in Mice With Mutations in Cardiac Troponin T. CIRCULATION 2004;110:2102-9. 28

Takeda S, Yamashita A, Maeda K, et al. Structure of the core domain of human cardiac troponin

in the Ca(2+)-saturated form. NATURE 2003;424:35-41. 29

Bers DM. Cardiac excitation-contraction coupling. NATURE 2002;415:198-205.

30

Sehnert AJ, Huq A, Weinstein BM, et al. Cardiac troponin T is essential in sarcomere assembly

and cardiac contractility. NAT GENET 2002;31:106-10. 31

Gersh BJ, Maron BJ, Bonow RO, et al. 2011 ACCF/AHA Guideline for the Diagnosis and

Treatment of Hypertrophic Cardiomyopathy. J AM COLL CARDIOL 2011;58:e212-60. 32

Maron BJ, Maron MS. Hypertrophic cardiomyopathy. LANCET 2013;381:242-55.

33

Frey N, Luedde M, Katus HA. Mechanisms of disease: hypertrophic cardiomyopathy. NAT REV

CARDIOL 2011;9:91-100.

24

1 2 3 4 5 6 7 8 9 10 11 12 13 14

34

Ertz-Berger BR, He H, Dowell C, et al. Changes in the chemical and dynamic properties of

cardiac troponin T cause discrete cardiomyopathies in transgenic mice. Proc Natl Acad Sci U S A 2005;102:18219-24. 35

Sweeney HL, Feng HS, Yang Z, et al. Functional analyses of troponin T mutations that cause

hypertrophic cardiomyopathy: insights into disease pathogenesis and troponin function. Proc Natl Acad Sci U S A 1998;95:14406-10. 36

Ho CY, Lever HM, DeSanctis R, et al. Homozygous mutation in cardiac troponin T: implications

for hypertrophic cardiomyopathy. CIRCULATION 2000;102:1950-5. 37

Lombardi R, Bell A, Senthil V, et al. Differential interactions of thin filament proteins in two

cardiac troponin T mouse models of hypertrophic and dilated cardiomyopathies. CARDIOVASC RES 2008;79:109-17. 38

Ferrantini C, Coppini R, Pioner JM, et al. Pathogenesis of Hypertrophic Cardiomyopathy is

Mutation Rather Than Disease Specific: A Comparison of the Cardiac Troponin T E163R and R92Q Mouse Models. J AM HEART ASSOC 2017;6:19.

15 16

25

1

Figure Legends

2 3

Figure 1. Schematic overview of the targeting strategy for the R92Q mutation

4

and sequencing verification. A, Schematic overview of the TALEN targeting

5

strategy. Nucleotide sequences in the blue frame indicate the specific binding sites for

6

the TALEN R92 L1/R1 within exon 10 of human TNNT2. The donor vector was

7

comprised of 5' homology arm with R92Q mutation, loxP-flanked PGK-puromycin 26

1

cassette and 3' homology arms. To prevent disruption of TNNT2 expression, the

2

PGK-PURO cassette was inserted in the adjacent intron downstream of the R92Q

3

mutation and was cut at last. Blue boxes: WT or R92Q mutant human TNNT2

4

sequence; Arrows: loxP pairs; PGK was the abbreviation of phosphoglycerol kinase

5

promoter while PURO for puromycin resistant gene. B, Sequencing of mutations in

6

heterozygous and homozygous TNNT2 R92Q hESCs. The mutation site was close to

7

the right end of the PCR products, reverse primer was used when DNA sequencing, so

8

both heterozygote and homozygote indicated CTG.

9

10 11

Figure 2. Maintenance of pluripotency in wild type and TNNT2 R92Q mutant

12

hESCs. A, All hESC lines expressed the pluripotency markers Oct4 and SSEA4. Red: 27

1

Oct4, Green: SSEA4, Blue: DAPI for nuclei. B, Teratomas with tissues representing

2

all three embryonic germ layers including gut epithelium (endoderm), cartilage

3

(mesoderm) and sebaceous gland (ectoderm). Scale bars, 50 µm. C, All the three cell

4

lines showed high expression of Nanog, Sox2, Oct4 and Klf4.

5

28

1

2

Figure 3. Myofilament structure and gene expression of R92Q mutant

3

cardiomyocytes. A, The cardiac differentiation protocol used in this study.

4

Representative immunostaining images of cardiac troponin T (cTnT) at one month

5

after differentiation showed that myofilament structure disorganized (B) and cell size 29

1

increased (C) in heterozygous and homozygous R92Q hESC-cardiomyocytes

2

compared with wild type cells. Scale bars, 50 µm. Red: cTnT, Blue: DAPI. D and E,

3

Levels of mRNAs for selected markers of cardiac hypertrophy such as MYH7/MHY6,

4

ANP, BNP, SERCA2a and MEF2c were shown.

5

6 30

1

Figure 4. Representative traces of MEA recordings and relative contractility for

2

WT and Heterozygous R92Q mutant hESC-cardiomyocytes. A, The electrical

3

signals recorded by MEA in response to epinephrine and metoprolol. X axes: time

4

(second), Y axes: Voltage (µV). B, Representative images showed the relative

5

contractility measured by Nikon edge detection system. Voltage of Y axes indirectly

6

reflected the relative contractility of beating cardiomyocytes. C, Beating rate recorded

7

by MEA. D, Relative contractility recorded by edge detection system.

8

31

1 2

Figure 5. Representative Ca2+ line scan images, waveforms and quantification of 32

1

calcium handling parameters of WT and Hetero R92Q mutant cardiomyocytes.

2

A, Ca2+ line scan images; B, waveforms; Quantification of calcium handling

3

parameters: (C) peak to peak time; (D) transient amplitude (△F/F0); (E) transient

4

duration 50; (F) time to peak; (G) decay time and (H) ratio of CMs with irregular

5

Ca2+ transients, n>20 in each group.

6

7

33

1

Figure 6. Decellularization of adult rat hearts and characteristics of the

2

engineered heart tissues. A, Pictures of rat hearts before decellularization, after

3

perfusion with deionized water and PBS each for 1 hour, perfusion with 1% SDS

4

solution for 2 hours and after perfusion with solution Amp/Str solution for 1hour. B,

5

Engineering heart tissues by combining extracellular matrix (ECM) with mature

6

hESC-cardiomyocytes. C, Representative image of an

7

cultured on the MEA probe. D, MEA recordings of the WT and Hetero- cTnT R92Q

8

mutant EHTs in response to epinephrine and metoprolol. X axes: time (second), Y

9

axes: Voltage (µV).

engineered

heart

tissue

10

11 12

Figure 7. R92Q mutant cardiomyocytes showed increased susceptibility to

13

caspofungin-induced cardiotoxicity. A, Representative immunostaining images of

14

thin filament in WT cardiomyocytes showed dose-dependent cardiotoxicity of 34

1

caspofungin. B, Percentage of cardiomyocytes exhibiting caspofungin-induced

2

myofilaments disorganization and dispersion in wild type hESC-cardiomyocytes at

3

different drug concentrations. C, R92Q mutant cardiomyocytes exhibited severe

4

damages at concentrations as low as 20 µg/mL caspofungin.

5

6

35

1

Figure 8. Calcium signaling and heart development related genes were changed

2

in R92Q mutant cardiomyocytes. A, Top non-redundant enrichment clusters were

3

shown using Metascape bar graph. Enriched clusters were exhibited, and discrete

4

color scale represented statistical significance. B, Network visualization of the

5

enriched clusters. Cluster annotations are shown in color code. C, Compared with WT

6

cardiomyocytes, fold changes of several key genes in calcium signaling and (D) heart

7

development.

36

Highlights


TNNT2 R92Q mutant hESCs were constructed using genome-editing technique.




TNNT2 R92Q hESC derived cardiomyocytes showed phenotypes of HCM.




R92Q mutant human EHTs responsed to electrophysiological drugs.




Dysregulation of calcium participated in the development of R92Q mutation induced HCM.

Conflict of interest form The authors declared that they have no conflicts of interest to this work. We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.