Low expression of the K280N TNNT2 mutation is sufficient to increase basal myofilament activation in human hypertrophy cardiomyopathy

Abstract

Background

Hypertrophic cardiomyopathy (HCM) is an autosomal dominant genetic disorder with patients typically showing heterozygous inheritance of a pathogenic variant in a gene encoding a contractile protein. Here, we study the contractile effects of a rare homozygous mutation using explanted tissue and human-induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) to gain insight into how the balance between mutant and WT protein expression affects cardiomyocyte function.

Methods

Force measurements were performed in cardiomyocytes isolated from a HCM patient carrying a homozygous troponin T mutation (cTnT-K280N) and healthy donors. To discriminate between mutation-mediated and phosphorylation-related effects on Ca²⁺-sensitivity, cardiomyocytes were treated with alkaline phosphatase (AP) or protein kinase A (PKA). Troponin exchange experiments characterized the relation between mutant levels and myofilament function. To define mutation-mediated effects on Ca²⁺-dynamics we used CRISPR/Cas9 to generate hiPSC-CMs harbouring heterozygous and homozygous TnT-K280N mutations. Ca²⁺-transient and cell shortening experiments compared these lines against isogenic controls.

Results

Myofilament Ca²⁺-sensitivity was higher in homozygous cTnT-K280N cardiomyocytes and was not corrected by AP- and PKA-treatment. In cTnT-K280N cells exchanged with cTnT-WT, a low level (14%) of cTnT-K280N mutation elevated Ca²⁺-sensitivity. Similarly, exchange of donor cells with 45 ± 2% cTnT-K280N increased Ca²⁺-sensitivity and was not corrected by PKA. cTnT-K280N hiPSC-CMs show elevated diastolic Ca²⁺ and increases in cell shortening. Impaired cardiomyocyte relaxation was only evident in homozygous cTnT-K280N hiPSC-CMs.

Conclusions

The cTnT-K280N mutation increases myofilament Ca²⁺-sensitivity, elevates diastolic Ca²⁺, enhances contractility and impairs cellular relaxation. A low level (14%) of the cTnT-K280N sensitizes myofilaments to Ca²⁺, a universal finding of human HCM.

Graphical abstract

Highlights

• •Protein-dosage effects are detrimental in hypertrophic cardiomyopathy progression.
• •Low (14%) mutant protein is sufficient to increase myofilament Ca²⁺-sensitivity.
• •Elevated myofilament Ca²⁺-sensitivity associates with high diastolic Ca²⁺
• •Mutations at the C-terminus of cTnT disrupt the thin-filament inactive state.

1. Introduction

Hypertrophic cardiomyopathy (HCM) is the most common inherited form of heart failure (1:200 prevalence) and is most frequently caused by mutations of genes encoding sarcomeric proteins that increase the Ca²⁺-sensitivity of the myofilaments [1], [2]. The most relevant clinical consequences are marked diastolic dysfunction progressing towards diastolic heart failure, and ventricular arrhythmias, with HCM being the most common cause of sudden cardiac death especially in young individuals [3], [4]. The dominant macroscopic feature of HCM is the extensive hypertrophy of the left ventricle (LV), which involves predominantly the interventricular septum. Most patients display asymmetric septal hypertrophy in combination with LV outflow tract obstruction (HOCM) at rest or during exercise. In later stages of disease, a small minority of individuals (~5%) ultimately develop systolic heart failure, characterized by reduced ejection fraction (<50%), cavity dilation, and regression of hypertrophy - i.e. dilated-hypokinetic (end-stage) evolution of HCM [5].

HCM is an autosomal dominant genetic disorder and most affected individuals are heterozygous, i.e., they carry one mutant and one wild-type allele. However, carrying a heterozygous sarcomere gene mutation cannot solely explain HCM pathology, because a mutation that causes HCM in one individual may be harmless in a sibling [6]. In addition, HCM typically develops at an older age (>20 years), while sarcomere gene mutations are present from birth. This indicates that secondary disease-modifiers are necessary to trigger and modulate the course of HCM development.

Having in mind the low penetrance between genotype and the clinical presentation of HCM, evidence suggests that it is the balance between mutant and wild-type protein that determines cardiomyocyte dysfunction in human myocardium [7], [8]. Because of the system's redundancy with recycling and rebalancing of myocellular proteins, dysfunction will not become evident as long as the amount of mutant protein is kept below the toxicity threshold. Proof for this effect comes from very rare cardiomyopathy cases in which both alleles are mutated. These homozygous or compound mutations are associated with a reduced life expectancy and heart transplantation at young age [9]. Notably, the expression level of the heterozygous mutant protein, i.e., toxic threshold, has been associated with varying phenotypes in HCM models in which higher expression levels of mutant proteins coincided with more severe disease phenotypes [7], [8].

Here, we provide experimental proof for the toxicity concept. Mutations in thin-filament proteins are characterized by an increased myofilament Ca²⁺ sensitivity, associated with development of concentric hypertrophy, and the occurrence of arrhythmias (reviewed in detail here [10]). Myofilament Ca²⁺-sensitivity was measured in membrane-permeabilized cardiomyocytes from a human HCM heart with a homozygous mutation (K280N) in the TNNT2 gene, which encodes the thin-filament protein cardiac troponin T (cTnT). In previous studies we observed that cardiomyocytes from this HCM heart show high myofilament Ca²⁺-sensitivity [11], [12]. This homozygous TNNT2 mutation results in 100% mutant cTnT and as such represents a unique tool to assess the level at which mutant protein perturbs sarcomere function. To assess how much mutant protein is sufficient to sensitize myofilaments to Ca²⁺, we performed the following troponin exchange experiments: ‘rescue’ experiments in which the cTnT-K280N was replaced by wild-type protein in the HCM patient sample, and incorporation of recombinant cTnT-K280N into non-failing donor samples. Using targeted troponin exchange in human cardiomyocytes we show that the toxicity threshold is achieved at relatively low levels, such that 14% of mutant cTnT-K280N levels is sufficient to increase myofilament Ca²⁺-sensitivity.

To further strengthen our view, we investigated the relationship between contractility and Ca²⁺ handling in human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs), where we gene-inserted the cTnT-K280N mutation using CRISPR/Cas9 technology. Homozygous cell lines of the hiPSC-CMs-K280N showed impaired cardiomyocyte relaxation evident from a significant increase in relaxation time, which associated with greater contractility and higher diastolic Ca²⁺. Increased contractility and higher diastolic Ca²⁺ levels were also identified in the heterozygous cell lines. However, the later changes did not coincide with alterations in cellular relaxation rates.

Altogether, these findings support the dominant effects of the TNNT2 K280N mutation on myofilament function. This effect is evident from the fact that low cTnT-K280N myofilament incorporation (14%) is sufficient to increase myofilament Ca²⁺-sensitivity and promote basal myofilament activation. Moreover, the cTnT-K280N mutation associates with alterations in Ca²⁺ handling evident from slowing of cellular relaxation in the hiPSC-CMs-K280N homozygous cell line. The K280N-mediated increase in myofilament Ca²⁺ sensitization is likely to precede perturbations in Ca²⁺ handling and consequently impaired cardiomyocyte relaxation.

2. Materials and methods

2.1. Cardiac tissue

Tissue from the free LV wall and interventricular septum (IVS) was obtained during cardiac transplantation surgery from a male diagnosed with HCM. Genotyping identified a homozygous TNNT2 mutation (K280N, TNNT2_mut) [13]. The patient had a Lebanese background and was 26 years at the time of heart transplantation surgery. He had undergone septal myectomy before heart transplantation surgery. Macroscopic analysis of the explanted heart revealed a septum thickness of 37 mm. The posterior LV wall measured 27 mm in thickness. Microscopic analysis revealed fiber disarray and a moderate diffuse degree of interstitial fibrosis. Non-failing cardiac LV tissue was obtained from donor hearts (N = 8; 7 males, 1 female; mean age 37 ± 6) when no suitable transplant recipient was found. The donors had no history of cardiac disease, a normal cardiac examination, normal ECG and normal ventricular function on echocardiography within 24 h of heart explantation. Donor hearts were perfused with cold cardioplegic solution and subsequently frozen in liquid nitrogen (~1 g transmural pieces). Samples were obtained after informed consent and with approval by both the St. Vincent's Hospital Human Research Ethics Committee (HREC approval: H03/118) and the Sydney Heart Bank at The University of Sydney HREC (2016/923).

2.2. Myofilament force measurements

Cardiomyocytes were mechanically isolated from small cardiac tissue samples as described previously [12]. In brief, tissue samples were thawed in isolating solution containing 5.95 mM Na₂ATP, 6.04 mM MgCl₂, 2 mM EGTA, 139.6 mM KCl and 10 mM Imidazole, pH 7.1. Thereafter, cardiomyocytes were mechanically isolated by tissue disruption and permeabilized by chemical digestion with an isolation solution containing 0.5% (v/v) Triton-X100 for 5 min. The composition of all solutions was calculated based on a computer program similar to that previously described by Fabiato [14]. The pH of all solutions was adjusted to 7.1 at 15 °C with KOH and ionic strength adjusted to 180 mM with KCl. The relaxing solution (10⁻⁹ μmol/L Ca²⁺) contained 6.48 mM MgCl₂, 5.89 mM MgATP, 6.97 mM EGTA, 100 mM BES and 14.5 mM PCr. Maximal saturating Ca²⁺-activating solution (32 μmol/L Ca²⁺) consisted of 6.28 mM MgCl₂, 5.97 mM MgATP, 7 mM EGTA, 100 mM BES and 14.5 mM PCr. Ca²⁺-activating solutions with lower free [Ca²⁺] were obtained by mixing saturating Ca²⁺-activating and relaxing solutions and assuming an apparent stability constant of the Ca²⁺-EGTA complex of 10^6.35. Demembraned cardiomyocytes were thereafter glued between a force transducer and a piezoelectric motor. Isometric force measurements were performed at maximal and submaximal [Ca²⁺] (ranging from 1 to 32 μmol/L) and stretched to a sarcomere length of ~2.2 μm.

Average sarcomere length was determined by means of a spatial Fourier transformation, as previously described [15]. Passive force (F_pas) was determined by shortening the myocyte in a relaxing solution by 30% of its length. Maximal developed force (F_max) was determined by activating the cardiomyocyte at saturating [Ca²⁺] (32 μmol/L), generating a total force value (F_total). F_max was obtained by subtracting F_pas from F_total (i.e. F_max = F_total-F_pas). Maximal tension (in kN/m²) was calculated as F_max normalized to cross-sectional area of the cardiomyocytes. Force-Ca²⁺ relations were fit to a modified Hill equation and myofilament Ca²⁺-sensitivity was denoted as EC₅₀ ([Ca²⁺] at which half of F_max was reached).

In several cells, force measurements were repeated after incubation for 40 min at 20 °C in relaxing solution containing alkaline phosphatase (AP; 2000 U/mL, New England BioLabs) or the catalytic subunit of protein kinase A (PKA; 100 U/mL, Sigma).

In addition, exchange of endogenous troponin by exogenous human recombinant troponin complex force was performed in single cardiomyocytes from mutant (with 0.25, 0.5 and 1 mg/mL of exogenous troponin) and donor samples (with 0.25, 0.5 and 1 mg/mL of exogenous troponin) and isometric force measured thereafter. Exchange of troponin complex, in single human cardiomyocytes, was performed as described previously [12]. Expression of cDNA encoding human wild-type cardiac troponin subunits (cTnC, myc-tag labeled cTnT (cTnT-myc), cTnI), purification and reconstitution were performed as described previously [12]. To determine the degree of exchange of endogenous troponin (mutant or wild-type) by recombinant wild-type cardiac troponin Western blotting was performed. Recombinant wild-type cTnT was labeled with a myc-tag, which allowed differentiation between endogenous and exogenous recombinant cardiac troponin complex. Proteins were separated on a 1D SDS-PAGE and blotted onto a nitrocellulose membrane. A specific monoclonal antibody was used against cTnT (Clone JLT-12, Sigma) to detect endogenous and recombinant cTnT by chemiluminescence (ECL, Amersham Biosciences).

2.3. Protein expression and phosphorylation

Small pieces of LV or IVS were taken from the HCM heart and five non-diseased donor hearts. Fragments were treated with trichloro acetic acid prior to protein analysis to preserve the endogenous phosphorylation status of the sarcomeric proteins. Proteins in homogenates were separated on 4–15% Tris-HCl gels (Bio-Rad). To detect protein phosphorylation. The gels were stained with ProQ Diamond and subsequently with SYPRO Ruby to determine total protein content. Pro-Q Diamond Phosphoprotein Staining of 1D gels allows simultaneous analysis of multiple phosphorylated myofilament proteins and reduces the required sample size. Phosphorylation signals obtained with Pro-Q Diamond were expressed relative to SYPRO Ruby-stained cMyBP-C bands as to account for differences in sample loading. Staining of the gels was visualized using the Las-3000 Image Reader (Fuji; 460 nm/605 nm) and analyzed with AIDA, as described before [16]. Two-dimensional gel electrophoresis (pH gradient of 4 to 7) was performed as described previously to study cTnT composition in more detail [16]. In brief, the 2D gel is able to detect the minute differences in the phosphorylation status of troponin T (TnT), myosin light chain 1 (MLC1) and myosin light chain 2 (MLC2) in great detail. MLC2 is composed of two isoforms (U and U*) with both phosphorylated forms detected in our study.

2.4. Genome engineering of hiPSC using the CRISPR/Cas9 system

2.4.1. Preparation of pSpCas9 (sgRNA) plasmid and repair template

The plasmid pSpCas9(BB)-2A-Puro V2.0 (PX459; plasmid derived from Streptococcus pyogenes containing the endonuclease Cas9 gene) was a gift from Feng Zhang (Addgene plasmid #62988) [17]. Single-guide RNA (sgRNA) targeting the region adjacent to 840G of TNNT2 was designed using the online CRIPSR design tool (http://crispr.mit.edu/) and the selected sgRNA minimized the likelihood of off-target cleavage. The sgRNA sequences used in the study were: forward: 5’-CACCGTTTAGCCTTCCCGCGGGTCT-3′ and reverse: 5’-AAACAGACCCGCGGGAAGGCTAAAC-3′. The sgRNA oligos were annealed and ligated into pSpCas9 as detailed in the Zhang lab protocol [17].

The 200-nt ssODN (single-stranded DNA oligonucleotide) repair template (standard-desalting, IDT Technologies) was designed with homologous arms flanking the pathogenic mutation. The sequence of ssODN was 5’-GCCTGCTCCT CTCCCCTTTG GCACCCCAGT CCTACCCCAG CCGCATGGTG ACCTACTACC CTGCCTGTGT CTCCATGTCA CTGCGTCCTG CTTCCCCTGC AGCTCTAAGA CCCGCGGGAA TGCTAAAGTC ACCGGGCGCT GGAAATAGAG CCTGGCCTCC TTCACCAAAG ATCTGCTCCT CGCTCGCACC TGCCTCCGGC-3′. A silent mutation was introduced at the NGG protospacer adjacent motif (PAM site) complementary site region (CCA), corresponding to codons 824C, 825C and 826A (CCA) with the silent mutation introduced to codon 825 T (CTA) to improve the efficiency of homology-directed repair. Note, underlined codon bases correspond to codons 838A, 839A and 840 T, responsible for encoding residue 280 of cTnT. Codon base 840 was mutated from G to T to encode the exchange of Lysine residue (K, AAG) to Asparagine (N, AAT) to generate cTnT-K280N. Mutation to AAT facilitated recognition of correct gene-inserted mutant cells with BsmI restriction enzyme (New England BioLabs) for later genotyping, since the enzyme creates overhang cuts at 5′-GAATG_CN^-3′ sequence regions.

2.5. HiPSC maintenance and transfection

Generation and characterization of human induced pluripotent stem cells were described in [18]. HiPSC lines were maintained on growth factor-reduced Matrigel-coated plates (1:200 dilution, DMEM/F12) in mTeSR1 medium (Stemcell Technologies). Cells were passaged every four days using 0.5 mmol/L EDTA (Life Technologies) in DPBS without Ca²⁺ or Mg²⁺ (Life Technologies). After passaging, 10 μmol/L Rho kinase inhibitor (Selleck Chemicals) was added to the mTeSR1 medium for overnight. Then, cells were fed with the regular mTeSR1 medium and maintained at 37 °C, with 5% CO₂ in a humidified incubator.

HiPSC cells were co-transfected with pSpCas9-sgRNA plasmid and ssODN using the Neon transfection system (Thermo Scientific). Briefly, 1 × 10⁶ cells were re-suspended in 130 μL electroporation mix with 10 μg pSpCas9 (sgRNA) plasmid and 10 μmol ssODN. Cells were electroporated for two pulses at 1200 V for 20 ms. The electroporated cells were plated on Matrigel-coated plates in mTeSR1. To select stable transduced cells, 0.5 μg/mL puromycin was applied 36 h post transfection for two days. We manually picked and assayed single-cell colonies. To screen for on-target indel in hiPSC colonies, we PCR amplified across CRISPR target sites in both alleles and sequenced these amplicons to confirm the zygosity knock-in status. Two cell lines were identified, one with the homozygous allele modification and another with the heterozygous allele modification.

2.6. Characterization of cTnT-K280N hiPSC lines

At day 12 post-transfection, cells were harvested by 0.5 mmol/L EDTA and pelleted by 200 ×g, 5 min. Genomic DNA was extracted using QuickExtract DNA Extraction Solution (Epicentre) following the recommended protocol. The genomic region flanking pathogenic mutation was amplified by PCR using primers below: forward: 5’-CTGCTTCAGCCCACAGGTTTC-3′ and reverse: 5’-TTATTACTGGTGTGGAGTGGGTG-3′. To identify clones with homology directed repair events, PCR products were digested with BsmI enzyme to screen for a novel restriction site introduced with the pathogenic and silent mutations and were analyzed by agarose gel electrophoresis. The zygosity of the pathogenic mutation in clones that had undergone incorporation of the silent mutation was determined by sequencing.

2.7. In vitro cardiac differentiation

HiPSC-derived cardiomyocytes were generated using an established small molecule method [19]. Briefly, hiPSCs were grown for four days until they reached 80% confluence. Cardiac differentiation was induced by 6 μmol/L CHIR99021 (Selleck Chemicals) in RPMI 1640 plus B27 supplement minus insulin (Life Technologies) as M1 medium. On day 2, the cells were growth in the M1 medium without CHIR99021. On days 3–5, the cells were treated with 5 μmol/L IWR-1 (Sigma) in the M1 medium. At day 10, the cells were kept in the RPMI 1640 without glucose (Life Technologies) plus B27 without insulin. On day 16, the medium was switched to RPMI 1640 medium (Life Technologies) plus 2% B-27 supplement (Invitrogen). On day 30 post-cardiac induction, hiPSC-CMs were dissociated and plated at low density on flexible Matrigel mattresses [19].

2.8. Matrigel mattress preparation and measurements of cellular contractility and Ca²⁺ transients

To investigate the relationship between contractility and Ca²⁺-handling in hiPSC-CMs with the cTnT-K280N mutation we used the Matrigel mattress method that enables the rapid generation of robustly contracting hiPSC-CMs, while improving cellular maturation [19]. This method provides a physiological load (5-7 kPa) on the hiPSC-CMs and as a result, the sarcomere length of hiPSC-CMs on the Matrigel Mattress is over 2 μm [19], close to the physiological levels of loaded cardiomyocytes in cardiac tissue. Besides, the extracellular matrix provided by Matrigel Mattress mimics the mechanical load conditions of cardiac tissue, while allowing the generation of rod-like hiPS cardiomyocytes with aligned myofilaments, with simultaneous increases in expression of key genes involved in EC-coupling [19]. These, include significant increases in the protein expression of cardiac TnI and decreases of the expression of slow skeletal TnI. Measurements of Ca²⁺ transients and cell shortening at field-stimulation (0.2 Hz) in Fura-2 AM loaded hiPSC-CMs were performed as described previously [19]. In brief, Matrigel Mattress was prepared the same day of CM dissociation and seeding. Mattresses were arrayed on a glass coverslip (Corning). Application of 1 μL lines of completely thawed, ice-cold, undiluted growth factor reduced Matrigel were arrayed in parallel. Matrigel was evenly pipetted at a 45° angle with a P10 pipet and corresponding tip. The Mattresses were allowed to incubate 8–10 min at room temperature at which 200 μL of RPMI 1640 medium (Life Technologies) and 2% B-27 supplement (Invitrogen), with the addition of 20,000 hiPSC-CMs were immediately added, halting Mattress polymerization. Each Mattress had a width of approximately 0.85 mm. The thickness of each Mattress was in the range of 0.40–0.88 mm and each line was approximately 23 mm long.

Video edge detection was used to assess cellular shortening of single hiPSC-CMs. Briefly, CMs were visualized using Olympus IX70 coupled to video microscopy system (IonOptix). Field stimulated contraction traces were recorded in 2 mmol/L Ca²⁺ Tyrode solution for 20s. For each cell and each experimental condition typical contraction parameters including percent cell shortening (i.e. percent of resting cell length) and contractile kinetics (i.e. time to peak 90% and time to baseline 90%) values were averaged. Contraction traces were recorded and analyzed using commercially available data analysis software (IonOptix, IonWizardTM Milton, MA).

HiPSC-CMs were incubated with 2 μmol/L Fura-2 AM (Molecular Probes Inc.) for 8 min at room temperature to load the indicator in the cytosol. CMs were washed twice for 10 min with Tyrode's solution containing 250 μmol/L probenecid to retain the indicator in the cytosol. Fura-2 AM-loaded Ca²⁺ transients were recorded during spontaneous beating or 0.2 Hz field stimulation in 2 mmol/L Ca²⁺ Tyrode's solution. For each cell and each experimental condition, tau (τ), amplitude and baseline values were averaged. Ca²⁺ transients were recorded and analyzed using commercially available data analysis software (IonOptix, IonWizardTM Milton, MA). All experiments were conducted at room temperature.

2.9. Data analysis

Data analysis and statistics were performed using Prism version 7.0 (Graphpad Software, Inc., La Jolla, CA) and SPSS version 22.0 (IBM, Armonk, NY). To take into account the repeated sample assessments within patient/donor groups, multilevel analysis was performed. All data were tested for normality using the Shapiro-Wilk Test. Normality was assumed when p > 0.05 and the variances were equal. Significance level was set to p < 0.05. Detailed statistical tests are shown in the figure legends.

3. Results

3.1. Increased myofilament Ca²⁺-sensitivity and altered myofilament protein phosphorylation in TNNT2_mut compared to controls

Force measurements were performed in single cardiomyocytes isolated from IVS and LV tissue from the TNNT2_mut transplanted heart and compared against eight non-failing donor control samples (Fig. 1). Maximal force was lower in both IVS and LV TNNT2_mut cardiomyocytes compared to donor (Fig. 1A), while an increased myofilament Ca²⁺-sensitivity was evident from reduced EC₅₀ values (Fig. 1B, 2.50 ± 0.08, 2.61 ± 0.13 and 3.07 ± 0.08 μM Ca²⁺, in IVS, LV and donor, respectively). Fig. 1C illustrates the leftward shift of the force-Ca²⁺ relation in IVS and LV cardiomyocytes compared to donor cells, indicating that TNNT2_mut cardiomyocytes contract at much lower Ca²⁺ levels than the corresponding controls. As intracellular Ca²⁺ raises from ~0.15 μM in diastole to a peak of ~1.6 μM Ca²⁺ during systole in vivo [20], the higher basal myofilament activation observed (Fig. 1C) is liable to dampen and slow myocardial relaxation.

Fig. 1.

A. Myofilament force measurements at a sarcomere length of 2.2 μm showed lower maximal force (F_max) in cardiomyocytes from IVS and LV TNNT2_mut samples compared to donor LV samples. B. TNNT2_mut cardiomyocytes showed higher myofilament Ca²⁺-sensitivity compared to donor cells evident by the lower EC₅₀ values and the (C) leftward shift of the force-Ca²⁺ relation in IVS and LV compared to donors. Number of cardiomyocytes used amounted to: IVS (n = 22 cells) and LV (n = 19 cells) from TNNT2_mut samples and DONOR (N = 8, n = 24 cells). Data are presented as mean ± SEM. *p < 0.05, vs. DONOR in Multilevel analysis.

The high myofilament Ca²⁺-sensitivity in TNNT2_mut (Fig. 1) may be due to reduced phosphorylation of downstream myofilament targets of the β-adrenergic receptor pathway [21], [22]. We therefore determined phosphorylation levels of myofilament target proteins of PKA (Fig. 2). No differences were observed in the phosphorylation levels of cardiac myosin-binding protein C (cMyBP-C) (Fig. 2B). Cardiac troponin I (cTnI) phosphorylation was reduced in LV and unaltered in IVS compared to donors (Fig. 2C). Moreover, cTnT phosphorylation was lower in both IVS and LV TNNT2_mut tissue (Fig. 2D), while the levels of myosin light chain 2 (MLC-2) phosphorylation were significantly higher in LV and IVS compared to donors (Fig. 2E).

Fig. 2.

A. HCM and donor LV samples were separated by one-dimensional gel electrophoresis and stained with ProQ Diamond phosphostain (right image) and SYPRO Ruby (left image) to correct for loading differences. B. ProQ Diamond analysis revealed no differences in phosphorylation of the protein kinase A (PKA)-target protein myosin binding protein C (cMyBP-C), while (C) the phosphorylation of the other PKA target, troponin I (cTnI), was lower in the LV sample, but not in the IVS sample from TNNT2_mut compared to donor. D. Lower troponin T (cTnT) phosphorylation was detected in both IVS and LV tissue from the TNNT2_mut compared to donor myocardium. E. Moreover, phosphorylation of myosin light chain 2 (MLC-2) was significantly higher in both IVS and LV compared to donor. Phosphorylation signals obtained with Pro-Q Diamond were expressed relative to SYPRO Ruby-stained cMyBP-C bands as to account for differences in sample loading. Values for protein phosphorylation in donor were set to 1 (a.u) relative to total. Abbreviations: IVS, interventricular septum; LV, left ventricle; PM, peppermint marker: two proteins, ovalbumin (45 kDa) and β-casein (24 kDa) are phosphorylated and stained by ProQ Diamond. Data are presented as mean ± SEM. *p < 0.05, vs. DONOR in post-test Bonferroni analysis.

Two-dimensional gel electrophoresis confirmed lower cTnT phosphorylation and higher MLC-2 phosphorylation in TNNT2_mut compared to donor tissue (Fig. 3).

Fig. 3.

Coomassie-stained two-dimensional gel electrophoresis (pH gradient of 4 to 7) was performed in human TNNT2_mut samples from the IVS and LV, and healthy LV samples. Myosin light chain 2 (MLC-2) is composed of two isoforms (U and U*), with both phosphorylated forms detected (P and P*, respectively). Phosphorylation of MLC-2 was higher in TNNT2_mut samples compared to donor. At the level of cTnT three protein spots are separated, which represent unphosphorylated (0P), monophosphorylated (1P) and bisphosphorylated (2P) cTnT. Abbreviations: IVS, interventricular septum; LV, left ventricle. N = 2 for each group.

3.2. High myofilament Ca²⁺-sensitivity is not restored to control levels upon exogenous PKA treatment in TNNT2_mut cardiomyocytes

In our previous studies using human IVS cardiomyocytes from HCM patients the high myofilament Ca²⁺-sensitivity was mostly explained by reduced protein phosphorylation of PKA-targets, including cMyBP-C and cTnI [12], [23], [24]. To determine if the lower cTnI phosphorylation in LV TNNT2_mut tissue (Fig. 2C) could partly explain the observed increases in myofilament Ca²⁺-sensitivity, force measurements were repeated after incubation with exogenous PKA (Fig. 4). Phosphorylation analysis showed that PKA increased phosphorylation of cTnI and cMyBP-C in both TNNT2_mut and donor (Fig. 4A and B), however, this was unable to correct the high myofilament Ca²⁺-sensitivity of IVS and LV TNNT2_mut tissue compared to donors (Fig. 4C and D, 2.75 ± 0.26, 2.82 ± 0.26 and 3.35 ± 0.14 μM Ca²⁺, in IVS, LV and donor, respectively).

Fig. 4.

Measurements performed after treatment with protein kinase A (PKA). A. Cardiac tissue from TNNT2_mut and LV donor samples was treated without (−) or with (+) PKA. Proteins were separated by 1D gel electrophoresis and stained with ProQ Diamond to detect phosphorylation of proteins. B. PKA increased phosphorylation of cMyBP-C and cTnI (donor values without PKA were set to 1; dashed line) in TNNT2_mut and donor tissue. Please note comparison was made against DONOR values (without PKA), which are set to 1. C—D. After PKA treatment, myofilament Ca²⁺-sensitivity was still higher in TNNT2_mut compared to donor. Abbreviations: cMyBP-C, cardiac myosin binding protein C; cTnT, cardiac troponin T; cTnI, cardiac troponin I; MLC-2, myosin light chain 2. Number of cardiomyocytes used amounted to: IVS (n = 22 cells) and LV (n = 19 cells) from TNNT2_mut samples and DONOR (N = 8, n = 24 cells). Data are presented as mean ± SEM. *p < 0.05, vs. DONOR in Multilevel analysis.

3.3. High myofilament Ca²⁺-sensitivity is not restored upon alkaline phosphatase treatment in TNNT2_mut cardiomyocytes

To determine if the low level of cTnT phosphorylation underlies the high myofilament Ca²⁺-sensitivity of TNNT2_mut cardiomyocytes, force measurements were repeated after treatment with alkaline phosphatase (AP), which specifically dephosphorylates cTnT (Fig. 5). After AP treatment, cTnT phosphorylation of donor cardiomyocytes substantially decreased to values present in TNNT2_mut tissue (Fig. 5A to D). However, after AP treatment the high myofilament Ca²⁺-sensitivity in IVS and LV TNNT2_mut samples remained higher compared to donor (Fig. 5E and F, 2.09 ± 0.09, 2.0 ± 0.17 and 2.70 ± 0.0.13 μM Ca²⁺, in IVS, LV and DONOR, respectively).

Fig. 5.

Measurements performed after treatment with alkaline phosphatase (AP). A. Cardiac tissue from TNNT2_mut and LV donor samples were treated without (−) or with (+) AP. Proteins were separated by 1D gel electrophoresis and stained with ProQ Diamond to detect phosphorylation of proteins. B. AP decreased cardiac troponin T (cTnT) phosphorylation after which the phosphorylation pattern of cTnT was similar in TNNT2_mut and donor myocardium. C. Coomassie-stained two-dimensional gel electrophoresis (pH gradient of 4 to 7) was performed in human TNNT2_mut samples from the IVS and non-healthy donor biopsies treated without (−) or with (+) AP. Phosphorylation of cTnT reveals three protein spots, which represent unphosphorylated (0P), monophosphorylated (1P) and bisphosphorylated (2P) cTnT. D. AP similarly decreased cTnT phosphorylation in donor samples to TNNT2_mut. E-F. AP significantly increased myofilament Ca²⁺-sensitivity in both groups and did not abolish the difference between TNNT2_mut and donor cardiomyocytes (E). Number of cardiomyocytes used amounted to: IVS (n = 5 cells) and LV (n = 5 cells) from TNNT2_mut samples and DONOR (N = 4, n = 12 cells). Data are presented as mean ± SEM. *p < 0.05, vs. DONOR in Multilevel analysis.

3.4. A relatively low amount of mutant cTnT-K280N increases myofilament Ca²⁺-sensitivity

We performed cardiac troponin exchange experiments to further establish that the elevated myofilament Ca²⁺-sensitivity of TNNT2_mut cardiomyocytes is not attributed to specific alterations in the phosphorylation state of myofilament proteins, but rather is due to a direct mutation effect. Because the homozygous TNNT2_mut results in 100% mutant cTnT it provides a unique tool to assess the amount at which mutant protein disturbs the myofilament contractile state. Exchange with increasing concentrations of wild-type (WT) human troponin complex (0.25, 0.5 and 1 mg/mL in the exchange solution) resulted in 62 ± 2%, 78 ± 1% and 86 ± 1% troponin exchange based on Western blot analyses of endogenous and myc-tag labeled wild-type cTnT (Fig. 6A, picture adapted from our previous work [12]). In other words, exchange of endogenous mutant troponin in TNNT2_mut cardiomyocytes with increasing concentrations of recombinant wild-type troponin resulted in increasing efficiency of exchange up to 86% (14% of mutant cTnT) (Fig. 6B). Exchange with (unphosphorylated) recombinant wild-type troponin significantly increased Ca²⁺-sensitivity compared to TNNT2_mut cells (100% cTnT-K280N) evident from lower EC₅₀ values in wild-type troponin exchanged cells compared to TNNT2_mut (Fig. 6B). Similarly, increased myofilament Ca²⁺-sensitivity was observed in donor cardiomyocytes exchanged with 1 mg/mL unphosphorylated recombinant wild-type troponin (83 ± 3% exchange) when compared to non-exchanged donor cells (Fig. 6B). As would be expected, exchange of endogenous WT (in donor cells) and mutant cTnT (in TNNT2 cells) by unphosphorylated recombinant troponin complex significantly enhanced responsiveness to exogenous PKA (Fig. 6C). PKA can phosphorylate recombinant cTnI to values observed in donor samples. After PKA, Ca²⁺-sensitivity of the myofilaments was restored in donor cardiomyocytes exchanged with 83% wild-type troponin, while it remained significantly higher in all troponin-exchanged cells compared to donor cells, indicating that 14% of mutant K280N is sufficient to increase Ca²⁺-sensitivity (Fig. 6D). To validate whether the relatively high MLC-2 phosphorylation (Fig. 2E) could have interfered with the high myofilament Ca²⁺ sensitivity of TNNT2_mut cardiomyocytes, donor cells were exchanged with recombinant cTnT-K280N mutant. Exchange (45% cTnT-K280N) with unphosphorylated recombinant cTnT-K280N complex increased Ca²⁺-sensitivity in donor cells (Fig. 6B), and this coincided with the expected large response to PKA (Fig. 6C). After PKA treatment, which normalizes cTnI phosphorylation to donor values, Ca²⁺-sensitivity remained high in cTnT-K280N exchanged donor cells compared to donor controls without troponin replacement (Fig. 6D). This experiment provides validation that the cTnT-K280N mutation by itself increases the sensitivity of myofilaments to Ca²⁺, irrespective of myofilament protein phosphorylation differences (Fig. 2). However, we cannot entirely rule out that other post-translational modifications, including glutathionylation, methylation, oxidation, etc., could also affect the interpretation of our findings, but the investigation of those are outside the scope of the present work.

Fig. 6.

Troponin exchange experiments were performed in demembranated single cardiomyocytes to assess the effect of mutant troponin T (cTnT) on Ca²⁺ sensitivity. Mutant cTnT in IVS TNNT2_mut cells was replaced by exogenous recombinant non-mutant cTnT, while in LV healthy cells endogenous cTnT was replaced by recombinant mutant cTnT (K280N) and wild-type cTnT (WT). Using different concentrations of cTnT complex in the exchange solution we replaced various amounts of mutant/control troponin with recombinant cTn complex. A. Quantification of troponin exchange in cardiomyocytes from the TNNT2_mut heart. Immunoblots stained with an antibody against cardiac troponin T (cTnT) that recognizes both endogenous cTnT (lower band) and recombinant myc-tag labeled cTnT (cTnT-myc; upper band). An example is shown of a suspension of cardiomyocytes from a TNNT2_mut heart exchanged with increasing concentrations of wild-type human recombinant troponin complex. Exchange with 0.25 mg/mL (lane 1), 0.5 mg/mL (lane 2) and 1 mg/mL (lane 3) troponin complex. TNNT2_mut heart without added recombinant troponin complex (lane 4). DONOR hearts were exchanged with 1 mg/mL wild-type (WT) and mutant (K280N) human recombinant troponin complex (not shown). Similar amounts were loaded on the blots (shown by Ponceau-stained actin) to allow cTnT analysis within the linear detection range. Please note this picture is adapted from our previous work [12] B. After exchange with the different cTn complexes, myofilament Ca²⁺-sensitivity increased in all cTnT-exchanged cells compared to baseline (both TNNT2_mut [100% mutant] and Donors). This may be explained by replacement of phosphorylated endogenous cTnI by unphosphorylated recombinant cTn complex (Please note that samples are compared vs 100% TNNT2_mut). C. Indeed, the response to PKA increased with increased percentage of troponin exchange and was highest in cells in which 83% of WT cTnT and 86% of mutant cTnT was replaced by recombinant cTn complex (Please note that samples are compared vs 100% TNNT2_mut). D. To normalize cTnI phosphorylation in all cells, measurements were repeated after treatment with PKA revealing that myofilament Ca²⁺-sensitivity was significantly higher compared to donor in all cells with low (14%) and high (100%) levels of mutant cTnT (Please note samples are compared vs DONOR). 100% IVS TNNT2_mut: n = 22 cells; 38% IVS TNNT2_mut: n = 10 cells; 22% IVS TNNT2_mut: n = 10 cells; 14% IVS TNNT2_mut: n = 10 cells; DONOR: N = 8, n = 24 cells; DONOR (45% K280N): n = 12 cells; DONOR (83% WT): n = 10 cells; *p < 0.05 vs 100% TNNT2_mut in in Multilevel analysis; Data are presented as mean ± SEM. #p < 0.05 vs DONOR in Multilevel analysis.

3.5. HiPSC-CMs expression of the human cTnT-K280N mutation

We subsequently investigated the relationship between contractility and Ca²⁺ handling in hiPSC cardiomyocytes where we gene-inserted the K280N mutation using CRISPR/Cas9 technology. Two hiPSC lines were generated: (1) One with both allele-insertions (homozygous); and (2) A second with a single allele-insertion (heterozygous) (Supplemental Fig. 1). Representative Ca²⁺ transients and cellular shortening are depicted in Fig. 7A. Cellular diastolic length was unaltered in both homozygous and heterozygous hiPSC-CMs-K280N lines (Fig. 7B), while contractility (i.e., myocyte shortening) was significantly increased in both hiPSC-CMs-K280N lines compared to isogenic controls (Fig. 7C). The increased contractility was not accompanied by changes in contraction time (Fig. 7D; time to peak cell shortening). Cardiomyocyte relaxation was impaired in the homozygous hiPSC-CMs-K280N line evident from the significantly increased time to baseline compared to isogenic controls (Fig. 7E). Heterozygous and homozygous hiPSC-CMs-K280N lines showed similarly increased diastolic Ca²⁺ levels (Fig. 8A) and slower Ca²⁺ decline rates (Fig. 8B). No alterations were observed in the Ca²⁺ transient amplitude or time to peak Ca²⁺ (Fig. 8C and D).

Fig. 7.

Measurement of Ca²⁺ transient and cell shortening of Fura2-AM-loaded control and HCM (TnT K280N mutations; heterozygous and homozygous) hiPSC-CMs. A. Representative traces of Ca transients and cell shortening in response to field-stimulation at 0.2 Hz. B—C. Diastolic cell lengths did not differ, whereas contractility was significantly increased in both hetero- and homo-K280N group. D-E. Cell shortening kinetics reveals that contraction time was not different in all groups. However, relaxation time (time to baseline 50, 90%) was significantly increased in homo-K280N iPSC-CMs. Data are presented as mean ± SDs. N = 17–45 cells; * p < 0.05,*** p < 0.001 vs control in post-test Bonferroni analysis.

Fig. 8.

Analysis of Ca²⁺ transients parameters obtained from control and HCM (TnT K280N mutations; Heterozygous and Homozygous) hiPSC-CMs. Stimulation frequency was set at 0.2 Hz. Diastolic Ca (A) and decay rate (B) were statistically increased in both hetero- and homo-K280N group. However, amplitude (C) and peak time (D) were not different between all groups. Data are presented as mean ± SDs. N = 17–45 cells; ** p < 0.01, *** p < 0.001 vs control in post-test Bonferroni analysis.

4. Discussion

4.1. Toxicity effects with the cTnT-K280N mutation

The human heart with a homozygous TNNT2 mutation enabled us to titrate the mutation-induced effects on sarcomere function by combining troponin exchange experiments in single cardiomyocytes with protein (de)phosphorylation assays. We provide evidence that the TNNT2 K280N mutation is by itself sufficient to increase myofilament Ca²⁺-sensitivity in human cardiomyocytes at a relatively low expression (about 14%) (Fig. 6D). Recently, we similarly found for other HCM-causing cTnT mutations (cTnT I79N and R94C) that low expression levels of the mutant protein are sufficient for the maximal mutation-mediated increase in myofilament Ca²⁺-sensitivity [25].

Differing penetrance effects of the cTnT-K280N mutation to cellular function is additionally strengthened by our studies in hiPSC-CMs lines. Heterozygous hiPSC-CMs-K280N showed comparable effects on Ca²⁺ transients and contractility to homozygous TNNT2 K280N cell lines, while depressed relaxation time (i.e., impaired cardiomyocyte relaxation (Fig. 7E)) was only observed in homozygous cells where full expression of the cTnT-K280N mutant occurs. Notably, however, it illustrates that heterozygous protein expression of the cTnT-K280N mutant protein, as observed in heterozygous HCM carriers, is sufficient to increase myofilament Ca²⁺-sensitivity and coincides with increased diastolic Ca²⁺ levels and increased contractility.

4.2. Phosphorylation profile of myofilament proteins cannot account for the elevated myofilament Ca²⁺-sensitivity of the cTnT-K280N mutation

Although phosphorylation of the archetypical target proteins downstream of the β-adrenergic receptors (i.e., cMyBP-C and cTnI) was largely similar in the TNNT2_mut and donor myocardium (Fig. 2A and B), phosphorylation of cTnT and MLC-2 differed from non-failing donor hearts suggesting secondary disease-related changes in protein phosphorylation during HCM development (Fig. 2D and E). The low cTnT phosphorylation may be the direct consequence of the TNNT2 mutation, but it was inadequate to alter myofilament Ca²⁺-sensitivity (Fig. 5). PKA-mediated phosphorylation of cTnI is known to decrease myofilament Ca²⁺-sensitivity, while phosphorylation of MLC2 increases myofilament Ca²⁺-sensitivity [26]. Indeed, treatment with exogenous PKA decreased sensitivity to Ca²⁺ in IVS and LV TNNT2_mut and donor cells from 2.50 ± 0.08, 2.61 ± 0.13 and 3.07 ± 0.08 μM Ca²⁺ to 2.75 ± 0.26, 2.82 ± 0.26 and 3.35 ± 0.14 μM Ca²⁺, respectively. However, PKA-mediated cTnI phosphorylation did not normalize myofilament Ca²⁺-sensitivity in TNNT2_mut cells to values observed in donor cells.

In a similar way, upon AP incubation cardiomyocytes desensitized to 2.09 ± 0.09, 2.0 ± 0.17 and 2.7 ± 0.13 μM Ca²⁺, in IVS, LV and donor, respectively. We also investigated whether MLC2 changes could account for the overall increases of Ca²⁺-sensitivity in the TNNT2 mutant samples by exchanging healthy non-failing donor cardiomyocytes with reconstituted cTnT-K280N (Fig. 6). Here, we also observed that the mutation itself (45% of exchange recombinant mutant protein cTnT-K280N), irrespective of phosphorylation, is sufficient to account for the elevated myofilament Ca²⁺-sensitivity.

4.3. Allosteric alterations to myofilaments mediated by the cTnT-K280N mutation

It is intriguing to consider how selective, but specific, amino acid alterations to cTnT could exert dramatic alterations to myofilament function and Ca²⁺ handling even at a very low mutant level. Several reconstruction models of thin-filament regulation agree that this region is particularly important to dock the troponin-tropomyosin complex onto the outer domain of actin at low cytoplasmic Ca²⁺, stabilizing the formation of the steric blockade, which is needed for normal cardiomyocyte relaxation (Supplemental Fig. 2A, red dashed circle) [27], [28]. In particular, the C-terminal region of cTnT is crucial for thin-filament inactivation [29], which is largely due to the presence of basic residues at the location [30]. Replacement experiments of all Lysines (K) at the very end of cTnT by the neutral amino acid alanine (A) show it to be sufficient to disrupt the thin-filament steric blockade [30]. Likewise, the cTnT-K280N mutation in our study replaces the 280 Lysine (K) residue at the very end of cTnT by the neutral amino acid Asparagine (N), and one can speculate a similar inhibition of the thin-filament blockade. A recently proposed mechanism considers that the C-terminal region of cTnT interacts with the N-domain of cardiac troponin C (cTnC) and modulates the interaction between cTnC and the cTnI switch peptide [31]. Along the lines, the cTnT-K280N mutation at the C-terminal segment end putatively lowers its affinity towards the cTnC-cTnI interaction and consequently disrupts the thin-filament inactive state, while increasing myofilament Ca²⁺-sensitivity. This is further supported by: (1) Solution studies with reconstituted HCM mutant troponin-tropomyosin proteins (in cTnT [32], [33], but also cTnI [34]) that show a disrupted steric-blockade by troponin mutations; and (2) Evidence that the cTnT-K280N indeed disrupts the actin-myosin blockade in the absence of Ca²⁺ [24]. By activating human cardiomyopathy muscle in ADP-containing solutions without Ca²⁺ we revealed that the cTnT-K280N promotes greater basal sarcomeric activation capable of impairing diastolic performance [24]. High basal activation was corrected by incorporation of wild-type troponin up to 85% [24]. These observations support that the negative effects of the cTnT-K280N are not restricted to Ca²⁺ alterations, but are additionally Ca²⁺ independent.

Notably, the C-terminal segment end of cTnT, in which the K280N mutation is located, is highly conserved in TnT gene isoforms and across species with relatively few myopathy mutations identified at the C-terminus [35], [36]. These strengthen the concept that this region of cTnT has a central role in sarcomere function/structure and mutations to the region are likely lethal and therefore will manifest rarely in the general population [35]. In agreement, ~86% of cTnI HCM-causing mutations are identified close to this contact region of cTnT (Supplemental Fig. 2A, red dashed circle) [37], [38], [39], supporting that the allosteric alterations caused by a single troponin mutation are likely propagated along the other two subunits and imped normal formation of the thin-filament inactive state.

4.4. Proposed model for propagation of mutant-effect along myofilaments

Equally fascinating we find here is the large degree of myofilament Ca²⁺-independent allosteric activation conferred by a very low mutant protein level (Fig. 6D). The high level of myofilament activation can potentially be explained by a large propagation of mutant-induced effects from thick-to-thin filaments (Supplemental Fig. 2B to D). In other words, the cTnT-K280N promotes large geometrical alterations onto the thin-filament via myosin thick-filament activation – irrespective of Ca²⁺, and may thereby exacerbate activation at low levels of Ca²⁺.

The thin-filaments are composed of individual functional units comprising of seven actin monomers spanned by one tropomyosin dimer and one troponin complex (A₇TmTn; Supplemental Fig. 2B toD). Upon cytosolic Ca²⁺ elevation, Ca²⁺-binding to cTnC promotes cTnI detachment from actin and potentiates tropomyosin movement with exposure of myosin-binding sites on the surface of F-actin (i.e. actin-myosin steric blockade at low Ca²⁺). This tropomyosin movement allows Ca²⁺-cooperative activation of the thin-filament with additional recruitment of strong-binding myosin (Supplemental Fig. 2C). Structural data suggest that, under normal operating conditions, individual strongly-bound cross-bridges bind to the regulatory unit (A₇TmTn spanning ~38.5 nm), and in a Ca²⁺-dependent manner, regulate tropomyosin movement up to ~3 units (covering ~115 nm) along the thin-filament [40]. These were validated in biochemical studies [41], [42]. Having in mind that ~26 A₇TmTn units occupy each half-sarcomere [24], and having as low as 14% endogenous mutant cTnT-K280N myofilament incorporation (Fig. 6D), one can estimate that ~4 cTnT-K280N A₇TmTn units are present per half-sarcomere (26 A₇TmTn × 0.14 cTnT-K280N; assuming homogenous distribution). This would imply that a mutant cTnT-K280N can promote substantial geometrical/allosteric alterations on the thin-filament at very low Ca²⁺ levels and propagate, upon the rise of intracellular Ca²⁺, activating effects via tropomyosin over a distance of up to 6.5 A₇TmTn units (26 A₇TmTn / 4 A₇TmTn K280N units). Although the latter assumption may be an inaccurate sarcomere representation, it does however provide guidance that the cTnT-K280N can allosterically/physically propagate alterations of the cTn-tropomyosin complex over far greater lengths than normal wild-type complexes at extremely low Ca²⁺ levels. Such propagating effects of the C-terminal region of cTnT under disease conditions, with residue alterations that cause much larger sarcomere activation changes, may underlie the observation that relatively few myopathies are identified to this region of cTnT.

4.5. Reduced force generation in the cTnT-K280N mutation

We previously showed that the cTnT-K280N sample has significantly lower myofibril density [43], which can readily explain the reduced maximal force generation in TNNT2 cardiomyocytes (Fig. 1A). In terms of cross-bridge modelling kinetics, a simple two-state model [44], states that the transitory changes of cross-bridges from non-force-bearing to force-bearing states is largely limited by slow detachment rates (g_app). An increase in cross-bridge detachment rate would reduce force generation capacity. Recent myofibril kinetics in the cTnT-K280N biopsy showed an increase in the slow detachment rate, but this coincided with an increase in activation kinetics (f_app), such that no alterations of maximal active force were detected in single myofibrils [13]. The latter supports that single myofibrils of the cTnT-K280N sample have similar maximal force generating capacity, but cardiac remodelling, i.e., reduced myofibril content in hypertrophied cardiomyocytes, rather than the mutation itself is responsible for the reductions in maximal force development at cardiomyocyte level.

4.6. Study limitations and conclusions

Our study was performed in cardiomyocytes from only a single patient with a homozygous TNNT2 mutation providing technical rather than biological replicates. As mutations in the TNNT2 gene are not frequent (frequency ~ 5.9%) (Harvard-Medical-School-Genetic-database), and homozygous patients are extremely rare, there was no possibility to include additional human samples with the TNNT2 K280N mutation. However, the homozygous mutation provides a unique model system to study mutation toxicity effects using the troponin exchange. To provide additional mechanistic data on the effect of the cTnT-K280N mutation, we complemented the ‘rescue’ experiments in the single cTnT-K280N patient sample with exchange of recombinant cTnT-K280N into non-failing donor samples. More, we generated two human-induced pluripotent stem cells (hiPSC) lines, where we co-inserted the cTnT-K280N mutation using CRISPR/Cas9. Analysis by nano-liquid chromatography mass spectrometry was unable to successfully quantify the total mutant cTnT-K280N expression in heterozygous K280N hiPSC-CMs, which therefore limits our interpretation of the lower threshold required for a mutation to be toxic in heterozygous cell lines. Also, the magnitude of the toxic threshold effect may vary from mutation to mutation, depending on mutant location and/or secondary effects, including protein phosphorylation. More, because hiPSC-CMs have much slower Ca²⁺ decay rate than human adult [45], we can only confidently measure the diastolic Ca²⁺ levels in hiPSC-CMs when paced at 0.2 Hz. This was nonetheless sufficient for the TnT-K280N mutation causing abnormal Ca²⁺ handling in hiPSC-CMs during 0.2 Hz stimulation. It is likely that our experimental setup underestimates the magnitude of some of those alterations.

We are however confident that the directionality and impact of the negative effects on heart function are potentially shared across all TNNT2 mutations identified in human HCM. Mutations in troponin subunits, specifically those in cTnT, have the dual effect of affecting Ca²⁺-independent (i.e., by decreasing steric-blockade at low Ca²⁺) and Ca²⁺-dependent activation (cooperative activation of thin-filament via cTnC), and thereby increase basal myofilament activation and increase cardiomyocyte contractility.

CRediT authorship contribution statement

The objective of our study was to demonstrate that low levels of the cTnT-K280N mutation are sufficient to replicate the main features of hypertrophic cardiomyopathy. All authors approved the final version of the manuscript. Conception and design of the experiments: V.S. devised, designed and performed the project, analyzed data and wrote the manuscript; L.W. performed experiments, analyzed data and revised the manuscript; P.J.M.W. performed experiments, analyzed data and revised the manuscript; K.K. performed experiments, analyzed data and revised the manuscript; J.R.P. performed experiments, analyzed data and revised the manuscript; C.d.R. collected the human cardiac samples, wrote and revised the manuscript; C.R. provided recombinant cTn complexes, wrote and revised the manuscript; B.C.K. supervised the iPSC experiments and revised the manuscript; J.v.d.V. supervised the project, wrote and revised the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary figures