ERRγ enhances cardiac maturation with T-tubule formation in human iPSC-derived cardiomyocytes

Abstract

One of the earliest maturation steps in cardiomyocytes (CMs) is the sarcomere protein isoform switch between TNNI1 and TNNI3 (fetal and neonatal/adult troponin I). Here, we generate human induced pluripotent stem cells (hiPSCs) carrying a TNNI1^EmGFP and TNNI3^mCherry double reporter to monitor and isolate mature sub-populations during cardiac differentiation. Extensive drug screening identifies two compounds, an estrogen-related receptor gamma (ERRγ) agonist and an S-phase kinase-associated protein 2 inhibitor, that enhances cardiac maturation and a significant change to TNNI3 expression. Expression, morphological, functional, and molecular analyses indicate that hiPSC-CMs treated with the ERRγ agonist show a larger cell size, longer sarcomere length, the presence of transverse tubules, and enhanced metabolic function and contractile and electrical properties. Here, we show that ERRγ-treated hiPSC-CMs have a mature cellular property consistent with neonatal CMs and are useful for disease modeling and regenerative medicine.

Fig. 1. Generation of the hiPSC TNNI1^EmGFP and TNNI3^mCherry dual reporter.

a Schematic for the construction of the reporter cell line and the homologous recombination strategy using the CRISPR/Cas9 system. FAM (6-carboxyfluorescein) and VIC (VIC™ fluorescent dye) are fluorescent probes for ddPCR. b Schematic representation of the TNNI1^EmGFP and TNNI3^mCherry dual reporter during cardiac differentiation. c ddPCR (FAM probe, blue bars; VIC probe, green bars) of TNNI1^EmGFP knock-in clones before and after treatment with piggyBac transposase. d Phase-contrast and EmGFP fluorescence images of hiPSCs (#4-12-20) and day-14 EBs. Scale bars: 500 μm. e Flow cytometric analysis and sorting on day 14 of the differentiation for EmGFP⁺ cells (top) and the EmGFP⁺-derived cTNT⁺ cell fraction (bottom). f Karyotype of the generated TNNI1-EmGFP knockin clone #4-12-20.

Fig. 2. Molecular screening for TNNI3^mCherry expression.

a Schematic representation of the cardiac differentiation protocol for cardiac maturation monitoring. b Bright field (left) and EmGFP fluorescence (right) images of day-14-derived EBs. Scale bars: 500 μm. c Flow cytometric analysis of day-14-derived EBs for TNNI1^EmGFP. d Compounds were selected by the median (MED) plus 3 MAD. e Screening outline of the chemical compounds. f High content imaging for mCherry dose-dependency intensity. n = 2 biologically independent samples per group. g Dose-dependent T112-derived fluorescence for TNNI1^EmGFP and TNNI3^mCherry. Scale bars: 100 μm. h Dose-dependent cell number per well after treatment. n.d. not detected. n = 2 biologically independent samples per group. i Dose-dependent T623-derived fluorescence for TNNI1^EmGFP and TNNI3^mCherry. Scale bars: 100  μm.

Fig. 3. T112 and T623 enhanced the cardiac maturation of hiPSC-CMs.

a Fluorescence images of day-16-derived EBs for TNNI1^EmGFP and TNNI3^mCherry expression. Scale bars: 500 μm. NT no treatment, PC positive control. b Flow cytometric analysis of day-16-derived cells for TNNI3^mCherry expression. n = 3 independent experiments per group. Data are the mean ± SEM; ****P < 0.0001 compared to DMSO using one-way ANOVA followed by Dunnett’s test. NT no treatment, PC positive control. c Principal-component analysis (PCA) in correlation with the maturation of CMs for more than 58,000 genes expressed in the four groups. Data are shown as average PCA for each group. n = 3 independent experiments per group. d Sample correlation matrix resulting from the consensus clustering analysis of mRNA data and showing transcriptomic differences among treatments. Data are shown as averages for each group. n = 3 independent experiments per group. e Volcano plots for DE genes in accordance with the cardiac maturation phenotype. f Summary of upregulated DE genes in either T623/T112-treated CMs or Combo-treated CMs compared to DMSO. The number of upregulated genes (at least 2-fold) is depicted. Left: 344 unique elements detected in the T623-treated CMs compared to DMSO. Middle: 947 unique elements detected in the T112-treated CMs compared to DMSO. Right: 1254 unique elements detected in the Combo-treated CMs compared to DMSO. g Analysis of the top-6 cardiac-related significant GO terms for the cellular components of T623-, T112- and Combo-treated CMs.

Fig. 4. Transcriptomic pathways and gene sets enrichment.

a Reactome pathway analysis for T623-, T112- and Combo-treated CMs. Enriched pathways were calculated using the binomial test over DE genes between individual conditions compared to DMSO. The top 15 statistically significant pathways are listed. P-values are displayed as logarithmic form (log). b Gene set analysis of cardiac maturation-specific genes.

Fig. 5. Metabolic and structural functionality of hiPSC-CMs.

a Mitochondrial respiration rates of hiPSC-CMs treated with DMSO (n = 4), T112 (n = 5), T623 (n = 5), or Combo (n = 5) (O, Oligomycin; F, FCCP, phenylhydrazone; AR, Antimycin A and Rotenone). n = biologically independent samples per group. Data are the mean ± SEM. b Basal and maximal OCR and ATP production of hiPSC-CMs treated with DMSO (n = 4), T112 (n = 5), T623 (n = 5), or Combo (n = 5). n = biologically independent samples per group. Data are the mean ± SEM. Statistical analysis was compared to DMSO using one-way ANOVA followed by Dunnett’s test. ****P < 0.0001. c top: Representative western blots of cTnI, ssTnI, and GAPDH proteins in hiPSC-CMs treated with DMSO, T112, T623, or Combo. For ssTnI, two bands were detected: the shorter one is wild type ssTnI and the longer one is ssTnI with F2A peptides. Bottom: cTnI/ssTnI relative ratio. n = 3 independent experiments per group. Data are the mean ± SEM. Statistical analysis was compared to DMSO using one-way ANOVA followed by Dunnett’s test. d Cell area of hiPSC-CMs treated with DMSO, T112, or T623 measured by high content imaging. For each group, n = 300 cells over three independent experiments. e Roundness of hiPSC-CMs treated with DMSO, T112, T623, or Combo. Each group, n = 3 independent experiments. Data are the mean ± SEM. Statistical analysis was done using one-way ANOVA followed by Tukey’s HSD test. f Sarcomere length in hiPSC-CMs treated with DMSO, T112, T623, or Combo. Boxes represent 25th–75th percentiles; whiskers represent the minimum and maximum ranges; horizontal lines indicate the median values. n = 15, one measurement per TEM image from 15 TEM images over two consecutive experiments; ****P < 0.0001 compared to DMSO, ‡P < 0.0001 compared to T623 using one-way ANOVA followed by Tukey’s HSD test. g Contractile properties of hiPSC-CMs treated with DMSO (n = 6), T112 (n = 7), T623 (n = 5), or Combo (n = 9). Top: Contraction velocity. Middle: Relaxation velocity. n = biologically independent samples examined over three independent experiments. bottom: Maximum conduction velocity. n = 3 independent experiments. Statistical analysis was compared to DMSO using one-way ANOVA followed by Dunnett’s test. ****P < 0.0001. h Representative AP recorded from hiPSC-CMs treated with DMSO or T112. i The fraction of APs with a notch (black). j Electrophysiology data of APA, Vmax and MDP in hiPSC-CMs treated with DMSO (n = 21) or T112 (n = 18) over 3 independent experiments. Boxes represent 25th–75th percentiles; whiskers represent the minimum and maximum ranges; horizontal lines indicate the median values. Statistical analysis was done using unpaired two-tailed t-test. k Typical high magnification TEM images of hiPSC-CMs treated with T112 or Combo. Scale bars: 500 nm. I: I-band; TT: T-tubule; Z: Z line.All results are from #8-5-3.

Fig. 6. Stage-specific causal role of ERRγ in regulating TNNI3-driven cardiac maturation.

a Fluorescence images of ERRγKO hiPSC-CMs treated with DMSO, T112, or T623 for TNNI1^EmGFP and TNNI3^mCherry expression. Scale bars: 500 μm. b Flow cytometric analysis of day-16 CMs derived from ERRγKO hiPSCs for TNNI3^mCherry expression. n = 4 independent experiments per group. Data are the mean ± SEM; Statistical analysis was compared to DMSO using one-way ANOVA followed by Dunnett’s test. NT no treatment. c Mitochondrial respiration rates of ERRγKO hiPSC-CMs treated with either DMSO or T112 (O Oligomycin; F FCCP, phenylhydrazone; AR Antimycin A and Rotenone). n = 5 biologically independent samples per group. d Basal and maximal OCR and ATP production of ERRγKO hiPSC-CMs treated with either DMSO or T112. Statistical analysis was done using unpaired two-tailed t-tests. n = 5 biologically independent samples per group. Data are the mean ± SEM; Statistical analysis was using unpaired two-tailed t-test.

Fig. 7. hiPSC cardiac maturation model.

Schematic representation of an ERRγ agonist inducing the cardiac neonatal-like phenotype.