Abnormal progenitor cell differentiation and cardiomyocyte proliferation in hypoplastic right heart syndrome

Pulmonary atresia with intact ventricular septum (PA-IVS) is a rare (4–8 per 100,000 live births) type of hypoplastic right heart syndrome (HRHS) where the right-sided structures in the heart are malformed ¹. In PA-IVS, the pulmonary valve does not open, resulting in no connection between the right ventricle (RV) and pulmonary arteries. After surgical or catheter-based intervention, PA-IVS patients have clinical outcomes that range from single-ventricle palliation (PA-IVS-1v) to a biventricular repair (PA-IVS-2v). Mechanisms underlying the spectrum of RV hypoplasia in PA-IVS are difficult to fully ascribe to an atretic pulmonary valve. In this study, we leverage PA-IVS patient-specific induced pluripotent stem cells (iPSCs) and single-cell RNA-sequencing (scRNA-seq) to elucidate etiologies of ventricular hypoplasia in PA-IVS.

Under an IRB-approved protocol at Nationwide Children’s Hospital and informed consent, we recruited individuals with PA-IVS-1v (n=3) and PA-IVS-2v (n=3) along with those without congenital heart disease (controls, n=3). Clinical outcomes of PA-IVS patients were confirmed by echocardiogram and/or review of medical records. Characterization of human iPSC lines including three germ layer differentiation, karyotyping, and cell line authentication were performed ². Human iPSC-derived cardiomyocytes (iPSC-CMs) were generated by sequential modulation of Wnt signaling ³. Immunofluorescence staining indicated the intercalated distribution of TNNT2 and α-actinin in the sarcomere structures in PA-IVS iPSC-CMs (Figure A).

Figure. Impaired cardiac differentiation and proliferation along with abnormal metabolic profiles in iPSC-derived cardiomyocytes (iPSC-CMs) from subjects with PA-IVS.

A, Sarcomere structures in PA-IVS and control D30 iPSC-CMs are illustrated by immunofluorescence staining using antibodies against cardiac troponin T (TNNT2, green) and α-actinin (red). Nuclei are stained with DAPI (blue). Scale bars: 100 μm. B, Quantitative analysis of percentages of proliferating cardiomyocytes (pHH3+ TNNT2+) in the presence or absence of CHIR99021 (One-way ANOVA with post hoc Tukey’s test for multiple comparison; n=10). C, Immunofluorescence staining using antibodies against TNNT2 (red) and pHH3 (green) to identify proliferating cardiomyocytes (shown by arrows). Scale bars: 100 μm. D, Heatmap shows 1,092 differentially expressed genes (DEGs) between control and PA-IVS-1v D13 iPSC-CMs (fold change >2, FDR <0.05). For each gene, expression values (log2 transformed values of fold change) were normalized to the median value of the controls. Red indicates upregulation, whereas blue indicates downregulation compared to the median control value. E, Gene Set Enrichment Analysis (GSEA) identifies downregulated pathways in PA-IVS-1v compared to control D13 iPSC-CMs. F, Immunofluorescence staining of PA-IVS and control D20 iPSC-CMs using antibodies against TNNT2 (green) and cell proliferation marker Ki67 (red) under static and cyclic stretch. Nuclei are stained with DAPI (blue). Arrows indicate representative proliferating cardiomyocytes. Scale bars: 50 μm. G, Quantitative analysis of percentages of proliferating cardiomyocytes under static and cyclic stretch (One-way ANOVA with post hoc Tukey’s test for multiple comparison; n=4). H, Schematic diagram illustrating time-course sample collections during cardiac differentiation of PA-IVS-1v and control iPSCs for scRNA-seq. I-J, UMAP plots show the topological structures of D5, D10, D14, and D30 cell populations visualized by samples (I) and cell types (J). K, UMAP plots of D10 differentiating cells which include epicardial progenitors (Epi), first heart field (FHF) progenitors, second heart field (SHF) progenitors, vascular endothelial cells, and early cardiomyocytes. Red circles highlight populations of SHF progenitors. L, Percentages of each progenitor population in PA-IVS-1v and control at D10 of differentiation. M, Violin plots show pathway enrichment scores (y-axis) in D14 iPSC-CMs between PA-IVS-1v and controls revealed by scRNA-seq. Each pair of violin plots represents a comparison between control-D14 and PA-IVS-1v-D14 samples. Each dot corresponds to a single cell and depicts pathway enrichment scores (y-axis) calculated by single sample GSEA; p-values indicate significance by Wilcoxon tests between sample groups. N, Seahorse analysis shows normalized oxygen consumption rate (OCR) measurements of basal and maximal respiration in control, PA-IVS-1v, and PA-IVS-2v D30 iPSC-CMs. O-P, Basal respiration (O) and ATP production (P) are elevated in PA-IVS-1v and PA-IVS-2v versus control iPSC-CMs measured by Seahorse assays (One-way ANOVA with post hoc Tukey’s test for multiple comparison; n=34 for control, n=84 for PA-IVS-1v, and n=41 for PA-IVS-2v). All data are shown as mean ± SEM. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Because ventricular hypoplasia is a major clinical manifestation in PA-IVS, we postulate that cardiomyocyte proliferation is impaired during embryonic heart development. Human iPSC-CMs resemble embryonic and fetal stage cardiomyocytes, and thus could potentially recapitulate cardiac proliferation defects in PA-IVS. We generated day (D)13 iPSC-CMs and found a lower percentage of proliferating cardiomyocytes (pHH3+ TNNT2+) in PA-IVS compared to controls (Figures B–C). In the presence of CHIR99021, a GSK3 inhibitor that promotes robust proliferation of human iPSC-CMs ⁴, both PA-IVS-1v and PA-IVS-2v iPSC-CMs were less proliferative than controls, suggesting cardiomyocyte proliferation defects in PA-IVS. Using bulk RNA-seq, we identified 1,092 differentially expressed genes (DEGs) between PA-IVS-1v and control D13 iPSC-CMs (FDR<0.05 and fold change >2) (Figure D). Top enriched pathways associated with DEGs were cell cycle progression, DNA replication, and mitosis (Figure E). We next assessed effects of cyclic stretch on the proliferation of PA-IVS iPSC-CMs. While cyclic stretch promoted cardiomyocyte proliferation (Figure F), PA-IVS-1v and PA-IVS-2v iPSC-CMs displayed less proliferative than controls under both static and cyclic stretch (Figure G), implying that genetic factors may contribute to a spectrum of ventricular hypoplasia in PA-IVS.

Next, we performed scRNA-seq on PA-IVS-1v and control cells during cardiac differentiation (D5, D10, D14, and D30) (Figure H). We constructed topological structures of single-cell transcriptomics (n=45,460) in PA-IVS-1v and controls using the Uniform Manifold Approximation and Project (UMAP) plot visualized by samples (Figure I) and cell types (Figure J). We annotated multiple cell types during cardiac differentiation, including first heart field (FHF) progenitor (ISL1-/NKX2–5+), second heart field (SHF) progenitor (ISL+/NKX2–5+), epicardial progenitor (WT1+/TBX18+), and early cardiomyocyte (TNNT2+) (Figure J). At D10, we identified four major cell types: FHF progenitors, SHF progenitors, epicardial progenitors, and early cardiomyocytes (Figure K). The population of SHF progenitors was dramatically diminished in PA-IVS-1v compared to controls (Figure K). The proportion of SHF progenitors was significantly reduced whereas those of FHF and epicardial progenitors were increased in PA-IVS-1v versus controls (Figure L). These results indicate cell lineage determination towards SHF progenitors is suppressed, with enhanced differentiation towards FHF and epicardial lineages in PA-IVS-1v. As SHF primarily contributes to RV, the abnormal lineage differentiation towards SHF progenitors may underlie the developmental etiology of RV hypoplasia in PA-IVS.

We further assessed biological pathways and metabolic functions that were dysregulated in early (D14) and fetal (D30) cardiomyocytes. For D14 cardiomyocytes, biological pathways associated with cell proliferation were downregulated, whereas those related to glycolysis and mitochondrial activity were upregulated in PA-IVS-1v versus controls (Figure M). Seahorse assays demonstrated elevated mitochondrial activity in PA-IVS D30 iPSC-CMs (Figure N). ATP production and respiration (basal and maximal) were significantly increased in PA-IVS compared to control iPSC-CMs (Figures N–P). Increased mitochondrial respiration in PA-IVS-1v iPSC-CMs may cause elevated ROS generation, which can result in oxidative DNA damage and cell cycle arrest in cardiomyocytes ⁵. These results indicate that PA-IVS iPSC-CMs are less proliferative and metabolically more advanced in mitochondrial respiration, which may lead to reduced cardiomyocyte proliferation and enhanced maturation. Data in this study are available from the corresponding author upon reasonable request.

In conclusion, we reveal that cardiomyocyte proliferation is compromised and pathways associated with cell cycle progression and cell proliferation are downregulated in PA-IVS iPSC-CMs. We uncover impaired differentiation of cardiac mesoderm towards SHF progenitors in PA-IVS-1v, implying a potential developmental etiology for HRHS. Elevated mitochondrial activity in early cardiomyocytes highlights metabolic dysfunction in PA-IVS, which may promote cardiomyocyte maturation at the expense of cardiac proliferation. Due to limitations in 2D monolayer differentiation, further work using 3D iPSC-derived cardiac organoids would elucidate chamber-specific defects in HRHS.

Nonstandard Abbreviations and Acronyms

CM

Cardiomyocyte

DEG

Differentially expressed gene

FHF

First heart field

HRHS

Hypoplastic right heart syndrome

iPSC

Induced pluripotent stem cell

PA-IVS

Pulmonary atresia with intact ventricular septum

RV

Right ventricle

SHF

Second heart field

UMAP

Uniform Manifold Approximation and Project