Left Ventricular Non-Compaction (LVNC) is a rare cardiomyopathy characterized by excessive trabeculae and intertrabecular recesses in the left ventricle of the heart. It is believed to be caused by abnormal trabecular compaction process, which involves trabecular remodelling to form a thick myocardial wall. Embryonic trabeculae consist of bipotent progenitors that differentiate into the ventricular conduction system (VCS) and contractile cardiomyocytes, yet it remains to be further understood how the pathways involved in trabecular cell specifications play a role in the compaction process. Here we present two transcription factors, Iroquois homeobox 3 (Irx3) and Irx4, playing cooperative and antagonistic roles during trabecular specification and maturation, thereby preventing LVNC. Our single-cell RNA sequencing analysis revealed that the differentiation of trabecular cells into VCS cells was accompanied by increased Irx3 but decreased Irx4 expression. In addition, we found that the loss of Irx4 led to both an increased expression level of Irx3 and an expansion in the number of Irx3 expressing cells. Conversely, the loss of Irx3 resulted in an increase in Irx4 expression within VCS cells. Next, to investigate the functional importance of the antagonistic relationship between Irx3 and Irx4 in trabecular differentiation, we generated Irx3 and Irx4 compound knockout (KO) mice and examined the structure and function of their hearts. Notably, unlike Irx3 KO hearts showing normal ventricular myocardium, Irx4 KO hearts exhibited mild LVNC, and Irx3 and Irx4 double knockouts (Irx3/4dKO) hearts displayed severe LVNC. Multiome sequencing of E14.5 ventricles revealed that Irx3/4dKO ventricles contained an increased number of trabecular cells, which expressed higher levels of cell proliferation markers but altered trabecular identity genes, compared to littermate controls. Collectively, our results demonstrate that balanced antagonism between Irx3 and Irx4 is crucial for establishing trabecular identity, highlighting the importance of proper differentiation of trabeculae into VCS and non-VCS contractile cells during ventricular compaction, thereby preventing noncompaction cardiomyopathy.

Congenital heart disease (CHD) affects 1% of live births and results from improper differentiation, migration, and specification of cardiac progenitor cells. The most common form of single ventricle CHD, Hypoplastic Left Heart Syndrome (HLHS), is characterized by underdevelopment of the left sided structures of the heart, including: hypoplastic left ventricle, stenotic or atretic valves, and aortic malformation. Despite advances in medical and surgical management of HLHS, there remain high levels of morbidity and mortality associated with the disease. Unfortunately, efforts into studying HLHS have been limited by shortcomings in experimental model systems. To address this, we utilized human induced pluripotent stem cell derived cardiac organoids to uncover mechanistic insights into the pathogenesis of HLHS in a cohort of HLHS-affected proband-parent trios. First, we generated an unaffected, WT atlas of cardiac organoid development comprising of scRNA/ATAC-sequencing across differentiation, and validated emergence of multiple cardiac cell types and recapitulation of key steps of early heart development. Importantly, upon genetic perturbation, we were able to identify expected cell-type specific differentiation defects. After establishing this model, we generated cardiac organoids from HLHS patient iPSCs and compared to cardiac organoids from their unaffected parents. Interestingly, we identified marked cell blebbing from HLHS-cardioids that we conjecture is indicative of inappropriate cell-cell interactions between co-developing cell types. We then performed scRNA/ATAC-seq on cardiac organoids from four HLHS-patient and parent lines. We identified aberrantly active cardiomyocyte-endocardial VEGF signaling which partially underlies the HLHS cardiac organoid phenotype. Additionally, we found enrichment of interferon signaling in HLHS-endocardium and identified a novel cardiac fibroblast cell state associated with the disease which may reflect altered specification. In summary, in this study we generated an atlas of cardiac organoid differentiation and utilized it to identify aberrant signaling networks in HLHS-patient derived cardiac organoids. Future studies utilizing chamber-specific protocols may provide further clarity on the pathogenesis of HLHS.

Congenital heart disease (CHD) is the most common birth defect, affecting ~1% of live births worldwide. Exome sequencing has identified hundreds of genes associated with CHD, and recent bulk and single-cell omics approaches by us and others have extensively characterized the molecular landscape of heart development, identifying putative drivers of cardiac development and CHD. However, functional validation of these candidates and an understanding of how their perturbation affects heart development remain missing. Yet, the large numbers of these candidates creates an enormous search space that is challenging to approach directly with individual functional studies, such as mouse knockout models. Here, we take advantage of the scalable human pluripotent stem cell (hPSC) differentiation model, coupled with CRISPR-based gene inhibition (CRISPRi), to assess the functional requirement of 1,626 candidates in atrial versus ventricular differentiation. A pilot screen with 30 candidates tested in undifferentiated hPSCs, mesoderm, and atrial and ventricular cardiomyocytes successfully identified known regulators (Nkx2-5, Mesp1) as hits, while accurately marking non-cardiac regulators as non-hits. Using this platform, we have screened 1,626 candidates including: differentially expressed genes between ventricular-fated and non-ventricular-fated cardiac progenitors, dynamically regulated genes during the transition from progenitor to differentiated cardiac cell types, differentially and highly expressed genes between atrial and ventricular cardiomyocytes, as well as genes associated with CHDs. While many of these candidates are associated with muscle and cardiovascular development, some are interestingly involved in lipoprotein assembley and remodeling, vitamin metabolism, ciliopathies, and neurogenesis. Our results will focus the large number of candidates to a defined set of critical regulators of heart development and disease, which can be mechanistically studied in animal models, with the overall goal to deepen our understanding of the complex mechanisms underlying CHD.

The heart functions in two parallel but asymmetric circuits, in which the right and left ventricles drive the pulmonary and systemic circulations, respectively. In the heterotaxy syndrome, abnormal left-right embryo patterning leads to a spectrum of severe congenital heart defects, including ventricle malposition. A postulate anchored in the clinical nomenclature, assumes that the looping direction of the embryonic heart tube determines ventricle position at birth. However, this has not been demonstrated experimentally. Here, we performed a unique longitudinal analysis of heterotaxy, using multi-modality imaging of Nodal mouse mutants. Based on direct correlations and advanced statistics, we dissected the contribution of heart looping variations to specific structural heart malformations. We uncovered unexpected plasticity of ventricle position after heart looping, in 30% of revertant samples. Genetic tracing and topological associations do not support molecular reprogramming of ventricles but rather point to a novel step of heart tube remodelling after heart looping. Our work reveals distinct asymmetric events shaping the heart, beyond initial symmetry breaking in the node.

Human cardiogenesis is a finely orchestrated process that involves dynamic genetic and morphological changes. However, its underlying cellular mechanisms remain obscure, due to limited access to human heart in utero and lack of effective experimental models. Advances in 3D bioprinting and stem cell technologies have enabled the fabrication of cardiac tissues with complex structures and microenvironment. Here we present a perfusable 3D bioprinted model of embryonic human heart tube (day 22), consisting of endocardium, cardiac jelly, and myocardial layers (Figure 1A-C). hiPSC-derived cardiomyocytes (hiPSC-CMs) and endothelial cells (ECs) were cocultured in bioprinted heart tubes (Figure 1F). Full endothelialization of the heart tube cavity, progression of myocardial tissue compaction and global contraction showed robust viability and cardiac function of the engineered heart. Further, single cell RNA-seq revealed a promoting effect of dynamic flow in cardiac tissue maturation and cell lineage commitment. Heterogeneous CM and EC subpopulations related to key biological processes during the linear heart tube stage (day 22) pre- vs. post- flow. Overall, this study establishes a perfusable model of human heart tube with robust function. Cellular activities related to cardiogenesis upon hemodynamic flow activation were observed, suggesting the great capacity of the 3D bioprinted model for studying cardiogenesis, disease, and prenatal therapeutic targets.

Hypoplastic left heart syndrome (HLHS) is a severe form of congenital heart disease that is uniformly lethal if left untreated. Protein damaging mutations in one allele of RBFOX2, a highly conserved RNA binding protein, are enriched in HLHS patients, suggesting they are causal. However, it remains unclear how RBFOX2 haploinsufficiency contributes to disease pathogenesis as heterozygous animal models appear healthy. Here, we generated and characterized RBFOX2 heterozygous and null human induced pluripotent stem cells (hiPSCs) and assessed their phenotypes in differentiated cardiomyocytes (iPSC-CMs). In 2D culture, we observed dose-dependent reductions in cell adhesion, size, maturation, respiration, calcium handling, and action potential duration with no significant effect on proliferation. Although myofibril alignment and contractility were completely abolished in null hiPSC-CMs, these features were largely preserved in heterozygous cells. To learn if altered function might be revealed in a more native microenvironment, we generated 3D engineered heart tissues (EHTs) from RBFOX2 heterozygous hiPSC-CMs. We found that RBFOX2 is in fact haploinsufficient for tissue maturation and contractility, suggesting that myocardial-intrinsic defects contribute to RBFOX2-mediated HLHS. To delineate the underlying molecular mechanisms, we performed bulk RNA-seq on hiPSC- CMs from each genetic cohort to reveal differential gene expression (DGE) and alternative splicing events (ASEs) and eCLIP-seq on WT hiPSC-CMs to detect RNAs directly bound by RBFOX2 in the CM lineage. Integration of these datasets uncovered significant enrichment of DGE and/or ASEs of RBFOX2-bound and unbound transcripts involved in adhesion between the extracellular matrix (ECM) and the actin cytoskeleton/sarcomere network. Included in the direct targets with altered expression and alternative splicing was the ECM ligand FIBRONECTIN 1 (FN1), which when downregulated was previously associated with HLHS. Overall, these data suggest that RBFOX2 sits atop an expression and splicing hierarchy that promotes optimal CM cell growth, adhesion, and contractility to protect the heart from HLHS pathogenesis.