Cardiovascular disorders, including cardiovascular disease and congenital heart defects (CHD), are leading causes of death and birth defects, respectively. Despite their importance, progress in creating human heart organoid models for cardiovascular disease research has lagged behind other organs. Human pluripotent stem cells (hPSCs) offer the potential to recapitulate developmental steps in vitro to produce specific cardiac cell types. However, current cell models lack the structural and cellular complexity of actual heart tissue, often focusing on isolated cell types and neglecting the intricate interactions between different cell types within the 3D matrix. There's a significant need for more accurate in vitro models of the human heart to study both healthy and diseased states for research and translational applications. Over the last decade, researchers have attempted to develop heart-on-a-chip or heart organoid models using tissue engineering techniques. While these approaches offer high control, they tend to be expensive, labor-intensive, and difficult to scale. Furthermore, they often don't accurately represent the original cell composition or organization of the heart. More recently, self-assembling organoid technologies have emerged, focusing on the differentiation of PSC embryo-like aggregates to recapitulate early cardiogenesis. Studies using mouse embryonic stem cells have provided insights into early heart development in vitro, but have limitations associated with mouse models. Human PSC-derived hybrid cardiac-foregut organoids and cardioids have also been reported, showing progress in self-assembly, but often relied on complex growth factor mixtures, co-culture approaches, and independent differentiation protocols. This study presents a small molecule-based methodology to create complex and physiologically relevant self-assembling human heart organoids (hHOs) using hPSCs by manipulating cardiac developmental programs. The protocol primarily uses three sequential Wnt modulation steps (activation/inhibition/activation) on suspension embryoid bodies (EBs) to produce heart-like structures. This method is cost-effective compared to growth factor-based approaches, simpler, automatable, scalable, and suitable for high-content/high-throughput pharmacological screenings. As a proof-of-concept, the organoid system was used to model the effects of pregestational diabetes (PGD) on the developing embryonic heart.

The existing literature reveals a significant gap in accurate in vitro models of the human heart for studying cardiovascular diseases and congenital heart defects. While previous attempts have used tissue engineering and self-assembly techniques, these methods often fall short in terms of cost-effectiveness, scalability, and physiological relevance. Mouse embryonic stem cell models have provided valuable insights into early cardiogenesis, but are limited by species differences. Recently developed human PSC-derived organoids have demonstrated increased complexity, but often rely on intricate and costly protocols involving multiple growth factors and co-culture approaches. This research aims to address these limitations by developing a simplified, efficient, and scalable method for generating human heart organoids that accurately reflect the structural and cellular complexity of the developing human heart.

The study employed a three-step Wnt signaling modulation strategy using small molecules to generate self-assembling human heart organoids (hHOs) from human pluripotent stem cells (hPSCs). The process began with the formation of embryoid bodies (EBs) from dissociated hPSCs via centrifugation in ultra-low attachment 96-well plates. These EBs were then subjected to a carefully timed sequence of Wnt pathway manipulations using CHIR99021 (a Wnt activator) and Wnt-C59 (a Wnt inhibitor). The optimal concentration of CHIR99021 for the initial activation was determined through experimentation, revealing that lower concentrations (1-4 μM) were more effective than the typically used 10-12 μM in monolayer methods, likely due to endogenous morphogen production within the 3D environment. A second Wnt activation step using CHIR99021 was introduced to promote proepicardial organ specification, further enhancing organoid complexity. The optimal concentration and duration of this second CHIR99021 exposure were also determined experimentally. The addition of BMP4 and Activin A was also tested to enhance chamber formation and vascularization, showing positive effects on organoid size and chamber interconnectivity. To characterize the developmental steps and cell types present in the hHOs, RNA sequencing was performed at multiple time points (days 0-19) during differentiation. The data was analyzed using k-means clustering, gene ontology analysis, and comparison with both monolayer differentiation data and human fetal heart tissue data from public repositories. Immunofluorescence staining was used to identify various cardiac cell lineages, including cardiomyocytes (TNNT2), epicardial cells (WT1, TJP1), cardiac fibroblasts (THY1, VIM), endocardial cells (NFATc1), and endothelial cells (PECAM1). Optical coherence tomography (OCT) was used to visualize the three-dimensional structure of the organoids, specifically their internal chambers. Transmission electron microscopy (TEM) was used for ultrastructural analysis. Electrophysiological activity was assessed using a multi-electrode array (MEA), and calcium transients were measured using live calcium imaging in organoids expressing GCaMP6f. Finally, to model pregestational diabetes (PGD)-induced CHD, the hHOs were cultured under normal glycemic conditions (NHOs) and diabetic-like conditions (PGDHOs) using media with varying glucose and insulin concentrations. The effects of these conditions were assessed through morphological analysis, electrophysiology, metabolic assays (Seahorse), and TEM.

This research successfully generated complex, self-assembling human heart organoids (hHOs) using a three-step Wnt signaling modulation protocol based on small molecules. These hHOs exhibited significant structural and functional similarities to age-matched human fetal heart tissue: * **Multicellular Complexity:** The hHOs contained a diverse array of cardiac cell types, including cardiomyocytes (approximately 59%), epicardial cells (16%), endocardial cells (14%), cardiac fibroblasts (12%), and endothelial cells (1.6%). The relative proportions of these cell types were comparable to those found in fetal hearts. * **Structural Organization:** The organoids spontaneously formed interconnected chambers lined with endocardial cells, a feature not commonly seen in previous in vitro models. They also displayed a well-defined epicardial layer adjacent to the myocardial tissue, mimicking the anatomical structure of the developing heart. * **Functional Activity:** The hHOs showed robust and coordinated beating as early as day 6 of differentiation, persisting for at least 8 weeks. Electrophysiological recordings using a multi-electrode array (MEA) revealed action potential waveforms similar to those observed in the human heart. Live calcium imaging confirmed regular calcium waves typical of cardiac muscle. * **Developmental Processes:** The hHOs recapitulated key developmental processes, including heart field formation (first and second heart fields represented by HAND1 and HAND2 expression) and atrioventricular specification (expression of MYL2 and MYL7). Transcriptomic analysis showed gene expression profiles highly similar to those of human fetal hearts, significantly different from those of monolayer cardiomyocyte differentiation methods. * **Metabolic Perturbations:** When cultured under diabetic-like conditions, the hHOs showed features consistent with pregestational diabetes-induced CHD, including increased size (macrosomia), irregular electrophysiological activity (arrhythmia), reduced mitochondrial density, altered lipid metabolism, and altered cardiomyocyte composition (reduced ventricular, increased atrial cardiomyocytes). This suggests the organoids can model the metabolic consequences of the disorder. The addition of BMP4 and Activin A to the differentiation protocol enhanced chamber interconnectivity and vascularization, further improving the physiological relevance of the hHOs.

This study successfully addresses the limitations of current in vitro models of the human heart by developing a robust, scalable, and cost-effective method to generate human heart organoids with unprecedented cellular and structural complexity. The hHOs closely mimic the development and cellular composition of the human fetal heart, including the formation of interconnected chambers lined by endocardial cells, and the presence of an epicardial layer in close proximity to the myocardial tissue. The functional activity of the hHOs, demonstrated through coordinated beating, electrophysiological recordings, and calcium transients, further validates their relevance as models of the human heart. The successful modeling of pregestational diabetes-induced CHD highlights the potential of the hHOs for studying the pathogenesis of congenital heart defects. The results provide novel insights into the molecular and cellular mechanisms underlying these defects and potentially enable high-throughput drug screening for therapeutic interventions.

This study presents a novel method for generating self-assembling human heart organoids that exhibit remarkable structural and functional complexity, closely resembling the developing human fetal heart. The ability to model pregestational diabetes-induced congenital heart defects demonstrates the organoids' utility for studying human heart development and disease. Future research should focus on further improving the vascularization of the organoids and applying these models to study a wider range of cardiovascular diseases and explore potential therapeutic strategies. The high-throughput nature of the method enables the screening of large numbers of compounds to identify potential treatments.

While the hHOs presented in this study show significant advancements in modeling the human heart, limitations remain. The organoids, like many other organoid models, may exhibit developmental deviations over extended culture periods. Although the organoids contain multiple cardiac cell types and show structural complexity, they may not perfectly recapitulate the full anatomical and functional features of the mature human heart. Further investigation into the maturity and functionality of the spontaneously formed vascular networks is warranted. Additionally, while the study provides a proof-of-concept for modeling PGD-induced CHD, further validation and exploration of other congenital heart disease models are needed to fully assess the platform's versatility.