Computational profiling of hiPSC-derived heart organoids reveals chamber defects associated with NKX2-5 deficiency

A human cardiac organoid system, coupled with single cell RNA sequencing and machine learning for transcriptional phenotyping, was developed. This allowed investigation of a genetic variant associated with Ebstein's Anomaly, a congenital heart disease with chamber defects. Heart organoids have the potential to generate primary heart-like anatomical structures and hold great promise as in vitro models for cardiac disease. However, their properties have not yet been fully studied, which hinders their wide spread application. Here we report the development of differentiation systems for ventricular and atrial heart organoids, enabling the study of heart diseases with chamber defects. We show that our systems generate chamber-specific organoids comprising of the major cardiac cell types, and we use single cell RNA sequencing together with sample multiplexing to characterize the cells we generate. To that end, we developed a machine learning label transfer approach leveraging cell type, chamber, and laterality annotations available for primary human fetal heart cells. We then used this model to analyze organoid cells from an isogeneic line carrying an Ebstein's anomaly associated genetic variant in NKX2-5, and we successfully recapitulated the disease's atrialized ventricular defects. In summary, we have established a workflow integrating heart organoids and computational analysis to model heart development in normal and disease states.

Automated summary

The researchers developed a human cardiac organoid system and used single cell RNA sequencing and machine learning to study the genetic variant associated with Ebstein's Anomaly. They found that the system could generate chamber-specific organoids and successfully recapitulated the disease's atrialized ventricular defects.