5 Benchmarking stem-cell derived models of cardiac cells against a high-resolution atlas of human heart development

<h3></h3> Human pluripotent stem cells (hPSCs) are a promising strategy for regenerative therapy, disease modelling, and personalised medicine. However, to realise their use in cardiovascular disease, researchers must overcome several challenges. In regeneration, evidence from engraftment of hPSC-derived cells into hearts suggests that cardiomyocytes may not integrate with host tissue. Additionally, disease modelling may not be informative when using the wrong cell types. This is further complicated as many hPSC-derived models of cardiomyocytes and other cardiac cells now exist without a clear indication of how they resemble their real-world equivalents. Models may still be optimised, and researchers and clinicians may have a difficult task of selecting the most appropriate model for their purposes. We hypothesise that the proximity between in vitro and in vivo systems can be determined using transcriptomics. We aim to benchmark and characterise the current hPSC-derived cardiac models using machine learning and statistical models trained on human developmental heart atlases.^{1 2} We constructed preliminary reference models to classify and stage up to 63 developmental cardiac cell types and applied these computational models to single-cell data from in vitro hPSC-cardiac models. Preliminary characterisation of hPSC-cardiomyocytes reveals a high fraction of cardiomyocytes and our approach captured increased cardiomyocyte compaction seen in protocols aimed at maturing cardiomyocytes, demonstrating a valid recapitulation of experimental results. However, our analyses suggested properties of in vitro models that may differentiate them from in vivo cardiomyocytes, such as a trend towards right rather than left cardiomyocytes, or an abundance of conductive-system cardiomyocytes in early culture. This research may ultimately be used as a guide for improving hPSC-derived cardiac models and for selecting appropriate protocols for translating stem cells into clinical application. <h3>References</h3> Bayraktar S, <i>et al</i>. High-resolution atlas of the developing human heart and the great vessels. <i>Preprint in bioRXiv</i> 2024. doi: 10.1101/2024.04.27.591127 Knight-Schrijver VR, <i>et al</i>. A single-cell comparison of adult and fetal human epicardium defines the age-associated changes in epicardial activity. <i>Nat Cardiovasc Res</i> 2022;<b>1</b>:1215–1229. doi: 10.1038/s44161-022-00183-w