Epicardioid single-cell genomics uncovers principles of human epicardium biology in heart development and disease

Abstract

The epicardium, the mesothelial envelope of the vertebrate heart, is the source of multiple cardiac cell lineages during embryonic development and provides signals that are essential to myocardial growth and repair. Here we generate self-organizing human pluripotent stem cell-derived epicardioids that display retinoic acid-dependent morphological, molecular and functional patterning of the epicardium and myocardium typical of the left ventricular wall. By combining lineage tracing, single-cell transcriptomics and chromatin accessibility profiling, we describe the specification and differentiation process of different cell lineages in epicardioids and draw comparisons to human fetal development at the transcriptional and morphological levels. We then use epicardioids to investigate the functional cross-talk between cardiac cell types, gaining new insights into the role of IGF2/IGF1R and NRP2 signaling in human cardiogenesis. Finally, we show that epicardioids mimic the multicellular pathogenesis of congenital or stress-induced hypertrophy and fibrotic remodeling. As such, epicardioids offer a unique testing ground of epicardial activity in heart development, disease and regeneration.

Single-cell roadmap of cardiac differentiation identifies roles for ZNF711 and retinoic acid in balanced epicardial and cardiomyocyte lineage commitment

“…Our findings corroborate these observa�ons and demonstrate that these popula�ons can emerge independently of RA treatment (Drakhlis et al, 2021;Silva et al, 2021). This was confirmed by the expression of markers of the juxta-cardiac field and pre-epicardial cells (Meier et al, 2023), including the expression of the newly iden�fied epicardial markers such as ITGA8 (Zawada et al, 2023). Paracrine signaling from endodermal and epicardial popula�ons has been reported to be beneficial for both the epicardial and CM cell popula�ons (Quijada et al, 2020).…”

“…Epicardial and anterior endoderm cell popula�ons have been observed in previous studies (Drakhlis et al, 2021;Ho�auer et al, 2020;Meier et al, 2023;Silva et al, 2021;Zawada et al, 2023).…”

“…Human pluripotent stem cells (hPSC) offer an immense potential to differentiate many cell types in vitro , thus presenting a relevant alternative in mimicking the earlier steps of human tissue development (Takahashi et al, 2007; Thomson et al, 1998; Yu et al, 2007). Various culture models have been established to obtain cardiomyocytes (CMs) (Burridge et al, 2012; Devalla & Passier, 2018; Mummery et al, 2012), with variations including an early step of retinoic acid (RA) treatment to enrich for atrial CMs (aCMs) (Devalla et al, 2015; Lee et al, 2017; Schwach et al, 2022; Zhang et al, 2010) or epicardial derivatives (Guadix et al, 2017; Hofbauer et al, 2021a; Meier et al, 2023). Over time, the cardiac cultures have increased in complexity (e.g.…”

“…Single-cell transcriptomic studies have elucidated the differentiation processes of hPSCs towards CMs (Ameen et al, 2022; Churko et al, 2018; Friedman et al, 2018; Grancharova et al, 2021; Hofbauer et al, 2021b; Lewis-Israeli et al, 2021; Meier et al, 2023; Mononen et al, 2020; Ruan et al, 2019; Sahara et al, 2019; Zawada et al, 2023). Several of these studies highlighted the presence of non-myocyte subpopulations within the culture (Churko et al, 2018; Friedman et al, 2018; Mononen et al, 2020), while others shed light on the specification of particular cardiomyocyte subtypes, such as atrial, right and left ventricular CMs, and node-like CMs (Funakoshi et al, 2021; Pezhouman et al, 2021; Wiesinger et al, 2022; Yang et al, 2022).…”

scCross: A Deep Generative Model for Unifying Single-cell Multi-omics with Seamless Integration, Cross-modal Generation, and In-silico Exploration

“…As the single-cell sequencing paradigm matures, we're seeing the convergence of multiple data modalities, offering a holistic view of cellular states [3,[12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. Effective single-cell multi-omics data integration remains challenging, particularly when integrating unmatched data.…”