Fig. 4: Communication specifically from the epicardium changes with age.
Specific epicardial markers were identified as selective for fetal, shared or adult epicardium through a parallel and stage-independent analysis shown in a mirrored volcano plot highlighting the top 10 genes in each subset (a) and a heat map of expression of the top 20, 10 and 10 gene subsets where markers with clear fetal bias, including WNT signaling genes, are shown with arrows (b). c, Top-scoring predictive markers for epicardium compared with other heart clusters and between epicardial subtypes. The predicted level of epicardial communication with other heart cell types in each stage was calculated using CellPhoneDB, showing the total volume of communication (d) and clustered epicardial-specific secretions filtered by upregulated genes predicted to interact with all other heart clusters of either fetal or adult source by tissue, color-coded by stage selectivity of secreted markers (e). In e, x-axis labels refer to resolution 2 cluster IDs: 1, Adipocytes; 2, Cardiomyocytes; 3, Endocardial; 4, Endothelial_Arterial; 5, Endothelial_Capillaries; 6, Endothelial_Other; 7, Endothelial_Venous; 11, Fibroblasts; 12, Immature_Cardiomyocytes; 13, Immature_other; 14, Lymphoid_Immune_Cells; 15, Myeloid_Immune_Cells; 16, Neuronal_Cells; 17, Pericytes; 18, Pericytes_Stromal; 19, Smooth_Muscle_Cells. The significance of epicardial markers in a was determined using two-sided Wilcoxon rank-sum tests and adjusted for multiple comparisons using Bonferroni correction.
