Extended Data Fig. 5: Overview of VEC Populations from Human and Mouse scRNA-seq and Spatial Location.
From: APOE–NOTCH axis governs elastogenesis during human cardiac valve remodeling
a. UMAP demonstrating the distribution of four valves within VECs; b. Left: Bar graph showing proportions of each valve within each VEC subtype. Right: Bar graph showing proportions of each VEC subtype within each valve; c. Feature plots of marker genes for each VEC subtype; d. UMAP visualization of mouse VEC subtypes and timepoint distribution (Hulin et al., 2019); e. Feature plots of marker genes for each VEC subtype from mouse dataset; f-g. UMAP demonstration of the similarity of VEC subtypes between human and mouse VECs through reverse projection; h-i. Immunofluorescence staining of FOXC2 in four human fetal valves at W15 (h). Zoom out of immunofluorescence staining of FOXC2, CD55 and RNA in situ hybridization of PTGDS in four human fetal valves at W15 (i). White arrows represent unidirectional flow directions. n = 3 different hearts; j. UMAP visualization of each VEC subtype (Upper), and violin plot of CD55, PTGDS, FOXC2 expression among four valves (Lower); k. RNA in situ hybridization of PTGDS, APOE, PECAM1, and immunofluorescence staining of CD55, APOE, and PECAM1 in four human fetal valves at W15. n = 3 different hearts; l. Quantification of PTGDS+ VECs from Fig. 4d; m. PCA Analysis comparing four different valves in each VECs subtype. For l: Data shown as mean ± SEM. ***p < 0.001. SL vs. AV. Statistics: Unpaired 2-tailed t-test (two groups). P value in l: p = 0.0007. SL valves: Semilunar valves, A-V valves: Atrioventricular valves.
