Figure 1. Proliferative split in basal cancer cells enables expansion of intermediate cells.
(A) Simplified schematic of single-cell RNA sequencing of WT and Ptenfl/fl ventral prostates. (B) UMAP of WT and Ptenfl/fl epithelial cells. Left, both conditions superposed; middle, WT only; right, Ptenfl/fl only. Epithelial cell types are demarcated by color (red = basal, light blue = luminal progenitor, dark blue = intermediate, green = differentiated). (C) Relative abundance of epithelial cells in WT (n=3) and Ptenfl/fl (n=2) mice. Y-axis shows the % composition of each sample by cell type (***p<0.001, negative binomial test). Data presented as +/-SD. (D) Top GSEA results enriched in Ptenfl/fl compared to WT for each epithelial subtype. Intermediate cells in Ptenfl/fl were compared to luminal progenitor cells in WT. All pathways are enriched with FDR <0.05. (E) Proliferation signature (CCP) composite score in epithelial cells, clustered by condition (Data presented as +/-SD, *p<0.05, n.s.=not significant, permutation test). N/A indicates missing data due to no cells being present in the condition. WT n=3, Ptenfl/fl n=2. (F) UMAP visualization of cell cycle phase assignment per cell, showing % cells in non-G1 (S or G2/M) (black border = WT basal cells, blue border = hypo-proliferative basal cells in Ptenfl/fl, and red border = hyper-proliferative basal cells in Ptenfl/fl. *p<0.05, chi-square test). (G) GSEA between hyper- and hypo-proliferative basal clusters in Ptenfl/fl. All pathways are enriched with FDR <0.05. (H) RNA velocity analysis of Ptenfl/fl epithelial cells; highlighted section shows intersection of basal and intermediate cells. (I) Pseudotime trajectories drawn by Palantir through the basal and intermediate compartments, with hypo-proliferating basal cells as the designated start point. (J) Expression of important cell fate and differentiation regulators along basal-intermediate trajectory.