Fig. 8: PDGF-AA can partially compensate for the absence of cytokines in co-aggregate cultures.

a, (i) Schematic showing the experimental strategy for co-aggregation of E11.5 UBC–GFP⁺ aortic endothelial cells with Mesp1^der PSCs or Wnt1^der PSCs in the presence and absence of cytokines or PDGF-AA. E11.5 UBC–GFP⁺ AGM re-aggregates were used as controls. (ii) CFU-C assays from 96-h co-aggregate cultures (E11.5 UBC–GFP⁺ aortic endothelial cells and Mesp1^der PSCs, n = 7–10 co-aggregates per culture condition; E11.5 UBC–GFP⁺ aortic endothelial cells and Wnt1^der PSCs; n = 9–13 co-aggregates per culture condition). BFU-E, burst-forming unit–erythroid; CFU-GM, colony-forming unit–granulocyte/macrophage; and CFU-GEMM, colony-forming unit–granulocyte/erythrocyte/macrophage/megakaryocyte. (iii) Percentage of GFP⁺ cells in the peripheral blood of primary transplant recipients (n = 5 co-aggregates per culture condition); e.e., embryonic equivalent. (iv) Percentage of GFP⁺ cells in the peripheral blood of secondary transplant recipients from the indicated primary recipients. One co-aggregate per recipient mouse; each point represents an individual recipient. b, Model incorporating the changing landscape of Mesp1- and Wnt1-derived PSCs in the AGM stroma and their role in generating endothelial and sub-endothelial cells as well as LT-HSCs. Pre-E9.5 and E9.5 Mesp1^der PSCs contribute to the aortic endothelium. E10.5 and E11.5 endothelium secretes PDGF-AA, which acts on PSCs that secrete haemogenic factors to promote EHT. Mesp1^der PSCs are replaced by Wnt1^der PSCs at E13.5. This is accompanied by the loss of high PDGF-AA in the AGM and interruption of a PDGFRA-mediated signalling axis involving Mesp1^der PSC-dependent induction of EHT. AoE, adult aortic endothelium; HE, heart endothelium; NT, neural tube; and N, notocord. Data were derived from biologically independent samples, animals and experiments (b(ii), n = 3; b(iii),(iv), n = 5). Data represent the mean ± s.d. A random-effects Poisson regression was used to compare colony counts (a(ii)) and an ANOVA was used to compare donor chimerism (a(iii),(iv)); **P < 0.01, ***P < 0.005. The precise P values are listed in the source data.

Source data