Fig 2: Low Concentrations of BMP9 Modulate the Calcineurin/NFATc1 Signaling Pathway(A) qRT-PCR analysis of BMP9 effects on the mRNA levels of transcription factors known to be essential for lymphatic differentiation. FLK-1+ vascular precursors at day 5 were plated on CellTracker CMFDA-stained OP9 stromal cells. BMP9 (0.3 ng/mL) was added at day 6. After 24 hr of treatment, the CMFDA-negative cell population was sorted and collected for gene expression analysis. Data are the mean ± SEM of five experiments.(B) NFATc1 protein expression analysis after BMP9 stimulation. FLK-1+ vascular precursors at day 5 were stimulated in suspension in serum-free medium with or without 0.3 ng/mL BMP9 for 20 min. Immunoblotting was performed using an antibody against NFATc1 or actin proteins as a loading control. Results of a representative experiment, out of three, are illustrated. Sizes of the molecular weight markers are indicated on the left.(C) Inhibition of NFATc1 signaling by cyclosporin A prevents the BMP9-induced increase in LYVE-1+ cell formation. LYVE-1-positive cell formation was measured by flow cytometry 4 days after addition of BMP9 in the absence or presence of cyclosporin A. Data are the mean ± SEM of three experiments.*p < 0.05, significantly different from control untreated cells (A) and significantly different from treatment with BMP9 alone (C) using unpaired Student's t test. See also Figure S3.

Fig 3: ESC-Derived Vascular Precursors Co-cultured on Murine Stromal OP9 Cells Are Able to Form Early Lymphatic Derivatives(A) Schematic of the differentiation protocol illustrating the main steps and specific treatments according to the experiment goals. EBs, embryoid bodies.(B) Morphological observations of endothelial cell clusters formed after 5 days of co-culture (day 10 of differentiation) in control conditions. The arrows point to cell clusters exhibiting an endothelial-like morphology.(C–G) Immunofluorescence staining of endothelial cell-like clusters obtained in unstimulated control conditions at day 10/11 with anti-CD31 (C), anti-LYVE-1 (D and G), and anti-PROX-1 (F) antibodies. Nuclei were counterstained with Hoechst 33258 (C and E). Scale bars, 100 µm.(H) Flow-cytometry dot plot of the LYVE-1 and CD31 double immunostaining of the co-cultures at day10/11 used for cell sorting. The different gates used are outlined: R1, CD31+/LYVE-1+ cells; R2, CD31+/LYVE-1- cells; R3, CD31-/LYVE-1- cells. Co-cultures were performed in the presence of 0.3 ng/mL BMP9 to obtain sufficient cell numbers in the LYVE-1+ and LYVE-1- cell fraction.(I) Relative Prox-1 mRNA expression levels. Data shown are the mean ± SD of triplicates from the qRT-PCR experiment performed with the RNAs extracted from the different cell populations gated on the dot plot of the experiment illustrated in (H).See also Figure S1.

Fig 4: Analysis of Differential BMP Type I Receptor Involvement in the BMP9 Low-Dose and High-Dose Effects by RNA-Silencing Experiments(A) Effect of Alk1 and Alk2 siRNA on LYVE-1-positive cell formation induced by 0.3 ng/mL BMP9 in FLK-1+ vascular precursor co-cultures on mouse OP9 cells. Alk knockdowns were performed by siRNA transfection of vascular precursors by either electroporation or by using Lipofectamine RNAiMAX, 24 hr before BMP9 stimulation. Data are the mean ± SEM of three (siAlk2) or four (siScr, siAlk1) different experiments.(B) High-dose BMP9-induced LYVE-1- endothelial cell formation is inhibited after prior Alk2 knockdown in OP9 cells. OP9 cells were first transfected by Lipofectamine RNAiMAX with the different siRNAs, 24 hr before FLK-1+ vascular precursor plating in co-culture on OP9 cells and 48 hr before stimulation with 10 ng/mL BMP9. LYVE-1- CD31+ cell formation is illustrated. Data are the mean ± SEM of three experiments.(C and D) Alk2 knockdown in OP9 cells prevents SMAD1/5 phosphorylation and VEGF-A secretion induced by 10 ng/mL BMP9. OP9 cells were transfected with the different siRNAs by Lipofectamine RNAiMAX, 48 hr before stimulation with 10 ng/mL BMP9. For SMAD1/5 phosphorylation studies, serum-starved subconfluent OP9 cells were stimulated for 1 hr in the absence (CTL) or presence of BMP9, before immunoblotting with anti P-SMAD1/5 and actin antibodies (C). VEGF-A levels were measured after 48 hr of treatment with or without 10 ng/mL BMP9, by ELISA (D). Data are mean ± SEM of three experiments.*p < 0.05, ***p < 0.001, significantly different from BMP9 stimulation obtained after transfection with siScramble (siScr), using unpaired Student's t test. See also Figure S4.

Fig 5: Hypothetical Model for BMP9 Regulation of ESC-Derived Vascular Precursor Differentiation in Co-culture with Murine OP9 Stromal CellsLow BMP9 concentrations induce the formation of LYVE-1+ early lymphatic-specified endothelial cells through ALK1 activation. BMP9 was observed to directly activate FLK-1+ vascular precursors and earlier stages of vascular committed cells through mobilization of the NFATc1 signaling pathway. Moreover, an increase in the proliferation of the LYVE-1-positive cell population was found to largely contribute to its expansion. At higher doses, BMP9 could, in addition to ALK1, also activate ALK2 present on OP9 stromal cells. This leads to a significant rise in the secretion of VEGF-A by OP9 cells, with as a resultant effect an increase in LYVE-1- endothelial cell population at the expense of the LYVE-1+ early lymphatic-specified endothelial cell formation.