Fig 2: FoxC1-induced the vascular microenvironment under hypoxia. ECs were transfected with FoxC1 vector (adFoxC1), FoxC1 siRNAs (siFoxC1), or control siRNA duplexes (Ctrl) and cultured under normoxia or hypoxia. A Tube formation of MSCs subjected to hypoxic culture in the absence or presence of adFoxC1, or siFoxC1. The number of tube formation was calculated (B). C Direct fluorescence staining with Dil-acLDL and FITC-UEA-1 in ECs culture system with or without transfection of adFoxC1 or siFoxC1, and counterstained with DAPI. E Factor VIII and α-SMA expression in MSCs, as determined using immunofluorescence with anti-factor VIII (red) and anti-α-SMA (green) antibodies, and the nuclei were stained blue using DAPI. Scale bars = 50 µm. Double-positive cells in D and F were expressed as a percentage of Dil-acLDL+ UEA-1+ DAPI+ (D) or Factor VIII+α-SMA+DAPI+ (F) relative to all DAPI+ cells. G–I ELISA assay of difference in pro-angiogenic factors, Ang1 (G), bFGF (H), and VEGF (I) at 48 h after nomorxic or hypoxic culture in the conditioned culture supernatants of the ECs transfected with FoxC1vector (adFoxC1), FoxC1 siRNAs (siFoxC1), or control siRNA duplexes (CON). All data are the means ± SEM. Welch ANOVA analyses were performed in B, G, H, and I, and one-way ANOVA analysis was used in D and F. p < 0.05: *versus the NO/HO group under the same culture conditions, †versus the NO + adFoxC1group, ‡versus the HO + adFoxC1 group (n = 10 per group). HPF, high-power field, acLDL, acetylated low-density lipoprotein, DAPI, 4′, 6-diamidino-2-phenylindole, FITC-UEA fluorescein isothiocyanate-labeled Ulex europoeus agglutinin I lectin

Fig 3: MSCs in FoxC1-mediated vascular microenvironments adopt endothelial cell fates after Oct4 overexpression. A Angiogenesis and fibrosis of MSCs subjected to transfection of adOct4, siOct4, or no intervention in the absence or presence of adFoxC1, as determined using immunoblotting with anti-factor VIII/α-SMA and collagen I/vimentin, respectively. B The histogram shows the quantification of the protein level relative to β-actin. C–F Double-positive cells in angiogenesis or fibrosis proteins were expressed as a percentage of factor VIII+CD31+ (C, D) or collagen I+vimentin+ (E, F) relative to all MSCs by FACS. G, H Direct fluorescence staining with VIII/α-SMA and collagen I/vimentin in MSCs culture system with or without transfection of adFoxC1, and counterstained with DAPI. Scale bars = 25 µm. I, J Correlation analysis of the double-positive rates of VIII/α-SMA (I) and collagen I/vimentin (J) in MSCs with the number of tubes/field in ECs. All data are the means ± SEM. p < 0.05: in ECs without adFoxC1, *versus MSCs treated with no intervention, †versus MSCs with adOct4, ‡versus MSCs with siOct4; in ECs with adFoxC1, §versus MSCs treated with no intervention, ‖versus MSCs with adOct4 (n = 10 per group). Welch ANOVA analyses were performed in B and D, and one-way ANOVA analysis was used in F. DAPI, 4′, 6-diamidino-2-phenylindole; FACS, fluorescent-activated cell sorting

Fig 4: FoxC1 overexpression in ECs promotes self-renewal and survival of MSCs under hypoxia. A Representative growth state and morphology of the MSCs without (monoculture) or with coculture with the ECs alone, adFoxC1-treated, and siFoxC1-treated ECs in 108 h of culture under hypoxia conditions. Transfection of adFoxC1 improves self-renewal of MSCs, and FoxC1 deficiency impaired this effect. Scale bars = 200 µm. B MSCs were cultured for 108 h under hypoxic conditions, and the number of cumulative population doublings was determined. C For cell cycle distribution, MSCs without (monoculture) or with coculture with ECs treated with or without adFoxC1 or siFoxC1 for 108 h were collected and detected by flow cytometry, the percentage of of MSCs at S phase of the cell cycle was statistically analyzed as a histogram. D The percentage of cells in S phase was determined in MSCs without (monoculture) or with coculture with the ECs treated with or without adFoxC1 or siFoxC1. E MSCs after 108 h of hypoxic culture were stained with anti-Ki67 antibodies and then counterstained with DAPI, followed by confocal microscopy. Scale bars = 200 µm. F The proliferation of MSCs cocultured with adFoxC1- treated ECs was remarkably higher than the other MSCs, as assessed using a CCK8 assay. G MSCs after 108 h of hypoxic culture were evaluate by using the TUNEL assay. TUNEL-positive cells were obviously decreased in the MSCs cocultured with adFoxC1- treated ECs as compared with the other MSCs. Scale bars = 200 µm. H Statistical analysis of the mean percentage of annexin V-positive cells relative to the total number of MSCs. FoxC1 overexpression in ECs reduced the apoptosis of MSCs, and FoxC1 deficiency increased apoptosis. All data are the means ± SEM. One-way ANOVA analysis was used in D, and Welch ANOVA analyses were performed in F and H. p < 0.05: *versus monoculture, †versus ECs alone, and ‡versus adFoxC1 + ECs (n = 10 in each group)

Fig 5: Oct4 enhances FoxC1-mediated MEndoT after MSC therapy. A Representative fluorescence microscopy images of tissue sections showing the retention of EGFP labeling MSCs at the injection site 30 days after transplantation. Scale bar: 100 μm. E Quantitative data showing the retention of EGFP+ MSCs in the ischemic hearts. B Representative phenotype of gated EGFP+ cells evaluated by FACS. F Quantitative analysis of the percentages of EGFP-positive cells (EGFP+) relative to the whole ventricular cell population in the ischemic hearts after 30 days of transplantation. C, G Endothelial cell immunostaining in EGFP-labeled MSCs (C, arrowheads) and quantitation of endothelial cell expression (G). Scale bar: 50 μm. D, H Vascular density images assessed by factor VIII immunostaining (D) and statistical analysis of blood vessels (H) in various groups 30 days after cell therapy, respectively. Scale bar: 100 μm. (I) Quantitation of infarct size. J, K, L The changes of LVFS (J), LVEDD (K), and LVEDV (L) prior to and 30 days after cell therapy. All graphs show means ± SEM. p < 0.05: in the ischemic hearts without FoxC1 transfection, *versus MSC therapy alone, †versus Transplantation of adOct4 transfected MSCs, ‡versus Transplantation of siOct4 transfected MSCs; in the adFoxC1 hearts, §versus MSC therapy alone, ||versus Transplantation of adOct4 transfected MSCs (n = 10 per group). Welch ANOVA analyses were performed in E–I and K, and one-way ANOVA analysis was used in J and L