Fig 2: Angpt1 deficiency increases lung metastasis of B16F10 melanoma experimental metastasis model. a Mean ± SEM for B16F10 subcutaneous tumor volume over 15 days comparing WT (n = 4) and Angpt1 Δ/Δ (n = 7) mice. b, c Quantification of lung metastasis of B16F10 melanoma cells 3 weeks after tail vein injection in Angpt1 Δ/Δ mice (n = 20) compared to WT mice (n = 19). d Quantification of lung metastasis of B16F10 melanoma cells 3 weeks after tail vein injection in Tek Δ/Δ mice (n = 14) compared to WT mice (n = 15). Data is shown as mean ± 95% CI. ***p < 0.001. e, f Size of lung metastatic foci in WT and Angpt1 Δ/Δ MMTV-PyMT mice (e) and B16F10 experimental metastasis model (f). Data is shown as mean ± 95% CI

Fig 3: Effect of Angpt1 on B16F10 cells. a Expression analysis of Angpt1, Angpt2 and Tek in B16F10 melanoma cells compared to lung tissue. b-d Quantification of migration assays (b) and representative images (c) of B16F10 cells in the presence or absence of Angpt1. In (b) every second value vas excluded for clarity. HD – high density, 40,000 cells/well, LD – low density, 20,000 cells/well

Fig 4: Angpt1 deficiency enhances initial stages of distant metastasis. a Quantification of B16F10 cells in the lungs 24 h after tail vein injection in WT (n = 6) and Angpt1 Δ/Δ mice (n = 6). b Quantification of B16F10 cells in the lungs 4 h after tail vein injection in WT and Angpt1 Δ/Δ mice (n = 4–5). c Representative fluorescent image of lung from WT and Angpt1 Δ/Δ mice 4 h after tail vein injection of GFP positive B16F10 cells. d Leakage experiments with cadaverin in WT and Angpt1 Δ/Δ mice, with and without tail vein injection of B16F10 cells. e Measurement of cross-sectional area of capillaries from micrographs of lungs from WT (n = 4) and Angpt1 Δ/Δ (n = 4) mice. f Quantification of microspheres in lungs 4 h after tail vein injection in WT (n = 5) and Angpt1 Δ/Δ (n = 5) mice**p < 0.01, ***p < 0.001. g Western blotting for FAK phosphorylation at Tyr397 in WT and Angpt1 Δ/Δ mice, with and without tail vein injection of B16F10 cells. h Quantification of B16F10 cell attachment to HUVECs with or without pre-treatment with Angpt1

Fig 5: Macrophage polarization is not affected in Angpt1 deficient. Individual lung metastases sizes were not different between WT and Angpt1 Δ/Δ in MMTV-PyMT mice (a) and B16F10 i.v melanoma injections (b), data is shown as mean ± 95% CI. (c) (a) FACS analysis of CD36 and CD206 expression in lung macrophages 3 weeks after B16F10 in WT and Angpt1 Δ/Δ mice. b, c Quantification of total number of macrophages did not show any differences at baseline or 4 h after B16F10 melanoma cell injection