Normal blood vessel morphology in E10.5 Tie1-Cre⁺;Rac1^fl/fl mouse embryos. (A) Whole-mount endomucin staining indicates similar vascular patterns in E10.5 Tie1-Cre^-;Rac1^fl/fl and Tie1-Cre⁺;Rac1^fl/fl embryos. Major vessels including dorsal aorta (arrowhead) and cardinal vein (arrow), perineural vascular plexus (Pn), endocardium (En) and intersomitic vessels (Isv) are present and well developed in the mutants. (B,C) High-magnification views of endomucin-stained (B) perineural vascular plexus and (C) intersomitic vessels reveal comparable patterns of vessel sprouting (dashed arrows) and branching (asterisks). Scale bars: 400 μm in A; 50 μm in B,C.

Embryonic and extra-embryonic blood vessels are normal in E12.5 Tie1-Cre⁺;Rac1^fl/fl embryos. (A) Macroscopic examination of E12.5 Tie1-Cre^-;Rac1^fl/fl and Tie1-Cre⁺;Rac1^fl/fl mouse embryos does not reveal any obvious morphological differences. (B) Whole-mount Pecam1 staining of E12.5 control and mutant forelimbs and intestines reveals comparable microvascular networks. (C) Whole-mount endomucin staining reveals comparable yolk sac vascular networks in both genotypes. Scale bars: 2.5 mm in A; 250 μm in B,C.

Sprouting angiogenesis does not rely on endothelial Rac1 expression in vivo. (A,B) The perineural vascular network was visualised at E12.5 by whole-mount Pecam1 staining on mouse hindbrains. Quantitation of (A) the number of sprouting vessels on the pial side and (B) the microvessel branch-points on the subventricular side in random areas of 0.25 mm² Tie1-Cre^-;Rac1^fl/fl and Tie1-Cre⁺;Rac1^fl/fl hindbrains shows no significant differences. n=8 embryos per genotype. Scale bars: 100 μm.

E13.5 Tie1-Cre⁺;Rac1^fl/fl embryos display gross morphological abnormalities and blood-filled lymphatic vessels. (A) Macroscopic examination shows that 39% of E13.5 Tie1-Cre⁺;Rac1^fl/fl mouse embryos display edema (arrow), tortuous blood vessels (white arrowhead), blood-filled lymphatics (black arrowheads) and haemorrhagic involuting forelimbs (white ↔). Insets show high-magnification views of edema and blood-congested lymphatics. (B) Transverse sections of H&E-stained back skin and forelimb and laminin-stained forelimbs. Arrows, epidermal blisters; arrowheads, haemorrhage; white dashed arrowheads, capillary collapse; open arrowhead, epidermal basement membrane. (C) Merged images of transverse sections double immunostained for laminin (green) and α-smooth muscle actin (α-SMA; red); endothelium (arrow) and mural cell coverage (bracket). Graph shows that aortic diameters are significantly smaller in mutant embryos. *P<0.001, n=6 embryos per genotype. (D) Serial transverse sections through the thoracic region of Tie1-Cre^-;Rac1^fl/fl embryos and of Tie1-Cre⁺;Rac1^fl/fl embryos without and with haemorrhage. Lyve1 identifies the endothelium in jugular lymph sacs (Jls) and endomucin identifies the endothelium in jugular vein (Jv) in both genotypes. Red blood cells (Rbc) are present in lymphatic vessels of mutants but not controls. Scale bars: 1 mm in A; 50 μm in B-D.

Abnormal lymphatic-blood vessel segregation in Tie1-Cre⁺;Rac1^fl/fl embryos. (A) Serial transverse sections through the jugular region of E12.5 Tie1-Cre^-;Rac1^fl/fl and Tie1-Cre⁺;Rac1^fl/fl mouse embryos. Lyve1 and podoplanin staining identify jugular lymph sacs (Jls) in both genotypes. Red blood cells (Rbc) are present in the lymphatic vasculature of mutants but not controls. Juxtaposition (double arrowheads) of jugular lymph sacs to the cardinal vein (Cv) is closer in Tie1-Cre⁺;Rac1^fl/fl embryos than in controls. Endomucin staining of blood vascular endothelium, including that of the cardinal vein, showed no apparent differences between genotypes. Graph of shortest vein-lymphatic vessel distances at comparable levels through the jugular region of E12.5 mutants and controls, showing that the juxtaposition is closer in mutants. *P<0.0001, n=5 embryos per genotype. (B) Unilateral 3D reconstruction of the cardinal vein (red) and lymphatic sac (blue) from E12.5 Tie1-Cre^-;Rac1^fl/fl and Tie1-Cre⁺;Rac1^fl/fl embryos. Snapshots from movies of reconstructions were taken at 0° (top), 45° (middle) and 180° (bottom) of rotational views on the y-axis to illustrate the juxtaposition distance between these two vessels in each genotype. Scale bars: 50 μm in A; 250 μm in B.

Endothelial Rac1 regulates the distribution patterns of budding Prox1-positive cells, Vegf-C156S-mediated motility and Vegfr3 expression. (A) Transverse sections at comparable levels of E11.5 control and Tie1-Cre⁺;Rac1^fl/fl mouse embryos immunostained for Prox1. Prox1 expression was observed in a subset of cardinal vein (Cv) endothelial cells (arrows) and in cells budding off into the mesenchyme (enclosed by black outline). (B) Schematic of the patterning of budding Prox1-positive cells in vivo. Black dot represents the closest point of cardinal vein to the area of budding Prox1-positive cells. W, maximum width; L, maximum length. Graph shows mean ± s.e.m. of width:length ratios. *P<0.005. Scale bars: 50 μm. (C) Percentage of Prox1-positive endothelial cells in the cardinal vein wall and number of Prox1-positive cells in the area of budding were significantly reduced in Tie1-Cre⁺;Rac1^fl/fl embryos (mean ± s.e.m.). *P<0.01. n=9 embryos per genotype. (D) Vegf-C156S-mediated migratory responses of scrambled (Con) or Rac1 siRNA-transfected lymphatic endothelial cells (LECs). Rac1 depletion in LECs modestly increased Vegf-C156S-stimulated cell speed and dramatically decreased persistence (*P<0.001). Bar charts represent mean ± s.e.m. n=175-200 cells per condition. Representative migration traces for Con and Rac1 siRNA-transfected LECs are given. (E) Con and Rac1 siRNA-transfected LEC extracts immunoblotted for Vegfr3 and Rac1 show that Rac1 depletion reduces Vegfr3 expression. Hsc70, loading control. Densitometry results (means ± s.e.m.) of relative protein levels. *P<0.01, n=3.