Sox17 is required for normal pulmonary vascular morphogenesis

Abstract

The SRY-box containing transcription factor Sox17 is required for endoderm formation and vascular morphogenesis during embryonic development. In the lung, Sox17 is expressed in mesenchymal progenitors of the embryonic pulmonary vasculature and is restricted to vascular endothelial cells in the mature lung. Conditional deletion of Sox17 in splanchnic mesenchyme-derivatives using Dermo1-Cre resulted in substantial loss of Sox17 from developing pulmonary vascular endothelial cells and caused pulmonary vascular abnormalities before birth, including pulmonary vein varices, enlarged arteries, and decreased perfusion of the microvasculature. While survival of Dermo1-Cre;Sox17Δ/Δ mice (herein termed Sox17Δ/Δ) was unaffected at E18.5, most Sox17Δ/Δ mice died by 3 weeks of age. After birth, the density of the pulmonary microvasculature was decreased in association with alveolar simplification, biventricular cardiac hypertrophy, and valvular regurgitation. The severity of the postnatal cardiac phenotype was correlated with the severity of pulmonary vasculature abnormalities. Sox17 is required for normal formation of the pulmonary vasculature and postnatal cardiovascular homeostasis.

Keywords: Dermo1-Cre; Endothelial; Lung; Sox17; Vascular morphogenesis.

Fig. 1

Sox17 is expressed in endothelial cells in the developing lung. (A–F) Immunofluorescence double labeling for Sox17 (green) and endomucin (red, A–C) or Pecam-1 (red, D–F) was performed on sections from E12.5 (A and D), E14.5 (B and E), and E18.5 (C and F) mouse lungs. Sox17 was detected in the nuclei of cells staining for endothelial cell markers in the developing mouse lung. (G–L) Immunofluorescence for GFP (G and J) and endomucin (H) or Pecam-1 (K) was performed on sections from E18.5 Sox17^GFP reporter knock-in mouse lungs. Sox17-expressing cells labeled with the GFP reporter were co-stained with endothelial cell markers. Nuclei are stained with DAPI (blue). Insets show higher magnification of boxed regions and dotted lines denote the basal surface of the airway epithelium. Pulmonary artery (inset, F; arrowhead, J–L). Peripheral microvasculature (insets, C, I, and L). Red blood cell autofluorescence in panel F was masked using Imaris software. Scale bars 40 μm.

Fig. 2

Deletion of Sox17 in the developing lung mesenchyme causes dilated pulmonary vessels. (A–B) H&E staining of tissue sections from E18.5 mouse lungs shows enlarged pulmonary vessels in Sox17Δ/Δ embryos (B) compared to controls (A). (C–H) Immunostaining for Sox17 on sections from E18.5 control (C, E, and F) and Sox17Δ/Δ (D, G, and H) lungs. Dermo1-Cre mediated deletion of Sox17 results in the loss of Sox17 expression in the majority of arterial, venous, and microvascular endothelial cells in the developing lung. (E–H) Higher magnification images of arteries (F and H) and veins (E and G) shown in C and D. a, Artery; v, vein. Scale bars, 50 μm.

Fig. 3

MicroCT analysis shows pulmonary vascular abnormalities in Sox17Δ/Δ embryonic lungs. (A–B) MicroCT images comprised of a merged stack of 50 2D microCT slices. Panels depict a sequential series through the developing lungs of E18.5 control (A–A″) and Sox17Δ/Δ (B–B″) mice. Pulmonary veins were dilated (red arrowheads) and vascular branching was aberrant in the lungs of Sox17Δ/Δ embryos. (C–D) Static 3D images of the developing pulmonary vascular network in E18.5 control (C) and Sox17Δ/Δ (D) lungs showing dilated pulmonary blood vessels (red arrowheads) and reduced staining of the peripheral microvasculature (red asterisks). PV, pulmonary vein; LA, left atrium.

Fig. 4

Pulmonary vascular endothelial cell differentiation is not affected in Sox17Δ/Δ mouse lungs. Staining for GSL-IB4 (A–B), Pecam-1 (C–D), CD34 (E–F), and EphB4 (G–H) was performed on sections from E18.5 control (A, C, E, and G) and Sox17Δ/Δ (B, D, F, and H) lungs. No differences in staining for these endothelial cell markers were observed between control and Sox17Δ/Δ embryo lungs. (G–H) EphB4-positive veins and EphB4-negative arteries were enlarged in Sox17Δ/Δ mouse lungs. Insets show higher magnification of veins (A–D, G–H) or arteries (E–F). Scale bars, 100 μm. a, Artery; v, vein.

Fig. 5

Sox17 deletion does not influence the expression of genes known to regulate vascular development. Gene expression was assessed by qPCR using mRNA isolated from CD146(+) endothelial cells harvested from E18.5 control (n=7) and Sox17Δ/Δ (n=5) mouse lungs. Sox17 and Camk4 expression was significantly decreased, and Sox7 mRNA was increased in the CD146(+) endothelial cell population from Sox17Δ/Δ lungs compared to controls. Expression of known regulators of vascular morphogenesis and blood vessel size was not affected.

Fig. 6

Postnatal Sox17Δ/Δ mice with pulmonary vascular abnormalites develop enlarged hearts. (A–C) Histological analysis of lungs from 2 week old control (A) and Sox17Δ/Δ (B–C) mice shows variable enlargement of pulmonary blood vessels. arrowheads, arteries; arrows, veins; b, bronchus. (D–I) Immunofluorescence staining for endomucin (red) and alpha-smooth muscle actin (green) was performed on lung sections from 2 week old control (D and G) and Sox17Δ/Δ (E and F, H and I) mice. (D-F) Whole mount staining of thick sections (150 μm) showed that the endomucin-positive microvascular network was variably reduced in the lungs of Sox17Δ/Δ mice. (G-I) Immunofluorescence staining of thin lung sections (10 μm) showed that paucity of the microvasculature was associated with alveolar simplification in severely affected Sox17Δ/Δ mice. (M) The area and volume of the pulmonary vascular network determined from endomucin staining were decreased in Sox17Δ/Δ mice. (J–L) Histological analysis of hearts from 2 week old control (J) and Sox17Δ/Δ (K and L) mice shows variable cardiac enlargement. (N) Ventricle weights were significantly increased in 2 week old Sox17Δ/Δ mice compared to controls. RV, right ventricle; LV+S, left ventricle and septum. The severity of pulmonary vascular abnormalities and heart enlargement in Sox17Δ/Δ mice were well correlated. Scale bars 500 μm (A–C and J–L); 100 μm (D–I). Asterisks indicated statistical significance as determined by Student's t-test (p ≤.05).