FOXO1 represses sprouty 2 and sprouty 4 expression to promote arterial specification and vascular remodeling in the mouse yolk sac

Abstract

Establishing a functional circulatory system is required for post-implantation development during murine embryogenesis. Previous studies in loss-of-function mouse models showed that FOXO1, a Forkhead family transcription factor, is required for yolk sac (YS) vascular remodeling and survival beyond embryonic day (E) 11. Here, we demonstrate that at E8.25, loss of Foxo1 in Tie2-cre expressing cells resulted in increased sprouty 2 (Spry2) and Spry4 expression, reduced arterial gene expression and reduced Kdr (also known as Vegfr2 and Flk1) transcripts without affecting overall endothelial cell identity, survival or proliferation. Using a Dll4-BAC-nlacZ reporter line, we found that one of the earliest expressed arterial genes, delta like 4, is significantly reduced in Foxo1 mutant YS without being substantially affected in the embryo proper. We show that FOXO1 binds directly to previously identified Spry2 gene regulatory elements (GREs) and newly identified, evolutionarily conserved Spry4 GREs to repress their expression. Furthermore, overexpression of Spry4 in transient transgenic embryos largely recapitulates the reduced expression of arterial genes seen in conditional Foxo1 mutants. Together, these data reveal a novel role for FOXO1 as a key transcriptional repressor regulating both pre-flow arterial specification and subsequent vessel remodeling within the murine YS.

Keywords: Arterial specification; Dll4; Foxo1; Sprouty.

Fig. 1.

Foxo1^ECKO results in vascular remodeling defects and lethality. (A) qRT-PCR for Foxo1 expression in E8.5 control and Foxo1^ECKO YSs (*P<0.05, **P<0.01). (B,C) FOXO1 protein expression in E8.5 control and Foxo1^ECKO YS (n=3) by western and quantification (*P<0.05). (D,E) Bright-field images of E10.5 littermate control and Foxo1^ECKO embryos. Scale bars: 500 µm. (F-K) Control and Foxo1^ECKO embryos within the YS at E8.5 (F,G), E9.5 (H,I) and E10.5 (J,K). Scale bars: 500 µm. (K) Pericardial edema (arrow); blood pooling in the heart (asterisk). (L-Q) Pecam1 staining in E8.5 and E9.5 control and CKO YSs. Data are mean±s.d.

Fig. 2.

Vascular remodeling defects in Foxo1^ECKO embryos do not result from reduced blood flow. Primitive erythroblasts in circulation in wild-type and CKO embryos were marked by crossing to an ε-globin-GFP transgenic reporter. Representative still images of E8.5 wild-type (A) and Foxo1^ECKO (B), and E9.5 wild type (C) and Foxo1^ECKO (D) embryos. (E-H) Individual blood cells from A-D were tracked and velocity profiles are plotted. (I,J) Quantification of the average blood velocity (Mann–Whitney U-test, **P=0.005). Average heart rates quantified in wild-type and Foxo1^ECKO embryos at E8.5 (K) and E9.5 (L) (Kruskal–Wallis test, **P=0.003). Data are mean±s.e.m.

Fig. 3.

FOXO1 regulates FLK1 expression without affecting other endothelial genes or EC viability prior to blood flow. (A) Expression levels of endothelial genes by quantitative RT-PCR. (B) Immunolabeling for endogenous FLK1 and PECAM1 in control and Foxo1^ECKO YSs. Scale bars: 50 µm. (C) Comparison of Pecam1 expression in MACS-sorted CD31⁻, CD31⁺ and combined control E8.25 YS cells by qPCR. (D,E) Relative Pecam1 (D) and Flk1 (E) expression between wild-type and Foxo1^null E8.25 MACS-sorted CD31⁻ and CD31⁺ YS cells. (F) YSs from control and Foxo1^ECKO at E8.25 DAPI stained and positive for Flk1-H2B::YFP transgene, which marks the EC nuclei (arrowheads). (G) Quantification of YFP⁺ cells relative to total number of cells labeled with DAPI from F. (H,I) Whole-mount phosphor-Histone-H3 (PH3) (H) or activated caspase 3 (I) staining of control and Foxo1^ECKO E8.25 YSs co-labeled with Flk1-H2B::YFP transgene and DAPI. Data are mean±s.d. (n=3). *P<0.05, **P<0.01, ***P<0.001.

Fig. 4.

Arterial marker expression is reduced in Foxo1^ECKO YS. (A) qRT-PCR expression analysis in control and Foxo1^ECKO YSs; arterial (red), venous (blue) and endoderm markers at E8.25. *P<0.05, **P<0.01; data are mean±s.d. (n=3). (B) Co-immunolabeling of eNOS, FLK1 and DAPI in control and Foxo1^ECKO YSs at E8.25. Scale bars: 20 µm.

Fig. 5.

Characterization of arterial defects in Foxo1^ECKO and germline mutants. (A,B,D,E) Using the Dll4-BAC-nlacZ reporter, nlacZ reporter activity was detected in the dorsal aorta (DA) and umbilical artery (UA) in E8.25 littermate control, Foxo1^ECKO and Foxo1-null embryos. Red arrows indicate YS ECs in the posterior region of the YS plexus. (C,F) qRT-PCR for Dll4 mRNA expression in littermate control, Foxo1^ECKO and Foxo1-null embryos and YSs (n>3; *P<0.05, ****P<0.0001). Data are mean±s.d. (G,H,K,L) nlacZ reporter activity in E9.5 littermate control, Foxo1^ECKO and Foxo1-null YS and embryo. VA, vitelline artery (arrows); insets in G and H show YSs only. (I,J,M,N) nlacZ reporter activity in E9.5 littermate control, Foxo1^ECKO and Foxo1-null embryos. EN, endocardium; DA, dorsal aorta; IAV, intersomitic arterial vessels; ACV, arterial cranial vasculature. Scale bars: 200 µm (E8.25); 500 µm (E9.5).

Fig. 6.

FOXO1 regulates Spry2 and Spry4 expression in the YS vasculature. (A,B) qRT-PCR analysis in littermate E8.25 control and Foxo1^ECKO YSs for (A) known FOXO1 targets and (B) Spry family members. (C) qRT-PCR of endogenous Spry1, Spry2, Spry3 and Spry4 expression relative to Foxo1. (D) Quantitative RT-PCR of Spry2 and Spry4 in MACS-sorted E8.25 CD31⁺ and CD31⁻ control and Foxo1^null YS cells. Data are mean±s.d. (n=3). *P<0.05, **P<0.01, ***P<0.001.

Fig. 7.

FOXO1 directly binds to endogenous Spry2 and Spry4 promoters, and represses Spry2 and Spry4 transcription. (A) Genomic locus of mouse Spry2 gene with the FOXO1-binding sites in blue, orange, red and purple. (B,E) FOXO1 ChIP-PCR using E8.25 YS chromatin. (C) Luciferase activity of FOXO1 on the Spry2 promoter in H1299 cells. (D) Genomic locus of mouse Spry4 gene with the FOXO1-binding sites in blue and red. (F) Luciferase activity of FOXO1 on the Spry4 promoter in H1299 cells. *P<0.05, **P<0.01, ***P<0.001. EV, empty vector. Data are mean±s.d.

Fig. 8.

Transient overexpression of Spry4 in ECs partially phenocopies Foxo1^ECKO mutants. (A) Schematic of Spry4 overexpression construct for pro-nuclei injection. (B) Bright-field image of E9.5 non-transgenic (control) and a transgenic embryo (TG). Confocal imaging of a TG embryo (TG′) showing YFP fluorescence in YS. Vessel remodeling in the control YS (arrow). (C,D) qRT-PCR of arterial markers in E8.25 control and TG YSs (C) and embryos (D) (n=3). *P<0.05, **P<0.01 and ***P<0.001. Data are mean±s.d.