Stepwise arteriovenous fate acquisition during mammalian vasculogenesis

Abstract

Arteriovenous (AV) differentiation is a critical step during blood vessel formation and stabilization. Defects in arterial or venous fate lead to inappropriate fusion of vessels, resulting in damaging arteriovenous shunts. While many studies have unraveled the molecular underpinnings that drive AV fate, surprisingly, the spatiotemporal emergence of arteries and veins in mammalian embryos remains unknown. Here, we examine artery and vein specification and differentiation during vasculogenesis. We show that the first intraembryonic vessels formed are arteries, which differentiate in a stepwise manner. By contrast, veins emerge later, progressively forming after embryonic turning. In addition, we demonstrate that hemodynamic flow is not required for arterial specification, but is required for maintenance of select arterial markers. Together, our results provide a first spatiotemporal analysis of mammalian AV cell fate establishment and anatomy, as well as a delineation of a molecular toolkit for analysis of arteries and veins during early vessel development.

Figure 1. Progressive formation of first embryonic vessels: dorsal aortae and vitelline vein primordia

β-galactosidase staining of Flk1-LacZ embryos from E7.5-E8.5 and cartoon schematics representing major arteries and veins during vasculogenesis. (A) Flk1 expression at 0s is restricted to vessels of the extraembryonic tissues or yolk sac. (B–C) At 1s, angioblasts began to align and coalesce into the paired dorsal aortae (red arrows), as well as the cardiac crescent. (D) At 2s, the dorsal aortae form cords that lack a lumen (inset). (E–F) At 5s, the paired aortae are lumenized (inset) and the vitelline veins (blue arrows) have begun to coalesce posterior to the cardiac crescent and sinus venosus. At this stage, the heart begins to beat weakly. (G) By 8s, the primitive heart tube, or endocardium (red asterisk), and sinus venosus (white asterisk) have formed. The heart beats more strongly. (H–I) At 9s, the embryo is initiating turning and the paired aortae are in closer proximity. Vitelline vein cords extending from sinus venosus are longer. (J–L) After the embryo completes turning, vitelline vein cords extend to the tail, while the anterior and posterior cardinal veins begin to appear. acv, anterior cardinal vein; aip, anterior intestinal portal; cc, cardiac crescent; da, dorsal aortae; e, endocardium; pcv, posterior cardinal vein; sv, sinus venosus; vv, vitelline vein. Scale bars represent 200μm.

Figure 2. Stepwise initiation of arterial genes in dorsal aortae prior to turning

In situ hybridization or β-galactosidase staining of established arterial makers from E8.0 to E8.5, as indicated. (A–J) During early vasculogenesis (4s–5s), Cx37, Hey1, and Dll4 are the first arterial genes expressed in the forming dorsal aortae. Ephrin-B2, Cx40, Hey2, Nrp1, Notch1/4 and Jag1 are not detected in endothelium. (A′–J′) At E8.25 (7s–8s), additional arterial genes have initiated, including Cx40 and Notch1/4. (A″–J″) Transverse sections through E8.25 embryos reveal presence or absence of expression in ECs of dorsal aortae. Somite stage for each embryo given at bottom right. Scale bars represent 200μm (A–E); 100μm (A′–E′, F–J′); 25μm (A″–E″, F″–J″).

Figure 3. Venous gene expression is largely absent from endothelium prior to E8.5

In situ hybridization or β-galactosidase staining of established venous markers as indicated from E8.0 to E8.25. (A–E) At E8.0 (4s–5s), expression of venous markers was not observed in distinguishable veins. EphB4 displayed punctate expression along the sinus venosus (inset) while CoupTFII, Nrp2, and APJ were expressed in the somites and lateral plate mesoderm (lpm). Flt4 and Nrp2 were observed in the vitelline vein primordia (blue arrows) while Flt4 was also at the dorsal aortae (red arrows). (A′–E′) At E8.25 (7s–8s), venous expression remained relatively similar for all markers examined, except EphB4, which marked more cells in the sinus venosus. (A″–E″) Transverse sections through E8.25 embryos reveal presence or absence of venous expression in ECs (outlined) of the sinus venosus. Scale bars represent 200μm (A–E); 100μm (A′–E′).

Figure 4. Stepwise acquisition of arteriovenous fate in early vessels

Co-immunofluorescence for VEGFR2 and arterial markers, at E8.25 (A–C) or E9.5 (D–E). Single channels for either Cx40 or Nrp1 or Nrp2 (A′–F′) or for endothelial Flk1 (A″–F″). Note expression of the arterial markers Cx40 (A) and Nrp1 (B) in dorsal aortic ECs, but also sinus venosus and vessels of the head/PNVP (white arrowheads) (A′,B′) at E8.25. Nrp2, by contrast, is strongly expressed in the sinus venosus, and head/PNVP vessels, but not in the dorsal aorta (C,C′). By E9.5, Cx40 (one of the two arterial markers assayed) had become restricted to the aorta (D). Nrp1, however, was still expressed in the venous ECs of the sinus venosus and the head/PNVP (white arrowheads) (D,D′). Nrp2, like Cx40, was restricted in its expression by E9.5 and exhibited expression only in the venous ECs of the sinus venosus. White arrows point to vessels of the head/PNVP; blue arrows point to sinus venosus; red arrows point to dorsal aorta.

Figure 5. Vitelline and cardinal vein formation follows that of the dorsal aortae

β-galactosidase staining of Flk1-LacZ embryos, at stages indicated. Schematics depict major vessels. Arteries, red; veins, blue. (A–C) The vitelline veins form lateral to the dorsal aorta and along the edge of the yolk sac plexus. At 6s, there are a few scattered angioblasts posterior to the vitelline ‘blind-ended’ vessels (white arrow), however by 10s more angioblasts have emerged and the vitelline vein has formed more posteriorly (blue arrow), thus lengthening. (D–F) Following turning, the anterior and posterior cardinal veins can be seen beginning to form as loose, lateral plexus like vessels, connected to the sinus venosus via the common cardinal veins. At 14s, the anterior cardinal vein is distinguishable from the heart to the head, and at 16s, both the anterior cardinal and posterior cardinal veins are longer and sprout intersomitic vessels. acv, anterior cardinal vein; ba, branchial arch; da, dorsal aortae; ccv, common cardinal vein; e, endocardium; isv, intersomitic vessel; pcv, posterior cardinal vein; sv, sinus venosus; vv, vitelline veins; ysp, yolk sac plexus. * denotes forming Duct of Cuvier. Scale bars represent 100μm (A–C) and 200μm (D–F).

Figure 6. Stepwise initiation of venous gene expression in forming vitelline and cardinal vein

In-situ hybridization and β-galactosidase staining of Flk1-LacZ and EphB4-LacZ embryos, at somite stages indicated. (A–C) In situ hybridization with Nrp2 (A), APJ (B), and Flt4 (C) show positive staining in forming vitelline veins (blue arrows). D) Flk1-LacZ embryo, focusing on the developing anterior cardinal vein (black arrow) and sprouting intersomitic veins (white arrow). In this forming vessel, both APJ (E) and Flt4 (F) initiate expression during turning, as angioblasts aggregate and form cords. (G–I) EphB4-LacZ expression initiated in ECs of the sinus venosus following turning. Expression expanded first in the anterior cardinal vein (H), and then the posterior cardinal vein (I) as they formed and became established. acv, anterior cardinal vein; ccv, common cardinal vein; isv, intersomitic vessels; sv, sinus venosus. Scale bars represent 100μm.

Figure 7. Blood flow is required for maintenance of arterial expression of Cx40, but not Dll4

β-galactosidase staining and in situ hybridization on cultured explants show that disruption in flow reduces expression of Cx40, but not Dll4. Explants were cultured for 10 or 18 hours. (A–A″) At 5s, the dorsal aortae were clearly outlined using Flk1-LacZ and the arterial markers Cx40 and Dll4. (B–B″) After 10 hours, the dorsal aortae were still clearly outlined and distinguishable using Flk1 and Dll4, however Cx40 expression decreased significantly. (C–C″) After 18 hours, dorsal aortae expression was still visible using markers Flk1 and Dll4, however Cx40 expression was almost completely absent in the aortae. In explants containing posterior portions of the embryos, similar trends were identified. (D–D″) At 5s, expression of Flk1, Cx40, and Dll4 was robust in the paired aortae. (E–E″) After 10 hours in culture, Flk1 and Dll4 continued to be expressed in the vessels, however Cx40 expression was relatively low and had diminished more rapidly than in the anterior portion. (F–F″) After 18 hours in culture, Flk1 and Dll4 expression remained, albeit the vessels appeared discontinuous, however Cx40 expression was completely lacking. Cc, cardiac crescent; red arrows point to dorsal aortae. Scale bar represents 100μm (A–F″).

Figure 8. Dependence of select arterial gene expression on blood flow

To test the dependence of arterial specification on flow, we analyzed arterial gene expression using in situ hybridization on E8.25 and E9.0 Rasip1 null embryos, which lack vascular lumens and blood flow. Expression of Cx40 was robust in the aortae of Rasip1^+/− at E8.25 (A) and E9.0 (C,C′). However, Cx40 expression in the Rasip^−/− was fainter and discontinuous at E8.25 (B), and subsequently downregulated by E9.0 (D,D′). Whole and (C,D) and transverse sections, E9.0 embryos (C′,D′); red arrows, aorta; black arrows vitelline artery. In contrast, Dll4 expression is not downregulated in Rasip1^−/− aortic cords at E8.25 compared to Rasip1^+/− (E,F) and expression continues at E9.0 (G,G′,H,H′). Transverse sections, E9.0 embryos (G′,H′). Red arrows point to dorsal aortae, note lack of lumens in Rasip1^−/− embryos (D′,H′). Stippled lines represent plane of section for (C′,D′,G′,H′). Scale bars represent 200μm (A,B,E,F) and 100μm (C′,D′,G′,H′).