Redundant Roles for Sox7 and Sox18 in Arteriovenous Specification in Zebrafish
The specification of arteries and veins is an essential process in establishing and maintaining a functional blood vessel system. Incorrect arteriovenous specification disrupts embryonic development but has also been diagnosed in human syndromes such as hypotrichosis–lymphedema–telangiectasia, characterized by defects in blood and lymphatic vessels and associated with mutations in SOX18. Here we characterize the role of sox7 and sox18 during zebrafish vasculogenesis. Sox7 and sox18 are specifically expressed in the developing vasculature, and simultaneous loss of their function results in a severe loss of the arterial identity of the presumptive aorta which instead expresses venous markers, followed by dramatic arteriovenous shunt formations. Our study identifies members of the Sox family as key factors in specifying arteriovenous identity and will help to better understand hypotrichosis–lymphedema–telangiectasia and other diseases.
- hereditary hemorrhagic telangiectasia
Arteriovenous (AV) specification and differentiation are two critical events required for the progression of vascular development and function, as evidenced by diseases such as hypotrichosis–lymphedema–telangiectasia and hereditary hemorrhagic telangiectasia, which have been associated with mutations in SOX18 and endoglin/activin-like receptor kinase-1, respectively.1–3 To study the process of AV specification, zebrafish embryos have proven particularly useful. Following the specification of arterial and venous cell types,4 endothelial cells coalesce into cord-like midline structures and subsequently reshape into tubes.5 However, although several signaling molecules and transcription factors4,6–8 have been implicated in these processes, we are still only beginning to understand their regulation. In an attempt to identify new factors involved in the regulation of AV specification and vasculogenesis, we analyzed the function of sox7 and sox18, which, together with the endodermally expressed sox17, form the Sox-F (Sry-related HMG box) family of DNA-binding proteins.9 Members of the Sox-F family play crucial roles during the formation of definitive endoderm,10 hematopoietic stem cell regulation,11 and cardiovascular development.12 Here we show temporal and spatial overlap of sox7 and sox18 expression and identify functionally redundant roles for these genes during vascular development in zebrafish embryos. Our results demonstrate a novel role for sox7 and sox18 in specifying the molecular identity of endothelial cells in their commitment to arterial tissues during vasculogenesis.
Materials and Methods
In situ hybridization and immuno-histochemistry were performed as described.5,13 The riboprobes used are specified in the online data supplement, available at http://circres.ahajournals.org.
Morpholino Injections and Microangiographs
Procedures are specified in the online data supplement.
Results and Discussion
Expression Analysis for sox7 and sox18
We first examined embryonic expression of sox7 and sox18. RT-PCR expression analysis (Figure I in the online data supplement) revealed that sox7 and sox18 are provided maternally. In situ hybridization showed that sox7 and sox18 transcripts localized to the lateral mesoderm at 12 hours post fertilization (supplemental Figure I). Reminiscent of migrating angioblasts, these presumptive precursor cells localized to the midline during somitogenesis (supplemental Figure I) and finally homed to the endothelium of the axial, head, and intersegmental vessels at 26 hpf (Figure 1A and 1B). Cells expressing sox7 and sox18 are likely endothelial based on the expression pattern of these genes at later stages and the absence of mesodermal sox7 and sox18 expression in cloche mutants and etsrp morphants,8 both of which lack the endothelial lineage (supplemental Figure II). In addition, sox7-expressing cells were found in rhombomeres at 26 hpf (Figure 1A, arrowhead), whereas sox18 expression was observed in the eye and retina (Figure 1B, arrowhead). The expression patterns of zebrafish sox7 and sox18 closely resemble the expression pattern of Sox18 in mice.12
Morpholino Knockdown Analysis of sox7 and sox18
Embryos injected with morpholinos (MOs) targeting sox7 or sox18 individually (two independent MOs for each gene; see supplemental Figure III for specificity tests) did not show any apparent morphological defects, or loss of endothelial cells (supplemental Figure I), or loss of circulation (supplementary Movies 2 to 3). Strikingly, on simultaneous injection of low amounts of both sox7- and sox18-MO, virtually all double knockdowns (dKDs) exhibited a loss of circulation in the posterior part of the embryo, whereas cardiac contractile function was normal (Figure 1E; supplementary Movie 1). Later, blood accumulated in a short circulatory loop near the heart leading to pericardial edema (>2 days post fertilization; not shown). Endothelial cells were specified in sox7/sox18-dKDs as demonstrated by fli1a:gfp expression, but we noticed poor segregation of artery and vein (compare supplemental Figure Ij and Im, insets). In addition, in microangiographs, the major axial vessels in the posterior part of sox7/sox18-dKDs were not filled with dye at 2 days postfertilization (Figure 1C and 1D). Dye injected into the sinus venosus drained from the heart into the posterior cardinal vein (PCV) rather than the dorsal aorta (DA) (Figure 1D, arrow). We conclude that combined loss of sox7 and sox18 function results in a severe disturbance of circulation.
Arteriovenous Specification and Vascular Tube Formation in sox7/sox18-dKDs
To further investigate this phenotype, we analyzed the expression of several molecular markers in sox7/sox18-dKDs compared with uninjected control embryos or silent heart morphants.15 No alteration was detectable in the primitive erythroid lineage marker gata1 (supplemental Figure IV), vegf receptors 2 and 4, or pan-endothelial markers like tie2, cdh5, and fli1a (not shown). However, we observed a dramatic decrease in the expression of arterial markers notch3, ephrinB2a (Figure 2A through 2D), and dll4 (supplemental Figure IV) and a concurrent increase in the expression of venous markers dab2 and flt4 in arterial tissues, such as the DA and intersegmental vessels (supplemental Figure IV and Figure 2C and 2F, respectively). These results suggest a key role for sox7 and sox18 in specifying the arterial fate of endothelial cells. A possible shift in AV identity attributable to the lack of circulation was excluded by analyzing silent heart morphants, which showed no alteration in marker gene expression (Figure 2C and supplemental Figure VI). This demonstrates that sox7 and sox18 are essential regulators of AV identity.
To better understand the lack of lumen formation observed in sox7/sox18-dKDs in microangiographs, we next examined transverse sections of sox7/sox18-dKDs, uninjected control embryos, and silent heart morphants. The nonvascular morphology in sox7/sox18-dKDs (Figure 3C) was completely normal, suggesting a vessel-specific phenotype. Uninjected controls and silent heart morphants exhibited normal segregation and lumenization of axial vessels (Figure 3A, 3B, 3D, and 3E). In all sox7/sox18-dKDs (n=16), we observed stretches of normal and physically separated axial vessels, alternating with regions where only a single PCV was present. At particular locations in sox7/sox18-dKDs, the DA apparently fused with the PCV (Figure 3C and 3F; see also supplemental Figure V). We conclude that the combined loss of Sox7 and Sox18 function disrupts AV specification and leads to severe shunt formation.
Our study provides novel insights into the molecular roles of sox7 and sox18, which are essential to the specification of the molecular identity of the dorsal aorta during embryogenesis and possibly during later stages of life. These findings, for the first time, offer direct insights into the molecular consequences of Sox function in endothelial cells at the in vivo level. Understanding the requirement for Sox7 and Sox18 in the process of arteriovenous specification might help to better understand syndromes such as hypotrichosis–lymphedema–telangiectasia and hereditary hemorrhagic telangiectasia.
We thank G. Soete, B. Hogan, J. Bussmann, J. Korving, J. Peterson-Maduro and L. de Windt for help with the manuscript and the anonymous reviewers for helpful suggestions.
Sources of Funding
This work was supported by a VIDI grant (H.J.D.) and the Royal Netherlands Academy of Arts and Sciences (S.S.-M.).
↵*Both authors contributed equally to this work.
Original received August 30, 2007; resubmission received October 15, 2007; revised resubmission received November 8, 2007; accepted November 14, 2007.