Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Jul 15.
Published in final edited form as: Dev Cell. 2013 Jul 3;26(1):45–58. doi: 10.1016/j.devcel.2013.06.007

ETS factors regulate Vegf-dependent arterial specification

Joshua D Wythe 1,2,3,†, Lan T H Dang 4,5,6, W Patrick Devine 1,3,7, Emilie Boudreau 4,5,6, Stanley T Artap 8, Daniel He 1, William Schachterle 3, Didier YR Stainier 3,9, Peter Oettgen 8, Brian L Black 3, Benoit G Bruneau 1,2,3,†, Jason E Fish 4,5,6,†
PMCID: PMC3754838  NIHMSID: NIHMS492229  PMID: 23830865

Summary

Vegf signaling specifies arterial fate during early vascular development by inducing the transcription of Delta-like 4 (Dll4), the earliest Notch ligand gene expressed in arterial precursor cells (aPCs). Dll4 expression precedes that of Notch receptors in arteries, and factors that direct its arterial-specific expression are not known. To identify the transcriptional program that initiates arterial Dll4 expression we characterized an arterial-specific and Vegf-responsive enhancer of Dll4. Our findings demonstrate that Notch signaling is not required for initiation of Dll4 expression in arteries, and suggest that Notch instead functions as a maintenance factor. Importantly, we find that Vegf signaling activates MAP kinase (MAPK)-dependent ETS factors in the arterial endothelium to drive expression of Dll4, as well as Notch4. These findings identify a Vegf/MAPK-dependent transcriptional pathway that specifies arterial identity by activating Notch signaling components, and illustrate how signaling cascades can modulate broadly expressed transcription factors to achieve tissue-specific transcriptional outputs.

Introduction

Arterial and venous blood vessels are anatomically, functionally and molecularly distinct. Establishing and maintaining these separate endothelial cell fates is critical to the proper function of circulatory networks in the embryo and the mature adult (Marchuk, 1998). In mice, angioblasts are specified in the lateral plate mesoderm during gastrulation. Bilateral streaks of VEGFR2/Flk1+-pre-aortic angioblasts, or arterial precursor cells (aPCs), form the dorsal aortae, followed by the formation of the embryonic veins by a distinct group of angioblasts (Chong et al., 2011). Arterio-venous (AV) specification is genetically determined before hemodynamic forces come into play. Hedgehog signaling induces expression of Vegf in a dorsal to ventral gradient in the early embryo (Lawson et al., 2002). Arterial cells are thought to receive a higher concentration of Vegf, which activates p42/p44 MAPK signaling (Hong et al., 2006) and induces the expression and activation of Notch signaling pathway components exclusively in the arterial endothelium (Lawson et al., 2002). Once activated, the intracellular domain of the cleaved Notch receptor (NICD) interacts with its transcriptional co-factor, RBPJk, in the nucleus to induce expression of Hey1/2, Hes1, EphrinB2 (Efnb2) and other downstream arterial genes (Yamamizu et al., 2010). The molecular events following Notch activation that maintain arterial identity are well understood (Swift and Weinstein, 2009), yet the initial transcriptional cues that function downstream of Vegf activation to induce the expression of Notch signaling components are unknown.

Dll4 is the first Notch ligand gene expressed in the arterial endothelium (Chong et al., 2011), and its expression is induced by Vegf (Lawson et al., 2001). Expression of Dll4 is initiated in the dorsal aorta before genes encoding its cognate receptors, Notch1 and Notch4, making Dll4 one of the earliest markers of the arterial lineage (Chong et al., 2011). Significantly, loss of only one copy of Dll4 produces AV specification defects and embryonic lethality in mice (Duarte et al., 2004; Gale et al., 2004; Krebs et al., 2004). Defining the transcriptional program responsible for early Dll4 expression will therefore provide key insights into arterial specification.

We describe the isolation of an arterial-specific enhancer of Dll4, which labels aPCs and differentiated arterial cells. We find that this enhancer is responsive to Vegf/MAPK signaling, and through deletion analysis identify a previously unappreciated role for ETS transcription factors in mediating the Vegf-responsiveness and arterial-specificity of this enhancer and endogenous Dll4. This Vegf/MAPK/ETS pathway also regulates the expression of Notch4, which encodes an arterial-specific receptor of Dll4. These studies provide mechanistic insight into the transcriptional program downstream of the Vegf receptor that mediates arterial specification through the induction of Notch signaling components.

Results

The genomic region 5′ of Dll4 does not drive arterial-specific expression

To understand the transcriptional basis of AV specification, we searched for evolutionarily conserved non-coding regions (ECRs) of the Dll4 locus. The region proximal to the Dll4 transcriptional start-site is regulated by β-catenin signaling through TCF/LEF sites and can also be activated by FOXC1/2 and RBPJk in vitro (Caolo et al., 2010; Corada et al., 2010; Seo et al., 2006). We cloned this 5-kb region (fragment 1, F1), and placed it upstream of a promoter-less β-galactosidase reporter (Figure 1A and S1A). The activity of this reporter construct was tested in transient transgenic mouse embryos, where it failed to direct any arterial expression (Figure 1A, S1A,B). This suggests that this region is not sufficient to mediate the arterial-specific expression of Dll4.

Figure 1. Identification of an intronic enhancer of Dll4 that drives arterial-specific expression.

Figure 1

Open in a new tab

(A) Conservation between murine and opossum Dll4 with location of fragment 1 (F1) and F2 indicated. Transgenic analysis of F1-lacZ and F2-hsp68-lacZ (E9.5) is below. Further analysis of F1-lacZ is shown in Figure S1A,B. (B) In situ hybridization of endogenous Dll4 (top) and expression in Dll4lacZ/+ (middle) and a stable Dll4-F2-hps68-lacZ (F2) reporter line (bottom). Dorsal aorta (arrows). (C) F2 expression in early arterial precursors (aPCs) and in cardiac crescent (CC). NT, neural tube. (D) Transverse sections of F2 expression. DA, dorsal aorta (arrow); CV, cardinal vein (caret). (E) A stable Dll4-F2-E1b:GFP transgenic zebrafish line demonstrates arterial-specific expression. kdrl:ras-mCherry marks all blood vessels. CA, caudal artery; CVP, caudal vein plexus; ISV, intersomitic vessel; DLAV, dorsal longitudinal anastomotic vessel. (F) Cross-section of axial vasculature of F2:GFP zebrafish. PCV, posterior cardinal vein. Scale bars: 500 μm (B), 100 μm (D), 50 μm (E), 10 μm (F). See also Figure S1.

Wnt signaling is not required for early Dll4 expression or artery specification

Endothelial-specific deletion of Ctnnb1, the gene encoding β-catenin, results in embryonic lethality at E12.5 (Cattelino et al., 2003). However, whether AV specification occurs normally in Ctnnb1 loss-of-function mice has not been previously assessed. Over-expression of a dominant-active allele of Ctnnb1 induces Dll4 expression (Corada et al., 2010), leading to the suggestion that Wnt/β-catenin plays an instructive role in arterial specification by inducing Notch signaling. However, we were unable to detect active canonical Wnt signaling in the arterial endothelium at E8.5 or E9.5 using multiple established Wnt reporter lines (BAT-gal, Axin2-d2EGFP, Axin2lacZ) (Figure S1C–E). To determine if Wnt signaling regulates early Dll4 expression, we ablated Ctnnb1 specifically in the endothelium. The dorsal aortae and cardinal vein were morphologically normal at E8.5 (data not shown) and E9.5 (Figure S1F), expression of Dll4 mRNA was unchanged (Figure S1G), and we did not observe any arteriovenous malformations (AVMs) in Cttnb1 mutants at E9.5 (Figure S1H). Collectively, these results demonstrate that Wnt/β-catenin signaling in the endothelium is dispensable for early artery formation and early Dll4 expression, and that the DNA region upstream of the promoter of Dll4 is not sufficient for artery-specific expression.

Identification of a Dll4 enhancer with activity in the developing arterial endothelium and endocardium

Another well-conserved ECR (fragment 2, F2) is located within the third intron of Dll4 (Figure 1A). This region can respond to FOXN4 (Luo et al., 2012), RBPJk/NICD, and β-catenin (Yamamizu et al., 2010) in both in ex-vivo and in vitro reporter analyses, but its in vivo activity has not been assessed. In transient transgenic embryos, F2 drove robust activation of a minimal promoter-lacZ reporter (hsp68-lacZ) in the arterial endothelium and endocardium, similar to endogenous Dll4 mRNA expression and to the β-galactosidase activity of Dll4lacZ/+ embryos (Figure 1B). Analysis of embryos at E7.5–7.75 from multiple stable transgenic founder lines demonstrated that this enhancer labeled aPCs prior to their coalescence into the cord-like structures of the dorsal aorta (Figure 1B,C). Examination of transverse sections confirmed the arterial-specificity of the enhancer from E8.5 through E10.5 (Figure 1D). F2 also drove strong expression in the endocardium, another tissue where Dll4 mRNA is observed (Figure 1B), suggesting that this enhancer recapitulates the entire developmental endothelial expression pattern of endogenous Dll4. F2, like endogenous Dll4, was active in the arteries of the post-natal retina, but it was not active in the vasculature of the adult retina, suggesting that F2 is a developmentally-regulated enhancer (data not shown). To determine whether this enhancer’s functionality was conserved in other vertebrate species, we injected an F2:GFP transgene into zebrafish embryos, and established stable transgenic lines. Expression of GFP was specific to the dorsal aorta, cardinal artery and intersomitic vessels, with very little expression detectable in the posterior cardinal vein or caudal vein plexus (Figure 1E,F). This intronic Dll4 F2 element is therefore a bona fide arterial-specific enhancer.

A 36-bp fragment necessary and sufficient for arterial-specific enhancer activity is regulated by ETS and RBPJk

Comparative genomic analysis revealed that the full-length F2 enhancer contains regions of high conservation among distantly-related mammals (Figure 2A). Using deletion analyses of the 802-bp F2 enhancer, we identified a 350-bp (F2-2) and a 225-bp fragment (F2-4) that both labeled the arterial endothelium, as well as a weak endocardial enhancer (F2-3) that failed to label the arterial endothelium in mice (Figure 2A,B). The proximal 176 bp of the F2 enhancer (F2-5), which is contained in F2 and F2-2, but absent in F2-4, was not sufficient to drive arterial expression (Figure 2A,B). F2, F2-2, and F2-4 share a common 100-bp region that, when deleted in the context of the entire 802-bp enhancer (F2Δ1), abolished arterial activity (Figure 2A,B). A well-conserved 36-bp element within this common region was necessary (F2Δ2) and, when concatamerized head-to-tail in triplicate (F2-6(3X)), was sufficient to direct arterial and endocardial expression (Figure 2A–C). Further characterization revealed that this minimal 36-bp enhancer (F2-6(3X)) was capable of directing robust and reproducible arterial expression at early stages (i.e. E8.5) of vascular development (Figure 2D). Importantly, transient transgenesis revealed that this 36-bp enhancer was also able to drive strong arterial-specific expression in the dorsal aorta and intersomitic vessels (Figure 2E), as well as the endocardium (data not shown), of zebrafish embryos. Venous expression was not observed in embryonic mice or zebrafish (Figure 2C,E).

Figure 2. Isolation of a minimal Dll4 enhancer element that drives arterial-specific expression.

Figure 2

Open in a new tab

(A) Sequence conservation of the F2 enhancer and deletion constructs used in transgenic analyses. Endocardial and arterial expression is indicated. (B) Representative transgenic embryos from panel (A) at E9.5. F2, F2-2, and F2-4 directed strong arterial-specific expression. F2-3 directed weak endocardial (EC) expression. Deletion of a highly conserved 100- and 36-bp region (F2Δ1 and F2Δ2, respectively) abrogated arterial activity of the F2 enhancer. The 36-bp region, F2-6, was sufficient, when arrayed in triplicate, F2-6(3X), to direct arterial expression. DA, dorsal aorta; AA, aortic arch; NT, neural tube; ICA, internal cerebral artery; HT, heart. Scale bar = 500 μm. (C) Transverse section of X-gal stained F2-6(3X)-lacZ embryo at E9.5. DA (arrow); cardinal vein (caret); EC (double arrow). Scale bar = 100 μm. (D) Whole-mount image of F2-6(3X)-lacZ embryo at E8.5. Scale bar = 500 μm. (E) Mosaic expression of F2-6(3x):GFP in the DA and intersomitic vessels (ISVs) of a transient transgenic 30 hpf zebrafish embryo. PCV, posterior cardinal vein. Scale bar = 50 μm. (F) Sequence comparison of F2-6 in human (Hs), mouse (Mm), opossum (Md), chicken (Gg) and zebrafish (Dr). Cis elements are indicated. (G) Each cis element was mutated in F2-6(3X):GFP and transient transgenics were assessed for arterial expression at 24 hpf. The percentage of embryos with expression in arteries (i.e. DA and/or ISV), expression elsewhere, and no expression, are indicated. * indicates a significant difference in arterial expression compared to wild-type (Chi-squared test). See also Figure S2.

The minimal enhancer contains conserved predicted binding sites for ETS, MEF2, E2F and RBPJk transcription factors (Figure 2F). We tested which sites were required for activity in vivo by separately mutating each DNA-binding site in the context of the F2-6(3X):GFP reporter and performing transient transgenesis assays in embryonic zebrafish. At 24 hours post fertilization (hpf), ~75% of embryos injected with the wild-type F2-6(3X) enhancer displayed arterial expression of GFP. Mutation of the E2F element did not affect enhancer activity, but mutation of the MEF2 element resulted in a modest, but significant reduction in arterial expression (Figure 2G). In mice, Mef2a and Mef2c are expressed within the endothelium (Lin et al., 1998; Wang et al., 2003) and loss of Mef2c leads to early embryonic lethality (~E9.5) and produces a range of vascular defects, including an absent or disorganized dorsal aorta (Bi et al., 1999; Lin et al., 1998). Although recombinant MEF2C protein bound to the MEF2 site of F2-6(3X) (Figure S2A), Mef2c is not required for either F2 activity or endogenous Dll4 expression (Figure S2B–C). Conversely, only 5% of zebrafish embryos injected with an ETS mutant F2-6(3X):GFP construct had arterial expression and mutation of the RBPJk-binding site abolished arterial expression (Figure 2G). These results suggest that ETS and RBPJk play a critical role in regulating F2-6(3X) enhancer activity in vivo, and imply that they may regulate endogenous Dll4.

Neither the RBPJk site in the F2 full-length enhancer nor Notch signaling are required for initiation of Dll4 expression

It has been proposed that activated Notch signaling can induce Dll4 gene expression via RBPJk-mediated transcriptional regulation through a feed-forward mechanism (Caolo et al., 2010). Given that the activity of the minimal F2-6(3X) enhancer was dependent upon a putative RBJPk-binding site, we sought to determine if Notch signaling is necessary or sufficient to regulate the activity of the full-length F2 enhancer. Forced induction of Notch signaling throughout the vasculature promotes the arterialization of veins (Krebs et al., 2010). Indeed, endothelial-specific over-expression of the intracellular domain of Notch1 produced AVMs in mouse embryos and expanded F2 reporter activity into the venous endothelium (Figure S2D), suggesting that the enhancer is Notch responsive. However, although RBPJk bound to the F2 enhancer (Figure S2E), mutation of the RBPJk site in the context of the full-length F2 enhancer failed to diminish arterial-specific expression in transient transgenesis analyses in zebrafish or mice (Figure A, B). We next assessed whether Notch signaling is required for F2 enhancer activity. Rbpjk knockdown reduced late (i.e. 26 hpf), but not early (i.e. 20 somites) enhancer activity (Figure 3C), consistent with a role for Notch signaling in the maintenance, but not the establishment of arterial Dll4 enhancer activity. In mice, global loss of Rbpjk, which leads to defects in somitogenesis and neural tube formation (Oka et al., 1995), diminished but failed to abolish F2 reporter activity at E8.25 or E8.5 (Figure 3D). Additionally, in embryos where Rbpjk was ablated specifically in the endothelium – bypassing global morphogenetic defects – the activity of the F2 reporter was unaffected at E8.5 or E9.5 (Figure 3E), demonstrating that RBPJk does not play a cell-autonomous role in initiating Dll4 enhancer activity. Finally, Dll4 mRNA was only moderately reduced in Rbpjk−/− mice compared to wild-type littermates at E8.5 (Figure 3F). Together, these results indicate that Notch signaling is not required for the early arterial expression of Dll4, but is essential for its maintenance.

Figure 3. Notch signaling is not required for early arterial expression of Dll4.

Figure 3

Open in a new tab

(A)Transient transgenesis of wild-type F2:GFP compared to a RBPJk-mutant construct (F2ΔRBPJk:GFP). DA, dorsal aorta; PCV, posterior cardinal vein; ISV, intersomitic vessel. (B) Whole-mount images (left), and transverse sections (right) of wild-type and RBPJk-mutated F2-lacZ transgenic embryos. (C) Expression of F2:GFP in control and rbpjk morphant (MO) embryos, demonstrating normal enhancer activity in the DA (arrow) at 20 somites, but decreased expression at 26 hpf. Expression of kdrl:GFP in the DA was unaffected. (D) F2-lacZ expression was reduced but present in the DA of Rpbjk−/− embryos at E8.5 (Rpbjk+/+, n=25; Rpbjk−/−, n=5) and E8.25 (Rpbjk+/+, n=14; Rpbjk−/−, n=4). (E) F2-lacZ expression was normal at E8.5 (left), and at E9.5 in whole-mount (center) and transverse sections (right) from embryos with endothelial cell-specific loss (ECKO) of Rbpjk, although mutants had a reduced DA diameter (bar) at E9.5, as previously reported (Krebs et al., 2004). E9.5, Rpbjkdel/fl, n=6; RpbjkECKO, n=4; E8.5, Rpbjkdel/fl, n=8; RpbjkECKO, n=3. (F) In situ hybridization shows that Rbpjk is not absolutely required for Dll4 expression at E8.5 (n=3 for each genotype). DA, arrows; asterisk, neural tube. Scale bars: 50 μm (A,C), 500 μm for whole-mounts and 100 μm for sections (B,D,E,F). See also Figure S2.

ETS factors regulate the activity of the Dll4 enhancer

ClustalW analysis revealed the presence of nine perfectly conserved minimal ETS elements (ETS sites A–G), including the site within F2-6(3X) (ETS-B) (Figure S3A). This was intriguing since several ETS family members are key regulators of endothelial development (De Val and Black, 2009). EMSA confirmed that six of these sites competed for binding of the ETS1 DNA binding domain (DBD) to a known ETS element (Figure S3B). Among these six sites, ETS site B (ETS-B) is present in all arterial-positive constructs (F2, F2-2, F2-4, and F2-6), and corresponds to a canonical ETS-1 binding site (cETS-1). ETS-B also directly and specifically bound recombinant ETS1-DBD (Figure S3C), and ETV2, an ETS family member (Figure S3D).

To determine the functional requirement of the ETS binding sites in vivo we generated transient transgenic zebrafish harboring the full-length F2 enhancer driving GFP in which ETS-B or all six of the ETS1 DBD-binding ETS sites were mutated. Mutation of the six ETS binding sites abolished F2 enhancer activity, and mutation of the ETS-B site alone severely reduced enhancer activity in arteries (Figure 4A); confirming the importance of this single ETS element. Given the necessity of ETS binding sites for enhancer activity, we wondered whether ETS factors were sufficient to induce Dll4 expression. Thus, we next determined whether over-expression of ETS factors was sufficient to induce expression of endogenous DLL4 in human umbilical vein endothelial cells (HUVEC), a venous cell type that expresses low levels of DLL4. Indeed, over-expression of two different endothelial-enriched ETS family members, ETV2 or ERG, induced endogenous DLL4 expression (Figure 4B). Induction of DLL4 was accompanied by binding of V5-ERG to the endogenous region of DLL4 corresponding to F2 (i.e. intron 3), but not to a distal region of the DLL4 locus, as assessed by ChIP (Figure 4C). Furthermore, we detected high levels of endogenous ERG occupancy at intron 3, but not a distal site in the Dll4 locus, in an arterial cell line (bovine aortic endothelial cells (BAECs)) (Figure 4D). By comparison, NICD was only modestly enriched at intron 3 in BAECs (Figure 4D). Collectively, these results demonstrate that ETS factors can induce expression of endogenous DLL4 mRNA, and that ETS factors directly occupy the endogenous DLL4 genomic region corresponding to F2 in arterial cells.

Figure 4. Dll4 is regulated by ETS factors.

Figure 4

Open in a new tab

(A) Representative images (left) and quantification (right) of wild-type, ETS-B and ETS-6x mutant F2:GFP embryos at 48 hpf. ISV, intersomitic vessel; DA, dorsal aorta; PCV, posterior cardinal vein. * indicates a significant difference in arterial expression compared to wild-type (Chi-squared test). (B) Expression of endogenous DLL4 in HUVEC electroporated with ETV2 or ERG expression constructs. (n=3, ETV2; n=5, ERG). (C) V5-ERG expression in electroporated HUVEC (left). ChIP in V5-ERG-electroporated HUVEC. Fold enrichment (V5 vs IgG) was measured at the Dll4 enhancer (intron 3) and Dll4 exon 11 (n=2). (D) ChIP in BAEC for RNA Polymerase II (RNA Pol II), ERG or the Notch intracellular domain (NICD) (n=3). * indicates significant enrichment over IgG control. (E) Expression of F2:GFP (arterial) and kdrl:GFP (pan-endothelial) in erg;fli1a morphants. Representative images (left) and quantification (n=5–10) (right). (F) Endogenous dll4, ephrinb2a, hey2 and kdr in control and erg;fli1a morphants assessed by qPCR (n=5–10 individual embryos). (G) Notch activity (Tp1:GFP) was diminished in the dorsal aorta of ~30% of erg/fli1a morphants. (H) Levels of arterial markers and a pan-endothelial marker (Pecam1) were quantified by qRT-PCR in E8.5 embryos (n=3 for wild-type, n=2 for ErgΔEx4/ΔEx4 embryos). (I) Endogenous Dll4 mRNA is down-regulated in the DA of ErgΔEx4/ΔEx4 embryos (arrows; n=3 embryos per genotype) at E9.5. (J) Efnb2 section in situ hybridization at 9.5 shows down-regulation of Efnb2 in the DA (arrows) of ErgΔEx4/ΔEx4 embryos (Erg+/+, n=3; ErgΔEx4/ΔEx4 n=2). Scale bars: 50 μm (A,E,G), 500 μm (I), 100 μm (J). All data is mean ± SEM. See also Figures S3 and S4.

To further define the role of ETS factors in the regulation of Dll4 expression we performed loss- and gain-of-function experiments in mouse and zebrafish. Fli1a and Erg are closely related ETS factors that play redundant roles in regulating angiogenesis during zebrafish vascular development (Liu and Patient, 2008). Combined morpholino knock-down of these two ETS factors decreased F2:GFP expression at both early and late stages of vascular development (Figure 4E), and endogenous dll4, efnb2a and hey2 expression were also diminished (Figure 4F). However, kdrl:GFP expression (a pan-endothelial marker) was unaffected (Figure 4E), as was kdr mRNA (Figure 4F). Conversely, injection of ERG mRNA increased F2:GFP expression within the dorsal aorta and in what appeared to be the forming posterior cardinal vein at early stages of arterial specification (i.e 20 somite-stage), and also promoted premature sprouting of intersomitic vessels (Figure S4A). However, expanded enhancer activity in the vein was not observed at later stages of vascular development (i.e. 24 hpf; not shown). Since Notch target genes (i.e. efnb2a, hey2) were decreased in erg/fli1a morphants, we assessed Notch transcriptional activity using a Notch reporter line (Tp1:GFP) (Parsons et al., 2009) in an erg/fli1a loss-of-function setting. Strong Tp1:GFP expression in the dorsal aorta and intersomitic vessels was observed in 94% of control embryos (n=83) compared to only 67% of erg/fli1a morphants (n=165), suggesting that arterial Notch signaling was compromised in a proportion of ETS morphants (Figure 4G).

Erg splice variants in mice, driven from a translational start site in Exon 4, are specifically expressed in the endothelium, and Erg Exon 4 null mouse embryos (ErgΔEx4/ΔEx4) die early in development (E10.5–11.5) with severe vascular remodeling defects in the yolk sac and impaired vascularization of the central nervous system (Vijayaraj et al., 2012). We found that the dorsal aortae formed normally in ErgΔEx4/ΔEx4 embryos (Figure S4B), but that Dll4 expression was modestly reduced at E8.5 (Figure 4H) and substantially down regulated at E9.5 (Figure 4I), confirming a role for this ETS family member in regulating Dll4 mRNA expression. Interestingly, several other arterial markers were also reduced in ErgΔEx4/ΔEx4 embryos at E8.5 (Figure 4H), and levels of Efnb2 were modestly reduced at E8.5 (Figure 4H; Figure S4C) and E9.5 (Figure 4J), but were severely altered at E10.5 (data not shown). However, we did not find evidence of gross morphological AVM defects in ErgΔEx4/ΔEx4 embryos at E9.5 (Figure S4B,D), and only rarely observed shunts in erg/fli1a morphant zebrafish embryos (data not shown). We take these results to indicate that ETS factor redundancy may functionally compensate for the loss of ERG or Erg/Fli1a within the endothelium. Together these data define a role for ETS factors in the regulation of the arterial-specific expression of Dll4, and indicate that ETS factors are involved in the specification of arterial fate.

Vegf signaling regulates the expression of Dll4 through the induction and activation of ETS factors

Arterial specification is largely Hh- and Vegf-dependent (Lawson et al., 2002; Vokes et al., 2004), and Dll4 expression is regulated by Vegf levels (Coultas et al., 2010; Liu et al., 2003). We therefore assessed whether F2 enhancer activity was Vegf-responsive. Vegfa knockdown in zebrafish dramatically reduced F2 arterial enhancer activity (Figure 5A,B), similar to endogenous dll4 mRNA (Figure 5C). However, the expression of a pan-endothelial marker (kdrl:GFP) in the axial vessels and levels of kdr and kdrl mRNA were largely unaffected (Figure 5A–C); confirming that vegfa does not regulate vasculogenesis in zebrafish. Global over-expression of Vegfa, achieved by co-injection of mRNAs encoding Vegfa121 and Vegfa165, dramatically increased F2 activity in arterial cells (Figure 5D). In addition, F2 activity was expanded into the vein, confirming that Vegf levels regulate the arterial-specific expression of Dll4. Additionally, we examined F2-lacZ activity in two mutant mouse backgrounds that affect Vegf signaling. Hh-mediated activation of Smoothened (SMO) is required for expression of Vegf in the rostral somites at E8.5 and loss of Smo leads to disorganized or collapsed anterior (but not posterior) dorsal aortae by E8.5; although Flk1+ cells are specified and present in normal numbers (Vokes et al., 2004). Significantly, Smo mutants have reduced Dll4 mRNA in the anterior, but not posterior, dorsal aorta at E8.5 (Coultas et al., 2010). Correspondingly, we found that F2 enhancer activity was reduced in the anterior, but not the posterior dorsal aorta of Smo mutants at E8.25, and activity was abolished in this region of the dorsal aorta by E8.5 (Figure S5A). Additionally, enhancer activity was absent in Vegfr2−/− embryos (Figure S5A), although this is likely the result of reduced vasculogenesis (Shalaby et al., 1995).

Figure 5. Vegf regulates the Dll4 enhancer through an ETS element.

Figure 5

Open in a new tab

(A) Expression of F2:GFP (arterial) and kdrl:GFP (pan-endothelial) in control or vegfa morphant embryos. Lateral dorsal aorta (arrows); dorsal aorta (DA). Scale bar = 50 μm. (B) Quantification of GFP intensity in the DA (n≥10 per group). (C) Expression (qRT-PCR) of dll4 (arterial), kdr and kdrl (pan-endothelial) in control and vegfa morphant embryos (n≥3 per group). (D) Over-expression of Vegfa enhanced expression of F2:GFP in arterial cells and expanded expression to venous cells. CA, caudal artery; CVP, caudal vein plexus. Quantification is shown (n=6–8 embryos per group). Scale bar = 50 μm. (E) Expression of DLL4 (qRT-PCR) in BAECs treated with VEGF or VEGF receptor inhibitor (SU5416) for 24 h (n=3). (F) Relative F2:luciferase activity in BAEC treated with VEGF or SU5416. A representative experiment with 3 technical replicates is shown. * indicates a significant difference in luciferase activity compared to all other constructs. # indicates a significant difference between RBPJk MT and ETS-B MT, 6x-ETS MT and RBPJk/ETS-B MT. (G) ChIP assays for RNA Pol II and ERG in BAEC treated with VEGF or SU5416 for 24 h (n=3). (H) Western blotting demonstrating equal expression of ERG protein in BAECs treated with VEGF or SU5416 for 24 h. (I) Intracellular localization of ERG was unchanged by VEGF treatment. Scale bar = 20 μm. (J) Expression of ERG was elevated in arterial cells (HUAEC) compared to venous cells (HUVEC). All data is mean ± SEM. See Figure S5.

To determine which cis elements within F2 mediate this Vegf responsiveness we performed luciferase assays in cultured arterial endothelial cells treated with VEGF or the VEGF receptor inhibitor, SU5416. Endogenous DLL4 mRNA was elevated in cells treated with VEGF compared to cells treated with SU5416 (Figure 5E). While wild-type F2-luciferase activity was enhanced upon VEGF-treatment, mutation of ETS-B abrogated Vegf responsiveness and this reduction was only slightly more pronounced in the ETS-6x mutant (Figure 5F). While mutation of the RBPJk site modestly reduced enhancer activity, this reduction was not as severe as mutation of the ETS elements. This suggests that the Vegf responsiveness of the Dll4 enhancer is largely ETS-dependent.

Control of ETS activity by Vegf signaling could be mediated by alterations in the expression and/or activity of ETS factors. We first assessed whether the recruitment of ETS factors to the enhancer at the endogenous DLL4 locus was dependent on Vegf signaling. RNA Polymerase II occupancy at the DLL4 enhancer in cultured arterial cells was enhanced in VEGF-treated cells compared to cells treated with VEGF inhibitor (Figure 5G), a finding consistent with transcriptional regulation of DLL4. Importantly, ERG occupancy of the enhancer was also increased in the presence of Vegf signaling (Figure 5G). However, Vegf treatment did not alter ERG protein levels (Figure 5H), or ERG subcellular localization in arterial cells (Figure 5I and Figure S5B). Interestingly, ERG protein levels were elevated in lysates from cells of arterial origin (human umbilical artery endothelial cells (HUAEC)) compared to corresponding venous cells (HUVEC) (Figure 5J), and expression of ERG was moderately enriched in the dorsal aorta compared to the cardinal vein in vivo (Figure S5C), suggesting that ERG is enriched in arterial cells. Although ERG levels were not affected by short-term inhibition of Vegf signaling in cultured arterial cells, the expression of several ETS factors are Vegf-responsive in vitro (Ghosh et al., 2012; Heo et al., 2010). Therefore, we examined the levels of endothelial-enriched ETS factors (Liu and Patient, 2008) to determine whether Vegf regulated their expression during vascular development. Compared to controls, qRT-PCR analyses revealed that levels of elf2a, elf2b, elk4, erg, ets1a and fli1a were significantly reduced in vegfa morphants (Figure S5D,E). In contrast, the expression of etv2, which acts early in the gastrulation-stage mesoderm to specify endothelial cells (Lee et al., 2008), was unaffected by Vegf inhibition (Figure S5D,E). Thus, expression of several endothelial ETS factors is modulated by Vegf signaling in vivo, and their decreased expression coincides with a reduction in dll4 expression.

To determine if ETS factors function downstream of Vegf signaling to induce Dll4 expression, we over-expressed ERG in vegfa morphants. Co-injection of ERG mRNA partially rescued Dll4 F2 enhancer activity, sprouting of ISVs, and endogenous dll4 expression in vegfa morphant embryos (Figure 6A,B). Interestingly, F2 enhancer activity was expanded into the vein in ERG over-expressing embryos. Conversely, knock-down of erg and fli1a (using a sub-phenotypic dose of morpholino) prevented the Vegf-mediated induction of F2:GFP in both the artery and the vein (Figure 6C). Increased expression of endogenous dll4 (Figure 6D) and efnb2a (Figure S6A) induced by Vegf over-expression was also blunted in erg/fli1a morphants. Collectively, our results show that Vegf regulates ETS factors in arterial endothelial cells by inducing their expression and recruitment to DNA. Vegf-regulated ETS factors in turn control F2 enhancer activity, endogenous Dll4 expression and contribute to arterial specification.

Figure 6. ETS factors act downstream of Vegf to induce dll4 expression.

Figure 6

Open in a new tab

(A) ERG mRNA (200 pg) injection partially rescued F2:GFP expression and sprouting (arrows) in vegfa morphants at 26 hpf. Representative images (left) and quantification (right, n=10–12). (B) dll4 mRNA expression (normalized to kdr expression) was assessed by qRT-PCR in individual 26 hpf embryos injected with ERG mRNA and/or vegfa morpholino (n=4). (C) Partial knock-down of erg and fli1a with a sub-phenotypic dose of morpholino inhibited the induction of F2:GFP expression in both the dorsal aorta (DA) and posterior cardinal vein (PCV) in Vegfa121/165 mRNA-injected embryos. Arrows indicate collapsed PCV. Quantification of 10-13 embryos per group is shown. (D) Endogenous dll4 mRNA was assessed by qRT-PCR in individual embryos at 24 hpf (n=5). All data is mean ± SEM. See Figure S5A.

The Vegf/MAPK pathway activates an arterial transcriptional program through activation of ETS factors

Vegf-mediated activation of the p42/p44 MAPK and PI3K signaling pathways are known to promote or antagonize arterial specification, respectively (Deng et al., 2013; Hong et al., 2006; Ren et al., 2010). Pharmacological inhibition of p42/p44 MAPK signaling in arterial cells in vitro decreased DLL4 expression (Figure 7A). An arterial-enriched Notch receptor gene, NOTCH4, was also reduced, but NOTCH1 and venous gene expression (EPHB4 and COUP-TFII), were unchanged (Figure 7A). Consistent with decreased Notch ligand (ie. DLL4) and receptor (i.e. NOTCH4) levels, expression of a Notch-activated transcription factor gene, HES1, and a Notch-dependent arterial marker, EFNB2, were also reduced (Figure 7A). These data suggest that MAPK signaling is required for the maintenance of Notch signaling in arterial endothelium. In contrast, inhibition of PI3K activity with the small molecule, LY29004, enhanced expression of DLL4, NOTCH4, HES1 and EFNB2 (data not shown). Since NOTCH4 was regulated similarly to DLL4 we examined ECRs within the NOTCH4 locus for conserved ETS elements. An ECR within intron 1 of NOTCH4 contained multiple conserved ETS sites, and this region had high ERG occupancy in arterial cells, as assessed by ChIP (Figure 7B). Over-expression of ERG in HUVEC induced NOTCH4 mRNA, suggesting that NOTCH4 is an ETS factor-regulated gene (Figure 7C). Similar to DLL4, NOTCH4 expression was down-regulated upon Vegf inhibition (Figure 7D), and this was accompanied by a decrease in ERG recruitment to intron 1 (Figure 7E). We also found that ERG recruitment to DLL4 intron 3 and NOTCH4 intron 1 was reduced in cells treated with a MAPK inhibitor (Figure 7F), suggesting that ERG recruitment to these enhancers is MAPK-dependent. Accordingly, Vegf-induced Dll4 F2-luciferase activity was suppressed by MAPK inhibition in arterial cells and mutation of the ETS-B site largely abrogated this responsiveness (Figure 7G). In contrast, inhibition of PI3K signaling enhanced Vegf-induction of F2-luciferase activity, and this was also dependent on the ETS-B site (Figure S6B), suggesting that PI3K inhibition may enhance ETS activity. To determine whether MAPK regulates Dll4 in vivo, we inhibited MAPK signaling beginning at the 10-somite stage and found a significant decrease in F2:GFP expression at 28 hpf (Figure S6C,D). Expression of endogenous dll4 and hey2 mRNA was also reduced, while efnb2a levels were modestly decreased (Figure S6E). However, kdrl:GFP and kdr expression were unaffected (Figure S6D,E). In agreement with the MAPK inhibitor data, over-expression of dominant active MEKK in endothelial cells resulted in ectopic expression of F2:GFP within venous cells in vivo (Figure S6F), implying a cell-autonomous role for MAPK signaling in the regulation of arterial Dll4 expression.

Figure 7. Vegf/MAPK signaling activates arterial genes by promoting ETS factor binding.

Figure 7

Open in a new tab

(A) Expression of arterial genes (DLL4, NOTCH4, EFNB2) and the Notch target gene, HES1, were decreased in VEGF-treated BAEC co-treated with the MAPK inhibitor, U0126 (n=4). NOTCH1 and vein markers were unaffected. (B) An ECR (human reference genome) present in intron 1 of NOTCH4 contains multiple conserved ETS elements and binds ERG, as assessed by ChIP assay (n=6). (C) Over-expression of ERG in HUVEC increased NOTCH4 expression by qRT-PCR (n=4). (D) Inhibition of Vegf signaling decreased NOTCH4 expression, but had no effect on NOTCH1 (n=3). (E) Inhibition of Vegf signaling decreased ERG binding to intron 1 of NOTCH4 (n=5). (F) Inhibition of MAPK signaling decreased ERG binding to DLL4 intron 3 and NOTCH1 intron 1 (n=3). (G) Inhibition of MAPK decreased F2:luciferase activity and mutation of the ETS-B site similarly decreased activity and abrogated MAPK-responsiveness. Shown is a representative experiment, triplicate determinations. (H) Model of Dll4 regulation during artery development. At early stages of vasculogenesis (left), ETV2 is expressed independent of Vegf signaling in angioblasts and activates the Dll4 enhancer. At later stages of development, activity of other ETS factors, such as ERG are induced downstream of Vegf/MAPK signaling in the arterial precursor cells (aPCs) (middle), and expression of Dll4 and Notch4 is activated. Once Dll4 and Notch receptors are both expressed, continued Notch signaling maintains Dll4 enhancer activity and activates and maintains the expression of other arterial genes (right). All data is mean ± SEM. See Figure S6.

Discussion

Hh/Vegf-dependent activation of Dll4 expression and Notch/RBPJk activity is a prerequisite for early artery specification and maintenance in the developing vertebrate embryo (Swift and Weinstein, 2009). The MAPK pathway activates the arterial program downstream of VEGFR2, while PI3K signaling antagonizes the MAPK pathway and arterial fate (Deng et al., 2013; Hong et al., 2006; Ren et al., 2010). However, the transcriptional mediators that are activated downstream of VEGFR2/MAPK that induce Dll4 expression, and hence Notch activation, remain unknown. Herein we report the identification of an arterial-specific enhancer (F2) of Dll4, one of the earliest markers of the coalescing aPCs that form the dorsal aortae. The activity of the Dll4 F2 enhancer described here recapitulates endogenous Dll4 expression in the arterial and endocardial endothelium of the developing embryo. Functional analysis of this enhancer reveals that a minimal 36-bp DNA element, when concatenated, can drive arterial-specific expression, and through systematic mutation of the cis elements in this enhancer, we identify a previously undescribed role for the ETS family of transcription factors in the Vegf- and MAPK-dependent initiation of Dll4 expression (Figure 7H). Critically, Notch4 is also regulated by this Vegf/MAPK/ETS pathway, suggesting that this genetic network may coordinately activate expression of Notch signaling components to initiate Notch signaling in the early artery.

Previous work implicated FOXC1/2 and β-catenin in the regulation of Dll4 expression (Corada et al., 2010; Seo et al., 2006; Yamamizu et al., 2010). However, our results argue against these factors directing the arterial-specific expression of Dll4. In the case of Foxc1/2-dependent regulation of Dll4, the conclusions rely largely upon the phenotype of Foxc1−/−;Foxc2−/− mice, which have AVMs and reduced arterial gene expression. However, these embryos also completely lack somites 1–8 at E8.5 (Seo et al., 2006), which are a critical source of Hh-induced Vegf that is required for artery specification (Coultas et al., 2010). Similarly, while prior studies suggested that canonical Wnt/β-catenin signaling is active in early arteries (i.e. E9.5), and that β-catenin drives Dll4 promoter activity (Corada et al., 2010), we find no evidence of active canonical Wnt/β-catenin signaling in early arteries (i.e. E8.5 and E9.5) and Dll4 expression in the dorsal aorta is intact when β-catenin is deleted from the endothelium, suggesting that β-catenin is dispensable for early artery specification. Finally, the genomic fragment that is located just upstream of Dll4 – where β-catenin/TCF/LEF and Foxc1/2 have been suggested to function – does not direct arterial-specific expression.

Vegf-mediated activation of Notch signaling, via NICD/RBPJk transcriptional activity is necessary and sufficient to induce downstream transcription factors, such as Hey2 and other arterial markers, such as Efnb2 (Swift and Weinstein, 2009). However, we postulate that Notch signaling is not required for the initiation of Dll4 expression and primitive dorsal aortae formation. Dll4 is the first Notch ligand expressed in the forming arteries in the mouse, and Dll4 expression (first detected at E8.0) precedes that of the genes encoding arterial Notch receptors, Notch1 and Notch4 (detected at E8.25) (Chong et al., 2011); calling into question how endothelial cell-autonomous Notch signaling could initiate Dll4 expression. Indeed, Notch1−/−;Notch4−/− mice have normal levels of Dll4 at E9.5 (albeit in an abnormal expression pattern), and the dorsal aortae are still formed (Krebs et al., 2000). Similarly, dorsal aortae are present when Rbpjk is deleted specifically in the endothelium (Krebs et al., 2004). While the arterial enhancer of Dll4 that we isolated (F2) contains a conserved functional binding site for RBPJk, we provide several lines of evidence demonstrating that Notch signaling is not required for the initial arterial expression of Dll4. First, mutation of the RBPJk site in the context of the full-length enhancer does not affect its arterial-specificity in mice or zebrafish embryos. Second, global loss of Rbpjk in embryonic mice or zebrafish fails to abolish enhancer activity early in arterial development. Third, early arterial expression of endogenous Dll4 mRNA is present, albeit at reduced levels, in Rbpjk−/− embryos. However, endothelial-specific ablation of Notch signaling does not diminish F2 activity at E8.5 or E9.5, suggesting a possible non-endothelial role for Notch signaling in regulating Dll4. Interestingly, Dll4 enhancer activity in the dorsal aorta is greatly reduced in later-stage rbpjk morphant zebrafish embryos, suggesting that Notch signaling maintains Dll4 expression. Collectively, our findings demonstrate that initiation of Dll4 expression during artery specification is Notch-independent. We propose that Dll4 expression is initiated during vasculogenesis (through Notch-independent mechanisms; see below), and that Notch signaling subsequently becomes activated by a Vegf/MAPK/ETS pathway that regulates both Dll4 and Notch4 (Figure 7H). Subsequently, Dll4 expression is maintained through a Notch-dependent positive feedback loop to sustain artery specification, as has been previously suggested (Caolo et al., 2010).

A highly conserved ETS site (i.e. ETS-B) is responsible for the majority of the activity of the arterial enhancer of Dll4, including its responsiveness to Vegf/MAPK signaling. How then do ETS factors, many of which are uniformly expressed in the vasculature, regulate the arterial specificity of Dll4? The Dll4 enhancer can be bound (Figure S3D) and activated (data not shown) by the angioblast-enriched ETS family member ETV2 (Lee et al., 2008). Interestingly, the first angioblasts to arise during murine development appear to be arterial (Chong et al., 2011). We therefore posit that ETV2 initiates Dll4 expression during vasculogenesis (Figure 7H). Our data further suggests that a second wave of Vegf-dependent ETS factors appears to reinforce Dll4 expression in the forming dorsal aortae (Figure 7H). We find that Vegf regulates the expression of several zebrafish ETS factors in the endothelium, including elf2a, elf2b, elk4, erg, ets1 and fli1a. Furthermore, ERG appears to be enriched in arterial cells, implying that some ETS factors may be arterial-enriched. More importantly, we discover that the recruitment of ERG to the Dll4 F2 enhancer and an ECR within Notch4 is induced by Vegf/MAPK signaling. In this case, differential ERG occupancy is not mediated by changes in total ERG levels or subcellular localization, suggesting that Vegf/MAPK signaling enhances the DNA binding activity of ERG. We have also observed Vegf-dependent recruitment of ETS1 to these same enhancers (unpublished results mined from data in (Zhang et al., 2013)), suggesting that additional ETS family members may be regulated similarly to ERG, and co-operatively control arterial specification. Importantly, the enhanced activation of the Dll4 enhancer and endogenous dll4 in response to vegfa over-expression can be attenuated by knock-down of erg and fli1a, suggesting that ETS factors are necessary for the Vegf-induced expression of Dll4. We also find reduced Dll4 expression in the dorsal aorta of ErgΔEx4/ΔEx4 mice and in erg/fli1a morphant zebrafish embryos, but failed to observe AVMs, suggesting that other ETS factors compensate in the specification of the early artery.

Vegf is known to selectively activate MAPK signaling in arteries compared to veins during embryogenesis (Corson et al., 2003; Hong et al., 2006). Interestingly, several ETS family members are phosphorylated by MAPKs (Hill et al., 1993; Murakami et al., 2011; Petrovic et al., 2003), and these modifications are known to affect their interaction with other transcription factors as well as their binding to DNA (Hollenhorst et al., 2011). Future experiments will examine ETS factor phosphorylation downstream of Vegf/MAPK signaling, and whether ETS factors functionally interact with other transcription factor families to establish arterial fate. Finally, we also demonstrate that a Vegf/MAPK/ETS pathway regulates Notch4, suggesting that this pathway lies upstream of induction of Notch signaling in the arterial endothelium. Indeed, we observe reduced Notch-regulated gene expression and Notch-dependent reporter activity in embryos with reduced ETS activity. Interestingly, Notch4 is not as sensitive as Dll4 to ERG over-expression, which may in part explain the delay in Notch4 expression during artery formation (Chong et al., 2011).

In summary, we have uncovered a genetic pathway that integrates Vegf signaling, ETS-dependent transcriptional regulation, and the induction of one of the earliest, essential ligands for artery specification and Notch pathway activation, Dll4. This same pathway also regulates the expression of Notch4, an arterial-specific receptor for Dll4, and appears to initiate Notch signaling in the artery. These findings may provide insight into congenital defects in AV specification and maintenance and suggest novel approaches to direct endothelial cells towards the arterial lineage.

Experimental Procedures

Bioinformatic analyses, cloning, mutagenesis, generation of transgenic mice and mouse lines used

See Supplemental Experimental Procedures for details.

Mouse experiments

All mouse protocols were approved by the Institutional Animal Care and Use Committee at UCSF and Harvard Medical School. Histology and in situ hybridization (Dodou et al., 2003; Wythe et al., 2011), ink injections (Krebs et al., 2004) and CD31 immunofluorescence (Coultas et al., 2010) were performed as described elsewhere.

Electrophoretic mobility shift assay

DNA binding assays were performed as described previously (Dodou et al., 2003). See Supplemental Experimental Procedures for details.

Cell Culture, Immunofluorescence, Luciferase Analysis and ChIP

Human umbilical vein endothelial cells (HUVECs) (ScienCell) and Bovine arterial endothelial cells (BAECs) (Lonza) were grown in endothelial media (ScienCell). BAECs were treated with 50 ng/ml recombinant Vegf-A165 (R&D Systems), 0.5 μM SU5416 (Sigma), 20 μM U0126 (Invivogen), or 10 μM LY29004 (Cell Signaling) for 24 h. Immunofluorescence of ERG and phalloidin was performed as described (Fish et al., 2011) using ERG antibody (Santa Cruz, C-20, 1:100 dilution). pCS2-6xMYC-Etv2 (De Val et al., 2008) has been described and pCMV-Sport6-ERG was from Open Biosystems (IMAGE clone 6052140). Cloning of pCS2-V5-ERG is described in Supplemental Experimental Procedures. Luciferase experiments were performed as described (Cheng et al., 2013). Electroporation of HUVEC was performed using a Lonza 4D Nucleofector with the P5 Primary Cell Kit with 2.5 μg of expression construct and 0.5 μg of pmaxGFP. ChIP was performed using the Imprint Kit (Sigma) with 1 μg of antibody: RNA Pol II (mouse monoclonal, Sigma), ERG (rabbit polyclonal, Santa Cruz, C-20), V5 (mouse monoclonal, Invitrogen), or NICD (rabbit polyclonal, ChIP grade, Abcam). IgG (mouse, Sigma) was used as a negative control. Fold enrichment was calculated by determining the fold-change of qPCR values for the specific antibody compared to IgG control. See Table S1 for primer sequences.

Zebrafish experiments

Zebrafish protocols were approved by the Animal Care Committee at the University Health Network and UCSF. The following transgenic lines were utilized: Tg(kdrl:ras-mCherry)s896 (Chi et al., 2008), Tg(kdrl:GFP)s843 (Jin et al., 2005), Tg(EPV.Tp1-MmuHbb:EGFP)um14 (Parsons et al., 2009). See Supplemental Experimental Procedures for generation of Tg(Dll4-F2-E1b:GFP) and experimental details.

Quantitative reverse-transcriptase real-time PCR (qRT-PCR)

qRT-PCR was performed as before (Fish et al, 2008). See Supplemental Experimental Procedures for details and Table S1 for primers sequences.

Western blot

Western blotting was performed as described previously (Fish et al., 2011) using ERG (rabbit polyclonal, Santa Cruz, C-20), Sin3a (Santa Cruz, AK-11), V5 (mouse monoclonal, Invitrogen) or GAPDH (mouse monoclonal, Santa Cruz) (mouse monoclonal, Abcam) antibodies. Cells were starved overnight, and treated with 50 ng/ml recombinant Vegf-A165 (R&D Systems) for 20 minutes before harvesting for sub-cellular fractionation, as previously described (Wythe et al., 2011). HUVEC and HUAEC lysates were purchased from Sciencell.

Statistics

All graphs depict the mean ± standard error of the mean (SEM) of at least 3 independent experiments, unless otherwise indicated. Statistical significance was performed using a Student’s T-test, ANOVA or Chi-squared test as appropriate.

Supplementary Material

01

Highlights.

  • Identification of a Vegf/MAPK-dependent arterial-specific enhancer of Dll4

  • Enhancer activity requires ETS transcription factors

  • Vegf/MAPK signaling drives recruitment of ETS factors to Dll4 and Notch4 enhancers

  • Notch signaling is required for maintenance, not specification, of arterial identity

Acknowledgments

The authors thank Hilary Clay for zebrafish core facility guidance, Dario Miguel-Perez and Claire Cutting for mouse care and husbandry. J.D.W. was supported by a Scientist Development Grant from the American Heart Association (12SDG12060353) and NIH (5T32-HL007544). W.P.D. was supported by the California Institute of Regenerative Medicine (TG2-01160) and NIH (5T32-HL007731-20). S.A. is an American Australian Association Morgan Stanley Pediatrics Fellow, and S.A. and P.O. were supported by the NIH (PO1 HL76540). D.Y.R.S. was supported by the NIH (HL54737) and the Packard Foundation. This work was supported by grants from the NHLBI (P01 HL089707 to B.G.B and B.L.B., and HL64658 to B.L.B.), an operating grant from the Canadian Institutes of Health Research (MOP-119506 to J.E.F.), an Early Researcher Award (Ontario Ministry of Economic Development and Innovation, to J.E.F.), a Leaders Opportunity Fund Award (Canada Foundation for Innovation, to J.E.F.) and a New Investigator Award (Heart and Stroke Foundation of Canada, to J.E.F.). B.L.B. is supported by an Innovative Science Award (12PILT12670005) from the American Heart Association, Western States Affiliate. B.G.B holds the Lawrence J and Florence A. DeGeorge Charitable Trust/American Heart Association Established Investigator Award. This work was also supported by an NIH/NCRR grant (C06 RR018928) to the J. David Gladstone Institutes, and by William H. Younger, Jr (B.G.B.).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bi W, Drake CJ, Schwarz JJ. The transcription factor MEF2C-null mouse exhibits complex vascular malformations and reduced cardiac expression of angiopoietin 1 and VEGF. Developmental biology. 1999;211:255–267. doi: 10.1006/dbio.1999.9307. [DOI] [PubMed] [Google Scholar]
  2. Caolo V, van den Akker NM, Verbruggen S, Donners MM, Swennen G, Schulten H, Waltenberger J, Post MJ, Molin DG. Feed-forward signaling by membrane bound ligand receptor circuit: The case of NOTCH-Delta like 4 in endothelial cells. J Biol Chem. 2010;285:40681–40689. doi: 10.1074/jbc.M110.176065. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cattelino A, Liebner S, Gallini R, Zanetti A, Balconi G, Corsi A, Bianco P, Wolburg H, Moore R, Oreda B, et al. The conditional inactivation of the beta-catenin gene in endothelial cells causes a defective vascular pattern and increased vascular fragility. J Cell Biol. 2003;162:1111–1122. doi: 10.1083/jcb.200212157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cheng HS, Sivachandran N, Lau A, Boudreau E, Zhao JL, Baltimore D, Delgado-Olguin P, Cybulsky MI, Fish JE. MicroRNA-146 represses endothelial activation by inhibiting pro-inflammatory pathways. EMBO Molecular Medicine. 2013 doi: 10.1002/emmm.201202318. E-Pub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chi NC, Shaw RM, De Val S, Kang G, Jan LY, Black BL, Stainier DY. Foxn4 directly regulates tbx2b expression and atrioventricular canal formation. Genes Dev. 2008;22:734–739. doi: 10.1101/gad.1629408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chong DC, Koo Y, Xu K, Fu S, Cleaver O. Stepwise arteriovenous fate acquisition during mammalian vasculogenesis. Developmental dynamics: an official publication of the American Association of Anatomists. 2011;240:2153–2165. doi: 10.1002/dvdy.22706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Corada M, Nyqvist D, Orsenigo F, Caprini A, Giampietro C, Taketo MM, Iruela-Arispe ML, Adams RH, Dejana E. The Wnt/beta-catenin pathway modulates vascular remodeling and specification by upregulating Dll4/Notch signaling. Dev Cell. 2010;18:938–949. doi: 10.1016/j.devcel.2010.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Corson LB, Yamanaka Y, Lai KM, Rossant J. Spatial and temporal patterns of ERK signaling during mouse embryogenesis. Development. 2003;130:4527–4537. doi: 10.1242/dev.00669. [DOI] [PubMed] [Google Scholar]
  9. Coultas L, Nieuwenhuis E, Anderson GA, Cabezas J, Nagy A, Henkelman RM, Hui CC, Rossant J. Hedgehog regulates distinct vascular patterning events through VEGF-dependent and -independent mechanisms. Blood. 2010;116:653–660. doi: 10.1182/blood-2009-12-256644. [DOI] [PubMed] [Google Scholar]
  10. De Val S, Black BL. Transcriptional control of endothelial cell development. Developmental cell. 2009;16:180–195. doi: 10.1016/j.devcel.2009.01.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. De Val S, Chi NC, Meadows SM, Minovitsky S, Anderson JP, Harris IS, Ehlers ML, Agarwal P, Visel A, Xu SM, et al. Combinatorial regulation of endothelial gene expression by ets and forkhead transcription factors. Cell. 2008;135:1053–1064. doi: 10.1016/j.cell.2008.10.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Delgado-Olguin P, Huang Y, Li X, Christodoulou D, Seidman CE, Seidman JG, Tarakhovsky A, Bruneau BG. Epigenetic repression of cardiac progenitor gene expression by Ezh2 is required for postnatal cardiac homeostasis. Nat Genet. 2012;44:343–347. doi: 10.1038/ng.1068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Deng Y, Larrivee B, Zhuang ZW, Atri D, Moraes F, Prahst C, Eichmann A, Simons M. Endothelial RAF1/ERK activation regulates arterial morphogenesis. Blood. 2013;121:3988–3996. doi: 10.1182/blood-2012-12-474601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dodou E, Xu SM, Black BL. mef2c is activated directly by myogenic basic helix-loop-helix proteins during skeletal muscle development in vivo. Mech Dev. 2003;120:1021–1032. doi: 10.1016/s0925-4773(03)00178-3. [DOI] [PubMed] [Google Scholar]
  15. Duarte A, Hirashima M, Benedito R, Trindade A, Diniz P, Bekman E, Costa L, Henrique D, Rossant J. Dosage-sensitive requirement for mouse Dll4 in artery development. Genes & development. 2004;18:2474–2478. doi: 10.1101/gad.1239004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fish JE, Wythe JD, Xiao T, Bruneau BG, Stainier DY, Srivastava D, Woo S. A Slit/miR-218/Robo regulatory loop is required during heart tube formation in zebrafish. Development. 2011;138:1409–1419. doi: 10.1242/dev.060046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gale NW, Dominguez MG, Noguera I, Pan L, Hughes V, Valenzuela DM, Murphy AJ, Adams NC, Lin HC, Holash J, et al. Haploinsufficiency of delta-like 4 ligand results in embryonic lethality due to major defects in arterial and vascular development. Proceedings of the National Academy of Sciences of the United States of America. 2004;101:15949–15954. doi: 10.1073/pnas.0407290101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ghosh S, Basu M, Roy SS. ETS-1 protein regulates vascular endothelial growth factor-induced matrix metalloproteinase-9 and matrix metalloproteinase-13 expression in human ovarian carcinoma cell line SKOV-3. J Biol Chem. 2012;287:15001–15015. doi: 10.1074/jbc.M111.284034. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  19. Heo SH, Choi YJ, Ryoo HM, Cho JY. Expression profiling of ETS and MMP factors in VEGF-activated endothelial cells: role of MMP-10 in VEGF-induced angiogenesis. J Cell Physiol. 2010;224:734–742. doi: 10.1002/jcp.22175. [DOI] [PubMed] [Google Scholar]
  20. Hill CS, Marais R, John S, Wynne J, Dalton S, Treisman R. Functional analysis of a growth factor-responsive transcription factor complex. Cell. 1993;73:395–406. doi: 10.1016/0092-8674(93)90238-l. [DOI] [PubMed] [Google Scholar]
  21. Hollenhorst PC, McIntosh LP, Graves BJ. Genomic and biochemical insights into the specificity of ETS transcription factors. Annu Rev Biochem. 2011;80:437–471. doi: 10.1146/annurev.biochem.79.081507.103945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hong CC, Peterson QP, Hong JY, Peterson RT. Artery/vein specification is governed by opposing phosphatidylinositol-3 kinase and MAP kinase/ERK signaling. Curr Biol. 2006;16:1366–1372. doi: 10.1016/j.cub.2006.05.046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jin SW, Beis D, Mitchell T, Chen JN, Stainier DY. Cellular and molecular analyses of vascular tube and lumen formation in zebrafish. Development. 2005;132:5199–5209. doi: 10.1242/dev.02087. [DOI] [PubMed] [Google Scholar]
  24. Krebs LT, Shutter JR, Tanigaki K, Honjo T, Stark KL, Gridley T. Haploinsufficient lethality and formation of arteriovenous malformations in Notch pathway mutants. Genes & development. 2004;18:2469–2473. doi: 10.1101/gad.1239204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Krebs LT, Starling C, Chervonsky AV, Gridley T. Notch1 activation in mice causes arteriovenous malformations phenocopied by ephrinB2 and EphB4 mutants. Genesis. 2010;48:146–150. doi: 10.1002/dvg.20599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Krebs LT, Xue Y, Norton CR, Shutter JR, Maguire M, Sundberg JP, Gallahan D, Closson V, Kitajewski J, Callahan R, et al. Notch signaling is essential for vascular morphogenesis in mice. Genes & development. 2000;14:1343–1352. [PMC free article] [PubMed] [Google Scholar]
  27. Kume T. Specification of arterial, venous, and lymphatic endothelial cells during embryonic development. Histol Histopathol. 2010;25:637–646. doi: 10.14670/hh-25.637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lawson ND, Scheer N, Pham VN, Kim CH, Chitnis AB, Campos-Ortega JA, Weinstein BM. Notch signaling is required for arterial-venous differentiation during embryonic vascular development. Development. 2001;128:3675–3683. doi: 10.1242/dev.128.19.3675. [DOI] [PubMed] [Google Scholar]
  29. Lawson ND, Vogel AM, Weinstein BM. sonic hedgehog and vascular endothelial growth factor act upstream of the Notch pathway during arterial endothelial differentiation. Dev Cell. 2002;3:127–136. doi: 10.1016/s1534-5807(02)00198-3. [DOI] [PubMed] [Google Scholar]
  30. Lee D, Park C, Lee H, Lugus JJ, Kim SH, Arentson E, Chung YS, Gomez G, Kyba M, Lin S, et al. ER71 acts downstream of BMP, Notch, and Wnt signaling in blood and vessel progenitor specification. Cell stem cell. 2008;2:497–507. doi: 10.1016/j.stem.2008.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lin Q, Lu J, Yanagisawa H, Webb R, Lyons GE, Richardson JA, Olson EN. Requirement of the MADS-box transcription factor MEF2C for vascular development. Development. 1998;125:4565–4574. doi: 10.1242/dev.125.22.4565. [DOI] [PubMed] [Google Scholar]
  32. Liu F, Patient R. Genome-wide analysis of the zebrafish ETS family identifies three genes required for hemangioblast differentiation or angiogenesis. Circ Res. 2008;103:1147–1154. doi: 10.1161/CIRCRESAHA.108.179713. [DOI] [PubMed] [Google Scholar]
  33. Liu ZJ, Shirakawa T, Li Y, Soma A, Oka M, Dotto GP, Fairman RM, Velazquez OC, Herlyn M. Regulation of Notch1 and Dll4 by vascular endothelial growth factor in arterial endothelial cells: implications for modulating arteriogenesis and angiogenesis. Mol Cell Biol. 2003;23:14–25. doi: 10.1128/MCB.23.1.14-25.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Luo H, Jin K, Xie Z, Qiu F, Li S, Zou M, Cai L, Hozumi K, Shima DT, Xiang M. Forkhead box N4 (Foxn4) activates Dll4-Notch signaling to suppress photoreceptor cell fates of early retinal progenitors. Proceedings of the National Academy of Sciences of the United States of America. 2012;109:E553–562. doi: 10.1073/pnas.1115767109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Marchuk DA. Genetic abnormalities in hereditary hemorrhagic telangiectasia. Curr Opin Hematol. 1998;5:332–338. doi: 10.1097/00062752-199809000-00005. [DOI] [PubMed] [Google Scholar]
  36. Murakami M, Nguyen LT, Hatanaka K, Schachterle W, Chen PY, Zhuang ZW, Black BL, Simons M. FGF-dependent regulation of VEGF receptor 2 expression in mice. J Clin Invest. 2011;121:2668–2678. doi: 10.1172/JCI44762. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Oka C, Nakano T, Wakeham A, de la Pompa JL, Mori C, Sakai T, Okazaki S, Kawaichi M, Shiota K, Mak TW, et al. Disruption of the mouse RBP-J kappa gene results in early embryonic death. Development. 1995;121:3291–3301. doi: 10.1242/dev.121.10.3291. [DOI] [PubMed] [Google Scholar]
  38. Parsons MJ, Pisharath H, Yusuff S, Moore JC, Siekmann AF, Lawson N, Leach SD. Notch-responsive cells initiate the secondary transition in larval zebrafish pancreas. Mech Dev. 2009;126:898–912. doi: 10.1016/j.mod.2009.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Petrovic N, Bhagwat SV, Ratzan WJ, Ostrowski MC, Shapiro LH. CD13/APN transcription is induced by RAS/MAPK-mediated phosphorylation of Ets-2 in activated endothelial cells. J Biol Chem. 2003;278:49358–49368. doi: 10.1074/jbc.M308071200. [DOI] [PubMed] [Google Scholar]
  40. Ren B, Deng Y, Mukhopadhyay A, Lanahan AA, Zhuang ZW, Moodie KL, Mulligan-Kehoe MJ, Byzova TV, Peterson RT, Simons M. ERK1/2-Akt1 crosstalk regulates arteriogenesis in mice and zebrafish. J Clin Invest. 2010;120:1217–1228. doi: 10.1172/JCI39837. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Seo S, Fujita H, Nakano A, Kang M, Duarte A, Kume T. The forkhead transcription factors, Foxc1 and Foxc2, are required for arterial specification and lymphatic sprouting during vascular development. Dev Biol. 2006;294:458–470. doi: 10.1016/j.ydbio.2006.03.035. [DOI] [PubMed] [Google Scholar]
  42. Shalaby F, Rossant J, Yamaguchi TP, Gertsenstein M, Wu XF, Breitman ML, Schuh AC. Failure of blood-island formation and vasculogenesis in Flk-1-deficient mice. Nature. 1995;376:62–66. doi: 10.1038/376062a0. [DOI] [PubMed] [Google Scholar]
  43. Swift MR, Weinstein BM. Arterial-venous specification during development. Circulation research. 2009;104:576–588. doi: 10.1161/CIRCRESAHA.108.188805. [DOI] [PubMed] [Google Scholar]
  44. Vijayaraj P, Le Bras A, Mitchell N, Kondo M, Juliao S, Wasserman M, Beeler D, Spokes K, Aird WC, Baldwin HS, et al. Erg is a crucial regulator of endocardial-mesenchymal transformation during cardiac valve morphogenesis. Development. 2012;139:3973–3985. doi: 10.1242/dev.081596. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vokes SA, Yatskievych TA, Heimark RL, McMahon J, McMahon AP, Antin PB, Krieg PA. Hedgehog signaling is essential for endothelial tube formation during vasculogenesis. Development. 2004;131:4371–4380. doi: 10.1242/dev.01304. [DOI] [PubMed] [Google Scholar]
  46. Wang L, Fan C, Topol SE, Topol EJ, Wang Q. Mutation of MEF2A in an inherited disorder with features of coronary artery disease. Science. 2003;302:1578–1581. doi: 10.1126/science.1088477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wythe JD, Jurynec MJ, Urness LD, Jones CA, Sabeh MK, Werdich AA, Sato M, Yost HJ, Grunwald DJ, Macrae CA, et al. Hadp1, a newly identified pleckstrin homology domain protein, is required for cardiac contractility in zebrafish. Disease models & mechanisms. 2011;4:607–621. doi: 10.1242/dmm.002204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Yamamizu K, Matsunaga T, Uosaki H, Fukushima H, Katayama S, Hiraoka-Kanie M, Mitani K, Yamashita JK. Convergence of Notch and beta-catenin signaling induces arterial fate in vascular progenitors. J Cell Biol. 2010;189:325–338. doi: 10.1083/jcb.200904114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zhang B, Day DS, Ho JW, Song L, Cao J, Christodoulou D, Seidman JG, Crawford GE, Park PJ, Pu WT. A dynamic H3K27ac signature identifies VEGFA-stimulated endothelial enhancers and requires EP300 activity. Genome Res. 2013 doi: 10.1101/gr.149674.112. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS492229-supplement-01.pdf (7.1MB, pdf)

RESOURCES