Article

• The EMBO Journal (2008) 27, 993 - 1004
• doi:10.1038/emboj.2008.46

Published online: 13 March 2008

• Subject Categories:

  • Cell and Tissue Architecture | Signal Transduction

VE-cadherin is a critical endothelial regulator of TGF- signalling

Noemi Rudini^1,7, Angelina Felici^1,7, Costanza Giampietro¹, MariaGrazia Lampugnani^1,2, Monica Corada¹, Kendra Swirsding¹, Massimiliano Garrè¹, Stefan Liebner³, Michelle Letarte⁴, Peter ten Dijke⁵ and Elisabetta Dejana^1,2,6

1. Vascular Biology Unit, FIRC Institute of Molecular Oncology, Milan, Italy
2. Mario Negri Institute for Pharmacological Research, Milan, Italy
3. Institute of Neurology, Medical School, University of Frankfurt/Main, Frankfurt, Germany
4. Molecular Structure and Function Program, The Hospital for Sick Children, Heart and Stroke Richard Lewar Center of Excellence, University of Toronto, Toronto, Canada
5. Molecular Cell Biology Department, Leiden University Medical Center, Leiden, The Netherlands
6. Department of Biomolecular Sciences and Biotechnologies, School of Sciences, University of Milan, Milan, Italy
7. These authors contributed equally to this work

Correspondence to:

Elisabetta Dejana, Vascular Biology Unit, FIRC Institute of Molecular Oncology, Via Adamello 16, Milan 20139, Italy. Tel.: +39 02 574303234; Fax: +39 02 574303244; E-mail: elisabetta.dejana@ifom-ieo-campus.it

Angelina Felici, FIRC Institute of Molecular Oncology, Via Adamello 16, Milan 20139, Italy. Tel.: +39 02 574303309; Fax: +39 02 574303244; E-mail: angelina.felici@ifom-ieo-campus.it

Received 27 August 2007; Accepted 20 February 2008

• Keywords:

  • endothelial cells,
  • signal transduction,
  • Smad,
  • TGF-,
  • VE-cadherin

Introduction

Transforming growth factor- (TGF-) is a multifunctional dimeric polypeptide growth factor that regulates proliferation, differentiation, migration, extracellular matrix production and survival of various cell types. TGF- mediates its cellular effects through ligand-induced heteromeric receptor complexes of type II and type I transmembrane serine-threonine kinases. Activated type I receptors phosphorylate cytoplasmic receptor-associated Smad proteins (R-Smads). Phosphorylated R-Smads dissociate from the receptor and form a complex with the common mediator Smad4. This complex accumulates in the nucleus where it regulates transcription by interacting with many specific DNA-binding proteins (Attisano and Wrana, 2002; Derynck and Zhang, 2003).

In endothelial cells (ECs), the TGF- type II receptor (TRII) and two distinct TGF- type I receptors (TRIs), the EC-restricted ALK1 and the broadly expressed ALK5, are expressed (Oh et al, 2000; Goumans et al, 2002). ALK1 activation induces the phosphorylation of Smad1/5/8, whereas ALK5 promotes Smad2/3 phosphorylation. In addition, TGF- also binds the co-receptor endoglin, a modulator of ALK1 and ALK5 signalling, highly expressed on rapidly proliferating ECs. In some experimental conditions, endoglin increases ALK1 signalling and promotes EC proliferation and angiogenesis (Goumans et al, 2002; Lebrin et al, 2004). In other experimental settings, lack of endoglin increases response to ALK1 and subsequently growth (Pece-Barbara et al, 2005). These conflicting results suggest that the cellular context may strongly influence the downstream effect of TGF- activation in ECs.

Hereditary haemorrhagic telangiectasia (HHT) is an autosomal dominant vascular dysplasia characterized by epistaxis, telangiectases, pulmonary and cerebral vascular malformations and later in life by gastrointestinal bleeding. Two main genes are mutated in this disease, ENG in HHT1 and ALK1 in HHT2 that account for about 80% of cases (Lesca et al, 2006; Prigoda et al, 2006), indicating that yet unidentified genes are also responsible for this disease.

Mice lacking either endoglin or ALK1 die in utero at embryonic day 10.0–10.5 of major defects in vascular and heart development (Bourdeau et al, 1999; Li et al, 1999; Arthur et al, 2000; Oh et al, 2000; Urness et al, 2000). The phenotype is characterized by haemorrhages in the yolk sac and embryo proper due to lack of maturation of the primitive vascular plexus with abnormal dilation of the lumen and vascular rupture. In the heart, the ECs of the endocardium fail to undergo mesenchymal transition required for their migration into the atrio-ventricular cushion. This defect strongly affects heart development and function (Bourdeau et al, 1999; Arthur et al, 2000; Sorensen et al, 2003).

The vascular phenotype of endoglin- and ALK1-deficient embryos presents striking similarities with that of embryos lacking proteins of endothelial cell-to-cell adherens junctions. These structures are formed by an endothelial-specific member of the cadherin family (VE-cadherin), which promotes homotypic cell–cell adhesion and which, through its cytoplasmic domain, links intracellular partners such as -catenin, p120 and plakoglobin (Dejana, 2004; Gumbiner, 2005). Adherens junctions are required for EC stabilization and homoeostasis as they promote contact inhibition of growth and reduce cell sensitivity to apoptotic stimuli. Embryos lacking VE-cadherin die in utero by embryonic day 10.5 from alterations in the vascular development of the yolk sac and embryo. Vessels cannot undergo remodelling, are fragile and haemorrhagic, tend to grow in an irregular way and eventually regress (Carmeliet et al, 1999). Consistently, EC-specific inactivation of the -catenin gene leads to embryonic death due to abnormal development of the vascular system with lacunae, enlarged and irregular lumen and multiple haemorrhages (Cattelino et al, 2003). Most strikingly, -catenin-null endocardial cells fail to invade the atrio-ventricular cushion (Liebner et al, 2004) leading to a defect comparable to that described in endoglin/ALK1 knockout embryos. In adult mice, transplantation of VE-cadherin-null ECs leads to pathological features, such as the formation of large haemangiomas, abnormal EC growth and frequent bleeding (Zanetta et al, 2005).

The mechanism of action of VE-cadherin/-catenin signalling is complex and several pathways may be affected by its inhibition (Dejana, 2004; Liebner et al, 2006). We found that VE-cadherin may associate with VEGFR-2 and reduce its proliferative signal, thus limiting EC growth and controlling vascular remodelling. This activity requires VE-cadherin clustering at junctions and association with DEP-1/CD148 phosphatase, responsible for VEGFR-2 receptor dephosphorylation (Lampugnani et al, 2003). However, the mechanism by which VE-cadherin signalling controls vascular stabilization is still poorly investigated. Previously, we demonstrated that -catenin-null ECs/endocardial cells fail to respond to TGF- and to undergo mesenchymal transition, suggesting the existence of a strict interplay between the VE-cadherin/-catenin complex and TGF- signalling (Liebner et al, 2004). In the present study, we investigated the role of VE-cadherin expression and clustering in TGF--induced biological responses. Our data demonstrate for the first time that VE-cadherin is a key positive regulator of TGF-/ALK/Smad signalling in ECs.

VE-cadherin knockdown inhibited TGF--induced Smad1/5 and Smad2/3 phosphorylation, reduced expression of Smad target genes and counteracted the inhibitory effect of TGF- on EC proliferation and migration. TGF- stimulation induced TRII association with VE-cadherin at cell-to-cell contacts where VE-cadherin clustering may promote TRII/TRI assembly and activation. Overall our data indicate that VE-cadherin participates in maximal activation of the TGF- pathway, acting as a positive and EC-specific regulator of TGF- response. This indicates that the response of ECs to growth factors is context dependent and that increased TGF- signalling may contribute to VE-cadherin-dependent stabilization and remodelling of the vascular endothelium.

VE-cadherin increases TGF-/Smad-dependent transcription

To examine whether VE-cadherin has a regulatory role in TGF- signalling in ECs, we used mouse embryonic stem cell-derived ECs, with homozygous null mutation of the VE-cadherin gene (VEC null) and reconstituted by retroviral transfer to express wild-type levels of VE-cadherin (VEC positive), as described previously (Balconi et al, 2000; Lampugnani et al, 2002, 2003; Iurlaro et al, 2004). We analysed these cells in a transcriptional response assay using a TGF--inducible reporter. The ALK5-dependent transcriptional reporter (CAGA)₁₂-luc (Dennler et al, 1998) was activated by TGF- in VEC-positive cells; however, the stimulatory effect on the reporter was greatly reduced in the VEC-null cells (Figure 1A). To validate these findings in different cells, we repeated the reporter assays in mouse embryo-derived ECs (MEECs) from VE-cadherin knockout and wild-type embryos and in TGF- responsive CHO cells transfected with VE-cadherin or control vector (Breviario et al, 1995). The presence of VE-cadherin led to higher TGF- response in all three cell types (Figure 1A).

Figure 1.

VE-cadherin enhances TGF--induced transcription. (A) VE-cadherin is important for optimal activation of TGF--induced reporter activity. VEC-null and -positive (Pos) cells, MEECs from VE-cadherin knockout (KO) and wild-type (WT) embryos, and CHO cells bearing a control vector (Vector) or VE-cadherin (VE) cDNA were transfected with a TGF--inducible reporter and stimulated with (+) or without (-) TGF-1. Data are represented as meansA.D. of six replicates from a representative experiment out of three performed. The statistical difference between the two cell lines is indicated (^*P<0.0001). (B) Effect of VE-cadherin expression on TGF- target genes PAI-1 and Id1. VEC-null (open bars) and -positive (closed bars) cells were stimulated with TGF-1 (2.5 ng/ml) for the indicated times. Fold changes in gene expression were calculated using the 2^-C_t method. Error bars represent s.e. (n=3) from a representative experiment out of two performed. The statistical difference between the two cell lines is indicated (^*P=0.003 and ^**P<0.001).

View full figure (29 KB)

We next tested the effect of TGF- on the expression of plasminogen activator inhibitor-1 (PAI-1), downstream of the ALK5 receptor and on the inhibitor of differentiation-1 (Id1), an ALK1-specific target gene (Goumans et al, 2002). Both genes were robustly upregulated in the VEC-positive cells compared with the null cells (Figure 1B).

These experiments indicate that VE-cadherin expression is a positive regulator of TGF--induced transcriptional responses in ECs.

VE-cadherin is essential for TGF--induced inhibition of EC cell growth and migration

To examine the effect of VE-cadherin on TGF--induced EC growth inhibition, we counted the number of viable cells present after 5 days of TGF- treatment (5 ng/ml) by direct counting (Figure 2A). Consistent with previous results, VEC-null cells proliferate faster than VEC-positive cells and achieve higher cell densities (Lampugnani et al, 2003). However, TGF- inhibitory effect on cell proliferation was consistently greater in VEC positive (50%) than in null cells (21%). We next measured the contribution of VE-cadherin to the TGF- inhibition of migratory behaviour of human umbilical vein ECs (HUVECs), as TGF- is known to tightly control EC migration during angiogenesis (Muller et al, 1987; Basson et al, 1992). VE-cadherin expression was ablated by siRNA (Supplementary Figure 1S) and migratory behaviour was assessed by wound assay and time-lapse microscopy (Figure 2B). Short time intervals (3–12 h) were selected, where the role of VE-cadherin in migration is not apparent. siRNA-mediated knockdown of VE-cadherin blocked the inhibitory effect of TGF- on HUVEC migration. HUVEC transfected with siRNA for VE-cadherin, either untreated or TGF--treated, displayed a constant migration speed of 12 m/h. Control siRNA-transfected cells also displayed a migration speed of 12 m/h, but reduced by TGF- to 8 m/h.

Figure 2.

VE-cadherin potentiates TGF- inhibitory effect on EC growth and migration. (A) VEC-null cells are less sensitive to TGF--induced growth inhibition than VEC-positive cells. VEC-null (Null) and -positive (Pos) cells were stimulated with (+) or without (-) TGF-1 for 5 days. The number of viable cells was determined by Trypan blue and haemocytometer counting. Columns are meansA.D. of triplicate from a representative experiment. ^*P=0.02; ^**P=0.002. (B) siRNA-mediated knockdown of VE-cadherin blocks TGF--induced inhibition of wound repair in HUVECs. Cells transfected with VE-cadherin (VE) or control siRNA (CTL) were starved, wounded and treated with (+) or without (-) TGF-1. Migration of cells out of the wounded edge was recorded at the indicated times. The mean migrated distance and A.D. (n=6) of a representative experiment is shown. ^*P=0.002; ^**P<0.001. (C) Lack of VE-cadherin blocks TGF--induced inhibition of wound repair in VEC cells. EC monolayers were wounded and stimulated with (+) or without (-) TGF-1. At the indicated times the cells were fixed and stained with crystal violet. The migration of the cell front was measured by ImageJ software. The mean migrated distance and A.D. (n=9) of a representative experiment out of three performed are shown. ^*P<0.001. Left panels show the migrating front 12 h after wounding the EC monolayer. Scale bar=40 m.

View full figure (125 KB)

We validated this observation in a different EC model system. VEC-null and -positive cells were subjected to a wound assay and migrated cells were visualized by crystal violet and analysed by contrast phase microscopy (Figure 2C). VEC-null cells exhibited a basally higher migratory behaviour than VEC-positive cells, in accordance with VE-cadherin ability to prevent cell detachment from a confluent monolayer (Navarro et al, 1995). More importantly, VEC-null cell motility was not inhibited by TGF-, whereas that of VEC-positive cells was, indicating that VE-cadherin positively regulates TGF--mediated inhibition of EC migration. The difference in cell migration was found at time intervals (12 and 24 h) where TGF- has no antiproliferative activities on VEC cells (not shown), suggesting that VE-cadherin specifically regulates TGF- effect on EC motility. Taken together, our results suggest that VE-cadherin expression is critical for TGF- inhibitory activity of EC proliferation and migration.

VE-cadherin is necessary for efficient ALK1- and ALK5-dependent Smad phosphorylation in TGF--treated ECs

TGF- elicits biological responses in ECs through two distinct type I receptors, ALK1 and ALK5, which induce the phosphorylation of Smad1/5/8 and Smad2/3, respectively, and their nuclear translocation (Goumans et al, 2002, 2003). To assess at which level of the signalling cascade VE-cadherin impacts TGF- pathway, we analysed whether VE-cadherin-positive effect on TGF--induced response correlated with increased phosphorylation of TGF--dependent Smads. In time-dependent experiments, Smad1/5 phosphorylation was more transient than that of Smad2/3 (Figure 3). In VEC-positive cells peak levels for Smad1/5 and Smad2/3 phosphorylation were reached after 45 min of stimulation with TGF-1 and were higher (2–2.5-fold) than in null cells, whereas total Smad levels were substantially unchanged (Figure 3). Shortly after TGF- stimulation the onset and intensity of both Smad1/5 and Smad2/3 phosphorylation was increased, whereas at later time points only the duration of Smad2/3 phosphorylation was prolonged by VE-cadherin expression (Figure 3). In dose-dependent experiments, VEC-null cells showed 3- to 4-fold decrease in magnitude of both Smad1/5 and Smad2/3 phosphorylation, at equivalent TGF- concentrations, indicating a significantly lower sensitivity of these cells to TGF- (Figure 4). The enhanced TGF- response of cells expressing VE-cadherin was not due to increased TRII, ALK5 and ALK1 expression as shown by the analysis of TGF- receptor levels (Supplementary Figure 2S).

Figure 3.

More rapid kinetics of TGF--induced Smad phosphorylation in VE-cadherin-expressing ECs. VEC cells were serum starved and stimulated with TGF-1 (0.5 ng/ml) for the indicated times. Whole cell extracts were fractionated by 10% SDS–PAGE and blotted with phospho-specific Smad1/5, Smad2 and Smad3 antibodies. Total Smad1/2/3 levels were also probed for loading control. Lower graphs show the intensities of pSmad bands quantified by densitometry scan, corrected for background and plotted as optical density (OD). These results were confirmed by Odyssey Infrared Imaging System Analysis (data not shown).

View full figure (102 KB)

Figure 4.

Dose–response of TGF--induced Smad phosphorylation is enhanced by VE-cadherin. Serum-starved VEC cells were stimulated with the indicated concentrations of TGF-1 for 1 h. Whole cell extracts were blotted and quantified by densitometry scan. Lower graphs show the intensities of pSmad bands corrected for background and plotted as optical density (OD). The experiments reported in this figure were performed three times with comparable results.

View full figure (112 KB)

To substantiate these findings in a different EC system, we repeated the analysis of Smad phosphorylation in HUVEC transfected with VE-cadherin or negative control siRNA. Silencing VE-cadherin expression resulted in a significant decline in TGF- ability to induce ALK1- and ALK5-dependent Smad phosphorylation in the presence of unchanged Smad total levels (Supplementary Figure 3S). This effect was less marked than in VEC-null cells likely due to incomplete VE-cadherin knockdown (80%) achieved by siRNA interference in these cells.

Taken together, these results indicate that VE-cadherin increases TGF-/ALK1- and TGF-/ALK5-mediated responses in ECs by acting upstream of Smad phosphorylation.

VE-cadherin forms a complex with TRs

In ECs, signalling by TGF- is initiated by binding to TRII followed by recruitment of TRI complexes and subsequent Smad phosphorylation (Goumans et al, 2003). As VE-cadherin increases TGF- signalling upstream of Smad phosphorylation, we hypothesized that it might do so through physical association with the TGF- receptor complex.

To test this hypothesis, we immunoprecipitated TRs from COS-1 cells co-transfected with VE-cadherin and either TRII, ALK5 or ALK1. In the immunoprecipitates, we not only recovered the components of the TGF- receptor complex but also VE-cadherin, indicating that VE-cadherin interacts with TRII, ALK5 and ALK1 in the absence of exogenously added ligand (Figure 5A).

Figure 5.

VE-cadherin associates with the TR complex. (A) VE-cadherin associates with transiently expressed TRs in COS-1 cells. Cells were transfected with the indicated constructs, immunoprecipitated (IP) with specific antibodies (TRII, ALK5 and ALK1) or isotype control IgG, and blotted (WB) as indicated. Total cell lysates (TCLs) were loaded to control for transfection efficiency. Apparent decrease in VE-cadherin/TRs association is due to a lower TR constructs expression induced by endoglin. ENG, endoglin; VE, VE-cadherin. (B) Surface interaction of endogenous TRII and VE-cadherin. VEC positive cells were serum starved, sulfo-NHS-biotinylated on ice, rinsed, solubilized, immunoprecipitated with anti-TRII or isotype control IgG and probed with streptavidin-horseradish peroxidase (Strep-HRP). Cell lysates immunoprecipitated with -catenin antibody (-cat) were used to control for biotinylated VE-cadherin. (C) Ligand-dependent interaction of VE-cadherin with endogenous TRII. VEC-positive cells were serum starved and stimulated with TGF-1 (5 ng/ml) for the indicated time intervals. Cell extracts were immunoprecipitated (IP) with TRII antibodies or control isotype IgG and blotted with VE-cadherin (VE) or TRII antibody. The intensities of TRII-bound VE signals were quantitated by densitometry scan and are represented as optical density (OD). (D) Endogenous ALK5/VE-cadherin association. VEC-positive cells stimulated with TGF-1 (5 ng/ml) for 45 min, followed by treatment with cross-linker for 20 min, were immunoprecipitated (IP) with ALK5 antibody or control isotype IgG and blotted with VE-cadherin (VE) or ALK5 antibody. ^*ALK5 doublet.

View full figure (125 KB)

We confirmed previous results (Barbara et al, 1999; Abdalla et al, 2000; Pece-Barbara et al, 2005) that endoglin is present in the TGF- receptor complex and interacts with TRII, ALK5 and ALK1 (Figure 5A). We assessed whether endoglin might associate with VE-cadherin and modulate its binding to TRs. We report here that endoglin binds VE-cadherin efficiently (Supplementary Figure 4S, panel A) but is not required for the VE-cadherin/TRs complex to form, although it is most likely present in the complex in vivo (Supplementary Figure 4S, panel B). To ascertain that the analysed VE-cadherin/TR complexes from transfected COS-1 cells occurred at the cell surface and not in internal compartments, both in vivo-formed immune complexes analysis and surface biotinylation and coimmunoprecipitation analysis were used. Using both approaches, VE-cadherin was recovered from TRII- and TRI-containing immunoprecipitates, indicating that the association between VE-cadherin and ectopically expressed TRs occurs between cell surface exposed molecules in transfected COS-1 cells (Supplementary Figure 5S).

Next, we examined endogenous TRII using surface biotinylation of VEC-positive cells followed by immunoprecipitation with anti-TRII and detection using streptavidin. We found that a surface protein comigrating with VE-cadherin was coimmunoprecipitated by TRII (Figure 5B). To confirm the association of endogenous TRII and VE-cadherin and assess whether their binding was TGF- dependent, we carried out immunoprecipitation and immunoblotting of TGF--treated VEC-positive cells. We found that endogenous TRII interacts with VE-cadherin and that this interaction slowly increases between 15 and 45 min of ligand stimulation (Figure 5C) and is still sustained at 2 h (data not shown). Finally, we immunoprecipitated endogenous ALK5 and found that VE-cadherin is an interacting partner also of this receptor in TGF--stimulated cells (Figure 5D).

Disruption of VE-cadherin clustering inhibits TGF--dependent Smad phosphorylation

Proper clustering of VE-cadherin at interendothelial junctions is required for optimal inhibition of VEGFR-2 phosphorylation and mitogenic signalling (Lampugnani et al, 2003). To test whether VE-cadherin clustering was essential to exert its positive effect on TGF- signalling, we first assessed whether transiently expressed TRII codistributed with VE-cadherin at junctions in cultured VEC-positive cells by indirect immunofluorescence and confocal microscopy (Figure 6A). VE-cadherin staining was mostly localized at discrete spots of adjacent plasma membranes, as expected for a junctional protein. TRII displayed a vesicular and cell surface localization. Consistent with a possible localization at intercellular adherens junctions, TRII showed colocalization with VE-cadherin at cell-to-cell contacts (Figure 6A). The quantification of the average number of colocalization events at the plasma membrane showed weak codistribution in the untreated state followed by a two-fold increase upon stimulation with 2 ng/ml TGF- for 90 min (Figure 6B). We examined the cell surface distribution of endogenous TRII/TRI complexes by in vivo treatment with biotinylated TGF-1 and FITC-conjugated avidin followed by immunofluorescence confocal microscopy. After stimulation of cells with biotinylated TGF-1, cell surface TRII/TRI active signalling complexes localized to puncta unevenly distributed over the plasma membrane. A subset of these active TGF- receptor complexes concentrated at cell contacts where they localized with VE-cadherin (Figure 6C).

Figure 6.

VE-cadherin colocalizes with TRs at cell contacts. (A) TGF--induced colocalization of transiently expressed TRII and VE-cadherin at cell–cell junctions. VEC-positive cells transfected with TRII, and stimulated with (TGF-) or without (w/o) TGF-1 (2 ng/ml), were stained with the polyclonal antibody anti-TRII (red), and the monoclonal anti-VE-cadherin (green). Overlay of the two images is shown (merge), with colocalization appearing as yellow. Areas marked by a rectangle are enlarged and shown as insets. Scale bar=10 m. (B) Quantification of colocalization between VE-cadherin and TRII. Colocalization events per cell were acquired by immunofluorescence confocal analysis and quantitated using the ImageJ colocalization plug-in. Values are displayed as meanss.d. of triplicates from one representative experiment. ^*P=0.007. (C) VE-cadherin colocalizes with endogenous TRs. Serum-starved VEC-positive cells were incubated with biotinylated TGF-1 and stained in vivo with FITC-conjugated avidin to detect the endogenous TRII/TRI active complexes (green). The cells were fixed and stained with the monoclonal VE-cadherin antibody (red). Overlay of the two images (merge) is shown. In the bottom right panels (magnification of the boxed areas) the arrows indicate TRs/VE-cadherin colocalization (orange dots) set upon the VE-cadherin background (grey). Scale bar=10 m.

View full figure (302 KB)

Thus, biochemical and morphological data indicating ligand-dependent association and colocalization of endogenous TRs with VE-cadherin suggest that VE-cadherin recruits TRs at cell-to-cell contacts during signalling.

To test whether proper clustering of VE-cadherin at interendothelial junctions was required for its effect on TGF- signalling, we treated confluent monolayers of VEC-positive cells with 5 mM EGTA during TGF- stimulation to fully disrupt VE-cadherin clustering at cell-to-cell contacts (Supplementary Figure 6S). Only low levels of phosphorylated Smad2 and Smad3 were detectable in EGTA-treated VEC-positive cells challenged with TGF-, in contrast to control cells treated with TGF- in the absence of EGTA, suggesting that a robust activation of Smad signalling in endothelial monolayers requires correct homophilic interactions of VE-cadherin molecules. We further confirmed this finding by a monoclonal antibody (BV9) known to block VE-cadherin-adhesive properties in a highly specific way (Corada et al, 1999, 2001). We treated HUVEC confluent monolayers with BV9 to dismantle VE-cadherin from intercellular junctions and induce its redistribution on the cell membrane (Supplementary Figure 7S). Consistent with the Ca²⁺ chelation results, we observed a substantial decrease of TGF--dependent Smad3 phosphorylation in confluent HUVECs treated with BV9 compared with isotype control-treated cells (Figure 7). Overall, these data indicate that VE-cadherin clustering at interendothelial junctions is required for efficient TGF--induced Smad phosphorylation.

Figure 7.

Dismantling of VE-cadherin clusters at endothelial contacts inhibits Smad3 phosphorylation. HUVECs grown to confluence were serum starved, incubated for 5 h with VE-cadherin blocking mAb (BV9), or isotype control IgG, and then treated with TGF-1 for the indicated time intervals. Total cell extracts were blotted with anti-pSmad3 or Smad3 antibodies (upper panels), quantitated by densitometry scan, corrected for background and plotted as pSmad3/Smad3 ratio (lower panel).

View full figure (46 KB)

Clustered VE-cadherin binds TRs and promotes their assembly into an active receptor complex

TRs exist as multimeric complexes on the cell surface and their assembly has a strong impact on mediating and regulating intracellular signals (Shi and Massague, 2003). To test whether VE-cadherin may induce complex formation between TRII and TGF- type I receptors, we immunoprecipitated endogenous ALK5 or ALK1 from VEC-positive and -null cells treated with 5 ng/ml TGF- for 45 min, and probed the immune complexes for the associated TRII (Figure 8). As expected, TRII was found to form a complex with ALK1 and ALK5 in the presence of ligand. However, in contrast to ALK5, we observed a consistently weaker ligand-induced association of ALK1 and TRII receptors. This effect, in the presence of signalling, as judged by Smad1/5 phosphorylation (data not shown), is possibly due to a lower ALK1 relative expression and/or affinity of the ALK1 antibody. Importantly, however, analysis of ALK5/TRII and ALK1/TRII complexes revealed that they were increased by VE-cadherin (Figure 8). Taken together, these data show that clustered VE-cadherin may enhance cellular responses to TGF- by facilitating ligand-dependent assembly of TRs into an active receptor complex, thus regulating their ability to induce R-Smad phosphorylation and signal propagation.

Figure 8.

TGF--mediated assembly of TRII/TRI heteromers is increased by VE-cadherin. (A) TRII/ALK5 complexes are enhanced in TGF--treated VEC-positive cells. Lysates from VEC-null (Null) and -positive (Pos) cells, unstimulated (-) or treated (+) with TGF-1 (5 ng/ml) for 45 min, were subjected to immunoprecipitation (IP) with a protein A-sepharose-conjugated ALK5 antibody or isotype control IgG, followed by immunoblotting (WB) with TRII and ALK5 antibody, as control for immunoprecipitation efficiency. Membranes were quantitated by densitometry scan, corrected for background and plotted as TRII/ALK5 ratio (lower panel). (B) Increased TRII/ALK1 complexes in TGF--treated VEC-positive cells. Cells were treated as described in (A) and ALK1/TRII association was detected by immunoprecipitation (IP) with a protein A-sepharose-conjugated ALK1 antibody, followed by immunoblotting (WB) with TRII and ALK1 antibody. Membranes were quantitated as described in (A) (lower panel).

View full figure (66 KB)

Allantoises from VE-cadherin-null embryos exhibit reduced TGF-/Smad signalling

To test the relevance of the in vitro observations in an ex vivo model of vasculogenesis and angiogenesis, we used the mouse allantois organ culture model (Drake and Fleming, 2000; Downs et al, 2001). The allantois is an extraembryonic tissue undergoing extensive neovascularization in the developing mouse embryo at 8.5 d.p.c. (Argraves et al, 2002). When explanted, 8.5 d.p.c. allantoises display a highly branched network of capillary-like vessels after 24 h of culture; this property was much reduced in allantoises derived from VE-cadherin knockout (VE^-/-) embryos (Crosby et al, 2005). To test whether the loss of VE-cadherin might impair TGF- signalling, we investigated by immunofluorescence and confocal microscopy the nuclear localization of ALK5-dependent Smads in capillary-like vessels from VE^-/- and VE^+/+ allantoises. Consistent with the abundant expression of TGF- and TRs in developing allantois (Jonker and Arthur, 2002), pSmad nuclear localization in ECs was much increased upon stimulation of starved allantoises with TGF- (Supplementary Figure 8S). However, we were able to detect a consistent and significant decrease of pSmad2/3 (27%) nuclear localization in TGF--treated VE^-/- vs VE^+/+vessels (Figure 9), implying a role for VE-cadherin as positive regulator of TGF- signalling levels also in this experimental setting. Together with the in vitro studies, these data implicate adherens junction signalling as a significant contributor to TGF--induced maturation of blood vessels in vivo.

Figure 9.

Loss of VE-cadherin reduces nuclear pSmad2/3 levels in TGF--treated allantois explants. (A) TGF--induced pSmad2/3 nuclear localization is reduced in VE-cadherin-null vessels. E8.5 allantoises from wild-type (VE^+/+) and VE-cadherin null (VE^-/-) mouse embryos were cultured ex vivo as described in Supplementary data. Localization of pSmad2/3 in ECs was visualized with antibodies to pSmad2/3 (green) and PECAM (red). Overlay of the two images is shown. Endothelial nuclei were visualized with DAPI (blue) and PECAM (red). Dotted lines mark endothelial nuclei. Areas marked by a rectangle are enlarged and shown as insets. Scale bar=20 m. (B) Quantification of endothelial pSmad2/3 signal from confocal microscopy images. Data are expressed as % decrease in phosphorylation compared with VE^+/+ cells referred as 100% (meanA.D. of 50 endothelial nuclei scored from three to four embryos per each genotype). ^*P=0.0001.

View full figure (239 KB)

In this study, we describe a novel pathway regulating TGF- signalling in the endothelium. We found that the expression of the EC-specific adherens junction protein VE-cadherin enhances TGF--induced biological responses in ECs. Several lines of evidence support this conclusion. First, EC lines from VE-cadherin-null mice or primary ECs treated with VE-cadherin siRNA show attenuated TGF--induced transcriptional, antiproliferative and antimigratory responses even when expressing normal TGF- receptors and Smad complement. Second, VE-cadherin is required for efficient ALK-dependent Smad phosphorylation as shown by faster and more potent activation of Smad1/5 and Smad2/3 signalling and more sustained propagation of Smad2/3 signals over time in VE-cadherin-expressing ECs. Third, specific inhibition of VE-cadherin clustering in endothelial monolayers results in decreased TGF-/Smad signalling. When junctions are dismantled by Ca²⁺ depletion or by VE-cadherin blocking antibodies, the TGF-/Smad phosphorylation is largely attenuated clearly demonstrating that TGF- signalling in ECs is critically dependent on proper VE-cadherin engagement at adherens junctions.

We found that cell surface TRII codistributes with VE-cadherin at intercellular contacts and that TGF- increases TRII accumulation at adherens junctions (Figure 6). Based upon previous studies showing that TRII can be recruited into tight junctions through binding to the structural component Par6 (Ozdamar et al, 2005), and the observed physical interaction of VE-cadherin with TRII independently of the TGF- type I receptors and vice versa (Figure 5), we suggest that at junctional contacts VE-cadherin favours the reciprocal association of TRII and TGF- type I receptors in the presence of ligand. Indeed, the association of TRII with ALK1 and ALK5 is increased by VE-cadherin explaining the increment of Smad1/5 and Smad2/3 phosphorylation and signalling.

Previous studies have shown that TGF-/ALK5/Smad2/3 pathway leads to inhibition of cell migration and proliferation, whereas signalling through TGF-/ALK1/Smad1/5/8 can promote cell growth and motility (Goumans et al, 2002, 2003). In our experimental setting, the net downstream effect of VE-cadherin-mediated increment in TGF- signalling is inhibition of cell growth and motility. This was unexpected as not only ALK5 but also ALK1 receptor was activated. However, we observed that Smad2/3 activation was more persistent than that of Smad1/5 in the presence of VE-cadherin, indicating that at later times of TGF- stimulation junctional VE-cadherin might tip the balance between these two pathways in favour of ALK5 signalling. As VE-cadherin associates with ALK1, which is also a signalling receptor for bone morphogenetic protein (BMP)-9 and -10, two cytokines with strong antiangiogenic activities in ECs (David et al, 2007; Scharpfenecker et al, 2007), VE-cadherin may have similar functions in BMP pathways although further studies are required to clarify this aspect.

The clinical manifestations and symptoms of HHT vary greatly even within a family, suggesting a potential role for modifier genes in the progression of this pathology. At this stage, a connection between the observations reported here and HHT is only speculative. However, it is conceivable that conditions which impair VE-cadherin expression or clustering at junctions may attenuate TGF- signalling. VE-cadherin activity is strongly downregulated by inflammatory cytokines or agents that increase permeability such as histamine, thrombin or VEGF (Dejana, 2004). It is therefore tempting to speculate that when players of TGF- signalling, such as endoglin and ALK1, are reduced, as in HHT, a further impairment of TGF- signalling by a local reduction of VE-cadherin may cause vascular fragility and haemorrhages. It is interesting that VE-cadherin-2, a member of the protocadherin family (Telo et al, 1998), was reported to be downregulated in HHT1 and HHT2 blood outgrowth ECs (Fernandez-Lopez et al, 2007).

The role of VE-cadherin expression and clustering in the modulation of growth factor receptor signalling was previously described for VEGFR-2 (Lampugnani et al, 2003). In this case however, and in contrast to TGF- receptors, the association of VE-cadherin with the receptor caused a decrease in its tyrosine phosphorylation and proliferative signals. Inhibition of VEGFR-2 signalling by VE-cadherin is mediated by the phosphatase DEP-1/CD148 which is co-clustered at junctions, dephosphorylates the receptor and limits its internalization and signalling (Lampugnani et al, 2006). In contrast to VEGFR-2, it is likely that the action of VE-cadherin on TGF- pathway relies on the capacity of the junctional protein to associate with all the components of the TGF- receptor complex enhancing, in this way, the interaction between TRII and TGF- type I receptors and increasing the downstream signalling. A general concept that may accommodate these observations is that VE-cadherin acts by inducing vascular stabilization. Through the clustering of this protein, ECs are able to sense the presence of other adjoining identical cells and react by limiting their growth and migration. This action may be exerted by VE-cadherin recruiting growth factor receptors and modulating their activity. These opposing functions of VE-cadherin, inhibitory on VEGFR-2 and stimulatory on TRII/ALK signalling, support a model (Figure 10) whereby EC response to growth factors is context dependent and ECs activated by the same growth factor may respond in opposite ways whether they are confluent and VE-cadherin organized at junctions or not. This model has important biological implications as it describes the quiescent phenotype of ECs as an active state mediated by continuous inhibition of growth and enhancement of stabilization signals.

Figure 10.

Proposed model of the role of VE-cadherin in quiescent endothelium. VE-cadherin regulates VEGF/ERK1/2 and TGF-/Smad signalling pathways in opposite ways and with identical biological outcome in ECs. The formation of the VE-cadherin–VEGFR-2 complex limits cell proliferation by preventing ERK1/2 MAPK activation (Lampugnani et al, 2006). The formation of VE-cadherin–TR complex promotes TGF--induced Smad phosphorylation leading to vascular stabilization.

View full figure (77 KB)

Transient transfections and transcriptional reporter assays

Transient transfections and reporter assays were performed as previously described (Felici et al, 2003). In all the reporter assays, a -galactosidase expression vector was used as an internal control to correct for transfection efficiency.

Quantitative real-time RT–PCR

Total RNA from VEC cells was isolated with the RNeasy mini kit (Quiagen Inc.). cDNA was synthesized, amplified and analysed on an ABI PRISM^®7900 as previously described (Spagnuolo et al, 2004).

EC proliferation assay

VEC cells were seeded in starvation medium (DMEM, 2% FCS) at a density of 5 10⁴ cells per well in 24-well plates. After 12 h, TGF-1 was added at 5 ng/ml. At the indicated time of culture, cells were trypsinized and the number of viable cells, in triplicate wells, was determined by Trypan blue and haemocytometer counting.

VE-cadherin gene silencing by siRNA

For VE-cadherin silencing, ON-TARGETplus SMARTpool duplex siRNAs were used as described in Supplementary data.

Wound-healing migration assay

EC monolayers were wounded with a yellow tip to make a 100 m scratch. The wound-induced cell migration in the presence of TGF-1 (5 ng/ml) was followed by time-lapse microscopy (HUVECs) or crystal violet staining (0.5% crystal violet in 20% methanol) and phase contrast microscopy (VEC cells). The migration of the cell front was analysed by ImageJ software.

Immunoprecipitation and western blot analysis

Immunoprecipitation of total TRs and cell surface biotinylated TGF- receptors were performed as described previously (Felici et al, 2003; Pece-Barbara et al, 2005) and in Supplementary data.

Immunofluorescence microscopy

Immunofluorescence analysis of VE-cadherin and ectopically expressed or endogenous TGF- receptors was performed as described previously (Felici et al, 2003; Lampugnani et al, 2006) and in Supplementary data.

VE-cadherin declustering

For disruption of VE-cadherin clustering, EC monolayers were treated as described previously (Corada et al, 1999, 2001) and in Supplementary data.

Allantois culture

Allantoises were dissected from 8.5 d.p.c. mouse embryos and cultured as described previously (Argraves et al, 2002) and in Supplementary data.

Statistical analysis

A Student's two-tailed non-paired t-test was used to determine the statistical significance. P<0.05 was considered statistically significant.

We dedicate this study to the memory of Dr Anita Roberts who passed away in the summer of 2006. We thank C Hill for plasmids, G Balconi for MEECs and A Zanetti for HUVECs. We are grateful to Midory Thorikay and Maarten van Dinther for excellent technical assistance and to A Cattelino and C Francavilla for valuable discussions. We thank N Offenhäuser for her critical reading of the paper and U Cavallaro for scientific advice. This study was supported by Associazione Italiana per la Ricerca sul Cancro, Association for International Cancer Research, the European Community (Integrated Project Contract no. LSHG-CT-2004-503573; NoE MAIN 502935; NoE EVGN 503254; EUSTROKE consortium; Angioscaff consortium), Istituto Superiore di Sanità, Italian Ministry of Health, MIUR (COFIN prot: 2006058482_002), Fondation Leducq Transatlantic Network of Excellence, Dutch Cancer Society (UL 2005–3371), EC project 'Angiotargeting Integrated Project' no. 504743 and Canadian Institute of Health Research (MT-6247). CG was supported by an AIRC fellowship (AIRC-SISAL).

Abdalla SA, Pece-Barbara N, Vera S, Tapia E, Paez E, Bernabeu C, Letarte M (2000) Analysis of ALK-1 and endoglin in newborns from families with hereditary hemorrhagic telangiectasia type 2. Hum Mol Genet 9: 1227–1237 | Article | PubMed | ChemPort |

Argraves WS, Larue AC, Fleming PA, Drake CJ (2002) VEGF signaling is required for the assembly but not the maintenance of embryonic blood vessels. Dev Dyn 225: 298–304 | Article | PubMed | ChemPort |

Arthur HM, Ure J, Smith AJ, Renforth G, Wilson DI, Torsney E, Charlton R, Parums DV, Jowett T, Marchuk DA, Burn J, Diamond AG (2000) Endoglin, an ancillary TGFbeta receptor, is required for extraembryonic angiogenesis and plays a key role in heart development. Dev Biol 217: 42–53 | Article | PubMed | ISI | ChemPort |

Attisano L, Wrana JL (2002) Signal transduction by the TGF-beta superfamily. Science 296: 1646–1647 | Article | PubMed | ISI | ChemPort |

Balconi G, Spagnuolo R, Dejana E (2000) Development of endothelial cell lines from embryonic stem cells: a tool for studying genetically manipulated endothelial cells in vitro. Arterioscler Thromb Vasc Biol 20: 1443–1451 | PubMed | ISI | ChemPort |

Barbara NP, Wrana JL, Letarte M (1999) Endoglin is an accessory protein that interacts with the signaling receptor complex of multiple members of the transforming growth factor-beta superfamily. J Biol Chem 274: 584–594 | Article | PubMed | ISI | ChemPort |

Basson CT, Kocher O, Basson MD, Asis A, Madri JA (1992) Differential modulation of vascular cell integrin and extracellular matrix expression in vitro by TGF-beta 1 correlates with reciprocal effects on cell migration. J Cell Physiol 153: 118–128 | Article | PubMed | ISI | ChemPort |

Bourdeau A, Dumont DJ, Letarte M (1999) A murine model of hereditary hemorrhagic telangiectasia. J Clin Invest 104: 1343–1351 | Article | PubMed | ISI | ChemPort |

Breviario F, Caveda L, Corada M, Martin-Padura I, Navarro P, Golay J, Introna M, Gulino D, Lampugnani MG, Dejana E (1995) Functional properties of human vascular endothelial cadherin (7B4/cadherin-5), an endothelium-specific cadherin. Arterioscler Thromb Vasc Biol 15: 1229–1239 | PubMed | ISI | ChemPort |

Carmeliet P, Lampugnani MG, Moons L, Breviario F, Compernolle V, Bono F, Balconi G, Spagnuolo R, Oostuyse B, Dewerchin M, Zanetti A, Angellilo A, Mattot V, Nuyens D, Lutgens E, Clotman F, de Ruiter MC, Gittenberger-de Groot A, Poelmann R, Lupu F et al (1999) Targeted deficiency or cytosolic truncation of the VE-cadherin gene in mice impairs VEGF-mediated endothelial survival and angiogenesis. Cell 98: 147–157 | Article | PubMed | ISI | ChemPort |

Cattelino A, Liebner S, Gallini R, Zanetti A, Balconi G, Corsi A, Bianco P, Wolburg H, Moore R, Oreda B, Kemler R, Dejana E (2003) The conditional inactivation of the beta-catenin gene in endothelial cells causes a defective vascular pattern and increased vascular fragility. J Cell Biol 162: 1111–1122 | Article | PubMed | ISI | ChemPort |

Corada M, Liao F, Lindgren M, Lampugnani MG, Breviario F, Frank R, Muller WA, Hicklin DJ, Bohlen P, Dejana E (2001) Monoclonal antibodies directed to different regions of vascular endothelial cadherin extracellular domain affect adhesion and clustering of the protein and modulate endothelial permeability. Blood 97: 1679–1684 | Article | PubMed | ISI | ChemPort |

Corada M, Mariotti M, Thurston G, Smith K, Kunkel R, Brockhaus M, Lampugnani MG, Martin-Padura I, Stoppacciaro A, Ruco L, McDonald DM, Ward PA, Dejana E (1999) Vascular endothelial-cadherin is an important determinant of microvascular integrity in vivo. Proc Natl Acad Sci USA 96: 9815–9820 | Article | PubMed | ChemPort |

Crosby CV, Fleming PA, Argraves WS, Corada M, Zanetta L, Dejana E, Drake CJ (2005) VE-cadherin is not required for the formation of nascent blood vessels but acts to prevent their disassembly. Blood 105: 2771–2776 | Article | PubMed | ChemPort |

David L, Mallet C, Vailhe B, Lamouille S, Feige JJ, Bailly S (2007) Activin receptor-like kinase 1 inhibits human microvascular endothelial cell migration: potential roles for JNK and ERK. J Cell Physiol 213: 484–489 | Article | PubMed | ChemPort |

Dejana E (2004) Endothelial cell–cell junctions: happy together. Nat Rev Mol Cell Biol 5: 261–270 | Article | PubMed | ISI | ChemPort |

Dennler S, Itoh S, Vivien D, ten Dijke P, Huet S, Gauthier JM (1998) Direct binding of Smad3 and Smad4 to critical TGF beta-inducible elements in the promoter of human plasminogen activator inhibitor-type 1 gene. EMBO J 17: 3091–3100 | Article | PubMed | ISI | ChemPort |

Derynck R, Zhang YE (2003) Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature 425: 577–584 | Article | PubMed | ISI | ChemPort |

Downs KM, Temkin R, Gifford S, McHugh J (2001) Study of the murine allantois by allantoic explants. Dev Biol 233: 347–364 | Article | PubMed | ChemPort |

Drake CJ, Fleming PA (2000) Vasculogenesis in the day 6.5–9.5 mouse embryo. Blood 95: 1671–1679 | PubMed | ISI | ChemPort |

Felici A, Wurthner JU, Parks WT, Giam LR, Reiss M, Karpova TS, McNally JG, Roberts AB (2003) TLP, a novel modulator of TGF-beta signaling, has opposite effects on Smad2- and Smad3-dependent signaling. EMBO J 22: 4465–4477 | Article | PubMed | ISI | ChemPort |

Fernandez-Lopez A, Garrido-Martin EM, Sanz-Rodriguez F, Pericacho M, Rodriguez-Barbero A, Eleno N, Lopez-Novoa JM, Duwell A, Vega MA, Bernabeu C, Botella LM (2007) Gene expression fingerprinting for human hereditary hemorrhagic telangiectasia. Hum Mol Genet 16: 1515–1533 | Article | PubMed | ChemPort |

Goumans MJ, Valdimarsdottir G, Itoh S, Lebrin F, Larsson J, Mummery C, Karlsson S, ten Dijke P (2003) Activin receptor-like kinase (ALK)1 is an antagonistic mediator of lateral TGFbeta/ALK5 signaling. Mol Cell 12: 817–828 | Article | PubMed | ISI | ChemPort |

Goumans MJ, Valdimarsdottir G, Itoh S, Rosendahl A, Sideras P, ten Dijke P (2002) Balancing the activation state of the endothelium via two distinct TGF-beta type I receptors. EMBO J 21: 1743–1753 | Article | PubMed | ISI | ChemPort |

Gumbiner BM (2005) Regulation of cadherin-mediated adhesion in morphogenesis. Nat Rev Mol Cell Biol 6: 622–634 | Article | PubMed | ISI | ChemPort |

Iurlaro M, Demontis F, Corada M, Zanetta L, Drake C, Gariboldi M, Peiro S, Cano A, Navarro P, Cattelino A, Tognin S, Marchisio PC, Dejana E (2004) VE-cadherin expression and clustering maintain low levels of survivin in endothelial cells. Am J Pathol 165: 181–189 | PubMed | ISI | ChemPort |

Jonker L, Arthur HM (2002) Endoglin expression in early development is associated with vasculogenesis and angiogenesis. Mech Dev 110: 193–196 | Article | PubMed | ISI | ChemPort |

Lampugnani MG, Orsenigo F, Gagliani MC, Tacchetti C, Dejana E (2006) Vascular endothelial cadherin controls VEGFR-2 internalization and signaling from intracellular compartments. J Cell Biol 174: 593–604 | Article | PubMed | ChemPort |

Lampugnani MG, Zanetti A, Breviario F, Balconi G, Orsenigo F, Corada M, Spagnuolo R, Betson M, Braga V, Dejana E (2002) VE-cadherin regulates endothelial actin activating Rac and increasing membrane association of Tiam. Mol Biol Cell 13: 1175–1189 | Article | PubMed | ISI | ChemPort |

Lampugnani M, Zanetti A, Corada M, Takahashi T, Balconi G, Breviario F, Orsenigo F, Cattelino A, Kemler R, Daniel TO, Dejana E (2003) Contact inhibition of VEGF-induced proliferation requires vascular endothelial cadherin, beta-catenin, and the phosphatase DEP-1/CD148. J Cell Biol 161: 793–804 | Article | PubMed | ISI | ChemPort |

Lebrin F, Goumans MJ, Jonker L, Carvalho RL, Valdimarsdottir G, Thorikay M, Mummery C, Arthur HM, ten Dijke P (2004) Endoglin promotes endothelial cell proliferation and TGF-beta/ALK1 signal transduction. EMBO J 23: 4018–4028 | Article | PubMed | ISI | ChemPort |

Lesca G, Burnichon N, Raux G, Tosi M, Pinson S, Marion MJ, Babin E, Gilbert-Dussardier B, Riviere S, Goizet C, Faivre L, Plauchu H, Frebourg T, Calender A, Giraud S (2006) Distribution of ENG and ACVRL1 (ALK1) mutations in French HHT patients. Hum Mutat 27: 598 | Article | PubMed

Li DY, Sorensen LK, Brooke BS, Urness LD, Davis EC, Taylor DG, Boak BB, Wendel DP (1999) Defective angiogenesis in mice lacking endoglin. Science 284: 1534–1537 | Article | PubMed | ISI | ChemPort |

Liebner S, Cattelino A, Gallini R, Rudini N, Iurlaro M, Piccolo S, Dejana E (2004) Beta-catenin is required for endothelial–mesenchymal transformation during heart cushion development in the mouse. J Cell Biol 166: 359–367 | Article | PubMed | ISI | ChemPort |

Liebner S, Cavallaro U, Dejana E (2006) The multiple languages of endothelial cell-to-cell communication. Arterioscler Thromb Vasc Biol 26: 1431–1438 | Article | PubMed | ChemPort |

Muller G, Behrens J, Nussbaumer U, Bohlen P, Birchmeier W (1987) Inhibitory action of transforming growth factor beta on endothelial cells. Proc Natl Acad Sci USA 84: 5600–5604 | Article | PubMed | ISI | ChemPort |

Navarro P, Caveda L, Breviario F, Mandoteanu I, Lampugnani MG, Dejana E (1995) Catenin-dependent and -independent functions of vascular endothelial cadherin. J Biol Chem 270: 30965–30972 | Article | PubMed | ISI | ChemPort |

Oh SP, Seki T, Goss KA, Imamura T, Yi Y, Donahoe PK, Li L, Miyazono K, ten Dijke P, Kim S, Li E (2000) Activin receptor-like kinase 1 modulates transforming growth factor-beta 1 signaling in the regulation of angiogenesis. Proc Natl Acad Sci USA 97: 2626–2631 | Article | PubMed | ChemPort |

Ozdamar B, Bose R, Barrios-Rodiles M, Wang HR, Zhang Y, Wrana JL (2005) Regulation of the polarity protein Par6 by TGFbeta receptors controls epithelial cell plasticity. Science 307: 1603–1609 | Article | PubMed | ISI | ChemPort |

Pece-Barbara N, Vera S, Kathirkamathamby K, Liebner S, Di Guglielmo GM, Dejana E, Wrana JL, Letarte M (2005) Endoglin null endothelial cells proliferate faster and are more responsive to transforming growth factor beta1 with higher affinity receptors and an activated Alk1 pathway. J Biol Chem 280: 27800–27808 | Article | PubMed | ISI | ChemPort |

Prigoda NL, Savas S, Abdalla SA, Piovesan B, Rushlow D, Vandezande K, Zhang E, Ozcelik H, Gallie BL, Letarte M (2006) Hereditary haemorrhagic telangiectasia: mutation detection, test sensitivity and novel mutations. J Med Genet 43: 722–728 | Article | PubMed | ChemPort |

Scharpfenecker M, van Dinther M, Liu Z, van Bezooijen RL, Zhao Q, Pukac L, Lowik CW, ten Dijke P (2007) BMP-9 signals via ALK1 and inhibits bFGF-induced endothelial cell proliferation and VEGF-stimulated angiogenesis. J Cell Sci 120: 964–972 | Article | PubMed | ISI | ChemPort |

Shi Y, Massague J (2003) Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell 113: 685–700 | Article | PubMed | ISI | ChemPort |

Sorensen LK, Brooke BS, Li DY, Urness LD (2003) Loss of distinct arterial and venous boundaries in mice lacking endoglin, a vascular-specific TGFbeta coreceptor. Dev Biol 261: 235–250 | Article | PubMed | ISI | ChemPort |

Spagnuolo R, Corada M, Orsenigo F, Zanetta L, Deuschle U, Sandy P, Schneider C, Drake CJ, Breviario F, Dejana E (2004) Gas1 is induced by VE-cadherin and vascular endothelial growth factor and inhibits endothelial cell apoptosis. Blood 103: 3005–3012 | Article | PubMed | ISI | ChemPort |

Telo P, Breviario F, Huber P, Panzeri C, Dejana E (1998) Identification of a novel cadherin (vascular endothelial cadherin-2) located at intercellular junctions in endothelial cells. J Biol Chem 273: 17565–17572 | Article | PubMed | ChemPort |

Urness LD, Sorensen LK, Li DY (2000) Arteriovenous malformations in mice lacking activin receptor-like kinase-1. Nat Genet 26: 328–331 | Article | PubMed | ISI | ChemPort |

Zanetta L, Corada M, Grazia Lampugnani M, Zanetti A, Breviario F, Moons L, Carmeliet P, Pepper MS, Dejana E (2005) Downregulation of vascular endothelial-cadherin expression is associated with an increase in vascular tumor growth and hemorrhagic complications. Thromb Haemost 93: 1041–1046 | PubMed | ChemPort |

Main navigation

• Search EMBO Journal

  Search

• Advanced search

• The EMBO Journal

  • Home
  • Aims and scope
  • Current issue
  • Advance online publication
  • Web focuses
  • Archive:
  • Browse by issue
  • Browse by subject
  • Browse by category
  • Free online sample issue
  • Press releases

• Authors and Referees

  • Editorial process
  • Guide for authors
  • Submit an article
  • Guide for referees
  • Editors and Editorial Board
  • Contact editorial office

• Customer Service

  • Subscribe
  • Order sample copy
  • Purchase articles
  • Reprints and permissions
  • Contact NPG
  • Advertising