The Role of Wnt Signaling in Physiological and Pathological Angiogenesis

Transcriptional Regulation of Endothelial Cell Differentiation

All the steps of endothelial cell differentiation are orchestrated by specific transcription factors (Figure 1). Some of the transcriptional mechanisms that determine endothelial identity and then arterial, venous, or lymphatic differentiation have been partially identified (reviewed elsewhere⁷). However, very little is known about the transcription factors that induce organ-specific differentiation of the endothelium.

During the last decade, several groups dedicated a considerable effort in trying to decode the transcriptional mechanisms directing cells of mesodermal origin toward a specific endothelial identity. In contrast to other tissues, endothelial cell differentiation is mediated by several transcription factors that are not cell-specific but act in concert to modulate expression of vascular markers. Through the analysis of the promoter regions of endothelial-specific genes, different transcription factors have been identified as putative inducers of a “common” endothelial cell identity⁸: these include members of the Ets family, several homeobox (Hox) transcription factors, and members of Soxs (Sry-related high-mobility group box), in particular, the F subgroup (Sox18, Sox17, and Sox7). However, the identification of the specific transcription factors within each class able to transactivate one or another gene has been only partially accomplished.

After acquisition of “panendothelial” characteristics, endothelial cells undergo further differentiation. It was believed that arterial and venous determination was mostly attributable to endothelial exposure to differing blood flow rates and pressure. However, several studies on different animal models suggest that venous and arterial determination precedes the onset of the circulation (reviewed elsewhere^2,3,7,–,9). Notch signaling and activation of downstream transcription factors (Hey1 and Hey2 [Hes-related proteins]) are required to promote arterial differentiation and to suppress venous fate choices.^10,11 In contrast, the transcription factor COUP-TFII is required for venous identity. This factor acts by inhibiting Notch signaling and other effectors of arterial specification.¹² Lymphatic endothelial cells differentiate from venous endothelium. The transcriptional factor Sox18, in cooperation with COUP-TFII, promotes expression of Prox-1, which plays a crucial role in final lymphatic determination and maintenance.¹³ Other transcription factors such as FOXC1 and FOXC2 act in lymphatic remodeling and valve formation.^5,14,–,16

Wnt Signaling in Endothelial Cells

Recent reports underline the important role of the Wnt system in vascular morphogenesis in the embryo and in organ-specific endothelial cell differentiation.^4,17,–,21

Wnt signaling regulates fundamental aspects of development including cell fate specification, proliferation, survival, and overall organogenesis (reviewed elsewhere^17,22,–,26). For an updated and comprehensive review of these aspects in the cardiovascular system, see the article by Rao and Kühl²⁷ in this same series of reviews. Wnt factors use different receptors and signaling pathways. The most-studied and best-understood Wnt signaling pathway is mediated by the transcriptional activity of β-catenin (Figure 2). In absence of Wnt, cytoplasmic β-catenin is degraded by the action of a multiprotein complex (reviewed elsewhere^22,–,28), and Wnt target genes are repressed. The binding of Wnt to the Frizzled/Lrp (low-density lipoprotein receptor–related protein) receptor complex inhibits β-catenin degradation complex and stabilizes β-catenin in the cytoplasm, promoting its translocation to the nucleus. In the nucleus, β-catenin interacts with T-cell factor (TCF)/Lef transcription factors and activates cell transcription.^22,29,–,33

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

β-Catenin–dependent/canonical Wnt signaling in the endothelium. The figure represents 3 contiguous cells. In the presence of Wnt, Frizzled receptors and Lrp5/6 complexes are triggered, and this leads to β-catenin stabilization in the cytoplasm and translocation to the nucleus. In the nucleus, β-catenin binds TCF/Lef transcription factors, inducing Wnt target gene transcription. In the retina vasculature, Frizzled receptors may bind Norrin and are multimerized with the contribution of the tetraspanin protein TSPAN21. R-spondin activates Wnt signaling through LRP binding, whereas DKK and SERPINA3K inhibit the binding of Wnt to these receptor elements. The figure is based on previously published reviews.^{17,22,–,24,26,28,54} β-Catenin is also linked to cadherins (VE-cadherin in endothelial cells) in confluent cells. When cells are migrating, as in sprouting vessels, or exposed to growth or permeability increasing agents, β-catenin may detach from the membrane and translocate to the nucleus.^37,51 This mechanism contributes further to β-catenin transcriptional signaling.

Of note, β-catenin also binds to cadherins (VE- and N-cadherin in endothelial cells) at cell-to-cell adherens junctions and stabilizes their interaction with the cytoskeleton (Figure 2). In this way, β-catenin acts by stabilizing cell-to-cell adhesion and tissue integrity.^34,35 However, when cells are migrating or in sparse conditions and junctions are partially dismantled, a pool of β-catenin^36,37 dissociates from junctions and translocates to the nucleus. In nonvascular cells, recent evidence suggests that there are 2 pools of β-catenin at the membrane.³⁸ One is linked to the cadherin complex, which is protected from degradation. A second pool of β-catenin associates to a membrane bound phosphodestruction complex formed by adenomatous polyposis coli (APC) and GSK3, which promotes its degradation. Cadherins can promote the binding of β-catenin to the complex and limits in this way Wnt signaling. Thus, Wnt signaling is indirectly modulated by intercellular junctions.

The Wnt factors can signal also through noncanonical β-catenin–independent pathways.^24,–,28

The Ca²⁺ pathway is induced by the binding of some members of the Wnt family to specific Frizzled receptors, which, in turn, activate the phospholipase C and increase intracellular Ca²⁺.^23,24

Another β-catenin–independent Wnt signaling pathway is the planar cell polarity (PCP) pathway, which activates the small GTPases RhoA and Rac and modulates cytoskeletal rearrangements and PCP.^23,24,39

Which Wnt ligand binds and activates which Frizzled receptor and whether all Frizzled receptors trigger β-catenin dependent or independent signaling remains an open issue. Different Wnt receptors are responsible for inducing one or another signaling pathway, although some receptors⁴⁰ may activate both canonical and noncanonical pathways. It is likely that the final response is cell context–dependent, ie, related to the type of receptors expressed and the presence or absence of signaling partners in a given cell type at a given moment.

Noncanonical β-Catenin–Independent Wnt Signaling Pathways in Endothelial Cells

Some recent publications underline the importance of noncanonical Wnt pathways in regulating endothelial cell growth and angiogenesis.^20,40,–,44 Wnt activation of the Ca²⁺ pathway induces cell proliferation and inhibits apoptosis in cultured endothelial cells, suggesting a proangiogenic activity in vivo.^41,42 PCP also plays a role in vascular morphology, vessel sprouting, and elongation, suggesting a morphogenetic role for this Wnt signaling pathway in vascular development.⁴⁴ Recently, a member of a subgroup of formins (DAAM1 [Dishevelled-associated activator of morphogenesis]) was found to act downstream of the Wnt/PCP pathway in inhibiting endothelial cell proliferation and angiogenesis.⁴⁵ The mechanism of action of DAAM1 is likely to coordinate microtubule assembly and stabilization in the endothelium.

Although the molecular basis of β-catenin–independent signaling pathways is still poorly understood, there is a general consensus that β-catenin–independent signaling antagonizes β-catenin pathway.²⁴ Crosstalk between the different Wnt signaling cascades is possible, and it is also possible that all Frizzled receptors may couple to β-catenin–dependent and –independent pathways. Samarzija et al⁴⁶ observed that Wnt3a induces proliferation and migration of endothelial cells via canonical and noncanonical Wnt signaling pathways.

Using an embryonic stem cells (ES) differentiation system in vitro, Yamashita et al⁴⁷ reported that activation of the cells with Wnt3a, which mostly signals through β-catenin, preferentially augments endothelial differentiation while inhibiting cardiomyocyte differentiation. However, Hwang et al,⁴⁸ using a similar system, found that Wnt5a, which most frequently signals through noncanonical pathways, directs endothelial differentiation. These observations underline the complexity of the system and how cell responses may be determined by the cellular context, number, and types of Wnt receptors which are activated.

Another β-catenin–independent pathway is mediated by phosphorylation of APC (a member of the destruction complex of β-catenin).⁴⁹ It has been reported that phosphorylated APC localizes at the tips of cellular extension during cell migration and is required for cell migration and cytoskeletal organization.⁴⁹ These effects are independent of β-catenin transcriptional activity.

Overall, data available on noncanonical Wnt signaling in vascular development are still scarce, and additional investigation is required to define which functions and at which steps of embryo development noncanonical Wnt pathways may act. The following part of the review is mostly focused on β-catenin–dependent signaling on which more data are currently available.

Canonical Wnt Signaling in the Embryo Vasculature

By a systematic RT-PCR analysis, it was found that cultured endothelial cells express quite a large array of Frizzled receptors and Wnt factors, together with Lrp5/6.⁵⁰ This suggests that endothelial cells may be particularly sensitive to Wnt signaling and possibly undergo autocrine stimulation. However, reporter transgenic mouse lines in which Lac-Z expression is under the control of β-catenin–TCF/Lef complex showed relatively low activation of β-catenin transcriptional signaling in the resting vasculature of adult mice. In contrast, detectable activation of Wnt signaling was observed during embryonic development⁵¹ in many types of vessels and in the brain microvasculature (see below). Thus, Wnt signaling is important at early steps of vasculogenesis and angiogenesis but is likely silenced in healthy, resting vasculature.

This possibility is confirmed by gene inactivation experiments of different members of the Wnt signaling pathway (for a comprehensive review of these mouse embryo phenotypes, see Franco et al¹⁷). Ablation of Frizzled-5⁵² and Wnt 2⁵³ leads to a defective placenta vascularization and, in the case of Frizzled-5, also to defective remodeling of the yolk sac vasculature. Similarly, both loss-of-function and gain-of-function mutations of Frizzled-4 and Lrp5 resulted in early lethality of the mouse embryo resulting from defects of vascular organization, remodeling, and mesenchymal cell coverage.⁵⁴

Inactivation of R-spondin, a coactivator of Wnt/β-catenin,⁵⁵ caused defective placentation and vascular patterning. It was suggested that these defects were attributable to lack of correct angiogenesis and altered angioblast and blood cell specification.

Endothelial-specific deletion of β-catenin affects the development of the embryonic vasculature and results in early lethality in utero.⁵⁶ The embryos present defective vascular remodeling and diffuse hemorrhages. In some regions, such as the intersomitic vessels or umbilical veins and arteries, the vessels show hyperbranching.^51,56 These alterations were accompanied by changes in the molecular architecture of endothelial adherens junctions, explaining, in part, the presence of hemorrhages, but defects in the transcriptional activity of β-catenin can also contribute to the development of this pathological phenotype.

Furthermore, loss of β-catenin impairs the development of the endocardial cushion and cardiac valves because of altered endothelial mesenchymal transition.⁵⁷

An even more dramatic phenotype was observed by inducing endothelial-specific stabilization (gain of function) of β-catenin.^51,58 The phenotype was characterized by lack of vascular remodeling, altered elongation of the intersomitic vessels, defects in branching, and loss of venous identity. Consistently, endothelial-specific gain-of-function mutation of Frizzled-4 caused similar alterations in the development and specification of the vasculature.⁵⁴ Taken together, these data indicate that β-catenin/Wnt signaling must be strictly controlled for harmonic vascular development. Too much or too little signaling may prevent a correct vascular organization and lead to early fetal lethality.

The Role of Wnt in Eye and Brain Vasculature

β-Catenin–dependent Wnt signaling is strongly implicated in the development and differentiation of retina and brain microvasculature. These particular vessels express many common functional and morphological characteristics, such as well-developed tight junctions and a specialized and selective system of membrane transporters.⁵⁹ This vasculature forms the so-called blood–brain barrier (BBB), which is characterized by a strict control of permeability between blood and the central nervous system. In reporter mice, Wnt signaling was particularly strong during early stages of both retina and brain vascularization.^60,–,62

In the retina, different strategies such as null mutations of Frizzled- 4, Lrp5, Lef-1,^54,60,63,64 and Norrin⁵⁴ (see below) were used to abrogate canonical Wnt signaling. All of these conditions induced similar vascular alterations, such as absence of intraretinal capillaries, defects in vascular patterning, arteriovenous anastomosis, and intraocular hemorrhages. Inactivation of the Wnt cofactor Nrarp also caused defects of retina vasculature, with decrease in branching and radial expansion.⁶⁰

In the embryonic brain, Wnt signaling was quite strong in vessels penetrating into the brain stroma and was still detectable during early postnatal period but declined in the adult.^61,65 Wnt signaling followed the same kinetics of brain angiogenesis and BBB development, supporting the idea that Wnts are involved in brain vascularization and induction of BBB properties. Further support of this concept derives from inactivation of Wnt signaling through different strategies, such as ablation of endothelial β-catenin^61,62 or of Wnt 7a and -b⁶⁶ or administration of the soluble inhibitor Frizzled-8 receptor construct.⁶² Without canonical Wnt signaling, the vessels in the brain parenchyma look enlarged and hemorrhagic. In some regions, endothelial cells grow in multilayers into the vascular lumen, forming multicellular structures.

When β-catenin was ablated postnatally (from postnatal day 1 to 11) vascular morphology was not significantly modified, but vessels presented lack of BBB differentiation and control of permeability.⁶¹ These data suggest that β-catenin signaling is important for a correct brain and retina angiogenesis during development and, at later stages, for inducing BBB properties in the brain microvasculature.

Wnt signaling plays opposite effects in the hyaloids vessels. These vessels form a capillary network within the developing eye, which must regress at late stages of development. Regression is mediated by Wnt 7b released by adjacent macrophages.⁶⁷ Consistently, abrogation of Frizzled-5 also inhibits regression of hyaloid vessels.⁶⁸

Wnt Mechanism of Action

An important issue is the definition of the mechanism of action of canonical Wnt signaling in modulating vascular morphogenesis. Several downstream cofactors and synergistic pathways have been described in endothelial cells of developing vessels.

Corada et al⁵¹ found that during vascular development in the embryo, an endothelial-specific gain-of-function mutation of β-catenin induces activation of the Notch pathway. More specifically, β-catenin increases the transcription of the Notch ligand Delta-like (Dll)4, which, in turn, activates Notch-1 and -4. Overexpression of β-catenin leads to alterations of vascular morphology comparable to those described by overexpression of Dll4,⁶⁹ which include a strong induction of endothelial cell arterialization and lack of venous specification. These effects of β-catenin were detectable only during embryonic vasculogenesis and angiogenesis but were lost in the adult. It appears, therefore, that Wnt and Notch pathways act synergistically in modulating vascular differentiation (Figure 3). On this line, a recent publication⁷⁰ showed that, using an embryonic stem cell differentiation system in vitro, endothelial cell expression of arterial markers is induced by the synergistic activity of Notch and β-catenin. β-Catenin forms a transcriptional complex with RBP-J transcription factor and the intracellular domain of Notch (NICD), which modulates the transcriptional upregulation of arterial genes. Furthermore, it was found⁶⁰ that in the retina vasculature, Notch increases the expression of Nrarp, a transcription factor that, from one side, inhibits Notch activation and, on the other, increases β-catenin signaling by activating Lef-1.⁶⁰ Overall, these data determine a sequential and direct link between β-catenin and Notch signaling systems to tune endothelial cell differentiation and vascular morphogenesis (Figure 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Crosstalk between Wnt and Notch signaling pathways in the endothelium. Notch signaling is mediated by binding of Notch receptors to ligands such as Jagged and Dll4. This induces cleavage of the Notch intracellular peptide NICD, which translocates to the nucleus and mediates cell transcription. Notch increases Nrarp expression; Nrarp, in turn, acts as a negative-feedback mechanism, inactivating Notch signaling while activating β-catenin signaling through interaction with Lef-1. Wnt signaling also synergizes with Notch by upregulating Dll4 and by β-catenin binding to NICD–RBP-J complex.^6,70

Ye et al⁵⁴ found that, in endothelial cells, a gain-of-function mutation of Frizzled-4, increases the level of the transcription factor Sox 17. Through in vitro assays, these authors observed that Sox 17 is required for Norrin/Frizzled-4/Lrp–mediated angiogenesis in a 3D matrix gel. Sox17 may bind β-catenin and influence its transcriptional activity,^71,72 but how and at which level these 2 transcription factors act on angiogenesis in vivo is still unknown.

TCF4/β-catenin complex increases von Hippel–Lindau tumor suppressor,⁷³ which is required for inactivation of the transcription factor hypoxia-inducible factor (HIF)1α, which, in turn, is a potent inducer of vascular endothelial growth factor (VEGF). In addition, HIF1 α competes with TCF4 for direct binding to β-catenin.⁷⁴ β-Catenin/HIF1α interaction occurs at the promoter regions of HIF1α target genes and increases their expression. Taken together, these data suggest that at low levels of oxygen, the interaction of β-catenin with HIF1-α may prevail and sustain the response of this transcriptional factor. When normoxia is restored, HIF1α is decreased and TCF4 effectively competes for binding to β-catenin, increases von Hippel–Lindau tumor suppressor, and further promotes HIF1α degradation.

Other partners of canonic Wnt signaling are the members of the FOXO (FOXO1, -2, and -3) subfamily of transcription factors.⁷⁵ These factors are critically involved in the regulation of apoptosis, proliferation, and oxidative stress.⁷⁶ Under oxidative stress, FOXOs are increased, and β-catenin binding to them enhances their transcriptional activity. In nonendothelial cells, FOXO competes with TCF4 for interaction with β-catenin, thereby inhibiting TCF 4 transcriptional activity.⁷⁵

In endothelial cells, Taddei et al³⁷ found that FOXO1 and β-catenin complex binds to claudin 5 promoter and reduces its synthesis. Claudin 5 is an endothelial cell–specific member of the claudin family, which is indispensable for a correct organization of tight junctions. This inhibitory activity of the complex can be released by VE-cadherin expression and clustering at adherens junctions. VE-cadherin can activate phosphoinositide 3-kinase and, in this way, inhibit FOXO1. In addition, VE-cadherin can bind β-catenin and, in confluent cells, recruit this protein at the membrane and limits its nuclear translocation. All of this leads to VE-cadherin–mediated increase in claudin 5, which, in turn, promotes tight junction organization and permeability control.

Another important β-catenin interactor is Smad 2. This transcription factor is associated with transforming growth factor (TGF)β receptor and, on receptor activation, is phosphorylated and translocates to the nucleus.^77,78 Smad 2–induced transcription is further dependent on coactivators that stabilize its association to DNA. TGFβ receptors can associate with α3β1 integrin and cadherins (VE-cadherin in endothelial cells^79,80 and E-cadherin in the epithelium⁷⁹. This association generates β-catenin and Smad2 complex, which triggers expression of genes involved in cell motility and epithelial–mesenchymal transition.

Prox1, a key transcription factor in lymphatic differentiation (see above), is also a target of β-catenin/TCF⁸¹ in colon cancer.

Taken together, these observations indicate that cell responses to canonic Wnt signaling is strongly influenced by the cell context. β-Catenin may interact with different transcription factors depending on the functional condition of the cells (such as hypoxia, TGFβ signaling, etc). The type of transcription complex formed dictates the downstream cellular responses.

The Complexity of Canonic Wnt Signaling in Endothelial Cells

As a further level of complexity, endothelial cells not only express different Frizzled receptors and Wnt ligands, but they also respond to non-Wnt types of ligands. The best studied to date is Norrin.⁶⁴

This cysteine knot protein, although structurally unrelated to Wnt factors, is able to bind Frizzled-4 and LRP5/6 complex and potently induce canonical signal (Figure 2). Partial or complete loss-of-function mutations of genes coding for Norrin, Frizzled-4, or Lrp5 induce retina hypovascularization.⁵⁴ Importantly, these mutations are responsible for the so-called Norrie disease, familial exudative vitreoretinopathy and osteoporosis-pseudoglioma that in human beings causes important defects in retina vascularization.^82,83

The mechanism of action of Norrin has been partially disclosed by the discovery of TSPAN12, a member of the tetraspanin protein family.⁸⁴ The members of this family generate specialized tetraspanin-rich microdomains that function as signaling platforms at the cell membrane.⁸⁵ It was found that Norrin multimers act together with TSPAN12 to promote Frizzled-4 multimerization and downstream signaling (Figure 2). These data are of particular interest because TSPAN12 is specifically enriched in endothelial cells in the retina but not in other type of vessels or cells. Furthermore, TSPAN12 does not influence Wnt signaling in general but only when is triggered by Norrin. This pathway, therefore, is an example of how β-catenin–dependent signaling may be regulated in a cell- and organ-specific manner.

Another example of a pathway that modulates Wnt signaling in the endothelium is histone deacetylase (HDAC)7.⁸⁶ It was recently reported that HDAC7 binds and prevents β-catenin nuclear translocation and signaling. VEGF treatment induces HDAC7 degradation and rescues β-catenin nuclear translocation and signaling. Thus, posttranslational modification of β-catenin, such as phosphorylation or acetylation, may strongly influence its signaling.

Other agents known to strongly increase vascular permeability, such as thrombin, were found to induce β-catenin redistribution from cadherins to cytoplasm and nucleus.^87,88 Lipid phosphate phosphatase 3 increases β-catenin signaling, which, in turn, induces fibronectin synthesis, endothelial cell migration, and vascular branching.⁸⁹ These observations support the concept that stimuli that do not belong to the Wnt family may also act through β-catenin signaling, modulating its nuclear availability.

Finally, Wnt signaling may influence endothelial cell responses indirectly. A typical example is given by Aicher et al.⁹⁰ These authors found that vasculogenic progenitor cell mobilization from the bone marrow is strongly increased by administration of Dkk1, which is an inhibitor of Lrp5/6 (Figure 2). This effect is mediated by osteoclasts, which are induced by Dkk1 to produce cathepsin K, which, in turn, cleaves the adhesive bonds between stem cells and stromal cells in the bone marrow and promotes vasculogenic progenitor mobilization. In this case, therefore, the effect of Wnt inhibition is indirect, involving other cell targets that induce endothelial progenitor mobilization and angiogenic activity.

Wnt and Vascular Pathology

Taking into consideration the different pathways and conditions that may influence or be influenced by Wnt, an important question is whether vascular pathologies may derive from altered Wnt signaling. Given the many circumstances in which Wnt signaling plays a critical function, it is conceivable that this occurs, and few examples have been reported in the literature.

Canonical Wnt signaling can promote retina vascularization. This effect is particularly important in diabetic retinopathy, where neovascularization is accompanied by increased β-catenin nuclear signaling. Interestingly, SERPINA3K, a serine proteinase inhibitor, was found to inhibit Wnt signaling by binding to Lrp6, thus inhibiting its dimerization with Frizzled receptors (Figure 2).⁹¹ SERPINA3K was also able to reduce diabetic retinopathy.⁹¹

Norrie disease indicates how abrogation of β-catenin signaling may cause profound effects in retinal vascular development (see above). It is tempting to assume that alteration of the Frizzled-4/Norrin system may also play a role in diabetic retinopathy or retinopathy of prematurity⁹² because it was shown in preclinical models.⁹³ Furthermore, Frizzled-4 expression is quite ubiquitous in endothelial cells of different types of vessels.⁵⁴ Venous insufficiency^94,95 or female infertility related to vascular problems^96,97 have been described in Norrie disease, suggesting a more general role of altered β-catenin signaling also in vascular regions outside brain or retina.

Secreted Frizzled-related protein is an inhibitor of Wnt signaling that acts by binding Wnt and antagonizes both canonical and noncanonical Wnt signaling. This protein inhibits angiogenesis in different settings including tumors, suggesting that Wnt may be implicated in proliferative diseases by inducing vascular proliferation.⁹⁸ Consistently, nuclear localization of β-catenin was observed in coronary endothelial cells in the ischemic heart.⁹⁹ In these conditions, β-catenin may constitute a defense mechanism to promote angiogenesis in case of hypoxia. Wnt4, produced by mesenchymal stem cells,¹⁰⁰ promotes vascular proliferation in ischemia.

Healing of diabetic ulcers appears to be improved by localization of vasculogenic progenitors able to release Wnt factors and activate endothelial cells.¹⁰¹ Canonical Wnt signaling can induce VEGF synthesis in colonic neoplasia, which, in turn, would increase angiogenesis.¹⁰² In other conditions, however, the Wnt inhibitor Dkk3 promotes angiogenesis in tumors,¹⁰³ underlying the difficulties in defining a unique pattern of responses in the different experimental conditions.

Idiopathic pulmonary arterial hypertension is a rare disease linked to mutated bone morphogenetic protein (BMP) receptor II, which causes pulmonary artery endothelial cell apoptosis and loss of small vessels. It was found that binding of BMP to this receptor activates canonical and noncanonical Wnt pathways, which, in turn, promote angiogenesis and inhibit vascular regression.¹⁰⁴

Several reports indicate that Wnt signaling plays a crucial role in brain vascularization and BBB differentiation.^{17,61,62,65,66} Astrocytes produce Wnt, and this factor regulates the architecture and function of the so-called neurovascular unit (see above). It is, therefore, conceivable that any condition that inhibits Wnt signaling during development or after birth may strongly affect brain perfusion and BBB differentiation.¹⁰⁵ Vice versa, gain-of-function mutation of Wnt may accelerate establishment of BBB and possibly alter permeability properties in the brain and in other vascular regions.^{61,62,65,66,105}

Sources of Funding

This work was supported by the Fondation Leducq Transatlantic Network of Excellence, Associazione Italiana per la Ricerca sul Cancro, Association for International Cancer Research UK, the European Community (Project Contracts: JUSTBRAIN; EUSTROKE contract 202213, OPTISTEM contract 223098, ANGIOSCAFF NMP3-LA-2008-214402 and ENDOSTEMCELLS Networks), Istituto Superiore di Sanita,' Italian Ministry of Health, and CARIPLO Foundation contract 2008.2463.

Disclosures

None.