?3-Integrin Regulates Vascular Endothelial Growth Factor-A–Dependent Permeability
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Cell Adhesion and Disease/Tumour Biology Laboratory (S.D.R., L.E.R., K.M.H.-D.), Cancer Research UK Clinical Centre, Queen Mary’s School of Medicine & Dentistry at Barts & The London, John Vane Science Centre, Charterhouse Square, London; Novartis Institute for Biomedical Research (L.W.), Angiogenesis Programme, Basel, Switzerland; and ImClone Systems Inc (D.J.H.), New York, NY.
ABSTRACT
Objective— ?3-integrin deficiency has been implicated in increasing levels of Flk-1 expression on endothelial cells and enhancing vascular endothelial growth factor (VEGF)-induced angiogenesis. We determined the role of ?3-integrin in mediating VEGF-A–induced blood vessel permeability through Flk-1.
Methods and Results— Using the Miles assay, we demonstrated that VEGF-A–induced plasma leakage was enhanced in ?3-null mice when compared with wild-type controls. This was not caused by any changes in blood vessel structure (as detected by light or electron microscopy) or by changes in endothelial cell–cell adhesion proteins (as determined by Western blot analysis, flow cytometry, and immunofluorescence). Circulating levels of VEGF, baseline blood vessel leakage, and leakage in response to an acute inflammatory stimulus were identical in wild-type and ?3-null mice. However, VEGF-A–induced leakage was abolished in ?3-null mice by the inhibition of Flk-1, indicating that the elevated levels of Flk-1 on ?3-null endothelial cells enhance VEGF-A–induced permeability.
Conclusions— ?3-integrin–deficiency increases the sensitivity of endothelial cells to VEGF-A by elevating Flk-1 expression and, as a consequence, enhances VEGF-A–mediated permeability.
Blood vessels in ?3-integrin–deficient mice are histologically normal and functionally intact. However, they express elevated levels of Flk-1 and are more sensitive than wild-type blood vessels to VEGF-A–induced permeability, but not to permeability induced by acute inflammatory agents.
Key Words: ?3-integrin ? VEGF ? Flk-1 ? permeability ? endothelium
Introduction
Endothelial cell interactions with the surrounding extracellular matrix (ECM) are mediated primarily by the integrin family of adhesion receptors. These are heterodimeric transmembrane glycoproteins, consisting of and ? subunits, possessing both adhesive and signaling properties. Ligand and signaling specificity is conferred by their heterodimeric composition.1 Endothelial cells have been shown to express a variety of integrins including v?3, a receptor for fibronectin, vitronectin, osteospontin, von Willebrand factor, laminin, and collagen.2,3
The ability of integrins to act as a bridge between the extracellular and intracellular environments allows them to mediate ECM-induced changes in vascular tone, permeability, and vessel remodeling, and, as such, they have been depicted as key receptors for detecting and responding to vascular injury after damage to the ECM.4 There is mounting evidence that interactions between integrins and the ECM are important in regulating many aspects of vascular biology. For example, we have shown that 5-integrin and fibronectin are essential for early blood vessel development.5 Several studies demonstrate that endothelial cells must be anchored to the matrix for survival, migration, and proliferation.6–8 Furthermore, changes in the matrix milieu that activate different integrins can regulate the levels of fibroblast growth factor and vascular endothelial growth factor (VEGF) receptors on endothelial cells.9
VEGF, also known as vascular permeability factor,10 is unique among growth factors in that it is an important regulator of angiogenesis and a potent vasodilator capable of increasing vascular permeability.11,12 The most widely studied form of VEGF belongs to the VEGF-A family and is a 38-kDa homodimeric peptide (VEGF165 in humans;VEGF164 in mice is henceforth referred to as VEGF-A) belonging to the larger family of VEGFs.13 The biological actions of VEGF are mediated through 2 tyrosine receptor kinases, VEGFR-1 (Flt-1) and VEGFR-2 (Flk-1), and a more recently identified receptor, neuropilin-1, which is believed to be a coreceptor for VEGF acting in concert with VEGFR-2.14
We have shown previously that VEGF-induced angiogenesis is enhanced in mice that are deficient in ?3-integrin and that this response is associated with increased levels of Flk-1.15 In this study we have addressed the role of ?3-integrin in regulating VEGF-A–mediated vascular permeability. By examining the leakage of Evans blue dye from the vascular system, we show that blood vessels in ?3-null mice are more sensitive to VEGF-A–induced permeability. By administering either cells or animals with the monoclonal antibody DC101 directed against Flk-1,16 we show, for the first time to our knowledge, that this enhanced response is mediated by the elevated Flk-1 on ?3-null endothelial cells. We see no morphological defects in ?3-null vessels when compared with wild-type controls, nor is baseline permeability altered, demonstrating that the absence of ?3-integrin does not affect normal blood vessel structure or integrity. These results indicate that ?3-integrin regulates levels of Flk-1 and thereby influences VEGF-A–mediated permeability.
Methods
Animals
?3-integrin–deficient17 and wild-type control mice were 8 to 12 weeks of age. Lectin visualization of blood vessels was performed on mice from a 129sv genetic background, whereas all other studies were performed on mice from a C57/BL6 background. All animals were used in accord with United Kingdom Home Office regulations.
Reagents
For a detailed list of reagent sources, please see http://atvb. ahajournals.org
Miles Assays
Please see http://atvb.ahajournals.org for a detailed protocol of these methods.
Lectin Staining
Please see http://atvb.ahajournals.org for a detailed protocol of these methods.
Morphometric Measurements of Vessels
Vessel area density, vessel number density, and vessel diameters were measured in lectin-stained whole mounts as described by Thurston et al.18
Transmission Electron Microscopy
Ultrastructural analysis of microvessels was from saline-injected skin (as described for the Miles assay). Fifteen minutes after injection, skin surrounding the sites of injection was dissected and processed for electron microscopy.
Western Blotting, Flow Cytometry, and Immunofluorescence
Western blots, flow cytometry, and immunofluorescence were performed as described by Reynolds et al.15
In Vitro Permeability Assay
Transwell permeability assays were performed on monolayers of mouse endothelial cells as described by Wójciak-Stothard.19 VEGF164 (30 ng/mL) in the presence or absence of DC101 (20 μg/mL) was added to both the upper and lower chambers of the transwell concomitant with fluorescein isothiocyanate (FITC)-dextran (1 mg/mL) addition to the upper chamber. Samples were collected and assayed 30 minutes after VEGF stimulation.
In Vivo DC101 Treatments
Pre-experimental treatments were two 0.8-mg intraperitoneal injections of DC101 (100 μL in phosphate-buffered saline ) or either an equivalent concentration of a rat IgG isotyope control (in PBS) or an equivalent volume of PBS only, on days 0 and 3. VEGF164 intradermal injections and Miles assays were performed on day 6, as described.
Statistical Analysis
Data are presented as means±SEM. Significant differences between means were evaluated by unpaired Student t test. P<0.05 was considered significant.
Results
VEGF-A–Induced Vascular Permeability Is Enhanced in ?3-Integrin–Deficient Mice
To analyze the role of VEGF-A–mediated permeability, we used the Miles assay20 to measure vascular leakage in the skin of ?3-integrin–deficient and wild-type control mice. Mice were injected intravenously with Evans blue dye followed by intradermal injections of either 400 ng VEGF-A or vehicle (saline) alone. Spectrophotometric analysis of specimens from 8-mm biopsy punches performed around the sites of injection (using the point of injection as the center of the punch) showed that within 15 minutes after VEGF-A treatment, blood vessels in ?3-null mice were 2.5-fold more permeable than vessels in wild-type mice (Figure 1A). At this suboptimal concentration of VEGF-A, vessels in wild-type animals exhibited minimal dye leakage when compared with vehicle-only controls. Moreover, a higher (600 ng) dose of VEGF-A was required to induce a significant degree of plasma leakage in wild-type mice, indicating that ?3-null mice are more responsive to VEGF-A.
Figure 1. VEGF-A–induced vascular permeability is enhanced in ?3-null mice. A, Permeability of skin vessels after intradermal injections of VEGF-A. Values are presented as the ratio of VEGF-induced Evans blue leakage/vehicle alone-induced Evans blue leakage (±SEM), n=6 to 9 animals per genotype. The photographs illustrate enhanced leakage of Evans blue dye in ?3-null skin in response to 400 ng of VEGF-A; nsd=no significant difference. Scale bar=5 mm. B, Circulating levels of VEGF. The bars represent serum concentrations of VEGF (±SEM), n=6 mice per genotype. C, Baseline vascular permeability in the skin of ears. Bars represent ng of Evans blue leakage per mg weight of ear (±SEM), n=8 to 9 animals per genotype. The photographs show minimal Evans blue dye leakage in whole ears in the absence of an inflammatory stimulus. Scale bar=10 mm. D, Mustard oil-induced permeability in the skin of ears. Bars represent ng of Evans blue leakage per mg weight of ear (±SEM), n=8 to 9 animals per genotype. The inserted graph represents fold stimulation of treated/vehicle alone in wild-type and ?3-null mice. The photographs illustrate extensive Evans blue dye leakage in inflamed ears of wild-type and ?3-null mice. Scale bar=10 mm.
The extravasated Evans blue measured in this type of experiment is caused by both vessel permeability and vessel dilation. Although we cannot distinguish between the two, vessel dilation usually accounts for <0.01% of leakage in the type of assay described.21,22
Baseline Permeability and Inflammation-Induced Permeability Are not Affected by ?3-Integrin Deficiency
Because VEGF-A-mediated vessel permeability was enhanced in ?3-null mice, we then asked if ?3-integrin–deficiency affected baseline vessel permeability. To address this, we first examined circulating VEGF levels in wild-type and ?3-null mice. Enzyme-linked immunosorbent assay results measuring serum levels of VEGF are presented in Figure 1B. No significant differences in VEGF levels were observed between wild-type and ?3-null mice, indicating that in a young, healthy population of mice, circulating levels of VEGF are not affected by the absence of ?3-integrin. In addition, no evidence of edema was observed in the ?3-null mice (data not shown).
We next compared vascular leakage in ear skin under baseline conditions. Mice were injected intravenously with Evans blue dye followed by topical application of mineral oil (a noninflammatory agent used as the vehicle in subsequent inflammatory assays) to the epidermis of the ear. Spectrophotometric analysis of dissected ears showed that within the 30-minute time frame of the experiment, no significant difference was observed in dye leakage between wild-type and ?3-null mice (Figure 1C).
We next investigated vascular leakage in ear skin in response to mustard oil, an inflammatory agent that induces acute plasma leakage in the skin.23 Mice were injected intravenously with Evans blue dye, followed by topical application of mustard oil to the epidermis of the ear, as described (for the baseline assays). Spectrophotometric analysis of dissected ears showed that 30 minutes after Evans blue administration, mustard oil treatment induced 2.5-fold increase in dye leakage (when compared with baseline levels) in vessels from both wild-type and ?3-null mice; no difference was observed when comparing levels of leaked dye between the 2 genotypes (Figure 1D). Although the data presented are from samples collected 30 minutes after initial mustard oil treatment, we have not seen any differences in vascular leakage between ?3-null and wild-type vessels at either earlier (15 minutes) or later (60 minutes) time points (data not shown). Taken together, these data indicate that baseline permeability and inflammation induced permeability of vessels in ?3-null mice is equivalent to that of vessels in wild-type mice.
Blood Vessels in the Skin of ?3-Null Mice Are Morphologically Normal
The vessels in the skin of ?3-null mice appeared normal. The skin was not reddened and we saw no overt signs of edema (data not shown). We have shown previously that there are no differences between wild-type and ?3-null mice in the number of platelet and endothelial cell adhesion molecule (PECAM)-positive vessels in normal back skin.15 The architecture of the microvasculature was examined in whole-mount preparations of ear skin from wild-type and ?3-null mice in which vessels were visualized with biotin-labeled BS-1 lectin, which binds to the luminal surface of endothelial cells.24 The vessels throughout the skin of the ear appeared normal (Figure 2A). A quantitative analysis revealed no differences in vessel area density (reflecting the overall number, length, and size of vessels per unit area), vessel number density (reflecting the total number of vessels traversing the ear per unit area, independent of vessel size), or in the average diameter of microvessels (Figure 2B).
Figure 2. Blood vessel morphology is normal in ?3-null mice. A, BS-1 lectin visualization of microvessel morphology near the margin of the ear (art, arteriole; cap, capillary; ven, venule). Scale bar=100 μm. B, Morphometric quantitation of blood vessels in BS-1 lectin-stained ear whole mounts. Values represent mean (±SEM) of 3 to 6 microscope fields per ear and 5 ears per genotype. C, Ricin lectin visualization of exposed basement membrane surrounding mustard oil-treated blood vessels near the margin of the ear. Arrowheads indicate sites of leakage. Scale bar=30 μm.
The amount of plasma leakage in inflammation depends, at least in part, on the number and size of sites for leakage in the endothelium.25 These leakage sites expose the underlying endothelial basement membrane and can therefore be visualized in whole-mount preparations with ricin lectin, which binds more avidly to components of the basement membrane than the endothelium.26 In untreated vessels, rhodamine-labeled ricin lectin binding to the endothelial luminal surface was weak and uniform in both genotypes (data not shown). In mustard oil-treated ears, ricin lectin bound very strongly to focal sites of exposed basement membrane in vessels presumed to be venules based on vessel diameter and various models of acute inflammation.27 The degree of ricin lectin binding was comparable in both genotypes. We did not see any differences in the number (data not shown) or location of exposed basement membrane sites in wild-type versus ?3-null vessels (Figure 2C).
Analysis by transmission electron microscopy confirmed an intact microvascular network in the skin of untreated ?3-null mice. ?3-null vessels, including capillaries, were unaltered in size and showed normal pericyte recruitment and cell–cell contacts between endothelial cells and supporting cells (Figure 3A and data not shown).
Figure 3. Cell–cell and cell–ECM associations appear normal in ?3-null blood vessels. A, Upper panel shows transmission electron microscopy of vessels in the skin (e, endothelial cell nucleus; s, supporting cell body). Scale bar=2 μm. Lower panel shows higher magnifications of similar sections to those in upper panel. Arrowheads indicate electron-dense endothelial cell–cell junctions. Scale bar=200 nm. B, Western blot analysis of occludin levels in lung endothelial cells. Blotting for HSC 70 provides a loading control. The bar chart represents averaged densitometry results (means±SEM) from multiple independent experiments that show wild-type and ?3-null endothelial cells express equivalent levels of occludin. C, The panels show representative flow cytometric analyses of PECAM and VECAD levels on lung endothelial cells. D, The micrographs show immunofluorescence for PECAM (upper panels) and VECAD (lower panels) in both blood vessels from skin and in cultured lung endothelial cells. Positive, bright staining is concentrated at the borders between endothelial cells in both genotypes. Scale bar=5 μm in skin sections and 3 μm in cultured cells.
Endothelial cell-to-cell junctions are complex structures formed by different adhesive molecules.28 Endothelial cells possess adherens junctions and tight junctions similar to those described in epithelial cells. Adherens junctions are ubiquitous along the vascular tree and are formed by transmembrane proteins belonging to the cadherin superfamily. Endothelial cells express a cell-specific cadherin known as vascular endothelial cell cadherin (VECAD). Tight junctions comprise 3 types of transmembrane proteins, occludins, claudins and junctional adhesion molecule. In addition, other adhesive proteins, such as PECAM, and endoglin are concentrated at intercellular contacts in the endothelium. As a measure of the molecular architecture of junctions in wild-type and ?3-null endothelial cells, we performed Western blot analysis for occludin (Figure 3B) on lung endothelial cells derived from wild-type and ?3-null animals; flow cytometry for PECAM and VECAD (Figure 3C) on lung endothelial cells derived from wild-type and ?3-null animals; and immunofluorescence (Figure 3D) on both frozen skin sections and isolated lung endothelial cells from both genotypes. Wild-type and ?3-null endothelial cells showed equivalent expression of all 3 cell–cell adhesion molecules, suggesting that adherens junctions, tight junctions, and other cell adhesions are normal in ?3-null cells.
Inhibition of Flk-1 Function Abrogates VEGF-A–Induced Permeability in ?3-Null Mice
Given that no differences in either vessel structure or junctional protein levels were observed when comparing wild-type and ?3-null blood vessels, but that VEGF-A–mediated permeability and Flk-1 levels (Figure 4A) were significantly elevated in ?3-null mice, we asked if elevated Flk-1 levels were directly responsible for the enhanced VEGF-A–mediated vessel permeability response. To address this in vitro, we cultured monolayers of wild-type and ?3-null endothelial cells in transwell chambers and measured the flux of FITC-dextran across the monolayers in response to treatment with either VEGF-A alone or VEGF-A in combination with the neutralizing anti–Flk-1 antibody DC10116 (Figure 4B). To address this in vivo, we administered DC101 to wild-type and ?3-null mice and measured VEGF-A–induced vessel leakage using the Miles assay (Figure 4C). Both analyses revealed that by inhibiting the function of Flk-1, VEGF-A–induced permeability was completely abolished in ?3-null endothelium. These results indicate that increased levels of Flk-1 in ?3-null mice are directly responsible for an increased sensitivity to VEGF-A–mediated blood vessel permeability.
Figure 4. Elevated VEGF-A–induced blood vessel permeability in ?3-null mice is via Flk-1. A, Western blot analysis of Flk-1 levels in lung endothelial cells. Blotting for HSC 70 provides a loading control. The bar chart represents averaged densitometry results (means±SEM) from multiple independent experiments that show ?3-null endothelial cells express 2- to 3-fold more Flk-1 than wild-type controls. B, Effect of ?3-integrin deficiency in endothelial barrier function in vitro. VEGF-mediated permeability is inhibited by DC101 in ?3-null endothelial monolayers when compared with wild-type controls. Bars show the percentage of FITC-dextran flux across the monolayer in comparison to untreated controls (±SEM), n=4 independent wells per genotype/per treatment. C, Permeability of skin vessels in vivo after intradermal injections of suboptimal doses of VEGF-A in mice pretreated with DC101 or control (rat isotype matched IgG or PBS, collectively shown as control treatment). DC101 inhibits VEGF-A–induced vascular leakage in ?3-null mice. Bars show the ratio of VEGF-induced Evans blue leakage/vehicle alone-induced Evans blue leakage (±SEM), n=6 to 14 animals per genotype/per treatment.
Discussion
The establishment and maintenance of a vascular network is critical for normal embryological development and has significant implications for pathological processes such as wound healing and tumorgenesis. Numerous investigations have demonstrated that interactions between integrins and the ECM are important to these processes.2,6,29–32
A major focus of these studies has been on the roles played by v?3-integrin and v?5-integrin in angiogenesis.33–37 During angiogenesis, v?3 expression is upregulated.38 Furthermore, its expression is linked to cell migration, invasion, and cell survival.38–40 Given these collective data, it is surprising that genetic ablation of v-integrin elicits only minor vessel defects.41 Although most v-deficient mice die soon after birth, a complex embryonic vascular network develops. In addition, ?3-deficient mice are viable and fertile and appear to produce a vascular network without obvious defects.17,42 A detailed morphological examination of skin blood vessels in ?3-null mice, presented here, indicates that the vasculature is intact and indistinguishable from that of wild-type mice, reinforcing the conclusion that ?3-integrin is not necessary for establishing a normal vasculature.
Endothelial cells attach to the ECM to form a semi-permeable barrier restricting movement of solutes across the vascular wall. Evidence that integrins are involved in maintaining this barrier comes from studies showing that treating endothelial monolayers or isolated coronary venules with RGD peptides (small molecules that disrupt interactions between integrins and the ECM) increase their permeability.43–45 The fact that the increased permeability in these in vitro studies is prevented by the addition of soluble fibronectin or vitronectin suggests that interactions between integrins and the ECM influence vessel permeability. From the data we present here, however, we are able to conclude that ?3-integrin is not essential for maintaining an intact and functional endothelial barrier. These data are as follows: (1) baseline vessel permeability is not altered in ?3-deficient mice; (2) a molecular comparison of the endothelial cell-to-cell contacts between the 2 genotypes revealed no differences in the expression of PECAM, VECAD, or occludin; and (3) wild-type and ?3-null blood vessels respond equally to mustard oil-induced plasma leakage. Collectively, these data indicate that ?3-integrin is not necessary for maintaining baseline integrity of the endothelial barrier.
In marked contrast to the lack of an apparent phenotype in vessel structure and baseline integrity, we see an increase in the sensitivity of ?3-null vessels to subthreshold doses of VEGF-A treatment. Changes in matrix composition and integrin activation can influence levels of, and signaling through, VEGF and fibroblast growth factor receptors on cultured microvascular endothelial cells.9,46,47 We show that the absence of ?3-integrin is associated with an increase in the levels of Flk-1 expressed on endothelial cells and, for the first time to our knowledge, that this increase in Flk-1 leads to an increased sensitivity to VEGF-A–mediated permeability in vitro and in vivo. Previously, we were able to draw correlations between only elevated Flk-1 levels and enhanced VEGF-mediated angiogenesis.15 With respect to permeability, however, we show that by inhibiting signaling through Flk-1, we abolish the enhanced VEGF-A response seen in ?3-null vessels. This demonstrates that the increased Flk-1 is functional and is entirely responsible for the elevated sensitivity to VEGF-A–induced permeability.
Conventional treatment of vascular leakage is rather aspecific and includes treatments such as the administration of glucocorticoids or nonsteroidal inflammatory drugs and antihistaminergic compounds. Recently, VEGF antagonists48 and VEGF receptor inhibitors49 have been used with some success in treating the increased vessel leakiness associated with macular degeneration, diabetic retinopathy, and ischemia–reperfusion injury. By genetically ablating ?3-integrin function, we have altered blood vessel sensitivity to VEGF-A–induced plasma leakage, indicating that integrins might be additional useful targets for clinically manipulating vessel permeability.
Antagonists of v?3 and v?5, some of which are in clinical trials,50 have been shown to block angiogenesis, suggesting that these integrins are required for angiogenesis. However, we have shown that ?3-integrin–deficient mice support enhanced pathological angiogenesis.15 The data we present here and in our previous studies suggest that ?3-integrin acts as a transdominant inhibitor of VEGF-mediated vascular functions, with its presence being enough to downregulate levels of Flk-1. We have previously demonstrated that the reintroduction of ?3-integrin into null endothelial cells restores Flk-1 levels to those seen in wild-type cells,15 suggesting that ?3-integrin can regulate Flk-1 expression. Hence, eliminating ?3-integrin interactions with the ECM might alter the response of endothelial cells to VEGF. The antagonist of v?3-integrin, LM609, has been shown to enhance vascular permeability.51 If inhibition of v?3-integrin has the potential of elevating Flk-1 levels in endothelial cells, both angiogenesis and permeability of tumor vessels could be increased, especially if, as many are,52 the tumor is a source of VEGF production. This might prove detrimental to the treatment of the tumor. We do not question the efficacy of v?3 antagonists as potentially useful therapeutic agents, but our results suggest we need a more thorough understanding of the interactions between integrins and other molecules that are involved in regulating angiogenesis and permeability.
Many vascular disorders are accompanied by alterations in production and/or degradation of ECM components and by alterations in integrin expression.4 Because of this, integrins are obvious therapeutic targets for treating a host of vascular diseases such as hypertension, diabetes mellitus, and restenosis.53–56 This idea is exciting and should be explored, but with the thought in mind that manipulating integrin expression might lead to an increase in growth factor receptor levels expressed by endothelial cells. As we show here, genetically ablating ?3-integrin function elevates Flk-1 levels on endothelial cells. This has the potential of affecting vessel growth and vessel permeability, both of which have important implications for current and future therapeutics. Vessel leakiness, for example, is known to contribute to the abnormal microenvironment of tumors and can affect tumor growth, metastasis, and therapeutic drug delivery.57 As long as we exercise caution, though, integrins may provide us with useful targets for treating a range of vascular disorders.
Acknowledgments
We thank Garry Saunders, Sue Watling, Colin Wren, and Stephen Gschmeissner for their technical assistance, and Ian Hart, Francesco Conti, Andrew Reynolds, and the other members of the Department of Tumor Biology for their support and criticism during this study. We also thank Mark Karaczun for his critical reading of the manuscript.
References
Hynes R. Integrins: versatility, modulation, and signalling in cell adhesion. Cell. 1992; 69: 11–25.
Luscinskas F, Lawler J. Integrins as dynamic regulators of vascular function. FASEB J. 1994; 8: 929–938.
Hynes R, Bader B, Hodivala-Dilke K. Integrins in vascular development. Braz J Med Biol Res. 1999; 32: 501–510.
Martinez-Lemus LA, Wu X, Wilson E, Hill MA, Davis GE, Davis MJ, Meininger GA. Integrins as unique receptors for vascular control. J Vasc Res. 2003; 40: 211–233.
Francis SE, Goh KL, Hodivala-Dilke K, Bader BL, Stark M, Davidson D, Hynes RO. Central roles of 5?1 integrin and fibronectin in vascular development in mouse embryos and embryoid bodies. Arterioscler Thromb Vasc Biol. 2002; 22: 927–933.
Madri J, Williams, S. Capillary endothelial cell cultures: phenotypic modulation by matrix components. J Cell Biol. 1983; 97: 153–165.
Form D, Pratt B, Madri J. Endothelial cell proliferation during angiogenesis. In vitro modulation by basement membrane components. Lab Invest. 1986; 55: 521–530.
Underwood P, Bennett F. The effect of extracellular matrix molecules on the in vitro behavior of bovine endothelial cells. Exp Cell Res. 1993; 205: 311–319.
Tsou R, Isik FF. Integrin activation is required for VEGF and FGF receptor protein presence on human microvascular endothelial cells. Mol Cell Biochem. 2001; 224: 81–89.
Senger D, Galli S, Dvorak A, Perruzzi C, Harvey V, Dvorak H. Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science. 1983; 219: 983–985.
Ferrara N. Role of vascular endothelial growth factor in regulation of physiological angiogenesis. Am J Physiol Cell Physiol. 2001; 280: C1358–C1366.
Bates DO, Hillman NJ, Williams B, Neal CR, Pocock TM. Regulation of microvascular permeability by vascular endothelial growth factors. J Anat. 2002; 200: 581–597.
Ruhrberg C. Growing and shaping the vascular tree: multiple roles for VEGF. Bioessays. 2003; 25: 1052–1060.
Cross MJ, Dixelius J, Matsumoto T, Claesson-Welsh L. VEGF-receptor signal transduction. Trends Biochem Sci. 2003; 28: 488–494.
Reynolds LE, Wyder L, Lively JC, Taverna D, Robinson SD, Huang X, Sheppard D, Hynes RO, Hodivala-Dilke KM. Enhanced pathological angiogenesis in mice lacking ?3 integrin or ?3 and ?5 integrins. Nat Med. 2002; 8: 27–34.
Prewett M, Huber J, Li Y, Santiago A, O’Connor W, King K, Overholser J, Hooper A, Pytowski B, Witte L, Bohlen P, Hicklin D. Anti-VEGF receptor (Flk-1) monoclonal antibody inhibits tumor angiogenesis and growth of several mouse and human tumors. Cancer Res. 1999; 59: 5209–5218.
Hodivala-Dilke K, McHugh K, Tsakiris D, Rayburn H, Crowley D, Ullman-Cullere M, Ross F, Coller B, Teitelbaum S, Hynes R. ?3-integrin deficient mice are a model for Glanzmann thrombasthenia showing placental defects and reduced survival. J Clin Invest. 1999; 103: 229–238.
Thurston G, Murphy TJ, Baluk P, Lindsey JR, McDonald DM. Angiogenesis in mice with chronic airway inflammation: strain-dependent differences. Am J Pathol. 1998; 153: 1099–1112.
Wójciak-Stothard B, Potempa S, Eichholtz T, Ridley AJ. Rho and Rac, but not Cdc42 regulate endothelial cell permeability. J Cell Sci. 2001; 114: 1343–1355.
Miles A, Miles E. Vascular reactions to histamine, histamine-liberator and leukotaxine in the skin of guinea-pigs. J Physiol. 1952; 118: 228–257.
Buckley T, Brain S, Williams T. Ruthenium red selectively inhibits oedema formation and increased blood flow induced by capsaicin in rabbit skin. Br J Pharmacol. 1990; 99: 7–8.
Hughes S, Brain S, Williams G, Williams T. Assessment of blood flow changes at multiple sites in rabbit skin using a 133Xenon clearance technique. J Pharmacol Toxicol Methods. 1994; 32: 41–47.
Baluk P, Thurston G, Murph TJ, Bunnett NW, McDonald DM. Neurogenic plasma leakage in mouse airways. Br J Pharmacol. 1999; 126: 522–528.
Plendl J, Neumuller C, Vollmar A, Auerbach R, Sinowatz F. Isolation and characterization of endothelial cells from different organs of fetal pigs. Anat Embryol (Berl). 1996; 194: 445–456.
Zweifach BW. Microcirculation. Annu Rev Physiol. 1973; 35: 117–150.
Bergman RK, Munoz JJ. Effect of Bordetella pertussis extract and vasoactive amines on vascular permeability. J Allergy Clin Immunol. 1975; 55: 378–385.
Svensjo E, Grega G. Evidence for endothelial cell-mediated regulation of macromolecular permeability by postcapillary venules. Fed Proc. 1986; 45: 89–95.
Dejana E. Endothelial cell-cell junctions: happy together. Nat Rev Mol Cell Biol. 2004; 5: 261–270.
Ingber D, Folkman J. Mechanochemical switching between growth and differentiation during fibroblast growth factor-stimulated angiogenesis in vitro: role of extracellular matrix. J Cell Biol. 1989; 109: 317–330.
Ingber D. Extracellular matrix and cell shape: potential control points for inhibition of angiogenesis. J Cell Biochem. 1991; 47: 236–241.
Klasgsbrun M. Regulators of angiogenesis: stimulators, inhibitors, and extracellular matrix. J Cell Biochem. 1991; 47: 199–200.
Kanda S, Tomasini-Johansson B, Klint P, Dixelius J, Rubin K, Claesson-Welsh L. Signaling via fibroblast growth factor receptor-1 is dependent on extracellular matrix in capillary endothelial cell differentiation. Exp Cell Res. 1999; 248: 203–213.
Byzova T, Rabbani R, Souza S, Plow E. Role of integrin v?3 in vascular biology. Thromb Haemost. 1998; 80: 726–734.
Friedlander M, Brooks P, Shaffer R, Kincaid C, Varner J, Cheresh D. Definition of two angiogenic pathways by distinct v integrins. Science. 1995; 270: 1500–1502.
Stromblad S, Cheresh, D. Integrins, angiogenesis and vascular cell survival. Chem Biol. 1996; 3: 881–885.
Eliceiri B, Cheresh D. The role of v integrins during angiogenesis: insights into potential mechanisms of action and clinical development. J Clin Invest. 1999; 103: 1227–1230.
Hodivala-Dilke KM, Reynolds AR, Reynolds LE. Integrins in angiogenesis: multitalented molecules in a balancing act. Cell Tissue Res. 2003; 314: 131–144.
Brooks P, Montgomery A, Rosenfeld M, Reisfeld R, Hu T, Klier G, Cheresh, D. Integrin v ? 3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell. 1994; 79: 1157–1164.
Felding-Habermann B, Cheresh D. Vitronectin and its receptors. Curr Opin Cell Biol. 1993; 5: 864–868.
Brassard D, Maxwell E, Malkowski M, Nagabhushan T, Kumar C, Armstrong L. Integrin (v)?(3)-mediated activation of apoptosis. Exp Cell Res. 1999; 251: 33–45.
Bader B, Rayburn H, Crowley D, Hynes R. Extensive vasculogenesis, angiogenesis and organogenesis precede lethality in mice lacking all v integrins. Cell. 1998; 95: 507–519.
Huang X, Griffiths M, Wu J, Farese R, Sheppard D. Normal development, wound healing, and adenovirus susceptibility in ?5-deficient mice. Mol Cell Biol. 2000; 20: 755–759.
Curtis T, McKeown-Longo P, Vincent P, Homan S, Wheatley E, Saba T. Fibronectin attenuates increased endothelial monolayer permeability after RGD peptide, anti- 5 ? 1, or TNF- exposure. Am J Physiol. 1995; 269: L248–L260.
Qiao R, Yan W, Lum H, Malik A. Arg-Gly-Asp peptide increases endothelial hydraulic conductivity: comparison with thrombin response. Am J Physiol. 1995; 269: C110–C117.
Wu M, Ustinova E, Granger H. Integrin binding to fibronectin and vitronectin maintains the barrier function of isolated porcine coronary venules. J Physiol. 2001; 532: 785–791.
Soldi R, Mitola S, Strasly M, Defilippi P, Tarone G, Bussolino F. Role of v?3 integrin in the activation of vascular endothelial growth factor receptor-2. EMBO J. 1999; 18: 882–892.
Byzova TV, Goldman CK, Pampori N, Thomas KA, Bett A, Shattil SJ, Plow EF. A mechanism for modulation of cellular responses to VEGF: activation of the integrins. Mol Cell. 2000; 6: 851–860.
van Bruggen N, Thibodeaux H, Palmer J, Lee W, Fu L, Cairns B, Tumas D, Gerlai R, Williams S, van Lookeren Campagne M, Ferrara N. VEGF antagonism reduces edema formation and tissue damage after ischemia/reperfusion injury in the mouse brain. J Clin Invest. 1999; 104: 1613–1620.
The T.E.S. Group. Preclinical and phase 1A clinical evaluation of an anti-VEGF peglyated aptamer (EYE001) for the treatment of exudative age-related macular degeneration. Retina. 2002; 22: 143–152.
Gutheil J, Campbell T, Prierce P, Watkins J, Huse W, Bodkin D, Cheresh D. Targeted antiangiogenic therapy for cancer using Vitaxin: a humanized monoclonal antibody to the integrin v?3. Clin Cancer Res. 2000; 6: 3056–3061.
Tsukada H, Ying X, Fu C, Ishikawa S, McKeown-Longo P, Albelda S, Bhattacharya S, Anderson Bray B, Bhattacharya J. Ligation of endothelial v?3 integrin increases capillary hydraulic conductivity of rat lung. Circ Res. 1995; 77: 651–659.
Mukhopadhyay D, Datta K. Multiple regulatory pathways of vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) expression in tumors. Semin Cancer Biol. 2004; 14: 123–130.
Roth T, Podesta F, Stepp M, Boeri D, Lorenzi M. Integrin overexpression induced by high glucose and by human diabetes: potential pathway to cell dysfunction in diabetic microangiopathy. Proc Natl Acad Sci U S A. 1993; 90: 9640–9644.
Regoli M, Bendayan M. Alterations in the expression of the 3 ? 1 integrin in certain membrane domains of the glomerular epithelial cells (podocytes) in diabetes mellitus. Diabetologia. 1997; 40: 15–22.
Intengan H, Thibault G, Li J, Schiffrin E. Resistance artery mechanics, structure, and extracellular components in spontaneously hypertensive rats: effects of angiotensin receptor antagonism and converting enzyme inhibition. Circulation. 1999; 100: 2267–2275.
Bishop G, McPherson J, Sanders J, Hesselbacher S, Feldman M, McNamara C, Gimple L, Powers E, Mousa S, Sarembock I. Selective (v)?(3)-receptor blockade reduces macrophage infiltration and restenosis after balloon angioplasty in the atherosclerotic rabbit. Circulation. 2001; 103: 1906–1911.
McDonald DM, Baluk P. Significance of blood vessel leakiness in cancer. Cancer Res. 2002; 62: 5381–5385.(Stephen D. Robinson; Loui)
ABSTRACT
Objective— ?3-integrin deficiency has been implicated in increasing levels of Flk-1 expression on endothelial cells and enhancing vascular endothelial growth factor (VEGF)-induced angiogenesis. We determined the role of ?3-integrin in mediating VEGF-A–induced blood vessel permeability through Flk-1.
Methods and Results— Using the Miles assay, we demonstrated that VEGF-A–induced plasma leakage was enhanced in ?3-null mice when compared with wild-type controls. This was not caused by any changes in blood vessel structure (as detected by light or electron microscopy) or by changes in endothelial cell–cell adhesion proteins (as determined by Western blot analysis, flow cytometry, and immunofluorescence). Circulating levels of VEGF, baseline blood vessel leakage, and leakage in response to an acute inflammatory stimulus were identical in wild-type and ?3-null mice. However, VEGF-A–induced leakage was abolished in ?3-null mice by the inhibition of Flk-1, indicating that the elevated levels of Flk-1 on ?3-null endothelial cells enhance VEGF-A–induced permeability.
Conclusions— ?3-integrin–deficiency increases the sensitivity of endothelial cells to VEGF-A by elevating Flk-1 expression and, as a consequence, enhances VEGF-A–mediated permeability.
Blood vessels in ?3-integrin–deficient mice are histologically normal and functionally intact. However, they express elevated levels of Flk-1 and are more sensitive than wild-type blood vessels to VEGF-A–induced permeability, but not to permeability induced by acute inflammatory agents.
Key Words: ?3-integrin ? VEGF ? Flk-1 ? permeability ? endothelium
Introduction
Endothelial cell interactions with the surrounding extracellular matrix (ECM) are mediated primarily by the integrin family of adhesion receptors. These are heterodimeric transmembrane glycoproteins, consisting of and ? subunits, possessing both adhesive and signaling properties. Ligand and signaling specificity is conferred by their heterodimeric composition.1 Endothelial cells have been shown to express a variety of integrins including v?3, a receptor for fibronectin, vitronectin, osteospontin, von Willebrand factor, laminin, and collagen.2,3
The ability of integrins to act as a bridge between the extracellular and intracellular environments allows them to mediate ECM-induced changes in vascular tone, permeability, and vessel remodeling, and, as such, they have been depicted as key receptors for detecting and responding to vascular injury after damage to the ECM.4 There is mounting evidence that interactions between integrins and the ECM are important in regulating many aspects of vascular biology. For example, we have shown that 5-integrin and fibronectin are essential for early blood vessel development.5 Several studies demonstrate that endothelial cells must be anchored to the matrix for survival, migration, and proliferation.6–8 Furthermore, changes in the matrix milieu that activate different integrins can regulate the levels of fibroblast growth factor and vascular endothelial growth factor (VEGF) receptors on endothelial cells.9
VEGF, also known as vascular permeability factor,10 is unique among growth factors in that it is an important regulator of angiogenesis and a potent vasodilator capable of increasing vascular permeability.11,12 The most widely studied form of VEGF belongs to the VEGF-A family and is a 38-kDa homodimeric peptide (VEGF165 in humans;VEGF164 in mice is henceforth referred to as VEGF-A) belonging to the larger family of VEGFs.13 The biological actions of VEGF are mediated through 2 tyrosine receptor kinases, VEGFR-1 (Flt-1) and VEGFR-2 (Flk-1), and a more recently identified receptor, neuropilin-1, which is believed to be a coreceptor for VEGF acting in concert with VEGFR-2.14
We have shown previously that VEGF-induced angiogenesis is enhanced in mice that are deficient in ?3-integrin and that this response is associated with increased levels of Flk-1.15 In this study we have addressed the role of ?3-integrin in regulating VEGF-A–mediated vascular permeability. By examining the leakage of Evans blue dye from the vascular system, we show that blood vessels in ?3-null mice are more sensitive to VEGF-A–induced permeability. By administering either cells or animals with the monoclonal antibody DC101 directed against Flk-1,16 we show, for the first time to our knowledge, that this enhanced response is mediated by the elevated Flk-1 on ?3-null endothelial cells. We see no morphological defects in ?3-null vessels when compared with wild-type controls, nor is baseline permeability altered, demonstrating that the absence of ?3-integrin does not affect normal blood vessel structure or integrity. These results indicate that ?3-integrin regulates levels of Flk-1 and thereby influences VEGF-A–mediated permeability.
Methods
Animals
?3-integrin–deficient17 and wild-type control mice were 8 to 12 weeks of age. Lectin visualization of blood vessels was performed on mice from a 129sv genetic background, whereas all other studies were performed on mice from a C57/BL6 background. All animals were used in accord with United Kingdom Home Office regulations.
Reagents
For a detailed list of reagent sources, please see http://atvb. ahajournals.org
Miles Assays
Please see http://atvb.ahajournals.org for a detailed protocol of these methods.
Lectin Staining
Please see http://atvb.ahajournals.org for a detailed protocol of these methods.
Morphometric Measurements of Vessels
Vessel area density, vessel number density, and vessel diameters were measured in lectin-stained whole mounts as described by Thurston et al.18
Transmission Electron Microscopy
Ultrastructural analysis of microvessels was from saline-injected skin (as described for the Miles assay). Fifteen minutes after injection, skin surrounding the sites of injection was dissected and processed for electron microscopy.
Western Blotting, Flow Cytometry, and Immunofluorescence
Western blots, flow cytometry, and immunofluorescence were performed as described by Reynolds et al.15
In Vitro Permeability Assay
Transwell permeability assays were performed on monolayers of mouse endothelial cells as described by Wójciak-Stothard.19 VEGF164 (30 ng/mL) in the presence or absence of DC101 (20 μg/mL) was added to both the upper and lower chambers of the transwell concomitant with fluorescein isothiocyanate (FITC)-dextran (1 mg/mL) addition to the upper chamber. Samples were collected and assayed 30 minutes after VEGF stimulation.
In Vivo DC101 Treatments
Pre-experimental treatments were two 0.8-mg intraperitoneal injections of DC101 (100 μL in phosphate-buffered saline ) or either an equivalent concentration of a rat IgG isotyope control (in PBS) or an equivalent volume of PBS only, on days 0 and 3. VEGF164 intradermal injections and Miles assays were performed on day 6, as described.
Statistical Analysis
Data are presented as means±SEM. Significant differences between means were evaluated by unpaired Student t test. P<0.05 was considered significant.
Results
VEGF-A–Induced Vascular Permeability Is Enhanced in ?3-Integrin–Deficient Mice
To analyze the role of VEGF-A–mediated permeability, we used the Miles assay20 to measure vascular leakage in the skin of ?3-integrin–deficient and wild-type control mice. Mice were injected intravenously with Evans blue dye followed by intradermal injections of either 400 ng VEGF-A or vehicle (saline) alone. Spectrophotometric analysis of specimens from 8-mm biopsy punches performed around the sites of injection (using the point of injection as the center of the punch) showed that within 15 minutes after VEGF-A treatment, blood vessels in ?3-null mice were 2.5-fold more permeable than vessels in wild-type mice (Figure 1A). At this suboptimal concentration of VEGF-A, vessels in wild-type animals exhibited minimal dye leakage when compared with vehicle-only controls. Moreover, a higher (600 ng) dose of VEGF-A was required to induce a significant degree of plasma leakage in wild-type mice, indicating that ?3-null mice are more responsive to VEGF-A.
Figure 1. VEGF-A–induced vascular permeability is enhanced in ?3-null mice. A, Permeability of skin vessels after intradermal injections of VEGF-A. Values are presented as the ratio of VEGF-induced Evans blue leakage/vehicle alone-induced Evans blue leakage (±SEM), n=6 to 9 animals per genotype. The photographs illustrate enhanced leakage of Evans blue dye in ?3-null skin in response to 400 ng of VEGF-A; nsd=no significant difference. Scale bar=5 mm. B, Circulating levels of VEGF. The bars represent serum concentrations of VEGF (±SEM), n=6 mice per genotype. C, Baseline vascular permeability in the skin of ears. Bars represent ng of Evans blue leakage per mg weight of ear (±SEM), n=8 to 9 animals per genotype. The photographs show minimal Evans blue dye leakage in whole ears in the absence of an inflammatory stimulus. Scale bar=10 mm. D, Mustard oil-induced permeability in the skin of ears. Bars represent ng of Evans blue leakage per mg weight of ear (±SEM), n=8 to 9 animals per genotype. The inserted graph represents fold stimulation of treated/vehicle alone in wild-type and ?3-null mice. The photographs illustrate extensive Evans blue dye leakage in inflamed ears of wild-type and ?3-null mice. Scale bar=10 mm.
The extravasated Evans blue measured in this type of experiment is caused by both vessel permeability and vessel dilation. Although we cannot distinguish between the two, vessel dilation usually accounts for <0.01% of leakage in the type of assay described.21,22
Baseline Permeability and Inflammation-Induced Permeability Are not Affected by ?3-Integrin Deficiency
Because VEGF-A-mediated vessel permeability was enhanced in ?3-null mice, we then asked if ?3-integrin–deficiency affected baseline vessel permeability. To address this, we first examined circulating VEGF levels in wild-type and ?3-null mice. Enzyme-linked immunosorbent assay results measuring serum levels of VEGF are presented in Figure 1B. No significant differences in VEGF levels were observed between wild-type and ?3-null mice, indicating that in a young, healthy population of mice, circulating levels of VEGF are not affected by the absence of ?3-integrin. In addition, no evidence of edema was observed in the ?3-null mice (data not shown).
We next compared vascular leakage in ear skin under baseline conditions. Mice were injected intravenously with Evans blue dye followed by topical application of mineral oil (a noninflammatory agent used as the vehicle in subsequent inflammatory assays) to the epidermis of the ear. Spectrophotometric analysis of dissected ears showed that within the 30-minute time frame of the experiment, no significant difference was observed in dye leakage between wild-type and ?3-null mice (Figure 1C).
We next investigated vascular leakage in ear skin in response to mustard oil, an inflammatory agent that induces acute plasma leakage in the skin.23 Mice were injected intravenously with Evans blue dye, followed by topical application of mustard oil to the epidermis of the ear, as described (for the baseline assays). Spectrophotometric analysis of dissected ears showed that 30 minutes after Evans blue administration, mustard oil treatment induced 2.5-fold increase in dye leakage (when compared with baseline levels) in vessels from both wild-type and ?3-null mice; no difference was observed when comparing levels of leaked dye between the 2 genotypes (Figure 1D). Although the data presented are from samples collected 30 minutes after initial mustard oil treatment, we have not seen any differences in vascular leakage between ?3-null and wild-type vessels at either earlier (15 minutes) or later (60 minutes) time points (data not shown). Taken together, these data indicate that baseline permeability and inflammation induced permeability of vessels in ?3-null mice is equivalent to that of vessels in wild-type mice.
Blood Vessels in the Skin of ?3-Null Mice Are Morphologically Normal
The vessels in the skin of ?3-null mice appeared normal. The skin was not reddened and we saw no overt signs of edema (data not shown). We have shown previously that there are no differences between wild-type and ?3-null mice in the number of platelet and endothelial cell adhesion molecule (PECAM)-positive vessels in normal back skin.15 The architecture of the microvasculature was examined in whole-mount preparations of ear skin from wild-type and ?3-null mice in which vessels were visualized with biotin-labeled BS-1 lectin, which binds to the luminal surface of endothelial cells.24 The vessels throughout the skin of the ear appeared normal (Figure 2A). A quantitative analysis revealed no differences in vessel area density (reflecting the overall number, length, and size of vessels per unit area), vessel number density (reflecting the total number of vessels traversing the ear per unit area, independent of vessel size), or in the average diameter of microvessels (Figure 2B).
Figure 2. Blood vessel morphology is normal in ?3-null mice. A, BS-1 lectin visualization of microvessel morphology near the margin of the ear (art, arteriole; cap, capillary; ven, venule). Scale bar=100 μm. B, Morphometric quantitation of blood vessels in BS-1 lectin-stained ear whole mounts. Values represent mean (±SEM) of 3 to 6 microscope fields per ear and 5 ears per genotype. C, Ricin lectin visualization of exposed basement membrane surrounding mustard oil-treated blood vessels near the margin of the ear. Arrowheads indicate sites of leakage. Scale bar=30 μm.
The amount of plasma leakage in inflammation depends, at least in part, on the number and size of sites for leakage in the endothelium.25 These leakage sites expose the underlying endothelial basement membrane and can therefore be visualized in whole-mount preparations with ricin lectin, which binds more avidly to components of the basement membrane than the endothelium.26 In untreated vessels, rhodamine-labeled ricin lectin binding to the endothelial luminal surface was weak and uniform in both genotypes (data not shown). In mustard oil-treated ears, ricin lectin bound very strongly to focal sites of exposed basement membrane in vessels presumed to be venules based on vessel diameter and various models of acute inflammation.27 The degree of ricin lectin binding was comparable in both genotypes. We did not see any differences in the number (data not shown) or location of exposed basement membrane sites in wild-type versus ?3-null vessels (Figure 2C).
Analysis by transmission electron microscopy confirmed an intact microvascular network in the skin of untreated ?3-null mice. ?3-null vessels, including capillaries, were unaltered in size and showed normal pericyte recruitment and cell–cell contacts between endothelial cells and supporting cells (Figure 3A and data not shown).
Figure 3. Cell–cell and cell–ECM associations appear normal in ?3-null blood vessels. A, Upper panel shows transmission electron microscopy of vessels in the skin (e, endothelial cell nucleus; s, supporting cell body). Scale bar=2 μm. Lower panel shows higher magnifications of similar sections to those in upper panel. Arrowheads indicate electron-dense endothelial cell–cell junctions. Scale bar=200 nm. B, Western blot analysis of occludin levels in lung endothelial cells. Blotting for HSC 70 provides a loading control. The bar chart represents averaged densitometry results (means±SEM) from multiple independent experiments that show wild-type and ?3-null endothelial cells express equivalent levels of occludin. C, The panels show representative flow cytometric analyses of PECAM and VECAD levels on lung endothelial cells. D, The micrographs show immunofluorescence for PECAM (upper panels) and VECAD (lower panels) in both blood vessels from skin and in cultured lung endothelial cells. Positive, bright staining is concentrated at the borders between endothelial cells in both genotypes. Scale bar=5 μm in skin sections and 3 μm in cultured cells.
Endothelial cell-to-cell junctions are complex structures formed by different adhesive molecules.28 Endothelial cells possess adherens junctions and tight junctions similar to those described in epithelial cells. Adherens junctions are ubiquitous along the vascular tree and are formed by transmembrane proteins belonging to the cadherin superfamily. Endothelial cells express a cell-specific cadherin known as vascular endothelial cell cadherin (VECAD). Tight junctions comprise 3 types of transmembrane proteins, occludins, claudins and junctional adhesion molecule. In addition, other adhesive proteins, such as PECAM, and endoglin are concentrated at intercellular contacts in the endothelium. As a measure of the molecular architecture of junctions in wild-type and ?3-null endothelial cells, we performed Western blot analysis for occludin (Figure 3B) on lung endothelial cells derived from wild-type and ?3-null animals; flow cytometry for PECAM and VECAD (Figure 3C) on lung endothelial cells derived from wild-type and ?3-null animals; and immunofluorescence (Figure 3D) on both frozen skin sections and isolated lung endothelial cells from both genotypes. Wild-type and ?3-null endothelial cells showed equivalent expression of all 3 cell–cell adhesion molecules, suggesting that adherens junctions, tight junctions, and other cell adhesions are normal in ?3-null cells.
Inhibition of Flk-1 Function Abrogates VEGF-A–Induced Permeability in ?3-Null Mice
Given that no differences in either vessel structure or junctional protein levels were observed when comparing wild-type and ?3-null blood vessels, but that VEGF-A–mediated permeability and Flk-1 levels (Figure 4A) were significantly elevated in ?3-null mice, we asked if elevated Flk-1 levels were directly responsible for the enhanced VEGF-A–mediated vessel permeability response. To address this in vitro, we cultured monolayers of wild-type and ?3-null endothelial cells in transwell chambers and measured the flux of FITC-dextran across the monolayers in response to treatment with either VEGF-A alone or VEGF-A in combination with the neutralizing anti–Flk-1 antibody DC10116 (Figure 4B). To address this in vivo, we administered DC101 to wild-type and ?3-null mice and measured VEGF-A–induced vessel leakage using the Miles assay (Figure 4C). Both analyses revealed that by inhibiting the function of Flk-1, VEGF-A–induced permeability was completely abolished in ?3-null endothelium. These results indicate that increased levels of Flk-1 in ?3-null mice are directly responsible for an increased sensitivity to VEGF-A–mediated blood vessel permeability.
Figure 4. Elevated VEGF-A–induced blood vessel permeability in ?3-null mice is via Flk-1. A, Western blot analysis of Flk-1 levels in lung endothelial cells. Blotting for HSC 70 provides a loading control. The bar chart represents averaged densitometry results (means±SEM) from multiple independent experiments that show ?3-null endothelial cells express 2- to 3-fold more Flk-1 than wild-type controls. B, Effect of ?3-integrin deficiency in endothelial barrier function in vitro. VEGF-mediated permeability is inhibited by DC101 in ?3-null endothelial monolayers when compared with wild-type controls. Bars show the percentage of FITC-dextran flux across the monolayer in comparison to untreated controls (±SEM), n=4 independent wells per genotype/per treatment. C, Permeability of skin vessels in vivo after intradermal injections of suboptimal doses of VEGF-A in mice pretreated with DC101 or control (rat isotype matched IgG or PBS, collectively shown as control treatment). DC101 inhibits VEGF-A–induced vascular leakage in ?3-null mice. Bars show the ratio of VEGF-induced Evans blue leakage/vehicle alone-induced Evans blue leakage (±SEM), n=6 to 14 animals per genotype/per treatment.
Discussion
The establishment and maintenance of a vascular network is critical for normal embryological development and has significant implications for pathological processes such as wound healing and tumorgenesis. Numerous investigations have demonstrated that interactions between integrins and the ECM are important to these processes.2,6,29–32
A major focus of these studies has been on the roles played by v?3-integrin and v?5-integrin in angiogenesis.33–37 During angiogenesis, v?3 expression is upregulated.38 Furthermore, its expression is linked to cell migration, invasion, and cell survival.38–40 Given these collective data, it is surprising that genetic ablation of v-integrin elicits only minor vessel defects.41 Although most v-deficient mice die soon after birth, a complex embryonic vascular network develops. In addition, ?3-deficient mice are viable and fertile and appear to produce a vascular network without obvious defects.17,42 A detailed morphological examination of skin blood vessels in ?3-null mice, presented here, indicates that the vasculature is intact and indistinguishable from that of wild-type mice, reinforcing the conclusion that ?3-integrin is not necessary for establishing a normal vasculature.
Endothelial cells attach to the ECM to form a semi-permeable barrier restricting movement of solutes across the vascular wall. Evidence that integrins are involved in maintaining this barrier comes from studies showing that treating endothelial monolayers or isolated coronary venules with RGD peptides (small molecules that disrupt interactions between integrins and the ECM) increase their permeability.43–45 The fact that the increased permeability in these in vitro studies is prevented by the addition of soluble fibronectin or vitronectin suggests that interactions between integrins and the ECM influence vessel permeability. From the data we present here, however, we are able to conclude that ?3-integrin is not essential for maintaining an intact and functional endothelial barrier. These data are as follows: (1) baseline vessel permeability is not altered in ?3-deficient mice; (2) a molecular comparison of the endothelial cell-to-cell contacts between the 2 genotypes revealed no differences in the expression of PECAM, VECAD, or occludin; and (3) wild-type and ?3-null blood vessels respond equally to mustard oil-induced plasma leakage. Collectively, these data indicate that ?3-integrin is not necessary for maintaining baseline integrity of the endothelial barrier.
In marked contrast to the lack of an apparent phenotype in vessel structure and baseline integrity, we see an increase in the sensitivity of ?3-null vessels to subthreshold doses of VEGF-A treatment. Changes in matrix composition and integrin activation can influence levels of, and signaling through, VEGF and fibroblast growth factor receptors on cultured microvascular endothelial cells.9,46,47 We show that the absence of ?3-integrin is associated with an increase in the levels of Flk-1 expressed on endothelial cells and, for the first time to our knowledge, that this increase in Flk-1 leads to an increased sensitivity to VEGF-A–mediated permeability in vitro and in vivo. Previously, we were able to draw correlations between only elevated Flk-1 levels and enhanced VEGF-mediated angiogenesis.15 With respect to permeability, however, we show that by inhibiting signaling through Flk-1, we abolish the enhanced VEGF-A response seen in ?3-null vessels. This demonstrates that the increased Flk-1 is functional and is entirely responsible for the elevated sensitivity to VEGF-A–induced permeability.
Conventional treatment of vascular leakage is rather aspecific and includes treatments such as the administration of glucocorticoids or nonsteroidal inflammatory drugs and antihistaminergic compounds. Recently, VEGF antagonists48 and VEGF receptor inhibitors49 have been used with some success in treating the increased vessel leakiness associated with macular degeneration, diabetic retinopathy, and ischemia–reperfusion injury. By genetically ablating ?3-integrin function, we have altered blood vessel sensitivity to VEGF-A–induced plasma leakage, indicating that integrins might be additional useful targets for clinically manipulating vessel permeability.
Antagonists of v?3 and v?5, some of which are in clinical trials,50 have been shown to block angiogenesis, suggesting that these integrins are required for angiogenesis. However, we have shown that ?3-integrin–deficient mice support enhanced pathological angiogenesis.15 The data we present here and in our previous studies suggest that ?3-integrin acts as a transdominant inhibitor of VEGF-mediated vascular functions, with its presence being enough to downregulate levels of Flk-1. We have previously demonstrated that the reintroduction of ?3-integrin into null endothelial cells restores Flk-1 levels to those seen in wild-type cells,15 suggesting that ?3-integrin can regulate Flk-1 expression. Hence, eliminating ?3-integrin interactions with the ECM might alter the response of endothelial cells to VEGF. The antagonist of v?3-integrin, LM609, has been shown to enhance vascular permeability.51 If inhibition of v?3-integrin has the potential of elevating Flk-1 levels in endothelial cells, both angiogenesis and permeability of tumor vessels could be increased, especially if, as many are,52 the tumor is a source of VEGF production. This might prove detrimental to the treatment of the tumor. We do not question the efficacy of v?3 antagonists as potentially useful therapeutic agents, but our results suggest we need a more thorough understanding of the interactions between integrins and other molecules that are involved in regulating angiogenesis and permeability.
Many vascular disorders are accompanied by alterations in production and/or degradation of ECM components and by alterations in integrin expression.4 Because of this, integrins are obvious therapeutic targets for treating a host of vascular diseases such as hypertension, diabetes mellitus, and restenosis.53–56 This idea is exciting and should be explored, but with the thought in mind that manipulating integrin expression might lead to an increase in growth factor receptor levels expressed by endothelial cells. As we show here, genetically ablating ?3-integrin function elevates Flk-1 levels on endothelial cells. This has the potential of affecting vessel growth and vessel permeability, both of which have important implications for current and future therapeutics. Vessel leakiness, for example, is known to contribute to the abnormal microenvironment of tumors and can affect tumor growth, metastasis, and therapeutic drug delivery.57 As long as we exercise caution, though, integrins may provide us with useful targets for treating a range of vascular disorders.
Acknowledgments
We thank Garry Saunders, Sue Watling, Colin Wren, and Stephen Gschmeissner for their technical assistance, and Ian Hart, Francesco Conti, Andrew Reynolds, and the other members of the Department of Tumor Biology for their support and criticism during this study. We also thank Mark Karaczun for his critical reading of the manuscript.
References
Hynes R. Integrins: versatility, modulation, and signalling in cell adhesion. Cell. 1992; 69: 11–25.
Luscinskas F, Lawler J. Integrins as dynamic regulators of vascular function. FASEB J. 1994; 8: 929–938.
Hynes R, Bader B, Hodivala-Dilke K. Integrins in vascular development. Braz J Med Biol Res. 1999; 32: 501–510.
Martinez-Lemus LA, Wu X, Wilson E, Hill MA, Davis GE, Davis MJ, Meininger GA. Integrins as unique receptors for vascular control. J Vasc Res. 2003; 40: 211–233.
Francis SE, Goh KL, Hodivala-Dilke K, Bader BL, Stark M, Davidson D, Hynes RO. Central roles of 5?1 integrin and fibronectin in vascular development in mouse embryos and embryoid bodies. Arterioscler Thromb Vasc Biol. 2002; 22: 927–933.
Madri J, Williams, S. Capillary endothelial cell cultures: phenotypic modulation by matrix components. J Cell Biol. 1983; 97: 153–165.
Form D, Pratt B, Madri J. Endothelial cell proliferation during angiogenesis. In vitro modulation by basement membrane components. Lab Invest. 1986; 55: 521–530.
Underwood P, Bennett F. The effect of extracellular matrix molecules on the in vitro behavior of bovine endothelial cells. Exp Cell Res. 1993; 205: 311–319.
Tsou R, Isik FF. Integrin activation is required for VEGF and FGF receptor protein presence on human microvascular endothelial cells. Mol Cell Biochem. 2001; 224: 81–89.
Senger D, Galli S, Dvorak A, Perruzzi C, Harvey V, Dvorak H. Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science. 1983; 219: 983–985.
Ferrara N. Role of vascular endothelial growth factor in regulation of physiological angiogenesis. Am J Physiol Cell Physiol. 2001; 280: C1358–C1366.
Bates DO, Hillman NJ, Williams B, Neal CR, Pocock TM. Regulation of microvascular permeability by vascular endothelial growth factors. J Anat. 2002; 200: 581–597.
Ruhrberg C. Growing and shaping the vascular tree: multiple roles for VEGF. Bioessays. 2003; 25: 1052–1060.
Cross MJ, Dixelius J, Matsumoto T, Claesson-Welsh L. VEGF-receptor signal transduction. Trends Biochem Sci. 2003; 28: 488–494.
Reynolds LE, Wyder L, Lively JC, Taverna D, Robinson SD, Huang X, Sheppard D, Hynes RO, Hodivala-Dilke KM. Enhanced pathological angiogenesis in mice lacking ?3 integrin or ?3 and ?5 integrins. Nat Med. 2002; 8: 27–34.
Prewett M, Huber J, Li Y, Santiago A, O’Connor W, King K, Overholser J, Hooper A, Pytowski B, Witte L, Bohlen P, Hicklin D. Anti-VEGF receptor (Flk-1) monoclonal antibody inhibits tumor angiogenesis and growth of several mouse and human tumors. Cancer Res. 1999; 59: 5209–5218.
Hodivala-Dilke K, McHugh K, Tsakiris D, Rayburn H, Crowley D, Ullman-Cullere M, Ross F, Coller B, Teitelbaum S, Hynes R. ?3-integrin deficient mice are a model for Glanzmann thrombasthenia showing placental defects and reduced survival. J Clin Invest. 1999; 103: 229–238.
Thurston G, Murphy TJ, Baluk P, Lindsey JR, McDonald DM. Angiogenesis in mice with chronic airway inflammation: strain-dependent differences. Am J Pathol. 1998; 153: 1099–1112.
Wójciak-Stothard B, Potempa S, Eichholtz T, Ridley AJ. Rho and Rac, but not Cdc42 regulate endothelial cell permeability. J Cell Sci. 2001; 114: 1343–1355.
Miles A, Miles E. Vascular reactions to histamine, histamine-liberator and leukotaxine in the skin of guinea-pigs. J Physiol. 1952; 118: 228–257.
Buckley T, Brain S, Williams T. Ruthenium red selectively inhibits oedema formation and increased blood flow induced by capsaicin in rabbit skin. Br J Pharmacol. 1990; 99: 7–8.
Hughes S, Brain S, Williams G, Williams T. Assessment of blood flow changes at multiple sites in rabbit skin using a 133Xenon clearance technique. J Pharmacol Toxicol Methods. 1994; 32: 41–47.
Baluk P, Thurston G, Murph TJ, Bunnett NW, McDonald DM. Neurogenic plasma leakage in mouse airways. Br J Pharmacol. 1999; 126: 522–528.
Plendl J, Neumuller C, Vollmar A, Auerbach R, Sinowatz F. Isolation and characterization of endothelial cells from different organs of fetal pigs. Anat Embryol (Berl). 1996; 194: 445–456.
Zweifach BW. Microcirculation. Annu Rev Physiol. 1973; 35: 117–150.
Bergman RK, Munoz JJ. Effect of Bordetella pertussis extract and vasoactive amines on vascular permeability. J Allergy Clin Immunol. 1975; 55: 378–385.
Svensjo E, Grega G. Evidence for endothelial cell-mediated regulation of macromolecular permeability by postcapillary venules. Fed Proc. 1986; 45: 89–95.
Dejana E. Endothelial cell-cell junctions: happy together. Nat Rev Mol Cell Biol. 2004; 5: 261–270.
Ingber D, Folkman J. Mechanochemical switching between growth and differentiation during fibroblast growth factor-stimulated angiogenesis in vitro: role of extracellular matrix. J Cell Biol. 1989; 109: 317–330.
Ingber D. Extracellular matrix and cell shape: potential control points for inhibition of angiogenesis. J Cell Biochem. 1991; 47: 236–241.
Klasgsbrun M. Regulators of angiogenesis: stimulators, inhibitors, and extracellular matrix. J Cell Biochem. 1991; 47: 199–200.
Kanda S, Tomasini-Johansson B, Klint P, Dixelius J, Rubin K, Claesson-Welsh L. Signaling via fibroblast growth factor receptor-1 is dependent on extracellular matrix in capillary endothelial cell differentiation. Exp Cell Res. 1999; 248: 203–213.
Byzova T, Rabbani R, Souza S, Plow E. Role of integrin v?3 in vascular biology. Thromb Haemost. 1998; 80: 726–734.
Friedlander M, Brooks P, Shaffer R, Kincaid C, Varner J, Cheresh D. Definition of two angiogenic pathways by distinct v integrins. Science. 1995; 270: 1500–1502.
Stromblad S, Cheresh, D. Integrins, angiogenesis and vascular cell survival. Chem Biol. 1996; 3: 881–885.
Eliceiri B, Cheresh D. The role of v integrins during angiogenesis: insights into potential mechanisms of action and clinical development. J Clin Invest. 1999; 103: 1227–1230.
Hodivala-Dilke KM, Reynolds AR, Reynolds LE. Integrins in angiogenesis: multitalented molecules in a balancing act. Cell Tissue Res. 2003; 314: 131–144.
Brooks P, Montgomery A, Rosenfeld M, Reisfeld R, Hu T, Klier G, Cheresh, D. Integrin v ? 3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell. 1994; 79: 1157–1164.
Felding-Habermann B, Cheresh D. Vitronectin and its receptors. Curr Opin Cell Biol. 1993; 5: 864–868.
Brassard D, Maxwell E, Malkowski M, Nagabhushan T, Kumar C, Armstrong L. Integrin (v)?(3)-mediated activation of apoptosis. Exp Cell Res. 1999; 251: 33–45.
Bader B, Rayburn H, Crowley D, Hynes R. Extensive vasculogenesis, angiogenesis and organogenesis precede lethality in mice lacking all v integrins. Cell. 1998; 95: 507–519.
Huang X, Griffiths M, Wu J, Farese R, Sheppard D. Normal development, wound healing, and adenovirus susceptibility in ?5-deficient mice. Mol Cell Biol. 2000; 20: 755–759.
Curtis T, McKeown-Longo P, Vincent P, Homan S, Wheatley E, Saba T. Fibronectin attenuates increased endothelial monolayer permeability after RGD peptide, anti- 5 ? 1, or TNF- exposure. Am J Physiol. 1995; 269: L248–L260.
Qiao R, Yan W, Lum H, Malik A. Arg-Gly-Asp peptide increases endothelial hydraulic conductivity: comparison with thrombin response. Am J Physiol. 1995; 269: C110–C117.
Wu M, Ustinova E, Granger H. Integrin binding to fibronectin and vitronectin maintains the barrier function of isolated porcine coronary venules. J Physiol. 2001; 532: 785–791.
Soldi R, Mitola S, Strasly M, Defilippi P, Tarone G, Bussolino F. Role of v?3 integrin in the activation of vascular endothelial growth factor receptor-2. EMBO J. 1999; 18: 882–892.
Byzova TV, Goldman CK, Pampori N, Thomas KA, Bett A, Shattil SJ, Plow EF. A mechanism for modulation of cellular responses to VEGF: activation of the integrins. Mol Cell. 2000; 6: 851–860.
van Bruggen N, Thibodeaux H, Palmer J, Lee W, Fu L, Cairns B, Tumas D, Gerlai R, Williams S, van Lookeren Campagne M, Ferrara N. VEGF antagonism reduces edema formation and tissue damage after ischemia/reperfusion injury in the mouse brain. J Clin Invest. 1999; 104: 1613–1620.
The T.E.S. Group. Preclinical and phase 1A clinical evaluation of an anti-VEGF peglyated aptamer (EYE001) for the treatment of exudative age-related macular degeneration. Retina. 2002; 22: 143–152.
Gutheil J, Campbell T, Prierce P, Watkins J, Huse W, Bodkin D, Cheresh D. Targeted antiangiogenic therapy for cancer using Vitaxin: a humanized monoclonal antibody to the integrin v?3. Clin Cancer Res. 2000; 6: 3056–3061.
Tsukada H, Ying X, Fu C, Ishikawa S, McKeown-Longo P, Albelda S, Bhattacharya S, Anderson Bray B, Bhattacharya J. Ligation of endothelial v?3 integrin increases capillary hydraulic conductivity of rat lung. Circ Res. 1995; 77: 651–659.
Mukhopadhyay D, Datta K. Multiple regulatory pathways of vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) expression in tumors. Semin Cancer Biol. 2004; 14: 123–130.
Roth T, Podesta F, Stepp M, Boeri D, Lorenzi M. Integrin overexpression induced by high glucose and by human diabetes: potential pathway to cell dysfunction in diabetic microangiopathy. Proc Natl Acad Sci U S A. 1993; 90: 9640–9644.
Regoli M, Bendayan M. Alterations in the expression of the 3 ? 1 integrin in certain membrane domains of the glomerular epithelial cells (podocytes) in diabetes mellitus. Diabetologia. 1997; 40: 15–22.
Intengan H, Thibault G, Li J, Schiffrin E. Resistance artery mechanics, structure, and extracellular components in spontaneously hypertensive rats: effects of angiotensin receptor antagonism and converting enzyme inhibition. Circulation. 1999; 100: 2267–2275.
Bishop G, McPherson J, Sanders J, Hesselbacher S, Feldman M, McNamara C, Gimple L, Powers E, Mousa S, Sarembock I. Selective (v)?(3)-receptor blockade reduces macrophage infiltration and restenosis after balloon angioplasty in the atherosclerotic rabbit. Circulation. 2001; 103: 1906–1911.
McDonald DM, Baluk P. Significance of blood vessel leakiness in cancer. Cancer Res. 2002; 62: 5381–5385.(Stephen D. Robinson; Loui)