Phase I Maintenance of Quiescence by the Basement Membrane

Insofar as angiogenesis is the process of growth of a preexisting vasculature, it is appropriate to begin our discussion with the ECM status of the quiescent, differentiated blood vessel. Basally, the capillary secretes and assembles a basement membrane scaffold, which exists in close contact with the EC that make up the capillary network (Figure 2, upper panel). The basement membrane contains several classes of glycoproteins and proteoglycans. The basement membrane conglomerate is considered to play multiple roles in the vascular system, from partitioning EC from the surrounding stroma and providing structural support to the vasculature, to creating a charge barrier between the capillary and adjacent cells and tissues. However, in addition to playing these roles, it is also clear that many of the basement membrane components may also act individually or in concert to exert significant activities on angiogenesis.

Basement Membrane Structure

The basement membrane is a highly organized, net-like molecular conglomerate. At least two modes of basement membrane construction may take place: self-assembly driven by the molecules themselves and cell-assisted processes. The two major components comprising the network are type IV collagen and laminin. Type IV collagen is proposed to self-assemble through multimerization via its N-terminal S domains and C-terminal noncollagenous domains, and lateral associations between its central col-

Figure 1 The Microvasculature and the ECM are Intimately Associated. This electron micrograph of an arteriole illustrates that EC that line the vessel lumen (Lu) are surrounded closely by circumferentially arranged smooth muscle cells (SMC) and interstitial ECM. A basement membrane, deemed to be phase I of angiogenic regulation in this review, lies between the EC and the underlying vessel components, but it is too thin to be resolved at this magnification. Interstitial collagen fibrils (Co), comprising phase II of angiogenic regulation, can be seen in abundance underneath the EC layer and throughout the media.

Figure 1 The Microvasculature and the ECM are Intimately Associated. This electron micrograph of an arteriole illustrates that EC that line the vessel lumen (Lu) are surrounded closely by circumferentially arranged smooth muscle cells (SMC) and interstitial ECM. A basement membrane, deemed to be phase I of angiogenic regulation in this review, lies between the EC and the underlying vessel components, but it is too thin to be resolved at this magnification. Interstitial collagen fibrils (Co), comprising phase II of angiogenic regulation, can be seen in abundance underneath the EC layer and throughout the media.

lagenous regions. Similarly, the laminin scaffold is thought to form through oligomerization of laminin monomers (M = 900 Kd), likely driven by low-affinity interactions between the N-terminal regions of the a, b, and g arms of the monomers, and may be further modulated by the local protein and calcium concentration. The other major components of basement membranes include the glycoprotein nidogen (M @ 80 Kd) and at least three proteoglycans, including perlecan (M @ 750 Kd), agrin, and type XVIII collagen. In addition, type XV collagen harbors chondroitin sulfate side chains. Nidogen is proposed to serve as the molecular glue that holds the basement membrane together. Thus, the C-terminal domain of nidogen binds with high affinity to the short arm of the laminin g chain, and through weaker affinity interactions, via its other domains, with type IV collagen and perlecan. The heparan sulfate (HS) chains of perlecan and, presumably, those of collagen XVIII and agrin, may also stabilize basement membrane structure because heparin-binding is exhibited by laminin and type IV collagen preparations. Depending on the tissue, other non-structural components of basement membranes may include type XVIII or XV collagens, the agrin proteoglycan, or non-structural components associated with the basement membranes such as basic fibroblast growth factor or matrix metalloproteinases (MMPs), proposed to bind to and be sequestered by the HS chain components of perlecan. All of the components of basement membranes are synthesized and secreted by EC, which may help regulate or direct its assembly. For example, disruption of the function of the cell surface b1 integrin receptor interferes with the ordered deposition of type IV collagen during basement membrane formation.

The Basement Membrane and EC Quiescence

In the quiescent capillary, EC are tightly associated with a basement membrane. It is proposed that the basement membrane may promote EC quiescence in two ways: (1) the intact matrix scaffold may exert strong differentiation-promoting activities on EC, and (2) individual basement membrane components or their proteolytic products may exert strong antiproliferative and antiangiogenic activities (Figure 2, upper and lower left panels).

Role of the Intact Basement Membrane

The most compelling evidence supporting the role of the intact basement membrane in promoting EC quiescence and differentiation is circumstantial in nature (i.e., the fact that quiescent capillaries assemble basement membranes and exist in tight association with them). Moreover, a hallmark of other differentiated cell types that, like the microvascula-ture, differentiate to form polarized, morphologically distinct tissues, is the production of and association with basement membranes. Not surprisingly, EC and many other cell types differentiate in vitro when cultured within or on top of basement membrane mixtures such as Matrigel. Further evidence in support of the role of the basement membrane in angiogenesis is supplied by studies that identified genes upregulated by at least two-fold in various angiogen-esis systems [1]. Coincident with capillary tube differentiation in response to either type I collagen or fibrin, mRNA for various basement membrane components including laminin chains a4, b1, and b2, type IV collagen, the type XVIII collagen aI chain, and nidogens 1 and 2, are increased along with other molecules directly or indirectly related to basement membrane structure and function. The latter include collagen synthesis enzymes, MMPs 1, 2, and 9, disintegrin, and ADAMs (A disintegrin-like and metalloproteinase domain) -9, 10, and 17. More work is needed to understand how the complex basement membrane scaffold acts to promote EC quiescence. In this regard, it may be useful to consider how the basement membrane regulates differentiation in mammary epithelial cells that respond to the three-dimensional matrix environment in very defined ways, as elucidated by Mina J. Bissel and co-workers. Thus, when these epithelial cells interact with laminin, they change their cell morphology, undergo growth arrest, and in the presence of prolactin, express the differentiation marker, the milk protein b-casein. A multiplicity of cell surface receptors, including the a6b4 integrin, b1 integrins, and an E3 laminin receptor, were found to mediate these cellular responses. Whether the intact basement membrane promotes EC quiescence and the expression of EC-specific differentiation in a similar manner remains to be determined.

Role of Individual Basement Membrane Components

Most of the research probing the function of basement membranes in angiogenesis has taken the reductionistic approach of assaying the activities of individual basement

Figure 2 The Two-Phase Matrix Model of Angiogenesis Regulation. (Phase I, Top of Diagram) It is proposed that the basement membrane phase, shown here on an orange background, promotes EC quiescence and maintains capillary integrity in two ways: (1) via stable interactions between cell surface receptors and various basement membrane components on the basal EC side, and (2) by the release of antiangiogenic fragments of basement membrane components through the continuous low-level proteolysis and basement membrane turnover that may occur during quiescence. For the sake of clarity, this latter activity is shown only on the EC apical side, although it is presumed to occur virtually anywhere the EC has contact with the basement membrane. Antiangiogenic fragments may inhibit EC growth and migration by binding to unique receptors (cell on left) or by binding to and interfering with integrin receptor function (cell on right). (Phase I, Bottom Left of Diagram) Upon growth factor stimulation, or during tissue injury, massive proteolysis, likely mediated by MMPs, causes the degradation and collapse of the basement membrane, and integrins or other cell surface receptors may mediate the migration of EC into the second matrix phase (intussusception). It is proposed that the fibrillar collagen or fibrin phase, shown on a blue background, catalyzes capillary tube construction in two ways: (1) by presenting a linear, rigid scaffold that helps align EC into presumptive tubes, and (2) by presenting regularly spaced ligand-binding sites that induce clustering of EC receptors such as the a2|31 integrin in the case of type I collagen, or VE-cadherin in the case of fibrin, that induce capillary lumen formation and perhaps the expression of genes required for the maintenance of EC differentiation. (see color insert)

membrane components, or their various domains on EC behavior. It is difficult to reconcile the relevance of these findings to angiogenesis, because it is not clear whether the various constituents of basement membranes are accessible to the EC during quiescence or throughout the various stages of capillary development. Nonetheless, these investigations will be summarized, and it will be assumed here, for the sake of simplicity, that the following findings are most relevant to two aspects of EC biology: (1) that basement membrane components may be exposed during the constant, low-level turnover of this ECM compartment, which may occur during EC quiescence (Figure 2, upper panel), or (2) alternatively, during the massive proteolysis and collapse of the matrix scaffold at the initiation of angiogenesis or blood vessel remodeling (Figure 2, lower left panel). Early approaches examined how EC responded to culture on a reconstituted basement membrane gel. It was found that tubes formed well within one day and that, although blocking laminin function in various ways inhibits tube morphogenesis, neither laminin nor type IV collagen gels alone support tube formation. These findings imply that a multiplicity of interactions among various cell surface receptors and basement membrane components are required for tube formation in response to the basement membrane, but again, it is questionable whether in vivo EC ever encounter these matrix components as individual entities rather than as complex supramolecular assemblies.


Several groups have mapped the regions of laminin that influence various EC behaviors. For example, an RGD pep-tide in the a chain of laminin promotes EC adhesion. The

YIGSR sequence in the b1 chain, when substrate-bound, induces cell-to-cell interactions and promotes tube formation, and soluble peptide inhibits that process; the IKVAV sequence from the a1 chain also promotes angiogenesis. Moreover, it has been shown that MMP-2 cleavage of laminin-5 exposes a cryptic site in the g2 subunit that promotes cell migration. Others examined the activity of a laminin-nidogen mixture and showed that it can promote EC sprouting in vitro at very low concentrations but is inhibitory at high concentrations, consistent with a quiescence-promoting function of the intact basement membrane.

Type IV Collagen

Type IV collagen contains multiple sites for EC surface binding, including to a1b1, a2b1, and a3b1 integrins. Non-collagenous (NC)-1 domains from the a1(IV), a3(IV), and ab6(IV) chains of collagen IV support integrin-dependent EC adhesion and migration. NC1 domains from a1(IV), a4(IV), a3(IV), and a6(IV) are potent inhibitors of angiogenesis in vivo. Angiogenesis inhibitors derived from these fragments include arresten, generated from a1(IV); canstatin, generated from a2(IV); and tumstatin, a product of a3(IV) degradation. A site exists in collagen IV that is normally cryptic but may be unmasked through exposure to MMP-2. Blocking the function of this site in vivo disrupts angiogenesis and tumor growth.


The a3b1 integrin has been reported to bind directly to this basement membrane component, and its fragments regulate cellular adhesion. Because nidogen may be critical in stabilizing basement membrane structure, it can be speculated that once it undergoes proteolysis, the basement membrane scaffold may fall apart or may be easily breached by EC, and that nidogen's fragments may regulate EC migration into the tissue stroma.


This predominant proteoglycan of the basement membrane can interact with cell surfaces via its core protein, as well as modulating cell proliferation through its sequestration or release of various growth factors. EC are proposed to bind to the protein core via the a1, b1, and b3 integrin receptors. Moreover, the activity of basic fibroblast growth factor (FGF2) on cell growth and vascular morphogenesis is potentiated through its associations with perlecan and other proteoglycans. Thus, perlecan, but not syndecans or glypi-cans, extracted from fibroblast cultures facilitates high-affinity FGF2 binding to Chinese hamster ovary cells deficient in endogenous HS and engineered to express the FGF receptor-1, and soluble FGF receptors. Perlecan encapsulated in alginate beads is capable of binding FGF2 and promoting extensive angiogenesis in vivo. It was proposed that perlecan functions as a low-affinity co-receptor by deliver ing FGF2 to its high-affinity cell surface receptors. Consistent with this model, it was shown using an antisense approach that blocking of perlecan expression by colon carcinoma cells attenuated their growth, which correlated with a reduction in their responsiveness to FGF-7. Furthermore, antisense-mediated inhibition of perlecan expression in tumor xenografts of colon carcinoma cells or allografts of mouse melanoma cells showed a reduced capacity for tumorigenesis and the promotion of neovascularization in nude mice. Perlecan-mediated potentiation of growth factor action may occur via sites on either or both of the gly-cosaminoglycan chains and the core protein. At the cell surface, cell-associated proteoglycans such as perlecan are proposed to contribute HS chains in the formation of a ternary complex with two FGF1 molecules, and one growth factor receptor chain, to activate the growth factor receptor. Perlecan's HS chains can also serve as reservoirs or binding sites for other angiogenic factors such as MMP-7, which, when released, may initiate or help sustain EC migration during angiogenesis.

Investigations into the functional role played by the per-lecan protein core in vascular development led to the discovery that the C-terminus potently inhibits EC migration and adhesion to type I collagen and capillary morphogenesis in vitro and in vivo. The C-terminal portion of human perlecan was named endorepellin to designate its antien-dothelial activity. This domain blocks EC-matrix adhesion without interacting directly with the matrix components themselves. Endorepellin also binds with high affinity to EC cell surfaces, as well as to endostatin, and counteracts its antiangiogenic activity. These latter observations imply that perlecan-type XVIII collagen interactions could play a structural role in promoting basement membrane stability in some tissues, but that during basement membrane dissolution, the presence of both endorepellin and endostatin would likely not interfere with the subsequent EC migration, growth, and capillary morphogenesis, because these factors may interact and thus neutralize each other. Iozzo and co-workers have shown that endorepellin may be liberated from the perlecan core by BMP-1/Tolloid-like MMPs, and various lines of evidence are consistent with the liberation of endorepellin by cell cultures in vitro, and of its presence in the circulation of humans in various pathological states.

Type XVIII and XV Collagens

These collagens are members of the mutiplexin collagen family, because they carry multiple triple-helical domains interrupted by noncollagenous regions. Type XVIII collagen is a hybrid collagen-proteoglycan molecule and is a constituent of basement membranes of blood vessels, kidney, skeletal muscle, and retina, and some other tissues. This collagen carries up to four HS chains, each attached to one of four nontriple-helical domains. The widely distributed basement membrane molecule type XV collagen is highly similar to type XVIII in primary structure and domain arrangement. Type XV collagen is also a collagen-PG

hybrid molecule, but carries chondroitin sulfate, not HS chains.

An 18Kd protein fragment of type XVIII collagen was isolated from hemangioma cell conditioned media and shown to potently inhibit EC growth in vitro, and tumor growth in vivo, and was thus named endostatin by its discoverers, Judah Folkman and coworkers. It is proposed to be liberated from the parent molecule through the action of proteases such as cathepsin L and MMPs. Endostatin was shown to potently inhibit FGF2-induced EC growth in vitro, and in the chick chorioallantoic membrane (CAM), and to nearly abolish the growth and associated angiogenesis of a variety of tumors implanted in mice. Examination of residual carcinomas in endostatin-treated mice revealed it to act not by inhibiting tumor cell growth but by increasing apoptosis sevenfold, via its inhibition of tumor-associated angiogenesis. Discontinuation of endostatin treatment resulted in the regrowth of the implanted tumors. Thus, endostatin was proposed to hold great clinical promise as a tumor dormancy factor. A later report, however, failed to observe an effect of endostatin on the FGF2-induced growth of bovine or human EC in vitro. However, these authors found endostatin to potently inhibit the VEGF-induced migration of human EC and, at very low doses, to inhibit the growth of human renal cell carcinomas implanted subcuta-neously in nude mice. A third group reported an inhibition of FGF2 but not VEGF-induced angiogenesis by endostatin. Finally, drastically different results were reported when the roles of the zinc- and heparin-binding domains of endostatin to its antiangiogenic action were examined. Notably, type XV collagen also contains a C-terminal noncollagenous domain with about 60 percent sequence homology to endo-statin, that does not interact with heparin or zinc, but nonetheless inhibits FGF2- or VEGF-induced angiogenesis in the CAM. In experiments where endostatin from type XVIII collagen and ES-XV were compared for their abilities to inhibit FGF2- or VEGF-induced angiogenesis, strikingly different activities were observed. Thus, the type XV collagen-derived fragments, ES-XV, and NC1 XV similarly inhibited VEGF-induced angiogenesis, but angiogenic stimulation by FGF2 was potently inhibited by NC1-XV but not by ES-XV. In contrast, whereas type XVIII endostatin and NC1 domains failed to inhibit VEGF-induced angiogenesis, endostatin, but not the type XVIII collagen NC1 domain, were active against FGF2-induced angiogenesis.

Mode of Angiogenesis Inhibition by ECM Fragments

Angiogenesis inhibitors are likely to act via distinct pathways primarily because of their diversity in structure and receptors used. For example, endostatin is proposed to interact with the a5b1 integrin that triggers a signaling cascade leading to inactivation of RhoA GTPase; this subsequently may cause disruption of the actin cytoskeleton and disassembly of focal adhesions. Thus, endostatin blocks the mobility of EC, thereby preventing the early stages of angio-genesis that require cells to migrate. This process requires an intact cytoskeletal organization. Cytoskeleton dynamics are of fundamental importance for the actions of another antiangiogenic factor derived from the C-terminus of per-lecan, endorepellin. Endorepellin is proposed to interact with another integrin, the a2b1 that is the major receptor for collagen I, and this may lead to an intracellular increase in c-AMP, activation of protein kinase A, and subsequent dissolution of actin stress fibers and focal adhesions. Notably, collagen I exerts the opposite effect, further indicating that the a2b1 integrin is directly involved in endorepellin-related signaling. Interestingly, tumstatin, a C-terminal fragment of collagen type IV, is believed to act differently by binding yet another integrin, avb3, thereby triggering a signaling cascade that leads to inhibition of EC proliferation. Therefore, a key question is whether common effector pathways are similarly modulated by the various antiangiogenic inhibitors. Endpoints for these pathways would presumably contain the most essential molecular targets for inhibition of angiogenesis. There is, thus, the need to investigate not only intracellular signaling pathways, but also the initial ECM/growth factor interactions with the surface receptors of the EC. In addition, we need to further understand the factors that alter or modulate the cell surface receptors that comprise the first lines of sensing of the angiogenesis environment by the EC.

Was this article helpful?

0 0
Essentials of Human Physiology

Essentials of Human Physiology

This ebook provides an introductory explanation of the workings of the human body, with an effort to draw connections between the body systems and explain their interdependencies. A framework for the book is homeostasis and how the body maintains balance within each system. This is intended as a first introduction to physiology for a college-level course.

Get My Free Ebook

Post a comment