Bone sialoprotein and the paradox of angiogenesis versus atherosclerosis.

Bone sialoprotein (BSP) is a protein thought to be highly specific for bone. BSP contains an arginine-glycineaspartic acid (RGD) cell attachment sequence involved in osteoclast adhesion to bone matrix via the vitronectin receptor and plays an important role in the early process of bone mineralization and resorption. The study by Bellahcène et al1 in this issue of Circulation Research indicates that BSP mediates adhesion and chemotactic migration of endothelial cells and promotes angiogenesis, suggesting that BSP may be an important factor in angiogenesis and the initiation of atherosclerosis, two processes that are probably related. In earlier studies, Folkman2 hypothesized that tumor growth beyond a few millimeters requires recruitment and growth of a new microcirculation, or angiogenesis, which is induced by tumors as lifelines for oxygen and nutrients. New blood vessels also provide exits for cancer cells to spread to other parts of the body. Angiogenesis is also involved in physiological conditions, such as embryogenesis, and other pathological conditions, such as wound healing. The process of angiogenesis requires a highly coordinated series of events, including endothelial cell proliferation, migration, tube and lumen formation, and, in some cases, recruitment of smooth muscle cells (SMCs) and other adventitial cells. Adhesive interaction of cells with components of the extracellular matrix is a recognized requirement for cell proliferation and migration. Evidence indicates that many of the adhesive interactions are mediated by members of the integrin family of heterodimeric adhesion receptors. Among these integrins, avb3, which is expressed by a variety of cell types, has been shown to play a key role in the cell migration involved in metastasis and angiogenesis.3 The work presented by Bellahcène et al1 indicates that BSP, a bone-associated protein that contains the RGD sequence, a common recognition sequence for most integrins, binds avb3 in endothelial cells and mediates the migration of such cells, extending the study by Byzova et al.4 Furthermore, Bellahcène et al1 show that another integrin, avb5, does not bind BSP, suggesting that there is a certain degree of specificity for the interaction between BSP and avb3. The in vivo data indicating that BSP promotes angiogenesis via its interaction with avb3 integrin and that such an angiogenic effect is probably even greater than that of basic fibroblast growth factor (bFGF)1 are intriguing. These data underscore the physiological and pathophysiological consequences of the interaction between BSP and avb3 integrin and define BSP as a novel angiogenic factor. These findings may provide an explanation for the previously established association of BSP expression levels in tumors with the development of bone metastases.5 Higher BSP expression in the tumor correlates with an increased risk of metastasis of carcinomas to bone tissue, which could be due to angiogenesis enhancement by BSP. In addition, BSP expression in the tumor may facilitate tumor cell migration and calcification. Several families of factors have been implicated in angiogenesis. These include angiogenic factors, such as vascular endothelial growth factor (VEGF) or bFGF, and antiangiogenic factors, such as angiostatin or endostatin.6 Translational research has now been initiated by several centers to test the hypothesis that local delivery of angiogenic agents, especially VEGF and bFGF, by various strategies, including viral vectors, naked DNA, or purified recombinant proteins, may improve blood flow to ischemic tissues in patients with advanced atherosclerosis.7 Similarly, clinical trials for local delivery of antiangiogenic factors, such as angiostatin, using various techniques, are underway to treat patients with malignant tumors.8 Although blockade of BSP interaction with avb3 may limit tumor growth and metastasis, the effect of local delivery of BSP to improve blood flow to ischemic tissues, in particular the myocardium, may be complicated by the fact that BSP-avb3 interaction may also initiate and aggravate atherosclerosis. Neointima formation, associated with both atherosclerosis and restenosis, is a complex process that involves SMC migration and proliferation. The molecular mechanisms governing intimal thickening have been a focus of intense research. Numerous studies have indicated a critical role for integrin heterodimers, including avb3, in mediating cell-matrix interactions and SMC adhesion and migration, implicating the potential involvement of physiological ligands containing RGD sequence, including BSP, for the activity of such integrins.3 Indeed, in vitro studies have shown that osteopontin, a bone-associated protein closely related to BSP, which contains the RGD sequence, supports SMC adhesion to avb3, avb5, and avb1 integrins and mediates SMC migration specifically via avb3. In vivo experiments have demonstrated that osteopontin and avb3 are expressed in vascular SMCs. In addition, osteopontin expression is upregulated in human atherosclerotic and restenotic lesions.10 Using animal models of neointima formation, several groups have observed increased expression of osteopontin and avb3 in such lesions.11 More direct evidence supporting the potential The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From the Heart and Lung Institute and Division of Cardiology, Department of Internal Medicine, College of Medicine and Public Health, Ohio State University, Columbus, Ohio. Correspondence to Pascal J. Goldschmidt-Clermont, 514 Medical Research Facility, 420 W 12th Ave, Columbus, OH 43210. E-mail [email protected] (Circ Res. 2000;86:827-828.) © 2000 American Heart Association, Inc.