Extracellular matrix (ECM) assembly is a critical aspect of new blood vessel formation in tumors. Hence identification of factors that control the interaction of ECM components with cells in the tumor microenvironment is of paramount importance for furthering our understanding of tumor angiogenesis. SPARC is a matricellular protein that participates in ECM remodeling and angiogenesis and is expressed prominently by stromal cells in many types of cancer. By studying the response of wild-type (WT) and SPARC-null animals to various challenges, including tumor implantation we have identified SPARC as a critical factor in assembly of ECM. Due in part to defects in ECM deposition, pancreatic tumors grown in SPARC-null mice are 3-fold larger and have a 3-fold increase in metastases in comparison to tumors grown in WT mice. We have established that tumors grown in the absence of SPARC show a 20% reduction in microvessel density and a 50% reduction in pericyte recruitment to blood vessels, as well as a substantial decrease in the level of collagen-associated decorin, a key factor in controlling TGF-? bioavailability. We predict that increased tumor burden in SPARC-null mice is supported by fewer blood vessels due to increases in vascular permeability and in perfusion. Preliminary studies also indicate that greater amounts of collagen associate with cell membranes after secretion from SPARC-null cells than WT counterparts. Based on these observations, we hypothesize that a primary function of SPARC is to control collagen-cell interaction and that increased ligation of collagen-binding integrins (?1?1 and ?2?1) on endothelial cells by collagens I and IV in the absence of SPARC results in 1) decreased ECM deposition accompanied by irregular vascular basement membrane formation, 2) inhibition of ?v?3 activation by a mechanism termed transdominant integrin inhibition, and 3) an increase in soluble TGF-?, which results in reduced pericyte recruitment. We will use an orthotopic pancreatic tumor model in WT and SPARC-null animals, as well as in vitro methods, to evaluate our hypothesis through three specific aims: (1) We will determine if SPARC decreases collagen I and/or IV binding to endothelial cell integrins and whether this results in inhibition of ?v?3 activation. (2) We will delineate specific differences in vascular basement membrane assembly and function resulting from the absence of SPARC. (3) We will determine whether increased soluble TGF-? is responsible for decreased pericyte recruitment in the absence of SPARC. Integrin signaling is a pivotal control point for angiogenesis and therefore characterization of the function of SPARC in the regulation of collagen interaction with integrins is essential for our understanding of the angiogenic process and the refinement of anti-integrin strategies for cancer therapy.
Showing the most recent 10 out of 22 publications
