DISSECTION OF THE ROLE OF NG2/CSPG4 PROTEOGLYCAN IN TUMOR PROGRESSION

To accomplish the metastatic process, disseminating tumor cells enter haematic and lymphatic conduits to reach distant sites where they egress the circulation by penetrating the vessel wall – a phenomenon denoted extravasation and thought to involve complex interactions between the tumor cells, the vascular cells and their associated ECM. More recent experimental data suggest that pericytes may play a particularly critical role in this process, while a plephora of cell surface components present on all the interacting cell types are believed to act as promoters of transvascular passage. Cell surface proteoglycans (PGs) are among the components believed to be central cell membrane- and ECM-associated factors in this context, although it is not fully understood how they operate. The transmembrane proteoglycan NG2/CSPG4 is widely documented to be a tumor-promoting agent capable of driving tumor spread through the promotion of intricate microenvironmental interactions and is thereby is a prime candidate for the regulation of the cellular and molecular interactions underpinning the intra- and extravasation processes. The present thesis work has therefore approached the putative role of NG2/CSPG4 in the control of tumor spread, with specific reference to its potential ability to mediate the cancer cells’ interaction with vascular structures. As starting point, the effectiveness of NG2/CSPG4 in promoting tumor growth in vitro and in vivo (xenogenic setting in athymic mice) was comparatively and quantitatively evaluated using a larger panel of melanoma, sarcoma and carcinoma cell lines with diverse, constitutive expression of NG2 and with immunosorted NG2/CSPG4-positive (representing putative cancer initiating cells) and NG2/CSPG4-negative subsets of some of these cancer cell lines. These experiments firmly corroborated the pro-tumorigenic role of NG2/CSPG4 and further highlighted its putative role in the control of cancer cell-host microenvironmental interplays. Global gene profiling evidenced marked gene expression differences between the two cell phenotypes, which are currently under investigation. Parallel antibody array-based phospho-proteomic analyses revealed NG2/CSPG4 phenotype-specific differences in the phosphorylation status of key molecules of the PI-3K and Rho/Rho-GAP pathways. The above cell types were then assayed under both static and shear-dependent conditions for their ability to interact with vascular cells and their matrices. To this end, a novel method for the in vitro derivation of native cell-free ECMs was developed with the aim to obtain matrices reproducing in vitro the in vivo exhibited compositional and supramolecular configuration, as defined by in vivo metabolic labelling, immunochemistry and AFM and eSEM analyses. NG2/CSPG4-expressing cells responded to differently to matrices derived from immature perivascular cells and in particular motility responses further differed between NG/CSPG4+ and NG2/CSPG- cell subsets. Involvement of the proteoglycan in cell adhesion on different substrates was confirmed through the use of propriety functional-blocking anti-NG2/CSPG4 antibodies and by RNAi-mediated knockdown of the molecule. The potential involvement of NG2/CSPG4 in the process of extravasation was addressed by “static” transmigration assays involving monolayers of endothelial cells and smooth muscle cells and by experiments under flow. Both the core protein of the PG and its glycan chains were found to be involved in these cellular interactions. To finally address the NG2/CSPG4 function in the interaction between cancer cells and the luminal blood vessel wall and the impact of the proteoglycan in the passage of cancer cells through the vasculature, assays involving the chick embryo chorionallontoic membrane were exploited in combination with fluorescence real-time imaging. Again a substantial contribution of the proteoglycan in this process could be observed. The findings of the study contribute to our understanding of the pro-tumorigenic role of NG2/CSPG4 and lend support to the idea that the proteoglycan could play an active role in the control of the metastatic process. The observations therefore reinforce the value of NG2/CSPG4 as a therapeutic target and encourage a more detailed pre-clinical evaluation of the anti-NG2/CSPG4 examined in this study as therapeutic agents.