Impact of tumor-derived fibroblast growth factors on tumor-associated endothelial cells

Colorectal cancer (CRC) cells over-express fibroblast growth factors (FGFs) that stimulate tumor growth and metastasis by autocrine and paracrine mechanisms. Specifically, they may stimulate tumor neoangiogenesis, thus interfering with therapeutic efforts that target angiogenesis specifically the vascular endothelial growth factor VEGF. If this is the case, targeting both VEGF and FGF-signaling may have synergistic effects and avoid development of resistance against VEGF-blocking therapy. An in depth analysis of the biological impact tumor- derived FGFs exert on tumor angiogenesis is essential to provide the scientific basis for such therapeutic strategies. It is therefore the objective of this project to investigate the impact of tumor-derived FGFs on the tumor- asscociated endothelial cells (TEC). For this purpose, the relationship between tumor-derived FGFs, neoangiogenesis and response to anti-angiogenic therapy in human CRC will be determined by morphometric analysis of paraffin-embedded tissue sections and by quantification of FGF serum -levels in patients receiving VEGF- blocking therapy. In addition, TEC will be isolated from CRC tissue and characterized as to their endothelial cell markers. They will be separated into blood- and lymph-endothelial cells based on their cell surface protein pattern and their capacity to grow, migrate and form vessels will be analyzed using established biological assays. Human tumor cell lines stably over-expressing FGFs will be constructed and their impact on endothelial cell biology will be studied. For this purpose combined tumor cell-TEC spheroids in vitro and xenotransplanted tumors in immune compromised mice will be used as 3-dimensional models that mimic angiogenesis in human CRCs. Finally, the impact of FGF-pathway inhibition will be studied in xenotransplanted tumors. The expected results will contribute to the design of combination therapies including blockade of FGFR-signaling that have the potential to improve therapy response and prognosis.

Malignant tumors consist of both cancer cells and stromal constituents recruited from the host. Tumor cells produce and secrete factors that attract and activate stromal cells. In turn, the activated environmental cells contribute to tumor progression and aggressiveness. In colorectal cancer transforming growth factor ß (TGFß) is high as well as fibroblast growth factors (FGF) 8 and 18. The project investigated how these factors impact on the tumor stroma and the resulting interactions between tumor and stroma. We have analysed human tumor tissues to identify proteins altered in abundance in the tumor and find correlations to tumor stage, vascularisation and prognosis. In addition we have used cell line models of cancer cells, endothelial cells and fibroblasts to investigate cellular interactions that may drive tumor development on the molecular level. Our results show that FGF 18 is capable of activating colon-associated endothelial cells to increase vessel formation. In addition, angiogenesis-related proteins were upregulated together with other TGFß targets in the colorectal cancer tissues - including thrombospondin B2 and endoglin. Synthesis of secreted protein acidic and cysteine-rich (SPARC) is increased in cancer-associated fibroblasts. SPARC is an extracellular matrix constituent that modulates the interactions between cells and matrix material. Our data demonstrate that increased levels of SPARC stimulated secretion of angiogenic factors by the fibroblasts and increased tumor cell migration and invasion. Patients, whose tumors expressed the highest levels of SPARC, had a significantly worse prognosis than other colorectal cancer patientss. Additional candidate markers deriving from the study are still being evaluated.