Growth factor pathways in human cancer : functional and therapeutic implications

Abstract: The multi-step development of tumors involves numerous changes at genomic level such as oncogene activation, loss of function of tumor suppressor genes and translocations resulting in fusion genes that encodes for chimeric proteins with tumorigenic functions etc. However, in the selection leading to cancer in somatic tissues it is likely that the cancer cells make use of the normal extracellular signaling for proliferation and/or antiapoptosis to create growth advantage over the normal cells. These signals are, in part, mediated by the growth factor receptors. This thesis aims to explore the mechanisms involved in expression and function of these receptors with special focus on insulin-like growth factor-1 receptor IGF-1R. The final goal is to identify some "Achilles' heel" in the growth factor pathways as a possible target in cancer therapy. N-linked glycosylation is crucial for expression of growth factor receptors at the cell surface. In Ewing's sarcoma cells, which carry the EWS-FLI-1 fusion gene, we found that inhibition of N-linked glycoproteins suppressed the EWS-FLI-1 protein leading to growth arrest. Since the fusion protein was demonstrated to not be a glycoprotein, we conclude that some other glycoproteins may be involved in regulation of EWS-FLI-1. Since growth factor receptors are N-linked glycoproteins and most N-linked glycoproteins are confined to the plasma membrane, the possibility of a link between cell surface expression of growth factor receptors and EWS-FLI-1 expression may be raised. We therefore tested different specific growth factor pathways regarding their potential influence on the EWS-FLI-1 protein. Our data indicate that the basic fibroblast growth factor (bFGF) pathway is important for up-regulation the EWS-FLI-1 protein. Other investigated growth factors pathways (e.g. IGF-1) seemed not to regulate the fusion protein. We investigated the functional impact of p53 for IGF-1R expression in malignant cells. Using three different system-(1) malignant melanoma cell lines expressing mutant p53, (2) malignant melanoma cell lines overexpressing wild type (wt) p53 and (3) BL41tsp53-2 cells (harboring a temperature-sensitive p53), we could demonstrate that induction of normal wt p53 or down regulation of the mutant type p53 impaired the IGF-1R expression. However, the melanoma cell lines expressing wt p53 also responded with decreased expression of IGF-1R upon p53 inhibition. We hypothesize that p53 may interfere with IGF-1R expression at posttranscriptional levels. Based on the aforementioned data we investigated the mechanism underlying the interaction between p53 and functional IGF-1R. Our data provides evidence that inhibition of p53 triggers Mdm-2- dependent ubiquitination and proteasomal dependent degradation of the lGF-1R. In fact we could demonstrate: a physical association of IGF-1R to Mdm-2; that inhibition of p53 expression, with maintained expression of Mdm-2, causes ubiquitination of IGF-1R; that co-inhibition of p53 and Mdm- 2 expression rescues the cells from IGF-1R down regulation and subsequent death; and that Mdm-2 ubiquitinates IGF-1R in cell-free systems. Finally, we identified potent and specific inhibitors of IGF-1R and demonstrated their potency in inhibition of malignant cell growth, both in vivo and in vitro.