- Background:
During tumor progression, beta1 integrin adhesion molecules play a major role in the acquisition of invasive capacity. We have recently investigated beta1 function in the ESb lymphoma model. Targeted disruption of both beta1 genes greatly reduced dissemination of these cells, which was restored upon transfection of intact beta1A. Cytoplasmic domain beta1 mutants that cannot bind to the beta1-associated protein ICAP-1, did not restore dissemination. To further explore the role of ICAP-1, we used yeast two-hybrid screening and identified three proteins that bind to it: the cytoskeletal protein vimentin, the antioxidant protein Aop1/MER5, and ROCK kinase which is activated by the RhoA small GTPase, a regulator of the actin cytoskeleton. This set of proteins is remarkable since ROCK was previously shown to bind to vimentin, and MER5 and vimentin were reported to be co-immune precipitated in a complex that also contains nm23-H1, a metastasis and motility suppressor. It is noteworthy that vimentin is not expressed in normal epithelial cells in the body, but is induced in epithelial cells in culture, in wounds and in carcinomas, and that transfection and knockout studies point to a role in cell migration. Activated ROCK phosphorylates vimentin, causing depolymerization of vimentin filaments. Vimentin oligomers, but not filaments, were recently shown to be linked by fimbrin to actin, and to be localized in filopodia. These oligomers can be observed as vimentin "dots" that rapidly move along microtubules to the leading edge of spreading cells. We found that a ROCK inhibitor impeded the invasion of lymphoma cells into a monolayer. Instead, the cells adhered to, and remained attached to, the dorsal surface of the monolayer. Finally, transfection of the identified small ICAP-1-binding domain of ROCK into 3T3 fibroblasts greatly reduced cell spreading on fibronectin.
Together these data indicate that beta1 can associate with vimentin in complexes with ROCK, MER5 and nm23, and suggest that such complexes are involved in the regulation of beta1 integrin function during spreading and migration of cells on beta1 ligands. A possible role for ROCK is to promote dissociation of proteins from this complex, upon activation by RhoA. PLIC proteins were shown to connect beta3 integrins with vimentin (via IAP/CD47) and to play a role in spreading, and we have preliminary results suggesting the existence of a distinct protein with similar function for beta2 integrins. Vimentin-integrin complexes with associated proteins may thus be of general importance for cell migration and tumor invasion.
- Aim:
The aim of this project is to investigate the potential role of this putative multiprotein complex in the migration and invasion of cells, in particular of highly invasive tumor cells. We will study the protein interactions using the yeast two-hybrid system and with recombinant proteins, and assess the (co)localization of these proteins, particularly in spreading and migrating cells. Based on fine mapping of the interaction sites, we will generate mutants and fragments that might act as dominant-negative inhibitors. To assess their effects on cell spreading, migration, invasion and metastasis, these putative inhibitors will be transduced into cells using highly efficient retroviral systems, and cells with high expression levels will be selected by FACS sorting, based on co-expressed GFP. (Co-)localization in living cells during spreading and migration will be determined using GFP-tagged proteins and confocal microscopy.
This project is expected to reveal essential mechanisms for migration of cells, relevant for invasion and metastasis of tumor cells. |
