Background Brain metastases are a ultimate oncologic challenge, and are assuming an increasing relevance since recent therapies improved patient outcomes, so that tumor intracranial localization is frequently a final step of their neoplastic progression. To date, pathophysiology of brain metastases remains elusive and progress in treatment has been hampered by various criticisms, including lack of human tissue samples or effective model systems. However, some molecular pathways specific to the development of brain metastases have been suggested, and mechanisms activating or selecting �brain �seeking� clones could bear relevance in the sequence of events leading to a cerebral secondary localization, as well as a newly acquired drug- and radio-resistance of neoplastic cells. However, the �brain sanctuary� could be violated not only by �seed� characteristics, but also by a �soil� facilitation, and brain cellular components could participate in mediating the �brain homing� process. In recent years, our group focused on Caveolin-1 (cav-1), a protein constituent of the membrane structures �caveolae�, whose role in cancer is still debated, since it may be coupled with various intracellular partners, including molecules involved in tumor cells invasiveness/migration or in multidrug resistance (MDR). We previously reported that in various subtypes of primary glial brain tumors cav-1 expression correlates to tumor aggressiveness and poor outcome, as well as that, in a limited series of lung carcinomas, cav-1 expression is increased in undifferentiated lesion and expressed ex novo in brain metastases.
Aims Aim of the project is to verify if cav-1 expression can be modulated at protein or gene level in a large series of primary human lung and breast carcinomas and melanomas, according to their metastatic potential and to the expression of drug and radio resistance related molecules. In addition, we aim to verify if brain metastases differ from their primary tumors or from extra-brain metastatic sites in cav-1 expression, suggesting the implication of a Cav-1 role in the �seed-soil� crosstalk involved in the brain-specific homing of neoplastic cells. These hypothesis will be tested by in vitro and in vivo studies using cav-1 silenced cells.
Materials and methods � Tissue studies: By immunohistochemistry, cav-1 expression will be tested in a series of 250 non metastatic or extra-brain metastatic lung and breast carcinomas and melanomas as well as in 100 lung carcinomas, 30 breast carcinomas and 15 melanomas with a matched brain metastasis. Also, its co-expression with the alpha1 subunit of Na/K ATPase, involved in MDR, and with YKL-40, involved in radio-resistance mechanisms in primary brain tumors and various carcinomas, will be studied. In the same cases, both cav-1 gene mutations (by direct sequencing) and cav-1 gene amplification (by fluorescence in situ hybridization) will be studied. � In vitro experiments: prior and after cav-1 silencing lung, breast carcinoma and melanoma cell lines invasiveness and migration will be tested in vitro � In vivo experiments: in nude mice, metastatic potential and intracerebral growth of cav-1+ and cav-1/silenced selected carcinoma cell lines will be tested after peripheral and intracranial injection.
Potential clinical relevance Potential clinical relevance of the project derives from the possible identification of a pheno- and geno-typical signature of primary tumors at high risk to develop brain metastases, which will therefore require a more aggressive therapeutic approach, ie prophylactic brain irradiation, or a specific molecular targeted therapy.