Président du jury Vassart, Gilbert
Promoteur Fuks, François
Publication Non publié, 2011-02-25
| Résumé : | It is increasingly recognized by the scientific community that the field of epigenetics is a key step for a better understanding of human biology in both normal and pathological states. Its implication in cancer, and in particular in breast cancer, is now well accepted. Breast cancer, responsible for more than 450,000 deaths worldwide yearly, is a heterogenous disease at the histological and clinical levels as well as at the molecular level. Despite considerable efforts to develop new treatments and improve patient management, patients with a same “profile” of breast cancer can respond differently to therapies and have completely different clinical outcomes. There is therefore a critical need to improve our understanding of breast cancer biology and diversity, in order to find new markers that should provide a better management of patients and the development of new therapies. An increasing number of biologists, pathologists as well as clinicians are currently working towards these goals. During my PhD, we have conducted two studies in order to gain new insights into the contribution of epigenetics in breast cancer biology. In the first study, by performing large genome-scale DNA methylation profiling of numerous breast tumors as well as of normal breast tissues, we first revealed the existence of six groups of breast tumors based on their DNA methylation profiles. Three of these groups showed a strong association with the basal-like, HER2 and luminal A breast cancer subtypes, previously identified by gene expression profiling. Interestingly, the three other groups were found to be a mixture of several gene expression-based subtypes, thus revealing the capacity of DNA methylation profiling to improve breast tumor taxonomy. Second, our study suggests that the establishment of DNA methylation patterns of breast tumors might help to determine their cell type of origin. Finally, we also showed that DNA methylation profiling can reflect the cell type composition of the tumor microenvironment and that a signature of T cell tumoral infiltration is associated with a good prognosis in particular categories of breast cancer patients. In the second study, we revealed the clinical relevance of the KDM5 histone demethylases in breast cancer. The expression of these histone demethylases was deregulated in the analyzed breast tumors as well as in the pre-invasive samples as compared to normal breast samples. This suggests that KDM5 enzymes might be good markers for early diagnosis of breast cancer. Moreover, we showed a prognostic value of the KDM5C histone demethylase. In conclusion, the above data should provide a better understanding of breast cancer biology and diversity, and this should bring new insights to improve breast cancer patient management. |
