Thesis: Fibrosis is identified as a chronic side effect occurring after radiotherapy for pelvic tumors in 5 to 10 % of patients. This pathological healing process is characterized by an accumulation of extracellular matrix synthesized by mesenchymal cells. Endothelial to mesenchymal transition (EndoMT), is a processes during which endothelial cells express mesenchymal markers in response to stress. EndoMT is identified as a source of mesenchymal cells taking part to fibrosis development in patients suffering from inflammatory bowel diseases. Then, this study focused on the potential participation of EndoMT in radiation-induced intestinal fibrosis and tried to identify new therapeutics targets. Interestingly, our results showed for the first time EndoMT in rectal tissues from patients who developed radiation proctitis following radiotherapy. We used an in vivo approach to follow the mesenchymal cells having an endothelial origin in a mouse model expressing the GFP under the control of an endothelial promoter, Tie2 (Tie2-GFP). Thereby, our results confirmed the existence of radiation-induced EndoMT in our preclinical model of radiation proctitis. In vitro characterization showed that irradiation induced a modulation of the endothelial phenotype through a mesenchymal profile, a hallmark of EndoMT. This project also focused on a potential molecular actor, Hey2. In this context, we generated a transgenic mouse model in which Hey2 gene expression is repressed specifically in the endothelial compartment and observed a decrease in radiation-induced mucosal damages and EndoMT frequency. Consequently, inhibiting Hey2 expression could represent a new interesting therapeutic strategy.