Role de la synthese de l'acide retinoique dans le controle de la proliferation et de la differenciation des cellules epitheliales mammaires

Retinoic acid (RA) is ligand of nuclear receptors RARs and RXRs that act as ligand-inducible transcription factors and mediate its biological effects. It was shown that RA has antiproliferative and prodifferenciating properties in mammary cells. A loss of RA sensitivity was associated with increased tumorigenicity in the mammary tissue, potentially facilitating the growth of tumors. It's believed that is was mainly due to deficiencies in the RA signaling pathway, probably caused by the loss of RAR and RXR expressions. However, some tumorigenic cell lines were still reported to be RA sensitive.;Keywords: retinoid acid, RALDH, breast cancer, stem cell;The role of RA synthesis in mammary tumorigenesis has been poorly characterized. RA is synthezised in target tissues from vitamin A (retinol) its precursor in blood. It was shown that mammary epithelial cells were able to synthesize RA from retinol in vitro. We show here for the first time that RALDH3, an enzyme involved in RA synthesis, is probably responsible for RA synthesis in normal mammary epithelial cells. Our result suggest that luminal cancer cells (that express ERalpha and represent 75 % of breast tumors) have a very low capacity of RA synthesis, probably due to a low estrogen-mediated RALDH3 expression. It might represent a new mechanism of estrogen-driven tumorigenenesis allowing RA sensitive tumors to proliferate in the presence of retinol in the blood. It was suggested that RALDH1, an other enzyme of the RA synthesis pathway that shares 70 % of identity with RALDH3, is a marker of mammary stem cells, of more aggressive tumors and higher occurance of metastasis. We shown that unlinke RALDH1, RALDH3 is a marker of a lower occurance of metastasis and probably a marker of differentiation, suggesting different roles in the mammary gland for these 2 enzymes. This is in good agreement with our results showing that they have very different enzymatic properties. All together our data suggest that RALDH1 and 3 might be markers of different populations of cells of in the mammary epithelium. We propose to use the differences between RALDH1 and 3 to rationally develop methods and tools to separate and isolate RALDH1- and 3-expressing cells that would help the understand the role of RA synthesis in the mammary gland...