Interaction between FGFR-2, Progesterone Receptor and STAT5 in breast cancer models

CERLIANI JUAN PABLO, GUILLARDOY TOMAS AND LANARI CLAUDIA

Washington

Congreso; AACR; 2010

We have developed a murine model of breast cancer in which hormone-dependent (HD) carcinomas transit through different stages of hormone responsiveness retaining the expression of estrogen and progesterone receptors (PR). Unlike HD carcinomas, hormone-independent (HI) tumors do not need the exogenous administration of medroxyprogesterone acetate (MPA) to grow. However, when tumor cells are cultured, there are no differences in hormone responsiveness between both types, suggesting the involvement of host factors regulating in vivo tumor growth. We have previously reported that carcinoma associated fibroblasts (CAF) stimulate HI cells inducing PR activation and that FGFR-2 plays a key role mediating this effect. In addition, MPA and FGF2 induce a direct interaction between FGFR-2 and PR in the nuclei of C4-HI and T47D cells. We have postulated that the stroma of HI tumors regulates HI tumor growth by releasing FGF-2, that acting as a paracrine growth factor, can activate PR in epithelial cells. It is also known that STAT5 can be activated either by FGF2 or by progesterone. Thus, the main goal of this study was to study the mechanisms by which FGF2 stimulates HI growth evaluating the interaction between PR, FGFR-2 and STAT5 using our murine tumor model and the human T47D breast cancer cells. C4-HI primary cultures or T47D cells were grown in chamber slides, starved using 1% charcoalized fetal calf serum for 24 hours and treated with FGF-2 (50 ng/ml), MPA (10 nM), FGF1 (20 ng/ml) or vehicle for 24 hs. Confocal imaging suggested an intimate interaction between STAT5 and PR when the cells were treated with FGF-2 or MPA rather than with FGF1. The interaction of both proteins in response to MPA or FGF2 was confirmed by co-immunoprecipitation assays using C4-HI tumor samples or T47D cells. Next, we performed gene transcription reporter assays in both, C4-HI and T47D cells using PRE or GAS-Luc vectors. Interestingly, in C4-HI cells, FGF2 was able to activate PRE and GAS reporter genes (p<0.05) while progestins activated mainly the PRE reporter gene. In T47D cells, however, MPA and FGF-2 activated both reporter genes (p<0.05). These results corroborate the crosstalk between both pathways. We also studied the expression of proteins coded from endogenous genes which have PRE or GAS sequences in their promoters such as Bcl-xl, CyclinD1, Tissue Factor and STAT5 itself. The expression of all these proteins was similarly increased after treatment with FGF2 or MPA (p<0.05) supporting the previous in vitro data. In summary, these results show for the first time that PR interact with FGFR-2 and with STAT5 in the nuclei of MPA- or FGF2-treated cells, and this is followed with the expression of target genes. These data support our hypothesis where FGF2 released by the tumor stroma activates FGFR-2 in the epithelial cells inducing PR and STAT5 activation, providing a possible mechanism for the acquisition of hormone independency.