Transcription factors involved in negative and positive gene regulation by glucocorticoids

Abstract: Steroid hormones, including glucocorticoids (GCs), control a wide variety of developmental and physiological responses in higher eukaryotic organisms. They pass through the plasma membrane and bind to specific receptor proteins inside the cell. Ligand binding promotes a conformational change, homodimerization and activation of the receptor to a DNA binding form. The glucocorticoid receptor (GR) translocates into the nucleus upon GC binding, where it stimulates or represses transcriptional activity of specific genes, either by binding to specific DNA sequences known as glucocorticoid response elements (GREs) or by interacting with other proteins. A key question is how the GR is able to both stimulate and repress transcription. To investigate some of these aspects, we studied (Papers I & II) a negative GRE (nGRE) from the bovine prolactin (PRL) gene promoter (PRL3), which mediates GC repression of the endogenous gene or a heterologous promoter. In the absence of GCs the PRL3nGRE confers an increased expression in both rat pituitary cells (GH3) and non-pituitary cells. We demonstrated that the Pit-1/GHF-1 from GH3 cells and the Oct-1 from non-pituitary cells were the main factors responsible for this enhanced activity. In addition, a second ubiquitously expressed protein, Pbx, was found to contribute to the augmented activity. We showed that binding of both the GR and the Pbx proteins to the PRO element were required for GC repression. Furthermore, only one GR moiety contacted the PRL3nGRE, which could explain the inability of GR to transactivate from this element. Finally, it was demonstrated that the GR displaced Pit-1/Pbx or Oct-1/Pbx binding to the PRL3nGRE element. As this displacement did not take place in the absence of Pbx, this protein has an important function in the mechanism of GC repression. In a subsequent study (Paper III), we examined whether M interfere with the transcriptional control of other promoter elements regulated by Pbx. We demonstrated that GCs potentiated RA-induced transcription from the Hoxb-1 gene promoter autoregulatory element (bl-ARE) recognized by Pbx1 and HOXB1 in P1 9 cells. The GR did not bind directly to the bl -ARE element. Instead we showed that the Pbxl/HOXB1 heterodimer was the target for the GC effect. Furthermore, we also demonstrated that the DNA-binding domain (DBD) but not the transactivating function of the GR was required for the synergism. GCs did not stimulate the expression of the Pbx1 or HOXB1 proteins in this cell line. In addition, RA was able to enhance GR-mediated transactivation via a GRE-controlled reporter gene in the same cell line. GR physically interacted with Pbx1 and HOXB1 proteins in vitro in GST-pull down assays, which may explain the mutual GC/RA synergy. In paper IV, we investigated the role of one of the recently discovered array of cofactors, RIP140, whose role as a coactivator has been questioned, in gene regulation by the GR. In our study this cofactor antagonized all GR mediated responses tested, including repression through a negative GRE, cross-talk with NF-[kappa]B (ReIA), activation through a classical GRE and the synergistic effects of glucocorticoids on AP-1 and Pbxl/HOXB1 responsive elements. The repressive activity of RIP140 required the ligand binding domain (LBD) of GR and did not occur when the GR was bound to the antagonist RU486. We also demonstrated a GR-RIP140 interaction in vitro by a GST-pull down assay. Overexpression of a coactivator, TIF-2, was able to partly restore the GR-dependent transactivation, suggesting that RIP140 may compete with other coactivators for interaction with the GR.