Ligand and coactivator identity determines the requirement of the charge clamp for coactivation of the peroxisome proliferator-activated receptor gamma

The activation function 2 (AF-2)-dependent recruitment of coactivator is essential for gene activation by nuclear receptors. We show that the peroxisome proliferator-activated receptor gamma (PPARgamma) (NR1C3) coactivator-1 (PGC-1) requires both the intact AF-2 domain of PPARgamma and the LXXLL domain of PGC-1 for ligand-dependent and ligand-independent interaction and coactivation. Although the AF-2 domain of PPARgamma is absolutely required for PGC-1-mediated coactivation, this coactivator displayed a unique lack of requirement for the charge clamp of the ligand-binding domain of the receptor that is thought to be essential for LXXLL motif recognition. The mutation of a single serine residue adjacent to the core LXXLL motif of PGC-1 led to restoration of the typical charge clamp requirement. Thus, the unique structural features of the PGC-1 LXXLL motif appear to mediate an atypical mode of interaction with PPARgamma. Unexpectedly, we discovered that various ligands display variability in terms of their requirement for the charge clamp of PPARgamma for coactivation by PGC-1. This ligand-selective variable requirement for the charge clamp was coactivator-specific. Thus, distinct structural determinants, which may be unique for a particular ligand, are utilized by the receptor to recognize the coactivator. Our data suggest that even subtle differences in ligand structure are perceived by the receptor and translated into a unique display of the coactivator-binding surface of the ligand-binding domain, allowing for differential recognition of coactivators that may underlie distinct pharmacological profiles observed for ligands of a particular nuclear receptor.     

Characterization of transactivational property and coactivator mediation of rat mineralocorticoid receptor activation function-1 (AF-1)

The autonomous activation function-2 (AF-2) in the mineralocorticoid receptor (MR) E/F domain is known to play a major role in the ligand-induced transactivation function of MR; however, it remained unclear about the transactivation function of its A/B domain. We therefore tried to characterize the MR A/B domain as the AF-1 and further studied the actions of known coactivators for AF-2 in the E/F ligand-binding domain in the function of the MR A/B domain. Deletion analyses of rat and human MRs revealed that the A/B domains harbor a transactivation function acting as AF-1. The MR mutant (E959Q) with a point mutation in helix 12, which causes a complete loss of MR AF-2 activity, still retained ligand-induced transactivation function, indicating a significant role for AF-1 in the full activity of the ligand-induced MR function. Among the coactivators tested to potentiate the MR AF-2, TIF2 and p300 potentiated the MR AF-1 through two different core regions [amino acids (a.a.) 1-169, a.a. 451-603] and exhibited functional interactions with the MR A/B domain in the cultured cells. However, such interactions were undetectable in a yeast and in an in vitro glutathione-S-transferase pull-down assay, indicating that the functional interaction of TIF2 and p300 with the MR A/B domain to support the MR AF-1 activity require some unknown nuclear factor(s) or a proper modification of the A/B domain in the cells.     

An LXXLL motif in nuclear receptor corepressor mediates ligand-induced repression of the thyroid stimulating hormone-beta gene

Nuclear receptor corepressor (N-CoR) regulates gene expression through interaction with DNA-bound nuclear receptors, recruiting multicomponent repressor complexes to the sites of target genes. We recently reported the presence of an LXXLL motif in N-CoR, and showed that this motif interacts in vitro and in vivo with retinoic acid receptor alpha (RARalpha) and thyroid hormone receptor beta (TRbeta). Transient transfection experiments now suggest that TRbeta and N-CoR act synergistically and may both be required for ligand-induced repression from the negative TR response element in the thyroid stimulating hormone-beta (TSHbeta) gene promoter. Mutation of the LXXLL motif in N-CoR abolished ligand-induced repression at this response element. Furthermore, in vitro binding of N-CoR to a complex between TRbeta and the negative TR response element was strictly ligand-dependent. We conclude that N-CoR and TRbeta cooperate in the regulation of the TSHbeta gene and that the ligand-dependent repression is mediated by the LXXLL motif in N-CoR.     

Ligand-induced recruitment of a histone deacetylase in the negative-feedback regulation of the thyrotropin beta gene.

We have investigated ligand-dependent negative regulation of the thyroid-stimulating hormone beta (TSHbeta) gene. Thyroid hormone (T3) markedly repressed activity of the TSHbeta promoter that had been stably integrated into GH(3 )pituitary cells, through the conserved negative regulatory element (NRE) in the promoter. By DNA affinity binding assay, we show that the NRE constitutively binds to the histone deacetylase 1 (HDAC1) present in GH(3 )cells. Significantly, upon addition of T3, the NRE further recruited the thyroid hormone receptor (TRbeta) and another deacetylase, HDAC2. This recruitment coincided with an alteration of in vivo chromatin structure, as revealed by changes in restriction site accessibility. Supporting the direct interaction between TR and HDAC, in vitro assays showed that TR, through its DNA binding domain, strongly bound to HDAC2. Consistent with the role for HDACs in negative regulation, an inhibitor of the enzymes, trichostatin A, attenuated T3-dependent promoter repression. We suggest that ligand-dependent histone deacetylase recruitment is a mechanism of the negative-feedback regulation, a critical function of the pituitary-thyroid axis.