Determination of nuclear receptor corepressor interactions with the thyroid hormone receptor

The thyroid hormone receptor (TR) recruits the nuclear corepressors, nuclear receptor corepressor (NCoR) and silencing mediator of retinoid and thyroid hormone receptors (SMRT), to target DNA elements in the absence of ligand. While the TR preferentially recruits NCoR, the mechanism remains unclear. The corepressors interact with the TR via interacting domains (IDs) present in their C terminus which contain a conserved motif termed a CoRNR box. Despite their similarity, the corepressor IDs allow for nuclear receptor specificity. Here we demonstrate that NCoR stabilizes the TR homodimer when bound to DNA by preventing its dissociation from thyroid hormone response elements. This suggests that NCoR acts to hold the repression complex in place on target elements. The TR homodimer recruits NCoR through two of its three IDs, one of which is not present in SMRT. This unique ID, N3, contains a CoRNR box but lacks the extended helical motif present in each of the other IDs. Instead, N3 contains an isoleucine just proximal to this motif. This isoleucine is also conserved in N2 but not in the corresponding S2 domain in SMRT. On thyroid hormone response elements and in mammalian cells this residue is critical in both N3 and N2 for high-affinity TR binding. In addition, this residue also controls specificity for the interactions of TR with NCoR. Together these data suggest that the specific recruitment of NCoR by the TR through a unique motif allows for stabilization of the repression complex on target elements.     

Thyroid hormone alters the DNA binding properties of chicken thyroid hormone receptors alpha and beta.

The effects of thyroid hormone agonists on thyroid hormone receptor (TR)/DNA complex formation was investigated to elucidate the mechanism by which TRs transactivate genes in response to ligand. The data, obtained from gel shift experiments, indicate that thyroid hormones alter the conformation of TRs bound to DNA, irrespective of if the element is occupied by monomeric TR, homodimeric TR/TR, or heterodimeric complexes with the retinoid receptors RAR or RXR. Furthermore, triiodo-thyronine (T3) prevents 2 TR molecules from binding to oligonucleotides containing direct repeats or inverted palindromes of the consensus AGGTCA motif, an effect that was not detected with palindromic elements. Heterodimers bound to direct repeats were less affected: RXR/TR were fully and RAR/TR complexes partially resistant to thyroid hormone. The data suggest that a ligand-induced conformational change in TR prevents double TR occupancy of a response element containing 2 direct repeats of the consensus binding motif, possibly by steric hindrance, whereas such an event does not prevent TR/RXR heterodimers from binding to DNA. Finally, our data show that a monomeric, liganded TR bound preferentially to the second half site in a AGGTCActcaAGGTCA element, and therefore indicate that nucleotides adjacent to the consensus half site contribute to binding specificity.     

3,5,3'-triiodothyronine receptor auxiliary protein (TRAP) enhances receptor binding by interactions within the thyroid hormone response element.

We have previously demonstrated that binding of in vitro synthesized thyroid hormone receptor (TR) to thyroid hormone response elements (TREs) is enhanced by the addition of nuclear extracts from several different cell types, suggesting that binding of TR is partially dependent on a T3 receptor auxiliary protein (TRAP). We have used the avidin-biotin complex DNA-binding assay to discriminate between regions of TREs that bind TR alone and sites that are influenced by interactions with TRAP. Mutations in the TREs from rat GH and glycoprotein hormone alpha-subunit genes show that a specific DNA sequence is required for TRAP-mediated enhancement of TR binding. Mutations in the B half-site of the rat GH TRE or in similar sequences [(T/A)GGGA] in the alpha-subunit TRE ablate the enhancement of TR binding by TRAP. Furthermore, binding of TR to a natural half-site in the TSH beta-subunit gene (bases -16 to 6), which lacks an additional AGGGA-like sequence, is not enhanced by the addition of TRAP. Binding of TR to TREs was also tested at physiological salt concentrations in the avidin-biotin complex DNA-binding assay. Binding of human TR beta to TREs decreases dramatically at 140 mM KCl compared to binding at 50 mM KCl; however, the addition of TRAP enhances the binding to almost 4-fold of basal binding, suggesting that TRAP may be important for stabilization of TR binding to TREs in the cell.