The thyroid hormone receptor variant α2 is a weak antagonist because it is deficient in interactions with nuclear receptor corepressors

The thyroid hormone receptor splice variant, alpha 2, is unable to bind thyroid hormone (T-3) and has been proposed to function as an endogenous inhibitor of T-3 action. In this report, we examined further the DNA sequence requirements for alpha 2 binding to thyroid hormone response elements (TREs) in an attempt to identify response elements that mediate potent inhibition by alpha 2. Heterodimers of alpha 2 and retinoid X receptor were found to bind to a subset of TREs (DR4, direct repeats spaced by 4 bp) in which selected flanking and spacer sequences enhanced interactions with the AGGTCA core binding sequence. Despite the optimization of the TRE-binding sites, alpha 2 remained a weak dominant negative inhibitor of TRE-driven transcription. A promoter interference assay was also developed for testing inhibition by alpha 2. In these studies, alpha 2 blocked gene transcription, but it required cotransfected retinoid X receptor, and it was not as potent as unliganded thyroid hormone receptors. These results led to the hypothesis that alpha 2 might be deficient in interactions with nuclear receptor corepressors. Consistent with this view, alpha 2 did not silence basal transcription in its native form or when linked to Gal4. alpha 2 also failed to interact with corepressors (NCoR and SMRT) in both gel shift assays and mammalian two-hybrid assays. We conclude that alpha 2 is a weak antagonist of thyroid hormone action because it binds weakly to a limited repertoire of response elements, and it does not interact with corepressors. Thus, alpha 2 may be able to compete with thyroid hormone receptors for binding to a limited group of target sites, but it is not able to actively inhibit transcription.