The malic enzyme (ME) gene is a target for both thyroid hormone receptors and peroxisome proliferatoractivated receptors (PPAR). Within the ME promoter, two direct repeat (DR)-1-like elements, MEp and MEd, have been identified as putative PPAR response elements (PPRE). We demonstrate that only MEp and not MEd is able to bind PPAR/retinoid X receptor (RXR) heterodimers and mediate peroxisome proliferator signaling. Taking advantage of the close sequence resemblance of MEp and MEd, we have identified crucial determinants of a PPRE. Using reciprocal mutation analyses of these two elements, we show the preference for adenine as the spacing nucleotide between the two half-sites of the PPRE and demonstrate the importance of the two first bases flanking the core DR1 in 5. This latter feature of the PPRE lead us to consider the polarity of the PPAR/RXR heterodimer bound to its cognate element. We demonstrate that, in contrast to the polarity of RXR/TR and RXR/RAR bound to DR4 and DR5 elements respectively, PPAR binds to the 5 extended half-site of the response element, while RXR occupies the 3 half-site. Consistent with this polarity is our finding that formation and binding of the PPAR/RXR heterodimer requires an intact hinge T region in RXR while its integrity is not required for binding of the RXR/TR heterodimer to a DR4.

While there is evidence that the two ubiquitously expressed thyroid hormone (T3) receptors, TR␣1 and TR␤1, have distinct functional specificities, the mechanism by which they discriminate potential target genes remains largely unexplained. In this study, we demonstrate that the thyroid hormone response elements (TRE) from the malic enzyme and myelin basic protein genes (ME TRE and MBP TRE ) respectively, are not functionally equivalent. The ME TRE , which is a direct repeat motif with a 4-base pair gap between the two half-site hexamers binds thyroid hormone receptor as a heterodimer with 9-cis-retinoic acid receptor (RXR) and mediates a high T3-dependent activation in response to TR␣1 or TR␤1 in NIH3T3 cells. In contrast, the MBP TRE , which consists of an inverted palindrome formed by two hexamers spaced by 6 base pairs, confers an efficient transactivation by TR␤1 but a poor transactivation by TR␣1. While both receptors form heterodimers with RXR on MBP TRE , the poor transactivation by TR␣1 correlates also with its ability to bind efficiently as a monomer. This monomer, which is only observed with TR␣1 bound to MBP TRE , interacts neither with N-CoR nor with SRC-1, explaining its functional inefficacy. However, in Xenopus oocytes, in which RXR proteins are not detectable, the transactivation mediated by TR␣1 and TR␤1 is equivalent and independent of a RXR supply, raising the question of the identity of the thyroid hormone receptor partner in these cells. Thus, in mammalian cells, the binding characteristics of TR␣1 to MBP TRE (i.e. high monomer binding efficiency and low transactivation activity) might explain the particular pattern of T3 responsiveness of MBP gene expression during central nervous system development. Thyroid hormone receptors (TRs)1 are transcription factors that belong to the steroid/thyroid nuclear receptor superfamily, which are encoded by two distinct genes, c-erbA␣ and c-erbA␤. Alternative splicing of the ␣ gene primary transcripts at their 3Ј extremity generates mRNAs encoding TR␣1, which binds thyroid hormone (triiodothyronine; T3), as well as the c-erbA␣2 and c-erbA␣3 variants, which do not. The ␤ gene expresses two T3-binding isoforms, TR␤1 and the minor form TR␤2 (reviewed in Ref. 1). In the adult animal, the ␣1 and ␤1 receptor mRNAs are found in all T3-sensitive tissues, albeit with differences in their relative abundance (2, 3). In contrast, expression of the two receptor isotypes is strikingly different during brain development. Whereas TR␣1 expression predominates at early stages and remains steady, TR␤1 expression is initially very low but exhibits a dramatic transient increase during the critical period of the central nervous system maturation (4 -6). At this moment, ␤1 transcripts prevail in the proliferation layer, while those of ␣1 are localized in the differentiation layer of the developing rat cerebral cortex. These patterns of expression are consistent with specific functions of the two receptors in brain development (7-9).TR transactivation activities are mediated thr...

The beta thyroid hormone receptor (TRbeta), but not TRalpha1, plays a specific role in mediating T(3)-dependent repression of hypothalamic TRH transcription. To investigate the structural basis of isoform specificity, we compared the transcriptional regulation and DNA binding obtained with chimeric and N-terminally deleted TRs. Using in vivo transfection assays to follow hypothalamic TRH transcription in the mouse brain, we found that TRbeta1 and chimeras with the TRbeta1 N terminus did not affect either transcriptional activation or repression from the rat TRH promoter, whereas N-terminally deleted TRbeta1 impaired T(3)-dependent repression. TRalpha1 or chimeras with the TRalpha1 N terminus reduced T(3)-independent transcriptional activation and blocked T(3)-dependent repression of transcription. Full deletion of the TRalpha1 N terminus restored ligand-independent activation of transcription. No TR isoform specificity was seen after transcription from a positive thyroid hormone response element. Gel mobility assays showed that all TRs tested bound specifically to the main negative thyroid hormone response element in the TRH promoter (site 4). Addition of neither steroid receptor coactivator 1 nor nuclear extracts from the hypothalamic paraventricular nuclei revealed any TR isoform specificity in binding to site 4. Thus N-terminal sequences specify TR T(3)-dependent repression of TRH transcription but not DNA recognition, emphasizing as yet unknown neuron-specific contributions to protein-promoter interactions in vivo.

The budding yeast Saccharomyces cerevisiae has been used extensively as a biological 'test tube' to study the regulation of the human estrogen receptor (ER) alpha. However, anti-estrogens, which are of great importance as therapeutic agents and research tools, fail to antagonize the activation by estrogen in yeast. Here, we have surveyed the antagonistic potential of five different anti-estrogens of diverse chemical nature. While they all act as agonists for wild-type ERalpha, we have established a novel yeast assay system for anti-estrogens, in which at least the commonly used anti-estrogen hydroxytamoxifen is a potent antagonist.

The beta thyroid hormone receptor (TRbeta), but not TRalpha1, plays a specific role in mediating T(3)-dependent repression of hypothalamic TRH transcription. To investigate the structural basis of isoform specificity, we compared the transcriptional regulation and DNA binding obtained with chimeric and N-terminally deleted TRs. Using in vivo transfection assays to follow hypothalamic TRH transcription in the mouse brain, we found that TRbeta1 and chimeras with the TRbeta1 N terminus did not affect either transcriptional activation or repression from the rat TRH promoter, whereas N-terminally deleted TRbeta1 impaired T(3)-dependent repression. TRalpha1 or chimeras with the TRalpha1 N terminus reduced T(3)-independent transcriptional activation and blocked T(3)-dependent repression of transcription. Full deletion of the TRalpha1 N terminus restored ligand-independent activation of transcription. No TR isoform specificity was seen after transcription from a positive thyroid hormone response element. Gel mobility assays showed that all TRs tested bound specifically to the main negative thyroid hormone response element in the TRH promoter (site 4). Addition of neither steroid receptor coactivator 1 nor nuclear extracts from the hypothalamic paraventricular nuclei revealed any TR isoform specificity in binding to site 4. Thus N-terminal sequences specify TR T(3)-dependent repression of TRH transcription but not DNA recognition, emphasizing as yet unknown neuron-specific contributions to protein-promoter interactions in vivo.