The patterns of binding of RAR, RXR and TR homo- and heterodimers to direct repeats are dictated by the binding specificites of the DNA binding domains.

We show here that, in addition to generating an increase in DNA binding efficiency, heterodimerization of retinoid X receptor (RXR) with either retinoic acid receptor (RAR) or thyroid hormone receptor (TR) alters the binding site repertoires of RAR, RXR and TR homodimers. The binding site specificities of both homo- and heterodimers appear to be largely determined by their DNA binding domains (DBDs), and are dictated by (i) homocooperative DNA binding of the RXR DBD, (ii) heterocooperative DNA binding of RXR/RAR and RXR/TR DBDs, and (iii) steric hindrance. No homodimerization domain exists in the DBDs of TR and RAR. The dimerization function which is located in the ligand binding domain further stabilizes, but in general does not change, the repertoire dictated by the corresponding DBD(s). The binding repertoire can be further modified by the actual sequence of the binding site. We also provide evidence supporting the view that the cooperative binding of the RXR/RAR and RXR/TR DBDs to directly repeated elements is anisotropic, with interactions between the dimerization interfaces occurring only with RXR bound to the 5' located motif. This polarity, which appears to be maintained in the full-length receptor heterodimers, may constitute a novel parameter in promoter-specific transactivation.     

Dimerization interfaces formed between the DNA binding domains determine the cooperative binding of RXR/RAR and RXR/TR heterodimers to DR5 and DR4 elements.

We have previously reported that the binding site repertoires of heterodimers formed between retinoid X receptor (RXR) and either retinoic acid receptor (RAR) or thyroid hormone receptor (TR) bound to response elements consisting of directly repeated PuG(G/T)TCA motifs spaced by 1-5 bp [direct repeat (DR) elements 1-5] are highly similar to those of their corresponding DNA binding domains (DBDs). We have now mapped the dimerization surfaces located in the DBDs of RXR, RAR and TR, which are responsible for cooperative interaction on DR4 (RXR and TR) and DR5 (RXR and RAR). The D-box of the C-terminal CII finger of RXR provides one of the surfaces which is specifically required for the formation of the heterodimerization interfaces on both DR4 and DR5. Heterodimerization with the RXR DBD on DR5 specifically requires the tip of the RAR CI finger as the complementary surface, while a 7 amino acid sequence encompassing the 'prefinger region', but not the TR CI finger, is specifically required for efficient dimerization of TR and RXR DBDs on DR4. Importantly, DBD swapping experiments demonstrate not only that the binding site repertoires of the full-length receptors are dictated by those of their DBDs, but also that the formation of distinct dimerization interfaces between the DBDs are the critical determinants for cooperative DNA binding of these receptors to specific DRs.     

The dimerization interfaces formed between the DNA binding domains of RXR, RAR and TR determine the binding specificity and polarity of the full-length receptors to direct repeats.

Heterodimers of retinoid X receptor (RXR) and retinoic acid receptor (RAR) bind preferentially to directly repeated elements with spacing of two (DR2) or five (DR5) base pairs, due to the specific heterocooperative interaction of their DNA binding domains (DBDs) on these elements. We have demonstrated in the accompanying paper that the heterodimeric DBD interface that is responsible for the cooperative binding to DR5 elements, specifically involves the D-box of the RXR CII finger and the tip of the RAR CI finger. We show here that a second type of dimerization interface, which specifically implicates the RAR T-box and the RXR CII finger to the exclusion of the D-box, determines the selective binding to DR2 elements. Interestingly, the same type of dimerization interface (RXR T-box and CII finger) is responsible for the cooperative binding of homodimers of the RXR DBD to DR1 elements. Based on the three-dimensional structure of the glucocorticoid receptor DBD, modeling of RXR/RAR, RXR/TR and RXR/RXR DBD cooperative interactions predicts that in all cases the DBD contributing the CII finger, i.e. that of RXR, has to be positioned 5' to its cooperatively bound partner. This binding polarity of the DBDs is conferred upon the full-length receptors, since crosslinking experiments indicate that RXR is always 5' to RAR in complexes between either DR5 or DR2 and RXR/RAR heterodimers. The possible significance of these observations for transactivation by retinoic acid receptors is discussed.     

Identification of the promoter elements involved in the stimulation of ASCT2 expression by glutamine availability in HepG2 cells and the probable involvement of FXR/RXR dimers

Expression of the glutamine transport protein ASCT2 in the human hepatoma cell line HepG2 is increased when cells are cultured in the presence of glutamine and this has been shown to be due to stimulation of the ASCT2 promoter. Analysis of a number of promoter constructs localised the activation site to be between bases -653 and -543. Gel shift assays identified an IR-1 repeat within a 24bp region of this sequence which bound at least two nuclear proteins. Protein binding to this site was significantly higher in cells grown in glutamine-containing medium than when glutamine was absent. The identity of the higher molecular weight species binding to this promoter element was likely to be FXR/RXR dimers. Simultaneous overexpression of FXR and RXR increased the promoter activity in cells grown without glutamine to the same extent as did glutamine addition; the effects of glutamine and FXR/RXR expression were not additive. Mutagenesis of the FXR/RXR binding site in the promoter construct abolished glutamine and FXR/RXR stimulation. Real-time PCR showed levels of FXR mRNA were significantly increased in response to glutamine. The activity of the FXR promoter was also increased in response to glutamine. These results show that the stimulation of ASCT2 expression in response to glutamine in part involves binding of FXR/RXR to the ASCT2 promoter.     

Modulation of retinoic acid receptor alpha activity by lysine methylation in the DNA binding domain

Metabolic labeling and detection with a methylated lysine-specific antibody confirm lysine methylation of RAR alpha in mammalian cells. We previously reported Lys (347) trimethylation of mouse retinoic acid receptor alpha (RAR alpha) in the ligand binding domain (LBD) that affected ligand sensitivity of the dissected LBD. Here we report two monomethylated residues, Lys (109) and Lys (171) identified by LC-ESI-MS/MS in the DNA binding domain (DBD) and the hinge region, which affect retinoic acid (RA) sensitivity, coregulator interaction and heterodimerization with retinoid X receptor (RXR) in the context of the full-length protein. Constitutive negative mutation at Lys (109), but not Lys (171), reduces RA-dependent activation. Methylation at Lys (109) plays a more dominant role than trimethylation at Lys (347) in terms of RA activation of the full-length receptor. Lys (109) is located in a homologous sequence (CEGC K GFFRRS) of the DBD in RARs and is conserved in the nuclear receptor superfamily even across the species boundary. This study uncovers a potential role for monomethylation at Lys (109) in coordinating the synergy between DBD and LBD for ligand-dependent activation of RAR alpha.     

Structural basis for retinoic X receptor repression on the tetramer

Retinoic X receptor (RXR) is a master nuclear receptor in the processes of cell development and homeostasis. Unliganded RXR exists in an autorepressed tetramer, and agonists can induce RXR dimerization and coactivator recruitment for activation. However, the molecular mechanisms involving the corepressor recruitment and antagonist-mediated repression of RXR are still elusive. Here we report the crystal structure of RXRα ligand-binding domain (LBD) complexed with silencing mediator for retinoid and thyroid hormone receptors (SMRT) corepressor motif. As the first structural report on the unliganded nuclear receptor bound to the corepressor motif, RXRαLBD-SMRT exhibits a significant structural rearrangement, compared with apoRXRαLBD tetramer. To elucidate further the molecular determinants for RXR repression by its antagonist, we also determine the crystal structure of RXRαLBD-SMRT complexed with the identified antagonist rhein. In the structure, two rhein molecules and two SMRT peptides are in the RXRαLBD tetramer, different from the case in RXRαLBD-SMRT structure, where four SMRT peptides bind to RXRαLBD tetramer. It seems that rhein induces a displacement of SMRT motif by activation function 2 (AF-2) motif binding to the receptor. Combining our current work with the published results, structural superposition of RXRαLBDs in different states reveals that RXR uses an overlapped binding site for coactivator, corepressor, and AF-2 motifs, whereas the AF-2 motif adopts different conformations for agonist or antagonist interaction and coactivator or corepressor recruitment. Taken together, we thus propose a molecular model of RXR repression on the tetramer.     

Evidence for a novel cardiac-enriched retinoid X receptor partner.

Recent studies indicate that retinoid-mediated pathways play a pivotal role in cardiac morphogenesis and function. To identify proteins that serve as interacting partners of the retinoid X receptor alpha (RXRalpha) in heart, DNA-protein binding studies were performed with an RXR-responsive element (NRRE-1) derived from the medium chain acyl-CoA dehydrogenase gene promoter and nuclear protein extracts prepared from adult rat heart. NRRE-1 is a pleiotropic RXR-responsive element comprised of three potential recognition sites for class II members of the nuclear receptor superfamily. Gel mobility shift assays performed with an NRRE-1 probe in the absence or presence of bacterially overproduced RXRalpha and nuclear protein extracts prepared from adult rat heart, liver, or brain identified a cardiac-specific, RXR-dependent DNA-protein interaction. The NRRE-1-RXR.cardiac-enriched RXR-interacting protein (CERIP) complex exhibited a distinct mobility compared with NRRE-1-RXR.peroxisome proliferator-activated receptor, NRRE-1-RXR.retinoic acid receptor, or NRRE-1-RXR.thyroid receptor complexes. Mutational analysis demonstrated that two of the three potential binding half-sites of NRRE-1 (an everted repeat separated by an 8-base pair spacer) are required for the NRRE-1-RXR. CERIP interaction. Gel mobility shift assays demonstrated that CERIP interacted with RXRalpha and RXRgamma but not with RXRbeta, indicating a receptor subtypespecific binding preference and suggesting an RXR AB region-dependent interaction. The RXR.CERIP complex did not form on NRRE-1 when a mutant GST-RXRalpha fusion protein lacking the NH(2)-terminal AB region (but containing the receptor dimerization domain) of RXRalpha was added in place of the full-length RXRalpha, confirming a role for the AB region in the RXR. CERIP interaction. DNA-protein cross-linking studies demonstrated that CERIP is a DNA-binding protein of approximately 110 kDa. These results provide evidence for the existence of a cardiac-enriched DNA-binding protein that interacts with RXRalpha via the AB region and suggest a mechanism whereby cardiac retinoid signaling is controlled in an RXR subtype-specific manner.     

Genome-wide profiling of liver X receptor, retinoid X receptor, and peroxisome proliferator-activated receptor α in mouse liver reveals extensive sharing of binding sites

The liver X receptors (LXRs) are nuclear receptors that form permissive heterodimers with retinoid X receptor (RXR) and are important regulators of lipid metabolism in the liver. We have recently shown that RXR agonist-induced hypertriglyceridemia and hepatic steatosis in mice are dependent on LXRs and correlate with an LXR-dependent hepatic induction of lipogenic genes. To further investigate the roles of RXR and LXR in the regulation of hepatic gene expression, we have mapped the ligand-regulated genome-wide binding of these factors in mouse liver. We find that the RXR agonist bexarotene primarily increases the genomic binding of RXR, whereas the LXR agonist T0901317 greatly increases both LXR and RXR binding. Functional annotation of putative direct LXR target genes revealed a significant association with classical LXR-regulated pathways as well as peroxisome proliferator-activated receptor (PPAR) signaling pathways, and subsequent chromatin immunoprecipitation-sequencing (ChIP-seq) mapping of PPARα binding demonstrated binding of PPARα to 71 to 88% of the identified LXR-RXR binding sites. The combination of sequence analysis of shared binding regions and sequential ChIP on selected sites indicate that LXR-RXR and PPARα-RXR bind to degenerate response elements in a mutually exclusive manner. Together, our findings suggest extensive and unexpected cross talk between hepatic LXR and PPARα at the level of binding to shared genomic sites.