A new set of combinatorial dimeric codes for nuclear hormone receptor–target gene specificity

In vertebrates and invertebrates, nuclear hormone receptors (NRs) recognize a consensus DNA element of AGGTCA in the promoters of target genes through a highly conserved 10 amino acid motif called the P-box. AGGTCA is arrayed as dimeric repeats with certain combinations of orientation and spacing, conferring binding and transcription specificities to cognate NR homo- or heterodimers. However, the combinations are fewer than NR dimers and may not explain all NR–target gene specificities. Here in Caenorhabditis elegans, we identify a response element 79/49RE for two HNF4α NRs: NHR-79 and NHR-49. 79/49RE is two repeats of GACTAC inverted and spaced by 3 bp (IR3). NHR-49’s P-box is highly conserved while NHR-79’s diverges in four amino acids; nonetheless, NHR-49 and NHR-79 each alone can bind IR3 but together bind IR3 preferably as a heterodimer. Upon fasting, NHR-79-NHR-49-79/49RE upregulates genes for peroxisome proliferation, peroxisomal β-oxidation, and lipid droplet catabolism. Moreover, direct, inverted, and everted repeats of GACTAC spaced by 0–6 bp are widespread in genes of other functional categories, some of which are uniquely regulated by NHR-49. Thus, the P-box motif is remarkably plastic to recognize the novel repeat, which constitutes a new set of combinatorial dimeric codes for NR–target gene specificities.