Identification and characterization of a novel corepressor interaction region in RVR and Rev-erbA alpha

Rev-erbA alpha and RVR are orphan nuclear receptors that function as dominant transcriptional silencers. Ligand-independent repression of transcription by Rev-erbA alpha and RVR is mediated by the nuclear receptor corepressors, N-CoR and its variants RIP (RXR interacting protein) 13a and RIP13 delta 1. The physical association between the corepressors and Rev-erbA alpha and RVR is dependent on the presence of a receptor interaction domain (RID) in the N-CoR family. Our previous study demonstrated that the E region of RVR and Rev-erbA alpha is necessary and sufficient for the in vivo interaction with the nuclear receptor corepressor, RIP13 delta 1. The present investigation demonstrates that two corepressor interaction regions, CIR-1 and CIR-2, separated by approximately 150 amino acids in the E region of RVR, are required for the interaction with N-CoR, RIP13a, and RIP13 delta A. The D region is not required for the physical interaction. In contrast, the D and E regions of Rev-erbA alpha were necessary for the interaction with the N-CoR and RIP13a-RIDs in vivo, suggesting that RIP13 delta 1 and N-CoR/RIP13a differentially interact with Rev-erbA alpha. Mutagenesis of CIR-1, a novel domain that is highly conserved between RVR and Rev-erbA alpha, demonstrated that the N-terminal portion of helix 3 plays a key role and is absolutely necessary for the interaction with RIP13 delta 1, RIP13a, and N-CoR. The phenylalanine residues, F402 and F441, in RVR and Rev-erbA alpha, respectively, were critical residues in supporting corepressor interaction. Cotransfection studies demonstrated that repression of a physiological target, the human Rev-erbA alpha promoter, by RVR was significantly impaired by mutation of CIR-1 or deletion of CIR-2. Furthermore, overexpression of either the N-CoR/RIP13a or RIP13 delta 1-RIDs alleviated RVR-mediated repression of the Rev-erbA alpha promoter, demonstrating that corepressor binding mediates the repression of a native target gene by RVR. A minimal region containing juxtapositioned CIR-1 and CIR-2 was sufficient for corepressor binding and transcriptional repression. In conclusion, our study has identified a new corepressor interaction region, CIR-1, in the N terminus of helix 3 in the E region of RVR and Rev-erbA alpha, that is required for transcriptional silencing. Furthermore, we provide evidence that CIR-1 and CIR-2 may form a single corepressor interaction interface.