Interferon regulatory factor-3 (IRF-3) functions as an essential transcription factor of type I IFNs induced by viral infection. IRF-3 activation is controlled by its phosphorylation, homo-dimerization and nuclear localization. So far, it has been reported that phosphorylation of C-terminal Ser/Thr residues including Ser396 modulate the transcriptional activity of IRF-3. However it is unclear which of Ser/Thr residues contribute to IRF-3 activation.
Based upon IRF-3 crystal structure, we proposed a model of head-to-tail dimer, which is essentially held between the C-terminal loop containing phospho-Ser386 and the pocket structure consisting of basic amino acid residues (basic pocket). Here we further inspected our model. First we re-investigated that IRF-3 Ser386 was phosphorylated in response to viral infection. IRF-3 Ser386Ala mutant abolished IRF-3 dimerization and IFN-β induction, suggesting that phosphorylation of IRF-3 Ser386 is critical for IRF-3 dimerization. In order to prove the presence of trans-interaction of IRF-3 via phospho-Ser386, we used recombinant IRF-3 protein and the synthesized peptide (IRF-3 peptide), which consists of the sequence of IRF-3 C-terminal loop containing Ser386. Phospho-IRF-3 peptide had a high affinity to recombinant IRF-3 protein compared to the non-phosphorylated counterpart in surface plasmon resonance analysis as well as pull-down assay. Next, we prepared the IRF-3 basic pocket-deficient mutants and confirmed that these mutants abolished IRF-3 dimerization. These recombinant mutants also reduced the affinity to phospho-IRF-3 peptide compared to WT IRF-3. Furthermore, phospho-IRF-3 peptide disrupted preformed IRF-3 dimer compared to the non-phospho-peptide. These results clearly indicate that phosphorylated Ser386 induces a specific binding to IRF-3 at its basic pocket, resulting in IRF-3 homo-dimer formation. The basic pocket appears to be a target for possible inhibitor for IRF-3-mediated gene activation, which mediate IFN-induced autoimmune disorders.