Transcription factor IRF-3 is post-translationally activated by Toll-like receptor (TLR) signaling and has critical roles in the regulation of innate immunity. Here we present the X-ray crystal structure of the C-terminal regulatory domain of IRF-3(175−427) (IRF-3 175C) at a resolution of 2.3 Å. IRF-3 175C is structurally similar to the Mad homology domain 2 of the Smad family. Structural and functional analyses reveal phosphorylation-induced IRF-3 dimerization, which generates an extensive acidic pocket responsible for binding with p300/CBP. Although TLR and Smad signaling are evolutionarily independent, our results suggest that IRF-3 originates from Smad and acquires its function downstream of TLR.
Various pathogen-associated molecules, such as bacterial lipopoly-saccharide (LPS), CpG DNA and viral double-stranded RNA, evoke host reactions termed innate immunity. These reactions result in the production of various cytokines such as type I interferons (IFN-/), which are critical in the operation of adaptive immunity1,
2,
3. IFN-/ binds to cell surface receptors in both autocrine and paracrine manners to exert multiple biological functions including the activation of antiviral defense and modulation of the immune system. IFN-/ production is mediated by interferon regulatory factor (IRF) family proteins. The IRF family now includes nine members (IRF-1 to IRF-9)4,
5. The highly conserved N-terminal DNA-binding domain is the hallmark of the IRF family; however, the C-terminal domains are divergent, consistent with the notion that most members function nonredundantly5 (Fig. 1a). Recent studies have shown that IRF-3 is a key regulator of IFN-/ genes induced by pathogens1,
2,
6. IRF-3 is ubiquitously accumulated in cytoplasm and is inactive for DNA binding. Stimuli such as LPS, dsRNA and viral infection induce phosphorylation of specific serine residues in the serine-rich region (SRR) of IRF-3, resulting in homodimerization. The homodimer then translocates to the nucleus and forms a complex with coactivator p300/CBP. This holocomplex is responsible for the primary activation of IFN-/ genes4,
7. The secreted IFN-/ subsequently induces IRF-7 expression, which has a role in the secondary amplification of IFN-/ gene expression in a phosphorylation-dependent manner. Notably, the C-terminal regulatory domain (RD) and SRR are conserved between IRF-3 and IRF-7 and these proteins share functional similarity8 (Fig. 1a,b).
Figure 1. Domain structure of IRF-3 and sequence alignment and characterization of IRF-3 175C.
(a) Domain structure of IRF-3. DBD (DNA-binding domain), NES (nuclear export signal), RD (regulatory domain) and SRR (serine-rich region) are ordered from N terminus to C terminus. (b) Structure-based sequence alignment of IRF-3 175C, its homologs (human, mouse, chicken) and human IRF-7. The secondary structural elements are indicated below the alignment. The Ser-Ser-Leu motif in the 2S site is blue and serine and threonine residues in the 5ST site are pink. Positively charged residues and the residues comprising the hydrophobic pocket in the LHR pocket are red and green, respectively. The aspartate and glutamate residues comprising the acidic pocket are yellow. Several loops connecting secondary structural elements and C-terminal SRR are also indicated. (c) Sedimentation equilibrium ultracentrifugation of IRF-3 175C. Measurements were taken using a Beckman Optima XL-A analytical ultracentrifuge. The protein was at a concentration of 2.8 mg ml-1 in 50 mM Tris, pH 7.9, 500 mM NaCl, and was centrifuged at 11,000 rpm, 9,500g, 4 °C. The lines of lower panel were obtained from global fits using a monomer model (green), monomer-dimer equilibrium model (red) and dimer model (blue). Residuals for each model are presented in upper panels.
The phosphorylation sites critical to IRF-3 activation have not been precisely determined. Fujita and coworkers demonstrated that point mutation to alanine of Ser385 or Ser386 (2S site) (Fig. 1b), part of the unique phosphorylation site of the Ser-Ser-Leu motif (SSL motif), abolished the phosphorylation and dimerization of IRF-3 (refs. 7,9,10). Moreover, mutation of the corresponding SSL motif on IRF-7 similarly blocked activation11, suggesting that these serine residues are phosphorylation targets. Meanwhile, Hiscott and colleagues proposed an autoinhibition model in which two regions corresponding to residues 380−427 (SRR) and 98−240 of IRF-3 mutually interact to form a closed structure in the repressed state that is opened by the introduction of massive negative charges by multiple phosphorylation. This model was supported by an observation that mutation of the serine and threonine residues between residues 395 and 407 (5ST site) (Fig. 1b) to phosphomimetic aspartate (IRF-3 5D) results in constitutively active IRF-3 (refs. 4,12).
Here we report the crystal structure of the C-terminal regulatory domain of IRF-3 at 2.3-Å resolution. The structure reveals the mechanisms of phosphorylation, dimerization and binding to p300/CBP and demonstrates the structural and functional similarity between IRF-3 and Smad protein.
Results Characterization of IRF-3 175C IRF-3 175C was identified as a structural domain by limited proteolysis of intact IRF-3 with trypsin. IRF-3 175C eluted from a size-exclusion column as two species, corresponding to monomer and dimer, respectively (data not shown), but we did not observe higher-order aggregates. Subsequently, we used ultracentrifuge analysis to study the oligomerization state of IRF-3 175C. Using sedimentation equilibrium analysis, we estimated the average molecular mass of IRF-3 175C at 36 kDa, suggesting the existence of higher-order oligomeric species other than the monomer (Fig. 1c). Considering the results of the size-exclusion chromatography, global fitting with the model for either monomer-dimer equilibrium or dimer alone was applied. The better fit was obtained with the model for monomer-dimer equilibrium, with an estimated self-dissociation constant of 300−400 M (Fig. 1c). Thus, IRF-3 175C has the potential to form a dimer, but under physiological conditions, it exists exclusively as a monomer. Dimer formation is expected to be enhanced by phosphorylation of serine residues at the 2S and/or 5ST sites. Indeed, IRF-3 175C forms a dimer in L929 cells in a virus infection−dependent manner (data not shown).
Overall structure IRF-3 175C forms a dimer in the asymmetric unit of the crystal (Fig. 2a). The core of the structure consists of a -sandwich with antiparallel -sheets of five and six strands each (Fig. 2b,c). One end of the -sandwich is capped by a four helix bundle with two additional -strands (H1, H4, H6, H7, 12 and 13, together called 4HB). The other end of the -sandwich is capped with a group of loops (R211 loop, L299 loop and the loop connecting 10 and H5), a long -helix (H3) and two short -helices (H2 and H5), and is called a loop-helix region (LHR). According to the autoinhibition model, both autoinhibitory elements are integrated in the structural core as a part of 4HB and the -sandwiches (H1, 1, 2 and 3), respectively (Fig. 2b,c).
Figure 2. Overall structure of IRF-3 and its structural similarity to Smad2.
(a) A stereo view of the C trace of the IRF-3 175C homodimer in the asymmetric unit. Subunits A and B are blue and red, respectively. Dotted circles represent the LHR-pocket. This figure was prepared with MolScript33. (b,c) Ribbon diagram representations of the IRF-3 175C monomer. SRR is cyan, and a disordered loop is indicated by a dotted line. b and c are related by a 90° rotation along the horizontal axis. (d) Comparison of the structures of IRF-3 175C (left) and Smad2 (right). The structurally similar region between IRF-3 175C and Smad2 is in cyan and others are in yellow. All figures except Figures 2a, 3b and 4 were prepared with MolScript and Raster3D34.
Notably, IRF-3 175C is topologically and structurally similar to the Mad homology domain 2 (MH2 domain) of the Smad family (according to searches using Dali13) in spite of the low sequence homology between them (Fig. 2d) (r.m.s. deviation between IRF-3 175C and Smad2 MH2 domain is 3.2 Å for C in the superimposable region)14,
15,
16. However, there is a structural difference between the two proteins in the region after H6 (H5 in Smad). In IRF-3 175C, the region after H6 (SRR) is integrated in the structural domain; structural units of H7, 12 and 13 form a part of 4HB together with H1, H4 and H6. The corresponding region of Smad2 is replaced by a three-helix bundle (H3, H4 and H5) and its C-terminal phosphorylation site (SSXS motif) is located on the flexible tail (Fig. 2d).
In TGF- signaling, Smad family proteins are phosphorylated by TGF- receptor kinase and activated to form oligomers in a phosphorylation-dependent manner, similar to the IRF-3 activation mechanism. Structural and functional similarities between IRF-3 and Smad family proteins suggest an evolutionary link.
Phosphorylation site Identification of the phosphorylation site responsible for IRF-3 activation by virus infection is a crucial issue. Based on the crystal structure, we investigated the chemical environment of the phosphorylation site in SRR. The 2S site is located on the SRR loop connecting H6 and 12. In 5ST site, Ser396 and Ser398 are on the loop connecting 12 and 13, Ser402 is in 13 and Thr404 and Ser405 are in the N-terminal region of H7. Thus, the 5ST site is dispersed over separate regions of the structural units (Figs. 1b and 2b,c). The side chain solvent accessibility of Ser385 and Ser386 at the 2S site in the monomer are 44.5% and 100%, respectively. In contrast, those of Ser402 and Ser405 at the 5ST site are at most 50% and those of the remaining residues are <25%. These data indicate that the 2S site is more exposed and more likely to be phosphorylated. However, the phosphorylation of the 2S site may induce the structural change of the 5ST site to be accessible for phosphorylation, which may induce IRF-3 activation.
Dimer interface Although IRF-3 175C was not phosphorylated, it forms a dimer in the asymmetric unit in the crystal. Thus, we analyzed the dimer interface of IRF-3 175C in the crystal in detail. Subunit B inserts its SRR loop into the pocket comprising the -sandwich and LHR of subunit A (Figs. 2a and 3a); Leu387 on the SRR loop of subunit B binds to the hydrophobic pocket lined with the main chain of Gly212 in R211 loop, the side chains of Phe209 in 1, Leu299 in the L299 loop and Trp345 in 10 of subunit A. These residues are highly conserved among the IRF 3 family and IRF-7 (Fig. 1b). The side chain of Leu387 is 100% solvent accessible in the monomer, the thermodynamically unfavorable hydrophobic surface would be stabilized through dimer formation. On the periphery of this hydrophobic pocket there are four basic residues, Arg211 on the R211 loop, Arg213 on 2, and Lys360 and Arg361 on H5 (Fig. 3a,b). These residues are also highly conserved among IRF-3 from different species and IRF-7 (Fig. 1b) and may provide the binding surface for phosphorylated Ser385 and Ser386 of subunit B. Considering that the S385D S386D mutant did not mimic the phosphorylated IRF-3 (ref. 17), formation of multiple hydrogen bonds between phosphate groups and guanidinium groups of arginine residues may have critical roles in IRF-3 activation. We named this hydrophobic pocket, together with the four peripheral basic residues, the LHR pocket (Fig. 2a).
(a) Dimer interface between subunit A (blue) and subunit B (red). Side chain carbon atoms of the LHR pocket in subunit A and the SSL motif in subunit B are shown in ball-and-stick model in purple and green, respectively. (b) Stereo view of the dimer interface. A 2Fo - Fc map of the hydrophobic pocket of subunit A (purple) and the SSL motif of subunit B (red) contoured at 1 . This figure was prepared with XtalView35 and PyMOL (http://www.pymol.org). (c) Phosphorylation of the p50-tagged IRF-3 mutants. The wild type and mutants were transiently expressed in 293T cells and metabolically labeled with 32P (top). The extracts were subjected to immunoprecipitation using anti-p50-tag. The precipitates were analyzed by SDS-PAGE and immunoblotting using anti-human IRF-3 NES (anti-NES) as a probe (bottom). The radioactivity of IRF-3 was analyzed by Image Analyzer (Fuji Film). (d) Critical role of the basic residues and Leu387 in dimer formation. The p50-tagged IRF-3 mutants were transiently expressed in 293T cells and the extracts equivalent to equal amount of the wild type and the mutants were subjected to native PAGE. The resolved IRF-3 was detected by immunoblotting using anti-p50-tag as a probe. Arrows indicate the IRF-3 monomer and dimer.
Mutational analysis of the dimer interface To examine the functional role of the Leu387 and the LHR pocket, we prepared several mutants in which Leu387 was substituted with alanine (L387A) and Arg211, Arg213, Lys360 and Arg361 were substituted pairwise with alanine (R211A R213A and K360A R361A) in the context of wild-type IRF-3. When expressed in cells, these mutants were neither phosphorylated nor activated even after viral infection, suggesting that these residues were critical for the signal-induced phosphorylation of IRF-3 under tested conditions (data not shown). To analyze the involvement of Leu387 and the basic residues in dimer formation, we employed the constitutively active 5D mutant for the mutational analysis. When expressed in 293T cells, 5D is hyper-phosphorylated as compared with wild-type IRF-3 (Fig. 3c) and forms a homodimer without viral induction (Fig. 3d). Notably, Ser385 and Ser386 were critical for the constitutive activity of 5D because a composite mutant 5D S385A S386A was phosphorylation and dimerization defective (Fig. 3c,d), indicating that Ser385 and Ser386 are phosphorylated in a signal-independent manner in the context of 5D. A similar constitutive phosphorylation mutant was reported for STAT5 (ref. 18). Next, we prepared composite mutants of 5D L387A, 5D R211A R213A and 5D K360A R361A to examine the functional roles of these residues. Although these mutants were hyperphosphorylated at comparable levels to 5D in 293T cells (Fig. 3c), they did not form a dimer (Fig. 3d). These results support that the interaction between Leu387 and the LHR pocket plays a critical role in homodimer formation. Phosphorylation of Ser385 and Ser386 could further stabilize the dimer through hydrogen bond formation and electrostatic interaction among the phosphoserine and the basic residues in the LHR pocket. Finally, 5A mutant, in which the 5ST site is converted to alanine residues, is strongly inducible by virus, but mutation of the 2S site completely inactivates IRF-3 (refs. 17,19).
Phosphorylation of IRF-3 by TBK-1 During revision of the present manuscript, TANK-binding kinase-1 (TBK-1) and IB kinase- (IKK) were identified as IRF-3 kinases by Hiscott20 and Maniatis21. We coexpressed TBK-1 with IRF-3 or the above-mentioned mutants in the background of IRF-3, not in the background of IRF-3 5D mutant in 293T cells. Here, we detected the dimerization of IRF-3 and phosphorylation of the 2S site with anti-tag-specific and anti-2S-phosphorylation-specific antibodies, respectively. In the presence of TBK-1, IRF-3 was phosphorylated and formed a dimer, whereas the S385A S386A mutant was not phosphorylated and did not form a dimer. Notably, the R211A R213A and K360A R361A mutants were phosphorylated by TBK-1 but did not form a dimer (Fig. 4). These results clearly demonstrate that TBK-1 phosphorylates the Ser385 and Ser386 residues of IRF-3; this results in dimer formation. Although TBK-1 phosphorylates the same critical residues of the mutants R211A R213A and K360A R361A, no detectable dimer formation was observed, demonstrating that the basic residues in the LHR pocket are critical in the specific interaction with phosphoserines at positions 385 and 386, hence the dimer formation. The 5D mutant is further phosphorylated by TBK-1 and forms a dimer (data not shown), indicating that the 5D mutant retains a similar structure to that of the wild type and is phosphorylated by a similar mechanism to that of the wild type. The present results suggest that the phosphorylation-induced dimerization model is relevant to the IRF-3 activation mechanism.
Figure 4. Phosphorylation of the 2S site and the dimerization of IRF-3 by TBK-1.
293T cells were transfected with the expression vectors for p50-tagged IRF-3 as indicated at the top of the figure in the presence of empty vector (-) or TBK-1 expression vector (+). Cell extracts were subjected to native PAGE and probed with either anti-tag or specific antibody raised against IRF-3 peptide with phosphoserines at positions 385 and 386 (anti-P-IRF-3). Positions of IRF-3 monomer and dimer are shown by arrows.
Acidic surface of IRF-3 175C dimer The IRF-3 175C monomer has an acidic surface, comprising Glu200, Glu201, Glu203, Glu205, Glu224, Glu298, Glu377 and Asp392, which is fused to form an extensive acidic pocket on the dimer interface (Fig. 5a). These acidic residues are conserved between human and mouse IRF-3 (Fig. 1b). To explore the functional role of the acidic pocket, a mutant in which Glu200, Glu201, Glu203 and Glu205 were substituted to alanine (IRF-3 E/A mutant) was prepared. The trans-activation potential of IRF-3 E/A was severely impaired as revealed by a reporter assay (data not shown). Wild-type IRF-3 and IRF-3 E/A were expressed in L929 cells and analyzed by native PAGE (Fig. 5b). Viral infection induced dimer formation of IRF-3 E/A at comparable levels to the wild type. However, IRF-3 E/A did not associate with p300/CBP (Fig. 5c). This suggests that the extensive acidic pocket is required for the binding of p300/CBP. Recently, the IRF-3-binding domain (IBiD) of CBP was identified and its structure was determined using NMR22. The structure of IBiD consists of three -helices and its surface is highly positively charged, suggesting the interaction with the acidic pocket. IBiD also binds to the p160 nuclear receptor coactivator23. The structure of IBiD in complex with the p160 coactivator was appreciably different from free IBiD, where the orientation of the three -helices is different. The orientation of -helices of IBID may change to adapt to the acidic charged surface of its binding partners. The above structural and mutational analyses are consistent with the experimental evidence that IRF-7 does not apparently associate with p300/CBP because it lacks acidic residues corresponding to Glu200, Glu201, Glu203, Glu224, Glu298 and Asp392 (Fig. 1b)17.
(a) The electrostatic surface potential representation of the IRF-3 homodimer. The surface corresponds to the opposite surface shown in Figure 2a. Figure 5a was prepared with GRASP36 and Raster3D34. (b) Dimer formation of the E/A mutant. Expression vectors for p50-tagged wild type and the E/A mutant of IRF-3 were transiently expressed in L929 cells. After mock treatment (-) or infection with NDV for 12 h (+), the extracts were prepared and subjected to native PAGE using anti-p50-tag as a probe. (c) Critical role of glutamate residues in the association of p50-tagged IRF-3 with p300/CBP. The extracts in Figure 5c were immunoprecipitated with anti-NES, and resolved by SDS-PAGE followed by immunoblotting with anti-p300/CBP (top) or anti-p50-tag (bottom).
Discussion Based on the crystal structure of IRF-3 175C and mutational analyses of IRF-3, we proposed the phosphorylation-induced dimerization as the activation mechanism of IRF-3. The present structure could facilitate understanding of the IRF-3 activation cascade. In a related study also published in this issue, Qin et al.24 propose that phosphorylation-induced rearrangement of the autoinhibitory domains leads to the activation of IRF-3. Further structural analysis of the phosphorylated form of IRF-3 is required to distinguish the two mechanisms. However, mutational analyses and experiments using TBK-1 in our work support the phosphorylation-induced dimerization model.
The partial structural relationship between the C-terminal domain of IRF family proteins and the MH2 domain of Smad family proteins has been implied in sequence analysis25. The present study revealed that the overall architecture of IRF-3 175C is similar to the MH2 domain of the Smad family; it also revealed functional similarity between two proteins in the phosphorylation-induced activation mechanism. Homologs of Smad proteins are found in Drosophila melanogaster (Mad) and Caenorhabditis elegans (such as Sma-2, Sma-3 and Sma-4), whereas those of the IRF family proteins are identified only in vertebrates. Notably, the forkhead-associated domain (FHA), found in prokaryotes, has a similar -sandwich structure but is devoid of flanking 4HB and LHR26. FHA likely represents the original structure; Smad in eukaryotes presumably develops flanking regions from this prokaryotic scaffold to mediate phosphorylation signaling; the IRF family of proteins further diverged, and IRF-3 participates in the TLR signaling mechanism and functions as the key regulator of innate immunity.
Methods Protein expression, purification, and crystallization. His6-tagged IRF-3 175C was coexpressed with molecular chaperones Gro-EL and Gro-ES in Escherichia coli BL21 (DE3) cells27 and was purified using a Ni-NTA (Qiagen) affinity column. The His6-tagged protein was cleaved with TEV protease (GIBCO/BRL) and was further purified with Resource Q-Sepharose (Amersham Pharmacia Biotech) and size-exclusion Superdex 75 columns (Amersham Pharmacia Biotech). The purified IRF-3 175C was concentrated to 10 mg ml-1 and buffer-exchanged to 50 mM Tris buffer, pH 7.9, 500 mM NaCl. Selenomethionine (SeMet)-substituted IRF-3 175C was also produced by expressing the protein in the methionine auxotroph E. coli B834(DE3) and purified by the same methods as native IRF-3 175C. Native IRF-3 175C crystals were grown at 4 °C by the sitting-drop vapor diffusion method by mixing the protein solution with an equal amount of reservoir solution containing 100 mM sodium acetate, pH 5.1, and 400 mM magnesium formate. Microseeding was also applied to standardize the quality of crystals. The crystals appeared within 2−3 d and grew to a maximal size of 300 300 100 m. The SeMet-substituted IRF-3 175C was crystallized using the same method as with native crystals except that reservoir solution containing 100 mM sodium acetate, pH 5.1, 300 mM magnesium formate was used. The crystal of IRF-3 175C belongs to tetragonal space group P4212, with unit cell dimensions a = 134.8 Å and c = 69.3 Å, and contains two molecules in the asymmetric unit.
Data collection, structural analysis and refinement. All data collection for native and four heavy-atom derivatives was carried out at 100 K. For cryoprotection, the native crystal was soaked for 12 h, whereas the SeMet-substituted crystal was soaked for 3 min in crystallization solution (100 mM acetate and 400 mM magnesium formate, pH 5.1) supplemented with 50% (v/v) glycerol. The heavy atom was soaked in the cryoprotect solution mentioned above, using one of the following heavy-atom metals: 1 mM thimerosal (for 3 h), 100 mM K2Pt(CN)6 (for 48 h), 10 mM K2PtCl6 (for 16 h). Native data were collected using synchrotron radiation on a PX210 (Oxford) CCD detector at BL44XU in SPring-8 and processed with MOSFLM28 and SCALA29 in the CCP4 program suite. Derivative data were collected using a Rigaku R-AXIS IV imaging-plate detector equipped with a Rigaku ULTRAX18 rotating anode generator operated at 50 mV and 100 mA, and processed with DENZO and SCALEPAK30. The structure was phased by multiple isomorphous replacement (MIR) with CNS31. The heavy-atom sites of the thimerosal derivative were found by inspection of the Patterson map, and those of other derivatives were identified by Fourier difference methods. Initial MIR phases had a mean figure of merit of 0.60 at 3.3 Å resolution, and were improved by density modification with CNS. A model was built into MIR electron density maps with TURBO-FRODO32. After rigid-body refinement and simulated annealing with CNS, several rounds of model building with TURBO-FRODO and refinement with CNS were carried out and the refinement converged to a final R-factor value of 21.7% (Rfree = 24.2%) for all data at a resolution of 2.3 Å. Diffraction data and refinement statistics are shown in Table 1.
Table 1. Data collection, phasing and refinement statistics
Site-directed mutagenesis. Site-directed mutagenesis was done using Gene Editor (Promega). Cell culture, transcription and protein analysis by native or SDS-PAGE were carried out as described9.
Antibody. Anti-P-IRF-3 was obtained by immunizing rabbit with peptide N-CRVGGA(pS)(pS)LEN-C, using KLH carrier.
Coordinates. The coordinates of the IRF-3 175C have been deposited in the Protein Data Bank (accession code 1J2F).
Received 17 January 2003; Accepted 15 September 2003; Published online: 12 October 2003.
REFERENCES
Kawai, T. et al. Lipopolysaccharide stimulates the MyD88-independent pathway and results in activation of IFN-regulatory factor 3 and the expression of a subset of lipopolysaccharide-inducible genes. J. Immunol.167, 5887-5894 (2001). | PubMed | ISI | ChemPort |
Doyle, S. et al. IRF3 mediates a TLR3/TLR4-specific antiviral gene program. Immunity17, 251-263 (2002). | PubMed | ISI | ChemPort |
Kaisho, T. & Akira, S. Dendritic-cell function in Toll-like receptor- and MyD88-knockout mice. Trends Immunol.22, 78-83 (2001). | Article | PubMed | ISI | ChemPort |
Mamane, Y. et al. Interferon regulatory factors: the next generation. Gene237, 1-14 (1999). | Article | PubMed | ISI | ChemPort |
Taniguchi, T., Ogasawara, K., Takaoka, A. & Tanaka, N. IRF family of transcription factors as regulators of host defense. Annu. Rev. Immunol.19, 623-655 (2001). | Article | PubMed | ISI | ChemPort |
Alexopoulou, L., Holt, A.C., Medzhitov, R. & Flavell, R.A. Recognition of double-stranded RNA and activation of NF-B by Toll-like receptor 3. Nature413, 732-738 (2001). | Article | PubMed | ISI | ChemPort |
Yoneyama, M., Suhara, W. & Fujita, T. Control of IRF-3 activation by phosphorylation. J. Interferon Cytokine Res.22, 73-76 (2002). | Article | PubMed | ISI | ChemPort |
Sato, M. et al. Distinct and essential roles of transcription factors IRF-3 and IRF-7 in response to viruses for IFN-/ gene induction. Immunity13, 539-548 (2000). | PubMed | ISI | ChemPort |
Yoneyama, M. et al. Direct triggering of the type I interferon system by virus infection: activation of a transcription factor complex containing IRF-3 and CBP/p300. EMBO J.17, 1087-1095 (1998). | Article | PubMed | ISI | ChemPort |
Iwamura, T. et al. Induction of IRF-3/-7 kinase and NF-B in response to double-stranded RNA and virus infection: common and unique pathways. Genes Cells6, 375-388 (2001). | Article | PubMed | ISI | ChemPort |
Marie, I., Durbin, J.E. & Levy, D.E. Differential viral induction of distinct interferon- genes by positive feedback through interferon regulatory factor-7. EMBO J.17, 6660-6669 (1998). | Article | PubMed | ISI | ChemPort |
Lin, R., Mamane, Y. & Hiscott, J. Structural and functional analysis of interferon regulatory factor 3: localization of the transactivation and autoinhibitory domains. Mol. Cell. Biol.19, 2465-2474 (1999). | PubMed | ISI | ChemPort |
Holm, L. & Sander, C. Protein structure comparison by alignment of distance matrices. J. Mol. Biol.233, 123-138 (1993). | Article | PubMed | ISI | ChemPort |
Wu, J. et al. Crystal structure of a phosphorylated Smad2. Recognition of phosphoserine by the MH2 domain and insights on Smad function in TGF- signaling. Mol. Cell8, 1277-1289 (2001). | PubMed | ISI | ChemPort |
Shi, Y., Hata, A., Lo, R.S. & Massague, J. & Pavletich, N.P. A structural basis for mutational inactivation of the tumour suppressor Smad4. Nature388, 87-93 (1997). | Article | PubMed | ISI | ChemPort |
Qin, B., Lam, S.S. & Lin, K. Crystal structure of a transcriptionally active Smad4 fragment. Structure Fold. Des.7, 1493-1503 (1999). | Article | PubMed | ISI | ChemPort |
Suhara, W. et al. Analyses of virus-induced homomeric and heteromeric protein associations between IRF-3 and coactivator CBP/p300. J. Biochem.128, 301-307 (2000). | PubMed | ISI | ChemPort |
Onishi, M. et al. Identification and characterization of a constitutively active STAT5 mutant that promotes cell proliferation. Mol. Cell. Biol.18, 3871-3879 (1998). | PubMed | ISI | ChemPort |
Yang, H. et al. Transcriptional activity of interferon regulatory factor (IRF)-3 depends on multiple protein-protein interactions. Eur. J. Biochem.269, 6142-6151 (2002). | Article | PubMed | ISI | ChemPort |
Sharma, S. et al. Triggering the interferon antiviral response through an IKK-related pathway. Science300, 1148-1151 (2003). | Article | PubMed | ISI | ChemPort |
Fitzgerald, K.A. et al. IKK and TBK1 are essential components of the IRF3 signaling pathway. Nat. Immunol.4, 491-496 (2003). | Article | PubMed | ISI | ChemPort |
Lin, C.H. et al. A small domain of CBP/p300 binds diverse proteins: solution structure and functional studies. Mol. Cell8, 581-590 (2001). | PubMed | ISI | ChemPort |
Demarest, S.J. et al. Mutual synergistic folding in recruitment of CBP/p300 by p160 nuclear receptor coactivators. Nature415, 549-553 (2002). | Article | PubMed | ISI | ChemPort |
Qin, B. et al. Crystal structure of IRF-3 transactivation domain: autoinhibition and virus-induced phospho-activation. Nat. Struct. Biol. advance online publication, 12 October 2003 (doi: 10.1038/nsb1002). | Article | PubMed |
Eroshkin, A. & Mushegian, A. Conserved transactivation domain shared by interferon regulatory factors and Smad morphogens. J. Mol. Med.77, 403-405 (1999). | Article | PubMed | ISI | ChemPort |
Durocher, D. et al. The molecular basis of FHA domain:phosphopeptide binding specificity and implications for phospho-dependent signaling mechanisms. Mol. Cell6, 1169-1182 (2000) | PubMed | ISI | ChemPort |
Yasukawa, T. et al. Increase of solubility of foreign proteins in Escherichia coli by coproduction of the bacterial thioredoxin. J. Biol. Chem.270, 25328-25331 (1995). | Article | PubMed | ISI | ChemPort |
Leslie, A.G.W. Mosflm User Guide, Mosflm Version 6.0 (MRC Laboratory of Molecular Biology, Cambridge, UK, 1999).
Collaborative Computing Project Number 4. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D50, 760-763 (1994). | Article | ISI |
Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol.276, 307-326 (1997). | Article | PubMed | ISI | ChemPort |
Brnnger, A.T. et al. Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D54, 905-921 (1991). | Article |
Roussel, A. & Cambillau, C. TURBO-FRODO Manual (AFMB-CNRS, Marseille, France, 1996).
Kraulis, P.J. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr.24, 946-950 (1991). | Article | ISI |
Jones, T.A., Zhou, J.Y., Cowan, S.W. & Kjeldgaard, M. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A47, 110-119 (1991). | Article | PubMed | ISI |
Nicholls, A., Sharp, K.A. & Honig, B. Protein folding and association: insights from the interfacial and thermodynamic properties of hydrocarbons. Proteins11, 281-296 (1991). | PubMed | ISI | ChemPort |
Acknowledgments This work was supported by grant-in-aids for CREST of Japan Science and Technology, Scientific Research on Priority Areas and National Project on Protein Structural and Functional Analyses from the Ministry of Education, Culture, Sports, Science and Technology, Japan to F.I.
Competing interests statement:
The authors declare that they have no competing financial interests.
/
), which are critical in the operation of adaptive immunity
M (
. This figure was prepared with XtalView
B kinase-
(IKK
300 
