Crystal structure of ATF-2/c-Jun and IRF-3 bound to the interferon-beta enhancer

Abstract

Transcriptional activation of the interferon-beta (IFN-beta) gene requires assembly of an enhanceosome containing the transcription factors ATF-2/c-Jun, IRF-3/IRF-7, NF-kappaB and HMGI(Y). These factors cooperatively bind a composite DNA site and activate expression of the IFN-beta gene. The 3.0 A crystal structure of the DNA-binding domains of ATF-2/c-Jun and two IRF-3 molecules in a complex with 31 base pairs (bp) of the PRDIV-PRDIII region of the IFN-beta enhancer shows that association of the four proteins with DNA creates a continuous surface for the recognition of 24 bp. The structure, together with in vitro binding studies and protein mutagenesis, shows that protein-protein interactions are not critical for cooperative binding. Instead, cooperativity arises mainly through nucleotide sequence-dependent structural changes in the DNA that allow formation of complementary DNA conformations. Because the binding sites overlap on the enhancer, the unit of recognition is the entire nucleotide sequence, not the individual subsites.

Figure 1

The ATF-2/c-Jun–IRF-3–DNA complex (drawn with MOLSCRIPT; Kraulis, 1991). ATF-2 is red, c-Jun is blue, IRF-3 domain A is green and IRF-3 domain B is yellow. (A, B) Two views of the complex, related by 90°, showing the proximity of the ATF-2/c-Jun heterodimer to loops L1 and L3 of IRF-3A and IRF-3B. (C) Electrostatic surface potential of the DNA-bound ATF-2/c-Jun/IRF-3 complex calculated with GRASP (Nicholls et al, 1991). The color scheme indicates the electrostatic potential from −10 to +10k_BT, where k_B is the Boltzmann constant and T is the temperature (blue, positive; red, negative). The orientation is as in (A). (D) Amino-acid sequences of the ATF-2, c-Jun, IRF-3 constructs, and of the 31 bp double-helical DNA fragment used in crystallization. Letters a–e above the ATF-2 sequence indicate the heptad repeat of the leucine zipper. Amino-acid numbering corresponds to the human protein sequences. The DNA sequence corresponds to the PRDIV–III region (−102 to −72 nucleotides from the start site of transcription) of the IFN-β enhancer. The core binding sequences are in red (IRF-3A and IRF-3B) and in blue for ATF-2/c-Jun.

Figure 2

Protein–DNA contacts. Schematic diagram of protein–DNA contacts generated using NUCPLOT (Luscombe et al, 1997). Residues from ATF-2 are in red, c-Jun in blue, IRF-3A in green and IRF-3B in yellow. The core binding sites for each protein are indicated in the corresponding colors. Blue lines indicate hydrogen bonds and red lines van der Waals contacts. Note that contacts by His 40 of IRF-3A and IRF-3B are water mediated.

Figure 3

(A) Side view of the complex showing bending of the DNA around the IRF-3A and IRF-3B domains. The red line shows the local helical axis and the black line a straight overall helical axis fit as calculated with the program Curves (Lavery and Sklenar, 1988). (B) Superposition of Cα residues in the leucine zipper region of ATF-2/c-Jun with those of GCN4 bound to a CRE binding site (1DGC.pdb). The superposed structure is rotated about 90° and viewed from below to reveal the conformational differences in the DNA of the two structures. The GCN4 α helices are shown in cyan and the DNA from the GCN4/CRE complex is shown in orange.

Figure 4

EMSA comparing ATF-2/c-Jun and IRF-3 binding to the 31-mer DNA duplex used for crystallization. The DNA contains either the wild-type interferon-β enhancer 5′-TGACATAG-3′ (left) or the CRE 5′-TGACGTCA-3′ variant (right). In lanes 6–12 and 13–17, 20 μM ATF-2/c-Jun heterodimer was preincubated with 20 μM substrate DNA for 10 min before addition of increasing concentrations of IRF-3 dimer. In lanes 1–5, IRF-3 dimer was mixed directly with 20 μM substrate DNA. IRF-3 dimer concentrations were 30 μM (lanes 5 and 12), 18 μM (lanes 4 and 11), 12 μM (lanes 3 and 10), 8 μM (lanes 2 and 9), 5.5 μM (lanes 1 and 8) and 3.6 μM (lane 7). On the CRE site substrate, IRF-3 dimer concentrations were 3, 6, 12 and 24 μM (lanes 14–17). ATF-2/c-Jun alone bound to the wild-type or CRE sequence is shown in lanes 6 and 13, respectively. Note that at the highest IRF-3 dimer concentrations, an extra band probably corresponding to binding of a second IRF-3 dimer appears (lanes 5 and 12).

Figure 5

The ATF-2/IRF-3 interface. ATF-2 is red, IRF-3 domain A is green and DNA is in gray. Residues in the interface that were tested by mutagenesis are shown.

Figure 6

Overlapping DNA contacts. Loop L1 of IRF-3A inserts into the minor groove opposite c-Jun, and loop L1 from IRF-3B inserts opposite to helix α3 of IRF-3A.