Fig. 3.

Molecular determinants of the binding of p50-p50 to a G-IRE sequence. (A) Ribbon model of the structure of p50-p50 on the G-IRE and IRE of Gbp1, based on previous x-ray crystallography structures of p65-p50 on κB sites (36). In both the ribbon model and the G-IRE sequence, the DNA in the vicinity of the first monomer is indicated in red, where as the DNA in the vicinity of the second monomer is in orange. Conserved tyrosine residues (Y) that contribute to the stability of the complex by inserting themselves between bases are indicated in green. Rotational flexibility between the DNA and the dimerization domains of the second monomer is indicated. (B) Space-filling model detailing the amino acid residues within the first monomer that make close contacts with the G-IRE sequence. Numbering is with respect to the center of a palindromic sequence that would apply to the κB consensus sequence. (C) Space-filling model revealing a gap between the protein surface and the second IRE sequence with in Gbp1. This gap is opened by the rotational flexibility between the dimerization and the DNA binding domains. Note that Tyr⁵⁷ stabilizes the complex because it is anchored by hydrogen bonds to the phosphate backbone and makes van der Waals contacts with the bases A2 and A3 of the second IRE. (D) Determination of the binding specificity of p50-p50 for IREs by EMSA by titrating recombinant p50 proteins (1, 5, 10, 20, 40, 80, 200, and 500 nM) with the G-IRE–containing probe derived from Gbp1 or indicated mutant variants thereof. Data are from, or representative of, at least three experiments.