Fig. 2
AP-1 monomers bind at unique loci that cannot be explained by differences in the DNA binding domain. a Hierarchical clustering of the relative strength of binding of each monomer at all AP-1 binding sites in Vehicle and 1-h Kdo2 lipid A (KLA) treatment conditions. b Representative browser shots of ChIP-seq peaks for Veh-specific monomers ATF3, Jun, and JunD. c Genome run-on sequencing at sites where ATF3, Jun, and JunD were lost, gained, or unchanged after 1 h KLA treatment. d Venn diagram of ATF3, Jun, and JunD peaks in Vehicle (left) and table indicating the de novo AP-1 motifs found in each subset of peaks and the percent of peaks in each subset that contain one of the two AP-1 motif variants (right). e Binding strength comparison of ATF3 chimeras. The ATF3 DNA binding domain (blue) is replaced by the DNA binding domains of Fos (yellow) or Jun (green) and then transduced into ATF3-deficient immortalized bone marrow-derived macrophage cells with a lentivirus vector (left). The binding of each chimera is shown as a heatmap of ChIP-seq tags centered on ATF3 chimera binding sites (replicates indicated in separate rows) that were found to be specific for ATF3 (blue) or Jun binding in thioglycollate elicited macrophages (green). f Heatmap showing the percent of unique binding sites for each monomer that contain a de novo motif calculated from each set of unique peaks. * indicates p < 0.01, for all comparisons between Lost in KLA, Unchanged Veh to KLA, and Gained in KLA for each AP-1 family member, independent T-test
