Fig. 3: AP1 binding motifs are the main active region of e1.

A The minimum active unit of e1 was measured by dual luciferase reporter in PANC-1, HPAC and ASPC-1 cells, respectively. The left panel shows a schematic representation of the TF binding density predicted by the JASPAR database for e1-2. B Confidence ranking of binding TFs on e1-2 identified by TRAP. C Confidence ranking of the top 3 TFs. FOXD1 motif ranked first; AP1 motif ranked second; GATA1 motif ranked third. D mRNA of DKK1 expression levels after interfering with candidate TFs by qRT-PCR. E The enhancer activity of e1-2 was measured by dual luciferase reporter after interfering with candidate TFs. D, E treated with 10 pM SR11302 and 50 pM siRNA. F The enhancer activity of e1-2 was measured by dual luciferase reporter assay after AP1 binding site deletion. A, E, and F used PGL4.10 vector as negative control. G Scatter plot of correlation coefficient between DKK1 and JUND/FOSL2 gene expression in PDAC patients. H Western blot to detect the binding of JUND and FOSL2 in DNA pull down products. Input represents nuclear proteins as positive control; e1-2 containing unbiotinylated e1-2 probe; e1-2-biotin containing biotinylated e1-2 probe and nuclear proteins; e1-2(del-all)-biotin containing nuclear proteins and e1-2 probe which deletion AP1 binding sites. The qRT-PCR data were normalized to the expression of GAPDH. Means of three biological replicates are shown. Error bars indicate SEMs. **P < 0.01; ***P < 0.001; ns. no significance by two-tailed Student’s t test.