DEK binds in vivo to the GCSFR3 promoter region in a C/EBPα pS21–dependent manner to enhance C/EBPα transactivation capacity. (A) Cells (293T) were cotransfected in triplicate with C/EBPα–WT, S21A, and S21D (0 or 50 ng); increasing quantities of DEK (0, 50, or 150 ng); and a luciferase reporter driven by the TK minimal promoter along with a repeat of 4 C/EBPα–binding sites corresponding to those on the GCSFR3 promoter (left) or a vector containing 4 mutated sequences of the C/EBPα binding site as a negative control (right). (B) Cells (293T) were cotransfected in duplicate with C/EBPα (0 or 50 ng), increasing quantities of DEK (0, 50, or 150 ng), and a luciferase reporter driven by a short sequence (+74/−67) corresponding to the native GCSFR3 promoter region (left) or the same GCSFR3 promoter sequence but with a mutated C/EBPα binding site as a negative control (right). Transfection efficiency was normalized by cotransfection with Renilla TK (7 ng; Promega). Error bars indicate the SD resulting from independent experiments (n = 3 or 2). (C-D) Cross-linked chromatin was prepared from C/EBPα mutants (WT-ER, S21A-ER, and S21D-ER) and control (ER) K562 cells treated with 1μM β-estradiol for 0, 2, and 6 hours, and then immunoprecipitated using either anti-DEK (C) or C/EBPα (D) Abs. The precipitated chromatin was subjected to quantitative PCR using primer pairs that spanned the promoters of human GCSFR3 and GAPDH, respectively. Immunoprecipitation with rabbit IgG served as a negative control. Quantitative PCR signals of the immunoprecipitated chromatin were normalized against that from the input chromatin (1:10 dilution). Error bars indicate the SD resulting from 2 independent experiments performed in triplicate. ***P < .001; **P < .01.