p21 is a direct target of SCL. (A) Analysis of p21 levels in MKs. (Left and middle panels) Quantitative reverse transcription PCR was performed using p21-specific primers. The y-axis represents the enrichment in cDNA sequences normalized to Gapdh gene control sequences. Data are the mean ± SD of 2 independent experiments. *P = .04. (Left panel) RNA was isolated from wild-type MkPs and MKs subsequently cultured in vitro for the indicated number of days (D1-D3) as well as from Cre;Sclfl/fl and Cre;Sclwt/wt MKs (SclΔ/Δ and Scl+/+, respectively). (Middle panel) RNA was isolated from sorted Cre;Sclfl/fl and Cre;Sclwt/wt MkPs. (Right panel) Western blot analysis of P21 levels in Cre;Sclfl/fl and Cre;Sclwt/wt MKs showing a 4-fold increase in P21 protein in mutant MKs. β-tubulin served as loading control. (B) Schematic representation of the mouse p21 proximal promoter and part of the first intron. The location of E-box (CANNTG, red and yellow boxes) and GATA (WGATAR, blue boxes) motifs is indicated in base pairs relative to the transcription start site (+1). The E-boxes E2 (+252) and E3 (+267) and the GATA motif (+578) are the sites mutagenized in the transactivation assays (see panel E). P1 to P6 show the location of the primer pairs designed for real-time PCR (not to scale). (C) ChIP analysis over the p21 locus using material isolated from MKD1 cells and antibodies as indicated in the figure. Data are the mean ± SD of 5 to 7 independent experiments; enrichment over no antibody. *P < .044. P1-P6 represents the genomic regions analyzed, as shown in panel B. (D) Same as in panel C, but chromatin was isolated from primary MKs derived from 5-FU–treated mouse bone marrow. Data are the mean ± SD of 3 independent experiments (enrichment over no antibody). *P = .006. (E) Transactivation assays in MKD1 cells. (Top) The luciferase gene is under control of a 736-bp fragment of the p21 promoter, including part of the first intron (p21-736). SA is a 225-bp splice acceptor sequence from the p21 gene subcloned upstream of the luciferase gene to allow for splicing. The constructs E2/E3, GATA, and E2/E3/GATA bear mutations in the 2 conserved E-boxes and GATA motif located in intron 1 as shown. The graph represents relative luciferase activity measured in MKD1 cells nucleotransfected with the wild-type reporter (P21-736, —) or mutated versions as indicated. The P21-736 construct was also assayed in MKD1 cells on Cre-mediated excision of the Scl floxed alleles (+Cre, Sclex/ex) and overexpression of ETO2 (+ETO2). Data are mean ± SD of 3 or 4 independent experiments performed in duplicate. *P < .01 (vs control cells transfected with p21-736 construct). (F; Top) PCR showing amplification of the floxed (fl) and excised (Δ) alleles in MKD1 cells (Sclfl/fl) and after Cre-mediated excision (Sclex/ex). (Bottom) Western blot analysis of ETO2 expression in MKD1 cells and after overexpression of ETO2 (+ETO2, 3-fold increase in ratio ETO2/histone H3). Histone H3 served as a loading control.