A novel, internal Pu.1 promoter resides within an ORR1A0 LTR element and is repressed by KLF3. (A) RNA from Klf3+/+ (WT) and Klf3−/− (KO) TER119+ fetal liver cells was subjected to 5′ RACE using a reverse primer specific for exon 3 of Pu.1 and analyzed by agarose gel electrophoresis. The smaller band in the Klf3KO lane was sequenced and found to contain a novel exon (exon 2b). (B) The sequence of the ORR1A0 LTR, in which Pu.1 exon 2b is shown in bold. Sequences which fit the KLF binding consensus 5′-NCN CNC CCN-3′ are boxed, and the TATA box at −30 is underlined. (C) Schematic of the murine Pu.1 locus showing the position of exon 2b. Exons are represented by blue boxes, transcription start sites by arrowheads and splicing events by broken lines. Start points of translation (ATGs) for the two alternative transcripts are also shown. (D) Real-time RT-PCR quantification revealing that transcripts containing exon 2b spliced to exon 3 of Pu.1 (that is, Pu.2 transcripts) are upregulated in Klf3−/− TER119+ E14.5 fetal liver cells compared to Klf3+/− (HET) and Klf3+/+. Values have been normalized to 18S rRNA and the Klf3+/+ sample has been set to 1.0. n = 3 for each genotype. **, P <0.005 compared to both Klf3+/+ and Klf3+/− (Student’s two-tailed t-test). (E) ChIPs were performed on Klf3+/+ and Klf3−/− E14.5 fetal livers (n = 2 or 3 of each genotype per IP). Data are represented as the fold-change enrichment in Klf3−/− cells compared to Klf3+/+. The Fam132a and Klf8 promoters have been included as positive controls while Serpina9, Gapdh, and MyoD are negative control regions. *, P <0.05 compared to Gapdh (Student’s one-tailed t-test). In (D and E), error bars represent standard error of the mean.
Mak et al. Genome Biology 2014 15:R58 doi:10.1186/gb-2014-15-4-r58