Transcriptional and DNA binding effects of BCL11A homomeric and BCL11A-BCL6 heteromeric complexes on artificial target genes. (A) Gal4DBD-BCL11A isoforms repress Gal4-mediated Firefly luciferase activity of pG5luc in BCL1 and M12.4 B cells or in HEK293 cells. Dual luciferase assays were performed 48 hr post-transfection into the indicated cell lines on whole cell lysates shown by Western blotting to contain approximately equal amounts of Gal4-BCL11A isoforms (data not shown). Values are the average of three independent experiments and are expressed as percent of those obtained using the reporter construct and Gal4DBD alone after normalization of transfection efficiency for the activity of Renilla luciferase. (B) Repression by BCL6, but not BCL11A, is TSA sensitive. pG5luc and Gal4DBD-BCL11A isoform fusions or FLAG-BCL6 along with a firefly luciferase reporter vector containing 5×-BCL6 binding sites upstream of the minimal SV40 promoter (5XBCL6-SV40-Luc; REF), and a Renilla luciferase control plasmid were transiently cotransfected into HEK293 cells. Indicated amounts of TSA were added into the cultures 24 hr after transfection, then 24 hr later Luciferase activity was measured. The values represent the average of three independent experiments. BCL6 does not affect BCL11A isoform repression in HEK293 and M12.4 cells when expressed at levels sufficient for co-immunoprecipitation. Transient co-transfections were carried out in triplicate, normalized and plotted as described above with pG5luc luciferase and Gal4DBD-BCL11A isoform fusions with or without FLAG-BCL6 or an irrelevant DNA (pCEP4) to equalize input DNA. (C) Reciprocal de-repression of BCL6 and BCL11A-XL in chicken DT-40 B cells. Co-transfection of FLAG-BCL6 decreases Gal4-mediated luciferase (GAL4-FF) repression of Gal4-BCL11A-XL isoforms. Co-transfection of HA-tagged BCL11A isoforms releases site-specific repression on 5XBCL6-SV40-Luc. Cells were transfected by electroporation and dual luciferase activities measured and normalized as above. (D) Identification of a specific DNA binding motif for BCL11A-XL. Cyclic amplification of targets (CASTing) for consensus sequences. In vitro translated N-terminal HA-tagged BCL11A-XL was subjected to 5 rounds of binding/PCR with a random 17-mer bridged by vector/cloning sites, as achieved by immobilization and immunoprecipitation with an anti-HA mAB and protein A beads. Cloned and sequenced candidates are aligned, with lower case letters denoting vector sequence and asterisks denoting sequences that contain >1 putative consensus (indicated at the bottom). (E) BCL6-BCL11A association results in enhanced binding of BCL11A-XL to it target sequence. Left, EMSA performed with the BCL11A consensus oligonucleotide probe (4× CASTing sites as determined in [D]) using the indicated protein extract amounts (μg) prepared from single transfections of FLAG-BCL6 or HA-BCL11A-XL into HEK293 cells. Right, EMSA with same probe, utilizing tagged-proteins from transfected DT40 B cells. The remaining HA-BCL11A-XL/DNA complex is ablated by anti-HA but not affected by anti-FLAG, indicating that no heteromeric complex is formed. (F) BCL6/BCL11A-XL association results in subtly decreased binding of BCL6 to specific DNA target sequence. The tandem BCL6 binding site repeat was isolated as a fragment from 5XBCL6-SV40-Luc, labeled with 32P and used as a probe in EMSA (detailed in Methods and Materials). The indicated amounts (μg) of proteins, prepared from nuclear extracts of untransfected HEK293 cells or BCL11A-XL transfected HEK293, and antibodies used in super-shifts were added (+) or not (-) to a constant amount (2 μg) of FLAG-BCL6 prepared in the same manner.
Liu et al. Molecular Cancer 2006 5:18 doi:10.1186/1476-4598-5-18