The germ line ɛ promoter is a high-affinity binding site of BCL-6. Two EBV-transformed human B-cell lines which differentially express BCL-6, Mutu I (BCL-6 positive) and Mutu III (BCL-6 negative), were cultured in the presence or absence of human recombinant IL-4 (10 U/ml) for 1 h prior to harvest. EMSAs were performed with 5 μg of whole-cell extract and an oligonucleotide probe corresponding to the germ line ɛ promoter Stat6 site (A) or the CD23b Stat6 site (B). The DNA binding complexes were identified upon supershift with antisera (Ab [antibody]) to the Stat6 (αStat6) or BCL-6 (αBCL-6) proteins. In panels C and D, the affinity of BCL-6 for the Stat6 site at −115 to −99 of the germ line ɛ promoter relative to other known binding sites of BCL-6 was assessed in cold competition assays. Five micrograms of whole-cell extract prepared from the BCL-6-positive M12.4.1 murine B-cell lymphoma line was incubated with a labeled probe generated from the Iɛ Stat6 site (−115 to −99) and increasing concentrations (10, 30, 60, or 100 ng) of cold competitor oligonucleotides. (C) The Iɛ Stat6 site (lanes 2 to 5) and B6BS, the in vitro defined binding site for BCL-6 (lanes 7 to 10). (D) The Iɛ Stat6 site (lanes 2 to 5), the Stat6 site of the CD23b promoter (lanes 7 to 10), and the Stat6 site of the putative IL-4 silencer (lanes 12 to 15).

BCL-6 and Stat6 protect overlapping regions of the germ line ɛ promoter. Binding of purified GST (400 ng) (lane 2), recombinant Stat6 (100 ng) (lane 3), or a fusion protein of GST with the DNA binding domain of BCL-6 (400 ng) (lane 4) to the murine germ line ɛ promoter was compared by DNase I footprinting analysis. A G+A sequencing reaction is shown in lane 1. The positions of the protected regions relative to the transcriptional start site (13) are indicated.

BCL-6 represses IL-4-inducible transcriptional activity. Mutu III cells (10⁷) were transfected with TK-luciferase reporter constructs driven by multimerized binding sites corresponding to either the Iɛ Stat6 site (Stat6-LUC) (A and D), the in vitro defined binding site of BCL-6 (B6BS-LUC) (B), or an HIV-κB binding site linked upstream of the minimal c-Fos promoter (κB-LUC) (C). These cells were cotransfected with increasing amounts of plasmid expressing either the full length BCL-6 (A to C) or the zinc finger DNA binding domain of BCL-6, HAZF (D). Following electroporation, the cells were either left untreated (B and C) or were divided and cultured either alone or in the presence of IL-4 (10 U/ml) for 24 h (A and D). Percent relative reporter gene activities are expressed as arbitrary units of luciferase activities normalized with β-galactosidase activities. The results are given as means ± standard deviations from three separate experiments.

Differential regulation of Iɛ and CD23b by BCL-6. M12.4.1 cells (5 × 10⁶) were cotransfected with a luciferase reporter driven by either −162 to +57 of the murine germ line ɛ promoter (A) or −183 to −33 of the murine CD23b promoter (B) and increasing concentrations of the indicated expression construct. Following electroporation, the cells were divided and cultured either alone or in the presence of IL-4 (10 U/ml) for 24 to 48 h. The results are given as means ± standard deviations from two (Iɛ) or three (CD23b) separate experiments and are normalized with β-galactosidase activities. Fold induction is given relative to unstimulated cells.

Repression of germ line ɛ transcription is dependent on an intact BCL-6 binding site. (A) Schematic diagram of mutations used in this study. These mutations were generated within the context of the germ line ɛ promoter (−167 to +55). (B) M12.4.1 cells were cultured in the presence of recombinant murine IL-4 for 1 h prior to harvest. EMSAs were performed with 5 μg of whole-cell extract and oligonucleotide probes corresponding to the germ line ɛ promoter BCL-6–Stat6 site mutants described in panel A: WT (lane 1), S2m1 (lane 2), S2m4 (lane 3), and S2m6 (lane 4). In panel C, the affinity of BCL-6 for the wild-type BCL-6–Stat6 site relative to the mutant BCL-6–Stat6 binding sites was assessed in cold competition assays. Five micrograms of whole-cell extract prepared in panel B was incubated with a labeled probe generated from the wild-type BCL-6–Stat6 site and increasing concentrations (3, 10, 30, or 100 ng) of cold competitor oligonucleotides: the wild-type BCL-6–Stat6 site (lanes 2 to 5), S2m1 (lanes 7 to 10), S2m4 (lanes 12 to 15), and S2m6 (lanes 17 to 20). In panel D, 5 × 10⁶ M12.4.1 cells were cotransfected with a luciferase reporter driven by the indicated variations on the germ line ɛ promoter (WT, S2m1, S2m4 or S2m6) and either control plasmid or 2.5 μg of BCL-6 expression vector. Following electroporation, the cells were divided and cultured either alone or in the presence of IL-4 (10 U/ml) for 24 to 48 h. The results are given as means ± standard deviations from three separate experiments and are normalized with β-galactosidase activities. Fold induction is given relative to unstimulated cells.

B cells from BCL-6^−/− mice produce higher levels of IgE in response to IL-4. IgM⁺ B cells from the spleens of 11 BCL-6^−/− mice or 10 wild-type littermates were cultured in the presence of LPS and cytokine (either 5,000 U of IL-4 per ml or 60 U of IFN-γ per ml) for 6 days. Supernatants from these cultures were assayed for secreted Ig isotypes by ELISA. Each diamond represents the levels from 1 mouse. The mean concentrations are noted by lines.

BCL-6^−/− Stat6^−/− mice do not produce increased levels of IgE. (A) Submandibular lymph node (top) and spleen (bottom) sections from BCL-6^Δ/Δ Stat6^+/+ (left panels) or BCL-6^Δ/Δ Stat6^−/− mice (right panels) were stained with anti-IgE or anti-IgG1 mouse antibodies, as indicated. The original magnifications were ×125 (top) and ×80 (bottom). (B) IgM⁺ B cells were isolated from the spleens of BCL-6^−/− mice, Stat6^−/− mice, BCL-6^−/− Stat6^−/− mice, or wild-type littermate controls. LPS cultures and ELISAs were performed as in Fig. 6. Each diamond represents the levels from one mouse. The mean concentrations are noted by lines.