Unique Immune Cell Coactivators Specify Locus Control Region Function and Cell Stage

Abstract

Locus control region (LCR) functions define cellular identity and have critical roles in diseases such as cancer, although the hierarchy of structural components and associated factors that drive functionality are incompletely understood. Here we show that OCA-B, a B cell-specific coactivator essential for germinal center (GC) formation, forms a ternary complex with the lymphoid-enriched OCT2 and GC-specific MEF2B transcription factors and that this complex occupies and activates an LCR that regulates the BCL6 proto-oncogene and is uniquely required by normal and malignant GC B cells. Mechanistically, through OCA-B-MED1 interactions, this complex is required for Mediator association with the BCL6 promoter. Densely tiled CRISPRi screening indicates that only LCR segments heavily bound by this ternary complex are essential for its function. Our results demonstrate how an intimately linked complex of lineage- and stage-specific factors converges on specific and highly essential enhancer elements to drive the function of a cell-type-defining LCR.

Keywords: B cell lymphoma; BCL6; CRISPR; MEF2B; Mediator complex; OCA-B; OCT2; enhancer; germinal center; locus control region.

Figure 1.. OCA-B and OCT2 are highly enriched at the BCL6 LCR

(A) Overlap of OCT2 and OCA-B ChIP-seq peaks in OCI-Ly7 cells. (B) Genomic distribution of OCT2 or OCA-B ChIP-seq peaks. (C) Density metaplots of OCT2 or OCA-B ChIP-seq peaks at primed, active, and super enhancers. (D) De novo motif analyses of OCT2 or OCA-B ChIP-seq peaks. (E) Loading of OCT2 or OCA-B on enhancers in OCI-Ly7 cells. (F) Genome browser view of the BCL6 locus with tracks for various ChIP-seq in NB, GCB, or OCI-Ly7 cells. Constituent enhancers (CEs) are shaded in grey. Differential expression of transcription factors and cofactors from NB to GCB is color coded. NB: naïve B-cells, GCB: germinal center B-cell. See also Figure S1.

Figure 2.. OCA-B binds to MEF2B to drive BCL6 LCR activity

(A, F) BCL6 mRNA level quantified by RT-qPCR in CRISPRi knock-down of OCA-B (A), OCT-2 (A) and MEF2B (F) in OCI-Ly7 cells. Error bars represent mean ± SD (n=3). (B, G) BCL6 protein level measured by immunoblot following CRIPSR knock-out of OCA-B (B), OCT2 (B) and MEF2B (G) in OCI-Ly7 cells. (C) OCA-B-associated proteins identified by IP-MS with anti-OCA-B antibody in OCI-Ly7 cells. Significances were measured by comparisons to signals from normal rabbit IgG antibody. (D) Association of MEF2B with OCA-B or OCT2 in OCI-Ly7 nuclear extracts measured by reciprocal co-IPs with indicated antibodies. (E) Association of MEF2B with OCT2+/OCA-B+ complex by double immunoprecipitation with indicated antibodies. (H) ChIP-qPCR for H3K27ac following CRISPRi knock-down of indicated factors in OCI-Ly7 cells. Error bars represent mean ± SD (n=3). (I) Schematic of reporter constructs in (J). (J) Luciferase reporter assays in 293T cells overexpressing indicated factors. Error bars represent mean ± SEM (n=3). See also Figure S2.

Figure 3.. OCT2, OCA-B and MEF2B bind cooperatively to the BCL6 LCR through octamer elements.

(A-B) Binding of OCT2 or MEF2B at 150 mM KCl (A) or both at 150, 200 and 300 mM KCl (B) on CE1 DNA with serially deleted fragments (Figure S3B) measured by ITA. (C) Cooperative binding of OCT2, OCA-B and MEF2B to CE1_O1–4 DNAs with wildtype (WT) or O3 octamer-mutated (mut) DNAs measured by ITA. (D) MEF2B and/or OCT2 binding to OCA-B measured by GST-pulldown assay. Arrowhead indicates degradation product of GST-OCA-B. (E-G) Measurement of stable (octamer-dependent) protein-DNA complexes by EMSA with a 50 bp DNA probe CE1_O3. Non-biotinylated DNA probes from wildtype O3 (wt), mutant O3 (mt) or an adjacent sequence of O3 (ns) were used as cold competitors in (F) and an O3 octamer-mutated (mut) probe (CE1_O3_mut) was included in (G). OCT2-CE1_O3 complex is visible on a longer exposed film at the bottom of (G). The anomalous competition by the mutant octamer probe in panel F, lane 6 reflects the use of an excessive (100x) amount of competitor, with octamer dependency being more clearly revealed in Panel G and in Figure S3H with mutant octamer probes and/or lower amounts (2–5x) of competitor, respectively. (H-M) Occupancy of MEF2B, OCT2, and OCA-B on CE1 measured by ChIP-qPCR analysis of MEF2B (H, I), OCT-2 (J, K) and OCA-B (L, M) binding to CE1 in wildtype, OCT2^−/− (H, L), OCA-B^−/− (I, K) or MEF2B^−/− (J, M) OCI-Ly7 cells. NR: negative control region 41 KB upstream of BCL6 TSS. (N) OCA-B and MEF2B co-occupancy measured by ChIP-re-ChIP-qPCR using anti-MEF2B or anti-OCT2 as 1^st antibodies, and anti-OCA-B antibody as 2^nd antibody. (H-N) Error bars represent mean ± SEM (n=3). See also Figure S3.

Figure 4.. OCA-B recruits the Mediator to BCL6 LCR constituent enhancers.

(A) Schematic of crosslinking-coupled binding assay. (B) Direct binding of OCA-B to the Mediator measured by crosslinking-coupled binding assay using recombinant OCA-B and affinity purified Mediator. (C) Direct binding of recombinant MED1 to OCA-B in GST-pulldown assay. (D) Association of MED1 with OCA-B in OCI-Ly7 nuclear extracts measured by co-immunoprecipitation using anti-OCA-B antibody. (E) MED1-dependent binding of Mediator to recombinant OCA-B in nuclear extract from wildtype (WT) versus Med1^−/− (KO) mouse embryonic fibroblasts. (F) MEF2B-facilitated, octamer-dependent, and OCA-B-dependent binding of MED1 to octamer-bound OCT2 on CE1_O2–4 measured by ITA. See also Figure S4.

Figure 5.. The Bcl6 LCR must be positioned in cis to allow GC formation in vivo

(A) Mating strategy used to obtain Bcl6 knock-out combined with Bcl6-LCR knock-out on separate alleles (i.e., in trans). (B-D) GC formation after sheep red blood cell (SRBC) immunization in wildtype, Bcl6 ^+/−, LCR ^+/−, LCR ^−/− and trans double heterozygous mice (the latter is highlighted in red) as shown by representative flow cytometry plots (B) and quantification of B220⁺ splenocytes (C) and GL7⁺FAS⁺ splenic GC B-cells (D). (E) Relative BCL6 mRNA levels in wildtype, CE1^+/−, or CE1^−/− OCI-Ly7 cells. (F) Genome browser view of the BCL6 locus with tracks for H3K27ac and OCA-B ChIP-seq in OCI-Ly7 cells, and 4C-seq anchored on the BCL6 promoter in GC B-cells. (G) Contact frequency between the BCL6 promoter (anchor) and the indicated LCR CE elements measured by 3C-qPCR in wildtype, CE1^−/− or OCA-B^−/− OCI-Ly7 cells. HEK293T cells were used as negative control. (H, I) Occupancy of RNA Pol II (RPB1, H) or Mediator (MED1, I) on the BCL6 promoter and CE1 region measured by ChIP-qPCR using indicated antibodies in wildtype vs. OCA-B^−/−, OCT2^−/−, MEF2B^−/− OCI-Ly7 cells. Error bars represent mean ± SEM (n=3). NR: negative control region 41 KB upstream of BCL6 TSS. See also Figure S4.

Figure 6.. BCL6 LCR architectural functions require OCA-B and ternary complex-bound enhancer elements

(A) Schematic representing screening strategy and overview of the screened region and H3K27ac ChIP-seq signal as a surrogate for regulatory elements. (B) Genome browser view of CRISPRi screening results along with H3K27ac ChIP (top). BCL6 LCR constituent enhancers (CEs) are categorized as essential enhancers if the mean growth score is lower than the 1^st percentile of control gRNA growth scores (1^st-99^th percentile indicated by horizontal lines, lower panel). TSS: BCL6 transcriptional start site. (C) Effect of CRISPRi-mediated perturbation of individual CEs, the TSS, and two negative control regions (deserts) on cell growth. gRNA-expressing cells are GFP positive and the fraction of GFP positive cells was monitored by flow cytometry over time. (D) Effect of CRISPRi-mediated perturbation of individual CEs, the TSS, and deserts on BCL6 expression as determined by qPCR. Error bars represent min/max values of biological replicates (n=2). Each region was targeted with two different gRNAs (C-D). See also Figure S5 and 6.