Fig 1.
Identification of the Ets1 TAD and sub-TAD.
Chromatin interaction maps of mouse B cell Hi-C data showing the TAD (A) and sub-TAD (B) that include the Ets1 gene. The aqua triangles in (B) denote the sub-TAD within the TAD and the interaction hotspot discussed in the text. Below the interaction maps is CTCF ChIP-seq data from splenocytes. The aqua highlights denote the TAD and sub-TAD boundaries that overlap with strong CTCF enrichment. FIMO was used to identify CTCF motifs at the peaks of CTCF binding and are depicted below the CTCF ChIP-seq profiles. The CTCF binding site located at the down-stream boundary of the sub-TAD lacks a consensus CTCF motif and hence no sequence is shown at the bottom for this site.
Fig 2.
Design of a BAC-derived reporter construct to identify sequences required for lymphoid Ets1 expression.
(A) Hi-C data from mouse B cells shows that the RP23-350A20 BAC (whose extent is denoted by the light beige block flanked by dotted lines) contains the entire hotspot of most frequent interactions (smaller blue triangle in the top part of the figure). The first exon of Ets1 was replaced in the BAC by eGFP cDNA via recombineering. eGFP reporter expression is under the control of the putative response elements contained with the BAC. (B) qPCR and (C) Western blot to show expression of Ets1 and GFP in various tissues of BACtg mice and non-transgenic mice.
Fig 3.
Expression of the BACtg eGFP in lymphoid cells.
(A) Flow cytometry analysis of intracellular Ets1 and eGFP in gated B220 + CD19 + B cells from spleen of wild-type non-transgenic (WT) and BAC transgenic (BACtg) mice. The top figure shows GFP staining while the bottom figure shows intracellular Ets1 staining. (B) Flow cytometry plots of intracellular Ets1 versus eGFP staining in splenic B cell subsets: follicular B cells (B220 + CD19 + CD23hiCD21lo), marginal zone B cells (B220 + CD19 + CD23loCD21hi), germinal center B cells (B220 + CD19 + Fas+PNA+), memory B cells (B220 + CD19 + CD80 + PDL2+), plasma cells (B220loCD138 + CD98+), CD4 T cells (CD3 + CD4+) and CD8 T cells (CD3 + CD8+).
Fig 4.
The Ets1 sub-TAD region contains several putative response elements.
(A) Genome Browser image depicting B cell epigenetic profiles within the Ets1 sub-TAD. Datasets shown include mouse B cell ATAC-seq (accessible chromatin) and ChIP-seq for H3K27ac (active response elements), H3K4me1 (active or poised response elements), CTCF (denoting the upstream sub-TAD boundary), and H3K4me3 (active promoter). Highlighted in light blue are sequences with ATAC-seq peaks and enrichment of H3K27ac and H3K4me1. Several transcript variants of the mouse Ets1 gene are shown, but the major isoform expressed in B cells is Ets1/NM_011808.3 (B) The human Ets1 gene locus in regions homologous to the mouse BACtg. DNAse I hypersensitivity is shown below. Also shown is H3K27ac ChIP-seq (active response elements) from GM12878 B lymphoma cells, which identifies seven discrete regions of enrichment (green highlights labeled 1-7). Two transcript variants are shown, but the major isoform expressed in B cells is Ets1/NM_005238.4.
Fig 5.
Differences in Ets1 expression in B cell subsets correlate with changes in chromatin accessibility.
(A) Relative expression levels of Ets1 in sort-purified cells at various stages in the differentiation from hematopoietic stem cells to mature B cells and to plasma cells based on RNA-seq data from the ImmGen immunocyte RNA-seq project (GSE109125). (B) ATAC-seq profiles for follicular (FO B) versus plasma cells (PC), based on ImmGen ATAC-seq data (GSE100738), within the sequences included in the RP23-350A20 BAC. Blue highlighted regions mark six areas with differential ATAC-seq peaks in FO B cells versus PC. Also shown are H3K27Ac peaks from bulk CD3-, B220 + , CD19 + splenic B cells.
Fig 6.
Identification of differentially-accessible regions (DARs) in response to BCR stimulation.
(A) Schematic of the experimental layout. Total splenic B cells from wild-type C57BL/6 mice were isolated and either left unstimulated or stimulated for 2 hours with BCR crosslinking antibody (anti-IgM). ATAC-seq was performed and DARs were identified. (B) Top seven biological processes identified by analyzing genome-wide DARs that either gained or lost accessibility upon BCR stimulation. Gene ontology analysis was performed on genes within 10 kb of DARs using the DAVID software. (C) Transcription factor binding motifs enriched in DARs that gained (left) or lost (right) accessibility with BCR stimulation. Genome-wide DAR sequences were analyzed by HOMER for de novo transcription factor motif discovery. Shown are the top 3 de novo motifs enriched in each set of DARs and the best matching known consensus motif.
Fig 7.
DARs in the Ets1 gene locus contain motifs for B cell-relevant transcription factors.
(A) Shown are the ATAC-seq profiles in the Ets1 locus (within the region of the BAC transgene) in unstimulated (NT) and stimulated (αIgM) B cells. The third track shows the difference between the two datasets, with the positive peaks indicating sites that are more accessible in unstimulated and negative peaks indicating sites that are more accessible in stimulated B cells. DARs showing at least 2-fold changes in accessibility are highlighted blue for those that gained accessibility and highlighted in red for those that lost accessibility. (B) Close-up views of the DARs and transcription factor motifs they contain. FIMO was used to scan for transcription factor motifs within the four DARs. Depicted are the five highest scoring motifs. (C) ATAC-seq profiles compared with AP-1 subunits (c-Jun and JunD) ChIP-seq data from CH12 cells. The blue highlight denotes the + 49 kb DAR.
Fig 8.
Individual putative response elements do not mediate strong activation or repression in transient transfection assays.
(A) ATAC-seq profiles of follicular (FO) B cells and plasma cells (PC) in the Ets1 locus (the region contained within the BAC transgene). Highlighted in light blue are larger fragments equivalent to Sites 1-3 and 5-7 that were incorporated into luciferase constructs. Highlighted in bright green are small regions containing differentially-accessible regions comparing follicular B cells to plasma cells (as shown in Fig 5B). Highlighted in pink is the region surrounding the proximal promoter and part of first intron that were previously tested in transgenic mice and shown to be insufficient for mediating lymphocyte-specific expression. (B) General schematic of the design of the luciferase constructs. (C-D) A20 cells were transfected with firefly luciferase plasmids containing the Ets1 promoter alone or with the indicated response elements, or with empty vector, along with eF1α promoter Renilla luciferase internal control plasmid. Luciferase activity was measured 24 hours after transfection. Firefly luciferase activity was normalized to Renilla luciferase activity, then values were set relative to the plasmid with the Ets1 promoter alone. Significance was determined by one-way ANOVA. N = 2-5 replicates for each transfection.