Figure 1.
Analysis of ChIP-seq data for EBNA 2 and EBNA 3 proteins.
(A) Pie chart showing the distribution of all significant binding sites for EBNA 2 relative to gene TSSs. (B) Distribution of EBNA 3 family binding sites. (C) The frequency of EBNA 2 or EBNA 3 protein binding sites plotted as distance from the TSS of the closest gene. (D) Pie chart showing the proportion of sites identified for EBNA 2 and EBNA 3 family proteins that are shared or unique. (E) Comparison of genes closest to EBNA 2 binding sites with genes closest to EBNA 3 binding sites for sites within 2 kb of a TSS. (F) Comparison of genes closest to an EBNA 2 or EBNA 3 binding site located any distance from a gene TSS.
Figure 2.
Colocalization of histone modifications and transcription factor binding at EBNA 2 and 3 binding sites.
(A) Heatmap of EBNA 2, EBNA 3 and histone modification ChIP-seq signals at the top 1000 EBNA 2 binding sites. EBNA 2 and 3 ChIP-seq data from Mutu III BL cells was aggregated with ENCODE histone modification ChIP-seq data from the GM12878 LCL using hierarchical clustering. Each window displays the ChIP-seq signal −/+ 1 kb around the EBNA 2 binding site midpoint. Clusters of active enhancers (H3K4me1+, H3K27ac+) and poised enhancers (H3K4me1+, H3K27ac−) are indicated. (B) Heatmap of EBNA 3, EBNA 2 and histone modification ChIP-seq signals at the top 1000 EBNA 3 binding sites. (C) Position weight matrix and TF consensus prediction generated from unbiased motif searching using the top 300 EBNA 2 binding sites. Numbers show the p-value for site enrichment. (D) Position weight matrix derived from motif searching using the top 300 EBNA 3 family sites. (E) Position weight matrix for motif searching using all shared sites.
Figure 3.
EBNA 2, 3A, 3B and 3C binding at the CTBP2 locus in EBV-infected cells.
(A) The number of EBNA 2 (green) and EBNA 3 (red) sequencing reads from immunoprecipitated Mutu III DNA are plotted per million background-subtracted total reads and aligned with the human genome. The direction of gene transcription is indicated by the red arrow. GM12878 LCL H3K27ac ChIP-seq data from ENCODE are shown at the bottom of the panel. Panels B-E show ChIP-QPCR carried out in Mutu III cells and panels F-I show data from the PER253 B95.8 LCL. Precipitated DNA was analysed using primer sets located at the binding site (set B) or regions on either side of the binding site (sets A and C). Primers spanning the transcription start site of the cellular gene encoding peptidylprolyl isomerase A (PPIA) that is not regulated or bound by the EBNAs provide a background binding control (indicated by dotted lines). (B) and (F) ChIP using anti-EBNA 2 antibodies. (C) and (G) ChIP using anti-EBNA 3A antibodies. (D) and (H) ChIP using anti-EBNA 3B antibodies. (E) and (I) ChIP using anti-EBNA 3C antibodies. Percentage input signals, after subtraction of no antibody controls, are shown as the mean +/− range of two independent experiments. (J) Q-PCR analysis of CTBP2 transcript levels using cDNA from wild-type LCLs (wt1, 2 and 3) and LCLs established from EBNA 3A knock-out viruses (mtB1, B2 and B3) in two different donor backgrounds (D2 and D3). Transcript levels were normalised to GAPDH levels and expressed relative to the level in D2 wt1 cells. (K) Q-PCR analysis of CTBP2 transcript levels using cDNA from wild-type LCLs infected with B95.8 virus (wt) and EBNA 3B knock-out LCLs (KO) in two different donor backgrounds (PER142 and PER253). Transcript levels were normalised to GAPDH levels and expressed relative to the level in wt cells for each donor. (L) Q-PCR analysis of CTBP2 transcript levels using cDNA from the ER/EB 2.5 LCL expressing EBNA 2 that is active in the presence of β-estradiol (+ est) and inactive in the absence of β-estradiol (−est). Cells were incubated in β-estradiol-free media for 4 days prior to re-addition of β-estradiol or DMSO control for 6 or 17 hrs. Transcript levels were normalised to GAPDH levels and expressed relative to the level in the absence of β-estradiol for each time course. All cDNA results (J–L) show the mean −/+ range of two independent QPCR reactions each performed in duplicate.
Figure 4.
The influence of EBNA 2 and 3A on chromosome looping at the CTBP2 locus.
(A) Diagram (not to scale) showing the EcoR1 restriction fragments at the CTBP2 locus that encompass the promoter (P), enhancer (E) and an intervening control region (con). The arrow indicates the direction of transcription. (B) Chromosome conformation analysis in LCLs infected with wild-type or EBNA 3A knock-out EBV using primer pairs that amplify across promoter-enhancer or promoter-control ligation junctions. Positive controls show PCR amplification from control digestion and ligation reactions carried out using PCR-amplified DNA fragments encompassing the promoter, enhancer and control regions. (C). Chromosome conformation capture analysis in the ER-EB 2.5 LCL expressing EBNA 2 that is active in the presence of β-estradiol (+ est) and inactive in the absence of β-estradiol (−est). (D) Model for the control of chromatin looping by EBNA 2 and 3 proteins at CTBP2. (E) Re-ChIP analysis using anti-EBNA 2 antibodies in the first round of ChIP followed by a second round of ChIP in absence of antibody or using anti-EBNA 2, EBNA 3A, EBNA 3B or EBNA 3C antibodies. Results show mean percentage primary input −/+ range of two independent Q-PCR reactions from a representative experiment. (F) Control re-ChIP analysis using anti-EBNA 3A antibodies in the first round followed by re-precipitation in the absence of antibody or using anti-EBNA 3A antibodies. (G) Control re-ChIP analysis using anti-EBNA 3B antibodies in the first round followed by re-precipitation in the absence of antibody or using anti-EBNA 3B antibodies. (H) Control re-ChIP analysis using anti-EBNA 3C antibodies in the first round followed by re-precipitation in the absence of antibody or using anti-EBNA 3C antibodies.
Figure 5.
EBNA 2 and EBNA 3 protein binding at the WEE1 locus in EBV infected cells.
(A) EBNA 2 (green) and EBNA 3 (red) sequencing reads at the WEE1 locus (displayed as described in Figure 3). Panels B–E show ChIP-QPCR carried out in Mutu III cells and panels F–I show data from the PER253 B95.8 LCL. Precipitated DNA was analysed using primer sets located at the binding sites (sets B, D, F, H and J) or regions adjacent to the binding sites (sets A, C, E, G and I). Binding signals at the CTBP2 binding site in the same ChIP experiments are shown as a positive control and primers spanning the transcription start site of the cellular gene PPIA provide a background binding control (indicated by dotted lines). (B) and (F) ChIP using anti-EBNA 2 antibodies. (C) and (G) ChIP using anti-EBNA 3A antibodies. (D) and (H) ChIP using anti-EBNA 3B antibodies. (E) and (I) ChIP using anti-EBNA 3C antibodies. Percentage input signals, after subtraction of no antibody controls, are shown as the mean −/+ range of two independent ChIP experiments. (J) Q-PCR analysis of WEE1 transcript levels using cDNA from BL31 parental cells and BL31 cells infected with wild-type recombinant EBV (wtBac-2 and 3), EBNA 3C knock-out EBV (3C KO-3 and 6) or EBNA 3C revertant EBV (3Crev-2 and 4). Transcript levels were normalised to GAPDH levels and expressed relative to the level in parental BL31 cells. * indicates a p-value of <0.01 (students t-test) compared to the wtBac-2 cell line. (K) Q-PCR analysis of WEE1 transcript levels using cDNA from the ER/EB 2.5 LCL expressing EBNA 2 that is active in the presence of β-estradiol (+ est) and inactive in the absence of β-estradiol (−est). Cells were incubated in β-estradiol-free media for 4 days prior to re-addition of β-estradiol for 6 or 17 hrs. Transcript levels were normalised to GAPDH levels and expressed relative to the level in the absence of β-estradiol. All cDNA results (J–L) show the mean −/+ standard deviation of three independent QPCR analyses from two independent cDNA preparations.
Figure 6.
The influence of EBNA 2 and 3C on chromosome looping at the WEE1 locus.
(A) Diagram (not to scale) showing the EcoR1 restriction fragments at the WEE1 locus that encompass the promoter (P), two downstream enhancers (E1 and E2) and an intervening control region (con). The arrow indicates the direction of transcription. (B) Chromosome conformation analysis in BL31 parental cells and BL31 cells infected with wild-type recombinant EBV (wtBac-2), EBNA 3C knock-out EBV (3CKO-3) or EBNA 3C revertant EBV (3Crev-4) using primer pairs that amplify across promoter-enhancer or promoter-control ligation junctions. Positive controls show PCR amplification from control digestion and ligation reactions carried out using PCR-amplified DNA fragments encompassing the promoter and enhancers. (C) Chromosome conformation analysis in BL31 parental cells and BL31 cells infected with wild-type recombinant EBV (wtBac-2) or EBNA 2 KO EBV. (D) Chromosome conformation capture analysis in the ER-EB 2.5 LCL expressing EBNA 2 that is active in the presence of β-estradiol (+ est) and inactive in the absence of β-estradiol (−est). (E) Model for the control of chromatin looping by EBNA 2 and 3 proteins at WEE1. (F) Re-ChIP analysis in Mutu III cells using anti-EBNA 2 antibodies in the first round of ChIP followed by a second round of ChIP in absence of antibody or using anti-EBNA 2 or EBNA 3A, 3B or 3C antibodies. Primers at peak 5 in enhancer 2 were used for analysis. Results show mean percentage primary input −/+ range of two independent Q-PCR reactions from a representative experiment. (G) Control re-ChIP analysis using anti-EBNA 3C antibodies in the first round followed by re-precipitation in the absence of antibody or using anti-EBNA 3C antibodies.
Figure 7.
EBNA 2 and EBNA 3 protein binding at the ITGAL promoter in EBV-infected cells.
(A) EBNA 2 (green) and EBNA 3 (red) sequencing reads from immunoprecipitated Mutu III DNA plotted as in Figure 3. Panels B–E show ChIP-QPCR carried out in Mutu III cells and panels F–I show data from the PER253 B95.8 LCL. Precipitated DNA was analysed using primer sets located at the binding sites (sets B, D, and F) or regions adjacent to the binding sites (sets A, C, and E). Binding signals at the CTBP2 binding site in the same ChIP experiments are shown as a positive control and primers spanning the transcription start site of the cellular gene PPIA provide a background binding control (indicated by dotted lines). (B) and (F) ChIP using anti-EBNA 2 antibodies. (C) and (G) ChIP using anti-EBNA 3A antibodies. (D) and (H) ChIP using anti-EBNA 3B antibodies. (E) and (I) ChIP using anti-EBNA 3C antibodies. Percentage input signals, after subtraction of no antibody controls, are shown as the mean −/+ range of two independent ChIP experiments.
Figure 8.
The Effect of EBNA 2 and 3 proteins on ITGAL expression.
(A) Q-PCR analysis of ITGAL transcript levels using cDNA from wild-type LCLs infected with B95.8 virus (wt) and EBNA 3B knock-out LCLs (KO) in two different donor backgrounds (PER142 and PER253). Transcript levels were normalised to GAPDH levels and expressed relative to the level in wt cells for each donor. Results show the mean −/+ range of two independent QPCR reactions each performed in duplicate. (B) Luciferase reporter assays carried out in DG75 cells transiently transfected with 2 µg of the control vector pGL3 basic, an ITGAL promoter-luciferase reporter (pGL3 ITGALp) or the EBV C promoter reporter (pCp1425GL2) (right panel) in the absence or presence of 10 or 20 µg of EBNA 2, 3A, 3B or 3C expressing constructs. Firefly luciferase signals were normalised to Renilla luciferase signals from the co-transfected control plasmid pRL-TK (1 µg). Results show the mean −/+ standard deviation of 3 independent experiments and are expressed relative to the pGL3 basic signal (left panel) or the Cp1425GL2 signal (right panel) in the absence of EBNA 2. Western blot analysis of EBNA 2, 3A, 3B and 3C expression levels in transfected cells. Each set of blots was also probed for actin as a loading control. (C) Re-ChIP analysis in Mutu III cells using anti-EBNA 2 antibodies in the first round of ChIP followed by a second round of ChIP in absence of antibody or using anti-EBNA 2, EBNA 3B or EBNA 3C antibodies. Primers at peak 3 were used for analysis. Results show mean percentage primary input −/+ range of two independent Q-PCR reactions from a representative experiment. (D) Control re-ChIP analysis using anti-EBNA 3B antibodies in the first round followed by re-precipitation in the absence of antibody or using anti-EBNA 3B antibodies. (E) Control re-ChIP analysis using anti-EBNA 3C antibodies in the first round followed by re-precipitation in the absence of antibody or using anti-EBNA 3C antibodies.
Figure 9.
EBNA 3 protein binding at the BCL2L11 promoter in EBV-infected cells.
(A) EBNA 3 sequencing reads from immunoprecipitated Mutu III DNA plotted as in Figure 3. Panels B–D show ChIP-QPCR carried out in Mutu III cells and panels F–G show data from the PER253 B95.8 LCL. Precipitated DNA was analysed using primer sets located at the binding site (set B) or regions on either side of the binding site (sets A and C). Binding signals at the CTBP2 binding site in the same ChIP experiments are shown as a positive control and primers spanning the transcription start site of the cellular gene PPIA provide a background binding control (indicated by dotted lines). (B) and (E) ChIP using anti-EBNA 3A antibodies. (C) and (F) ChIP using anti-EBNA 3B antibodies. (D) and (G) ChIP using anti-EBNA 3C antibodies. Percentage input signals, after subtraction of no antibody controls, are shown as the mean −/+ range of two independent experiments.
Figure 10.
EBNA 3 protein binding at the ADAM28/ADAMDEC1 intergenic enhancer in EBV-infected cells.
ChIP-QPCR carried out in Mutu III cells (A–C) and the PER253 B95.8 LCL (D–F). Precipitated DNA was analysed using primer sets located at the centre of the binding site (set B) or the edges of the binding site (sets A and C). Binding signals at the CTBP2 binding site in the same ChIP experiments are shown as a positive control and primers spanning the transcription start site of the cellular gene PPIA provide a background binding control (indicated by dotted lines). (A) and (D) ChIP using anti-EBNA 3A antibodies. (B) and (E) ChIP using anti-EBNA 3B antibodies. (C) and (F) ChIP using anti-EBNA 3C antibodies.
Figure 11.
The influence of EBNA 3C on chromosome looping at the ADAM28/ADAMDEC1 locus.
(A) Diagram (not to scale) showing the HindIII restriction fragments around the ADAM28 locus that encompass the promoter (P), the ADAM enhancer (E, located downstream of ADAM28) and two intervening control regions (con1 and con2). The arrow indicates the direction of transcription. (B) Chromosome conformation analysis of the ADAM28 locus in the pz1 control BJAB cell line (−) and the E3C-3 stable EBNA 3C expressing cell line (+) using primer pairs that amplify across promoter-enhancer or promoter-control ligation junctions. Positive controls show PCR amplification from control digestion and ligation reactions carried out using PCR-amplified DNA fragments encompassing the promoter, enhancer and control regions. (C) Diagram (not to scale) showing the AciI restriction fragments around the ADAMDEC1 locus that encompass the promoter (P), the ADAM enhancer (E, located upstream of ADAMDEC1) and an intervening control region (con). The arrow indicates the direction of transcription. (D) Chromosome conformation analysis of the ADAMDEC1 locus in the pz1 control BJAB cell line (−) and the E3C-3 stable EBNA 3C expressing cell line (+) using primer pairs that amplify across promoter-enhancer or promoter-control ligation junctions. Positive controls show PCR amplification from control digestion and ligation reactions carried out using PCR-amplified DNA fragments encompassing the promoter, enhancer and control region.