The Epstein-Barr Virus Regulome in Lymphoblastoid Cells

Abstract

Epstein-Barr virus (EBV) transforms B cells to continuously proliferating lymphoblastoid cell lines (LCLs), which represent an experimental model for EBV-associated cancers. EBV nuclear antigens (EBNAs) and LMP1 are EBV transcriptional regulators that are essential for LCL establishment, proliferation, and survival. Starting with the 3D genome organization map of LCL, we constructed a comprehensive EBV regulome encompassing 1,992 viral/cellular genes and enhancers. Approximately 30% of genes essential for LCL growth were linked to EBV enhancers. Deleting EBNA2 sites significantly reduced their target gene expression. Additional EBV super-enhancer (ESE) targets included MCL1, IRF4, and EBF. MYC ESE looping to the transcriptional stat site of MYC was dependent on EBNAs. Deleting MYC ESEs greatly reduced MYC expression and LCL growth. EBNA3A/3C altered CDKN2A/B spatial organization to suppress senescence. EZH2 inhibition decreased the looping at the CDKN2A/B loci and reduced LCL growth. This study provides a comprehensive view of the spatial organization of chromatin during EBV-driven cellular transformation.

Keywords: CDKN2A/B; ChIA-PET; EBV regulome; EBV super-enhancers; Epstein-Barr virus; MYC.

Copyright © 2017 Elsevier Inc. All rights reserved.

Figures

Figure 1. RNAPII ChIA-PET linked EBV enhancers to their direct target genes

(A) ChIA-PET links between EBV enhancers and known EBNA2 target genes MTA1, SLC7A5, and IRAK1 and effects of enhancer CRISPR knock out on their expression. ChIP-seq tracks for the indicated TFs were listed on top. In ChIA-PET tracks, each magenta line represented a pair of genomic interaction mediated by RNAPII. ChIA-PET linked genes were in red. gRNA targeted regions were indicated by arrows. LCLs stably expressing Cas9 were transduced with gRNA or control lentiviruses and selected with puromycin. Total RNAs were prepared from these cells and qRT-PCR was used to determine the expression level normalized against b-actin as loading control. The levels of control gRNA treated cells were set to 1. Error bars represented standard error. *p<0.05. (B) Intersection of genes essential for LCL growth and survival and EBV enhancer targeted genes. (C) ChIA-PET links between EBV enhancers and genes essential for LCL survival, CDK6, CCND2, ZNF143, CTBP, RBPJ and a well characterized EBNA2 target CR2. All data are represented as mean+/−SE. See also Figure S1 and Table S1 and S2.

Figure 2. RNAPII ChIA-PET linked ESEs to their direct targets

(A) Snapshot of a 5 mb region on chromosome 18. ChIP-seq tracks for EBV TFs were shown on the top. H3K27ac track was shown in the middle followed by RNAPII and CTCF ChIA-PET. Genes in the loci was listed with MALT1, PMAIP1(NOXA), and BCL2 highlighted. Red arrows indicated ESEs at the loci. (B) Snapshots of genome browser with ChIP-seq tracks and ChIA-PET links and cartoons representing the complexities of EBV mediated enhancer-promoter interactions. Top, ChIP-seq tracks. Middle red boxes indicated ESEs. RNAPII and CTCF ChIA-PET links were below ESEs followed by gene positions. Left panel: 1 ESE regulating 1 gene. Middle Panel: 1 ESE regulating 2 genes. Right panel: 3 ESEs regulating 1 gene. In the cartoons, red arrows indicated ESEs and blue arrows indicated ESE associated genes. Magenta lines represented ChIA-PET links. (C) ESE and many other EBV enhancers were linked to CFLAR. Orange arrows indicated the EBV enhancers, red arrow indicated ESE and green arrow indicated CFLAR gene body. See also Figure S2 and Table S3.

Figure 3. MYC ESEs were essential for MYC expression and cell growth

(A) RNAPII ChIA-PET linkages at the MYC locus. ChIP-seq tracks were shown on the top. MYCSE1 (-525kb from TSS) and MYCSE2 (-428kb from TSS) were indicated by red box below the tracks. RNAPII ChIA-PET links were shown by magenta lines. Orange boxes indicated the regions targeted by CRISPR. (B) Validation of MYCSE1 Deletion. Cas9 stable LCLs or BJAB were transduced with paired gRNA targeting the edges of ESE. After puromycin selection, genomic DNA was prepared from LCLs and BJAB cells transduced with dual gRNA. The targeted region was PCR amplified. The presence or absence of the MYC SE1 was shown. (C) MYCSE1 deletion reduced MYC mRNA level. 2 independent pairs of gRNAs both decreased MYC mRNA by qRT-PCR normalized to b-actin. Control gRNA treated cells were set to 1. (D) Growth of MYCSE1 deleted LCLs and LCLs with deletion in a control region. (E) MYCSE2 deletion also reduced MYC mRNA level. 2 independent pairs of gRNAs both decreased MYC mRNA by qRT-PCR. All data are represented as mean+/−SE. See also Figure S3.

Figure 4. EBNA2, EBNA3A, and EBNA3C were essential for MYC ESEs looping to MYC TSS

(A) qPCR primers spanning the MYC locus including the MYCSE1 and MYCSE2 enhancers for 3C were indicated on the top. The anchor primer was near MYC TSS. LCLs conditional for EBNA2, EBNA3A and EBNA3C were grown under conditions permissive or non-permissive for their expression. Cells were fixed with formaldehyde and lyzed. Chromatin was digested with EcoRI, diluted, and ligated. After reverse crosslinking, DNA was purified. qPCR was used to determine the interaction frequency. The amount of DNA used in qPCR was normalized with GAPDH and the qPCR primer efficiencies were normalized using ligated DNA from BAC clones covering the entire genomic region. EBNA2, EBNA3A, and EBNA3C inactivation significantly decreased MYC ESEs looping to MYC TSS (*p<0.05). Data are represented as mean+/−SE. (B) RBLs were infected with wild type B958 or EBNA2 deletion mutant P3HR1 EBV for the time indicated. 3C-qPCR indicated that wild type EBV induced MYC ESEs looping to MYC TSS as early as two days post infection and continued to increase at day 7. However, EBV mutant delete for EBNA2 and the EBNALP Y1Y2 exons failed to induce the looping. Data are represented as mean+/−SE. (C) ATAC-seq signals at the MYC ESEs. RBLs were infected with wild type or P3HR1 mutant EBV for the indicated time. Nuclei were first prepared and mixed with transposase reaction. The DNA was then purified and PCR amplified. Deep sequencing was used to determine the accessibility of the MYC ESE regions. Published GM12878 LCLs ATAC-seq were also included. (D) ATAC-seq signals at ESEs and the neighboring +/− 2kb windows during the RBL infection time course experiment. LCL signals were indicated. See also Figure S4.

Figure 5. EBNA3C reduced looping at the CDKN2A/B loci were

(A) RNAPII ChIA-PET linked the p14ARF, p15INK4B promoter and p16INK4A gene body in LCLs. CDKN2A/B loci were also linked to MTAP promoter region. (B) EBNA3C inactivation increased looping at the CDKN2A/B loci. Primers for 3C were listed on the top. Anchor primer was also indicated. EBNA3C conditional LCLs were grown under permissive or non-permissive conditions. 3C assays were done and the interaction frequencies between the anchor and other primers were determined by qPCR. The efficiencies of PCR primer pairs were normalized with BAC DNA. GAPDH was used to normalize the amount of DNA used in qPCR. All data are represented as mean+/−SE. See also Figure S5.

Figure 6. EBNA3A induced looping at the CDKN2A/B loci

(A) RNAPII ChIA-PET linked EBNA3A peaks >80k upstream of CDKN2A/B loci to the MTAP promoter and gene body. MTAP promoter and gene body were then linked to the CDKN2A/B loci. Major EBNA3C (orange) or EBNA3A (purple) ChIP-seq peaks were indicated. (B) Primers used for 3C at the ~100 kb region of the CDKN2A/B loci. Anchor primer near p16INK4A was indicated. 3C assay was done in GM12878 LCL (left) or EBNA3A conditional LCLs (right) grown under permissive or non-permissive conditions. qPCR was used to determine the interaction frequencies. EBNA3A inactivation significantly reduced EBNA3A site looping to the CDKN2A/B loci (*P<0.05). All data are represented as mean+/−SE. See also Figure S6.

Figure 7. EZH2 inhibitor decreased CDKN2A/B looping and stopped LCL growth

(A) GM12878 LCLs were grown in 1% FBS and treated with EZH2 inhibitor 20uM EPZ6438 for 4 Days. 3C-qPCR was used to determine the interaction frequencies at the CDKN2A/B loci. (B) GM12878 LCLs grown in 1% FBS were treated with 20uM EPZ6438 or 10uM UNC1999 for 4 days. Cell growth was determined using Cell Titer Glo. All data are represented as mean+/−SE. See also Figure S7.