Epigenetic Activation of the EBV Protein Kinase Determines Antiviral Drug Response in Central Nervous System Lymphoma

Clinical Data

Patients with EBV+ CNSL who were treated with the GARD regimen at the Ohio State University (OSU) Wexner Medical Center between January 1998 and June 2024 were selected for this study. Research was approved by the OSU institutional review board record (IRB# 2023C0117) and clinical data was accessed retrospectively from the electronic medical records (Table 1, Table 2). Pathology was reported according to the original diagnosis and pathology reports. Complete response (CR), unconfirmed complete response (CRu), partial response (PR), and progressive disease (PD) were defined using guidelines from the International Primary CNS Lymphoma Collaborative Group (19). The GARD regimen consisted of 14 days of induction treatment with intravenous GCV (5mg/kg twice daily), AZT (1,500mg twice daily), and dexamethasone (10-40 mg). Four doses of rituximab (375 mg/m²) were given on days 1, 8, 15, and 22. Following induction, patients were prescribed maintenance oral valganciclovir (450mg twice daily) and AZT (300 mg twice daily).

Sample Collection and Consent

Tissue samples at diagnosis were obtained from the OSU Tissue Archive Services at the Department of Pathology. Patients provided consent to research via Ohio State University institutional review boards (IRB) protocols. Donor lymphoblastoid cell lines (LCLs) were generated from peripheral blood mononuclear cells (PBMCs) from EBV-seropositive healthy individuals with donor consent as described previously (20).

DNA and RNA isolation

Tissue biopsy DNA and RNA was isolated with QIAamp DNA FFPE Tissue Kit (QIAGEN) and RNeasy FFPE Kit (QIAGEN) according to manufacturer instructions. DNA and RNA from cell culture was isolated with QIAamp DNA Mini Kit (QIAGEN) and TRIzol reagent (Invitrogen), respectively. Plasmid and lentiviral vector DNA were prepared with ZymoPURE Plasmid Miniprep Kit (Zymo Research) and PureYield Plasmid Midiprep System (Promega).

Cell culture

Cells were grown in a humidified cell culture incubator at 5% CO2 and 37°C. All LCLs and hematopoetic cell lines were maintained in RPMI 1640 with 10% (v/v) fetal bovine serum (FBS) supplemented with penicillin/streptomycin (Gibco) and GlutaMAX (Gibco). LCLs were generated as described (20). HEK293, MEC1/2 and Raji lines were obtained from Christoph Plass’s group (German Cancer Research Center, Heidelberg, Germany). YT cells were obtained from Bethany Mundy’s group (The Ohio State University, Columbus, OH). Lenti-X HEK293T cells were obtained from Takara Bio. M81 EBV-carrying HEK293 cells (21) were a gift from Henri-Jaques Delecluse’s group (German Cancer Research Center, Heidelberg, Germany). HEK293/293T cells were maintained in DMEM media (Gibco) with 10% (v/v) FBS, penicillin/streptomycin and GlutaMAX. Burkitt lymphoma lines Kem, Mutu and Rael were obtained from Lisa Giulino-Roth’s group (Weill Cornell Medical College, New York, NY). All cell lines were confirmed mycoplasma-free via the MycoAlert Mycoplasma Detection Kit (Lonza) every two weeks during experiments.

Lentiviral transduction

IDH1 R132H plasmid (gift from Eric Campeau (22), Addgene plasmid #81686) was cloned into pLenti-PGK Blasti DEST lentiviral vector (gift from Jesse Boehm (23), Addgene plasmid #19065) using the Gateway cloning system (Invitrogen) in One Shot Stbl3 Chemically Competent E. coli (Invitrogen). Lentivirus particles were produced in HEK293T cells using standard methods. Viral supernatant was concentrated using Lenti-X Concentrator solution (Takara Bio) following the manufacturer protocol. Concentrated lentivirus preparations were mixed with 1.0e6 live HEK293 cells to yield a MOI of 5. Cells were selected with Blasticidin S HCl (10 µg/ml, GoldBio) for at least 14 days.

CRISPR-Cas9 gene knockout model

TET2 knockout was achieved with CRISPR-Cas9-mediated gene disruption using the TrueTag system (Invitrogen). Two homology templates (Puromycin/Blasticidin resistance-CMV-GFP/RFP) were designed together with a guideRNA targeting the TET2 locus in both alleles with the TrueDesign Genome Editor software (Invitrogen) and PCR-amplified. Homology templates (1250ng each) were mixed with 6.2µg of TrueCut HiFi Cas9 Protein (Invitrogen), 1200ng of guideRNA (Invitrogen, Suppl. Table 2) and 5e6 HEK293 cells in 100µl total volume of Electroporation buffer (MaxCyte) and electroporated using a MaxCyte GTx device (MaxCyte). Cells were selected under Puromycin Dihydrochloride (1µg/ml, Gibco) and Blasticidin S HCl (10µg/ml, GoldBio) for at least 14 days. GFP/RFP double-positive cells were then sorted on a FACSAria device (BD Biosciences) and maintained under Puromycin and Blasticidin.

qPCR and qRT-PCR

EBV was detected in DNA samples via primers amplifying the EBNA1 and normalized to human genome equivalents (ACTB locus, Suppl. Table 2), using PCR with Fast SYBR Green Master Mix (Applied Biosystems). RNA expression of EBV genes was assessed using quantitative reverse-transcriptase PCR (qRT-PCR). 1µg of total isolated RNA was reverse-transcribed using Superscript IV reverse transcriptase (Invitrogen) according to manufacturer protocol with random hexamer oligonucleotides (QIAGEN). qPCR with primers for EBV BZLF1, BXLF1, BGLF4 and LMP1 was carried out and normalized to human ACTB signal (Suppl. Table 2). qRT-PCR was performed using TaqMan Fast Advanced Master Mix (Applied Biosystems) and PrimeTime qPCR Probe Assays with FAM/ZEN/IBFQ probes (Integrated DNA technologies). Samples were run in technical duplicates and only included in analysis when both readouts yielded gene expression values within 15% range of each other. qPCR was performed on a QuantStudio 7 Pro real-time qPCR system (Applied Biosystems).

Western Blot

Western Immunoblots were carried out with standard methods using SDS-PAGE (Bio-Rad). Samples were blotted on PVDF membranes (Bio-Rad) and antibodies against IDH1 (Thermo Scientific, Cat.-N.: A304-161A-T), IDH1 R132H (Thermo Scientific, Cat.-No.: Z2010RT), TET2 (Thermo Scientific, Cat.-No.: A304-247A-T), GFP (ABclonal, Cat.-No.: AE012) and ACTB (Cell Signaling Technologies, Cat.-No.: 8H10D10) as loading control. Signals were visualized on an Odyssey CLx (LI-COR Biotech) using IRDye 800CW goat-anti-rabbit and IRDye 680RD goat-anti-mouse secondary antibodies (LI-COR Biotech).

DNA methylation assays

DNA methylation analysis of the LMP1, BGLF4, BXLF1 and BZLF1 regions (Suppl. Table 2) was carried out using the EpiTYPER assay as described (5). In short, DNA was bisulfite converted using EZ DNA methylation kit (Zymo Research) and EBV regions were PCR-amplified with primers specific for bisulfite DNA (Suppl. Table 2). PCR products were in-vitro transcribed and fragmented with RNase A. Fragments were analyzed via Matrix-Assisted Laser Desorption/Ionization-Time of Flight (MALDI-ToF) mass spectrometry (Agena Bioscience). Ratios of unmethylated versus methylated mass peaks were used to calculate the percentage of DNA methylation.

Global EBV DNA methylation analysis was carried out using the methylation-iPLEX assay as described (24). iPLEX capture and extension primers were designed to amplify and target EBV CpGs from bisulfite-converted DNA sequences using the Typer4.0 software (Agena Bioscience, Suppl. Table 2). All Samples were dispensed onto 384-well SpectroCHIP arrays using the RS1000 Nanodispenser and analyzed using the MassARRAY Analyzer4 system (Agena Bioscience).

EBV infection

For production of EBV viral particles, cells were co-transfected with EBV BZLF1, BRLF1 and BALF4 (gp110) mammalian expression plasmids described before (21). Virus collection and EBV infection was carried out with lentiviral transduction protocols described in the methods. After infection, cells were selected with Hygromycin B (200µg/ml, GoldBio) for at least 14 days.

Luciferase reporter assays

Luciferase assays were carried out in reporter vector pCpG-free promoter-lucia (Invivogen) as published previously (25). BGLF4 fragments were PCR amplified (Suppl. Table 2) and cloned into the HindIII/SdaI restriction sites using standard techniques. Reporter vectors were in-vitro methylated using M.SssI CpG methyltransferase (Thermo Fisher Scientific). Luciferase reporters were transfected into HEK293 cells with TransIT-LT1 reagent (Mirus Bio), and luciferase activity was read out after 48h in a Synergy HT plate reader (Biotek).

5’RACE

BGLF4 5’RACE was carried out based on published protocols (26) with modifications. 5µg of total RNA was reverse-transcribed using GSP-RT primer (Suppl. Table 2), poly-A tailed and amplified using nested PCRs with GSP (gene specific), Qt, Qo (outer) and Qi (inner)-primers (Suppl. Table 2 and (26)) with Q5 DNA polymerase (New England Biolabs). Products were gel-purified and cloned into pDONR221 (Invitrogen) with Gateway BP clonase II (Invitrogen) followed by Sanger sequencing at the OSU Genomics Shared Resource. EBV sequences were mapped to the EBV reference genome (RefSeq ID NC_007605.1) to determine 5’ends as transcription start sites.

Data Analysis and Statistics

Statistical significance was determined with unpaired, two-tailed t-test unless stated otherwise. Receiver-operating characteristics (ROC) curves and sensitivity/specificity were determined with easyROC ver. 1.3.1. Unsupervised clustering was carried out with Cluster3 (27) Using Euclidean Distance and visualized with Treeview. EBV genomic data tracks were generated from EBV reference genome (Refseq ID: NC_007605.1) using Integrative Genomics Viewer. Violin Plots were generated with BoxPlotR (28), with center lines showing medians, box limits indicating the 25th and 75th percentiles. Overall survival (OS) was defined as the time from start of treatment until death from all causes while progression free survival (PFS) was defined as the time from start of treatment until disease progression. OS and PFS were visualized with GraphPad Prism 10. Treatment related mortality (TRM) and was defined as death occurring within ninety days of completion of GARD regimen that is not due to progressive disease.

Long-term follow up reveals durable responses to GARD treatment of EBV+ CNSL

A total of n=24 patients with EBV+ CNSL were treated at OSU with GARD between 1998 and 2024, with a subset previously described (13). Here we report additional epigenetic studies and follow-up on these patients. This cohort included patients with PTLD PCNSL as well as patients on iatrogenic immune suppression for autoimmune conditions, secondary PTLD CNSL with peripheral PTLD involvement and two EBV+ PCNSL cases with no evidence of immunosuppression (Table 1). The mean age at diagnosis was 56±19 years (Suppl. Table 1). EBV tumor status was verified via EBER1 staining of biopsy tissue or detectable EBV viral load in the cerebrospinal fluid in cases where a biopsy was unavailable. Cytopenia and other side effects like nausea and weight loss generally improved by holding induction or reducing maintenance therapy. Diffuse large B-cell lymphoma was the most common histology (n=11; 45.8%) followed by grade 3 lymphomatoid granulomatosis (LYG) (n=5; 20.8%) and other B-cell lymphoproliferative disease (n=5;20.8%). Median survival time was 5.4 years, and n=15 patients (63%) achieved a complete response (CR) or CRu, with an overall response rate of 79%. Two-year survival was 63% (95% CI 46-85%) and five-year survival was 53% (95% CI 37-78%) (Figure 1, Table 2). We show rapid resolution of CNSL lesions upon GARD induction (Fig. 1A). Long-term follow up over 15 years post-treatment revealed durable responses and favorable overall and progression-free survival in CNSL patients treated with GARD (Fig. 1B,C).

• Download figure
• Open in new tab

Figure 1. Clinical GARD response characteristics of EBV+CNSL.

(A) Contrast MRIs from two patients with sequential imaging. Both patients achieved complete and durable that persisted until the patients passed from causes unrelated to PTLD or treatment. (B,C) Kaplan-Meier curves representing overall survival (B) and progression-free survival (B) in n=24 CNSL patients treated at OSU. (D,E) Representative IHC images for lytic EBV proteins BXLF1 (D) and BGLF4 (E).

DNA methylation loss in BGLF4 is associated with BGLF4 expression in CNSL

We have reported widespread expression of the GCV/AZT target kinases in CNSL tumor biopsies (Fig. 1D,E and (13)). However, the molecular determinant for viral kinase expression remained unknown. In particular, it has remained uncertain if EBV BXLF1/BGLF4 presence was due to lytic replication of EBV. In order to address this question, we carried out gene expression analysis of lytic and latent EBV transcripts in available tumor tissue biopsies of CNSL (n=12, Table 1) and peripheral PTLD patients (n=24, Suppl. Table 1). BGLF4 showed expression in CNSL significantly above that of peripheral PTLDs, with BXLF1 expression differences close to statistical significance (Fig. 2A, Suppl. Fig 2A). RNA expression of the essential immediate-early lytic transcript BZLF1 (Suppl. Fig. 2B), an indicator of lytic cycle entry (2) showed that this transcript was undetectable in many peripheral PTLDs (9/21, 41%) and all but one CNSL (11/12, 92%). The remaining biopsies displayed markedly low BZLF1, with the exception of a subset of peripheral PTLDs (6/21, 27%). RNA expression of the EBV latent membrane protein 1 (LMP1, Suppl. Fig. 2C) was detectable in the majority of peripheral PTLDs (16/21, 73%) and CNSLs (9/12, 75%), indicating latent EBV.

• Download figure
• Open in new tab

Figure 2. Lowered EBV BGLF4 DNA methylation and increased BGLF4 expression in CNSL.

(A) RNA expression (qRT-PCR) and (B) EBV BGLF4 DNA methylation determined via EpiTYPER in peripheral PTLD and CNSL tissue biopsies. (C) correlation of BGLF4 RNA expression and DNA methylation. (D) Average EBV DNA methylation determined via iPLEX assay. Data points show average methylation across 31 EBV CpG sites. (E) Average LINE1 DNA methylation determined via EpiTYPER. (F) correlation of BGLF4 RNA expression and average EBV DNA methylation from (D), R²: correlation coefficient. *p<0.05, ***p<0.001, t-test (B,D,E: Exact Wilcoxon-Mann-Whitney test)

To distinguish the specific epigenetic state of EBV loci, we next carried out quantitative DNA methylation profiling of BGLF4, BXLF1, BZLF1 and LMP1 upstream sequences using the EpiTYPER assay (29) in tissue biopsy DNA. We revealed significantly lowered BGLF4 methylation in CNSL compared to peripheral PTLD (p= 0.0006; Fig. 2B, Suppl. Fig. 2D), but no significant differences in BXLF1, BZLF1 or LMP1 (Suppl. Fig. 2D-F), highlighting that these EBV loci have distinct DNA methylation states. Comparison of RNA expression and methylation showed inverse correlation for BGLF4 (Fig. 2C). We next sought to determine whether the significant EBV methylation differences between CNSL and peripheral PTLDs were the result of global methylation alterations. For this purpose we read out n=31 distinct methylation sites evenly distributed in the EBV genome to establish a representative global EBV methylation state (Fig 2D, Suppl. Fig. 2G). We also read out human genomic LINE1, LTR and Alu repeat element methylation as a surrogate of global host cellular methylation (Fig. 2E, Suppl. Fig. 2H,I). With the exception of a low (avg. 3.2%) methylation difference for LINE1, results indicate that global EBV and host cellular DNA methylation are not significantly different in CNSL versus peripheral PTLD, thus indicating a site-specific regulation at BGLF4 and BXLF1. In line with this observation, we found no significant correlation between BGLF4 mRNA expression and average EBV methylation (Fig. 2F). Lytic EBV activation is characterized by a rapid expansion of EBV DNA copies within a cell, loss of global EBV DNA methylation (2) and widespread lytic gene expression including BGLF4 (30). We assessed these lytic EBV characteristics by correlating BGLF4 expression with EBV DNA loads determined in tissue samples via qPCR (Suppl. Fig. 2J) and clinically determined EBV qPCR from serum in peripheral PTLDs (Suppl. Fig. 2K). These comparisons showed no evident correlations, indicating again that BGLF4 expression is not associated with overall EBV lytic activation. We also assessed the correlation of overall EBV methylation and EBV load (Suppl. Fig. 2L) and again saw no correlation between these readouts. Together, these observations revealed strikingly low DNA methylation at BGLF4 in CNSL. This lowered methylation did not align with correlates of overall methylation or lytic virus activity (EBV load, EBV methylation, BZLF1 expression), thus indicating a highly site-specific methylation event while retaining multiple features of viral latency.

Low DNA methylation at three individual BGLF4 upstream loci

In order to understand the specificity of lowered methylation in CNSL, we carried out a detailed assessment of DNA methylation in an approximately 1 kilobase upstream region of BGLF4 previously shown hold transcription initiation sites (TSS) (31). Using the EpiTYPER assay, we derived quantitative, single CpG resolution methylation maps for our CNSL and peripheral PTLD biopsies (Fig. 3A, Suppl. Fig. 3A). We found that significant methylation reduction in CNSL is localized specifically to three differentially methylated regions (DMRs), with each DMR having at least one single CpG dinucleotide with pronounced methylation loss when comparing CNSL and peripheral PTLD (Fig. 3B). Localized BGLF4 methylation loss was also observed in a subset of peripheral PTLDs (Suppl. Fig 3A), with over one third (9/24, 38%) having at least one DMR with a methylation level below 50%. Comparison of histologic subtypes in CNSL and peripheral PTLDs highlighted a predominance of Diffuse Large B cell lymphoma (DLBCL)-type PTLDs, with significant methylation differences at BGLF4 CpGs retained when comparing CNS- and peripheral DLBCLs only (Suppl. Fig. 3B-D). All three DMR CpGs displayed inverse correlation with BGLF4 RNA expression (Fig. 3C-E), thus serving as correlates of the initially observed differential methylation. When categorizing PTLD tissue samples based on the number of hypomethylated (<50% methylation) BGLF4 DMRs, we found that BGLF4 RNA expression increased with the number hypomethylated CpG sites (Fig. 3F). Presence of two or more BGLF4 DMR CpGs below 50% methylation called high (4^th quartile) BGLF4 expression with high sensitivity and specificity (Suppl. Fig. 3E). In summary, methylation analysis of the BGLF4 upstream region pointed to highly site-specific methylation reduction associated with BGLF4 expression.

• Download figure
• Open in new tab

Figure 3. Site-specific methylation loss at EBV BGLF4.

(A) Expanded methylation analysis of the BGLF4 upstream locus using EpiTYPER assay. EBV genomic coordinates and statistical significance for individual CpGs are annotated above. Data show mean±s.e.m. DNA methylation in CNSL (n=12) and peripheral PTLDs (n=24) cohorts. Lower panel: CpG islands (green), EBV transcripts (blue) and DMRs (orange) are annotated. (B) DNA methylation at three individual DMR CpGs marked in (A). (C-E) correlation of DNA methylation and BGLF4 RNA expression in three DMR CpGs of the BGLF4 upstream locus. (F) Classification of BGLF4 methylation based on number of DMRs with lowered methylation and corresponding BGLF4 RNA expression. *p<0.05, **p<0.01, ****p<0.0001, t-test. (A,B: Exact Wilcoxon-Mann-Whitney test).

BGLF4 DMRs have gene promoter activity and mark three distinct BGLF4 transcription start sites

The identification of three DMRs upstream of the BGLF4 open reading frame and their association with gene expression pointed to the presence of gene regulatory elements in that region. We tested the functional properties of BGLF4 upstream elements in a dual luciferase reporter assay and found that all three DMR regions carried gene promoter activity significantly above that of a control minimal promoter (Fig. 4A). These activities were frequently increased when utilizing a HEK293 model carrying latent EBV (Suppl. Fig. 4A). We confirmed the influence of DNA methylation on promoter activity by using in vitro CpG-methylated reporter constructs. DNA methylation significantly decreased DMR promoter activity (Fig. 4B). In order to identify whether the observed regions of promoter activity give rise to RNA transcripts, we next carried out 5’ rapid amplification of cDNA ends (RACE). Using amplification of cDNA ranging from the BGLF4 start codon into upstream sequences, we identified three distinct BGLF4 TSS within DMR2 and DMR3 in n=5 primary CNSL samples (Fig. 4C-G) and n=4 EBV cell models (Suppl. Fig. 4B-E, Suppl. Table 3). Combining data from our EBV cell models and CNSL biopsies, we showed that TSS activity was strongly associated with CpG methylation immediately adjacent each TSS (Fig. 4H). This association was confirmed in all three TSS sites individually (Suppl. Fig. 4F) and when including methylation in a 25bp window around TSS (Suppl. Fig. 4G). Together, our data revealed promoter activity in BGLF4 upstream DMR elements that give rise to distinct viral BGLF4 transcripts in CNSL.

• Download figure
• Open in new tab

Figure 4. BGLF4 DMRs possess gene promoter activity.

(A,B) Luciferase reporter assay of BGLF4 upstream DNA elements alone (A) or after in vitro DNA methylation or reporter constructs (B). DNA sequences representing DMRs are marked in orange. EBV genomic coordinates, CpG islands (green) and EBV transcripts (blue) are indicated above. Bar graphs show mean±standard deviation of quadruplicate readouts in HEK293 cells. (C-G) Results from 5’RACE in n=5 CNSL samples with high BGLF4 RNA expression. Plots show DNA methylation across the BGLF4 upstream region in each sample with detected TSS marked by vertical lines. Lower panels: CpG islands (green), EBV transcripts (blue), DMRs (orange) and sequenced transcripts from 5’RACE (grey). (H) Violin plot depicting DNA methylation at CpG sites closest to TSS with either detectable (n=14) or undetectable (n=13) transcript. Data are aggregate readouts from n=5 CNSL samples from (C-G) and an additional n=4 EBV+ cell lines (Suppl. Fig. 4B-E). *p<0.05, ****p<0.0001, t-test (H: Exact Wilcoxon-Mann-Whitney test).

TET activity is required for the establishment of lowered BGLF4 DMR methylation and BGLF4 expression

We next sought to determine which host epigenetic factors may be required for the establishment of BGLF4 DNA methylation patterns. Site-specific loss of DNA methylation in the human genome is mediated by the ten-eleven translocation (TET) family of DNA demethylases (32). TET2 was reported to have profound impact on EBV latency and global gene expression patterns (33,34). To test whether TET2 may also be required for methylation regulation at BGLF4, we implemented a model that recapitulates the establishment of DNA methylation patterns on EBV following host cell infection and latency establishment. For that we used HEK293 cells and the epitheliotropic EBV strain M81 (Suppl. Fig. 5A). We tested the impact of biallelic CRISPR knockout of TET2 in that setting (Suppl. Fig. 5B). We again used the EpiTYPER assay to interrogate DNA methylation at BGLF4. Comparison of EBV infection in control or biallelic TET2 knockout (TET2−/−, Fig. 5A) revealed that a region around the active BGLF4 TSS significantly gained DNA methylation in a TET2−/− setting (Fig. 5B). This coincided with significantly decreased BGLF4 RNA expression (Fig. 5C). We found that TET2 knockout did not significantly alter the global EBV methylation average (Fig. 5D, Suppl. Fig. 5C). We next validated these findings by establishing a second model of TET deficiency that relies on overexpression of mutant isocitrate dehydrogenase 1 (IDH1 R132H) which impairs all three TET family members by generating an alternative metabolite, alpha-hydroxyglutatrate (35). Based on this mechanism, IDH1 R132H provides a distinct mode of TET inhibition that targets enzymatic activity directly. We found that lentiviral overexpression of IDH1 R132H (Suppl. Fig. 5D) followed by EBV M81 infection in HEK293 cells elevated DNA methylation and lowered RNA expression at BGLF4 (Suppl. Fig. 5E-G). In contrast to TET2 knockout, IDH1 R132H expression also resulted in a modest but significant increase in global EBV methylation (Suppl. Fig. 5H,I). Taken together, we deployed two distinct models of TET deficiency and found that the site-specific decrease of BGLF4 methylation requires TET2.

• Download figure
• Open in new tab

Figure 5. TET2 knockout increases BGLF4 promoter DNA methylation and lowers BGLF4 expression.

(A) Methylation analysis of the BGLF4 upstream locus using EpiTYPER assay in M81 EBV HEK293 cells (blue: parental, green: lytic induction, brown: EBV infected HEK293 wild type control, red: EBV infected HEK293 TET2−/−). EBV genomic coordinates and statistical significance for individual CpGs are annotated above. Lower panel: CpG islands (green), EBV transcripts (blue) and DMRs (orange) are annotated. (B,C) BGLF4 DNA methylation average from DMR3 CpGs (B, ChrEBV:111757, 111785, 111818) and BGLF4 RNA expression (C) in parental, wild-type reinfected control or TET2−/− re-infected HEK293 cells. (D) EBV DNA methylation average from n=31 EBV CpG sites determined via iPLEX assay in parental, lytic induced, re-infected or TET2−/− re-infected HEK293 cells. Data show mean±standard deviation of biological triplicates. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, t-test (Fig 4A: Exact Wilcoxon-Mann-Whitney test).