Subgrouping of Burkitt lymphoma variants by DNA methylation is driven by an EBV‐associated epigenotype

Introduction: Burkitt lymphoma (BL) is an aggressive mature B-cell lymphoma, which traditionally has been separated into three epidemiologic variants: endemic (eBL), sporadic (sBL) and immunodeficiency-associated BL (iBL). This subgrouping is strongly confounded by EBV-infection. Indeed, there is evidence from molecular studies that subtyping of BL based on EBV status might better reflect the pathogenetic heterogeneity than epidemiology. As EBV is known to influence epigenetic regulation, we here aimed at a comparative DNA methylation (DNAme) profiling of BL variants with respect to their geographic origin and EBV status. Methods: We collected DNAme profiles (Infinium HumanMethylation450 and Infinium MethylationEPIC BeadChip) of 116 BL patients (80 sBL: 7 EBV+; 29 eBL: 27 EBV+; 7 iBL: 4 EBV+), 17 BL cell lines and 6 EBV-transformed LCLs. For group determination unsupervised clustering algorithms (K-means, PGMRA, UMAP) were applied. To investigate potential consequences of EBV-associated DNAme we performed a protein-protein interaction network analysis on genes within the EBV-specific DNAme pattern. Moreover, we mined gene expression values from RNAseq in 21 EBV- sBL and 5 germinal-center B-cell populations. Results: Unsupervised clustering algorithms revealed two distinct DNAme phenotypes, which differentiate EBV+ and EBV- BL based on 1,266 CpGs. This separation is accompanied by a characteristic hypermethylation in EBV+ BLs. Additionally, we identified a subgroup of EBV+ BL with a similar DNAme pattern like the LCLs, possibly indicating latency phase differences within EBV+ BLs. Investigation of the EBV-specific DNAme pattern revealed a network with UBC as a key interactor, indicating the potential importance of the ubiquitin-proteasome system (UPS) in EBV-induced lymphomagenesis. Among genes differentially methylated between EBV+ and EBV- BL, 19 genes have been reported as recurrently mutated in BL. All except 3 of these genes are expressed in sBL and germinal-center B-cells. All these genes were significantly hypermethylated within regulatory regions in EBV+ BLs (σ/σmax = 0.4, q ≤ 0.01), including genes with a lower mutational frequency in EBV+ versus EBV- BLs. Conclusion: The findings of this study show the EBV status to be strongly associated with DNAme subgroups in BL, supporting the subtyping of BL variants based on their EBV status rather than their geographic origin. Moreover, the pattern of DNAme in EBV+ BL might suggest epigenetic silencing and deregulated ubiquitination as means alternative to gene mutation to be involved in the pathogenesis of EBV+ BL. Keywords: Aggressive B-cell non-Hodgkin lymphoma, Genomics, Epigenomics, and Other -Omics, Pathology and Classification of Lymphomas No conflicts of interests pertinent to the abstract.