Modulation of Cellular CpG DNA Methylation by Kaposi's Sarcoma-Associated Herpesvirus

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV, HHV-8) is a gammaherpesvirus associated with several human malignancies. DNA methylation at CpG dinucleotides is an epigenetic mark dysregulated in many cancer types and in KSHV-infected cells. Several previous studies have analyzed in detail the CpG methylation of the KSHV episomal genomes, but little is known about the impact of KSHV on the human genome. Our knowledge of cellular CpG methylation in the context of KSHV infection is currently limited to four hypermethylated human gene promoters. Therefore, we undertook a comprehensive CpG methylation analysis of the human methylome in KSHV-infected cells and KSHV-associated primary effusion lymphoma (PEL). We performed Infinium HumanMethylation450K and MethylationEpic BeadChip arrays and identified panels of hyper- and hypomethylated cellular promoters in KSHV-infected cells. We combined our genome-wide methylation analysis with high-throughput RNA sequencing (RNA-seq) to add functional outcomes to the virally induced methylation changes. We were able to correlate many downregulated genes with promoter hypermethylation and upregulated genes with hypomethylation. In addition, we show that treating the cells with a demethylating agent leads to reexpression of these downregulated genes, indicating that, indeed, DNA methylation plays a role in the repression of these human genes. Comparison between infection and PEL suggests that the virus induces initial hypermethylation followed by a slow increase in genome-wide hypomethylation. This study extends our understanding of the relationship between epigenetic changes induced by KSHV infection and tumorigenesis. In cancer cells, certain promoters become aberrantly methylated, contributing to the phenotype of the tumor. KSHV infection seems to modify cellular CpG methylation, but only a few methylated promoters have been identified in KSHV-infected cells. Here, we investigated the CpG methylation of the human genome in KSHV-associated primary effusion lymphoma (PEL) and KSHV-infected cells. We have identified many hyper- and hypomethylated gene promoters and correlated their methylation with cellular gene expression. These differentially methylated cellular promoters can distinguish KSHV-positive cells from uninfected cells and may serve as the foundation for the use of these differentially methylated regions as potential biomarkers for KSHV-associated malignancies. Drugs that reverse these cancerous methylation patterns have the potential to inhibit tumor growth. Here, we show that treating PEL cells with a demethylating drug (5-aza-2'-deoxycytidine) led to inhibition of cell growth, raising the possibility of testing this drug for the treatment of PEL.

High-throughput sequencing analysis of a “hit and run” cell and animal model of KSHV tumorigenesis

“…It is important to mention that the goal of this study was to identify the changes during tumorigenesis and not following infection, therefore the hyper-methylation immediately following infection is already present in the KSHV (+) cells. These results are in agreement with a previous study [20] that tried to address this question indirectly by comparing KSHV-negative BJAB to de-novo infected BJAB (BJAB.219) or to chronically infected PEL cells, and detected very similar hyper-methylation between de-novo infection and PEL, but dramatic difference towards hypo-methylation in PEL. While our study directly evaluates the methylation changes during transformation, both studies support the notion that hypo-methylation is the major epigenetic change during this process.…”

“…An additional mechanism by which KSHV might modify the human methylome is via the Polycomb complex that creates the histone mark histone H3 trimethylated on Lys27 (H3K27me3) and can direct cellular CpG methylation via its interaction with DNMTs [ 16 , 17 ]. The pattern of CpG DNA methylation in chronically infected primary effusion lymphoma (PEL) cells and during de-novo in-vitro infection was investigated for the KSHV episomal genome [ 17 – 19 ] and for the host cellular genome [ 20 ].…”

RNA-sequencing analysis of a multistep and hit-and-run cell and animal model of KSHV tumorigenesis reveal the roles of mutations, CpG methylation, and viral-infection footprints in oncogenesis

“…It is important to mention that the goal of this study was to identify the changes during tumorigenesis and not following infection, therefore the hyper-methylation immediately following infection is already present in the KSHV (+) cells. These results are in agreement with a previous study (Journo et al, 2018) that tried to address this question indirectly by comparing de-novo infection (in BJAB cells) to chronically infected PEL cells, and detected very similar hyper-methylation between de-novo and PEL, but dramatic difference towards hypo-methylation in PEL. While our study directly evaluates the methylation changes during transformation, both studies support the notion that hypo-methylation is the major epigenetic change during this process.…”

“…An additional mechanism by which KSHV might modify the human methylome is via the Polycomb complex that creates the histone mark histone H3 trimethylated on Lys27 (H3K27me3) and can direct cellular CpG methylation via its interaction with DNMTs (Gunther and Grundhoff, 2010; Schlesinger et al, 2007). The pattern of CpG DNA methylation in chronically infected primary effusion lymphoma (PEL) cells and during de-novo in-vitro infection was investigated for the KSHV episomal genome (Chen et al, 2001; Gunther and Grundhoff, 2010; Shamay et al, 2012a) and for the host cellular genome (Journo et al, 2018).…”

Genome-Wide Identification of Direct RTA Targets Reveals Key Host Factors for Kaposi’s Sarcoma-Associated Herpesvirus Lytic Reactivation

“…Further studies will be required to determine which host genes are directly induced by RTA or rather RTA-independent mechanisms at later time points of KSHV reactivation. In addition, the modifications of cellular chromatin, DNA methylation of the host genome, the presence of different host and viral TFs, and the depletion of RNA polymerase II on the host genome during KSHV lytic reactivation can all influence RTA binding on the host genome, whether or not RTA can induce its host target genes, and when it can induce them in the course of KSHV reactivation (17,23,26,28,35,42,(64)(65)(66)(67). Importantly, we found that several of the core RIGs are involved in the Notch signaling pathway (e.g., HEY1, HES1, and DLL4), which has been shown to play a role in KSHV pathogenesis (44,45,68).…”