Stochastic mutagenesis at lncBART-KDM5B-bound genomic loci in Epstein-Barr virus-associated nasopharyngeal carcinoma

Introduction: The mechanisms linking Epstein-Barr virus (EBV) infection to mutagenesis in nasopharyngeal carcinoma (NPC) remain poorly understood. Methods: By integrating whole-genome sequencing of 216 NPC tumors with multi-omics profiling, including chromatin isolation by RNA purification (ChIRP)-seq for lncBARTs binding sites, CUT&RUN for the binding sites of KDM5B and CTCF and histone mark H3K27ac, 4C-seq for virus-host chromatin interactions, and ATAC-seq for chromatin accessibility, we uncover a multi-layered model of virus-driven mutagenesis. Results: We first show that cellular epigenetic regulators CTCF and KDM5B independently define genomic domains of significantly elevated mutation burden (2-fold more than matched controls, P<0.0001). Crucially, the viral lncBART does not increase mutation rates but fundamentally reshapes their distribution, randomizing mutations into a stochastic, Poisson-distributed pattern within KDM5B-bound domains (p=0.3665), which otherwise exhibit clustered, non-random mutagenesis (p<0.001). Furthermore, the lncBART binding sites enriches specific mutation signatures, namely G>T transversions (2.1-fold increase, p=0.005) and G>A transitions (1.3-fold increase, p=0.05), indicative of oxidative damage and replication stress, without altering the overall mutation density. We developed a computational model that predicts mutation patterns based on CTCF, KDM5B, and lncBART binding together with enhancer status and chromatin accessibility, demonstrating that these factors partially define the landscape of genomic instability. Incorporation of the data from ChIRP-seq and Capture RNA-Protein interactions (CARPID) following mass spectrometry analysis revealed that lncBARTs may act as a central scaffold that directly co-recruits key members of NuRD complex to create chromatin states against large processive mutagens and blockade of the base excision repair (BER) machinery. Conclusions: Our work establishes a novel paradigm in which a virus co-opts and modulates host epigenetic machinery to create a context-dependent mutagenic environment, providing a new framework for understanding virus-driven cancer evolution. Acknowledgement: This work was supported by General Research Fund (17101122) and Theme-based Research Scheme (T12-703/22-R; T123-70323-N) from Research Grant Council and Health@InnoHK from Innovation and Technology Commission in the government of Hong Kong (SAR), P. R. China