Deciphering Virus-Mediated Remodeling of the Lymphoma Tumor Microenvironment Through Spatial Multi-Omics

Abstract

Oncogenic viruses are associated with ~10% of cancers and play key roles in tumor development, yet how they remodel the tumor microenvironment to drive immune evasion and shape treatment responses remains poorly understood. Consequently, viral status remains mostly diagnostic rather than prognostic, leading to identical treatments against virus-associated cancers, despite growing evidence of different treatment responses. This dissertation addresses this gap by focusing on Epstein-Barr Virus (EBV), the prototypical human oncogenic virus that infects >90% of the population and is associated with several aggressive lymphoid malignancies. These include classic Hodgkin Lymphoma (cHL) and Diffuse Large B-cell Lymphoma (DLBCL), with EBV positivity in ~25% and 5-15% cases respectively. Through spatial proteomics and transcriptomics, EBV was found to be major factor impairing antitumor immunity by spatially reorganizing the tumor microenvironment. In cHL, EBV-positive tumors were distinctly enriched in cytotoxic CD8 T-cells surrounding malignant cells that retained MHC Class I, yet these T-cells were reprogrammed into terminal exhaustion through direct engagement with malignant cells, a process associated with the viral oncogene LMP1. In DLBCL, EBV disrupted a macrophage:CD4:tumor immunomodulatory axis leading to down-regulated MHC Class II and suppressed CD4 T-cell activity, a process also linked to LMP1. In parallel, several technological advancements were made: IN-DEPTH, a same-slide spatial multi-omics method enabling simultaneous proteomics and transcriptomics profiling; SGCC, a computational tool developed by colleagues at Ohio State University that uses graph Fourier transform to detect spatially coordinated gene-regulatory programs between cell populations; and the incorporation of proximity labeling to detect molecular interactions in situ. The findings presented herein demonstrate how EBV exploits distinct immunoevasive mechanisms across lymphomas, revealing therapeutic vulnerabilities overlooked by current virus-agnostic treatment approaches. The methodological innovations further expand the spatial biology toolkit with broadly applicable experimental and computational strategies. By integrating virology, immunology, cancer biology, and spatial biology, this dissertation also represents a framework for future systems-level investigations across diverse contexts.