Within-Host Diversity of Kaposi Sarcoma-associated Herpesvirus

Abstract

Kaposi sarcoma (KS) is a progressive, incurable soft tissue disease that is the most common cancer of men in regions of sub-Saharan African and the second most common among people with HIV. Kaposi Sarcoma-associated Herpesvirus (KSHV) is the infectious cause of KS, but why only a small fraction of those infected develop KS is not understood. Immune status, viral co-infection, host genetics and environmental factors all are likely to play a role. My thesis sought to identify a possible viral genetic component of KS, through an analysis of the diversity of whole KSHV genomes within individuals afflicted by KS. Strain variation or de novo mutations occurring in other oncoviruses have been associated with variations in disease risk, disease manifestation and clinical course. These variations have the potential to become diagnostic biomarkers and help reveal insights into the pathogenicity of the virus. To identify potential tumor-associated mutations in KSHV I analyzed the ~131 kb unique sequence regions of 43 whole KSHV genomes from tumors and oral swabs from 22 individuals and then screened additional tumors by targeted sequencing and ddPCR to better assess the frequency of identified mutations. In total, 65 KS tumors and 18 oral swabs were evaluated from a cohort of 30 study participants with KS, all from the Uganda Cancer Institute. In addition, the major internal repeat regions of the KSHV genome were examined from 16 individuals. The newly developed technologies of duplex Unique Molecular Identifier (dUMI) and single molecule real time sequencing with UMI (SMRT-UMI) were employed to provide an unprecedented accuracy and depth of study of herpesvirus populations within individuals. These studies revealed recurring tumor-associated mutations, specifically an overrepresentation of a specific region encompassing the K5 and K6 genes and inactivating mutations in the K8.1 gene, both of which were found in at least a third of the cohort and had associations with late-stage tumor characteristics. In contrast, the translation potential of full-length Kaposin proteins from IR2 was lost in a majority of individuals, but no tumor association or clinical correlate was identified. Rearrangement breakpoints were sometimes shared between different tumors from the same person, indicating spread by metastases, helper viruses or residual infectivity. In summary, these studies revealed the presence of selection for tumor- and clinical-stage specific changes the KSHV genome. Hypotheses on whether these changes play a role in KS tumor formation, immune evasion and driving persistence in the host are put forth along with proposed experiments to test them. Among them, findings from my thesis suggest that targeted screening of the K5-K6 and K8.1 gene regions can be helpful in tumor classification.