A-108 HTLV-1/BLV genomic approaches to understanding ATL

Abstract

Human T-cell Leukemia Virus (HTLV-1) was the first infectious agent discovered to be the direct cause of cancer and is the most carcinogenic of all oncoviruses. The retrovirus causes Adult T-cell leukemia (ATL) an aggressive CD4+ T-cell malignancy which has a median survival of 8–10 months for the acute sub-types, despite all advances in therapy. HTLV-1 remains a strong threat to individual and community health in endemic regions, and even more so to global health because of the accelerated rate of human migration in recent times. Bovine Leukemia Virus (BLV), a close relative of HTLV-1, induces a similar disease affecting the B-cell lineage in cattle. Both viruses produce a chronic infection in their respective host that evolves into full-blown leukemia/lymphoma in ∼5% of infected individuals after several decades of latency. While not a natural host, it is possible to infect sheep with BLV. In contrast to cattle, all infected sheep develop tumors at an accelerated rate (∼24 months), providing a unique model for studying early asymptomatic stages. Historically, research into both viruses has primarily focused on virus-encoded transcripts/proteins, especially Tax and HBZ. However, increasing evidence from genomic studies suggests that both the proviral integration site and somatic alterations within the host genome play a critical role in oncogenesis, as only a subset of infected individuals, following a long period of asymptomatic infection, develop a tumor. Using genomic approaches, we demonstrated that HTLV-1 and BLV integrate in the vicinity of host cancer driver genes, which they perturb either by provirus-dependent transcription termination or as a result of viral antisense RNA-dependent cis-perturbation via virus-host chimeric transcripts. This was revealed by applying novel NGS-based methods to primary tumors, enabling (1) genome-wide mapping of proviral integration sites and (2) targeted capture RNA-seq with increased sensitivity. To address the role of proviral integration and insertional mutagenesis at early stages, we set up a retrospective longitudinal study of infected individuals who developed aggressive leukemia, tracking back their evolution during asymptomatic stages in the BLV animal model. NGS clonality at polyclonal time points revealed hotspots of proviral integration and allowed us to monitor clonal architecture over time, providing novel insights into tumor evolution. The novel NGS clonality approach includes several critical modifications, overcoming some of the limitations of previously published protocols mainly in terms of sensitivity, cost and hands-on time, providing a new molecular tool for monitoring patients in a clinical setting. In a longitudinal study of ATL patients, we found that the optimized NGS mapping approach outperformed any other currently available method, enabling a better estimate of molecular response in leukemia patients, the detection of refractoriness to first-line therapy and a better prediction of relapse. Overall, we provide a proof-of-concept for the integration of a quantitative molecular HTLV-1-clonality signature into response assessment criteria for the monitoring of ATL patients and the follow-up of clinical trials that remain critical in the future.