Precise identification of viral-host integration events in HPV-positive cervical cancers by targeted long-read sequencing.

Human papillomaviral (HPV) integrations into host human genome, a frequently observed event in HPV associated cervical cancer, are currently mapped through expensive Whole Genome sequencing (WGS) or RNA sequencing (RNA-seq) methodologies. This study aims to develop a targeted sequencing assay to determine HPV integrations in cervical tumors without the need for WGS or RNA-seq. We employed a library preparation strategy using tiled single primers that bind to HPV genome as a template and possibly extend HPV sequences into adjacent host human genomic sequences resulting in HPV and human chimeric sequences. Using this strategy, we sequenced known HPV integrations in HPV18 positive HeLa and HPV16 positive SiHa cell lines. We further used this method to detect HPV integration sites in four HPV-positive cervical cancer patients and confirmed these integration breakpoints by WGS and Sanger sequencing. Functional impact of HPV integrations was explored through differentially expressed genes within or near topologically associating domain (TAD) boundaries, possibly disrupted by respective integration events in these patients. We found ZFP36L1, CPA3, CPB1 and CXCL8 as some of the differentially expressed genes within disrupted TADs, which are known cancer associated genes. Our approach also reduced the cost of HPV integration detection by 90 % compared to WGS while also minimizing sequencing data volume. We believe that this method captures HPV integrations at significantly reduced costs and lesser sequencing data volume leading to better understanding of disease progression and monitoring cancer treatment.