Abstract B010: Long-read sequencing of high-grade serous ovarian carcinoma genomes: Insights into structural variants and allele-specific DNA methylation

Abstract High-grade serous ovarian carcinomas (HGSOC) have one of the highest burdens of focal copy number alterations amongst cancers studied by The Cancer Genome Atlas. These focal copy number alterations are driven by chromosomal instability and rearrangements. Our current understanding of structural variants (SVs) in HGSOC come largely from a combination of macroscale karyotyping and base-level short-read Illumina sequencing. However, short-read sequencing is inherently hindered in accurate SV calling due to limited information of exact breakpoints captured using this technology. With the development of long-read sequencing, including Oxford Nanopore Technologies (ONT) sequencing used here, SVs are resolved with higher accuracy. Additionally, ONT sequencing allows better haplotype phasing and information about 5-methylcytosine (5mC; gene silencing) and 5-hydroxymethylcytosine (5hmC; gene activation). Here we sought to examine SVs and allele-specific methylation (ASM) in HGSOC cell lines, organoids, and tumors. Our cell line data, from 7 HGSOC cell lines, sequenced the genome to an average coverage of ∼35x. Using the epi2me human variation pipeline (ONT), supplemented by additional long-read SV caller cuteSV and filtering with MAVIS, we identified Insertions, Deletions, Translocations, Duplications, and Inversions. Because our samples lacked germline calls, we leveraged coverage information from the gnomAD database across the Ten-Thousand Genomes Project to remove regions of frequent variation in the population, which greatly filtered the number of insertions and deletions. A strong peak of 300bp was observed in insertions and deletions, which was nearly entirely repetitive elements (SINE, LINE, LTR, and simple repeats), making up over 50% of these SVs. MAVIS annotates SVs predicted to make non-synonymous changes to protein sequences; of 95 non-synonymous SVs across 7 cell lines, 57 were examined by PCR, with successful validation of 85. 9%. Long-read sequencing was also haplotype phased, and DSS was used to determine regions of differential methylation between haplotypes. A mean of 828 promoters with ASM were observed, with about 40% shared between at least two cell lines. While 91 different tumor suppressor genes (TSG) had promoters with ASM were identified (mean 25 TSG per cell line), none were confirmed to have a mutation to the expressed haplotype. We are continuing to confirm whether ASM leads to bias in allelic expression of RNA, as well as expanding our study to examine long-read sequencing in organoids and tumor datasets. This work adds granularity to existing data on chromosomal rearrangements in HGSOC. Future directions of this work will examine how translocations between chromosomes influence novel enhancer-gene interactions. Citation Format: Jessica D. Lang, Ryan Moreno, Kristin M. Adams, Lauren Kersten. Long-read sequencing of high-grade serous ovarian carcinoma genomes: Insights into structural variants and allele-specific DNA methylation abstract. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Ovarian Cancer Research; 2025 Sep 19-21; Denver, CO. Philadelphia (PA): AACR; Cancer Res 2025;85 (18Suppl): Abstract nr B010.