Genome-Wide Assessment of DNA Methylation Fidelity Identifies Representative Ovarian Cancer Cell Line Models

Epithelial ovarian cancer comprises multiple distinct histotypes with different molecular, clinical, and therapeutic characteristics, yet ovarian cancer cell lines are often broadly classified as 'ovarian carcinoma' or ungraded 'serous'. While studies have classified cell lines using mutations, transcriptomics, or immunophenotypes, their DNA methylation profiles remain poorly characterised. Given the role of DNA methylation in epithelial ovarian cancer, histotype-specific methylation profiling is essential to determine how accurately cell lines model primary tumours and to guide their appropriate use. The Infinium 450 K and EPIC DNA methylation profiles of 60 unique ovarian cancer cell lines and 108 ovarian tumours were evaluated though a correlation and effect size analysis, focusing on genetic and epigenetic domains. DNA methylation patterns were found to be highly conserved between cell lines and tumour tissues at key transcriptional regulatory domains (CpG islands, shores, and 5' regulatory regions), with these CpG-dense loci more accurately reflecting in vivo conditions than CpG-sparse regions. We identified specific cell lines with representative epigenetic signatures of high-grade serous, clear cell, endometrioid, mucinous and mixed ovarian carcinoma histotypes. This study provides a framework to optimise the selection and application of ovarian cancer cell line models in DNA methylation research.