Background: DNA methylation (DNAm) influences both sex differences and cancer development, yet the mechanisms connecting these factors remain unclear.
Methods: Utilizing data from The Cancer Genome Atlas, we conducted a comprehensive analysis of sex-related DNAm effects in nine non-reproductive cancers, compared to paired normal adjacent tissues (NATs), and validated the results using independent datasets. First, we assessed the extent of sex differential DNAm between cancers and NATs to explore how sex-related DNAm differences change in cancerous tissues. Next, we employed a multivariate adaptive shrinkage approach to model the covariance of cancer-related DNAm effects between sexes, aiming to elucidate how sex impacts aberrant DNAm patterns in cancers. Finally, we investigated correlations between the methylome and transcriptome to identify key signals driving sex-biased DNAm regulation in cancers.
Results: Our analysis revealed a significant attenuation of sex differences in DNAm within cancerous tissues compared to baseline differences in normal tissues. We identified 3,452 CpGs (Pbonf < 0.05) associated with this reduction, with 72% of the linked genes involved in X chromosome inactivation. Through covariance analysis, we demonstrated that sex differences in cancer are predominantly driven by variations in the magnitude of shared DNAm signals, referred to as "amplification." Based on these patterns, we classified cancers into female- and male-biased groups and identified key CpGs exhibiting sex-specific amplification. These CpGs were enriched in binding sites of critical transcription factors, including P53, SOX2, and CTCF. Integrative multi-omics analyses uncovered 48 CpG-gene-cancer trios for females and 380 for males, showing similar magnitude differences in DNAm and gene expression, pointing to a sex-specific regulatory role of DNAm in cancer risk. Notably, several genes regulated by these trios were previously identified as drug targets for cancers, highlighting their potential as sex-specific therapeutic targets.
Conclusions: These findings advance our understanding of how sex, DNAm, and gene expression interact in cancer, offering insights into the development of sex-specific biomarkers and precision medicine.
Keywords: Cancer; DNA methylation; Gene expression; RNA-seq; Sex differences.
Sex disparities in non-reproductive cancers are well-documented across various aspects, including incidence, survival, mortality, and treatment outcomes. A deeper understanding of these differences could support the development of personalized therapeutic strategies. In this study, we conducted a comprehensive analysis of sex-related DNA methylation (DNAm) effects in nine non-reproductive cancers, comparing cancer tissues with paired normal adjacent tissues (NATs). Our findings revealed that DNAm differences between males and females were significantly reduced in cancerous tissues. The CpGs associated with this reduction were linked to pathways involving the tumor microenvironment. Additionally, we found that these sex differences in cancer were primarily driven by variations in the effect sizes of shared DNAm signals, allowing us to classify cancers into male-biased (BLCA, THCA, KIRP, LUAD, and HNSC) and female-biased (LUSC, LIHC, COAD, and KIRC) groups. Notably, we found that differential expression in cancers was correlated with differential methylation in a sex-specific manner, through concordant magnitude differences within each sex. Several of these genes, regulated by DNAm changes, were already targets of cancer drugs (e.g., ECSCR, GATA2, and ERBB3), highlighting the potential for developing sex-specific treatments. Overall, this research enhances our understanding of the role of DNAm in cancer and could contribute to more personalized therapies for both males and females.
© 2024. The Author(s).