Pan-cancer nucleosome landscapes reveal tumor heterogeneity and inform cell-free DNA-based detection

Abstract

Background: Nucleosome positioning and occupancy are fundamental features of chromatin organization, yet their variation across human cancers and potential utility for noninvasive cancer detection remain poorly characterized. Methods: We systematically characterized nucleosome organization across 410 primary tumors spanning 23 cancer types using ATAC-seq, establishing a high-confidence pan-cancer nucleosome atlas validated by independent MNase-seq data. We characterized shared and cancer-type-enriched nucleosome patterns, investigated their associations with tumor heterogeneity and regulatory alterations, and evaluated tissue-derived nucleosome signatures for cancer classification and detection using circulating cell-free DNA (cfDNA). Results: The atlas identified shared and cancer-type-enriched nucleosomes associated with distinct regulatory programs. Notably, nucleosome occupancy provided information complementary to chromatin accessibility for tumor stratification and revealed molecular heterogeneity within individual cancers, including subgroups associated with Basal-like breast cancer. In hepatocellular carcinoma, differential nucleosome analysis revealed cancer-associated chromatin remodeling involving regulatory pathways, transcription factor motifs, and candidate genes. Importantly, tissue-derived nucleosome signatures were detectable in cfDNA, enabling classification among three cancer types (AUROC = 0.877) and colorectal cancer detection (AUROC = 0.930), with an AUROC of 0.773 in an independent validation cohort. We also developed OncoNucAtlas, a publicly accessible database facilitating exploration of these nucleosome landscapes. Conclusions: Our study cancer-associated nucleosome organization as a complementary source of epigenomic information and demonstrates the potential of tissue-derived nucleosome signatures for cfDNA-based cancer detection. The pan-cancer atlas provides a resource for investigating tumor heterogeneity and developing noninvasive cancer detection approaches.