Methylation Fingerprints of Colorectal Cancer: From Normal Tissue to Liver and Brain Metastasis
Annual Congress of the European Association for Cancer Research (EACR 2026), Budapest, Macaristan, 8 - 11 Haziran 2026, ss.1, (Özet Bildiri)
- Yayın Türü: Bildiri / Özet Bildiri
- Basıldığı Şehir: Budapest
- Basıldığı Ülke: Macaristan
- Sayfa Sayıları: ss.1
- Karadeniz Teknik Üniversitesi Adresli: Evet
Özet
Introduction
Colorectal
cancer (CRC) progression from normal mucosa to metastasis involves extensive
epigenetic reprogramming. Beyond genetic alterations, DNA methylation is
critical for tumor initiation and the acquisition of organotropic metastatic
phenotypes. This study systematically defines conserved and divergent
methylation signatures across the adenoma-carcinoma-metastasis sequence to
identify core gene sets and biomarkers for organ-specific progression.
Material
and method
DNA
methylation data (GSE28094) were retrieved from GEO. Samples included normal
colon (n=97), adenoma (n=12), primary tumor (n=110), liver metastasis (n=32),
and brain metastasis (n=13). Pairwise differential methylation was assessed
using the limma package (FDR<0.05). Core targets shared across all disease
groups were analyzed via GO-BP/KEGG enrichment, STRING-based protein-protein
interaction (PPI) networks and hub gene identification. Diagnostic performance
was evaluated using receiver operating characteristic (ROC) analysis.
Results
and discussion
Significant
differentially methylated genes (DMGs) relative to normal tissue were 935 for
adenoma, 743 for primary tumor, 772 for liver metastasis, and 991 for brain
metastasis. Adenoma was predominantly hypomethylated (819/935), whereas
hypermethylation was more frequent in primary tumors and liver metastases.
Brain metastases showed significant divergence from both primary and liver
lesions (707 and 720 targets, respectively), supporting organotropic epigenetic
differentiation. We identified a core set of 114 Hyper and 138 Hypo targets
shared across all stages. Enrichment analysis linked the Hyper set to
epithelial proliferation and MAPK signaling, while the Hypo set was associated
with gland morphogenesis, ECM remodeling, and hypoxia. PPI networks revealed
Hyper hubs centered on angiogenesis axes (FGF8/FGFR1, KDR/FLT4) and Hypo hubs
driven by EGFR, CTNNB1, and IL6. ROC analysis identified FGF8 as a superior
Hyper hub (AUC=0.958) with CTNNB1 (AUC=0.836) and HGF (AUC=0.803) as leading
Hypo hubs.
Conclusion
Our
findings demonstrate a conserved core epigenetic signature across CRC stages.
The distinct methylation profiles in brain metastasis suggest that metastatic
adaptation is epigenetically regulated. These hub genes and high-AUC regions
represent robust biomarker candidates for early diagnosis and stage-specific
targeted therapies.
Acknowledgements
We
acknowledge the authors who generated and deposited GSE28094 (Fernandez et al.)
and GEO/NCBI for providing access to the publicly available data used in this
study.