A recent breakthrough in cancer research comes from the Garvan Institute, where scientists have used artificial intelligence to discover potentially cancer-causing elements in regions of the genome once considered “junk DNA.” These non-coding regions, which make up 98% of our DNA, could transform how we diagnose and treat cancer.
Image d’illustration Pixabay
The researchers published their findings in the journal Nucleic Acids Researchhighlighting mutations in these overlooked regions of the genome. These mutations are thought to be involved in the formation and progression of several cancers, including prostate, breast and colon. This discovery paves the way for early diagnosis and new treatments for many types of cancer. Dr. Amanda Khoury, lead author of the study, explained that non-coding DNA, once considered useless, could offer a universal approach to treating cancer. By targeting these common mutations, researchers hope to develop more effective therapies.
The scientists focused on CTCF protein binding sites, which play a crucial role in the three-dimensional organization of the genome. Their disruption could lead to gene dysregulation and promote cancer development. To test this hypothesis, they used a machine learning tool called CTCF-INSITE to predict persistent CTCF binding sites in 12 types of cancer.
Using this tool, they analyzed more than 3,000 tumor samples, revealing that CTCF binding sites were rich in mutations in all cancer samples. These mutations would give cancer cells a survival advantage, allowing them to proliferate and spread.
Professor Susan Clark, Head of the Cancer Epigenetics Laboratory at the Garvan Institute, believes the findings could have broad implications for the treatment of many cancers. Current treatments must be tailored to specific mutations, which are often rare across different tumour types, but CTCF binding sites provide a common target for many cancers.
Next steps in the research will include large-scale experiments using CRISPR gene editing to study how these mutations affect the 3D structure of the genome and potentially promote cancer growth. This approach could lead to the development of biomarkers for early cancer detection or targets for new treatments.
