Methylation supernegative colorectal cancers - a key to unlocking the secrets of the DNA (de)methylation machinery

Although the cells in our body basically contain the same DNA, they have different functions and appearances depending on which genes they express. DNA methylation is a chemical modification that helps regulate the gene expression and ultimately decides whether they are turned on or off. In cancer, including in colorectal cancer, hundreds to thousands of genes are usually inactivated by such DNA methylation. In the present project, we have identified a small subgroup of colorectal cancers which seems to have unusually little DNA methylation. These `methylation supernegative` cancers represent a unique opportunity for gaining insights into the inner workings of our DNA methylation machinery. In order to provide comprehensive insights into this intriguing subgroup, we have combined cutting-edge technologies with advanced computational methods. In the discovery phase of the project, we have generated DNA methylation data of the entire genome, using the state-of-the-art Whole-Genome Bisulfite Sequencing technology, in collaboration with Johns Hopkins University. We have confirmed that a subset of colorectal tumors display genome-wide reduced DNA methylation levels compared to both 'regular' colorectal tumors and samples from healthy donors, and constitute a novel subgroup. To further characterize the relevant features associated to the methylation supernegative subgroup and identify candidate biomarkers, i.e. 'labels' of the supernegative tumors that can be used for stratifying colorectal cancers, we have generated additional data from a larger set of samples using an alternative technology (reduced representation bisulfite sequencing). With this approach, we are targeting regions of the genome which are functionally important, i.e. regions which are involved in gene regulation, at a reduced cost, allowing larger sample series to be analyzed. In collaboration with the research group of Ragnhild Lothe and the K.G Jebsen center for colorectal cancer we have combined this data with gene expression profiles (microarrays) for the same cancers. We can demonstrate that the reduced levels of DNA methylation rate are associated with reduced expression of DNA methyltransferases, a family of enzymes responsible for adding methyl groups to the DNA. Interestingly, our data also suggests that methylation supernegative tumors are 'immunogenic', i.e. they are able to induce an immune response, which is in contrast to most colorectal cancers. This implies that the supernegative colorectal cancer patients may benefit from immunotherapy. In parallel, we have established in the lab a new and valuable technology in form of the highly sensitive and absolute quantitative digital droplet PCR (ddPCR). To ensure a highly-standardized pipeline for downstream biomarker validation, we have optimized ddPCR for targeted analysis of DNA methylation levels and provided an accompanying algorithm for data analysis (Pharo et al. 2018, Clinical Epigenetics; PMID 29484034, Pharo H. et al, Scientific Reports, 2016; PMID 27671843). This work has resulted in two patent applications and a user-friendly desktop application is work in progress. The results from this project, in addition to bringing novel insights into DNA methylation and cancer biology, may provide new guidance on the clinical management of colorectal cancer patients. Methylation supernegative biomarkers can be used for improving tests for early detection and prognostication of colorectal cancer, and thereby potentially improve the survival of this patient group. Furthermore, we have demonstrated that the methylation level in colorectal cancers is associated with patient survival (Vedeld HM. et al, Int J Cancer, 2017; PMID: 28542846). This identifies a group of patients that could benefit from a more aggressive cancer treatment.

The project has advanced our knowledge of colorectal cancer and enabled to characterize a new epigenetic phenotype. It could establish the bases for improving tests for early detection and prognostication of colorectal cancer, and thereby potentially improve survival of this patient group. In addition, data indicates immunogenicity among the supernegative patients, implying that these patients may benefit from immunotherapy. These findings may have a significant impact in the field, both for the scientific community and also for the society and the patients who could have access to better health care. The supernegative subtype could be extended to other cancer types and contribute to improved stratification of cancer patients in general (future perspectives). The project has also required method optimization and standardization. In particular, we are providing the research community with a new computational tool for normalization and analysis of ddPCR methylation data.

Cytosine methylation is essential for the proper epigenetic regulation of gene expression and maintenance of genomic integrity. We aim at expanding existing knowledge about the mechanism of the DNA methylation and demethylation machineries by a systematic analysis of a novel and intriguing subgroup of cancers, which have unusually little promoter hypermethylation. These tumors represent a unique opportunity for direct insight into the molecular and cellular (dys)functions of the DNA (de)methylation machinery and will be studied at several levels by a multi-disciplinary team, using state-of-the art technology and bioinformatics. In addition to novel biological insights, the results from this project may be useful in the clinical management of colorectal cancer patients. Recurrent aberrations among the methylation supernegative tumors represent novel biomarkers that can improve the sensitivity of existing DNA methylation based test for early detection of colorectal cancer, and ultimately the survival of this patient group. Finally, we aim at identifying novel prognostic markers that may improve the stratification of colorectal cancer patients