Dissecting Regulatory Mechanisms Underlying the Poor Prognosis Colon Cancer Subtype

Description

Colorectal cancer, or colon cancer, is a form of cancer that occurs in the colon or rectum. It isthe 3rd most common cancer in men and 2nd in women worldwide, accounting for ~8% of allcancer death. According to figures published by the World Health Organization, bothincidence and mortality have an increasing trend in eastern Asian countries during the lastfew decades. Thus, apart from better detection and prevention at precancerous and earlycancer stages, it becomes urgent to comprehensively understand the molecular biology ofcolon cancer. One key challenge lies in the fact that colon cancer is molecular and clinicaldiverse, i.e., there are multiple types of colon cancers that differ in molecular properties,clinical outcomes as well as response to drugs. How to stratify colon cancer patients tomolecular distinct subgroups, in relation to clinical features, is critical for decision makingfor therapy and more targeted drug design. In our previous work, we have reported such astratification based on unsupervised classification on whole-genomegene expression profilesof colon cancer patients. We identified three colon cancer subtypes, of which the thirdsubgroup (CCS3) is associated with much poorer prognosis, but cannot be identified bycharacteristics such as KRAS, BRAF and TP53 mutations.The ultimate goal of this project is to obtain a full landscape of the regulatory mechanismsunderlying the poor prognosis colon cancer subtype (CCS3). Step by step, we aim to achievethe following major objectives in the next 3 years: (1) perform a comprehensive molecularcharacterization of CCS3 colon cancer based on integrative analysis of multiple types of(epi)genomic data. (2) identify genetic and epigenetic prognostic biomarkers for coloncancer, based on survival analyses of molecular characteristics resulted from the firstobjective. (3) Reconstruct the regulatory network underlying CCS3 colon cancer, taking intoaccount transcriptional, posttranscriptionaland posttranslationalregulations. (4) Prioritizepotential drug targets for CCS3 colon cancer based on probabilistic modeling of causaleffects of (epi)genetic aberrations on downtream pathways driving colon cancer development.From public databases, we have prepared genomic, clinical and drug response data for a largecohort of over 4000 colon cancer patients. In order to achieve these major objectives, we willintegrate bioinformatic analyses, probabilistic network modeling, computational simulation,as well as experimental validation fully supported by our close collaborators at the AcademicMedical Center Amsterdam.?