Author ORCID Identifier
Date of Graduation
Human and Molecular Genetics
Doctor of Philosophy (PhD)
James E. Hixson
Donald W. Parsons
Paul A. Scheet
Colorectal cancer (CRC) is the third most commonly diagnosed cancer among men and women in the United States, with 3 to 5 percent of the cases diagnosed in the background of a hereditary form of the disease. Biologically, CRC is divided into two groups: microsatellite instable (MSI) and chromosomally unstable (CIN). Genomic and transcriptomic characterization of CRC has emerged from large-scale studies in recent years due to the advancement of next-generation sequencing technologies. These studies have identified key genes and pathways altered in CRC and provided insights to the discovery of therapeutic targets. Despite the wealth of knowledge acquired in the carcinoma stage, there have been insufficient efforts to systematically characterize premalignant lesions at the molecular level, which could lead to a better understanding of neoplastic initiation, risk prediction, and the development of targeted chemoprevention strategies. The challenge in characterizing premalignancy has always been the limited availability of sample material. This challenge is tackled by getting more samples, integrating public datasets, deploying better technology that use less amount of nucleic acids and in-silico tools to extract multi-layer information from the same experiment.
My genomic study consisted of whole exome sequencing (WES) and high-depth targeted sequencing on 80 premalignant lesions bulk tissue and crypts to assess clonality and mutational heterogeneity. WES results showed the presence of multiple clone in premalignancy based on clustering somatic mutation allele frequency. In addition, I determined that multiple clones originate from independent crypts harboring distinct APC and KRAS alterations. In my second study, I performed immune expression profiling and assessment of mutation and neoantigen rate of 28 premalignant lesions with DNA mismatch repair (MMR) deficient and proficient background using RNAseq. My results showed an activated immune profile despite low mutational and neoantigen rate, which challenges the canonical view in MMR-deficient carcinoma stage that immune activation is largely due to high mutation and neoantigen rate. In the last study, I performed transcriptomic sub-classifications of 398 premalignant lesions that associate them with different carcinomas subtypes, and clinical and histopathological features. My results revealed two major findings: prominent immune activation and WNT and MYC activation in premalignancy.
In summary, my large-scale genomic and transcriptomic analyses of colorectal adenomas have identified key molecular characteristics in early colorectal tumorigenesis and provide a foundation for discovering novel preventive strategies.
colorectal cancer, polyps, premalignancy, Lynch syndrome, Familial Adenomatous Polyposis, immunology, neoantigen, consensus molecular subtypes, clonality, serrated adenoma
Available for download on Sunday, May 05, 2019