Cancer is caused by the sequential accumulation of driver mutations in the genome of a single cell, allowing the clonal expansion of this cell. Several factors can increase the risk of developing cancer, such as old age, genetic predisposition, exposure to sunlight, alcohol consumption, tobacco smoking, viral infections, and obesity. However, for the majority of the risk factors it remains unclear how they contribute to the development of (specific types of) cancer. Furthermore, humans are more likely to develop cancer in certain tissues and the exposure to risk factors alone cannot explain the variation in cancer incidence across tissues.
To gain insight into the mutational processes that contribute to the accumulation of driver events, and ultimately to cancer development, genome-wide patterns of the accumulated somatic point mutations can be characterized. In this thesis, I have investigated mutational processes in tissue-specific adult stem cells (ASCs), which are believed to be the cells-of-origin of many cancer types. We developed a novel technique, which utilizes the organoid culturing technique to select and to clonally expand single ASCs, until sufficient DNA can be obtained to perform whole-genome sequencing and to reliably measure mutations present in the cell-of-origin. Using this technique, we characterized the genome-wide accumulation of mutations in human ASCs from liver, small intestine, and colon during life. Surprisingly, we observe a gradual accumulation of point mutations with age at similar rate in all three ASC types (~40 point mutations per year), despite a substantial difference in cancer risk and number of stem cell divisions. Chronic alcohol abuse, which increases the risk of developing liver cancer, does not affect the number of mutations in liver ASCs either, confirming that the accumulation of point mutations alone cannot explain tumor incidence.
Liver ASCs do have a different mutational profile compared to small intestinal ASCs and colon ASCs, suggesting tissue-specific activity of mutational processes. In the intestine, mutations are predominantly caused by deamination of methylated cytosines, whereas the mutational process in liver ASCs remains to be determined. We also used our new technique to characterize the genome-wide mutational consequences of deficiency of DNA-repair pathway Nucleotide Excision Repair (NER) on ASCs of mouse liver and intestine. In the mouse liver, NER-deficiency causes an increase in mutations, whereas the number of mutations remains similar in the intestine. Nevertheless, both NER-deficient liver and NER-deficient intestinal ASCs show an increase in a specific mutational profile (Signature 8).
Although the number of observations should be increased, our results suggest that the accumulation of complex structural variations and the selection of precancerous stem cells (by a tumor-promoting microenvironment) might play a pivotal and underestimated role in tumor development. Ultimately, a tissue-specific combination of mutational processes, epigenetic variation, and cellular selection is most likely involved in tumor initiation in each patient. Although many questions still remain unanswered, such as the role of epigenetic variation in cancer risk, the research described in this thesis brings us one step closer to understanding the origin of cancer.