Colorectal cancer (CRC) is the third most common cancer and accounts for 10% of all new cancers. Approximately 85% of CRCs are microsatellite stable (MSS) while 15% exhibit microsatellite instability (MSI). Of MSI CRCs, 80% are sporadic cases and 20% associate with Lynch syndrome. MSI results from DNA mismatch repair (MMR) deficiency that leads to the accumulation of a high number of mutations, especially small insertions and deletions (indels) in short repeats called microsatellites. In the coding region, these may lead to premature termination codons and truncated protein products. In MSI CRC, also the point mutation rate is elevated. Genes that provide growth advantage to cells via loss of function mutations in microsatellites are called MSI target genes. Many criteria have been suggested to distinguish these genes from incidentally mutated passengers, one of them being simple mutation frequency. To date, numerous genes have been published as candidate MSI target genes and putative tumor suppressors, often based on their mutation frequency alone, but the functional data remains sparse. On the contrary, only few oncogenes are known in MSI CRC, and they are most often flagged by somatic mutation hot spots. Due to the high mutation frequency in MSI CRC, the challenge of distinguishing driver mutations from passengers is augmented. Therefore, when mutation frequencies are evaluated as evidence for selection in MSI CRC, considering the high background mutation rate is of paramount importance. At the same time, the high mutation count makes these tumors a sensitive system for the study of mutation accumulation and selection. The aim of this study was to identify novel driver genes in MSI CRC and to investigate a potential single marker for MSI. In the first study, we sought to identify novel oncogenes with somatic mutation hot spots. We utilized exome sequencing data from 25 MSI CRCs and respective healthy tissues and identified 15 novel candidate oncogenes with confirmed mutation hot spots. Three of the genes (ZBTB2, PSRC1, and RANBP2) showed mutations also in the validation set of 86 additional MSI CRCs and thus emerged as our top candidate oncogenes. The interactomes of ZBTB2 and PSRC1 were studied with liquid chromatography-mass spectrometry and were found to consist of many proteins previously linked to cancer. An increase in cell proliferation was observed for the CRC-associated form of mutant ZBTB2. Additional functional validation is warranted to confirm the role of these genes in MSI CRC tumorigenesis. In the second study, we exploited the high number of mutations from the exome sequencing data of 24 MSI CRCs and the corresponding normal samples to comprehensively catalogue all indels targeting mononucleotide microsatellites. We developed a new statistical model for the somatic background indel rate in mononucleotide microsatellites. We combined mutation significance from our statistical model with mutation clonality in the validation set of 93 additional MSI CRCs to identify the most likely MSI CRC target genes. Two genes – AASDH and SLC9A8 – emerged as our top novel candidates that to our knowledge had not been implicated in MSI cancers before. Further functional validation is required to confirm the pathological role of these genes in MSI CRC. In the third study, we aimed to identify new cancer genes driven by point mutations in MSI CRC. We utilized exome or whole genome sequencing data from 36 MSI CRCs and with MutSigCV and OncodriveFML algorithms we ranked genes based on mutation significance. Two genes – SMARCB1 and STK38L – were selected for further functional validation. A comprehensive interactome analysis revealed enrichment for carbon metabolism in the SMARCB1 binding partners exhibiting altered interaction with the mutant SMARCB1 protein. The mutant STK38L protein in turn exhibited altered interactions with many proteins previously linked to cancer. In a colony formation assay, the SMARCB1 mutant protein increased the formation of drug resistant colonies. Also, in order to continue on our efforts in study I, we identified genes with recurrent somatic mutation hot spots. Seven genes displayed hot spot mutations also in the validation set of 93 additional MSI CRCs, and therefore emerged as our candidates for novel oncogenes in this tumor type. However, additional functional validation is warranted to further understand the role of the genes identified in this study in MSI CRC development. In the fourth study, we examined a 16 T/U mononucleotide microsatellite in the 3’UTR of the EWSR1 gene that was previously reported to show perfect sensitivity and specificity in detecting MSI in colorectal, endometrial, and gastric cancers in two independent populations. We analyzed the repeat with PCR and subsequent fragment analysis in 213 MSI CRCs from two independent populations, 148 MSS CRCs, and the respective normal samples. We found the repeat to be altered in 212/213 (99.5%) MSI CRCs. All the 148 MSS CRCs were wild type for the locus. Therefore, in our data, the repeat showed nearly perfect sensitivity for MSI and thus represents a potential single marker for MSI in CRC. Characterizing the genetic changes underlying cancer development is fundamental in basic cancer research. A profound understanding of cancer biology is a prerequisite for the development of targeted therapies and personalized medicine. In recent years, MSI cancers in particular have shown considerable potential for targeted immunotherapies.
|Tila||Julkaistu - 2019|
|OKM-julkaisutyyppi||G5 Tohtorinväitöskirja (artikkeli)|
- 3111 Biolääketieteet