The Effects of Mutator Phenotype (Elevated Mutation Rate) in Cancer Evolution.

10/01/2018 - 15:00 - 16:00

During their evolution, many (but not all) tumors acquire a mutator phenotype of some kind (such as an elevated rate of single nucleotide or short insertions and deletions or copy number changes or structural variations). While a mutator phenotype is a known cancer hallmark, it also presents a challenge to the tumor, as an elevated mutation rate leads to more deleterious mutations that decrease cell fitness. Whole genome doubling, which happens in about 40% of tumor cases, is one mechanism that elevates the cancer cell mutation rate. We showed that while it is true that whole genome doubling decreases cell fitness, it simultaneously benefits the cancer cells by increasing their adaptation rate, thus enabling the cancer to develop in the short time it has.

Targeting DNA repair mechanisms is another common strategy for increasing mutation rates in cancer. Under mismatch repair deficiency (MMRd) conditions, short repetitive regions in the DNA (AKA microsatellites), are strongly affected, and develop a high number of insertions and deletions – termed Microsatellite Instability (MSI). Although microsatellite indels occurs frequently, it is commonly thought of as a marker for MMRd and not as an independent driver of tumor progression. We hypothesized that the technical difficulties of current software to identify mutations in repetitive DNA regions may be masking the true contributions of MSI to cancer. Thus we developed two software tools, MSMuTect for calling indels in MS loci, and MSMutSig,  for identifying significantly mutated microsatellite loci. Applying these tools to 6,700 tumors from the TCGA cohort, we uncovered seven driver genes with such microsatellite loci.  We supported their functional role by pathway analysis of the mutated vs. non-mutated patients.

Our tools have taken on increased significance since the FDA recently approved a new drug (Pembrolizumab) based on MSI status rather than tumor type for all solid tumors.  

Finally, I will examine a unique set of tumors, which in addition to MMRd, have a mutation in the exonuclease domain of the POLE gene, leading to extremely high mutation rates. We show the unique effects of the combined mutator phenotype, and the advantages of having such an elevated mutation rate in spite of decreased cell fitness. 

Yosef Maruvka, Broad Institute
BIU Engineering Building 1103, Room 329