Affiliations: 1) University of Washington; 2) Fred Hutchinson Cancer Research Center
Keywords: Population history
Mutations provide essential raw material for evolutionary change. Genetic factors that increase mutation rates, i.e. mutator alleles, have been observed to arise during experimental evolution, facilitating adaptation to the lab environment. However, little is known about how prevalent mutator alleles are in natural populations and how they have historically contributed to evolution due to the rarity of mutations under normal conditions. To tackle this challenging question, we have established a framework that utilizes the mutation spectrum—the relative frequencies of different types of mutations, calculated from natural polymorphisms—to determine potential historical impacts of Saccharomyces cerevisiae mutation rate modifiers that affect certain mutation types disproportionally. We combined this with efficiently accumulating de novo mutations in a reporter gene using a modified fluctuation assay of natural isolates to identify S. cerevisiae populations that have experienced recent mutation rate and spectrum changes. We discovered a 10-fold range of mutation rate variation among 16 haploid strains from diverse populations. Two strains from the Mosaic beer clade have excess C>A mutations in both de novo and rare natural polymorphisms, indicating a recent occurrence of at least one mutator allele common to the two strains. We further identified a mutator allele in OGG1 that partially explains these strains’ mutator phenotype. Our mutation spectrum analysis of polymorphisms also indicates that additional mutators have likely been influencing the evolution of natural populations beyond the Mosaic beer strains. We discovered that strains from the African beer population are even more conspicuous mutation spectrum outliers, and a subset of French dairy strains show intermediate mutation spectra between the African beer and the rest of the strains. Analysis of these French dairy strains suggests that there are likely mutator alleles introgressed from African beer strains affecting the subsequent mutation spectra. We have engineered a reporter into the African beer and French dairy strains for the modified fluctuation assay, and we are measuring de novo mutation spectra in these strains to test this hypothesis. In summary, our framework has proven useful in identifying mutator alleles in natural populations of budding yeast, and will reveal how mutator alleles contribute to evolution.