Mutability of mononucleotide repeats, not oxidative stress, explains the discrepancy between laboratory-accumulated mutations and the natural allele-frequency spectrum in C. elegans

Important clues about natural selection can be gleaned from discrepancies between the properties of segregating genetic variants and of mutations accumulated experimentally under minimal selection, provided the mutational process is the same in the laboratory as in nature. The base-substitution spectrum differs between laboratory mutation accumulation (MA) experiments and the standing site-frequency spectrum, which has been argued to be in part owing to increased oxidative stress in the laboratory environment. Using genome sequence data from MA lines carrying a mutation ( - ) that increases the cellular titer of reactive oxygen species (ROS), leading to increased oxidative stress, we find the base-substitution spectrum is similar between - , its wild-type progenitor (N2), and another set of MA lines derived from a different wild strain (PB306). Conversely, the rate of short insertions is greater in - , consistent with studies in other organisms in which environmental stress increased the rate of insertion-deletion mutations. Further, the mutational properties of mononucleotide repeats in all strains are different from those of nonmononucleotide sequence, both for indels and base-substitutions, and whereas the nonmononucleotide spectra are fairly similar between MA lines and wild isolates, the mononucleotide spectra are very different, with a greater frequency of A:T → T:A transversions and an increased proportion of ±1-bp indels. The discrepancy in mutational spectra between laboratory MA experiments and natural variation is likely owing to a consistent (but unknown) effect of the laboratory environment that manifests itself via different modes of mutability and/or repair at mononucleotide loci.