Investigating the landscape and genetic architecture of germline mutagenesis

PROJECT SUMMARY Variation in the efficacy of DNA proofreading and repair can alter germline and somatic mutation rates and cause ripple effects on many important determinants of health, including the frequency of birth defects and the likelihood of developing cancer. It is becoming feasible to measure mutation rate variation within and between species, but we lack a solid frame of reference for its interpretation–for example, the clinical meaning of an ele- vated mutation rate in a human family or the relevance of mutation rate to the survival prognosis of an endangered species. The goal of this research program is to investigate why mutation rates vary within and between species and to what extent this variation is shaped by selection to preserve longevity or population fitness. Since selection can only act on variation that is genetically determined, it will first be crucial to disentangle the heritable compo- nent of mutation rate variation from the effects of environmental mutagens or reproductive life history. We thus propose several complementary strategies for identifying genetic variants that perturb mutation rates in germline and somatic tissues. First, we will investigate the genetic architecture of somatic mutagenesis in the BXD mice, a collection of recombinant progeny of the lab strains C57BL/6J and DBA2/J that we previously used to map quantitative trait loci affecting germline mutagenesis. We will measure the rates and spectra of somatic mutations occurring in aged BXD mouse tissues and test for associations between inherited alleles and the rate of mutation accumulation over time. We will pay special attention to any somatic mutator effects of known germline mutator alleles, including a Mutyh variant whose human homolog is linked to heritable colorectal cancer. Second, we will scan for germline mutator allele activity in human whole genome sequence biobanks. We will cluster haplotypes based on evidence that they were inherited identical-by-descent from the same common ancestor and identify clusters whose mutation spectrum is inconsistent with the genomic background, indicating the likely presence of a germline mutator allele that increases the relative rates of certain mutation types. After applying this method to the UKBiobank, which includes many biometric and biomedical phenotype measurements, we will then test whether mutator alleles tend to be associated with traits such as disease status, hormone levels and fertility. Third, we will use a similar haplotype clustering method to identify mutations in whole-population pedigrees of endangered Florida scrub jays and black rhinos and test for mutation spectrum shifts that appear consistent with the presence of mutator alleles. We will then evaluate whether these data support a longstanding theoretical pre- diction that selection against mutator alleles is inefficient in threatened populations with low effective population sizes. Finally, we will measure the age dependence of de novo mutation rates and spectra in three strains of the short-lived killifish Nothobranchius furzeri and test whether the rate of reproductive aging appears to have accelerated in the populations that have evolved the shortest lifespans. Together, these studies will illuminate how variation in DNA proofreading and repair efficacy may contribute to variation in health and longevity.