Why does the target-recognition site in PRDM9 have unusually high mutation rates?

Abstract

Gene PRDM9, responsible for the location of recombination hotspots in mammals (including humans), exhibits one of the highest mutation rate in the human genome. This high mutation rate has a direct impact on the life expectancy of recombination hotspots and the variability of the recombination landscape in mammals. Why would PRDM9 evolve such a high mutation rate? Why is PRDM9 one of the fastest evolving genes in the human genome? While the evolution of PRDM9-guided recombination hotspots has received much attention, the evolution of mutation rates at PRDM9 has received little attention. Here we explore the evolution of mutation rates at PRDM9. We formulate and analyse population genetics models for modifiers of mutation rates at PRDM9 and modifiers of mutation rates at PRDM9's target sites. By letting mutation rates evolve we find that natural selection favours the evolution of high mutation rates at PRDM9 and low mutation rates at PRDM9's target sites. However, considerations about linkage disequilibrium and genetic drift suggest that only high mutation rates at PRDM9 are likely to evolve. Our predictions match observations, with mutation rates at PRDM9 being two orders of magnitude higher than the average for the genome. Our modifiers of mutations at PRDM9 are most likely to evolve in close linkage with the PRDM9 locus. This is consistent with the binding domain in PRDM9 adopting a zinc-finger structure -molecular structure with high intrinsic mutation rates. Interestingly, our results are an exception of the reduction principle in modifier theory. We draw parallels between our results and another exception to the reduction principle: when genes are evolving in fluctuating environments. We suggest that the red-queen dynamics that characterise the co-evolution between PRDM9 and its target can generate a fluctuating genomic environment that drives the evolution of our modifier of mutation rates at PRDM9.