Gene drives—selfish alleles that bias inheritance in their favor—have great potential for solving epidemiological and agricultural problems. Seeding populations with transgenic individuals containing drive elements linked to a trait of interest, such as resistance, has been proposed as strategy to control disease vectors and agricultural pests. The most predictable, and consequently useful, of these elements are targeted endonucleases that recognize and insert themselves into homologous alleles in which they are absent by cleaving a target sequence and allowing homologous recombination to result in gene conversion. Homing endonucleases and CRISPR endonucleases have been demonstrated to be effective knockout drivers with reliable activity and specificity in laboratory populations (Windbichler et al. 2011; DiCarlo et al, 2015; Gantz & Bier, 2015), and thus are possible tools for this type of population level genetic manipulation in the field.
Models of gene drive suggest that the gains from distorting inheritance can overcome selection such that even strongly deleterious alleles can spread through drive (Deredec et al. 2008; Unckless et al. 2015). These studies, however, do not consider coevolutionary dynamics between drive elements and host genomes. Homing endonucleases rarely exert deleterious effects in their native hosts, suggesting that coevolution readily occurs in response to drive to lessen intragenomic antagonism. Changes in response to selection on the part of either drive elements or host genomes could potential derail the intended effects of synthetic gene drives. To date, no studies have assessed the efficacy and reproducibility of gene drive in evolving populations or to the predictability of host genome response.
Here we propose to directly test the efficacy of gene drive in evolving populations of the budding yeast, Saccharomyces cerevisiae. We have successfully constructed several drive elements. Alleles have been designed so that we can vary the fitness cost of drive across experimental replicates. Drive and target alleles will be fluorescently tagged to quantify their frequencies over time. Whole genome sequencing will be used to characterize reproducibility of host genome responses. We predict that host genomes will evolve in response to drive and that fates of drive elements will be predictable as a function of increasing selection, but unpredictable between replicate populations of the same selective regime due to the stochastic nature of molecular adaptation.