Gene drives can potentially fixate in a population by biasing inheritance in their favor, opening up a variety of potential applications in areas such as disease-vector control and conservation. CRISPR homing gene drives have shown much promise for providing an effective drive mechanism, but they typically suffer from the rapid formation of resistance alleles. Even if the problem of resistance can be overcome, the utility of such drives would still be limited by their tendency to spread into all areas of a population. To provide additional options for gene drive applications that are substantially less prone to the formation of resistance alleles and could potentially remain confined to a target area, we developed several designs for CRISPR-based gene drives utilizing toxin-antidote (TA) principles. These drives target and disrupt an essential gene with the drive providing rescue. Here, we assess the performance of several types of TA gene drive systems using modeling and individual-based simulations. We show that Toxin-Antidote Recessive Embryo (TARE) drive should allow for the design of robust, regionally confined, population modification strategies with high flexibility in choosing drive promoters and recessive lethal targets. Toxin-Antidote Dominant Embryo (TADE) drive requires a haplolethal target gene and a germline-restricted promoter but should enable the design of both faster regional population modification drives and even regionally-confined population suppression drives. Toxin-antidote dominant sperm (TADS) drive can be used for population modification or suppression. It spreads nearly as quickly as a homing drive and can flexibly use a variety of promoters, but unlike the other TA systems, it is not regionally confined and requires highly specific target genes. Overall, our results suggest that CRISPR-based TA gene drives provide promising candidates for further development in a variety of organisms and may allow for flexible ecological engineering strategies.
METHODSDeterministic model. To analyze the dynamics of TA systems, we developed a deterministic, discrete-generation modeling approach. These models were used to demonstrate frequency trajectories and to calculate parameters such as introduction thresholds. Initially, drive/wild-type heterozygotes are added to a population of wild-type individuals at a specified introduction frequency. In each generation, frequencies were tracked for each genotype. Females select a mate randomly, with each male's chance of being selected being proportional to his fitness value. Females then generate a number of potential offspring equal to twice their fitness value.