Fig. 2

Prototypical Design and Behavior of a Cis-acting Self-Elimination Mechanism. A) Expected behavior of a GD demonstrating rapid spread followed by subsequent removal and restoration of the wild-type phenotypes. B) Simplified mathematical mechanism underlying the SEM design, in the absence of resistance mechanisms. The top panel (“Gene Drive”) depicts the initial behavior of the drive, which cleaves with probability p (remaining wild-type with probability 1-p). Homologous recombination (HR) occurs with probability q, leading to a GD homozygote (“GG”), or with probability 1-q non-homologous end-joining occurs (NHEJ), leading to an unviable genotype. Application of a small-molecule initiates drive (“G”) cleavage (lower panel, “Self-Excision”), with probability a, which may be resolved by single strand annealing (SSA) (with probability b) or by NHEJ (probability 1-b). SSA may result in cut-resistant wild-type alleles (“V”) with probability c, or susceptible wild-type alleles with probability 1-c (“W”). NHEJ also may result in an SEM-resistant allele, “S”. C) Hypothetical biological arrangement and mechanism of a small molecule activated SEM. The GD transgene (green) is linked to another endonuclease (“Endonuclease ‘X’”) and the associated target site (“X’s target”), flanked by direct repeats (“DR”). Under normal conditions, only the GD is active, and the entire construct propagates through allelic conversion. However, application of a control molecule activates Endonuclease ‘X’, cleaving X’s target site and initiating repair via single strand annealing repair pathway, facilitated by the DRs. The intended outcome is fixation of wild-type alleles that are resistant to drive-mediated cleavage. D) Simulated behavior of a simple small molecule activated SEM construct. A single, panmictic population of 2000 mosquitoes was simulated for ~ 2 years. Mosquitoes followed a simple cis-acting SEM pattern, where “W” is a wild-type allele, “G” is an active gene drive, “H” is an active SEM, and “V” is the excised cut-resistant wild-type allele that remains. At ~ 3 months of simulation time, we perform 4 weekly releases of 100 SEM-homozygous male mosquitoes. The drive has a cleavage rate of 50% and an HR rate of 100%, generating no resistance alleles. At the 1-year mark, daily small-molecule spraying begins, continuing for ~ 6 months. The activated SEM cleaves at a rate of 50% and repairs alleles 100% of the time, again generating no resistance alleles. Average allele proportions are plotted with 95% empirical quantiles