Population seasonality and release timing significantly affect the probability of establishment for small releases of gene drive mosquitoes

Nikolov, MO, A. L.; Beaghton, A. K.; Beaghton, P. J.; Wenger, E. A.; Burt, A.; Welkhoff, P. A.,  American Journal of Tropical Medicine and Hygiene,  99:367-367. 2018.

Highly efficient CRISPR/Cas9 gene-drive systems have recently been developed, targeting reproductive-capacity and malaria-competency loci of malaria transmitting vector species, such as An. gambiae. The resulting drive systems aim to either suppress the local wild-type population or alter its genome, conveying desirable phenotypes such as P. falciparum refractoriness. The potential for sustained spread of gene drive constructs as proposed for malaria and a variety of other applications (pest control, tick borne diseases, dengue) has raised concerns for unintentional or unauthorized organism release outside approved and strictly-regulated trial sites. Previous analyses posit that as few as one or two gene drive organisms carrying efficient gene drive cassettes may establish a permanent (sub)population of genetically-modified (GM) mosquitoes with probability >50%. While these results are broad and cautionary, we show that seasonality is a fundamental environmental characteristic to consider when modeling decision variables. For the first time, we investigate the impact of gene-drive release timing and numbers on the establishment probability of GM vectors in the context of realistic seasonal population variation. We model a male sex bias, driving-Y population suppression gene drive, targeting An. gambiae, since these are among the first field trials candidate constructs. We analyze gene-drive establishment in geographies of different seasonality and spatial vector population features. We show that releasing a small number of gene-drive mosquitoes over the few weeks in the beginning of the wet season facilitates population founder effects and high establishment probability: between 60% – 80% for releases of as few as one or two mosquitoes. However, releasing genedrive mosquitoes outside this time results in much lower establishment probability, typically <20%. Our findings address crucial ethical and environmental concerns, which may guide whether, how, and where to set up gene-drive trials.