Resistance to a CRISPR-based gene drive at an evolutionarily conserved site is revealed by mimicking genotype fixation

Abstract

CRISPR-based homing gene drives can be designed to disrupt essential genes whilst biasing their own inheritance, leading to suppression of mosquito populations in the laboratory. This class of gene drives relies on CRISPR-Cas9 cleavage of a target sequence and copying ( homing) therein of the gene drive element from the homologous chromosome. However, target site mutations that are resistant to cleavage yet maintain the function of the essential gene are expected to be strongly selected for. Targeting functionally constrained regions where mutations are not easily tolerated should lower the probability of resistance. Evolutionary conservation at the sequence level is often a reliable indicator of functional constraint, though the actual level of underlying constraint between one conserved sequence and another can vary widely. Here we generated a novel gene drive in the malaria vector Anopheles gambiae, targeting an ultra-conserved target site in a haplosufficient essential gene (AGAP029113) required during mosquito development, which fulfils many of the criteria for the target of a population suppression gene drive. We then designed a selection regime to experimentally assess the likelihood of generation and subsequent selection of gene drive resistant mutations at its target site. We simulated, in a caged population, a scenario where the gene drive was approaching fixation, where selection for resistance is expected to be strongest. Continuous sampling of the target locus revealed that a single, restorative, in-frame nucleotide substitution was selected. Our findings show that ultra-conservation alone need not be predictive of a site that is refractory to target site resistance. Our strategy to evaluate resistance in vivo could help to validate candidate gene drive targets for their resilience to resistance and help to improve predictions of the invasion dynamics of gene drives in field populations. Author summaryGene drives have the potential to be applied as novel control strategy of disease-transmitting mosquitoes, by spreading genetic traits that suppress or modify the target population. Many gene drive elements work by recognising and cutting a specific target sequence in the mosquito genome and copying themselves into that target sequence allowing the gene drive to increase in frequency in the population. Like other mosquito control interventions, efficacy will greatly depend on minimising the development of resistance to the gene drive mechanism - most likely via a change in the target sequence that prevents further cutting. One strategy to reduce resistance is to target sequences that are highly conserved, which implies that changes cannot easily be tolerated. We developed a strategy that simulates high selection pressure, under which resistance is most likely to emerge, and therefore provides a stringent test of its propensity to arise. Unlike previous results with another gene drive, we recovered a resistant allele within a few generations of gene drive exposure and at high frequency. Our results show that conserved sequences can vary hugely in ability to tolerate mutations and highlights the need to functionally validate future candidate gene drive target sites for their robustness to resistance.