Closing the gap to effective gene drive in Aedes aegypti by exploiting germline regulatory elements

Abstract

CRISPR/Cas9-based homing gene drives have emerged as a potential new approach to mosquito control. While attempts have been made to develop such systems in Aedes aegypti, none have been able to match the high drive efficiency observed in Anopheles species. Here we generate Ae. aegypti transgenic lines expressing Cas9 using germline-specific regulatory elements and assess their ability to bias inheritance of an sgRNA-expressing element (kmo^sgRNAs). Four shu-Cas9 and one sds3-Cas9 isolines can significantly bias the inheritance of kmo^sgRNAs, with sds3G1-Cas9 causing the highest average inheritance of ~86% and ~94% from males and females carrying both elements outcrossed to wild-type, respectively. Our mathematical model demonstrates that sds3G1-Cas9 could enable the spread of the kmo^sgRNAs element to either reach a higher (by ~15 percentage point) maximum carrier frequency or to achieve similar maximum carrier frequency faster (by 12 generations) when compared to two other established split drive systems.

Fig. 1. Cas9 expressed by sds3 and shu regulatory elements causes inheritance bias of the kmo^sgRNAs element.

a Crossing scheme for determination of Cas9-induced inheritance bias. b Illustration of the difference between WT-, mosaic-, or white-eyed phenotype. c Inheritance rate of the kmo^sgRNAs in P₂ larvae, scored by AmCyan fluorescence. Total number of screened larvae from each cross is presented above corresponding data points. Error bars are the Wilson confidence intervals for the binomial proportion. The confidence intervals are calculated from the pooled progeny count and cannot account for potential over dispersal due to parent by parent ‘batch’ effects. Statistical significance was estimated using Fisher’s two-sided exact test relative to the control inheritance rates represented by the dotted line (p ≥ 0.05^ns, p < 0.05*, p < 0.01**, and p < 0.001***). d Percentage of P₂ larvae exhibiting WT or mosaic/white-eye phenotype, according to genotype.^†The integration sites for EwaldC1-Cas9, EwaldC2-Cas9, and nosA2-Cas9, isolines are likely linked in trans to kmo^sgRNAs indicated by the low representation of the trans-heterozygote and non-transgenic genotypes in P₂. Mosquito figures obtained from Ramirez^,. Source data are provided as a Source data file. pro = promoter, ♂ = male, ♀ = female.

Fig. 2. Inheritance bias of the kmo^sgRNAs element could still occur with maternal deposition of Cas9 and/or sgRNAs.

a Crossing scheme for determination of Cas9-induced inheritance bias. b Inheritance rate of the kmo^sgRNAs in P₂ larvae, scored by AmCyan fluorescence. Total number of screened larvae from each cross is presented above corresponding data points. Error bars are the Wilson confidence intervals for the binomial proportion. The confidence intervals are calculated from the pooled progeny count and cannot account for potential over dispersal due to parent by parent ‘batch’ effects. Statistical significance was estimated using Fisher’s two-sided exact test relative to the control inheritance rates represented by the dotted line (p ≥ 0.05^ns, p < 0.05*, p < 0.01**, and p < 0.001***). c Percentage of P₂ larvae exhibiting WT or mosaic/white-eye phenotype, according to genotype. Mosquito figures obtained from Ramirez^,. Source data are provided as a Source data file. pro = promoter, ♂ = male, ♀ = female.

Fig. 3. Drive characteristics of individual kmo^sgRNAs;sds3G1-Cas9 trans-heterozygotes.

a Crossing scheme for determination of sds3G1-Cas9-induced inheritance bias. b, c Boxplots of kmo^sgRNAs inheritance (b) and eye phenotype (c) rates of I₂ larvae. The median value is indicated by a thick black line and the maximum and minimum ends of whiskers represent the most extreme data point that is no more than 1.5 times the interquartile range. The size of each symbol is scaled to the amount of I₂ progeny scored. Total number of screened larvae from each cross is presented below corresponding data points. Mosquito figures obtained from Ramirez^,. Source data are provided as a Source data file. pro = promoter, ♂ = male, ♀ = female.

Fig. 4. Model predicted split drive dynamics using sds3G1-Cas9, nup50-Cas9 and bgcn-Cas9.

Red lines represent carrier frequencies for the Cas9 elements, while blue lines represent the resulting carrier frequencies of the kmo^sgRNAs or w^U6b-GDe elements. Here, results for kmo^sgRNAs;bgcn-Cas9 (dashed lines) are identical to those previously reported and the computational simulations were shown to approximate the experimental cage data. Results for nup50-Cas9;w^U6b-GDe (dotted line) are based on parameter values listed in Li et al.. and kmo^sgRNAs;sds3G1-Cas9 (solid lines) are predictions based on experimental data generated here. Solid lines marked with dots display hypothetical results for a kmo^sgRNAs;sds3G1-Cas9 system in which kmo^sgRNAs fitness costs are reduced to levels observed for w^U6b-GDe (marked with * in the figure legend). Parameter values for all scenarios are detailed in Supplementary Note 1.