Modeling-informed Engineered Genetic Incompatibility strategies to overcome resistance in the invasive Drosophila suzukii

Abstract

Engineered Genetic Incompatibility (EGI) is an engineered extreme underdominance genetic system wherein hybrid animals are not viable, functioning as a synthetic speciation event. There are several strategies in which EGI could be leveraged for genetic biocontrol of pest populations. We used an agent-based model of Drosophila suzukii (Spotted Wing Drosophila) to determine how EGI would fare with high rates of endemic genetic resistance alleles. We discovered a surprising failure mode wherein field-generated females convert an incompatible male release program into a population replacement gene drive. Local suppression could still be attained in two seasons by tailoring the release strategy to take advantage of this effect, or alternatively in one season by altering the genetic design of release agents. We show in this work that data from modeling can be utilized to recognize unexpected emergent phenomena and a priori inform genetic biocontrol treatment design to increase efficacy.

Keywords: agent-based modeling; genetic biocontrol; incompatible insect technique (IIT); resistance; spotted wing drosophila.

Figure 1

Genetic design and cross-compatibility of Engineered Genetic Incompatibility (EGI) in presence of natural resistant variants (A) EGI agents are homozygous for haplosufficient Cas9-activators targeting a developmental regulator, whose ectopic expression is lethal. A haploinsufficient mutation in the promoter target provides resistance. Orthogonal EGI can be generated by targeting different developmental regulators (middle). A single EGI line can have multiple independent targets (bottom). (B) Hybrids with wild-type contain both the Cas9-activator and a sensitive allele leading to lethal overexpression. (C) Summarized cross compatibility between strains assuming no natural resistance. (D) Detailed cross compatibility of between strains when natural resistant alleles exist in the target population. Large boxes indicate parental strains. Inset boxes represent which if any natural resistant SNPs are deposited by the parental strains (inset right).

Figure 2

Modeling one season of SWD in the presence of EGI based IIT and natural sequence diversity generated resistance. (A) Six male only EGI release strategies were tested using the pyr EGI (blue), hh EGI (red), and dual EGI strains (purple). (B) Total counts of adult female SWD agents over the course of the simulation of one season with initial seeding of resistant alleles, each independently generated at 1% frequency. Resistant alleles were seeded on April 1st. (C) Expanded Punnett square showing genetic outcomes of mating between a resistant agent and an EGI male. The mating allows for generation of EGI females that can then drive a population replacement. P=Engineered resistant promoter, p=wild-type promoter, r=resistant SNP, T=PTA, t=wild-type sequence X=recessive female phenotype, Y=dominant male phenotype) (D, E) Data collected from simulation of one season with initial seeding of resistant alleles, each independently generated at 1% frequency. (F, G) Data collected from simulation of one season with initial seeding of double homozygous resistant agents at 1% frequency. (D, F) Ratiometric tracking of EGI relative to the total number of agents at each timestep throughout the season. (E, G) Steepness parameter from logistic curve fit to data in (D, F). (See also Supplementary Note 4 .). ANOVA was used to identify suppression values that significantly differ from each other. A Student's t-test was used to calculate p-values for those that did.

Figure 3

Second season of EGI IIT treatment assuming overwintering. A second season was simulated with 5, 50, or 95% of the starting population being either pyr EGI or dual EGI. (A) A log scale heatmap with cumulative counts of adult females in the simulation averaged from 10 replicates. Grey boxes indicate simulations that were not run. (B–E) A mapping of genotypes from selected conditions over the course of the full season. Black triangles indicate the timepoint that resistant alleles were generated. New EGI was released five timesteps later. “_r” stands for “reverse”, and denotes strategies where pyr and hh EGI are switched for all releases (i.e. when pyr EGI is release in strategy iv, hh EGI is release in strategy iv_r and vise versa).

Figure 4

L-SSIMS overcomes resistance mechanisms. (A) L-SSIMS genetic design. Sex specific splicing leads to expression of rTA only in females. In absence of tetracycline rTA binds tetO and through positive feedback causes lethal overexpression of rTA. (B) Adult females over one season of treatment with male pyr L-SSIMS agents. The FL construct and PTA are never lost. (C) Adult females over one season of treatment with male pyr L-SSIMS agents with FL and PTA resistance. The FL construct and PTA are independently reverted to wildtype in 10% of eggs.