A synthetic gene drive system for local, reversible modification and suppression of insect populations

Abstract

Replacement of wild insect populations with genetically modified individuals unable to transmit disease provides a self-perpetuating method of disease prevention but requires a gene drive mechanism to spread these traits to high frequency. Drive mechanisms requiring that transgenes exceed a threshold frequency in order to spread are attractive because they bring about local but not global replacement, and transgenes can be eliminated through dilution of the population with wild-type individuals. These features are likely to be important in many social and regulatory contexts. Here we describe the first creation of a synthetic threshold-dependent gene drive system, designated maternal-effect lethal underdominance (UD(MEL)), in which two maternally expressed toxins, located on separate chromosomes, are each linked with a zygotic antidote able to rescue maternal-effect lethality of the other toxin. We demonstrate threshold-dependent replacement in single- and two-locus configurations in Drosophila. Models suggest that transgene spread can often be limited to local environments. They also show that in a population in which single-locus UD(MEL) has been carried out, repeated release of wild-type males can result in population suppression, a novel method of genetic population manipulation.

Figure 1. Schematic of the UD^MEL drive system and illustration of how UD^MEL rewires developmental gene expression

The UD^MEL system is composed of two constructs. UD^MEL-1 consists of maternal toxin A (red) and zygotic antidote B (green) and UD^MEL-2 consists of maternal toxin B (purple) and zygotic antidote A (light blue) (A). In wildtype mothers, maternal transcripts from gene A and gene B (gray line) are required for normal embryonic development. The toxins, multimers of miRNAs, degrade one or both of these mRNAs (red line for UD^MEL-1 toxin A targeting Gene A, and purple line for UD^MEL-2 toxin B targeting Gene 2), to which they are complementary. Embryos lacking one or both mRNAs and/or their products (orange lines), depending on whether the mother is heterozygous or transheterozygous, respectively, for the miRNA multimers, die. Progeny inheriting the other construct, or both constructs, respectively, survive because they express miRNA-resistant versions of the mRNAs (blue and green) in the early zygote at levels sufficient to rescue embryonic development. Dashed line (dark blue) corresponds to the initiation of zygotic transcription(B). Single- and two-locus UD^MEL configurations are illustrated, with the two different constructs being indicated by the blue and red boxes, and homologous and non-homologous chromosomes indicated by the positions of their centromeres (blue circles). Wildtype chromosomes lack boxes (C).

Figure 2. UD^MEL single- and two-locus systems are predicted to show threshold-dependent gene drive and bring about local population replacement

The threshold frequency above which a UD^MEL drive system spreads into a population, and below which it is eliminated from the population, was calculated using a deterministic model and graphed. Release thresholds are calculated for two single locus scenarios: a single, all-male release of transheterozygotes (A) and two all male releases of transheterozygotes in the first and second generation (B), for elements with zero fitness cost (s). For X-autosome two-locus UD^MEL (C) and autosome-autosome two-locus UD^MEL (D) single releases of doubly homozygous males are illustrated. Introduction frequencies/transgene frequencies represent the fraction of individuals in the total population carrying at least one UD^MEL construct. Two-way migration occurs between population 1, illustrated as a group village of houses, which has undergone population replacement, and population 2, which is separated from population 1 by a barrier (vertical line), and is initially all wildtype (E). Plots depict the dynamics of single- and two-locus UD^MEL under two-population models in which migrants are exchanged between population 1 (blue line), which has been seeded with transgenics, and population 2 (red line), which initially consists only of wildtypes (F–Q). Migration occurs either prior to mating (F,G,J,K,N,O) or after mating (H,I,L,M,P,Q). For single-locus UD^MEL, when migration occurs before mating and the migration rate is 1%, transgenics spread to high levels in population 1, and reach a frequency of 2.0% in population 2 (F); when the migration rate is 4%, transgenes are ultimately eliminated from both populations (G). When mating occurs before migration and the migration rate is 0.2%, transgenics spread to high frequency in population 1, and reach a frequency of ~1% in population 2 (H); a migration rate of 0.25% or higher results in loss of transgenes from both populations (I). For X-autosome two-locus UD^MEL, when migration occurs before mating and the migration rate is 1%, transgenics spread to high frequency in population 1, and reach a frequency of 8% in population 2 (J); a migration rate of 1.8% or higher (2.5% is illustrated), results in spread to fixation in both populations (K). When mating occurs before migration and the migration rate is 1%, transgenics spread to high frequency in population 1 and reach a frequency of 8% in population 2 (L); a migration rate of 1.7% or higher (1.8% is illustrated) results in spread to fixation in both populations (M). For autosome-autosome, two-locus UD^MEL, when migration occurs before mating and the migration rate is 1%, transgenics spread to high frequency in population 1, and reach a frequency of 8.8% in population 2 (N); a migration rate of 1.65% or higher (2% is illustrated), results in spread to fixation in both populations (O). When mating occurs before migration and the migration rate is 1%, transgenics spread to high frequency in population 1 and reach a frequency of 11% in population 2 (P); a migration rate of 1.4% or higher (1.5% is illustrated) results in spread to fixation in both populations (Q).

Figure 3. Introduction of wildtype individuals into UD^MEL-replaced populations can result in population suppression, and/or loss of transgenes from the population

Two releases of 10,000 males transheterozygous for single-locus UD^MEL constructs with no fitness cost into a wildtype population of 10,000 results in UD^MEL transgene fixation at ~ generation 12 (blue line). Six releases of 10,000 wildtype males into this population during generations 31–36 results in a population crash. Subsequent release of 200 wildtype males and females into the wild at generation 37 results in recovery of the total population to its wildtype, pre-transgenic numbers within 12 generations (red line). This is associated with loss of the UD^MEL chromosomes, which have fallen below their threshold frequency for spread (A). One release of 10,000 AAX^B Y males (B), or AABB males (C), into a wildtype population of 10,000, results in population replacement. Releases of 5,000 males and 5,000 females into the replaced population during generations 31 and 33 results in loss of transgenes from the population, but does not lead to population suppression.

Figure 4. Synthetic UD^MEL chromosomes show maternal-effect lethal and zygotic rescue, underdominant behavior, and drive population replacement

Crosses between parents of specific genotypes, either widltype or heterozygotes for the same UD^MEL construct (indicated in the two leftmost columns) were carried out, and progeny survival to crawling first instar larvae quantified (rightmost 6 columns). The + indicates wildtype. The chromosome each UD^MEL construct is inserted on is indicated by colour of the horizontal line (second chromosome, green; third chromosome, blue) (A). Crosses between parents of different genotypes (indicated to the left) were carried out. The maternal copy number of BicC-driven miRNAs targeting myd88, dah, or o-fut1 (toxin, n), and the zygote copy number of the bnk-driven, miRNA-resistant versions of myd88, dah, or o-fut1 (antidote, n) are indicated, as are the predicted and observed rates of embryo survival. The genotypes of embryos expected to survive are indicated in black, and those of embryos expected to die in red (B). *(embryo survival data were normalized to those of wild-type (w¹¹¹⁸), which was 93.80± 3.19. Plots depict frequency of transgenics over generations in populations of +/+ (wildtypes) into which single-locus UD^MEL transheterozygous (C) or two-locus UD^MEL double homozygotes (D) were introduced. The % release indicates the fraction of the total population, post release, consisting of UD^MEL individuals. All populations were followed for 15 generations, or until transgenic individuals were lost from the population. Thin lines represent experimental data. The thick line and surrounding shading represent a best-fit analysis of the data. The dashed line represents the predicted behavior of elements carrying no fitness cost.