Forty years of erratic insecticide resistance evolution in the mosquito Culex pipiens

Abstract

One view of adaptation is that it proceeds by the slow and steady accumulation of beneficial mutations with small effects. It is difficult to test this model, since in most cases the genetic basis of adaptation can only be studied a posteriori with traits that have evolved for a long period of time through an unknown sequence of steps. In this paper, we show how ace-1, a gene involved in resistance to organophosphorous insecticide in the mosquito Culex pipiens, has evolved during 40 years of an insecticide control program. Initially, a major resistance allele with strong deleterious side effects spread through the population. Later, a duplication combining a susceptible and a resistance ace-1 allele began to spread but did not replace the original resistance allele, as it is sublethal when homozygous. Last, a second duplication, (also sublethal when homozygous) began to spread because heterozygotes for the two duplications do not exhibit deleterious pleiotropic effects. Double overdominance now maintains these four alleles across treated and nontreated areas. Thus, ace-1 evolution does not proceed via the steady accumulation of beneficial mutations. Instead, resistance evolution has been an erratic combination of mutation, positive selection, and the rearrangement of existing variation leading to complex genetic architecture.

Figure 1. Time Variation of the Different Alleles' Maximum Frequency

The evolution of the maximum frequency (coastal frequency) for each ace-1 resistance allele estimated from the descriptive model is presented. The blue triangles and red squares represent, respectively, the maximum frequency of R and D (= D₂ + D₃) estimated for each cline independently (blue and red bars represent the confidence intervals, respectively), the blue and red lines representing the frequency estimated from the complete dataset conjointly (see text). Note that no tools are currently available to distinguish D₂ and D₃ in the field, only their total frequency being thus estimated (see Materials and Methods).

Figure 2. Sample Site Locations in the Northwest Southeast Transect in the Montpellier Area

Samples are indicated with black circles. The dashed line represents approximately the border between treated and untreated areas (modified from [16]). C. pipiens is present in the whole area.

Figure 3. Frequency of Homozygotes (D/D) Compared to Heterozygotes (D/S) during Development

The frequency of (D/D) individuals is indicated for second instar and emerging adults on the left. On the right, adults are divided into early and late emerging adults. (i) MAURIN-D ((D₂/D₂) versus (D₂/S), green), (ii) BIFACE-D ((D₃/D₃) verus (D₃/S), orange), and (iii) BIFACE-DFix x MAURIN-D ((D₂/D₃) versus (D₃/S), red). The bars indicate one error-type. Significant frequency differences between the different stages are indicated (n.s., p-value > 0.05; *, p-value < 0.05; **, p-value < 0.01; and ***, p-value < 0.001).

Figure 4. Frequency of Heterozygotes (D/S) and (D/R) Compared to Homozygotes (R/R) during Development

The frequency of (D/S) and (D/R) individuals is indicated on the left for two developmental stages, first instar and emerging adults. On the right, emerging adults are divided into early and late emerging adults. (i) (D₃/S) versus (R/R) (green), (ii) (D₃/R) versus (R/R) (orange). The bars indicate one error-type. Significant frequency differences between the different stages are indicated (n.s., p-value > 0.05; *, p-value < 0.05; **, p-value < 0.01; and ***, p-value < 0.001).