Adaptive deletion in resistance gene duplications in the malaria vector Anopheles gambiae

Abstract

While gene copy-number variations play major roles in long-term evolution, their early dynamics remains largely unknown. However, examples of their role in short-term adaptation are accumulating: identical repetitions of a locus (homogeneous duplications) can provide a quantitative advantage, while the association of differing alleles (heterogeneous duplications) allows carrying two functions simultaneously. Such duplications often result from rearrangements of sometimes relatively large chromosome fragments, and even when adaptive, they can be associated with deleterious side effects that should, however, be reduced by subsequent evolution. Here, we took advantage of the unique model provided by the malaria mosquito Anopheles gambiae s.l. to investigate the early evolution of several duplications, heterogeneous and homogeneous, segregating in natural populations from West Africa. These duplications encompass ~200 kb and 11 genes, including the adaptive insecticide resistance ace-1 locus. Through the survey of several populations from three countries over 3-4 years, we showed that an internal deletion of all coamplified genes except ace-1 is currently spreading in West Africa and introgressing from An. gambiae s.s. to An. coluzzii. Both observations provide evidences of its selection, most likely due to reducing the gene-dosage disturbances caused by the excessive copies of the nonadaptive genes. Our study thus provides a unique example of the early adaptive trajectory of duplications and underlines the role of the environmental conditions (insecticide treatment practices and species ecology). It also emphasizes the striking diversity of adaptive responses in these mosquitoes and reveals a worrisome process of resistance/cost trade-off evolution that could impact the control of malaria vectors in Africa.

Keywords: adaptive trajectory; fitness cost; gene duplication; genome evolution; insecticide resistance; malaria vector.

Figure 1

Anopheles gambiae ace‐1 gene duplicated alleles, genotypes, and phenotypes. (a) The various alleles revealed using the two tests are symbolized: the small boxes represent the ace‐1 alleles, green for alleles carrying 119G (susceptible), and red for alleles carrying 119S (resistant); the large boxes represent the amplicons (different colors are used to represent the various duplicated alleles although the amplicons are similar as far as we know); the internal deletion (ID) present in one of the amplicons of the R^x* allele is indicated. (b) For each test (Res‐test or ID test), or the combination of the two (two‐test), the various PCR profiles, that is, phenotypes, and associated genotypes are indicated, with conserved color code for each allele. Note that even this combination of tests does not allow complete genotype discrimination

Figure 2

Duplication structure and primer positions of the internal deletion test (ID test). (a) Amplicon structure. The whole amplicon is represented by the box. The predicted genes are represented by gray dots, except for ace‐1, which is indicated by the black line (see Assogba et al., 2016 for details). The white box represents the area deleted in some amplicons, that is, the internal deletion (ID). The blue arrows represent the ID test primers positions. (b) PCR results of the ID test for different genotypes (NB: This image has been produced by merging two parts of a single photograph). Only those containing an R^x* copy are amplified ([ID+]). M is the size marker

Figure 3

Evolution of the number of ace‐1 and ID copies in [RR] individuals from Baguida (Togo). Box plot represents the distributions of the copy numbers ([a] ace‐1, [b] ID) in individuals sampled in Baguida in 2013, 2014, and 2016. The bold line represents the median, the box and whiskers, respectively, represent the 25% and 75%, and 5% and 95% quartiles, and the dots represent outliers

Figure 4

Allele frequencies. The cumulated frequencies of the R^x, R^x*, D, and S alleles are presented for each sample. The locality and year of collection are also indicated (bottom), as well as the number of analyzed individuals (N) and the species (top). Note that only samples with more than 10 individuals were considered to estimate the allelic frequencies using the maximum‐likelihood approach (see text and Table S3). Colors are the same than in Figure 1