Mosaic genome architecture of the Anopheles gambiae species complex

Abstract

Background: Attempts over the last three decades to reconstruct the phylogenetic history of the Anopheles gambiae species complex have been important for developing better strategies to control malaria transmission.

Methodology: We used fingerprint genotyping data from 414 field-collected female mosquitoes at 42 microsatellite loci to infer the evolutionary relationships of four species in the A. gambiae complex, the two major malaria vectors A. gambiae sensu stricto (A. gambiae s.s.) and A. arabiensis, as well as two minor vectors, A. merus and A. melas.

Principal findings: We identify six taxonomic units, including a clear separation of West and East Africa A. gambiae s.s. S molecular forms. We show that the phylogenetic relationships vary widely between different genomic regions, thus demonstrating the mosaic nature of the genome of these species. The two major malaria vectors are closely related and closer to A. merus than to A. melas at the genome-wide level, which is also true if only autosomes are considered. However, within the Xag inversion region of the X chromosome, the M and two S molecular forms are most similar to A. merus. Near the X centromere, outside the Xag region, the two S forms are highly dissimilar to the other taxa. Furthermore, our data suggest that the centromeric region of chromosome 3 is a strong discriminator between the major and minor malaria vectors.

Conclusions: Although further studies are needed to elucidate the basis of the phylogenetic variation among the different regions of the genome, the preponderance of sympatric admixtures among taxa strongly favor introgression of different genomic regions between species, rather than lineage sorting of ancestral polymorphism, as a possible mechanism.

Figure 1. The 42 microsatellite markers used in the current study.

Figure 2. The STRUCTURE results.

Consistent STRUCTURE results for 4≤K≤7 across the simulation lengths (5×10⁵, 10⁶ and 2×10⁶) for plot A. Each of the 414 individuals in the analysis is represented by a thin vertical line, and is partitioned into K colored segments that represent the individual's probability of belonging to one of the K populations. Mosquitoes were positioned in the figure according to the collection site (Burkina Faso (BF), Mali, Kenya and Senegal) and to their species/forms (M molecular form (M1 and M2), S-form (S1, S2 and S3), A. arabiensis (A1 and A2), A. merus (R) and A. melas (L). In plot B, the M1 is further separated from M2 for K = 2 with simulation length 5×10⁵.

Figure 3. PCA and CA results of the 414 individuals at 36 loci.

PCA analysis of the 414 individuals in a genome-wide study for 36 microsatellites (A) and for seven loci on the X chromosome (B). CA analysis of the nine populations in the Xag inversion region (C), near the X centromere region (D), on chromosomes 2 (E) and 3 (F). Individuals and populations are color-coded as described (see legend where numbers in brackets next to the populations name indicate the number of mosquitoes in each group). The contribution of each PC or CA is indicated as a percentage in brackets on the axis. In plots A and B, each point represents a single mosquito, while in plots C to F, each point represents a population. In plots C to F, microsatellite markers are localized by black bars. The first three letters of their name (AGN, where N is chromosome X, 2 or 3) were removed.

Figure 4. PCA and CA results of the 262 A. gambiae s.s. at 42 loci.

PCA results of the 262 A. gambiae s.s. at 42 microsatellites (A), and CA results of the five populations of A. gambiae s.s. in the Xag inversion region and near the X centromere (B), and on autosomes (C). Each individual (in plot A) or each population (in plots B and C) is represented by a single point. See Figure 3 for legend. Only the most significant microsatellites markers are indicated.