Whole-genome sequencing of major malaria vectors reveals the evolution of new insecticide resistance variants in a longitudinal study in Burkina Faso

Abstract

Background: Intensive deployment of insecticide based malaria vector control tools resulted in the rapid evolution of phenotypes resistant to these chemicals. Understanding this process at the genomic level is important for the deployment of successful vector control interventions. Therefore, longitudinal sampling followed by whole genome sequencing (WGS) is necessary to understand how these evolutionary processes evolve over time. This study investigated the change in genetic structure and the evolution of the insecticide resistance variants in natural populations of Anopheles gambiae over time and space from 2012 to 2017 in Burkina Faso.

Methods: New genomic data have been generated from An. gambiae mosquitoes collected from three villages in the western part of Burkina Faso between 2012 and 2017. The samples were whole-genome sequenced and the data used in the An. gambiae 1000 genomes (Ag1000G) project as part of the Vector Observatory. Genomic data were analysed using the analysis pipeline previously designed by the Ag1000G project.

Results: The results showed similar and consistent nucleotide diversity and negative Tajima's D between An. gambiae sensu stricto (s.s.) and Anopheles coluzzii. Principal component analysis (PCA) and the fixation index (F_ST) showed a clear genetic structure in the An. gambiae sensu lato (s.l.) species. Genome-wide F_ST and H12 scans identified genomic regions under divergent selection that may have implications in the adaptation to ecological changes. Novel voltage-gated sodium channel pyrethroid resistance target-site alleles (V402L, I1527T) were identified at increasing frequencies alongside the established alleles (Vgsc-L995F, Vgsc-L995S and N1570Y) within the An. gambiae s.l.

Populations: Organophosphate metabolic resistance markers were also identified, at increasing frequencies, within the An. gambiae s.s. populations from 2012 to 2017, including the SNP Ace1-G280S and its associated duplication. Variants simultaneously identified in the same vector populations raise concerns about the long-term efficacy of new generation bed nets and the recently organophosphate pirimiphos-methyl indoor residual spraying in Burkina Faso.

Conclusion: These findings highlighted the benefit of genomic surveillance of malaria vectors for the detection of new insecticide resistance variants, the monitoring of the existing resistance variants, and also to get insights into the evolutionary processes driving insecticide resistance.

Keywords: An. gambiae; Genomic surveillance; Insecticide resistance; Malaria.

Fig. 1

Map of the sampling sites around Bobo-Dioulasso city

Fig. 2

Average of genetic diversity statistics within the An. gambiae s.l. populations in the study sites from 2012 to 2017; The Y-axis shows the diversity statistic, the X-axis shows the sampling periods and the colours show the populations. A Nucleotide diversity (θ_π), B Watterson theta (θ_w), C Tajima’s D, error bars correspond to the 95% confidence interval for each year

Fig. 3

Genetic differentiation and population structure of An. gambiae s.l. using the SNPs identified in the 3L chromosome. Sour: Souroukoudinga; A Pairwise F_ST between An. gambiae s.l. populations collected in different villages; no differentiation between populations of the same species collected in two different villages; B PCA showing the genetic structure of An. gambiae s.l. populations

Fig. 4

Heat map showing the frequencies of the non-synonymous SNPs frequencies (max freq > 0.05 in at least one population) in the VGSC gene in the An. gambiae s.l. populations over time. The X-axis shows the An. gambiae s.l. populations and the sampling periods. The Y-axis shows the non-synonymous SNPs positions in the chromosome 2L and the corresponding amino acid change. The gradient colour bar shows the distribution of the allelic frequencies

Fig. 5

Dynamics of vgsc-L995F, vgsc-V402L + I1527T and N1570Y allele frequencies in the An. coluzzii populations; vgsc-V402L and vgsc-I1527T are shown to be linked and exhibited the same frequencies from 2012 to 2017, the X-axis shows the sampling periods and the Y-axis shows the allelic frequencies

Fig. 6

Heat map of the cytochrome P450 genes showing high CNVs frequencies in the An. gambiae s.l. populations. The X-axis shows the An. gambiae s.l. populations and the sampling periods. The Y-axis shows the positions of the P450 genes ID and the CNV type (del: deletion or amp: amplification). The gradient colour bar shows the distribution of the CNV frequencies

Fig. 7

Evolution of variant frequencies of the ACE1 (Ace1-G280S, AGAP001356 (ACE1) amp) and Diacylglycerol Kinase (AGAP000519 amp) genes in the An. gambiae s.s. populations. The X axis shows the sampling periods and the Y axis, the allele frequencies of the genetic variants

Fig. 8

Genome-wide scan for recent selection using Garud’s H12 across chromosome X and arm 2R (5 kb window) of the An. gambiae s.s. and An. coluzzii populations from 2012 to 2017 showed positive selection signals mainly in the genomic region involved in insecticide resistance. The X-axis shows the genome (chromosome) positions and the Y-axis shows the Garud’ H12 values. The top and bottom left panels show the selection signal in An. gambiae s.s. populations, and the top and bottom right panels show the signal in An. colozzii populations. The colours represent the sampling periods, where the navy colour is 2012 and the brown colour is 2017. High values of H12 indicate a signal of positive selection in the corresponding genomic region within the populations: Ace1: Acetylcholinesterase gene; Cyp6p: Cytochrome P450 gene; Coeae60: Carboxyl-esterase gene; Keap1: Kelch-like ECH-Associated Protein 1; Dgk: Diaglycerol Kinase (AGAP000519) gene; Cyp9k1: Cytochrome P450 gene

Fig. 9

Genome-wide selection scan using genetic differentiation (F_ST) across chromosome X (10 kb window) between populations of An. gambiae s.s. and An. coluzzii collected in 2012 and in 2017. The X-axis shows the positions of the chromosome X and the Y-axis shows the values of the 10 kb windowed F_ST. Signals of positive selection (high genetic differentiation, F_ST > 0) were observed in the genomic regions corresponding to Cyp9k1 and DGK genes within the An. gambiae s.s populations