Insecticide resistance in the malaria vector An. gambiae s.l. threatens insecticide-based control efforts, necessitating regular monitoring. We assessed resistance in field-collected An. gambiae s.l. from Jinja, Uganda using WHO biosassays. Only An. gambiae s.s. and An. arabiensis (≈70%) were present. Female An. gambiae exhibited extremely high pyrethroid resistance (permethrin LT50 >2h; deltamethrin LT50 >5h). Female An. arabiensis were resistant to permethrin and exhibited reduced susceptibility to deltamethrin. However, whilst An. gambiae were DDT resistant, An. arabiensis were fully susceptible. Both species were fully susceptible to bendiocarb and fenitrothion. Kdr 1014S has increased rapidly in the Jinja population of An. gambiae s.s. and now approaches fixation (≈95%), consistent with insecticide-mediated selection, but is currently at low frequency in An. arabiensis (0.07%). Kdr 1014F was also at low frequency in An. gambiae. These frequencies preclude adequately-powered tests for association with phenotypic resistance. PBO synergist bioassays resulted in near complete recovery of pyrethroid susceptibility suggesting involvement of CYP450s in resistance. A small number (0.22%) of An. gambiae s.s. x An. arabiensis hybrids were found, suggesting the possibility of introgression of resistance alleles between species. The high levels of pyrethroid resistance encountered in Jinja threaten to reduce the efficacy of vector control programmes which rely on pyrethroid impregnated bednets or indoor spraying of pyrethroids.

Metabolic resistance to pyrethroid insecticides is widespread in Anopheles mosquitoes and is a major threat to malaria control. DNA markers would aid predictive monitoring of resistance, but few mutations have been discovered outside of insecticide-targeted genes. Isofemale family pools from a wild Ugandan Anopheles gambiae population, from an area where operational pyrethroid failure is suspected, were genotyped using a candidate-gene enriched SNP array. Resistance-associated SNPs were detected in three genes from detoxification superfamilies, in addition to the insecticide target site (the Voltage Gated Sodium Channel gene, Vgsc). The putative associations were confirmed for two of the marker SNPs, in the P450 Cyp4j5 and the esterase Coeae1d by reproducible association with pyrethroid resistance in multiple field collections from Uganda and Kenya, and together with the Vgsc-1014S (kdr) mutation these SNPs explained around 20% of variation in resistance. Moreover, the >20 Mb 2La inversion also showed evidence of association with resistance as did environmental humidity. Sequencing of Cyp4j5 and Coeae1d detected no resistance-linked loss of diversity, suggesting selection from standing variation. Our study provides novel, regionally-validated DNA assays for resistance to the most important insecticide class, and establishes both 2La karyotype variation and humidity as common factors impacting the resistance phenotype.

Functionally-constrained genes are ideal insecticide targets because disruption is often fatal, and resistance mutations are typically costly. Synaptic acetylcholinesterase (AChE) is an essential neurotransmission enzyme targeted by insecticides used increasingly in malaria control. In Anopheles and Culex mosquitoes, a glycine-serine substitution at codon 119 of the Ace-1 gene confers both resistance and fitness costs, especially for 119S/S homozygotes. G119S in A. gambiae from Accra (Ghana) is strongly associated with resistance, and, despite expectations of cost, resistant 119S alleles are increasing significantly in frequency. Sequencing of Accra females detected only a single Ace-1 119S haplotype whereas 119G diversity was high overall but very low at non-synonymous sites; evidence of strong purifying selection driven by functional constraint. Flanking microsatellites showed reduced diversity, elevated linkage disequilibrium and high differentiation of 119S, relative to 119G homozygotes across up to two megabases of the genome. Yet these signals of selection were inconsistent and sometimes weak tens of kilobases from Ace-1. This unexpected finding is attributable to apparently ubiquitous amplification of 119S alleles as part of a large copy number variant (CNV) far exceeding the size of the Ace-1 gene, whereas 119G alleles were unduplicated. Ace-1 CNV was detectable in archived samples collected when the 119S allele was rare in Ghana. Multi-copy amplification of resistant alleles has not been observed previously and is likely to underpin the recent increase in 119S frequency. The large CNV compromised localization of the strong selective sweep around Ace-1, emphasizing the need to integrate CNV analysis into genome scans for selection.

The spread of resistance to insecticides in disease-carrying mosquitoes poses a threat to the effectiveness of control programmes, which rely largely on insecticide-based interventions. Monitoring mosquito populations is essential, but obtaining phenotypic measurements of resistance is laborious and error-prone. High-throughput genotyping offers the prospect of quick and repeatable estimates of resistance, while also allowing resistance markers to be tracked and studied. To demonstrate the potential of highly-mulitplexed genotypic screening for measuring resistance-association of mutations and tracking their spread, we developed a panel of 28 known or putative resistance markers in the major malaria vector Anopheles gambiae, which we used to screen mosquitoes from a wide swathe of Sub-Saharan Africa (Burkina Faso, Ghana, Democratic Republic of Congo (DRC) and Kenya). We found resistance association in four markers, including a novel mutation in the detoxification gene Gste2 (Gste2-119V). We also identified a duplication in Gste2 combining a resistance-associated mutation with its wild-type counterpart, potentially alleviating the costs of resistance. Finally, we describe the distribution of the multiple origins of kdr resistance, finding unprecedented diversity in the DRC. This panel represents the first step towards a quantitative genotypic model of insecticide resistance that can be used to predict resistance status in An. gambiae.

Insecticide resistance is a major impediment to the control of vectors and pests of public health importance and is a strongly selected trait capable of rapid spread, sometimes even between closely-related species. Elucidating the mechanisms generating insecticide resistance in mosquito vectors of disease, and understanding the spread of resistance within and between populations and species are vital for the development of robust resistance management strategies. Here we studied the mechanisms of resistance in two sympatric members of the Anopheles gambiae species complex – the major vector of malaria in sub-Saharan Africa – in order to understand how resistance has developed and spread in eastern Uganda, a region with some of the highest levels of malaria. In eastern Uganda, where the mosquitoes Anopheles arabiensis and An. gambiae can be found sympatrically, low levels of hybrids (0.4%) occur, offering a route for introgression of adaptively important variants between species. In independent microarray studies of insecticide resistance, Gste4, an insect-specific glutathione S-transferase, was among the most significantly up-regulated genes in both species. To test the hypothesis of interspecific introgression, we sequenced 2.3kbp encompassing Gste4. Whilst this detailed sequencing ruled out introgression, we detected strong positive selection acting on Gste4. However, these sequences, followed by haplotype-specific qPCR, showed that the apparent up-regulation in An. arabiensis is a result of allelic variation across the microarray probe binding sites which artefactually elevates the gene expression signal. Thus, face-value acceptance of microarray data can be misleading and it is advisable to conduct a more detailed investigation of the causes and nature of such signal. The identification of positive selection acting on this locus led us to functionally express and characterise allelic variants of GSTE4. Although the in vitro data do not support a direct role for GSTE4 in metabolism, they do support a role for this enzyme in insecticide sequestration. Thus, the demonstration of a role for an up-regulated gene in metabolic resistance to insecticides should not be limited to simply whether it can metabolise insecticide; such a strict criterion would argue against the involvement of GSTE4 despite the weight of evidence to the contrary.