Anopheles funestus is a major malaria vector in Tanzania is particularly dominant in the north-western region and southeastern regions (7, 8, 41). Yet, its resistance to key public health insecticides has been, at best, patchy in comparison to the other malaria vectors such as An. arabiensis and An. gambiae for which the resistance data is more commonly collected. This study provides the first large-scale contemporaneous dataset on both the genotypic and phenotypic expression of resistance in An. funestus to a range of insecticides, including two pyrethroids, deltamethrin and permethrin, an organochlorine, DDT, a carbamate, bendiocarb and an organophosphate, pirimiphos methyl.
The findings show widespread multiple resistance in wild An. funestus mosquito populations in mainland Tanzania and provides a basis for decision-making for national malaria control programs on the management of insecticide resistance and selection of appropriate interventions. The findings show pyrethroid resistance to be particularly common in An. funestus with high resistance intensities to deltamethrin in most locations. Similar pyrethroid resistance profiles in this vector have been shown in earlier but separate studies in the south and north of Tanzania (7, 8). Continentally, phenotypic expression of pyrethroid resistance in An. funestus has been reported in other countries including Uganda (42), Mozambique (43), Malawi (44, 45), Ghana (24, 46), and Cameroon (47), with most countries still lacking bioassay data on this vector species. The synergistic effect of PBO, leading to the restoration of susceptibility in pyrethroid-resistant An. funestus mosquitoes was evident in all tested populations, pointing to the predominant involvement of metabolic-mediated resistance mechanisms. Malaria vectors of the An. gambiae complex has also shown similar restoration of pyrethroid-susceptibility with PBO(9). In epidemiological studies evaluating ITNs treated with both PBO and pyrethroids in Tanzania (48) and Uganda (49), increased community protection against malaria was observed, compared to protection accrued with pyrethroid-only nets. Our results therefore add to the body of evidence on the efficacy of PBO-ITNs against pyrethroid-resistant malaria vector populations (50, 51).
The study also shows An. funestus across Tanzania to be susceptible to the IRS insecticide pirimiphos-methyl unlike in An. gambiae, where resistance has been confirmed in parts of the country (52). However, the observed resistance to the carbamate (bendiocarb) in parts of the country is concerning and should be factored in by malaria control programs if IRS campaigns are planned in future. Though Tanzania has recently discontinued the use of IRS (53), this remains an important tool and could be especially useful in high transmission settings or emergency situations to curb outbreaks. Phenotypic resistance to DDT was isolated to the Kilombero valley in south-eastern Tanzania and has been linked to the recent discovery of knockdown resistance (kdr) in An. funestus in the same location (54). The kdr evolution in Kilombero is thought to be due to agricultural DDT contamination and a legacy of past extensive DDT stockpiles (54). Without proactive management of resistance, the observed resistance situation in the country is likely to worsen, as similar insecticide classes are heavily used for agricultural pest control (55).
As vector control evolves in sub-Saharan Africa and as new or improved insecticidal interventions are introduced, it is crucial to monitor the genetic basis of insecticide resistance and track its evolution within and beyond Tanzania. The genotype frequencies of CYP6P9a/b and SV 6.5kb IR genes have previously been estimated as low-moderate in Tanzania but fixed in other southern African countries (18, 20, 30). Here, we found the resistant alleles of these three genes to be either fixed or near fixation across most localities in the eastern region of Tanzania, indicating how these mutants have recently selectively swept through An. funestus populations in Tanzania. The high frequencies of homozygous susceptible genotypes of CYP6P9a-SS in the northern Tanzania districts and the evidence of mixed genotypes in west of Tanzania (near Lake Tanganyika) could suggest a gene flow barrier in the country (possibly including landscape features such as the Great Rift Valley); or an ongoing south-north directional selection of the alternative homozygous resistant genotype CYP6P9a-RR arising from the southern Africa region (56). This potential gene flow restriction could further explain the contrasting distribution of structural variants, with SV 4.3kb fixed and restricted in the northern Tanzania districts (Missenyi and Misungwi) and SV 6.5kb in the rest of the country. The cause of high non-amplifications of SV 6.5kb and CYP6P9b in northern Tanzania is unclarified and warrants investigation. Nonetheless, it has been postulated to be due to the SV 4.3kb (29). Both structural variants have been demonstrated to exacerbate resistance in An. funestus by increasing the overexpression of CYP6P9a and CYP6P9b (29, 30). Recent analysis had similarly linked such copy number variants with increased expression of the genes encoding metabolic resistance in An. gambiae and An. coluzzii (57, 58). The lack of statistical association of CYP6P9a/b, SV 4.3kb and SV 6.5kb genes with pyrethroid resistance in this study is likely due to their already strong selection and fixation in the An. funestus populations in Tanzania. However, this observation could also have resulted from matching the number of samples genotyped which did not consider the natural frequencies of these genes in the populations hence lacking sufficient statistical power to detect a true effect.
L119F-Gste2, which confers cross resistance to DDT and pyrethroids in An. funestus have previously been identified in West and Central Africa (25, 26, 28) and Uganda (42) as the only previous detection in eastern Africa. However, findings of this study now establish that the mutation has now spread to eastern Africa, with low frequencies of the L119F-Gste2 resistance alleles detected in the Tanzanian districts bordering the Democratic Republic of the Congo (Tanganyika), Mozambique (Mtwara rural), and Malawi (Tunduru). The strong association observed between this genotype and deltamethrin resistance phenotypes is concerning and warrants close monitoring of its spread and evolution. Though it is not possible to conclusively describe the evolutionary path of this resistance allele in the region, its strong association with pyrethroid resistance could allude to a recent selection.
Overall, this study has highlighted the importance of broadening the resistance monitoring efforts to include all principal malaria vectors, including not just the members of the An. gambiae complex, but also An. funestus. This is particularly important in areas where this vector species dominates malaria transmission. One complementary outcome of this study was the extension to develop a user-friendly online platform for stakeholders, particularly scientists and public health officials to access the data. This platform lays the groundwork for informed decision making regarding both the management of insecticide resistance and the actual deployment of effective interventions. The study also has the advantage of providing contemporaneous data from multiple sites with both genetic and phenotypic observations. However, it will be essential to regularly update this information, preferably with greater granularity in terms of both the geographical coverage at regional and sub-regional levels, and in terms of the temporal frequency of data collection.
One limitation of this study was the utilization of adult wild female mosquitoes of unknown ages in the bioassays. It is standard practice to utilise age-synchronized, 2-3-day-old mosquitoes in resistance bioassays, to avoid aging-related weaknesses (59) and environmental stresses. However, despite efforts to characterize breeding habitats of An. funestus(13), this species is notoriously difficult to locate in most ecological settings making it challenging to apply the standard approach. This also explains why resistance data on An. funestus has been largely missing previous Tanzania’s sentinel surveys (9). In this study, given the difficulties of finding the larvae, and the logistical challenges due to the geographical expanse covered, a decision was made to use adult samples. Therefore, as insecticide susceptibility increases with age in mosquitoes (60) we cannot completely rule out the potential confounding effect of aging in our mortality data, which could have underestimated the resistant levels of our samples. However, this issue was mitigated by a number of steps including a) using only unfed-non-gravid adults, b) holding the mosquitoes for at least 24 hrs to eliminate any mortality associated with mosquito handling, c) normalising excess mortalities using a set of controls, where natural mortality due to senescence was considered, and d) screening for both the genotypic and phenotypic expression of resistance in the mosquitoes.
Another significant constraint of this study was the potentially low statistical power to discern the associations between specific gene variants and phenotypic resistance. Given the sample sizes used, there was a predisposition toward detecting only variants with large effect sizes, while more subtle or moderate effects likely remained unidentified. Lastly, the study design and samples sizes did not permit an exploration of the cumulative or epistatic effects exerted by multiple gene mutations within the same individual mosquito. It is conceivable that a combination of mutations across the five genes investigated, rather than the impact of each mutation in isolation, could more accurately account for the variations observed in the phenotypic resistance. This complexity underscores the need for future studies to consider larger sample sizes and more sophisticated statistical and genetic analyses that can capture these nuances.
One area of improvement is that future tests should incorporate some of the newly approved insecticides for vector control, including the pyrrole, chlorfenapyr, which is used in the dual-active Interceptor G2, nets, and neonicotinoids, such as clothianidin, which is also approved by WHO for IRS (61). Such expansion will provide essential data for the national malaria control program to guide their selection of interventions in the context of the evolving resistance patterns.