Background
In recent years, substantial achievements have been made in global malaria control which has led to an estimated 22% and 29% decrease in malaria incidence and mortality, between 2000 and 2015, respectively [
1]. These impressive achievements have largely resulted from the rapid scale-up of diagnosis, treatment and vector control interventions, particularly indoor residual spraying (IRS) and long-lasting insecticidal nets (LLINs). Unfortunately, the expansion of IRS coverage and concomitant mass LLIN distributions have placed high levels of selection pressure on
Anopheles mosquito populations to evolve resistance to the thirteen insecticides belonging to the four main classes approved for public health use: pyrethroids, carbamates, organophosphates, and organochlorines [
2,
3]. Moreover, the rate of decline in malaria case incidence has begun to stall and has even reversed in some regions since 2014 [
4]. Insecticide resistance is now a pervasive phenomenon that has been reported in approximately two-thirds of countries with ongoing malaria transmission [
1,
3,
5]. In addition, many vector populations are resistant to multiple insecticides from different chemical classes; of the 73 countries that provided monitoring data from 2010 onwards, 50 reported resistance to two or more insecticide classes [
1]. The continued spread of resistance could threaten malaria control progress achieved thus far and ultimately lead to operational failure of prevailing control measures [
3]. In response, the current recommendations for insecticide resistance management rely on tactical deployment of the active ingredients used for IRS and on LLINs in rotation, combinations (particularly LLINs), mosaics and mixtures [
3,
6]. Unfortunately, these management strategies are restricted in their potential effectiveness by the limited choice of available insecticides. The urgent need for new chemicals with novel modes of action has been the impetus driving the evaluation of established agricultural insecticides to control resistant mosquito vector populations [
7].
Historically, the chemical industry has not focused on the development of new public health insecticides because of high market uncertainties and low profit margins in comparison to the agricultural sector [
8,
9]. However, current efforts are focusing on testing insecticides already available to the agricultural industry as potential public health tools. The ideal insecticide for IRS and LLINs should possess the following properties: intrinsic toxicity, chemical and physical properties that facilitate effective uptake upon contact, long residual efficacy, toxicity to specific mosquito species at low dosages, easy application to the desired substrate, stability, low volatility, and low mammalian toxicity (including to other non-target species). Chlorfenapyr and clothianidin are two such insecticides, belonging to different chemical classes with distinct modes of action.
Chlorfenapyr is a pyrrole class insecticide commonly used against mites and termites, which functions as an oxidative phosphorylation uncoupler. This compound disrupts the proton gradient across mitochondrial membranes, interrupting ATP synthesis and ultimately resulting in death of the organism [
10,
11]. A susceptibility survey of 19 pesticides tested against insectary colonies of
Aedes aegypti, Culex quinquefasciatus and
Anopheles quadrimaculatus indicated that chlorfenapyr was most effective against
An.
quadrimaculatus and least against
Culex quinquefasciatus [
12]; other toxicity screens have reported that chlorfenapyr can have a lethal effect against field populations of the latter species [
13]. Furthermore, multiple phase II trials in Tanzania and Benin have highlighted the effectiveness of chlorfenapyr as an adjunct to pyrethroid-treated nets against
Anopheles arabiensis, Anopheles gambiae sensu stricto (s.s.), and
Culex quinquefasciatus [
14‐
18] and as a candidate for IRS in Benin [
8,
10].
Clothianidin is one of seven insecticides within the neonicotinoid class; it has low mammalian toxicity and is primarily used against piercing–sucking insects of major crops [
19,
20]. The basic mode of action is to target the nicotinic acetylcholine receptor (nAChR) in the insect central nervous system [
19,
21]. Compared with chlorfenapyr, this class of insecticide has been through less rigorous study in relation to vector control. At a molecular level, each neonicotinoid has been characterized by differential activity against the nAChR protein subunit of
An. gambiae, suggesting that these compounds may have differential efficacies against target insects [
21]. In a toxicity survey examining 25 different synthetic insecticides, clothianidin was among the group of six neonicotinoids inducing the highest mortality levels against
Culex quinquefasciatus [
13]. Toxicity of six neonicotinoids was tested alone and in combination with deltamethrin and the synergist piperonyl butoxide (PBO) against
Ae. aegypti and
An. gambiae and all compounds had poor individual efficacies but induced higher levels of insecticidal action when in combination, possibly due to the amelioration of oxidase and/or esterase activity by PBO [
22]. Clothianidin, both alone and in mixtures, demonstrated some of the lowest mortality rates among the six compounds. However, it is noteworthy that this study did not evaluate a range of doses for each insecticide, nor monitor mosquito mortality past 24 h; given the slower acting nature of these compounds and that each neonicotinoid can produce its own unique range of sub-lethal effects, further testing is warranted [
22].
Both chlorfenapyr and clothianidin have been manufactured into new commercial vector control formulations. Chlorfenapyr is one of two active ingredients in a combination LLIN with alpha-cypermethrin, produced by BASF (Interceptor
® G2), which received interim approval from the World Health Organization (WHO) in 2017 [
23]. It is also under evaluation as an IRS product (Sylando
® 240SC). Clothianidin has been developed by Sumitomo into an IRS formulation (SumiShield
® 50WG), which has been pre-qualified by the WHO [
24], and Bayer has also incorporated clothianidin in a mixture IRS product with deltamethrin (Fludora™ Fusion WP-SB) [
25]. To date, phase II trials of these respective interventions have reported promising results with multiple resistant vector species [
8,
10,
14,
17,
22,
26,
27]. Furthermore, a recent phase III trial in India, demonstrated greater reductions in the density of indoor, pyrethroid-resistant
Anopheles culicifacies with SumiShield
® 50WG, compared to Actellic
® 300CS (pirimiphos-methyl) [
28]. Prospective community-level implementation of such interventions, as part of National Malaria Control Programmes (NMCPs), is first predicated on demonstrable efficacy against local vector populations.
Insecticide resistance is a major public health concern in Ethiopia, where intensive DDT spraying since 1959 and mass LLIN distributions over the past 10–15 years have resulted in highly focal and heterogeneous patterns of insecticide resistance across the country [
29]. In 2016, decreased susceptibility to pyrethroids (alpha-cypermethrin, deltamethrin, etofenprox, lambda-cyhalothrin and permethrin) and incipient resistance to bendiocarb, pirimiphos-methyl and propoxur, the three chemicals routinely used for IRS, were detected in multiple regions [
29]. The aim of this study was to determine the diagnostic doses of chlorfenapyr and clothianidin, defined as the concentration of insecticide that kills 100% of susceptible mosquitoes within a given time [
30], against a susceptible laboratory strain of
An. arabiensis and screen for resistance in a field-collected multi-resistant
An. arabiensis population from Oromia region, Ethiopia. Establishment of the diagnostic doses will provide a critical starting point to monitor future resistance and define the suitability of these insecticides for intervention deployment. In addition, improved understanding of the cross-resistance between these novel chemicals and currently used insecticides will aid the NMCP and other stakeholders in making informed choices regarding the most appropriate tools for malaria vector control and insecticide resistance management.
Discussion
Given the limited number of insecticides available for public health use, there is an urgent need to evaluate alternate chemicals with new modes of action to improve the control of resistant vector populations [
7]. This study determined the diagnostic doses of clothianidin (neonicotinoid) and chlorfenapyr (pyrrole) for resistance monitoring of field populations of
An. arabiensis in Ethiopia and also investigated the prevalence of cross-resistance to other chemicals already in use by the NMCP. Diagnostic doses for chlorfenapyr at 100 µg/bottle and clothianidin at 2%/filter paper (both using 60 min exposures) against the susceptible laboratory-reared
An. arabiensis DZ strain achieved complete mortality by 24 h and 72 h, respectively. These testing conditions are therefore appropriate for this laboratory population and can be used to screen for susceptibility to these compounds in the field, as well as to monitor residual efficacy of interventions longitudinally.
By comparison, field collected
An. arabiensis exposed to 2%/filter paper clothianidin, demonstrated significantly higher rates of knock-down at 30 and 60 min and reached 100% mortality 24 h earlier than the laboratory DZ strain. It is conceivable that this increased susceptibility may result from a fitness cost or negative cross-resistance incurred by the presence of other resistance traits. A similar phenomenon has been described for the neonicotinoid, dinotefuran, which was more toxic against carbamate-resistant
Culex quinquefasciatus, compared to a susceptible strain, in both larval bioassays and topical applications to adults [
39]. These results were attributed to insensitive acetylcholinesterase being less efficient at degrading nicotinic substrates, such that in the presence of dinotefuran, the concentration of acetylcholine could increase more rapidly at the synaptic level in carbamate resistant individuals, leading to earlier saturation of nicotinic receptors and higher levels of mortality [
39]. Additional laboratory studies have reported that an
Anopheles stephensi colony which was resistant to DDT, malathion and deltamethrin, required lower lethal concentrations of the neonicotinoids imidacloprid, thiacloprid and thiamethoxam, compared to a susceptible counterpart [
40], while a tri-mixture of PBO, deltamethrin and dinotefuran was more effective than a combination of PBO and deltamethrin in killing a pyrethroid resistant
An. gambiae strain (VKPR: homozygous for
kdr) [
26]. In the latter case, it was proposed that PBO blocked mixed-function oxidase (MFO) detoxification of deltamethrin, allowing the rate of miniature excitatory postsynaptic potentials and resulting acetylcholine release to increase, with dinotefuran competitively inhibiting the inactivation of nicotinic receptors.
In the present study, additional testing demonstrated that the wild
An. arabiensis population had higher tolerance to the organophosphate malathion as well as intense resistance to the pyrethroids deltamethrin and permethrin, with survivors at five and ten times the diagnostic insecticide dose. Complete susceptibility to the carbamates bendiocarb and propoxur was also observed. This field population has previously been characterized by moderate L1014F
kdr allele frequencies, an absence of the G119S-
Ace-
1R mutation and partial synergism to pyrethroids following PBO exposure [
29,
32,
41]. To date, the underlying metabolic mechanisms conferring resistance to organophosphates and pyrethroids await elucidation but the putative negative cross-resistance to clothianidin render this insecticide a promising candidate for inclusion in local vector control campaigns.
The efficacy of chlorfenapyr against resistant mosquito species has been reported from multiple settings [
8,
10,
11,
15‐
18,
32,
42‐
47]. Because the mode of action for chlorfenapyr differs from that of other neurotoxic insecticides used for malaria vector control, it is anticipated that there is minimal risk for cross-resistance to evolve from currently known metabolic resistance pathways [
18]. Furthermore, laboratory studies have demonstrated that PBO can act as an antagonist with chlorfenapyr, reducing mosquito mortality, owing to the involvement of MFOs in the initial conversion step of the pro-insecticide to its active toxic form [
11,
48,
49]. This property highlights the potential of this insecticide to manage resistant populations characterized by elevated oxidases, one of the predominant mechanisms conferring insecticide resistance across sub-Saharan Africa [
50]. However, in this study, complete mortality was not achieved with the wild
An. arabiensis population at any dose within 72 h indicating that a higher concentration or longer holding period may be required. Previous laboratory evaluations have highlighted temperature and time of bioassay, as two factors which can influence mortality; activation of chlorfenapyr and disruption of respiratory pathways is enhanced when the mosquito is more metabolically and behaviourally active [
51]. All of the current testing was conducted during the daytime, at constant, regulated temperatures, which may account for lower levels of lethality; in this dataset it is not possible to ascribe incomplete mortality to this scenario or the presence of a small proportion of more tolerant individuals. Study findings reinforce the need to adapt standardized laboratory testing guidelines to take into account the idiosyncrasies of non-neurotoxic insecticides [
44,
51].
During testing, a number of limitations were encountered and areas of improvement were identified. Due to issues of crystallization and achieving uniform coating of Wheaton bottles with technical grade clothianidin diluted in acetone, formulated SumiShield 50WG was used to treat filter papers for WHO susceptibility tests. In addition to introducing batch variability between ‘home-made’ filter papers, other additives contained within this formulation may have also contributed an unmeasurable degree toward mosquito mortality. In future studies, it may be appropriate to evaluate alternate methodologies and solvents to facilitate the use of technical grade insecticide in bioassays. In the clothianidin bioassays, maintaining control mosquitoes for 7 days presented issues, with mortality exceeding 20% by day 5 in a number of replicates, possibly because of old age or starvation due to a lack of blood meal. Given that complete mortality in both wild and field populations was achieved within 3 days, longer holding periods may not be needed for future testing of these particular populations. The clothianidin results have been interpreted relative to reports of other neonicotinoids; however, observations may not be generalizable as each compound is known to vary in binding potency to the nAChR receptor and may, therefore, exert distinct effects between insect species [
21]. Unfortunately, due to insufficient sampling, the wild
An. arabiensis used in chlorfenapyr testing was not collected at the same time as those used to assess the diagnostic doses of clothianidin or other insecticides, preventing complete comparisons between these groups. However, the resistance profiles of this population have been shown to be consistent over the last couple of years [
29]. Finally, levels of insecticide resistance of wild Ethiopian mosquitoes are highly focal and heterogenous across the country [
29,
41] and ideally such screening (including different concentrations and exposure times) and cross-resistance testing would be extended to other field populations with different resistance intensities and underlying molecular and metabolic mechanisms. Future studies may also consider including synergists to explore possible metabolic mechanisms that can explain the observed cross-resistance, especially for clothianidin which is known to have a synergistic effect with PBO [
22,
26]. This work only examined lethal effects of exposure, while some neonicotinoids are known to modify insect behaviour at sublethal concentrations; if these compounds impact feeding behaviour, fecundity and/or egg hatchability, these could also be important contributors to reducing vectorial capacity. The range of sub-lethal concentrations identified herein provide potential starting points for such studies.
Authors’ contributions
KA, SI, DY and LAM designed the study and were responsible for data analysis and interpretation. REW provided statistical expertise during data analysis. JS and DY coordinated entomological collections and provided materials and equipment. KD, SI, REW and LAM drafted the manuscript which was revised by co-authors. All authors read and approved the final manuscript.