Susceptibility of Anopheles gambiae sensu lato from Côte d’Ivoire to insecticides used on insecticide-treated nets: evaluating the additional entomological impact of piperonyl butoxide and chlorfenapyr

Pyrethroid-treated mosquito nets are currently the mainstay of vector control in Côte d’Ivoire. However, resistance to pyrethroids has been reported across the country, limiting options for insecticide resistance management due to the paucity of alternative insecticides. Two types of insecticide-treated nets (ITNs), ITNs with pyrethroids and the synergist piperonyl butoxide (PBO), and Interceptor ® G2 nets, a net treated with a combination of chlorfenapyr and alpha-cypermethrin, are believed to help in the control of pyrethroid-resistant mosquitoes.

the An. gambiae s.l. populations from any of the sites surveyed. However, PBO pre-exposure did increase mortality for all three pyrethroids, particularly deltamethrin (from 4.4% to 48.9%). Anopheles gambiae s.l. from only one site (Bettie) were susceptible to chlorfenapyr at the dose of 100 µg active ingredient (a.i.)/bottle. At the dose of 200 µg (a.i.)/bottle, susceptibility was only recorded in 10 of the 15 sites.

Conclusion
Low mosquito mortality was found for pyrethroids alone, and while PBO increased mortality, it did not restore full susceptibility. The vector was not fully susceptible to chlorfenapyr in one third of the sites tested. However, vector susceptibility to chlorfenapyr seems to be considerably higher than for pyrethroids alone or with PBO. These data should be used cautiously when making ITN procurement decisions, noting that bioassays are conducted in controlled conditions and may not fully represent eld e cacy where the host-seeking behaviors which include free-ying activity are known to enhance proinsecticide chlorfenapyr intoxication to mosquitoes.

Page 4/29
The recent malaria control success recorded in several malaria endemic countries is largely attributed to the scale up of vector control tools, particularly the implementation and high coverage of insecticidetreated nets (ITNs) and indoor residual spraying (IRS) [1]. Pyrethroid insecticides have been used for several years both for ITNs and IRS because of their effectiveness in killing mosquitoes and also for their repellent properties, providing users additional personal protection and safety during sleeping hours [2,3]. Due to their low cost, long residual activity and safety, pyrethroids remain the recommended insecticides for treating ITNs [4]. Unfortunately, the spread of target-site insensitivity and metabolic resistance mechanisms against pyrethroid insecticides occurring in malaria vectors are a threat to vector control and could lead to resurgence of malaria in endemic countries [5][6][7].
Resistance to all the insecticides previously used in public health for adult malaria vector control has been reported across Côte d'Ivoire [8], limiting the country's malaria vector control toolkit. Côte d'Ivoire was the rst West African country to report the knockdown resistance (kdr) mutation in the 1990s, before it was detected all over the continent [9,10]. Moreover, due to the intense exposure of Anopheles mosquitoes to insecticides used in both agriculture and public health, An. gambiae s.l. have developed resistance in several parts of the country [8,11,12].
A rst attempt to manage insecticide resistance using ITNs was the development of ITNs that incorporate a pyrethroid and a synergist, piperonyl butoxide (PBO), which has been shown to increase the mortality of malaria vectors with metabolic resistance involving mono-oxygenases. Currently, ve ITNs types with PBO have been developed, all of which received WHO pre-quali cation and are available on the market.
Three of these are treated with a combination of PBO and deltamethrin (PermaNet 3.0, Tsara Plus, Tsara Boost), one with PBO and permethrin (Olyset Plus), and one with PBO and alpha-cypermethrin (Veeralin) [13][14][15]. The effectiveness of these ITNs in a given area depends on the extent of the involvement of metabolic resistance mediated by mono-oxygenase enzymes in vector populations. A countrywide distribution of such nets therefore requires baseline entomological studies to show the involvement of enzyme activities in pyrethroid resistance of An. gambiae s.l. and the effect of the synergist in restoring susceptibility of the targeted vector population to pyrethroids [16].
Recent research exploring other classes of insecticides identi ed chlorfenapyr-based tools as a promising option for vector control. Furthermore, WHO Pre-Quali cation (WHO PQ) has approved chlorfenapyrbased nets and additional studies are ongoing under the UNITAID-funded New Nets Project to demonstrate the public health impact of those nets for wide use. The different mode of action of chlorfenapyr is expected to be effective in the presence of pyrethroid resistance developed by mosquitoes through several mutations [17,18]. The oxidative removal of the N-ethoxymethyl group of the proinsecticide chlorfenapyr leads to a toxic form of the molecule identi ed as CL 303268 which functions to uncouple oxidative phosphorylation and disrupt metabolic pathways of ATP production in mitochondria [19]. This molecule has low mammalian toxicity and is classi ed as a slightly hazardous insecticide by the WHO. Due to its novel mode of action, chlorfenapyr is unlikely to show any cross-resistance with standard neurotoxic insecticides as observed in Anopheles mosquitoes [20,21]. Chlorfenapyr has been evaluated in laboratory bioassays and in experimental hut studies for ITNs in several countries including Cote d'Ivoire, Burkina Faso, Tanzania, and for both ITNs and IRS in Benin [22][23][24][25][26]. Results from those studies suggest that chlorfenapyr is promising as a new molecule for ITNs and IRS to control populations of pyrethroid-resistant malaria vectors [27,28].
With regards to widespread resistance to the insecticides previously used and the aim of introducing new generation vector control tools in Côte d'Ivoire, ITNs with chlorfenapyr and PBO represent options to overcome and manage widespread resistance. The current study was undertaken within a countrywide insecticide resistance monitoring effort to support the National Malaria Control Programme (NMCP) in making vector control decisions in Côte d'Ivoire and also to contribute to the knowledge base for chlorfenapyr more broadly.

Study sites
Côte d'Ivoire is divided into four ecological zones including the forest zone in the South and West, the transitional zone in the Centre, the savanna zone and the Sudan savanna zone in the North. The climate in the southern forest zone is equatorial, with annual rainfall between 2,100 mm and 2,500 mm and subequatorial in the western forest with a range of average rainfall between 1,600 mm and 2,300 mm per year. In the central transitional zone of the country, the climate is tropical, with an average of 1,200 mm rainfall per year. The climate is also tropical in the northern savanna and Sudan Savanah zone, with an annual average rainfall of 900 mm. Insecticide resistance surveys were carried out in fteen sites distributed across the country, including Aboisso, Adzopé, and San Pedro in the South, Bouaké, Béoumi, Dabakala, Sakassou and Yamoussoukro in the Centre, Odienné in the North-West, Bouna and Nassian in the North-East, Daloa and Gagnoa in the west and Abengourou and Bettié in the East (Fig. 1). Of Those sites, ten represented the NMCP sentinel sites were all malaria control activities and impact were monitored all year round. The additional sites (Bettie, Gagnoa, Sakassou, Dabakala and Béoumi) were surveyed based on their malaria endemicity and targeted for indoor residual spraying site selection. Furthermore, the country has a history of two mass campaigns of pyrethroid-only long lasting insecticide treated net (LLIN) distributions conducted in 2014 and 2017.
The study was conducted between June and September 2019. Anopheles gambiae s.l. larvae and pupae were collected from several urban and rural larval habitats in each site using the dipping method, pooled and reared to adulthood in a eld laboratory at each site.
Pre-weighed preparations of technical-grade active ingredient of chlorfenapyr (BASF Corporation USA, Batch N 2130H070HV) provided by the Centers for Disease Control and Prevention (CDC), Atlanta, were used to prepare 50 ml of two doses (100 µg (a.i)/bottle and 200 µg (a.i) /bottle) by adding the prescribed volume of acetone solvent. The solution obtained was vortexed to ensure that the solution was uniformly mixed. All insecticide solutions and papers were kept at 4ºC before and after each test.
Prior to the eld susceptibility testing, two impregnated papers received were randomly selected from each box and tested against a laboratory-maintained susceptible strain of An. gambiae Kisumu to verify the effectiveness of the paper. Both chlorfenapyr solutions (100 µg (a.i.)/bottle and 200 µg (a.i.)/bottle) were also tested using the CDC bottle assay against An. gambiae Kisumu to check the e cacy on a susceptible population of mosquitoes.
WHO susceptibility test, intensity and synergist assays Insecticide susceptibility tests were conducted using WHO tube tests on 2-4 day old adult female An. gambiae s.l. [29,30]. The diagnostic concentrations of deltamethrin (0.05%), permethrin (0.75%), alphacypermethrin (0.05%) were tested alone and in combination with PBO concurrently in all sites. For the synergist assays, mosquitoes were pre-exposed to PBO for one hour before being exposed to the pyrethroid insecticides for one hour. Resistance intensity at ve and ten times the diagnostic concentration of deltamethrin, permethrin and alpha-cypermethrin was also tested in each site using WHO susceptibility kits. Two tubes lined with silicone oil treated papers were set in parallel during each test and served as controls. The start and end temperatures during all tests were recorded at 9 sites, and Hobo data loggers (HOBO UX100-003, Onset Computer Co., Bourne, MA, USA) were used to track temperatures every 10 minutes in the laboratory in two sites (Beoumi and Sakassou).

CDC bottle assay with chlorfenapyr
Due to non-availability of a WHO recommended protocol for testing chlorfenapyr, two doses (100 µg (a.i.)/bottle and 200 µg (a.i.)/bottle) were selected based on a literature review [21,27,28,31] and used to coat the CDC bottles. The bottles were coated following the protocol described by Brogdon et al. [32], with 1 ml of chlorfenapyr diluted in acetone at the concentration of 100 µg (a.i.)/bottle and 200 µg (a.i.)/bottle. The coated bottles were wrapped in aluminum foil and dried overnight at room temperature. A modi ed protocol was designed following the procedures described by Brogdon et al. [32]. Fifteen to 20 adult mosquitoes (2-4 days old) that emerged from eld collected larvae were aspirated into 250 ml chlorfenapyr-coated Wheaton bottles for 1 hour, after which the mosquitoes were released into a clean cage, aspirated back to paper cups and fed with 10% sugar solution. Mortality was recorded at 24, 48, and 72 hours after exposure. Two bottles coated with acetone only were prepared similarly to serve as controls.
Species identi cation of the An. gambiae s.l. population A subsample of about fty An. gambiae s.l. mosquitoes was randomly selected among the population tested per site and processed for molecular identi cation of the species of the An. gambiae complex following the SINE PCR protocol described by Santolamazza et al. [33]. Additionally, the presence of knocked-down resistance (kdr)-West and East resistance mechanisms was characterized within each An. gambiae s.l. population using the Taqman protocol described by Bass et al. [34] and the presence of acetylcholinesterase (ace-1) gene was determined following the protocol described by Weill et al. [35].

Statistical analysis
Insecticide resistance status was de ned following WHO criteria [27] with mortality after 24 hours for pyrethroids and 72 hours for chlorfenapyr < 90% as con rmed resistance, between 90% and < 98% as possible resistance, and ≥ 98% as susceptible. Mortality was corrected using Abbott's formula when the mortality of the control tubes was above 5% and less than 20%.
Corrected mortality of: 98-100% at 5x the diagnostic dose indicates low resistance intensity Less than 98% at 5x diagnostic dose implies testing the 10x diagnostic dose 98-100% at 10x the diagnostic dose con rms a moderate resistance intensity.
Less than 98% at 10x the diagnostic doses indicates high resistance intensity.
For the synergist assays, an increase in the mortality after pre-exposure to PBO compared to the diagnostic dose of the insecticide alone indicates the involvement of enzymes such as P450s in the population tested. The mortality of mosquitoes exposed to PBO + pyrethroid was also compared with that of the insecticides alone and plotted using Graph Pad Prism 5 software. The total percentage mortality of all sites against each insecticide alone and those of the PBO + insecticides were analyzed and compared using a chi-squared test with Stata 14 (Stata Corporation, College Station, Texas, USA).

WHO susceptibility test, intensity and synergist assays
The An. gambiae s.l. population tested was primarily composed of Anopheles coluzzii (>95%) in Abengourou, Adzopé, Béoumi, Bettié, Dabakala, Daloa, Gagnoa, Sakassou and San Pedro, while An. gambiae sensu stricto (s.s.) was the only member of the complex found in the country northern sites : Bouna, Nassian and Odienné. However, both species were in sympatry (≈ 50% each) in Aboisso, Bouaké and Yamoussoukro. Furthermore, high kdr-West resistance allele frequency was recorded in all the 15 sites ranging between 34% in Gagnoa and 97% in Odienné. The kdr-East was also observed in few sites but at a lower frequency compared to the kdr-West. The highest frequency was observed in Sakassou with 10% kdr-East resistance allele frequency. The ace-1 frequency was relatively high in eight of the fteen sites (between 28% in Odienné to 65% in Gagnoa) (Fig.2) Full susceptibility (100% mortality) was recorded for all insecticides tested against An. gambiae Kisumu and no sign of compromised quality and/or e cacy of the insecticide impregnated papers and the chlorfenapyr-coated bottles were observed.
The start temperatures in the 9 sites where start and end temperatures were recorded had a mean of 25 When all sites were considered together, synergist assays using PBO pre-exposure yielded statistically signi cant increases in the mean mortality for all three pyrethroids (p<0.0001) (Fig. 6). Average permethrin mortality increased from 2.9% when tested alone to 26% mortality with PBO pre-exposure. Deltamethrin mortality was 4.3% mortality without PBO and 47.7% mortality with PBO pre-exposure. Alphacypermethrin mortality was 3.3% without PBO and 33.7% with PBO pre-exposure.
Chlorfenapyr CDC bottle assay Susceptibility (98-100%) was recorded for chlorfenapyr after 24h and 48h delayed mortality for the dose of 200 µg/bottle and 100 µg/bottle, respectively, when tested against the An. gambiae Kisumu strain.

Discussion
Resistance of An. gambiae s.l. mosquitoes to pyrethroids observed in this study is consistent with the previously described trends in Côte d'Ivoire involving various and high insecticide resistance mechanisms such as the kdr-West and East and the ace-1 mutations [8,11,12,36,37]. Testing the intensity of the resistance from 15 sites showed that resistance intensity to all three pyrethroid insecticides is high. This presents a challenge for malaria control as it limits the country's options for insecticide-based vector control interventions. The level of resistance observed could be attributed to the different applications of insecticide in the country, either for agriculture or public health. Agriculture accounts for 24% of GDP in Côte d'Ivoire, one of the largest producers and exporters of coffee, cocoa, cashew, and palm oil worldwide. A recent survey by the ministry of agriculture revealed that more than 20,000 tons of chemicals including pesticides and insecticides were applied in the country in 2018 for agricultural purposes including 40% of this amount entering illegally [38]. This likely contributed to the increasing resistance of malaria vectors to insecticides commonly used for insect control in general and malaria vector control speci cally [36,37,39]. In Côte d'Ivoire, the two major malaria vectors include An. gambiae and An. coluzzii [40]. Both species generally breed in fresh water pools resulting from human activities such as irrigated rice elds and vegetable gardens, which are often treated with insecticides [36,37,39,41]. As described by Chouaibou et al. [42], agricultural pesticides contain several insecticide molecules and particularly pyrethroids and carbamates, which contribute to the increased and widespread resistance of the vectors [39]. Furthermore, studies conducted in Western and Eastern Africa have shown a signi cant increase in the frequency of genes associated with pyrethroid resistance immediately following the implementation of an ITN campaign [7]. Moreover, the National Strategic Plan (NSP) developed in 2018 by the Côte d'Ivoire National Malaria Control Programme (NMCP) prioritized universal ITN coverage in households across the country through mass distribution campaign of pyrethroid-only ITNs.
In this study, pre-exposure to PBO before the different pyrethroids tested showed statistically signi cant but still insu cient increases in the mortality of the An. gambiae s.l. populations. These results would support the use of PBO-treated vector control tools in select areas where signi cant reversal of resistance was observed after pre-exposure to PBO. In some locations, PBO has resulted in increased mortality of even highly resistant mosquito populations [43]. Out of several combinations of PBO ITNs, combinations with the type II pyrethroids (deltamethrin and alpha-cypermethrin) showed higher performance than those with the type I pyrethroid (permethrin), consistent with several other studies comparing the different types of PBO ITNs [44,45]. These data could inform the selection and strati cation of the PBO ITNs for resistance management in Côte d'Ivoire.
Higher mortality of mosquitoes was observed against chlorfenapyr particularly in areas of probable metabolic resistance where higher mortality was noted after pre-exposure of the mosquitoes to PBO before the different pyrethroids. It is known that signi cant increase in mortality after exposure to PBO indicates the involvement of P450 enzymes in the insecticide resistance of the vectors. The trends observed were consistent with previous observations, when assessing the lethality of chlorfenapyr in areas of high metabolic resistance in Aedes aegypti after mosquitoes are exposed or not to PBO [17].
Chlorfenapyr is a protoxin requiring activation, via cytochrome P450 monooxygenases, to exert its toxic effects, via uncoupling of oxidative phosphorylation [17,19]. The expression in An. gambiae s.l. of some cytochrome P450s involved in oxidative metabolism are under circadian control and more strongly expressed at night when Anopheles ight and host-seeking activity are higher [46,47]. Such a model could suggest that exposure at night could likely produce higher mortality results compared to what was observed using traditional testing methods. The supposition that lab-based exposures to technical grade active ingredient or any other formulation where behaviors are excluded, likely negates attributes of oxidative uncoupling by chlorfenapyr when compared to host-seeking behaviors which include free-ying physiologically heightened states known to enhance pro-insecticide chlorfenapyr intoxication to mosquitoes as demonstrated by Oxborough et al. [31], Furthermore, previous work showed that testing completed during the night induces signi cantly greater mortality than daytime bioassays, indicating the possibility of higher effectiveness of chlorfenapyr-based ITNs such as Interceptor® G2 since the peak biting hours of the vectors are typically late in the night [31]. Moreover, several studies are still ongoing to understand and delineate the appropriate diagnostic doses and times for testing chlorfenapyr susceptibility [31,48,49]. This study will therefore contribute to the evidence base for supporting the determination of an appropriate diagnostic dose. With regards to the data generated, in ten out of 15 sites where the tests were conducted, 98% or greater mortality was recorded at the dose of 200 µg/bottle against An. gambiae s.l., indicating that chlorfenapyr could be considered as an option to manage the resistant population of malaria vectors in Côte d'Ivoire, where resistance has continuously increased over the past several decades.
Several trials and studies have been conducted on the Interceptor® G2 net. As described by Bayili  WHO susceptibility and intensity assays of deltamethrin against An. gambiae s.l. of the different sites surveyed. Error bars represent the standard errors at the Y axis.    WHO susceptibility and intensity assays of alpha-cypermethrin against An. gambiae s.l. of the different sites surveyed. Error bars represent the standard errors at the Y axis.

Figure 6
Comparative percentage increment in susceptibility of An. gambiae s.l. using the synergist piperonyl butoxide (PBO) + pyrethroids in relation to pyrethroids alone across all sites surveyed. Signi cant increment in susceptibility was observed with the addition of PBO (p < 0.001) for all three pyrethroids.

Figure 6
Comparative percentage increment in susceptibility of An. gambiae s.l. using the synergist piperonyl butoxide (PBO) + pyrethroids in relation to pyrethroids alone across all sites surveyed. Signi cant