Efficacy of a novel mode of action of an indoor residual spraying product, SumiShield® 50WG against susceptible and resistant populations of Anopheles gambiae (s.l.) in Benin, West Africa

Background Scale-up of the distribution of long-lasting insecticide-treated bed nets and indoor residual spraying with insecticides over the last decade have contributed to the considerable decrease of malaria morbidity and mortality in sub-Saharan Africa. Due to the increasing pyrethroid resistance intensity and the spread of carbamate resistance in Anopheles gambiae (s.s.) mosquitoes and the limited number of insecticides recommended by the WHO for vector control, alternative insecticide formulations for IRS with long-lasting residual activity are required to sustain the gains obtained in most malaria-endemic countries. Methods SumiShield 50WG (clothianidin 300 mg ai/m2) developed by Sumitomo Chemical was evaluated alongside deltamethrin 25 mg ai/m2 (K-Othrine 250 WG) against a pyrethroid resistant Anopheles gambiae (s.l.) population in experimental huts in Covè, Benin. Residual activity was also tested in cone bioassays with the susceptible An. gambiae “Kisumu” strain and the local wild resistant population. Results The results showed very low toxicity from deltamethrin (mortality rates ranged between 1–40%) against host-seeking resistant Anopheles populations. SumiShield in contrast gave an overall mean mortality of 91.7% at the 120 h observation across the eight- month observation period following spraying. The residual activity measured using cone tests was over the 80% WHO threshold for 24 weeks for resistant wild Anopheles population and 32 weeks for the susceptible strain “Kisumu” after the spraying. Conclusions SumiShield is a good candidate for IRS in areas of permanent malaria transmission and where Anopheles populations are resistant to other conventional insecticides such as pyrethroids. It would be interesting to complete experimental huts studies by assessing the efficacy and residual effect of SumiShield 50WG at community level (small-scale field testing) in an area where vectors are highly resistant to insecticides.


Background
The first lesson learnt from the monitoring and evaluation of indoor residual spraying (IRS) in Africa is the variation in the residual life of the main insecticides used for indoor residual spraying: bendiocarb (carbamate) and pirimiphos methyl (organophosphate). Actellic EC (pirimiphos methyl) and Ficam VC (bendiocarb) have shown a short residual effect in several countries such as Benin, Mali, Rwanda and Equatorial Guinea [1][2][3][4]. In previous IRS programs, pirimiphos methyl capsule suspension (CS) showed a longer residual effect in Ghana compared to other countries such as Benin, Ethiopia, Liberia, Mali, Senegal and Zimbabwe [4]. The second problem is the emergence of carbamate and organophosphate resistance [5]. Pyrethroid resistance is widespread in most malaria endemic countries and since carbamate resistance has arisen in some countries, only organophosphates, the remaining class of insecticides recommended by the WHO, is used by National Malaria Control Programmes for indoor residual spraying and for resistance management. For the National Malaria Control Programmes and the international community, it is judicious to prevent the rapid and widespread resistance to carbamates and organophosphates by evaluating new insecticide formulations with different modes of action and a long residual effect.
The solution proposed by the WHO for resistance management [6] appears difficult to implement due to the limited number of insecticides recommended for IRS. The mitigation of the spread of vector resistance is to avoid subjecting Anopheles to the same products or those with the same mode of action over several years or to reduce resistance selection pressures in order to weaken the resistance gene carriers toward these products. In the Atacora, Alibori and Donga areas in Benin where pirimiphos methyl CS is the insecticide used for IRS, the NMCP recommends its usage for not more than two or three years and suggests the implementation strategy for resistance management based on the rotational use of two or three insecticides with different modes of action.
Overall, given the limited number of insecticides recommended for LLINs and IRS since the 1970s and malaria vectors developing resistance to most insecticide classes, formulations of insecticides with new modes of action are needed for resistance management. The Innovative Vector Control Consortium (IVCC) has developed novel active ingredients. These new candidates should be on the market by or around 2022 for IRS. Also, new formulations and new classes of insecticides could be recommended for LLINs impregnation.
SumiShield® 50WG (clothianidin) developed by Sumitomo, is a good candidate as needed by all NMCPs in Africa, and has prequalification listing. SumiShield® contains clothianidin, a neonicotinoid insecticide not previously used for vector control. As a result, it is expected that pyrethroid resistant mosquito populations should be killed by SumiShield 50WG. This study was implemented to evaluate the efficacy and residual effect of SumiShield 50WG applied at 300 mg ai/m 2 against wild pyrethroids resistant Anopheles population in semi-field conditions in experimental huts in Benin, West Africa.

Study design
The study was conducted in the experimental station at Covè, Southern Benin, over a period of 8 months (October 2015 to May 2016). The huts were built alongside a paddy field which constitutes the permanent breeding sites for Anopheles mosquitoes. Adult volunteers were recruited among the inhabitants of the location (Covè). Pregnant and breast-feeding women were not involved in the study. After having announced through the district for volunteers, selection was done after approval was granted by the traditional head of the study site.
Sleepers (mosquito collectors) were rotated randomly among huts each night following a Latin square design. They entered the huts at dusk (21:00 h) and remained inside until dawn. In the morning at 6:30 h, dead mosquitoes were collected from the floor of the huts and resting mosquitoes from the walls, roof and exit traps using aspirators. Mosquitoes were scored by location as dead or alive and as fed or unfed. Live mosquitoes were placed in small cups and provided with 10% sugar solution for 24, 48, 72, 96 and 120 h to assess delayed mortality. All wild An. gambiae (s.l.) mosquitoes caught per night and per hut were taken into account. In each month, sleepers collected free-flying host-seeking Anopheles gambiae (s.l.) from Covè during four Latin square weeks per month (24 days). One Latin square week was equal to six days. A sleeper slept once in each of the six huts used during six successive nights and rested on the seventh night. The wild mosquito populations were complemented with the susceptible An. gambiae "Kisumu" strain which were reared in the insectary of the Centre de Recherche Entomologique de Cotonou (CREC) and released in the huts. At least 20 females of An. gambiae "Kisumu" strain aged 5-8 days were released in each hut over one night per month; these numbers were kept reasonably low to prevent sleepers from receiving too many mosquito bites.
The primary parameters measured in these experimental huts according to the duration of the trial were: (i) deterrence: reduction in hut entry relative to the control huts (untreated hut); (ii) induced exophily: the proportion of mosquitoes that exited early and were found in exit traps; (iii) blood-feeding inhibition: the reduction in blood-feeding compared with that in the control hut; (iv) immediate mortality: the proportion of mosquitoes that were killed at the end of the exposure time; and (v) delayed mortality: the proportion of mosquitoes that were killed after 24, 48, 72, 96 and 120 h.
The huts used were designed to simulate local and typical West African households (Fig. 1). The description and how the host-seeking mosquitoes were captured as described by [7]. Six experimental huts (smooth cement) were used in this study: 2 huts for SumiShield 50WG (contains 50% w/w clothianidin), 2 huts for K-Othrine 250WG (Deltamethrin 250 WG, contains 25% w/w deltamethrin) and 2 untreated (control).

Insecticide treatment
The maximum safety instructions and protective measures were observed. The operators who treated the walls wore full protective clothing: long-sleeved shirt and trousers, hat, rubber boots, gloves and a particlefiltering half-mask.

Dilution process
Standard nozzles recommended by the WHO for malaria vector control are types 8002 or 8001. A standard nozzle type 8002 was used. This nozzle delivers a volume of 760 ml of insecticide solution / min (for 19 m 2 ). If a sprayer does not have a CFV (control flow valve), an average application rate of 40 ml/m 2 is assumed. In the present study, the nozzle was accompanied by a red CFV (1.5 bar pressure) which delivers a volume of 570 ml of insecticide solution/ min (for 19 m 2 ) or 30 ml/m 2 .
(i) Dilution of SumiShield 50WG. The capacity of the sprayer used (Hudson Xpert, Chicago, USA) was 6 l. The sprayer was rinsed with clean water after each treatment. For a volume of 6 l, the dilution was: Q = (300 mg × 6000 ml) / 30 ml = 60,000 mg of clothianidin in 6 l of water. The SumiShield 50WG contained 50% clothianidin. Therefore 120 g (120,000 mg) of SumiShield 50WG was diluted in 6 l of water to treat 200 m 2 . To manage the remaining diluted insecticide after treatment, each insecticide solution was adjusted to each hut surface area. For hut 1 (17 m 2 ), the dilution was 10,200 mg SumiShield 50WG in 510.0 ml of water. For hut 5 (16.74m 2 ), dilution was 10,044 mg SumiShield 50WG in 502.0 ml of water (ii) Dilution of K-Othrine 250WG. For a volume of 6 l of the sprayer (Hudson Xpert) used, the dilution was: Q = (6000 ml × 25 mg) / 30 ml = 5000 mg of deltamethrin in 6 l of water. K-Othrine 250WG contained 25% of deltamethrin; therefore, the dilution was 20 g (20,000 mg) of K-250 WG Othrine in 6 l of water to treat 200 m 2 . To manage the rest of diluted insecticide after treatment, each insecticide solution was adjusted to each hut surface area.For hut 3 (16.86 m 2 ), the dilution was 1686 mg of K-250 WG Othrine in 506 ml of water. For hut 4 (16.00 m 2 ), the dilution was 1600 mg of K-250 WG Othrine in 480 ml of water. Huts 2 and 6 were used as control (untreated).

Measurement of pH of the wall substrates
The purpose of this activity was to intermittently follow the evolution of the pH of the wall surfaces and the efficacy of the impregnated insecticides as well. A small quantity (5 g of substrate) from the wall surface of each hut was scraped into a Petri dish before, during and at the end of the study. The substrates were dissolved in distilled water for the determination of the pH in CREC.
Status of resistance of the wild An. gambiae population of the study area Susceptibility tests using WHO tubes were performed using 2-5 day-old aged adult female mosquitoes. Detection of Leu-Phe kdr mutation was performed with untested mosquito samples using PCR following the protocol described by [8].

WHO cone bioassay
For each month, WHO cone bioassays were performed. A laboratory colony of An. gambiae "Kisumu" strain which were fully susceptible to all insecticides and a wild population of An. gambiae (s.l.) resistant to pyrethroids sampled around the study site were used. The WHO cone bioassays [9] were carried out in week 1, and each month after spraying. During the whole study, cones were placed at fixed level (0.5, 1, 1.5 and 2 m) on the walls (treated and untreated huts). Each level was randomly selected and labelled to one wall surface of each hut. For each month, at least 10 females of the An. gambiae susceptible reference strain "Kisumu" and wild An. gambiae (s.l.) mosquitoes aged 3-5 days were introduced per cone for 30 min exposure. After this exposure period, knockdown mosquitoes were recorded and those alive were kept in observation for 24, 48, 72, 96 and 120 h to score delayed mortalities. Mosquitoes exposed to unsprayed substrates were used as controls. When the mortality in the control was between 5 and 20%, corrected mortality was determined using Abbot's formula.

Statistical analysis
The raw data were managed with Microsoft Excel and the statistical analyses were performed using R software. The WHO criteria were used to classify the level of resistance of tested mosquitoes from Covè to insecticides. The WHO bio-efficacy threshold was used for the analyses of the residual effect in time per treatment. Others analyses on the significance between each of the bio-efficacy measured parameters in treated huts were made and compared between treatment and control (untreated) using R software.

Analysis of pH of samples of wall substrates
Analysis of the pH of wall samples of the huts was done 7 days prior to spraying, and 4 and 8 months after spraying in the huts. The mean pH observed is indicated in Table 1. No significant variation of mean pH was observed between insecticide treatments at the observation times (4 and 8 months after spraying).
Resistance level of An. gambiae (s.l.) population from the experimental huts station Prior to spraying, the resistance status of the population of An. gambiae (s.l.) from the study site was investigated. The results showed that the Anopheles population from the experimental hut station was resistant to deltamethrin 0.05% (mortality rate: 27.45%) and susceptible to pirimiphos methyl 0.25% (mortality rate: 100%). A resistance to bendiocarb 0.1% was also suspected (Fig. 2).

Resistance mechanisms in Anopheles population
PCR results showed a high frequency (0.95) of knockdown resistance (kdr) and very low frequency (0.03) of acetylcholinesterase insensitive mutation (Ace-1) in the Anopheles population from the experimental hut station ( Table 2).
Efficacy of clothianidin 50WG against host-seeking Anopheles populations in the experimental huts station Over the period of eight months of evaluation of the efficacy of clothianidin 50WG in comparison to that of deltamethrin 250WG, 24,135 host-seeking An. gambiae (s.l.) were collected from the six experimental huts used in this study. Among these mosquitoes, 35. 7% (8609), 31.4% (7577) and 33.0% (7949) were collected from the untreated huts (control huts), the huts treated with clothianidin and those treated with deltamethrin, respectively. The results of the measured parameters are shown below.

Exophily
Induced exophily is the proportion of mosquitoes that exited the huts and were thus found in the exit trap veranda. Overall, all treatments induced in the hostseeking An. gambiae (s.l.) population an exit behavior from the treated surfaces to the exit trap's veranda (Table 3). However, significantly higher exophily rates were observed with deltamethrin (P < 0.0001).

Blood-feeding
Blood-feeding inhibition describes the reduction in bloodfeeding compared with that in the control huts. Irrespective of whether the huts were treated or not, no difference was observed on the blood-feeding of mosquitoes in the experimental huts (Table 4). The blood-feeding rates were over 80% in the treated and control huts.

Toxicity effect
The toxicity or lethal property of deltamethrin was low (ranging between 1 and 40%). The mortality rates recorded with this insecticide decreased after one month (Fig. 3). However, high mortality rates (mean mortality rate of 91.7% at the 120 h observation point) were recorded over a period of eight months on the evaluation with clothianidin. While good initial mortality was observed, mortality continued to increase across observation times to 120 h indicating the additional killing properties of clothianidin.
Efficacy of clothianidin 50WG against laboratory susceptible Anopheles populations in the experimental huts station The results from releasing activities are summarized in Tables 5, 6 and Fig. 4. The exophily and blood-feeding rates observed in the huts treated either with clothianidin or deltamethrin were not significantly higher than those observed in the untreated huts (Tables 5, 6). However, clothianidin induced high mortality rates (about 100%) post-120 h observation time over a period of seven months of the evaluation. The induced mortality rate by deltamethrin decreased to about 80% at month 8. The additional killing effect of clothianidin over time against the susceptible Anopheles population released in the huts was also noticed (Fig. 4).
Residual effect of clothianidin 50WG against susceptible strain "Kisumu" and wild in experimental huts

Cone bioassay results
The efficacy of each treatment in terms of mortality rate observed after exposure of mosquitoes to treated walls was compared to the WHO bio-efficacy threshold [9], 80% and to that recorded in the untreated huts. The mortality rates over 5% in control huts were corrected using Abbott's formula. During the eight months of the evaluation, the huts treated with clothianidin showed the mortality rates over 80% at the 120 h observation point against susceptible Anopheles population "Kisumu" (Fig. 5). Concerning the wild An. gambiae (s.l.) population from Covè (experimental hut station) the mortality rates over 80% were observed until six months in the huts treated    with clothianidin ( Fig. 6. The mortality rate decreased a little below 80% from the seventh month (77.9%) to the eighth month (60.3%). Mortality rates increased each month across observation times demonstrating the additional time-dependent killing effect of clothianidin.

Knockdown effect
The efficacy of each treatment in terms of knockdown rate obtained after 30 min exposure in cone bioassays of mosquitoes to treated walls was compared to the WHO threshold bio-efficacy [9] which is 95% for LLINs, and to that observed in the untreated huts. Overall, the knockdown rates in the structures treated with clothianidin were very low (< 95% for most) compared with those obtained in the huts treated with deltamethrin (Fig. 7).

Discussion
The present study assessed the efficacy and residual effect of a novel mode of action Indoor Residual Spray product, SumiShield 50WG under semi-field conditions in West Africa (Covè, Bénin) where An. gambiae (s.l.) is resistant to pyrethroids. SumiShield was evaluated against a population of An. gambiae characterized by a high frequency (0.95) of knockdown resistance (kdr) and elevated oxidase and glutathione S-transferase (GST) activities. After eight months of evaluation, SumiShield 50WG showed a better efficacy and lasting residual effect compared to deltamethrin. The results were encouraging since deltamethrin was ineffective against a pyrethroid resistant host-seeking Anopheles population from Covè during the whole study period. The exophily of host-seeking wild An. gambiae (s.l.) recorded with deltamethrin was significantly higher than what was observed in the control huts. This is not unexpected for pyrethroids. During the evaluation period, the blood-feeding rates were high in both treated and control huts. Irrespective of the huts treated or not, the blood-feeding behavior was the same. This is, however, not worrying; many phase II and phase III evaluations implemented in Benin continuously demonstrated that despite the treatment of the houses with insecticides, the majority of mosquitoes successfully enter these treated houses and take their blood meal of their host before resting on the treated walls [10,11]. Indeed, this is not unexpected  for IRS treatments except for DDT that has some spatial action as well as contact activity. This finding explains why it was proposed to Benin NMCP to always educate communities who are protected by IRS to additionally sleep under LLINs to supplement and maximize malaria control efforts. Such combination strategy implicates an increase in the cost of malaria prevention, but it is desirable for areas with high levels of malaria transmission. Fortunately, 95% of blood-fed An. gambiae (s.l.) died after resting on the treated walls in huts treated with clothianidin 50WG. This will keep them from staying alive to continue malaria transmission. Furthermore, community-wide use of SumiShield 50WG in IRS will produce a "mass effect" on the reduction of the density of infective mosquitoes in the area and, consequently, protecting the whole community including those whose houses are not treated.
Given that SumiShield 50WG has demonstrated at least eight months (study period) efficacy in this study, this insecticide could be a good solution for IRS in areas of permanent malaria transmission. Clothianidin is a novel neonicotinoid insecticide acting as an agonist of the nicotinic acetylcholine receptor (nAChR). This  [8]. The knockdown rates observed with clothianidin were low: this is not unexpected given the mode of action of this insecticide. Additionally, clothianidin has been intensively used in agriculture. Several studies have demonstrated that clothianidin is highly active not only against hemipteran insects but also coleopteran, thysanopteran, dipteran and some lepidopteran pests [12]. Because of its broad spectrum of insecticidal activity, good systemic properties and low mammalian toxicity, clothianidin is a compound that is considered to be compatible with integrated pest management strategies [13]. The use of the new cost-effective, long-lasting IRS insecticides with a new mode of action such as SumiShield for malaria control or elimination in  Residual effect of SumiShield 50WG and deltamethrin 250 WG represented by the mortality rates observed following cone bioassays against laboratory susceptible strain "Kisumu" in the experimental hut of Covè. Sumi24h, Sumi48h, Sumi72h, Sumi96h and Sumi120h refer to the mortality rates observed each month 24, 48, 72, 96 and 120 h observation times in the huts treated with SumiShield, respectively. Delta24h, Delta48h, Delta72h, Delta96h and Delta120h refer to the mortality rates observed each month 24, 48, 72, 96 and 120 h observation times in the huts treated with deltamethrin, respectively endemic countries will help support resistance management and the optimization of vectors control strategies. However, NMCPs should recommend the judicious use of these new insecticides in preventing the early selection and development of resistance in these malaria parasite vectors.
The additional killing effect of clothianidin over time against both susceptible and resistant Anopheles populations was noticed. Given this, it is important to study the impact of this action on the fertility of female malaria vectors subjected to clothianidin in terms of the number of eggs laid and the viability of the embryo from the egg. Furthermore, as the performance of clothianidin as observed in this study was beyond six months, it would also be important to implement another study at community level (small scale Phase III trial) to assess the efficacy and residual effect of SumiShield 50WG under field conditions.