Insecticidal and sterilizing effect of Olyset Duo®, a permethrin and pyriproxyfen mixture net against pyrethroid-susceptible and -resistant strains of Anopheles gambiae s.s.: a release-recapture assay in experimental huts

In the context of the widespread distribution of pyrethroid resistance among malaria vectors, we did a release-recapture trial in experimental huts to investigate the insecticidal and sterilizing effects of a novel long-lasting net (LN), Olyset® Duo, incorporating a mixture of permethrin (PER) and the insect growth regulator (IGR), pyri-proxyfen (PPF). An LN containing PPF alone and a classic Olyset® Net were tested in parallel as positive controls. The effect of progressive number of holes (6, 30, or 150) that may accrue in nets over time was simulated. We used two laboratory Anopheles gambiae s.s. strains: the susceptible Kisumu strain and the pyrethroid-resistant VK-Per strain having solely kdr as resistance mechanism. The effect of these nets on the reproductive success of blood-fed females that survived the different LNs conditions was recorded. Regardless of the mosquito strain, the LNs containing PPF alone with as many as 30 holes drastically reduced the number of eggs laid by females succeeding in feeding, i.e. fecundity by 98% and egg hatching rate (fertility) by 93% relative to untreated control net. Very few of the resistant females blood fed and survived under the Olyset® Duo with similar number of holes (up to 30) but of these few, the inhibition of reproductive success was 100%. There was no evidence that the Olyset® Duo LN with 150 holes impacted fecundity or fertility of the resistant colony. The efficacy of Olyset® Duo is encouraging and clearly illustrates that this new net might be a promising tool for malaria transmission control and resistance management.


Introduction
During the last decade, expansion of financial support and significant scale-up of malaria control tools have led to a large reduction in malaria incidence and/or mortality [30]. Despite this major reduction of the burden, malaria is still a major public health concern, with an estimated 207 million cases and 627,000 deaths that occurred in 2012 [30].
Moreover, the rapid and widespread resistance to insecticides and antimalarial drugs represents a serious threat to sustainability of current malaria control tools [2,17,29,30]. Major mechanisms of insecticide resistance involve either an alteration in the rate of insecticide detoxification (metabolic mechanisms) and/or mutations within the target site of the insecticide. There two main mechanisms confer phenotypic resistance to the four insecticide families used for malaria vector control (carbamates, organophosphates, organochlorines, and pyrethroids) [10,27]. The pyrethroid insecticides are the only class deployed for bed net impregnation because they are highly efficacious, fast-acting against insects at low doses, and have acceptably low toxicity to mammals [32]. The development of alternative tools against resistant mosquitoes is therefore an urgent need.
One strategy to prevent or delay the development of resistance is to use at least two insecticides having unrelated modes of action in combination on the same bed net. Many mixtures and combinations have been studied in the last decade with the objective of impregnating bed nets for use in the future [3,11,18,22,24,25]. Among these, the currently available combination products are long-lasting insecticidal treated nets (LNs) impregnated with pyrethroid and synergist, e.g. piperonyl butoxide (PBO). These LNs were developed to specifically target malaria vectors bearing metabolic resistance mechanisms, with encouraging results [4,16,23,28].
Another candidate for combining with insecticides for vector control is an insect growth regulator (IGR), e.g. pyriproxyfen (PPF), that impedes insect morphogenesis, embryogenesis, and reproduction [5]. Due to its unique mode of action that sterilizes adult malaria vectors and the lack of reported resistance to PPF in mosquitoes, PPF could complement the insecticidal arsenal to better control malaria transmission [9]. In this context, Sumitomo Chemical Ó developed a new LN impregnated with both permethrin (PER) and PPF. This new product showed promising performances in the laboratory (Rossignol M, pers. comm.) and against one wild An. gambiae population [19]. Here, its performance was investigated against both susceptible and pyrethroid-resistant laboratory strains of An. gambiae through a release-recapture experiment in experimental huts in Benin. The aim of this trial was to measure the efficacy of the new LN incorporating both PER and PPF in comparison with nets impregnated with PER or PPF alone (as positive controls) and untreated net (as negative control). The impact of these nets on the reproductive success of susceptible and resistant An. gambiae was of interest and therefore investigated when accrued numbers of holes appear in them in the field.

Study area
The study was carried out in four experimental huts situated in Akron (6°30 0 N and 2°47 0 E), a village on the periphery of Porto Novo, the administrative capital of Benin. The site is a horticultural area covering approximately 20 ha.

Mosquito strains
Two strains of Anopheles gambiae s.s. were used: -Kisumu is a standard susceptible strain originating from Kenya, maintained for many years in the laboratory and free of any detectable insecticide resistance mechanisms (VectorBase, http://www.vectorbase.org, KISUMU1). The susceptibility of Kisumu mosquitoes to insecticides is checked monthly using WHO test kit and the absence of resistance mechanisms was also checked by PCR and biochemical assays. -VK-Per originated from Burkina Faso and was exposed to constant permethrin selections in the laboratory until becoming homozygous for the kdr mutation (L1014F). The fixed status of the kdr mutation in this strain is checked monthly. VK-Per strain displayed the same expression level of metabolic resistance enzyme as Kisumu [17].

Design of huts
The huts are made from concrete bricks, with a corrugated iron roof, a ceiling of thick polyethylene sheeting, and a concrete base surrounded by a water-filled channel to prevent entry of ants. Mosquito access is via four window slits constructed from pieces of metal, fixed at an angle to create a funnel with a 1 cm wide gap. Mosquitoes fly upward to enter through the gap and downwards to exit; this precludes or greatly limits exodus through the aperture enabling the majority of entering mosquitoes to be accounted for. A single veranda trap made of polyethylene sheeting and screening mesh measuring 2 m long, 1.5 m wide, and 1.5 m high, projects from the back wall of each hut. Movement of mosquitoes between hut and veranda is unimpeded during the night.
Before testing in the experimental huts, the nets (including control) were deliberately holed with three different holing conditions to mimic possible wear and tear under natural condition of use: -6 holes of 4 cm · 4 cm as per WHO guidelines, corresponding to 0.07% of the surface of a family size Olyset Ò Net, distributed proportionally on each side of the net; -30 holes of 4 cm · 4 cm corresponding to 0.36% of the surface of a family size Olyset Ò Net, distributed proportionally on each side of the net; -150 holes of 4 cm · 4 cm corresponding to 1.83% of the surface of a family size Olyset Ò Net, distributed proportionally on each side of the net.
In order to even out the surface content of active ingredient on the different types of nets, nets were washed three times to remove initial content according to the WHOPES standard washing methods and active ingredient left to regenerate over one week prior to the release-recapture trial in huts [31].
For each treatment and each holing condition, four release sessions (50 mosquitoes each) of An. gambiae Kisumu and four of VK-Per strains were run in each hut, respectively, allowing a complete rotation of treatments in the huts.

Volunteer participants' recruitment and mosquito collections
Adult volunteers recruited among inhabitants of villages close to the field site signed a written informed consent form. Their vaccination status against yellow fever was checked before enrolment and they were medically followed up throughout the study. Sleepers were rotated among huts and treatments each night of the study following a Latin square design. They entered into the hut at 21:00 and remained inside until dawn. The Ethics Committee of the Health Ministry of Benin approved the protocol for the study.
All window slits were blocked to prevent entry/exit of any free wild mosquitoes. Batches of 50 unfed mosquito females 3-5 days old were released in the evening (21:00) when sleepers got under the net. In the morning (06:00), the sleeper collected dead mosquitoes from the floor of the hut, the veranda traps, and inside the nets. Resting mosquitoes were collected using glass tubes from inside the net, on the walls, and ceiling of the hut and exit trap. Mosquitoes were scored by location as dead or alive and as fed or unfed. Live mosquitoes were placed in small cups with access to honey solution for 24 h to assess delayed mortality.
The primary outcomes measured were [31]: -induced exophily (the proportion of mosquitoes found in the veranda relative to the total collected mosquitoes); -blood-feeding inhibition (the reduction in blood feeding of mosquitoes in the treatment hut relative to the control huts); -immediate and delayed mortality (the proportion of dead mosquitoes at the time of collection and those dying after 24 h).
All outcomes were calculated as the proportions of recaptured mosquitoes.

Fecundity and fertility assessment
Living blood-fed females collected from all compartments of the huts were counted and brought back to the laboratory. For each treatment, blood-fed females were put by batch of up to 20 females in cardboard cups (450 mL). We previously filled the bottom of the cardboard cups with a 1 cm high layer of wetted cotton, covered with a filter paper disk to allow females to lay their eggs. The females in cardboard cups were maintained with honey solution in an observation room at 28°C and 80% RH for 5 days. After 5 days, we checked the mortality of females in cups. We took a picture of each filter paper to count the eggs using Egg Counter software [15]. All eggs from the control and treated batches were immersed in water. After 6 days, the number of larvae (3rd and 4th instars) was checked to determine the hatching rates. Efficacy was measured in terms of fecundity (number of eggs laid per female) and fertility (% of eggs that hatched).

Statistical analysis
The software ''R'' was used for the statistical analyses [26]. The proportions of mosquitoes that attempted to exit were killed within the hut and had blood fed successfully, were analyzed using a logistic regression model. The ''brglm'' function from the brglm package was used for the analysis [13]. It allowed the fitting of binomial-response regression models using the bias-reduction method developed by Firth [7]. These procedures return estimates with improved frequentist properties (bias, mean squared error) that are always finite, even in cases where the maximum likelihood estimates are infinite (data separation).
Count data related to the actual number of eggs laid by batches of surviving females and those hatching (i.e. number of larvae per surviving female) were analyzed using negative binomial regression. Differences in mortality and KD rates A. Djènontin et al.: Parasite 2015, 22, 27 between the treatments from the WHO cone bioassays were compared using the v 2 test.

Insecticidal efficacy of treatments
The trial was carried out between June 21, 2012 and August 2, 2012. Four release-recapture replicates were run for each holing configuration and each strain, leading to a total of 4800 mosquitoes released. Of these, 1883 An. gambiae s.s. Kisumu and 2087 An. gambiae s.s. VK-Per were recaptured. The results for the treatment-induced exophily, blood-feeding inhibition, and induced mortality are summarized in Table 1 for LNs condition with 6, 30, and 150 holes.

Control
The proportions of the resistant VK-Per mosquitoes that blood fed under the untreated net (84-98%) were significantly higher than with the Kisumu strain (36-87%) (p < 0.001).
The blood-feeding rates increased with the number of holes. The mortality rates in the control hut were universally low (<3%), regardless of the strain or the number of holes, indicating that there was no contamination or carryover of active ingredient from one hut to another during the trial.

PPF-treated net
Regardless of the hole configurations and the mosquito strains, PPF-LN alone induced less than 4% mortality. Blood feeding of Kisumu strain ranged from 40% (6 holes) to 86% (150 holes) and that of VK-Per ranged from 80% (6 holes) to 100% (150 holes). The induced exophily was not significantly different from the control treatment except for Kisumu strain with 150 holes.

Olyset Ò Net
Blood-feeding rates were highly inhibited with the Kisumu strain (74-79%) and inhibited to a lesser extent with the VK-Per strain (16-34%). Mortality rates corrected for control were nearly 100% with Kisumu and significantly lower with VK-Per mosquitoes (51-82%). With Olyset Ò Net, fewer Kisumu mosquitoes were collected in the veranda trap compared to the control hut, while significantly more VK-Per mosquitoes were collected in the veranda trap with Olyset Ò Net with 6 holes and 150 holes (p < 0.05).

Olyset Ò Duo
Olyset Ò Duo strongly inhibited blood-feeding rates of the Kisumu strain by 79-100%, regardless of the number of holes. With both strains, the BFI rates for Olyset Ò Duo were The numbers in the same column sharing the letter superscript do not differ significantly (p ! 0.05).
significantly higher than those of the Olyset Ò Net (p < 0.05) although within each hole category. The impact on Kisumu (79-100%) was greater than for VK-Per (27-87%). BFI decreased with the number of holes in the nets. The decrease was more pronounced with the VK-Per than the Kisumu strain: between 79% and 91% of Kisumu were still prevented from blood-feeding under the LNs with 150 holes compared to only 16-27% for VK-Per under the same type and condition of LNs. Olyset Ò Duo killed 100% of Kisumu mosquitoes whereas it killed between 73% and 92% of VK-Per. Overall, Olyset Ò Duo killed significantly greater proportions of the VK-Per mosquitoes than did Olyset Ò Net (p < 0.05).

Fecundity and fertility of surviving blood-fed females
During the study of fecundity and fertility, we scored 51,790 eggs laid by 1682 surviving blood-fed females collected from the release-recaptures and 41,328 derived larvae. The results of fecundity and fertility relative to the LN products are summarized in Table 2 for LNs conditions with 6, 30, and 150 holes.
Firstly, the proportions of surviving blood-fed females were analyzed. Both the Olyset Ò Net and Olyset Ò Duo killed all blood-fed Kisumu females and drastically decreased survivorship of blood-fed VK-Per females (4-12% survival) compared to the control (92%) (p < 0.001). There was no significant difference in survivorship between the PPF LN and the control. The highest proportions of live blood-fed females were observed in huts with LNs having 150 holes (43% for the Olyset Ò Net and 25% for Olyset Ò Duo).
Analysis of fecundity in live blood-fed Kisumu females showed that the PPF-LN reduced the mean number of eggs per female by >98% (p < 0.001).
The Olyset Ò Net never impacted the mean number of eggs per live VK-Per female (p > 0.05). None of the very few (6-7 blood-fed VK-Per Anopheles) that survived the Olyset Ò Duo with 6 and 30 holes laid eggs, leading to 100% fecundity reduction. More VK-Per females blood-fed and survived the Olyset Ò Duo LN with 150 holes but with this level of holes there was not a significant effect on the mean number of eggs produced per female compared to the control net with the same number of holes (p > 0.05) ( Table 2).
The analysis of the fertility of the eggs laid by blood-fed females that survived showed that PPF LN alone significantly reduced the hatching rate for both Kisumu and VK-Per strains, regardless of the number of holes. There was no clear evidence that the Olyset Ò Duo significantly impacted the hatching rate of mosquitoes (p > 0.05).
The overall IGR effect is illustrated by the mean number of larvae obtained per surviving blood-fed female ( Table 2). Table 2. Summary results of fecundity and fertility among Kisumu and VK-Per mosquitoes exposed to the various treatments.

Discussion
The present study investigated the insecticidal and sterilizing efficacy of the new product, Olyset Ò Duo, on both susceptible and kdr-resistant laboratory An. gambiae s.s. mosquitoes in comparison with LNs impregnated with permethrin (Olyset Ò Net) or PPF alone.
Against the susceptible An. gambiae s.s. Kisumu strain, both Olyset Ò Duo and Olyset Ò Net killed 100% of exposed mosquitoes. By contrast, a reasonable number of kdr-resistant An. gambiae s.s. mosquitoes survived exposure to these LNs, confirming that Kdr mutation by itself confers some level of phenotypic resistance. Differences in behavior close to the bed net might induce a variable contact time with the impregnated surface. It is interesting to note that the killing effect of permethrin increased when it was used in combination with pyriproxyfen on Olyset Ò Duo relative to the Olyset Ò Net for VK-Per. Two hypotheses might explain this pattern: either there is a difference in the bleed rate of permethrin in the combination LN relative to the Olyset Ò Net [19] or an additive/synergistic interaction between permethrin and pyriproxyfen, although no evidence of such an interaction is available in the literature [8,14]. It is also possible that the presence of PPF on Olyset Ò Duo interacts with the irritancy and/or repellence property of permethrin, thereby allowing mosquitoes to rest for longer thus picking up higher doses than Olyset Ò Net alone.
The results clearly showed that pyriproxyfen LN alone reduced the mean number of eggs per blood-fed females. These findings confirmed the results of a previous study demonstrating fecundity and fertility reduction in adult mosquitoes exposed to bed nets treated with PPF [1,20,21]. We showed that Olyset Ò Duo with a number of holes as high as 30 still greatly reduced the mean number of eggs (fecundity) of blood-fed pyrethroid-resistant An. gambiae s.s. that survived this LN. It is worth noting that when hole number increased to 150, Olyset Ò Duo failed to have an impact on fecundity, although this number of holes in PPF-treated net did reduce the mean number of eggs per VK-Per female. This suggests that the irritant effect of permethrin limited the contact between the female mosquitoes and the treated surface, leading to reduced exposure to pyriproxyfen and thereby limited impact on fecundity. The good efficacy of Olyset Ò Duo with 6 or 30 holes to sterilize females was encouraging and clearly illustrated that this new LN might be a promising tool to control pyrethroid-resistant An. gambiae s.s. and has potential for resistance management. Nevertheless, 5 days after exposure (the end of the period given to lay eggs), only 6 out of 22 and 6 out of 13 VK-Per females collected in the experimental hut with the Olyset Ò Duo LN with 6 and 30 holes respectively, were alive. It would be important to replicate the trial to strengthen our results. Moreover, it is also crucial to study the efficacy of such new tools against pyrethroid-resistant mosquitoes bearing other resistance mechanisms such as metabolic alone and combined with the kdr mutation.
The limitation of this study was the number of surviving females. We suggest to use the LN with 30 holes (instead of WHOPES-standard six holes) to investigate the efficacy of Olyset Ò Duo against multi-resistant mosquitoes, in order to increase the number of blood-fed females that survive the Duo net without decreasing the time of contact with the LN too significantly (as suggested by the results with Olyset Ò Duo with 150 holes). In order to better design future experiments, it would also be very useful to study the relationship between tarsal contact time and pyriproxyfen effects as the rationale behind the overall technology of Olyset Ò Duo is to be able to sterilize pyrethroid-resistant mosquitoes that probe through pyrethroid-treated net and survive, to possibly transmit malaria elsewhere. In adult mosquitoes, the period of exposure to pyriproxyfen in relation to mosquito blood meals affects the mosquitoes' ability to produce viable offspring [12]. A recent study showed that only female mosquitoes that blood-fed one day prior to pyriproxyfen exposure produced no viable offspring during that gonotrophic cycle [9]. This requires an unlikely scenario in which mosquitoes would rest more than 24 h on bed nets after feeding. In the present study, released mosquitoes were probably in contact with the pyriproxyfen-treated bed net before their blood meal and just after it. The use of PPF as a complementary vector control tool on mosquito resting places might also be an interesting opportunity. Deploying IRS with pyriproxyfen to complement the protective effect of LN within homes might also be an attractive alternative strategy to explore within the framework of managing insecticide resistance in malaria vectors [6].
We conclude that the new Olyset Ò Duo LN mixing permethrin and pyriproxyfen showed promising results in terms of protective and sterilizing effects against both susceptible and Kdr-resistant An. gambiae s.s. The next step is to study its efficacy against wild multi-resistant An. gambiae s.s. populations from different ecological settings before implementing a larger-scale study with epidemiological outputs.

Conflict of interest
This work was supported financially by Sumitomo Chemical. Funders participated in the study design and the decision to publish, but they have no role in data collection, analysis, and preparation of the manuscript. The authors declare that they have no competing interests.