Insecticide susceptibility of Aedes aegypti and Aedes albopictus in Central Africa

Background Aedes aegypti (Linnaeus, 1762) and Aedes albopictus (Skuse, 1894) are the main vectors of dengue (DENV) and chikungunya (CHIKV) viruses worldwide. As there is still no vaccine or specific treatment for DENV and CHIKV, vector control remains the cornerstone of prevention and outbreak control. Unfortunately, vector control programs are facing operational challenges with mosquitoes becoming resistant to commonly used insecticides in several areas through the world. Throughout Central Africa no recent data are available susceptible/resistant status of either vector species since the introduction/arrival of Ae. albopictus in this area. We therefore studied the level of resistance of these two major vectors to insecticides commonly used in Africa for mosquito control. Results Aedes aegypti and Ae. albopictus were sampled in six urban localities of Cameroon (Garoua, Bertoua, Yaoundé, Bafia, Buea) and Gabon (Libreville). Larval bioassays, carried out to determine the lethal concentrations (LC50 and LC95) and resistance ratios (RR50 and RR95) suggested that both vector species were susceptible to Bti (Bacillus thuringiensis var israeliensis) and temephos. Bioassays were also performed on adults using WHO diagnostic test kits to assess phenotypic resistance to deltamethrin, DDT, fenitrothion and propoxur. These experiments showed that one population of Ae. aegypti (Libreville) and two populations of Ae. albopictus (Buea and Yaoundé) were resistant to DDT (mortality 36% to 71%). Resistance to deltamethrin was also suspected in Ae. albopictus from Yaoundé (83% mortality). All other field mosquito populations were susceptible to deltamethrin, DDT, fenitrothion and propoxur. No increase in the knockdown times (Kdt50 and Kdt95) was noted in the Yaoundé resistant population compared to other Ae. albopictus populations, suggesting the possible involvement of metabolic resistance to deltamethrin and DDT. Conclusion In view of the recent increase in dengue and chikungunya outbreaks in Central Africa, these unique comparative data on the insecticide susceptibility of Ae. aegypti and Ae. albopictus could help public health services to design more effective vector control measures.


Background
Dengue virus (DENV, Flaviviridae, Flavivirus) and chikungunya virus (CHIKV, Togaviridae, Alphavirus) are mosquito-borne viruses of medical concern in most tropical regions. With about 50-100 million reported cases annually, including 500 000 severe cases of dengue haemorrhagic fever (DHF) or dengue shock syndrome (DSS), DENV is the most prevalent mosquito-borne human virus worldwide [1]. In West and Central Africa, where DENV epidemics remained limited until 5 years ago, the number of outbreaks or case reports has increased significantly in several areas, including Cameroon in 2006 [2,3], Gabon in 2007 [4], and West African countries such as Mali in 2008, and Cape Verde and Senegal in 2009 [5]. Similarly, CHIKV, which previously caused only sporadic outbreaks in sub-Saharan Africa [6], has recently emerged in several urban epidemic foci in Central Africa [4].
Aedes aegypti (Linnaeus, 1762) and Aedes albopictus (Skuse, 1894) are the main epidemic vectors of DENV and CHIKV worldwide [2,7]. Both species occur in sub-Saharan Africa: Ae. aegypti is native to Africa and Ae. albopictus has recently invaded several Central African countries [8]. Recent observations in Cameroon [9] and Gabon [10] indicate that these two species infest urban environments.
As there is still no vaccine or specific treatment for DENV or CHIKV, vector control remains the cornerstone of prevention and outbreak control. Conventional control strategies rely on the reduction of larval sources by eradicating water-holding containers that serve as larval habitats, and by using larvicides (e.g. temephos and Bacillus thuringiensis var israeliensis [Bti]) in natural and/or domestic breeding sites [11]. Space spraying (using pyrethroids or organophosphates) is generally used when larval source reduction fails to reduce the density of adult mosquitoes, and also during outbreak situations [12]. Unfortunately, many vector control programs are threatened by the development of insecticide resistance in Ae. aegypti and Ae. albopictus [13][14][15][16]. Resistance to multiple insecticides (e.g. pyrethroids and organophosphates) has been reported in Ae. aegypti in South-East Asia [17,18], South America and the Caribbean [19][20][21]. There have been fewer reports of insecticide resistance in Ae. albopictus [15,22], although a significant decrease in susceptibility to permethrin, malathion and temephos has been observed in Thailand [14]. In West and Central Africa, most available data on pesticide resistance concern Ae. aegypti but date back more than 30-40 years [23], while virtually no such data are available for Ae. albopictus, owing to its recent introduction in Africa. In order to implement effective and sustainable arbovirus vector control measures, there is an urgent need to determine the susceptibility of the two major vectors of DENV and CHIK to insecticides commonly used for mosquito control. We therefore examined the distribution and insecticide susceptibility of both larval and adult Ae. aegypti and Ae. albopictus populations in sentinel sites from Central Africa (Cameroon and Gabon) using WHOPES-approved procedures [24,25].

Mosquito sampling
Aedes aegypti and Ae. albopictus mosquitoes were sampled as larvae or pupae in six urban localities ( aegypti population (Bénoué) was sampled in the Bénoué National Park (08°15'N; 13°49'E). The latter population, which is sylvan and thus considered as naïve to pesticide, was used as a susceptible reference strain. Immature stages (field generation, F0) were collected from water in domestic (e.g. jars, tanks), peri-domestic (e.g. tires) and natural environments (e.g. tree holes). For each sampling site, larvae or pupae from 2-6 larval breeding places were collected, stored in plastic boxes and transferred to insectaries for rearing to the adult stage. Once identified as Ae. aegypti or Ae. albopictus, mosquitoes of the same species and from the same locality were pooled to prevent inbreeding. Mosquitoes were reared to generation F1 for larval and adult bioassays. Mosquito populations were maintained at insectaries conditions (27°C +/-2°C; relative humidity 80% +/-10%) and females were fed on rabbits to complete their gonotrophic cycle.

Larval bioassay
The susceptibility of larvae to temephos (organophosphate, Sigma Aldrich-Pestanal, Seelze, Germany) and Bti (VectoBac, Bayer CropScience, Monheim, Germany) was assessed using standard WHO bioassays [24]. Stock solutions and serial dilutions were prepared in 95% ethanol (temephos) or mineral water (Bti) and were stored at +4°C. Each bioassay used 25 late-third or early-fourth instars in plastic cups with 99 ml of mineral water and 1 ml insecticide solution at the required concentration. We used 5 concentrations in the range of activity of each insecticide and 50 to 100 larvae per concentration (with two to four replicates, depending on the sample and the number of larvae available). Each bioassay included a control group which received 1 mL of ethanol (for temephos) or 1 mL of mineral water (for Bti) only. Tests were run at 27°C ± 2°C, and mortality was assessed after 24 h of insecticide exposure. For each mosquito strain, one replicate corresponded to a single rearing batch.

Adult bioassay
The procedure used for adult bioassays followed the standard WHO protocol [25]. Papers were impregnated using acetone solutions of insecticide and silicone oil as the carrier at the "Laboratoire de Lutte contre les Insectes Nuisibles" (LIN), Montpellier, France (WHO Collaborating Centre) using diagnostic concentrations defined for Aedes laboratory strain (Bora) susceptible reference strain [17]. Four technical-grade compounds representing the 4 major classes of insecticide (carbamate, organochlorine, pyrethroid and organophosphate) were used as follows: 0.3% propoxur (Bayer CropScience, Monheim, Germany), 4% DDT (Hindustan Insecticides Limited, New Delhi, India), 0.06% deltamethrin (Agrevo Environmental Health, Berkhamsted UK) and 0.5% fenitrothion (SUPELCO, Bellefonte, USA). Two to four batches of 25 non-blood-fed females (2-4 days of age) were introduced into exposure tubes containing impregnated filter papers for 60 minutes. The number of knocked-down (Kd) mosquitoes was recorded every 5 minutes. The mosquitoes were then transferred to a recovery tube containing 10% glucose solution and maintained at 27°C ± 2°C with 80% ± 10% relative humidity. Mortality was recorded 24 hours post-exposure.
For each strain a batch of 25 mosquitoes was used as control.

Data analysis
Larval and adult mortality rates were corrected using Abbott's formula (Abbott 1925) in case of control mortality >5% but less than 20%. Data from larval bioassays were analyzed with Win DL 2.0 software [26], which uses the iterative method of maximum likelihood to fit a linear regression between the log of the insecticide concentration and the probit of mortality. Goodness-of-fit was estimated with the Pearson χ 2 test. Win

Larval bioassay
Larval bioassays using Bti showed that LC 50 and LC 95 for Ae. albopictus ranged from 0.18 mg/l for the Bénoué strain (susceptible reference strain) to 0.69 mg/L for the Yaoundé and Buea populations. Against Ae. aegypti, Bti acted in a similar range of concentrations, with LC 50 and LC 95 ranging from 0.18 mg/L for Garoua to 0.91 mg/L for Bafia (Table 1). RR 50 ranged from 1.00 (Garoua) to 1.50 (Bertoua) and RR 95 from 1.13 (Libreville) to 2.84 (Bafia). These values and their 95%CIs indicated that all Ae. aeggyti and Ae. albopictus populations were similarly susceptible to Bti.
The estimated knockdown times (KdT) for Ae. aegypti mosquitoes exposed to deltamethrin indicated that the KdT 50 and KdT 95 values of the field populations were similar (overlapping 95%CIs) to those of the reference strain (Table 4). KdT 50 values recorded in Yaoundé and Buea Ae. albopictus strains were higher than for the reference strain. Despite the ICs calculated for these two strains did not overlapped with the IC of the reference, the increase in KdT seemed no significant (P > or = 0.05). The DDT KdT 50 for all Ae. albopictus samples and the KdT 95 for Ae. albopictus Bertoua were significantly higher than the corresponding values for all Ae. aegypti samples. These differences in KdT between Ae. aegypti and Ae. albopictus may be physiological and/or species-dependent.

Discussion
All Ae. aegypti and Ae. albopictus samples collected in Cameroon and Libreville (Gabon) were susceptible to Bti and temephos. Resistance to Bti has been described in Culex pipiens in the USA [28] and suspected in Aedes rusticus in France [29], but no resistance to this pesticide has been reported to date in Ae. aegypti or Ae. albopictus worldwide. Temephos is the most widely used larvicide for dengue vector control. Resistance to this organophosphate has been recorded in Ae. aegypti in Asia [30][31][32][33][34] and South America [35][36][37], and in Ae. albopictus in Malaysia [38] and Thailand [14]. Mouchet et al. [23] observed full susceptibility of Ae. aegypti larvae to temephos in Cameroon and Gabon (Libreville and Yaoundé) during the 1970s. Up to now, larvicides targeting these two species have rarely been used in African countries, thus explaining the persistent susceptibility to temephos and Bti in this part of the world. WHO bioassays carried out on adult mosquitoes showed DDT resistance in one Ae. aegypti population from Gabon (Libreville) and two Ae. albopictus populations from Cameroun (Buea and Yaoundé). In addition, DDT resistance was suspected in the Ae. albopictus sample from Bertoua. DDT resistance has been widely found in Ae. aegypti worldwide, including in Cameroon [23]. Kdr mutation of the voltage-gated sodium channel has been found to confer resistance to DDT in Ae. aegypti [39] but detoxifying enzymes such as glutathione S-transferases (GSTs) can also play a key role in DDT metabolism and resistance [40]. Although DDT resistance has also been recorded in Ae. albopictus in Thailand and   50 and LC 95 , are expressed in mg/liter; 95%CI: 95% confidence interval; p > 0.05 suggests a well-fitting model, P < 0.05 suggests an invalid model population.
This study also showed full susceptibility of Ae. aegypti and Ae. albopictus to deltamethrin, except in Yaoundé where the Ae. albopictus population showed mortality rates of around 80%, strongly suggesting resistance. Pyrethroid resistance in Ae. aegypti is widespread [43,15] and is usually associated with an altered amino acid sequence of the target protein, the sodium channel that confers knockdown resistance when altered (the so-   called Kdr mutation) [39,44]. Pyrethroid resistance has also been linked to an increase in metabolic process through over-transcription of detoxification genes [21]. Although the number of studies reporting insecticide resistance in Ae. albopictus has increased this decade [17,41,42,45,46], few have shown evidence of pyrethroid resistance [14]. The Kdr mutation has been found in Ae. aegypti [39,16] but has never been reported in Ae. albopictus. Further molecular and biochemical studies are needed to identify the genetic basis of pyrethroid resistance in Ae. albopictus in Yaoundé. The absence of cross-resistance between DDT and deltamethrin in all mosquito populations (except Ae. albopictus collected in Yaoundé) suggests the involvement of metabolic resistance rather than a target site modification (Kdr mutation) in the sodium channel. It is unclear whether cross-resistance to deltamethrin and DDT in the Ae. albopictus population in Yaoundé is due to common or different resistance mechanisms. The source of DDT and deltamethrin resistance observed here in DENV and CHIKV vectors is unclear. Indeed, in Central Africa, insecticide treatment specifically targeting Ae. aegypti or Ae. albopictus is extremely limited. It is possible that insecticides used to control other insects of medical or agricultural interest exert indirect selection pressure on these two mosquito species. For example, indoor residual spraying of DDT for malaria control [47] was suspected of favouring the selection of DDT resistance in Anopheles [48] as well as in Aedes [42,45]. Contamination of larval breeding places by insecticides used in agriculture (cotton, vegetables) has also been shown to select for DDT and pyrethroid resistance in malaria vectors [48][49][50][51][52]. Concerning Ae. albopictus, which recently spread to Cameroon [8], one cannot exclude the possibility that the invading population possessed a resistance background.

Conclusion
This is the first study of the susceptibility status of Ae. aegypti and Ae. albopictus to various insecticides in Central Africa. The increase in dengue and chikungunya outbreaks in this region calls for more robust surveillance and vector control. Our observations could help to guide insecticide-based strategies, although broader monitoring of insecticide resistance in Aedes mosquitoes is needed. Further molecular and biochemical studies are also needed to determine the mechanisms involved in insecticide resistance among arboviruses vectors in Central Africa.