Increase in susceptibility to insecticides with aging of wild Anopheles gambiaemosquitoes from Côte d’Ivoire

Insecticide-based interventions remain the principal vector control measure in malaria endemic countries. Insect proteins targeted by insecticides generally become insensitive to these compounds through point mutations such as the kdr mutation, altered acetylcholinesterase and GABA mutations [1–3]. Metabolic resistance involves biochemical transformation of an insecticide through mechanisms including enhanced detoxification or sequestration, ultimately reducing the capacity of the insecticide to interact with a target insect protein and cause mortality. Many studies have described the primary involvement of esterases, cytochrome P450 monooxygenases and glutathione-S-transferases in conferring resistance [4, 5]. Major progress has been made in the last decade in identifying genes associated with insecticide resistance and structural changes affecting their functions [1–3]. However, knowledge on vector insecticide resistance status, changing trends of resistance in target vectors and their operational implications remain basic requirements to guide insecticide use in disease control programmes. This knowledge can provide a basis for selecting insecticides and for ascertaining continued susceptibility to insecticides already in use. Standardization of test methods for monitoring insecticide resistance is important as it ensures comparability of data from different sources, for different test populations and assessment periods. Standardized test include the World Health Organization (WHO) susceptibility test [6]. There are three general factors that can affect outcomes from WHO susceptibility test: (i) the physiological status of mosquitoes used in assays, i.e., whether adult females are unfed, blood fed, semi-gravid or gravid; (ii) the age of the adult mosquitoes used; and (iii) the temperature/ humidity under which insecticide exposure is conducted. The WHO guidelines [6] for performing assays advise the use of non-blood fed adult females at 24–48 hours post-emergence, within a test environment of 23-27°C and 70 – 80% relative humidity. However, in many field studies, the age of mosquitoes used varies widely [7–9]. Reduced phenotypic resistance in older mosquitoes has been observed in several laboratory studies [10, 11], but very few studies have explored this phenomenon in natural populations.

Since outcomes from resistance tests can influence decisions in insecticide-based intervention programs, we investigated the extent that different ages of wild population of An. gambiae adult mosquitoes can influence phenotypic resistance to insecticides of four different classes used in malaria control, and whether the current methodology is likely to result in variable classification of resistance status.

Anopheles pre-imaginal stages (L1 to L4 instars) were collected via ladles within rice farms from the village of Tiassalé in Côte d’Ivoire in April 2010, a time of year corresponding to the transition from the dry to the rainy season. Tiassalé is located 110 km northwest of Abidjan and is surrounded by irrigated rice fields. Rice farming is practiced throughout the year with heavy insecticide use for crop protection. Due to that the farms are irrigated, breeding sites are present throughout the year and we therefore assumed that the larvae collected in the study period were representative of the population that could be found during other periods of the year. Larvae collected from multiple breeding sites were pooled together then re-distributed evenly in development trays containing tap water. Larvae were provided access to powdered TetraFin® fish food, and were reared to adults under insectary conditions of 25-28°C and 70-80% relative humidity at Centre Suisse de Recherches Scientifiques en Côte d’Ivoire (CSRS) located in Adiopodoumé. Non blood-fed adult female An. gambiae s.l.[12] at 1, 2, 3, 5 and 10 days post-emergence were tested using standard WHO susceptibility test kits [6] with insecticides belonging to the four classes used in public health, namely deltamethrin and permethrin (pyrethroids), DDT (organochlorine), malathion (organophosphate), and propoxur (carbamate). Non blood-fed female adults at 1–2 days post-emergence from the susceptible An. gambiae KISUMU strain were used as a control for each insecticide and each age group; only this age group was selected because this is the recommended age for WHO susceptibility assays [6]. Tests were performed following the standard protocol for adults at 25°C and relative humidity of 70-80%. Each complete bioassay was performed with five batches of 20–25 unfed females: four batches were exposed to impregnated filter papers and one non-exposed batch served as a control. The number of mosquitoes knocked down was recorded at 5 min intervals during the 1 hour exposure period and the mortality was determined 24 hours post-exposure. Bioassays were also performed with the susceptible reference KISUMU strain of An. gambiae maintained in the insectary at CSRS. Following the exposure, mosquitoes were supplied with 10% honey solution and kept overnight under laboratory conditions prior to noting the 24 hour mortality rates. WHO [6] criteria were used to indicate the susceptibility status. Results were compiled and analyzed using EpiInfo Version 6 [13] to test for any significant difference in mortality rates between the different age groups via Mentel-Haenszel Chi square test. The length of the exposure time at which 50% of the test population were knocked down (LT₅₀) was determined using WinDL version 2.0 computer software (CIRADCA/MABIS, Montpellier, France), based on Finney (1971) log-probit model [14]. The resistance ratio (RR₅₀) was determined relative to the KISIMU susceptible strain. This was obtained by dividing the LT₅₀ of wild strain to the LT₅₀ of the susceptible strain.

The impact of adult mosquito age on KD effect and mortality following exposure to various insecticides in WHO susceptibility tests, and also the resulting susceptibility status according to current WHO criteria [6], was assessed for a wild An. gambiae strain from Côte d’Ivoire.

Age-related variations in susceptibility to a particular insecticide were consistent for both knockdown and mortality rates. For the pyrethroids (i.e. deltamethrin and permethrin), a general increase in KD effect with mosquito age was observed, though this was non-linear for deltamethrin. A similar observation was made by Hodjati et al., [10], who showed a reduction in mean KD times of 10 day old adult An. stephensi and An. gambiae, relative to newly emerged mosquitoes. Another study has indicated that 14 day old mosquitoes were more rapidly knocked down than three day old mosquitoes, regardless of the strain or species [11]. For mortality, as also shown by others [15, 16], older mosquitoes were more susceptible to insecticides. The impact of age was particularly marked for permethrin and propoxur and to a lesser degree with deltamethrin, with younger individuals being the least susceptible. For DDT, the influence of age on the mortality could not be assessed as all the age groups exhibited high resistance. The converse situation applied to malathion, in which all the age groups assessed were almost fully susceptible. Nevertheless, susceptibility to malathion has been shown to vary by mosquito age for An. stephensi from Pakistan [17]. Despite the consistently high mortality following exposure to malathion, observed differences in RR₅₀ by mosquito age suggest that this is a sensitive indicator of reductions in susceptibility to malathion relative to a susceptible strain.

The long term and extensive use of DDT over the last few decades in Côte d’Ivoire [18] would have led to a strong selective pressure on exposed mosquito populations and could explain the high resistance to DDT observed in the study population. This observation highlights a concern that although pyrethroids are still effective, extensive and exclusive use of these compounds could lead to a situation where even the oldest individuals are no longer susceptible. This may have serious consequences for insecticide-based vector control programmes.

Cross-resistance between DDT and pyrethroids has been extensively demonstrated to be associated with kdr point mutations in the common target site of these insecticides [1, 3, 19]. In the current study, the large disparity in mortality rates after exposure to DDT versus the two pyrethroids suggests that kdr alone is likely to confer greater resistance to DDT compared to pyrethroids (at least in this population) and/ or several different mechanisms may confer resistance in the studied mosquito population. This was confirmed by the detection in this population of the kdr mutation at very high allelic frequency (94.5%) via molecular assays as well as indications of the involvement of up-regulated non-specific esterases (NSE) and mixed function oxidases (MFO) demonstrated by synergist tests via CDC bottle bioassays (Chouaibou et al., personal communication). Furthermore, insensitive acetylcholinesterase (ace-1 ^R ) due to the G119S mutation responsible for resistance to carbamates and organophosphates has also been described in this population with an allelic frequency of 50% [20]. In addition to the likely involvement of NSE, MFO, ace-1 ^R and kdr mutations, studies using other An. gambiae populations have shown that a number of other mechanisms, such as oxidative stress reduction, can be involved in insecticide resistance [21–25]. The involvement of GST’s also merits further investigation as this could contribute to the higher level of DDT resistance relative to the pyrethroid resistance observed. Co-existence of all of these mechanisms would result in a large fitness cost, which may ultimately weaken older mosquitoes and render them more vulnerable to insecticides. Thus, there might be a trade-off between energy used for defense against insecticides and the onset of senescence, however, this requires further investigations. Moreover, changes in mosquito physiology that is not specifically associated with insecticides but that occurs with senescence such as an increase in the rate of cuticle permeability or a decrease in the rate of xenobiotic excretion, could also lead to an increase in susceptibility to insecticides. Increases in insecticide susceptibility in mosquitoes with aging may therefore explain observations relating to a lack of impact of observed phenotypic resistance on malaria epidemiological outcomes in areas with high insecticide treated net coverage [26, 27], since the infective stage of the malaria parasites are more often harbored by the older Anopheles spp. that are presumably the most susceptible to insecticides. This hypothesis was recently confirmed by Jones et al. [28], who showed that as malaria vectors aged they became increasingly susceptible to pyrethroid based long-lasting insecticidal nets. Results from the current study showed that 1 to 3 days old adult females from the test population from Tiassalé, were the most resistant, thereby supporting the WHO recommendation to conduct assays on mosquitoes at 24–48 hours post-emergence [6]. The use of these younger and less susceptible mosquitoes in assays for phenotypic resistance should allow detection of resistance when the highest rate of the resistant phenotype is present.

Using the WHO recommended age for insecticide susceptibility testing will enable detection of resistance where the highest rate of the resistant phenotype is present. However, the operational impact of that resistance may be less than often assumed, as the older mosquitoes that transmit malaria may still be affected by insecticide-based vector control. The current study highlighted the need adhere to standard procedures described in the WHO susceptibility test in order to generate accurate and exploitable data to more accurately inform choice of appropriate insecticidal vector control interventions.