Absence of knockdown resistance suggests metabolic resistance in the main malaria vectors of the Mekong region

Background As insecticide resistance may jeopardize the successful malaria control programmes in the Mekong region, a large investigation was previously conducted in the Mekong countries to assess the susceptibility of the main malaria vectors against DDT and pyrethroid insecticides. It showed that the main vector, Anopheles epiroticus, was highly pyrethroid-resistant in the Mekong delta, whereas Anopheles minimus sensu lato was pyrethroid-resistant in northern Vietnam. Anopheles dirus sensu stricto showed possible resistance to type II pyrethroids in central Vietnam. Anopheles subpictus was DDT- and pyrethroid-resistant in the Mekong Delta. The present study intends to explore the resistance mechanisms involved. Methods By use of molecular assays and biochemical assays the presence of the two major insecticide resistance mechanisms, knockdown and metabolic resistance, were assessed in the main malaria vectors of the Mekong region. Results Two FRET/MCA assays and one PCR-RFLP were developed to screen a large number of Anopheles populations from the Mekong region for the presence of knockdown resistance (kdr), but no kdr mutation was observed in any of the study species. Biochemical assays suggest an esterase mediated pyrethroid detoxification in An. epiroticus and An. subpictus of the Mekong delta. The DDT resistance in An. subpictus might be conferred to a high GST activity. The pyrethroid resistance in An. minimus s.l. is possibly associated with increased detoxification by esterases and P450 monooxygenases. Conclusion As different metabolic enzyme systems might be responsible for the pyrethroid and DDT resistance in the main vectors, each species may have a different response to alternative insecticides, which might complicate the malaria vector control in the Mekong region.


Background
In the Mekong region, the malaria vector control relies on the use of insecticides for impregnation of bed nets and for indoor residual spraying and is mainly focussed on three vector species complexes: Anopheles dirus sensu stricto (s.s.) [1], Anopheles minimus sensu lato (s.l.) [2] and Anopheles sundaicus s.l. [3]. Hence, the emergence of insecticide resistance in vector species may have important implications for the effectiveness of insecticide based vector control measures. This is particularly true for the Mekong region, where vector control efforts have significantly contributed to decrease the malaria burden in recent years [4].
WHO bioassays done in the framework of a three year survey (from March 2003 until July 2005) on insecticide resistance in the Mekong region (Vietnam, Laos, Cambodia and Thailand) have shown that different levels of pyrethroid and DDT resistance occur in the main malaria vectors of the Mekong region [5]. Anopheles dirus s.s., the main vector in forested malaria foci, was permethrin susceptible throughout the region. In central Vietnam, it showed possible resistance to type II pyrethroids. In the Mekong Delta, Anopheles epiroticus was highly pyrethroidresistant. It was susceptible to DDT, except near Ho Chi Minh City where it showed possible DDT resistance. In Vietnam, pyrethroid-susceptible and tolerant An. minimus s.l. populations were found, whereas the An. minimus s.l. populations of Cambodia, Laos and Thailand were susceptible [5].
As each insecticide resistance mechanisms may have a different impact on the effectiveness of the pyrethroid-based control programmes, knowledge on the involved pyrethroid resistance mechanisms is necessary to guide the insecticide use in the vector control programmes. A resistance mechanism against pyrethroids and DDT, known as knockdown resistance (kdr), has been linked to mutations in the para-type sodium channel gene. Knockdown resistance has been described in several insect species [6]. In the African malaria vector, Anopheles gambiae s.s., two substitutions at codon 1014 (L1014F and L1014S) of domain II of the sodium channel gene have been associated with knockdown resistance [7,8]. Beside knockdown resistance, metabolic resistance mechanisms have been found in pyrethroid-and/or DDT-resistant Anopheles populations. Biochemical assays have been developed to measure levels of monooxygenases, esterases and glutathione-S-transferases (GST) in mosquitoes and elevated levels of such enzymes may enhance insecticide tolerance in Anopheles populations [9][10][11]. The present study aims to explore the involvement of knockdown and metabolic resistance mechanisms in providing DDT and pyrethroid resistance in the main malaria vectors of the Mekong region.

Mosquito samples
In the framework of a three-year survey (from March 2003 until July 2005) on insecticide resistance in the Mekong region (Vietnam, Laos, Cambodia and Thailand), adult mosquitoes were collected by different collection methods (indoor and outdoor human landing collection, collection on cattle and morning resting collections inside houses) and their susceptibility was assessed against DDT and pyrethroids by use of the WHO bioassays. The bioassay results were summarized in three resistance classes as defined by WHO [12]: (1) susceptible when mortality was 98% or higher, (2) possible resistant when mortality was between 97 and 80%, and (3) resistant when the mortality was lower than 80%. The results of the bioassays were described in Van Bortel et al [5]. The pyrethroid-and/ or DDT-resistant Anopheles populations, given in Figure 1, were further analysed on the presence of kdr mutations. Although the An. dirus s.s. populations, were DDT and pyrethroid susceptible, these populations showed high levels of variation in mean knockdown time (KDT50) (ranged from 8 till 31 minutes for DDT, and from 8 till 24 minutes for pyrethroids) and were also further analysed for the presence of kdr mutations. After the bioassays, the mosquitoes exposed to control papers were stored in liquid nitrogen for subsequent biochemical analyses to assess the involvement of metabolic insecticide resistance mechanisms in pyrethroid-and/or DDT-resistant Anopheles populations. Insecticide exposed mosquitoes were dried on silica gel.

DNA extraction
One to six legs of the dried individual mosquitoes were used for DNA extraction, applying the procedure described in Collins et al [13]. DNA was resuspended in 25 μl TE buffer (10 mM Tris-HCl pH 8; 1 mM EDTA). After the biochemical assays, the remaining mosquito homogenate of the control mosquitoes was spotted on filter papers. DNA of these mosquitoes was obtained by incubating a filter paper punch overnight at 4°C in 1 ml of 0.5% saponin in phosphate buffered saline (PBS). The punches were washed for 1 h in PBS at 4°C and transferred to a new tube containing 200 μl of a 20% chelex solution (Biorad, Hercules, USA). The samples were further extracted by the chelex method described in Verhaeghen et al [14]. A negative control was included with every set of extractions.

Molecular identification
observed in Cambodia and Vietnam [15,16]. Anopheles minimus s.l. is a complex of at least two isomorphic species, which occur in sympatry in this area [17]. Morphologically identified An. minimus s.l. mosquitoes were analysed by using a slightly adapted version of the PCR-RFLP developed by Van Bortel et al [17]. The restriction enzyme BsiZi was replaced by its isoschizomer Cfr13I. The reaction mixture contained 17 μl sterile water, 2.5 μl Tan-goTM buffer (provided by manufacturer), 0.5 μl of Cfr13I (10 units/μl) (Fermentas, St-Leon Rot, Germany) and 5 μl of the PCR product. The mixture was incubated for 2 h at 37°C. After incubation, the specimens were electrophoresed on a 3% mixed agarose gel (1.5% agarose and 1.5% small fragment agarose) and visualized under UV light after ethidium bromide staining.

PCR-RFLP for the detection of kdr mutations in An. dirus s.s
The DIIS6 region of the para-type sodium channel gene was amplified by an adapted version of the protocol by Martinez-Torres et al [7]. Amplification was performed in a 50 μl reaction containing 1 μl of template DNA, 1 × Qiagen PCR buffer, 1 mM MgCl 2 , 200 μM of each dNTP, 100 nM of the primers Agd1 and Agd2 and 1 unit Taq DNA polymerase (Taq PCR core kit, Qiagen, Hilden, Germany). The cycling conditions were as follows: initial denaturation at 94°C for 3 minutes, 40 cycles of 1 minute denaturation at 94°C, 30 seconds annealing at 47°C and 30 seconds extension at 72°C followed by a final extension of 10 minutes at 72°C. Amplification products were checked on a 2% agarose gel and visualized under UV light after ethidium bromide staining. Direct PCR sequencing was performed by the VIB genetic service facility (University of Antwerp, Belgium).
Based on the DIIS6 sequence, a restriction enzyme, FspBI was found suitable to screen for the presence of a kdr mutation in An. dirus s.s. The restriction mixture contained 15.5 μl sterile distilled water, 2.5 μl buffer Tan-goTM, 1 μl enzyme FspBI (5'-C^TAG-3', 10 units/μl) (Fermentas, St-Leon Rot, Germany) and 6 μl of the PCR amplified template. The mixture was incubated overnight at 37°C. After incubation, the samples were electrophoresed on a 3% small fragment agarose gel and visualized. In both assays, a primary PCR was performed with the primers Agd1 and Agd2 (FRET/MCA I) or Agd1Mi and Agd2H (FRET/MCA II) to amplify the DIIS6 region of the para-type sodium channel gene [7,18] (Table 1). The reaction and cycling conditions given for the amplification of the DIIS6 region of An. dirus s.s. were used. The DIIS6 fragment was subsequently used in an amplification reaction (20 μl), which contained 1 × iQ supermix (Biorad, Hercules, USA), 300 nM of a forward ROX labelled primer, 60 nM of a reverse primer and 1 μl of a 10-fold dilution of the primary PCR product. The PCR was performed on an iCycler with a 490/20X FAM excitation and a 620/30M ROX emission filter (Biorad, Hercules, USA) following the cycling conditions as given in Table 1. After amplification, a FAM labelled probe was added in a final concentration of 200 nM, and a melt curve was performed. Cooling to 48°C will allow the FAM labelled probe to anneal adjacent to the ROX fluorophore of the PCR product. Subsequently, the temperature is slowly increased, while the ROX fluorescence resulting from FRET, is continually monitored. When the melting temperature of the probeamplicon hybrid is reached, the ROX fluorescence will decrease as FRET can no longer occur between the FAM label of the probe and the ROX label of the PCR product.
The change in the ROX fluorescence appears as a positive peak on the plot of the first negative derivative of the fluorescence versus temperature function. The iCyclerTM iQ Optical system software version 3.1 was used for data analysis (Biorad, Hercules, USA) and all experiments were performed in duplicate to verify reproducibility.
Control mutant and wild type plasmids were used on each plate as positive controls. The wild type plasmids used in FRET/MCA I and FRET/MCA II originated respectively from a sequenced wild type An. epiroticus and An. minimus s.s. specimen of Vietnam. The plasmids were constructed by ligation and transformation of the primary PCR product (primers Agd1-Agd2 or Agd1Mi-Agd2H), by use of the Original TA cloning kit according to the manufacturer's instructions (Invitrogen, Carlsbad, California). For each FRET/MCA, two mutant plasmids were constructed by gene synthesis (Genscript, Piscataway, New Jersey). For the FRET/MCA I, a 153 bp fragment similar to the DIIS6 region of the para-type sodium channel gene of An. epiroticus (fragment from primer sundF-ROX to Agd2) was made. The triplet TTA, encoding the L1014 in the wild type sequence, was replaced by TCA (L1014S) or TTT (L1014F) to represent the two kdr mutations in An. gambiae s.s. The mutant plasmids used in FRET/MCA II contained a 132 bp fragment of the DIIS6 region of An. minimus s.s. (fragment from primer MinF-ROX to Fun-minR). Similarly, the triplet TTA was replaced by TCA (L1014S) or TTT (L1014F). All mutant plasmids were cloned in a pUC57 vector and transformed in TOP10 cells (Invitrogen, Carlsbad, California). Five millilitres of each clone was purified on a column (QIAprep Spin Miniprep kit, Qiagen, Hilden, Germany) and DNA was resuspended in 50 μl water. One microlitre of a 100-fold dilution of the plasmids was used in the secondary PCR of the FRET/MCA assay.

Biochemical assays
In order to obtain a complete resistance profile for the main malaria vectors, biochemical assays were performed on resistant An. minimus s.l., An. epiroticus and An. subpictus populations of Vietnam. The main vector An. dirus s.l. could not be tested in the biochemical assays because the collection sites in Central Vietnam could not be accessed with a liquid nitrogen container. Only mosquitoes exposed to control papers during the WHO bioassays were used for these assays.
To ensure that the presence of blood could not interfere with the biochemical assays, the abdomen of the individual mosquitoes was removed. The remaining head-thorax portion was homogenized in 200 μl distilled water and centrifuged at 13,000 g for 2 min. The monooxygenase, esterase (para-nitrophenyl acetate as substrate), GST (1-chloro-2,4-dinitrobenzene as substrate) and protein assay were carried out, in duplicate, on the supernatant as described by Penilla et al [19].
For this study, no fully susceptible reference colony strains for different test species were available. Therefore, an An. minimus s.s. field population of VHBA (Hoa Binh Province, northern Vietnam; figure 1), fully susceptible in the WHO bioassays, was taken as reference. The two-sample Kolmogorov-Smirnov Z test (SPSS 14) was used to compare the results of the biochemical assays with the reference An. minimus s.s. population of VHBA.

Molecular identification
Anopheles minimus s.l. is a complex of at least two isomorphic species, which occur in sympatry in Vietnam [17]. All An. minimus s.l specimens used to study target-site insensitivity and metabolic resistance, were molecular identified (Table 2). In Cambodia, most of the specimens identified as An. minimus s.l. (60 of 62; 96.8%) proved to be An. minimus s.s. In Vietnam, 2,425 of the 2,480 (97.8%) morphologically identified An. minimus s.l. specimens belong to the species complex. Among the mosqui-  Figure 2). In total, 927 An. dirus s.s. specimens (Table 3) originated from nine Cambodian and 11 Vietnamese collection sites were tested with this PCR-RFLP. Of these 927 specimens, 854 showed a complete digestion corresponding to the wild type triplet (CTA), whereas 73 specimens showed an incomplete digestion with the restriction enzyme FspBI. The partial restriction occurred among control, alive and death insecticide exposed mosquitoes. The DIIS6 region of the para-type sodium channel gene was sequenced for all specimens with a partial digestion, but no kdr mutation was observed among these 73 specimens.  (Table 3). One An. harrisoni mosquito used as control in the WHO bioassay and orig-

Biochemical assays
Non-specific esterases, monooxygenase and GST assays were performed on different Anopheles populations of Vietnam (Table 4). A significant increase in esterase activity was measured in the An. minimus s.s. populations from northern Vietnam (VLSA, VNAB, VQNA, VQNB). All these populations were possibly pyrethroid-resistant in the WHO bioassays (  (Table 5).
All An. epiroticus populations of the Mekong Delta, except the population of VHCB, had an increased esterase activity (1.5 to 4 fold higher) compared to the An. minimus s.s. population of VHBA (Table 4). In the bioassays, these populations were characterized by pyrethroid resistance (Table 5). In An. subpictus, the esterase activity was significantly higher (2 times higher) in the VKGB population. The esterase activity was not significantly increased in the possible permethrin-resistant (97% mortality) VLAA An. subpictus population (Table 5).
Almost no differences in GST activity were observed between the different Anopheles populations. Only, the possible DDT-resistant An. subpictus populations from VKGB and VLAA were characterized by a significant higher GST activity (2.5 fold higher) (Tables 4 and 5).
In general, the estimates of the monooxygenase levels in An. minimus s.l., An. epiroticus and An. subpictus populations were significantly lower than those obtained for An. minimus s.s. of VHBA. A significant increased monooxygenase level (2 fold) was only detected in the An. minimus s.s. population from VNAB (Table 4). In the WHO bioassay, this population was only tested with alpha-cypermethrin and had a 24 h mortality of 90% (Table 5).

Discussion
In the 1960s, malaria control in Vietnam was primarily based on the use of DDT for residual house spraying.  [5].
Insects may survive the toxic effect of insecticides by different physiological mechanisms including target-site insensitivity and elevated detoxifying enzyme production. Knockdown resistance, caused by a mutation at codon 1014 of the para-type sodium channel gene, is described in several Anopheles species including An. gambiae s.s.,  Because no kdr mutation was observed in the main vectors of the Mekong region, metabolic resistance mechanisms were studied by use of biochemical assays. Biochemical assays point towards metabolic resistance mechanisms, by indicating whether there is a general P450 monooxygenase, glutathione-S-transferase or esterase response. High GST activities were found in An. subpictus. Glutathione-Stransferases catalyse the dehydrochlorination of DDT to DDE and have been reported to play a significant role as DDT resistance mechanism in many insects including An. gambiae and Anopheles crascens of Thailand [30,31]. In addition, GSTs can mediate insecticide resistance by conjugation of glutathione to the insecticide or its primary toxic metabolic product. GSTs were found to play a minor role as pyrethroid resistance mechanism in An. funestus [11]. Because An. subpictus was possible DDT-resistant, it

PCR-RFLP for the detection of kdr mutations in An. dirus s.s
Sequences of the DIIS6 region of the para-type sodium channel gene obtained for An. harrisoni mosquitoes with respectively a Tm of 63.9°C (wild type, WT) and 62.1°C is likely that the high GST activity detected in this study confers DDT resistance.   [9,32] reported monooxygenase and esterase based resistance mechanisms alone or in combination, in permethrin-resistant Anopheles albimanus from Guatemala. In an An. minimus s.s. colony of Thailand, deltamethrin-resistant was primarily associated with increased detoxification by P450 monooxygenases [33].
Many biological processes may affect gene expression making the interpretation of biochemical assays not straightforward. It is always possible that the measured insecticide detoxification levels represent other processes than insecticide resistance. Therefore biochemical assays should only be used to formulate hypothesis on the possible metabolic resistance mechanisms involved. Synergist studies should be further performed to confirm the role of the detected metabolic enzyme levels in the detoxification of pyrethroids and/or DDT. In addition, as our biochemical assays were performed on field populations of Anopheles species; differences in species, age and blood feeding status might induce an extra variability in the met- abolic enzyme data [34,35]. To account for differences in enzyme levels due to age, WHO recommends using nonblood fed, adult mosquitoes of the same age, which can only be obtained by collecting larvae and rearing to adults. However, collecting an appropriate number of larvae of the major Anopheles species in the Mekong region is problematic due to the scattered nature of their breeding sites. In the current study, biochemical assays were performed on field collected adults used as control in the WHO tube bioassay. To ensure that the presence of blood could not interfere in the biochemical assays, the abdomen of the mosquitoes was removed. Despite the limitations of these biochemical assays, the presence of a systematic increase of a certain enzyme system in a large number of DDT and/or pyrethroid-resistant Anopheles populations of a certain species (e.g. systematic increase of esterases in pyrethroid-resistant An. epiroticus) might indicate the involvement of this metabolic enzyme system as insecticide resistance mechanism.
Another drawback of the biochemical assays was the absence of fully susceptible reference strains for each species tested. population showed reduced mortality against the type II pyrethroids (alpha-cypermethrin 96% and lambda-cyhalothrin 94% mortality) and the non-ester pyrethroid etofenprox (95% mortality). Taking this into account, it is likely that the pyrethroid resistance in An. minimus s.l. could be conferred to an increased detoxification by both P450 monooxygenases and esterases, whereas in An. epiroticus and An. subpictus the pyrethroid resistance could be conferred to an esterase mediated detoxification. However, additional WHO bioassays performed on 1-2 days old An. epiroticus mosquitoes of VBLA and VBLB (Mekong Delta) revealed a low mortality against the non-ester pyrethroid etofenprox [5]. This means that it is likely that beside an esterase mediated detoxification also other pyrethroid resistance mechanisms are involved in An. epiroticus of the Mekong Delta.
Knowledge on the different resistance mechanism is necessary to guide the insecticide use in the vector control programmes as each insecticide resistance mechanism may have a different impact on the effectiveness of the pyrethroid based control programmes. Studies in West Africa have shown that ITNs remain effective against kdrresistant An. gambiae s.s. [37][38][39]. Chandre et al [40] demonstrated that large proportions of kdr homozygous resistant (L1014F/L1014F) females were killed by prolonged contact with pyrethroids due to diminished sensitivity to the excito-repellent effect of the insecticide. However, recent studies have shown a reduced efficacy of ITNs when the L1014F kdr allelic frequency is high [41,42]. In South Africa, metabolic pyrethroid resistance in An. funestus required a switch back from pyrethroid insecticides to DDT for house spraying to restore the malaria vector control [43]. Elevated levels of metabolic enzymes in An. gambiae s.s. of Cameroon did not influence the personal protection afforded by ITNs, but did reduce the vector mortality which could in turn limit the mass effect [36]. As in the Mekong region, metabolic resistance mechanisms are involved in the main malaria vectors, the operational impact should be further studied as it may have implications for the sustained efficacy of the ITN based control programmes. As different metabolic enzyme systems might be responsible for the pyrethroid and DDT resistance in the main vectors, each species may have a different response to alternative insecticides. Furthermore, assessing the susceptibility status of the main malaria vectors to organophosphates and carbamates will provide information on the usefulness of these insecticides in the vector control programmes.

Conclusion
A three-year survey on insecticide resistance in the main malaria vectors of the Mekong region showed that the main vector An. epiroticus was pyrethroid-resistant in the Mekong delta. Anopheles minimus sensu lato was pyrethroid-resistant in northern Vietnam, whereas An. dirus sensu stricto was possible resistance to type II pyrethroids in central Vietnam. The secondary vector An. subpictus was both DDT and pyrethroid-resistant in the Mekong Delta.
As the operational implications of insecticide resistance will largely depend on the resistance mechanisms involved, the role of both knockdown and metabolic resistance was assessed in the main vectors of the Mekong region. Two FRET/MCA assays and one PCR-RFLP were developed to detect kdr mutations, but no kdr mutation was observed in any of these vectors albeit many samples of different populations from a wide geographical area were screened. Biochemical assays suggested the involvement of metabolic resistance mechanisms. As different metabolic enzyme systems might be responsible for the pyrethroid and DDT resistance in the main vectors, each species may have a different response to alternative insecticides, which might complicate the malaria vector control in the Mekong region.