Selection for and biochemical characterization of DDT resistance in laboratory strains of Anopheles arabiensis

Resistance to conventional insecticides still constitutes a major obstacle to control of malaria vectors. Xenobiotic pollutants encountered by aquatic stages of natural populations of malarial vector species in agricultural and domestic environment are often selected due to resistance to various insecticides. The Laboratory Matatuine (MAT) and Kayamba (KGB) strains of Anopheles arabiensis were subjected to controlled dosages of DDT for over twenty generations. WHO insecticide susceptibility protocols was used to monitor changes in mortality between generations. The selected lines of both strains developed resistance to DDT and cross resistance to permethrin. Polymerase Chain Reaction (PCR) detection of knock-down resistance (kdr) gene and sequencing revealed absence of L1014F mutations. Biochemical analysis of detoxification enzymes showed significant Glutathione S transferase (GST) activity in the selected lines [MAT: 0.236 (P>0.001) and KGB: 0.221 (P>0.014)], thus suggesting the presence of GSTbased resistance mechanism.


INTRODUCTION
Anopheles arabiensis is the second most efficient malaria vector species of the An. gambiae complex and it occurs in sympatry with An. gambiae sensu stricto in most areas. The two species form the most efficient malaria vectorial system in Africa (Powell et al., 1999;Coetzee et al., 2013). It often breeds in pesticide contaminated rice irrigation ecosystems found at malaria endemic areas in East and West Africa and adapt faster to man-made ecological habitats in urban cities (Coluzzi et al., 1979;Ijumba and Lindsay, 2001;Kamau and Vulule, 2006). It contributes considerably to malaria transmission in South Africa, Sudan, Nigeria and is the major malaria vector in Tanzania (Onyabe and Conn, 2001;Nardini et al., 2013;Matawo et al., 2014;Tarig et al., 2018). Current global strategy for control of malaria is based on chemoprophylaxis, treatment of diagnosed infected *Corresponding author. Email: yayoabdulsalami@gmail.com.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License persons using effective anti-plasmodial drugs and insecticidal based methods to control the malaria vectors. The World Health Organisation (WHO) Pesticides Evaluation Scheme (WHOPES) approves eleven insecticides including permethrin and DDT to be used in public health, particularly malaria control programmes (WHO, 1998). Development of resistance to insecticides by malaria vector species still remains the major obstacle to malaria control globally (Karunaratne et al., 2018;Ranson et al., 2011). Resistance is the ability of an insect to withstand toxic effects of an insecticide by means of natural selection and mutations and is a heritable genetic trait passed down through generations (Davidson, 1957;Corvel and Nguessan, 2013). Increasing selection pressure on malaria vector populations caused by xenobiotic pollutants presence in agricultural and domestic environments necessitates selection for resistance to various insecticides (Nkya et al., 2012;Matawo et al., 2015). Resistance to DDT in field populations of An. arabiensis began to appear in South Africa (Hargreaves et al., 2003), Nigeria (Kristian et al., 2003), and subsequently, in other countries. Resistance to deltamethrin, permethrin and DDT have been reported in Ethiopia (Balkew et al., 2010;Yewhalaw et al., 2011) and Sudan (Abdullah et al., 2008).
The resistance to DDT in mosquitoes is generally associated with one of two mechanisms; increasing DDT dehydrochlorination catalysed by GSTs or decreased target site sensitivity (Hemingway and Ranson, 2000;Ranson et al., 2011;Karunaratne et al., 2018). Some of the previous studies on insecticide resistance in An. arabiensis have used populations selected in the laboratory for many generations (Hemingway, 1983;Matambo et al., 2007). The use of laboratory strains (MAT and KGB) in comparison with field strains to study resistance mechanisms advantageously exclude factors such as effect of temperature, larval diet and exposure to agricultural pesticides that can confound diagnosis (Nardini et al., 2013). The present study was aimed at selecting MAT and KGB laboratory reared strains of An. arabiensis for DDT resistance. The specific objectives were to (i) test for succeptibility of parental lines of MAT and KGB strains to DDT (ii) select subsequent generations for resistance to DDT (iii) test the selected lines for cross resistance to permethrin, and (iv) compare levels of activity of the detoxifying enzymes GSTs, esterases and monooxygenases in mosquitoes sampled from parental and selected lines of both strains.

Establishment of An. arabiensis colony
Eggs of An. arabiensis MAT strains were collected from a field site at Matatuine located 10 km from Maputo in Mozambique.

Mosquito rearing
The colonies of MAT and KGB An. arabiensis mosquito strains were maintained in the insectaries at the Liverpool School of Tropical Medicine. The mosquitoes were reared at temperature range of 27 to 28°C , 80 to 85% relative humidity with 12-h day/night light regime and 45-min dusk/dawn cycles. The duration of development from eggs to emerging adults ranged from 7-12 days amongst both strains. All mosquito larvae were fed on Tetramin fish food flakes, using ground-up flakes for the first instar. Extreme care was taken to avoid contamination between the different lines and strains of An. arabiensis. All larval trays were cleaned with hot water after each rearing cycle, when all the pupae had emerged into adults. Pipettes, egg pots and larval trays were colour-coded for each line of An. arabiensis strains. Adult mosquitoes were constantly provided with cotton wool soaked in saturated 10% sugar solution formed by using tap water. Females of both strains were given guinea pig blood twice a week (Hunt et al., 2005). Samples of the adult mosquitoes were identified to species using the polymerase chain reaction method described by Scott et al. (1993).

Susceptibility to WHO bioassays
Adult mosquitoes from both colonies of the parental lines of An. arabiensis MAT and KGB strains were tested for susceptibility to DDT. Bioassays were performed according to WHO protocols using standard WHO susceptibility test kits and 4% DDT impregnated papers (WHO, 1998). Survivors from each test were placed in a separate cage and used to establish subsequent generations.

Selection for resistance to DDT
Mosquitoes from An. arabiensis MAT and KGB strains, which survived previous exposures to DDT, were reared and their progeny subjected to selection using 4% DDT. Adult mosquitoes from An. arabiensis MAT strain were maintained under selection pressure with 4% DDT continuously for eight months and selection was interrupted for five months due to a crush in the colony, but thereafter the selection pressure was continued further for eight months. Mosquitoes from An. arabiensis KGB strain were similarly selected for resistance to DDT for a period of twenty three months without interruption.

Testing for cross resistance to permethrin
Batches of mosquitoes from An. arabiensis MAT and KGB selected lines were tested for resistance to DDT and cross-resistance to permethrin. Samples of adult mosquitoes from F10 and F20 selected generations of An. arabiensis MAT and An. arabiensis KGB respectively were exposed to 4% DDT and 0.75% permethrin for one hour. Knocked-down mosquitoes were recorded at intervals of ten minutes. The KDT50 and KDT90 knock-down times were calculated by probit analysis. The data was entered into Minitab 14 and LDP line software programmes for the analysis. test for the presence of the typical kdr mutations in individual mosquitoes sampled from parental and selected lines of both strains.

Biochemical assays
Unexposed mosquito samples taken from the parent stock and F15 selected generation of An. arabiensis MAT and KGB strains were kept at -80°C for biochemical analysis. Biochemical assays were performed according to the standardized procedures described in the manual by Hemingway (1998). Batches of 22 one-day old, frozen mosquitoes were individually homogenized in 200 µl of distilled water in 1.5 ml Eppendorf tubes. The crude homogenate was spun at maximum speed of. 10,000 rev minfor two minutes in a microfuge. After centrifugation, the supernatant from each Eppendorf tube was then transferred to a well of a micro titre. Esterases, monooxygenases and GST assays were carried out in line with WHO (1998).

Protein assay
Protein assays were conducted according to the method of Bradford (1976). Microfuged homogenate (10 µl) from each mosquito was added to 300 µl Bio-Rad Protein assay reagent (diluted 5 times from stock), incubated for 5 min and end point absorbance measured at 570 nM. Protein concentration was determined by converting the absorbance into concentration based on a bovine serum albumin standard curve.

WHO susceptibility assays
The susceptibility levels to DDT of adult mosquitoes of the original parental populations of the An. arabiensis MAT and KGB strains were determined. A total of 338 adult mosquitoes aged two to three day old from the parental An. arabiensis MAT strains were exposed to 4% DDT for different time periods (Table 1). The 87% mortality after exposure to DDT for one hour indicates the presence of low level of resistant genotypes in the MAT parental colony.
In the KGB strain, 389 adult mosquitoes in batches of 20 to 25 were exposed at the four different time points and mortality was recorded 24 h post exposure as shown ( Table 2).
The 81.6% mortality after exposure to DDT for 1 h suggests higher level of resistant genotypes in KGB than in the MAT colony.

Selection of resistant genotypes
The mortality decreased from 73.5% in the F3 to 51.4% in the F6. Due to rearing problems, selective pressure was not applied in generations from F7 to F13. The mortality rose back to 61.8% in F14 but decreased gradually to 48.3% in generation 16. Selection at 45 min exposure period raised the mortality to 69.4% but subsequently decreased to 53% in F20 (Figure 1).
The KGB colony did show a similar pattern of response to DDT. However, the selection pressure was gradually increased from 30 min in F1 generation to 60 min over 20 generations. Over these selected generations, the mortality decreased from 56.4% in F6 to 28.4% in F14 (Figure 2).  (Table 3). In the KGB strain, the LT50 values were 33.5 min and 50.8 min and the corresponding slopes of the regression lines were 3.26 and 1.7 in the parental and selected populations respectively (Table 3). The change in slope of regression lines between KGB selected and parental indicates increased resistance in the selected population.

Cross-resistance to permethrin
The populations of An. arabiensis DDT selected in both MAT and KGB parental strain were also tested against the diagnostic dosage of 0.75% permethrin to check for cross resistance or increased tolerance (Figure 3). Significantly more mosquitoes were knocked down by permethrin at 30 min, 40 min, (P < 0.001) and at 50 min (P = 0.072) in the parental line than in the DDT selected   population but both showed > 97% mortality at 24 h after exposure (Figure 3).

Glutathione Stransferase activity
The geometric mean GST activity was significantly higher (p < 0.001, 0.014) in populations under selection pressure than in the parental (unselected) populations in An. arabiensis MAT and the ninth generation KGB strains respectively (Table 4).

Esterase activity
The geometric mean values of α-esterase and β-esterase activities were significantly higher (p < 0.001) in the selected than in the parental lines of the MAT strains. In the KGB strain, the esterase activities were higher in the KGBS9 compared to KGBP6 (p < 0.001). However, the activities in KGBSG15 were lower than that in KGPG6 (P < 0.004) ( Table 5).

Monooxygenases activity
The geometric means of monooxygenases in the selected and parental lines in the An. arabiensis MAT and KGB strains were not significantly different (p = 0.148) ( Table 6).

DISCUSSION
The results of this study indicate that a low level of physiological resistance to DDT in An. arabiensis is developed under selection pressure in the laboratory. The LT50 and LT90 values of DDT increased significantly over 15 generations of selection pressure in both MAT and KGB An. arabiensis strains. The LT50 and LT90 recorded for the twentieth generation of KGB selected line are similar to the values reported for the DDT resistant field populations of An. gambiae sl (Tarig et al., 2018). In the selection of An. arabienisis MAT strain, high variation characterised the mortality values during the first four generations. This might have been due to the error made initially in the selection process by putting the survivors from the first two selection experiments back in the same cage with the parental colony. Mating between the two populations might have resulted in dilution of the selected resistant genotypes; therefore making the population more susceptible to DDT as was earlier hypothesized (Prasittisuk and Curtis, 1982). Alternatively, it has been suggested that high variation in mortalities is perhaps typical of populations in early stages of selective pressure (Theeraphap et al., 2002). The general patterns of the selection for DDT resistance are similar in both strains. This observation is typical of most laboratory regimes which tend to select within existing phenotypic distributions often at 80 -90% mortality in order to provide survivors for the next generation (Martins et al., 2012;Roush and McKenzie, 1987). The dosage for selection was closely controlled between 30 to 45 min and 30 -60 min for the MAT and KGB respectively to permit discrimination among similar genotypes within the physiological distribution of phenotypes (Roush and McKenzie, 1987). However, in previous similar studies, 3 laboratory colonies G1, SENN and MBN of An. arabiensis have been selected for resistance to DDT at higher doses and adults were reported to have survived exposure to DDT for 8 h (Hemingway, 1981;Matambo et al., 2007). This suggests that the KGB and the MAT strains were at comparatively low level of DDT resistance. Theoretically, a susceptible colony comprising of totally susceptible individuals will produce the highest slope for a regression line of doseresponse data. With selective pressure from the exposures to insecticides, a population will become heterozygous for resistant genotypes and as the frequency of resistant genotypes increases, the slope of the regression line will shift to the right (Brown and Brogdon, 1976). There was a shift to the right in regression lines from dose response data for the populations under DDT selection in both the MAT and KGB strains and the slopes of regression lines based on the data from these experiments continuously declined over time in the two strains. This suggests that the resistance to DDT in the selected populations was not due to vigour tolerance but reflects true physiological resistance (Oliver and Brooke, 2014;Brown and Brogdon, 1976).
Evidence for cross resistance to permethrin was observed in the DDTselected colony of An. arabiensis MAT strain Various previous studies have shown some evidence for resistance to permethrin in colonies of An. arabiensis strains selected for DDT resistance (Nardini et al., 2013;Matambo et al., 2007). In addition, cross resistance between pyrethroids and DDT has been reported in natural populations of An. arabiensis (Abdulla et al., 2008), An. gambiae (Matawo et al., 2015), and An. funestus (Tchouakui et al., 2019). The similar mode of action of DDT and pyrethroids can result in crossresistance if the mechanism is due to kdr mutations in the sodium ion channel target sites (Martinezz-Torre et al., 1998;Tene et al., 2013). The West African L1014F mutation has previously been reported in An. arabiensis from Sudan (Abdullah et al., 2008) and SENN-DDT resistant laboratory strain (Matambo et al., 2007), although in both the correlation between the L1014F genotype and DDT resistant genotype, it was not clear.
In this study, analysis of sequence data for the gene revealed absence of kdr mutations in both MAT and KGB colonies. Similarly, the kdr mutation has not been observed in the M form of An. gambiae ss and An. arabiensis despite high levels of resistance to pyrethroid and DDT although it was found in the S form (Diabete et al., 2002). The kdr mutations have been documented in DDT resistant field populations of An. coluzzi, An. gambiae ss and An. arabiensis (Cisse et al., 2015). The combined effects of detoxifying enzymes and potential mutations have been associated with resistance to multiple insecticides in An. funestus and An. arabiensis (Menze et al., 2016;Matawo et al., 2014) Nevertheless, the absence of the kdr mutations in the colonies of An. arabiensis studied here is not conclusive considering the low number of samples used in the assay.
The results of biochemical analysis have shown that more individuals with high GST activity are present in the selected than in the parental lines of both the MAT and KGB strains. The order of magnitude of change in GST activity observed in the selected populations of MAT and KGB strains is consistent with our recent report on involvement of the epsilon gste2 gene in DDT resistance in the colonies (Yayo et al., 2018(Yayo et al., , 2019. Previous studies have severally associated resistance to DDT with increased levels of GST activities in several species of mosquitoes including An. subpictus (Hemingway et al., 1991) and An. gambiae (Karunaratne et al., 2018). Elevated esterase activity was also detected in the MAT populations under DDT selection compared to the unselected population.
The results in absolute unit for the alpha and beta esterase were similar to those of Hargreaves et al. (2003). Casimiro et al. (2006) found lower average esterase activities with the two substrates in DDT susceptible populations of An. arabiensis from Mozambique. The monooxygenase activity was low in the selected populations in KBG strain, suggesting that the p450 enzyme system may not be involved in DDT resistance in this strain. However, the monooxygenase activity was significantly higher in some individuals from selected line compared to the parental line in MAT strain, but the small sample size was low to derive a conclusion.

Conclusion
Two laboratory strains of An. arabiensis exposed to controlled doses of DDT have developed resistance to DDT and cross resistance to permethrin. Analyses of the detoxification enzymes have shown significantly high GST activity in the selected line of both strains. Preliminary investigations revealed absence of kdr suggesting the possible role GST-based DDT resistance mechanism in the colony.