Allelic variants of cytochrome P450 monooxygenases: Constitutive and insecticide-mediated expression in a Malaysian strain of the dengue vector, Aedes aegypti (Diptera: Culicidae)

Cytochrome P450s (P450s) involved in insecticide resistance reduce the effi cacy of insecticide-based vector control by rendering vector control ineffective. They are recorded in many species of vectors and have various constitutive and insecticide induction profi les. In this study, the isolation and prediction of the structure of a P450 from a strain of Aedes aegypti originating from Malaysia is reported. Quantitative mRNA expression of this gene and a previously reported P450, CYP4H28v2, in the developmental stages of the mosquito after exposure to sub-lethal concentrations of insecticides is also reported. The isolated P450, CYP4H31v2, is an allelic variant of CYP4H31 and contains several conserved motifs of P450s. The secondary structure of the protein is mostly made up of alpha helices and random coils. The tertiary structure was generated using homology modeling and was of good quality based on structure validation using protein structure assessment tools. CYP4H28v2 and CYP4H31v2 were differentially expressed in the developmental stages of the vector, with a signifi cantly increased expression in adult males. The genes were signifi cantly over-expressed in larvae exposed to deltamethrin and permethrin for 6 h. In the DDT-treated larvae, only CYP4H31v2 was signifi cantly over-expressed after a 6 h exposure. Under-expression of the genes was predominant in larvae treated with the organophosphates malathion and temephos. Though the functions of these P450s are unknown, their response to induction by exposure to insecticides indicates the likely involvement of these genes in insecticide tolerance. * Corresponding author; e-mails: swavicor@usm.my / wintuma@live.com INTRODUCTION Viruses transmitted by Aedes mosquitoes are among the most devastating arboviruses affecting human health, causing morbidity and mortality in many countries. These mosquitoes are responsible for diseases like chikungunya and dengue fever. Though an integrated vector management approach is practiced in many parts of the world, a dependency on conventional insecticides is the most common means of controlling mosquito-borne diseases (van den Berg et al., 2012). However, resistance to these chemicals is reported (Vontas et al., 2012; Ishak et al., 2015). The extensive development of insecticide resistance has necessitated the inclusion of insecticide resistance management as an integral part of mosquito control. This is aimed at monitoring the resistance status of mosquito populations to detect incipient resistance and also manage resistance in populations where it is present. The fi rst report of dengue in Malaysia was in Penang in the 1900s (Skae, 1902), but as reported in Mudin (2014), the fi rst recorded outbreak in Malaysia was in 1962 in Penang, resulting in 41 cases and 5 deaths. Since then several Eur. J. Entomol. 113: 507–515, 2016 doi: 10.14411/eje.2016.067


INTRODUCTION
Viruses transmitted by Aedes mosquitoes are among the most devastating arboviruses affecting human health, causing morbidity and mortality in many countries.These mosquitoes are responsible for diseases like chikungunya and dengue fever.Though an integrated vector management approach is practiced in many parts of the world, a dependency on conventional insecticides is the most common means of controlling mosquito-borne diseases (van den Berg et al., 2012).However, resistance to these chemicals is reported (Vontas et al., 2012;Ishak et al., 2015).The extensive development of insecticide resistance has necessitated the inclusion of insecticide resistance management as an integral part of mosquito control.This is aimed at monitoring the resistance status of mosquito populations to detect incipient resistance and also manage resistance in populations where it is present.
The fi rst report of dengue in Malaysia was in Penang in the 1900s (Skae, 1902), but as reported in Mudin (2014), the fi rst recorded outbreak in Malaysia was in 1962 in Penang, resulting in 41 cases and 5 deaths.Since then several of 30 min.The mixture was subjected to a heat shock at 42°C for 2 min and then placed on ice for 10 min.It was incubated (37°C and 180 rpm) in LB broth (600 μl) for 2 h and then centrifuged at 12,000 g for 1 min.The pellet was resuspended in LB broth (50 μl), plated on LB agar supplemented with 100 μg/ml ampicillin, 40 μg/ml X-gal and 0.5 mM IPTG and incubated overnight at 37°C.Plasmid isolation was done using Wizard ® Plus SV Minipreps DNA Purifi cation System (Promega ® ) and sequenced using the universal primers pUC/M13 forward, pUC/M13 reverse, SP6 and T7 (First Base Laboratories Sdn.Bhd.).

Rapid Amplifi cation of cDNA Ends (RACE)
Gene-specifi c primers (Table 1) for 3'-and 5'-RACE were designed based on the partial nucleotide sequences of the cloned fragments.3'-RACE PCR (Table 1) solution contained 0.2 μM of each primer [SGP1 and Oligo (dT) 25 V], 0.4 μg cDNA, 1 × On-eTaq Hot Start Master Mix with Standard Buffer and deionized water in a reaction volume of 25 μl.The products were used in a nested PCR with similar constituents and conditions except for substituting SGP2 for SGP1.The products were analyzed on an agarose gel, purifi ed, cloned and sequenced as described above.5'-RACE ready cDNA and PCR were performed using SMART-er™ RACE cDNA Amplifi cation Kit (Clontech ® ) and Advantage ® GC 2 PCR Kit (Clontech ® ).The PCR (Table 1) solution was prepared by adding 1 × Advantage GC 2 PCR Buffer, 0.5 M GC-Melt, 1 × dNTP Mix, 1 × Universal Primer A Mix (UPM) and 0.2 μM primer specifi c gene (SGPR), 1.25 μg 5'-RACE cDNA, 1 × Advantage GC 2 Polymerase Mix and PCR-grade deionized water.The products were purifi ed, cloned and sequenced.

Isolation of full-length gene
cDNA was synthesized from RNA as described earlier and used with primers designed using the RACE PCR products in a PCR (Table 1) using Advantage ® GC 2 PCR Kit (Clontech ® ) to amplify the full-length of the gene.The products were purifi ed, cloned and sequenced.

Mosquito samples for constitutive expression
Larval (1 st -4 th instar) and adult stages were used to determine the basal expression of the P450s in these life stages.The adult mosquitoes were 3 days old and non-blood fed.

Insecticide treatment
Insecticide solutions were obtained from the VCRU, a WHO Collaborating Center for the supply of insecticide resistance test kits.Concentrations and induction periods that induced < 5% mortality in the 4 th instar larvae were used.The larvae were treated with deltamethrin (0.01 ppm), dichlorodiphenyltrichloroethane (DDT) (4 ppm), malathion (0.0003125 ppm), temephos (0.0000625 ppm) and permethrin (0.05 ppm) by placing larvae (25) in a paper cup containing 100 ml of distilled water and the respective insecticide at the required fi nal concentration.The paper cups were maintained at a room temperature of 25 ± 2°C.Controls were larvae kept in insecticide-free distilled water.There lies (Feyereisen, 2012), with families 4, 6 and 9 usually implicated in insecticide resistance (Amenya et al., 2008;Poupardin et al., 2010;Gong et al., 2013).The induction of P450s by insecticides is an approach commonly used to identify P450 genes potentially involved in mediating insecticide resistance (Poupardin et al., 2010;Feyereisen, 2012).Though insecticides are widely used in Malaysia, there is a paucity of information about their effects on the P450 genes of A. aegypti.The status of A. aegypti as a major vector of diseases and the extensive use of insecticides to control it emphasizes the importance of studying insecticide-responsive P450s in this mosquito as a fi rst step towards identifying putative insecticide-mediating P450s.This study investigated family 4 P450s in a Malaysian strain of A. aegypti by isolating a P450 cDNA and profi ling expression levels of this gene and another P450 gene (CYP4H28v2) (Elgarj & Wajidi, 2013) in the different life stages of this mosquito and in those exposed to sub-lethal concentrations of larvicides.Though CYP4H28v2 has been isolated (Elgarj & Wajidi, 2013), its expression profi le in the different life stages of this mosquito and in response to insecticides has not yet been determined.

Mosquito colony
The mosquitoes used in this study were subcultured at 27 ± 2°C and 80 ± 10% relative humidity (El-Garj et al., 2015) from a laboratory colony of A. aegypti, which originated from the Malaysian state of Penang in the 1980s and reared since then at the Vector Control Research Unit (VCRU) of Universiti Sains Malaysia during which time they were not exposed to insecticides.Fourth instar larvae were used for the experiments unless stated otherwise.

RNA isolation, cDNA synthesis, polymerase chain reaction (PCR) and cloning
RNA was isolated from the larvae (0.25 g) using the phenolchloroform method (Jowett, 1986).cDNA was synthesized from the RNA (5 μg) using a RevertAid™ Premium First Strand cDNA Synthesis Kit (Fermentas ® ).The use of commercial kits in this study followed the manufacturers' instructions unless stated otherwise.Forward (F001) and reverse (R459) primers based on the heme binding region (FXXGXXXCXG) of a partial length P450 sequence of A. aegypti (GenBank accession no.AY205085.1)(Elgarj & Wajidi, 2013) were used in a PCR (Table 1) containing 1 × OneTaq ® Hot Start Master Mix with Standard Buffer (New England Biolabs ® ), 0.2 μM of each primer, 0.4 μg cDNA and deionized water to amplify P450 fragments.The PCR products were separated on a 1% agarose gel and purifi ed using Wizard ® SV Gel and PCR Clean-Up System (Promega ® ).The purifi ed products were ligated into pGEM ® -T Easy Vector (Promega ® ) and transformed into Escherichia coli JM 109 using 1 × Transformation and Storage Solution (TSS) (Chung et al., 1989) as described below.The media used were Luria-Bertani (LB) agar (20 g/l) and LB broth (35 g/l).Single colonies of E. coli were inoculated into the LB broth and incubated overnight at 37°C and 180 rpm.The overnight culture was used to inoculate freshly prepared LB broth at a ratio of 1 : 100 and incubated (37°C and 180 rpm) for about 2 h or until the E. coli cells were growing exponentially and the OD 600 = 0.6.The cells were centrifuged at 12,000 g for 1 min and the pellet placed on ice for 5 min.The pellet was resuspended in 1 × TSS, which was placed on ice for 30 min, after which the ligation product (5 μl) was added, followed by a further period on ice were three replicates of each treatment.Surviving larvae were collected 1, 3 and 6 h after inducing the expression of the P450s.

Quantitative mRNA expression
Reverse Transcription-quantitative PCR (RT-qPCR) was used to investigate the constitutive and insecticide-mediated expression of CYP4H28v2 and CYP4H31v2 in this mosquito.Total RNA was extracted from the samples (about 30 mg each) using an RNeasy ® Mini Kit (Qiagen ® ) after manual homogenization using a sterile plastic pestle.The RNA was treated with DNase I (Fermentas ® ) to remove possible genomic DNA contamination as follows; incubation of 1 × reaction buffer with MgCl 2 and 1U DNase I, RNase-free per 1 μg RNA in a 10 μl reaction volume at 37°C for 15 min, then 5 mM EDTA was added and the mixture incubated at 65°C for 15 min.RNA integrity was evaluated on a 1.2% denaturing formaldehyde agarose gel by determining the 28S : 18S RNA ratio.A Nanodrop (ASP-2680) Spectrophotometer was used to measure RNA concentration and purity.RNA with a 28S : 18S ratio of 2 : 1 and purity of 1.8-2.0based on A 260 / A 280 was used to synthesize cDNA.cDNA synthesis for RT-qPCR was done using iScript TM Reverse Transcription Supermix kit (Bio-Rad ® ) with 1 μg total RNA, 1 × iScript TM RT Supermix and nuclease-free water.Aedes aegypti 40S ribosomal protein S17 (rpS17) (GenBank accession no.AY927787.2) was used as the normalization gene (El-Garj et al., 2016).Primers (Table 1) (El-Garj et al., 2016) were designed using Primer3Plus (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi/) and analyzed for secondary structure, melting temperature and complementarity using an Oligo Calc: Oligonucleotide Properties Calculator (http://www.basic.northwestern.edu/biotools/oligocalc. html).The expected product lengths of the genes were 214 bp for CYP4H28v2, 186 bp for CYP4H31v2 and 117 bp for rpS17.Optimization of the primer concentration and annealing temperature were done using concentration and thermal gradients of 0.1-1 μM and 50-65°C, respectively.Serial dilution of cDNA was conducted to determine standard curves for calculating amplifi cation effi ciencies of the genes based on three replicates for each dilution after fi ve intervals of time and a no-template control.Amplifi cation effi ciencies of the primers were within the acceptable range (98.7% for rpS17, 104.4% for CYP4H28v2 and 102.0% for CYP4H31v2) in accordance with MIQE guidelines (Bustin et al., 2009).Correlation coeffi cients (r 2 ) and slopes derived from the standard curves were 0.994 and -3.354 (rps17), 0.987 and -3.221 (CYP4H28v2) and 0.991 and -3.274 (CYP4H31v2).
Gene quantifi cation was done in a CFX96 Real-Time System (Bio-Rad ® ) using 0.2 μg cDNA, 1 × iQ TM SYBR ® Green supermix, 0.5 μM of each primer and nuclease-free water at the thermal condition specifi ed in Table 1. Assay specifi city was assessed using a melting curve recorded from 55.0°C to 95.0°C at increments of 0.5°C for 5 s.Quantifi cation cycle (Cq) for each reaction was determined using the CFX96 Real-Time System (Bio-Rad ® ).There were three biological replicates and three technical replicates for each sample.Fourth instar larvae and non-insecticide treated larvae (4 th instar) were used as calibrators for the life stage and insecticide-mediated expression, respectively.Relative expression of the genes was computed using the method of Pfaffl (2001).One-way analysis of variance with a 0.05 significance level was used to analyze the expression data.IBM SPSS 20.0 was used for the statistical analysis.Following Strode et al. (2008), a relative expression greater than or equal to 2 was considered as signifi cantly over-expressed while less than or equal to -2 was signifi cantly under-expressed.

Prediction of secondary and tertiary structures
The predicted secondary structure of the protein reveals the distribution of various structural components (Fig. S1) and using GOR IV, the distribution was as follows; alpha helix (35.25%), extended strand (20.79%) and random coil (43.96%).The 3D structure was generated using SWISS-MODEL after template search/identifi cation, templatetarget alignment and model building.The 3D model of CYP4H31v2 (Fig. 2) was generated using a human P450, CYP3A4 (PDB ID: 5A1P), which has a 31.22%sequence identity with CYP4H31v2 as the template.Ramachandran plot analysis indicated that 98.3% of the residues of this model were distributed within the most favored, additional allowed and generously allowed regions (Fig. S2).Analysis of the overall model quality using ProSA-web resulted in a z-score value of -7.56 (Fig. S2).

Constitutive expression and insecticide induction of P450s
Expression analysis of the P450s revealed signifi cant differences (p < 0.05) between the larvae and adults of A. aegypti (Fig. 3).Signifi cant over-expression of CYP4H28v2 and CYP4H31v2 were detected in the males.The larvae expressed lower levels of the P450 genes compared to the adults (Fig. 3).The symbols ▲ and ▼ indicate signifi cantly over-expressed and signifi cantly under-expressed genes, respectively.Male: 3 day old adult male; Female: 3 day old non-blood fed adult female; 1st larvae: 1st instar larvae, 2nd larvae: 2nd instar larvae; 3rd larvae: 3rd instar larvae.
Insecticide induction after different intervals of time revealed different expression levels of CYP4H28v2 and CY-P4H31v2 in the fourth instar larvae (Fig. 4).CYP4H28v2 was not signifi cantly increased by exposure to malathion, temephos and DDT at any of the three time intervals.In fact, malathion signifi cantly suppressed the expression of CYP4H28v2.Deltamethrin suppressed the expression of CYP4H28v2 after 3 h but caused a signifi cant over-expression of this gene after 6 h.Permethrin induced CYP4H28v2 and this increased as the exposure time increased (Fig. 4).Signifi cant under-expression of CYP4H31v2 was caused by the insecticides after different periods of exposure.However, signifi cant over-expression of CYP4H31v2 was detected in larvae exposed to DDT and pyrethroids for 6 h.Exposure to DDT and deltamethrin initially suppressed the expression of CYP4H31v2 after 3 h, but after 6 h, the expression level increased.When exposed to permethrin, CYP4H31v2 was suppressed after 1h but then its expression increased as the treatment period increased and was signifi cantly over-expressed after 6 h of exposure (Fig. 4).

DISCUSSION
P450s have been extensively studied in insects as they have innumerable roles in insect development, physiology and survival (Feyereisen, 2012).Their effect on insect survival, especially by detoxifying insecticides, makes them prime candidates for insecticide resistance studies, usually revolving around identifying P450s with putative roles in insecticide resistance (Poupardin et al., 2010;Liu et al., 2011;Gong et al., 2013).Though P450s may mediate insecticide resistance in Malaysian A. aegypti (Ishak et al., 2015), neither identifi cation of individual P450s with potential roles in insecticide resistance nor characterization of already identifi ed P450s from other strains under insecticide-induced conditions have been conducted for Malaysian strains.This study reports the cloning of a P450 cDNA encoding CYP4H31v2 from the VCRU strain of A. aegypti.This gene is a variant of the CYP4H31 gene in the Liverpool strain.Another P450 variant, CYP4H28v2, has also been recorded in the VCRU strain (Elgarj & Wajidi, 2013), though not much is known about it.The CYP4H28 and CYP4H31 genes are expressed in various proportions when exposed to permethrin and temephos in different A. aegypti strains (Saavedra-Rodriguez et al., 2012, 2014).Though the two A. aegypti strains are from different geographical areas, there is a very high conservation of the P450 sequences recorded for them, implying a similarity in properties and functions of the alleles from the Malaysian and Liverpool strains.However, variants of the same P450 gene could also respond differently to insecticide pressure.The genes CYP6A5 and CYP6A5v2 have a protein sequence identity of 99% and according to Zhu & Liu (2008), while CYP6A5v2 was over-expressed in a pyrethroid resistant strain of house fl y compared to a susceptible strain, there was no signifi cant difference in expression of CYP6A5 in both strains.There was also a remarkable difference in the expression levels of two genes, CYP4D42v1 (3.2-fold) and CYP4D42v2 (7.0-fold) in permethrin selected Culex quinquefasciatus assessed using qPCR (Reid et al., 2012).Hence, it is important to characterize the P450 variants and determine their insecticide induction response.
Secondary structures of P450s are made up of various components (Werck-Reichhart & Feyereisen, 2000) and in this study, alpha helices and random coils constituted the majority of the structural components of the protein, similar to the predicted secondary structure of an insect P450, CYP6AE25 (Zhang et al., 2011).However, it should be noted that the secondary structure components of CY-P4H31v2 are computational predictions and may not be a 100% representation of the actual secondary structure of the protein.There are no crystal structures of insect P450s; hence computational models are used to generate 3D structures for insights into insect P450 structures and for further computational investigations.The 3D model of CYP4H31v2 has the conserved folds of P450s (Werck-Reichhart & Feyereisen, 2000).The conserved core is made up of a bundle consisting of the helices D, E, I and L, the J and K helices, two β-sheets and the meander region (Werck-Reichhart & Feyereisen, 2000).These structures also contain P450 motifs and the catalytic heme binding region of P450s (Werck-Reichhart & Feyereisen, 2000;Feyereisen, 2012).Ramachandran plot of the CYP4H31v2 model showed that 1.7% of the residues were located in the disallowed regions.This compared favorably with residues in the disallowed regions of models of Anopheles minimus CYP6AA3, CYP6P7 and CYP6P8 and human CYP4A22, which ranged from 0.5 to 2.9% (Gajendrarao et al., 2010;Lertkiatmongkol et al., 2011).The ProSA z-score indicated that the overall quality of the model was good and comparable to the z-score of experimentally determined proteins of similar size.
Constitutive expression profi les of CYP4H28v2 and CY-P4H31v2 differed in the different developmental stages of this mosquito.The P450s were suppressed in the larvae compared to the adults.High expression of the P450s in the adult stage compared to the larval stage suggests that the genes may be involved in adult stage activities, which could include survival when exposed to adulticidal compounds.In the adults, the genes were over-expressed in the males but not the females, though a more than 1-fold expression was detected in the females.Survival by insecticide detoxifi cation and reproduction are important activities needed for the perpetuation of a population.However, resources may be limiting for performing both tasks, hence an optimal resource allocation occurs, which is refl ected in a trade-off between the competing demands of reproduction and insecticide detoxifi cation resulting in the suppression of P450s in females (McGraw et al., 2004).
Insecticide-based vector control relies on the effi cacy of the chemical agents used to control the target species (Vontas et al., 2012;Avicor et al., 2013Avicor et al., , 2015;;El-Garj et al., 2015).However, the development of resistance lowers the effi cacy of insecticides and diminishes the effectiveness of insecticide-based vector control (Temu et al., 2012).Therefore, identifying P450s with putative roles in insecticide resistance is key to resistance detection and management for an effective vector control system.Insecticides can induce P450s and the over-expression of P450s in insecticide-resistant and treated strains sometimes provide cues for identifying putative insecticide detoxifying P450s (Amenya et al., 2008;Poupardin et al., 2010;Huang et al., 2013).Some studies on the induction potential of insecticides on P450s use sub-lethal concentrations as in this present study and other previous works (Poupardin et al., 2010;Avicor et al., 2014) to record the adaptive response of insects to toxic compounds.Differential expression of CYP4H28v2 and CYP4H31v2 in A. aegypti larvae in this study was dependent on the type of insecticide and exposure time.This pattern is consistent with previous studies on other insects, including species of Aedes (Poupardin et al., 2010;Liu et al., 2011;Huang et al., 2013;Avicor et al., 2014).CYP4H28v2 and CYP4H31v2 were strongly induced by permethrin after a 6 h exposure.A similar pattern was also recorded in larvae exposed to deltamethrin for 6 h.Pyrethroid-induced expression of the A. aegypti P450s is consistent with previous studies on P450 induction in pyrethroid-treated insects (Jiang et al., 2010;Poupardin et al., 2010;Liu et al., 2011;Huang et al., 2013).Signifi cant up-regulation of CYP4H28v2 and CYP4H31v2 in this strain implies the potential involvement of both genes in pyrethroid tolerance, corroborating the suspected involvement of P450s in resistance of A. aegypti from this region based on synergistic assays (Ishak et al., 2015).Larvae exposed to DDT signifi cantly expressed only CYP4H31v2 after a 6 h exposure.Over-expression of the genes after a longer exposure may be attributed to a delay in the response of the mosquito to the insecticide.Therefore, longer exposure could strongly induce P450s but this can also cause an inverse response as indicated in the under-expression of CYP9M8 in permethrin exposed A. aegypti as induction time increased from 6 to 48 h (Poupardin et al., 2010).
Mosquito strains vary in their capacity to express P450s in response to insecticide stress (Liu et al., 2011;Saavedra-Rodriguez et al., 2014) and in this study; malathion and temephos suppressed the expression of CYP4H28v2 and CYP4H31v2 in the larvae regardless of the exposure time.In a previous study (Huang et al., 2013), malathion does not over-express family 4 P450s in Bactrocera dorsalis and signifi cantly suppresses the expression of some of the genes at longer post-treatment sampling periods.Saavedra-Rodriguez et al. ( 2014) also reported the suppression of CYP4H28 and CYP4H31 in some fi eld strains of A. aegypti following temephos treatment, similar to the results of this study, however, over-expression of the genes was also noticed in some strains in that study.Though El-Garj et al. ( 2016) recorded an increase in expression of CYP4H28v2 and CYP4H31v2 in adult females, which were briefl y exposed to paper impregnated with DDT (4%), malathion (0.8%) and permethrin (0.25%), organophosphate in the present study suppressed the expression of these genes in larvae.This could be due to the low basal levels of CY-P4H28v2 and CYP4H31v2 in larvae compared to adults.Suppression of CYP4H28v2 and CYP4H31v2 in organophosphate-treated larvae in the present study is indicative of their limited effect in detoxifying these insecticides at this stage of the insect's development.It is also likely to be an adaptive response that regulates the internal system of an organism that has been altered by insecticidal stress (Morgan, 2001).
In conclusion, the cloning and characteristics of a P450 gene, CYP4H31v2, including its predicted 3D model was reported.mRNA expression profi les of CYP4H28v2 and CYP4H31v2 revealed high levels of expression of both genes in adult males.Subjecting the larvae to insecticide stress induced differential expression of the genes with signifi cant over-expression of CYP4H28v2 in permethrin-and deltamethrin-treated larvae and signifi cant over-expression of CYP4H31v2 in DDT-, permethrin-and deltamethrintreated larvae.This study highlights the potential importance of these genes in the tolerance of A. aegypti of the aforementioned insecticides and the need to use genetic linkage analysis and metabolism assays to understand and pinpoint their functional roles in insecticide metabolism.

Fig. 1 .
Fig. 1.Sequence of full-length CYP4H31v2 from Aedes aegypti.The sequence contains 1671 nucleotides and has an open reading frame of 505 amino acids.P450 motifs (WXXXR, EXXR, PXXFX-PXRF and FXXGXRXCXG) and the commonly found conserved sequence EVDTFMFEGHDTT of family 4 P450s are underlined.There are 62 and 91 nucleotides in the 5'-Untranslated Region (UTR) and 3'-UTR, respectively.The canonical polyadenylation signal AATAAA and the poly-A tail (25 nucleotides long) are in lowercase bold.

Fig. 2 .
Fig. 2. Predicted tertiary structure of CYP4H31v2.Structural components are indicated on the model, with alpha helices and beta strands beginning with the prefi xes α and β, respectively.

Fig. 3 .
Fig. 3. Quantitative analysis of CYP4H28v2 and CYP4H31v2 in different developmental stages of Aedes aegypti (p < 0.05).The relative expression value is the ratio of expression in the specifi ed developmental stage compared with that in the fourth instar larvae.Relative expression is shown as mean ± standard error.The symbols ▲ and ▼ indicate signifi cantly over-expressed and signifi cantly under-expressed genes, respectively.Male: 3 day old adult male; Female: 3 day old non-blood fed adult female; 1st larvae: 1st instar larvae, 2nd larvae: 2nd instar larvae; 3rd larvae: 3rd instar larvae.

Fig. 4 .
Fig. 4. Quantitative analysis of (A) CYP4H28v2 and (B) CYP4H31v2 in fourth instar larvae of Aedes aegypti after insecticide induction for different lengths of time (p < 0.05).The relative expression value is the ratio of expression in insecticide-induced fourth instar larvae compared to that in non-insecticide induced fourth instar larvae after a specifi ed period of time.Relative expression is shown as mean ± standard error.The symbols ▲ and ▼ indicate signifi cantly over-expressed and signifi cantly under-expressed genes, respectively.

Fig. S2 .
Fig. S2.Structure validation of CYP4H31v2 model.(A) Ramachandran plot analysis of CYP4H31v2 model.Red, yellow, cream and white regions are the most favored, additional allowed, generously allowed and disallowed regions, respectively.Distribution of the residues within the regions is as follows; 86.7% in the most favored regions, 10.9 % in additional allowed regions, 0.7 % in generously allowed regions and 1.7 % in disallowed regions.(B) Z-score of CYP4H31v2 model.The X-axis shows the number of residues of protein structures obtained fusing X-ray or NMR while the Y-axis shows their z-score values.Z-score of CYP4H31v2 indicates the quality of the generated CYP4H31v2 model in comparison with native proteins of similar size experimentally determined using X-ray or NMR.

Table 1 .
Primers and thermal conditions used for PCR and qPCR.