Báo cáo Y họcMolecular characterization of the amplified aldehyde oxidase from insecticide resistant Culex quinquefasciatus docx

Molecular characterization of the amplified aldehyde oxidase from insecticide resistant Culex quinquefasciatus Michael Coleman, John G. Vontas and Janet Hemingway Liverpool School of Tropical Medicine, UK Primary structural information including the complete nucleotide sequence of the first insect aldehyde oxidase (AO) was obtained from the common house mosquito Culex quinquefasciat us (Say) through cloning and sequencing o f both g enomic DNA and cDNA. The d educed amino-acid sequence encodes a 150- kDa p rotein of 1266 amino-acid residues, which is consistent with the expected monomeric subunit size of AO. The Culex AO sequence contains a molybdopterin cofactor binding domain and two iron–sul- fur centres. A comparison of the partial sequen ces of AO from insecticide resistant and susceptible strains of C. quin- quefasciatu s shows two distinct alleles of this enzyme, one of which is amplified in the insecticide resistant strain on a 30-kb DNA amplicon alongside two resistance-associated esterases. The amplified AO gene results in elevated AO activity in all life stages, but activity is highest in 3rd instar larvae. The elevated enzyme can b e seen as a separate band on polyacrylamide gel electrophoresis. The role of AO in xenobiotic oxidation in mammals and t he partial inhibition of elevated AO activity by a range of insecticides in Culex , suggest that this AO may play a role in insecticide resistance. Keywords: aldehyde oxidase; mosquito; insecticide resis- tance. Culex mosquitoes are major vectors of filariasis and Japanese encephalitis as well as a general biting nuisance. Amplification of nonspecific esterases accounts for > 90% of known insecticide resistance in Culex populations. In the C. pipiens complex, distribution of amplified esterase alleles is geographically restricted, except for esta2 1 and estb2 1 , which are coamplified on a single DNA amplicon and occur world-wide. The rapid spread of this amplicon through Culex populations already containing alternative esterase alleles on resistance-associated amplicons suggests that the esta2 1 /estb2 1 amplicon has a very strong selective advantage over other esterase-based resistance mechanisms [1]. A comparison of esterases from resistant C. quinquefasciatus strains with different amplified esterases suggests no s elec- tive advantage of esta2 1 and estb2 1 over other resistant strains based on their enzyme activity alone [2]. The selective advantage observed for strains, such as PelRR, with this amplicon, must therefore be due to some other factor. Recently, w e have reported that the PelRR amplicon contains a third complete gene, putatively aldehyde o xidase (AO) [3]. It is possible that either the AO or esterases on the amplicon affect mosquito viability in the presence of filarial parasites, as we have shown that parasite survival and insecticide resistance status are highly negative ly correlated [4], which may contribute to the lack of correlation between mosquito biting rates, prevalence of microfilaraemia and disease which has been noted in several studies [5]. Alternatively, the AO m ay have a direct role in insecticide resistance. Cytochrome P450s have traditionally been thought of as the sole enzyme system associated with increased l evels of insecticide oxidation. However, in mammals oxidation of xenobiotics by the molybdopterin family of enzymes is now well documented [6,7]. AO is a molybdenum-containing enzyme belonging to this family. The tissue localization and physiological roles of this enzyme are still not fully understood. The broad substrate specificity of AO makes it a useful mammalian prodrug activator [8,9], and AO functions in herbivores to protect them against plant toxins [6]. The amplification of AO in insecticide-resistant insects may therefore have a functional significance, which h as been overlooked to date. There is no known sequence data for AO in insects, although t he enzyme has a number of diverse physiological functions [10–12] and its distribution patterns in Drosophila [13,14] and Musca domestica have been studied histochem- ically [15]. Here we report the first genomic, cDNA and deduced protein sequences of AO from the mosquito Culex quinquefasciatus, demonstrate that the allele present on the amplicon is expressed in insecticide-resistant insects and that it interacts with insecticides. EXPERIMENTAL PROCEDURES Mosquito strains C. quinquefasciatus larvae were collected from Peliyagoda, Sri Lanka in 1984. Th is population, which had been under fenthion selection pressure for several years, was selected i n the laboratory to produce two strains: an insecticide susceptible strain, PelSS, with the nonamplified esterases esta3 and estb1 2 , and an organophosphorus insecticide- resistant strain, PelRR, with the two coamplified esterases esta2 1 /estb2 1 [16,17]. The C. quinquefasciatus str ain TemR Correspondence to J. Hemingway, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, UK. Fax: + 44 151 7088733, Tel.: + 44 151 7089393, E-mail: hemingway@liverpool.ac.uk Abbreviations: AO, aldehyde oxidase; ALDH, aldehyde dehydrogen- ase; XDH, xanthine dehydrogenase. (Received 18 July 2001, revised 14 November 2001, accepted 16 November 2001) Eur. J. Biochem. 269, 768–779 (2002) Ó FEBS 2002 was obtained from G. Georghiou, University of Califor- nia, Riverside, USA. TemR is resistant to organophos- phates due to the amplification of a single esterase gene, estb1 1 [18]. Genomic DNA sequence A genomic library of PelRR fourth-instar larvae was constructed in the kGEM-11 vector (Promega) and probed with a partial esta2 1 cDNA as previously described [19,20]. The sequence downstream of the esta2 1 gene from the resultant positive bacteriophage clone (A2) suggested a third ORF with high homology to the molybdenum containing enzymes xanthine dehydrogenase (XDH) and AO [3]. Bacteriophage A2 was produced for analysis by inoculating 400 mL of Luria–Bertani broth [0.1% (w/v) bacto-tryptone, 0.05% (w/v) bacto yeast extract, 0.1% (w/v) NaCl] with 6mLofEscherichia c oli LE392 culture, grown overnight in Luria–Bertani broth + 0.2% maltose and incubated for 20 min, 37 °C, 225 r.p.m. The culture was inoculated with 10 9 plaque forming units of bacteriophage A2. After allowing the mixture to stand for 20 min at 37 °C, the culture was grown at 37 °C, 225 r.p.m. for 6 h. Chloroform (2 mL) w as added to lyse any remaining cells. Ten grams NaCl, RNase 1 m gÆmL )1 and DNase 1 mgÆmL )1 were added and the mixture incubated for 1 h at room temperature. The cell d ebris was removed by centrifugation at 12 000 g,4°C for 10 min 10% (w/v) of poly(ethylen e glycol) 6000 was gently dissolved into the s upernatant, and the mixture incubated at 4 °C for 10 min and resuspended in 5 mL S M [0.58% ( w/v) NaCl, 0.2% (w/v) M gSO 4 ,5% 1 M Tris/HCl pH 7.5, 0.01% gelatin]. Chloroform extraction was carried out and the superna- tant removed. CsCl 2 (0.75 g per mL) was dissolved into the supernatant a nd the m ixture centrif uged at 100 000 g, 10 °C for 24 h. The DNA band was dialysed overnight against 10 m M NaCl, 50 m M Tris/HCl pH 8.0, 10 m M MgCl 2 EDTA. After dialysis, proteinase K (50 mgÆmL )1 ) and SDS (0.5%) were added and the mixture incubated at 65 °C for 1 h . Phenol and chloroform extractions were carried out before precipitating the DNA in ethanol and resuspending in Tris/HCl pH 8.0. Restriction digests and subcloning of the A2 insert was undertaken to analyse the AO sequence. A2 was digested with BamHI and SacI and run o n a 1% agarose gel (Bio- Rad). The three resultant bands were extracted from the gel with the Wizard DNA Clean-up System (Promega) and subcloned into pBluescript (Stratagene). The ligation prod- ucts were used to transform E. coli XL-1Blue (Stratagene) and recombinant plasmids were isolated from a mplified bacterial colonies using a standard miniprep method (Qiagen). PCR was used to produce sequence spanning across the three subcloned fragments. P rimer X6 (5¢-GGTGTACA ACGTGCAGGA-3¢)andY4(5¢-GAGCGAGAACGAG CCGGAAC-3¢)wereusedtoPCRbetweenplasmids AO1 and AO2. Primer Y6 (5¢-GCCGAAATGTGATTAT TTG-3¢)andA1(5¢-TTAGCCC GAACCGCGGCC-3¢) were used to PCR across plasmids AO2 and AO3. These PCR products were ligated into p GEMT-easy (Promega) and positive colonies were s elected and prepared as a bove. A contig of the complete bacteriophage insert was made b y combining these sequence data. Synthesis of cDNA and sequencing of AO Total RNA was isolated from 1 g of fourth instar larval PelRR using TRI reagent (Sigma) according to the man- ufacturer’s instructions. Reverse transcription of first strand cDNA from mRNA was accomplished with SuperScript TM RT (Gibco BRL) according to the manufacturers instruc- tions with an oligo(dT) adaptor primer [5¢-GACTCG AGTCGATCGA-(dT) 17 -3¢]. Primers were designed to the putative 5¢,F1(5¢-A TG GAAGTCATATTTACGAT-3¢)and3¢,F2(5¢-TTG TAGTTTAAACTGTTC-3¢) ends of AO based on the genomic sequence. The 50-lL PCR reaction contained 20 ng of first strand cDNA, 150 ng o f each primer, 0.5 m M dNTPs, 2.5 m M MgCl 2 ,1.25UPfu DNA polymerase (Stratagene) 5 U of Amplitaq Gold DNA polymerase (PerkinElmer) and Taq DNA polymerase buffer. After one cycle of 95 °C for 10 min to activate the Amplitaq,35cycles of amplification were carried out as follows: 95 °C, 45 s; 50 °C, 45 s; 72 °C, 7.5 min. 3¢ RACE RACEwasusedtoobtainthe3¢ UTR of AO. Primer A20 (5¢-CCGAGAACTTGATCTACAG-3¢)designedtothe3¢ end o f t he cDNA was used in conjunction with an adaptor primer (5¢-GACTCGAAGTCGACATCGA-3¢)inaPCR reaction. The PCR reaction was carried out as above, without Pfu DNA polymerase and with an extension time of 2 min. 5¢ PCR The p artial 5¢ UTR was obtained using a primer designed from the genomic DNA 102 base pairs upstream of the transcription start codon, UTR5 (5¢-GCACTGTTTAACT CAGTTCG-3¢) and a primer (X7) designed to t he 5¢ end of the cDNA (5¢-TCCTGCACGTTGTACACC-3¢). The PCR reaction was carried out as for the 3¢ RACE. PCR products were isolated using a Wizard DNA Clean- up System (Promega) and subcloned into pGEMT-Easy (Promega) for sequencing. DNA sequencing Initially the inserts were sequenced using universal M13 forward and reverse primers complimentary to the plas- mids. Internal primers were synthesized based on this initial sequence data. Sufficient primers were synthesized to allow sequencing of both strands of the AO gene at least twice. Sequencing was carried out with an ABI Automatic Sequencer (PerkinElmer). Sequence data were analysed with the DNASTAR (Lasergene) program. Genomic Southern blots Genomic DNA was extracted by the method of Vaughan et al. [21]. Ten milligrams of PelRR, PelSS and TemR DNA was digested with Eco RV restriction enzyme a nd the products separated by electrophoresis on a 0 .8% agarose gel. The DNA was denatured, neutralized [22] and trans- ferred to a nylon membrane (Amersham) using the Hybaid Vacu-Aid. The DNA was fixed to the membrane with UV Ó FEBS 2002 AO from C. quinquefasciatus (Eur. J. Biochem. 269) 769 light. Membranes were prehybridized at 65 °Cfor1hin hybridization buffer (6 · NaCl/Cit, 0.1% (w/v) SDS, 0.1% (w/v) sodium pyrophosphate, 5% poly(ethylene glycol), 5 · Denhardt’s solution). The genomic clone AO1 was digested with HincII and PstI. Products were separated by electrophoresis and the 0.65-kb band with high homology to AO was extracted from the agarose using a Wizard TM kit (Promega). It was labelled with 32 P (specific activities > 2 · 10 6 c.p.m.Æmg )1 ) by random priming with a Pharmacia oligonucleotide labeling kit and used a s a probe. The probe was hybridized to t he phage DNA overnight at 60 °Cin hybridization buffer. Final washes were in 0.1 · NaCl/Cit, 0.1% (w/v) SDS at 60 °C for 45 min. Aldehyde oxidase assay Individual mosquitoes were assayed for AO activity by a method adapted from Moura & Barata [23] and Mira et al. [24]. Briefly, individual larvae were homogenized in 40 lL potassium p hosphate buffer pH 7.8 with 1 m M EDTA. Two replicates of 10 lL were transferred to a microplate and 200 lL of reaction mixture containing 0.1 mgÆmL )1 phenazine methosulfate, 0.1 mgÆmL )1 2,6-dichloroindophe- nol, 50 l M allopurinol, and 0.1 m M of a ‘neat’ m ixture of aldehyde substrates (1 : 1, v/v, acetaldehyde/benzaldehyde) was added. AO a ctivity was determined by measuring the rate of 2,6-dichloroindophenol reduction at 600 nm (e ¼ 21 m M )1 Æcm )1 ) as aldehyde is enzymatically oxi- dized. Kinetics were read immediately, by following the decrease in absorbance at 650 nm for 5 min. Specific activities are given in UÆmg )1 protein where a unit corre- sponds to 1 lmol o f 2,6-dichloroindophenol reduced per min, under the assay conditions used. All assays were compared to controls of id entical composition lacking substrate (aldehydes) or homogenate. Measurement of aldehyde dehydrogenase (ALDH) activity The ALDH assay was performed by the method of Tasayco & Prestwich [25], under anaerobic conditions. A 1-mL aliquot of potassium phosphate buffer pH 7.8 saturated with nitrogen, containing 1 m M NAD + and 10–15 mg protein of pooled crude homogenate was added to a 2-mL cuvette. Four millilite rs of 10 m M aldehyde in ethanol was added anaerobically and the appearance of NADH (e ¼ 6.22 m M )1 Æcm )1 ) was recorded continuously for 10 min at 340 n M . Specific activities are given in UÆmg protein )1 . A unit corresponds to the production of 1 nmol NADH in 1 min. Protein assay Protein content was determined b y the method of Bradford [26]. Protein values in mg ÆmL )1 were calculated from a standard curve o f a bsorbance of known concentrations o f bovine serum albumin. Inhibition of AO activity by pesticides and inhibitors Crude homogenates for the pesticide and effector experi- ments were prepared in ice-cold 0.1 M phosphate buffer (pH 7 .8) with 5% (v/v) glycerol. S topped time inhibition assays were performed. Solutions of AO inhibitors and various pesticides were prepared in either phosphate buffer or acetonitrile depending on their solubility (acetonitrile concentration of the medium never exceeded 1%, v/v). Each effector was preincubated with the crude homogenate for 15 min at 20 °C. AO or ALDH residual activity were then measured as described above in t he presence of each effector, except that phenazine methosulfate was omitted from the AO activity reaction mixture. Electrophoresis Electrophoresis of native protein samples was performed in a Phastsystem (Pharmacia). Crude homogenates were prepared as described above. Two microliters of each sample ( 5 lg p rotein) w as applied to a 8–25% gradie nt Phastgel (Pharmacia) with standard molecular mass mark- ers (Sigma) and subjected to native PAGE Phastsystem (400 V, 10 mA, 2.5 w, 10 °C, 390 Vh). Gels were divided to v isualize standard proteins by Comassie Blue R250 staining and AO activity bands using a formazan staining solution, prepared according to Tasayco & Prestwich [25]. Briefly, 20 mg Nitro Blue tetrazolium, 0.8 m g phenazine methosulfate, 50 mg allopurinol (to inhibit xanthine oxidase and XDH) and 1 mL of aldehyde substrates (acetaldehyde, benzaldehyde, Dimethyl-amino- benzaldehyde or heptaldehyde), were added in 50 mL 0.1 M Tris/borate buffer, pH 8.0 with 1 m M EDTA. All native gels were compared to controls of identical composition with aldehyde omitted. RESULTS A complete ORF (ORF) coding for a putative AO enzyme was obtained f rom genomic DNA of the C. quinquefascia- tus mosquito strain PelRR (Fig. 1). The computer-based analysis of the 5¢ flanking region of the AO gene identified several potential transcription factor-binding sites. Amongst them were three Barbie boxes located at )1137 to )1122, ) 666 to )652 and )539 to )525 to this ORF. These sequences have been previously found in the promoter regions of most GSTs with the core sequence o f AAAG common in all of them [27,28]. This element might be responsible for the induction of the GST genes by the drug phenobarbital and may play a role in drug resistance during cancer treatment. The common arthropod initiator sequence, TCAGT, occurs at both positions 13–18 (AI1) and 25–29 (AI2), with a possible TATA box located at )12 to )3, relative to the +1 of the 5¢ UTR, suggesting that this ORF codes for a functional gene. Primers designed to the 5¢ and 3¢ ends of the AO genomic DNA sequence were used to obtain a full-length cDNA of 3798 nucleotides from fourth instar Culex larvae. This cDNA sequence was completely homologous to the predicted exon regions of the genomic DNA sequence and the structure is superimposed on the genomic sequence in Fig. 1. A presumptive polyadenylation site, AATAA, is located at 5884–5889 (7540 on genomic sequence). The cDNA ORF pred icts a protein of 1266 amino-acid residues and an molecular mass o f 150 kDa. This is consistent with the expected monomeric subunit size for AO. 3¢ RACE was used to obtain a 3¢ UTR of 67 nucleotides that includes the presumptive polyadenylation site. A 102-bp 5¢ UTR w as 770 M. Coleman et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Ó FEBS 2002 AO from C. quinquefasciatus (Eur. J. Biochem. 269) 771 772 M. Coleman et al. (Eur. J. Biochem. 269) Ó FEBS 2002 obtained by PCR usin g primers designed to the genomic DNA sequence. The cDNA encodes a protein with two predicted iron– sulfur [2Fe)2S] centres between amino-acid residues 33–62 and 138–164. There is a highly conserved molybdopterin cofactor binding site between residues 706–752 (Fig. 2 ), and a region with high homology to the NADH binding site of various Drosophila enzymes [29] is located at 343–515. This region of the PelRR AO sequence has little homology with the related molybdopterin enzyme XDH and there is no putative NAD + binding domain in the mosquito sequence, as would be expected if this ORF coded for XDH. The PelRR AO has a high d egree of homology to XDH from Drosophila melanogaster (51%), Bombyx mori (51%) and D. subobscura (50% ) which is expected, as AO and XDH are closely related enzymes. Assigning this gene as an AO is supported by the three conserved active site centers and the high homology to bovine AO (51%), human AO (51%) and Arabidopsis AO (50%) (Fig. 2). A Southern blot of genomic DNA from various mosquito strains (Fig. 3), shows that this AO is amplified in PelRR, in contrast to PelSS and the insecticide resistant strain TemR, which has an amplicon containing estb1 1 .The complete AO ORF in PelRR is located on the same insecticide resistance-associated amplicon as the esterase genes esta2 1 /estb2 1 . The coding regions of the genomic AO and esta2 1 DNA sequences overlap at their extreme 3¢ ends in the PelRR amplicon. There are five introns within the ORF. All are small introns of < 200 bp with the exception of intron 3 which is  1700 bp in length (Fig. 1). A partial sequence of an AO from PelSS DNA is conserved at the 3¢ end of the gene (329/330 nucleotides) and has lower conservation in t he mid r egion (263/276 nucleotides) when PelRR and PelSS sequences are c ompared, suggesting that the two strains carry different alleles of this gene. The 3¢ end of the gene contains the conserved active site centers hence the lack of variability in this region i s expected. On native PAGE gels it is difficult to distinguish between ALDH and AO [30]. Allopurinol was used to inhibit xanthine oxidase and XDH bands which will stain with AO substrates. Figure 4 indicates that the upper band is xanthine oxidase and the two lower bands in PelRR are either AO or ALDH. Three bands occurred in PelRR adults in contrast to two in P elSS when equal amounts of protein were loaded (Fig. 5). Size estimation of the novel AO band was obtained under native conditions using its relative mobility and the linear regression equation for the molec- ular mass markers. The apparent molecular m ass of the amplified AO was estimated in its native form as approx- imately 302 kDa, which is in line with t he predicted molecular mass from the translated amino-acid sequence for a homodimeric AO, as recorded in other insects and mammals. The novel AO band in PelRR is stage specific (Fig. 6), with highest activity during t he larval stages, peaking in the third and fourth instars and decreasing to low levels in the pupal and adult stages. The specific AO activity of PelRR larval crude homo- genate using this assay was approximately fourfold higher than that in PelSS crude homogenates (Fig. 7). However, this fourfold elevation in the aldehyde oxidizing activity can not be readily compared to the degree of the AO genomic amplification (which is estimated as being much higher), as a many fold AO upregulation may only increase the overall activity a few fold. Adult PelRR and PelSS had similar levels of AO activity (data not shown), due to the lower expression of the amplified AO in this life stage, in comparison with other aldehyde oxidizing enzymes (such as ALDH). The identification of the novel PelRR enzyme as an AO was confirmed by measuring AO substrate oxidation anaerobically. These conditions suppress AO activity and any remaining enzymatic activity is due to ALDH. PelRR and PelSS had similar ALDH specific activities (Table 1). Any differences between the activity of the two strains with Fig. 1. Nucleotide sequence of C. quinquefasciatus PelRR AO genomic sequence, deduced after complete sequencing of both DNA strands (Accession no. AF 202953). A PelRR AO cDNA was a lso sequenced in both directions and e xactly matched the underlined exon sequences of the genomic DNA. The potential arthropod initiator sequences and poly A site are indicated as AI1, AI2 and PA, respectively. Some of t he p rimers used in PCR have been indicated. The 5¢ UTR and 3¢ UTR sequences are also shown. Ó FEBS 2002 AO from C. quinquefasciatus (Eur. J. Biochem. 269) 773 774 M. Coleman et al. (Eur. J. Biochem. 269) Ó FEBS 2002 AO substrates is therefore due to differences in AO activity. The novel PelRR AO had a high affinity for the substrate heptaldehyde, which is a good AO, but poor ALDH substrate, supportin g our earlier suggestion from anaerobic assays that this novel band is not an ALDH. Both bands were prominent with acetaldehyde. Because t he true kinetic parameters (V max or K m )and inhibition constants are not readily obtainable for a mixed Fig. 3. Southern blot of genomic DNA from the PelRR, PelSS and TemR strains of Culex quinquefasciatus dige sted with EcoRV demon- strating genomic amplification of AO in resistant PelRR, but not in resistant TemR. Fig. 4. Specific staining and inhibition of the xanthine oxidase (XO) stained band. Equal crude h omogen ate protein samples from C. quin- quefasciatus PelRR adults were loaded to each well of a gradient PhastGel (8–25%) and subjected to native PAGE P hastsystem. The formazan staining solution was prepared as described in materials and methods, with acetaldehyde/benzaldehyde substrates and the xanthin e oxidase inhibitor allopurinol (lane 1) and the specific xanthine oxidase substrate hypoxanthine (lane 2). The positions of the molecular mass markers are indicated on the left. Fig. 5. The amplified AO stained band in the insecticide resistant C. quinquefasciatus PelRR strain. (A)NativePAGEoftheinsecticide susceptible PelSS (lane 1) and the insecticide resistant PelRR (lane 2) Culex quinque fasciat us adults. Equal amounts of crude homogenate of pooled samples were loaded to each well o f a gradien t P hastGel (8– 25%) and a neat mixture of aldehyde substrates [1 : 1 (v/v) a cetalde- hyde/benzaldehyde] was used a s substrate. (B) The molecular mass was estimated by bilogarithmic plotting of molecular masses of the stan- dards against T%, which was the total polyacrylamide concentration reached by each protein afte r electroph oresis. The stan dard markers were as follows: (a) urease (hexamer, M r 545 0 00); (b) urease (trimer, M r 272 000); (c) albumin bovine s erum (dimer, M r 132 000); (d) albumin bo vine seru m (mono mer, M r 66 000); (e) albumin chicken egg (M r 45 000). Fig. 2. Comparison of the putative AO f rom PelRR with the amino-acid sequences of bovine, human and maize AO with XDH from Drosophila melanogaster. Common residues between sequences are boxed. Ó FEBS 2002 AO from C. quinquefasciatus (Eur. J. Biochem. 269) 775 enzyme system a simple inhibition study with various AO inhibitors and pesticides w as performed. Table 2 shows the percentage inhibition for each chemical tested on t he standard AO activity assay. Methadone, a potent inhibitor of rat AO, was the most potent inhibitor of PelRR AO activity. All four pesticides at concentrations of 0.05– 0.1 m M produced partial inhibition of AO activity, as did two commonly used herbicides a t concentrations of 1 m M . Only the triazine h erbicides had an inhibitory effect on anaerobically measured ALDH activity (data not shown), suggesting that inhibition is due to interaction with the AO enzyme. DISCUSSION Cytochrome P450s have traditionally been considered as the only enzymes to oxidize insecticides. Other oxidizing enzymes, such as AO and XDH have only recently been recognized as important in the oxidation of many drugs and xenobiotics [8]. A O is capable of utilizing a wide r ange of substrates such as, N-heterocyclics, aldehydes (which includes a number of drugs), azo dyes and N-oxides. Hepatic AO in humans mediates the oxidation of a large number of such compounds [31]. Bovine AO is expressed at high levels in the liver and lungs and is implicated in the detoxification of environmental pollutants [32]. The pres- ence of an amplified AO on the insecticide resistance- associated amplicon of C. quinquefasciatus opens up the possibility that this enzyme, may play a role in insecticide Table 2. Influence of pesticides and inhibitors on the AO activity of pooled C. quinquefasciatus PelRR larvae. AO activity was measured with a reaction mixture containing 2,6,d ichloroindo phen ol and 1 : 1 acetaldehyde/benzaldeh yde (0.1 m M ). The data are means ± SD of three separate experiments each of which was performed in duplicate. Pesticides Inhibitor concentration (m M ) Remaining activity (%) Herbicides S-Triazine 1 73.9 ± 2.0 Atrazine 1 67.4 ± 3.2 Insecticides Thiabendazole 0.05 76.2 ± 4.5 Methidathione 0.5 82.5 ± 3.5 Diazinon 0.1 85.7 ± 8.4 Parathion 0.1 89.7 ± 5.6 AO inhibitors Menadione 1.0 56.6 ± 4.2 Methadone 0.1 32.1 ± 2.5 SKF-525A (Profidane) 1 64.0 ± 3.8 Table 1 . ALDH specific activities in insecticide resistant PelRR and insecticide susceptible PelSS C. quinquefasciatus larvae. Assays were performed anaerobically and activities were estimated for pooled C. qu inque fasciatus PelRR and PelSS fourth instar larvae. Strain ALDH activity (nmolÆmin )1 Æmg protein )1 ) PelRR 1.24 ± 0.2 PelSS 1.15 ± 0.2 Fig. 6. AO enzymatic activity in different life stages of insecticide resistant C. quinquefasciatus PelRR. Upper panel: AO specific activity was measured in pooled crude homogenates of isogenic lines. Activity means were d etermin ed f or each of three indepe ndent i sogenic s amples at each time point. Results are means ± SD. Lower panel: native AO stained bands from equal loading of crude homogenate proteins of different life stages of an isogenic C. qu inqu efasc iatus PelRR line. Fig. 7. Ranges of AO activity from individual PelRR and PelSS larvae. Ranges of AO specific activity from individual C. quinque fasciat us PelRR and PelSS fourth instar larvae, respectively (n is the number of individuals tested). 776 M. Coleman et al. (Eur. J. Biochem. 269) Ó FEBS 2002 resistance. This may account for the selective advantage of insects carrying the esta2 1 /estb2 1 amplicon over those with other esterase containing amplicons. An AO gene has been cloned from the resistance- associated amplicon of the PelRR strain of C. quinquefas- ciatus. It contains the t hree conserved active site centres expected of an AO enzyme. This is the first reported AO sequence from any insect, hence absolute identification of the enzyme through DNA sequence homology alone is difficult [33]. However, the Culex enzyme clearly differs from XDH, a similar molybdenum containing enzyme in this family, as it lacks the NAD + domain which is essential for XDH activity. The predicted amino-acid sequence of the Culex ORF e ncodes fo r two [ 2Fe)2S] centres, a n NADH binding site and a molybdenum binding domain, which is consistent with the primary structure of AO from a range of species. There are several complete AO sequences on the database; within these there is a high degree of homology between human and bovine AO sequence, but little homology between these and Arabidopsis AO [9,32]. The Culex sequence has similar levels of homology with all three AOs. The lack of structural c onservation between AO of different species is suggested by Southern blots where bovine AO cDNA probes did not cross hybridize with Drosophila or to ad DNA, but did cross hybridize with lizard, chicken, mouse, rat and h uman DNA. AO may be less conserved than XDH, as a Southern blot using bovine XDH cross-hybridized to DNA from all the organisms above [9]. Identification of this enzyme as an AO is further supported by the gene structure. Three introns are con- served in all 31 insect XDHs recorded to date and a fourth is conserved in all but one insect species. There are three narrowly distributed novel introns, one in the medfly and two in the Willistroni group of Drosophila, one of which is shared by a second Drosophila group [35]. All of these introns, along with numerous others, are found in the genomic DNA encoding mammalian X DHs and AOs. Of the five in trons in the Culex AO genomic DNA, introns 1 and 5 occur in all known XDH and AO sequences. Introns 3 and 4 occur in all mammalian AO and XDH sequences but in no insect XDH sequence, and intron 2 is novel to this Culex AO sequence (Fig. 8). TheamplifiedAOoccursonthecommonesta2 1 /estb2 1 amplicon, but does not occur on either of the two estb1 amplicons in the TemR or COL Culex strains [3]. We have previously shown a ladder of truncated AO bands in the COL strain [3], and the current study shows TemR has no amplified AO sequence, further confirming the differences in the amplicons of these two strains despite both having an amplified estb 1 [18,36]. To influence fitness of mosquitoes carrying the amplicon, and p lay a role in insecticide detoxification, the amplified AO needs to be expressed. Multiple allelic variants of the esterase genes occur in Culex, making it easy to identify transcripts from the esterases on the insecticide re sistance-associated amplicons. PCR analysis was undertaken to see whether a similar level of allelic variation occurred in the AO locus. Comparison of the AO from resistant PelRR and susceptible PelSS strains of Culex show that there is allelic variation at this locus. The amplified AO diverges significantly from its nonamplified counterparts a t its 5¢ end, although all allelic variants are highly conserved at their 3¢ ends in line with other known AO sequences [32,34]. The variability be tween the nonamplified and a mplified alleles of AO from PelSS and PelRR, coupled with the complete homology between the PelRR genomic exons and the cDNA sequence from PelRR, suggests that the AO cDNA cloned from PelRR was transcribed from the amplicon and not from an un-amplified AO elsewhere in the genome. The strength of PCR product in PelRR also suggests that the rates of AO expression are higher in this strain than in PelSS. The normal physiological role of AO in mosquitoes is, a s yet, unknown, hence it i s difficult to p redict what effect if any the over-expression of this enzyme w ould have on the fitness of the mosquito carrying the AO containing ampli- con. Histochemical studies on the patterns of AO content in imaginal wing discs in Minute mutants o f hybrids of D. melanogaster [33] an d hybrids of M. domestica [15] suggest that it plays a role in larval development. In male moths AO assists in the catabolism of pheromones for location of female moths [10,12]. In some f emale moths their response to aldehydes in plant material is mediated by AO [12]. To further characterize the effe cts of the amplified AO, we analysed AO activity in resistant PelRR and s usceptible PelSS m osquitoes. The active site of AO includes an extended lipophilic active site [37], which can accommodate the diphenyl ring systems of methadone and SKF-525A. SKF-525A (Profidane) and its related analogues and methadone are potent inhibitors of rat liver cytsolic AO [38]. Methadone was a potent inhibitor o f mosquito AO activity, whilst SKF-525A was a poorer inhibitor. The absence of clear trends in structural requirements for substrates of AO, has been attributed to the flexibility of its substrate binding sites and its multiple productive orienta- tion [39]. This makes AO an effective detoxifying enzyme for a broad range of substrates in higher vertebrates. The broad substrate specificity of AO, including the oxidative or reductive metabolism of a wide variety of nitrogen or sulfur containing heterocyclic xenobiotics has been well docu- mented [6,7]. This suggests that insecticides and their metabolites may make good substrates for AO. AO from higher vertebrates is involved in the oxidative metabolism of neurotoxins [40], substituted quinazolines and pthalazines [41], chinhona antimalarials [42], purines and their ana- logues [39], quinoloinium cations a nd quinines [43,44], Fig. 8. Schematic diagram of the intron positions of PelRR AOscom- pared to AOs and XDHs from a range of species. Introns at positions A andGarecommontoPelRRAOandinsectXDHs. Intron* is novel to the Culex AO, while the four remaining introns are at positions in common with mammalian AO and XDH . Ó FEBS 2002 AO from C. quinquefasciatus (Eur. J. Biochem. 269) 777 [...]... liver: specificity toward purines and their analogs Arch Biochem Biophys 251, 36–46 40 Yoshihara, S & Ohta, S (1998) Involvement of hepatic aldehyde oxidase in conversion of 1-methyl-4-phenyl-2,3-dihydropyridinium (MPDP+) to 1-methyl-4-phenyl-5,6-dihydro-2-pyridone Arch Biochem Biophys 360, 93–98 41 Beedham, C., Critchley, D.J & Rance, D.J (1995) Substrate specificity of human liver aldehyde oxidase toward... Antennalspecific pheromone-degrading aldehyde oxidases from the moths Antheraea polyphemus and Bombyx mori J Biol Chem 265, 19712–19715 Tasayco, M.L & Prestwich, G.D (1990) Aldehyde oxidases and dehydrogenases in antennae of five moth species Insect Biochem 20, 691–700 Sprey, T.E & Kuhn, D.T (1987) The regulation of aldehyde oxidase in imaginal wing discs of Drosophila hybrids: evidence for cis- and trans-acting... in the nitroreductase activity based metabolic pathways of major insecticides, such as the conversion of parathion to aminoparathion, which is 100–300-fold less toxic than the parent compound [53] The partial inhibition of Culex AO activity by a range of insecticides, including 0.1 mM parathion, suggests that it may recognize this insecticide as a substrate The ability of AO to oxidize a range of xenobiotics... catalysed by aldehyde oxidase Biochem J 220, 67–74 44 Knox, W.E (1946) The quinne-oxidizing enzyme and liver aldehyde oxidase J Biol Chem 163, 699–711 45 Fowler, A.M., Chissick, H.H., Frearson, M & Wilson, K (1995) The role of aldehyde oxidase in the in vivo metabolism of benzothiazole Biochem Soc Trans 23, S604 46 Sugihara, K., Kitamura, S & Tatsumi, K (1995) Strain differences of liver aldehyde oxidase. .. S.H.P.P., Jayasuryia, H.T.R., Kalpage, K.S.P & Hemingway, J (2000) Insecticides and mosquito-borne disease Nature 407, 961–962 Gasarasi, D.B (2000) The transmission dynamics of bancroftian filariasis: the distribution of the infective larvae of Wuchereria bancrofti Culex quinquefasciatus and Anopheles gambiae and its effect on parasite escape from the vector Trans Roy Soc Lond Trop Med Hyg 94, 341–347... Biophys 242, 213–224 49 Yoshihara, S & Tatsumi, K (1986) Kinetic and inhibition studies on reduction of diphenyl sulfoxide by guinea pig liver aldehyde oxidase Arch Biochem Biophys 249, 8–14 50 Wolpert, M.K., Althaus, J.R & Johns, D.G (1973) Nitroreductase activity of mammalian liver aldehyde oxidase J Pharmacol Exp Ther 185, 202–213 51 Kitamura, S & Tatsumi, K (1984) Involvement of liver aldehyde oxidase. .. distribution of bovine liver aldehyde oxidase J Biol Chem 270, 31037–31045 van Zijll Langhout, B.W & Sprey, T.E (1988) Minute mutations of Drosophila melanogaster change aldehyde oxidase and pyridoxal oxidase distribution patterns in imaginal wing discs Dev Genet 9, 167–180 Sekimoto, H., Seo, M., Dohmae, N., Takio, K., Kamiya, Y & Koshiba, T (1998) Cloning and molecular characterization of plant aldehyde oxidase. .. Garattini, W (1998) Isolation and characterization of the human aldehyde oxidase gene: conservation of intron/exon boundaries with the xanthine oxidoreductase gene indicates a common origin Biochem J 332, 383–393 Rybczynski, R., Reagan, J & Lerner, M.R (1989) A pheromonedegrading aldehyde oxidase in the antennae of the moth Manduca sexta Neuroscience 9, 1341–1353 Rybczynski, R., Vogt, R.G & Lerner, M.R... thiocarbamate, phosphorothioate, phosphoramide and other heterocyclic pesticides However, the role of oxidative enzymes, such as AO, in the metabolism of these insecticides has not been investigated to date Inhibition studies on PelRR larvae suggested that many of these pesticides may interact with mosquito AO AO is involved in the reductive metabolism of a broad variety of nitro or sulfur compounds that can act... C quinquefasciatus (Eur J Biochem 269) 779 37 Beedham, C., Bruce, S.E., Critchley, D.J., Al-Tayib, Y & Rance, D.J (1987) Species variation in hepatic aldehyde oxidase activity Eur J Drug Metab Pharmacok 12, 307–310 38 Robertson, I.G.C & Gamege, R.S.K.A (1994) Methadone: a potent inhibitor of rat liver aldehyde oxidase Biochem Pharm 47, 584–587 39 Hall, W.W & Krenitsky, T.A (1986) Aldehyde oxidase from . characterization of the amplified aldehyde oxidase from insecticide resistant Culex quinquefasciatus Michael Coleman, John G. Vontas and Janet Hemingway Liverpool. m M of a ‘neat’ m ixture of aldehyde substrates (1 : 1, v/v, acetaldehyde/benzaldehyde) was added. AO a ctivity was determined by measuring the rate of