Portal de Periódicos da CAPES

Mosquito Carboxylesterase Estα21 (A2).

1995; Elsevier BV; Volume: 270; Issue: 28 Linguagem: Inglês

10.1074/jbc.270.28.17044

1083-351X

Ashley M. Vaughan, Janet Hemingway,

Computational Drug Discovery Methods

Organophosphorus insecticide resistance in Culex mosquitoes is commonly caused by increased activity of one or more esterases. The commonest phenotype involves elevation of the esterases Estα2 (A2) and Estβ2 (B2). A cDNA encoding the Estα2 esterase has now been isolated from a Sri Lankan insecticide-resistant mosquito (Culex quinquefasciatus, Say) expression library. In line with a recently suggested nomenclature system (Karunaratne, S. H. P. P.(1994) Characterization of Multiple Variants of Carboxylesterases Which Are Involved in Insecticide Resistance in the Mosquito Culex quinquefasciatus. Ph.D. thesis, University of London), as the first sequenced variant of this esterase, it is now referred to as Estα21. The full-length cDNA of estα21 codes for a 540-amino acid protein, which has high homology with other esterases and lipases and belongs to the serine or B-esterase enzyme family. The predicted secondary structure of Estα21 is similar to the consensus secondary structure of proteins within the esterase/lipase family where the secondary and tertiary structures have been resolved. The level of identity (∼47% at the amino acid level) between the estα21 and the various Culex estβ (B1 and B2) cDNA alleles that have been cloned and sequenced suggests that the two esterase loci are closely related and arose originally from duplication of a common ancestral gene. The lack of a distinct hydrophobic signal sequence for Estα21 and two possible N-linked glycosylation sites, both situated close to the active site serine, suggest that it is a nonglycosylated protein that is not exported from the cell. Southern and dot blot analysis of genomic DNA from various insecticide-resistant and susceptible mosquito strains show that the estα21 gene, like estβ21, is amplified in resistant strains. The restriction fragment length polymorphism patterns, after probing Southern blots of EcoRI-digested genomic DNA with estα21 cDNA, show that the amplified and nonamplified estα alleles differ in the resistant and susceptible Sri Lankan mosquitoes. Organophosphorus insecticide resistance in Culex mosquitoes is commonly caused by increased activity of one or more esterases. The commonest phenotype involves elevation of the esterases Estα2 (A2) and Estβ2 (B2). A cDNA encoding the Estα2 esterase has now been isolated from a Sri Lankan insecticide-resistant mosquito (Culex quinquefasciatus, Say) expression library. In line with a recently suggested nomenclature system (Karunaratne, S. H. P. P.(1994) Characterization of Multiple Variants of Carboxylesterases Which Are Involved in Insecticide Resistance in the Mosquito Culex quinquefasciatus. Ph.D. thesis, University of London), as the first sequenced variant of this esterase, it is now referred to as Estα21. The full-length cDNA of estα21 codes for a 540-amino acid protein, which has high homology with other esterases and lipases and belongs to the serine or B-esterase enzyme family. The predicted secondary structure of Estα21 is similar to the consensus secondary structure of proteins within the esterase/lipase family where the secondary and tertiary structures have been resolved. The level of identity (∼47% at the amino acid level) between the estα21 and the various Culex estβ (B1 and B2) cDNA alleles that have been cloned and sequenced suggests that the two esterase loci are closely related and arose originally from duplication of a common ancestral gene. The lack of a distinct hydrophobic signal sequence for Estα21 and two possible N-linked glycosylation sites, both situated close to the active site serine, suggest that it is a nonglycosylated protein that is not exported from the cell. Southern and dot blot analysis of genomic DNA from various insecticide-resistant and susceptible mosquito strains show that the estα21 gene, like estβ21, is amplified in resistant strains. The restriction fragment length polymorphism patterns, after probing Southern blots of EcoRI-digested genomic DNA with estα21 cDNA, show that the amplified and nonamplified estα alleles differ in the resistant and susceptible Sri Lankan mosquitoes. Insecticide resistance is a significant problem in many insect pests. Elevation of carboxylesterase activity is the major mechanism of resistance to the organophosphorus insecticides in a wide range of insect species(1Chang C.K. Jordan T.W. Pestic. Biochem. Physiol. 1983; 19: 190-195Crossref Scopus (17) Google Scholar, 2Herath P.R.J. Hemingway J. Weerasinghe I.S. Jayawardena K.G.I. Pestic. Biochem. Physiol. 1987; 29: 157-162Crossref Scopus (33) Google Scholar, 3Kao L.R. Motoyama N. Dauterman W.C. Pestic. Biochem. Physiol. 1985; 23: 228-239Crossref Scopus (30) Google Scholar, 4Malkenson N.C.D. Wood E.J. Zerba E.N. Insect Biochem. 1984; 4: 481-486Crossref Scopus (20) Google Scholar, 5Matsumura F. Sakai K. J. Econ. Entomol. 1968; 61: 598-605Crossref Google Scholar, 6Motoyama N. Kao L.R. Lin P.T. Dauterman W.C. Pestic. Biochem. Physiol. 1984; 21: 139-147Crossref Scopus (36) Google Scholar, 7Parker A.G. Russell R.J. Delves A.C. Oakeshott J.G. Pestic. Biochem. Physiol. 1991; 41: 305-318Crossref Scopus (37) Google Scholar). However, the resistance mechanism has been studied in depth at the biochemical and/or molecular level in only a few species. In the mosquito Culex quinquefasciatus and the aphid Myzus persicae, amplification of at least one esterase gene underlies the increase in esterase activity, with up to 250 copies of an esterase gene per genome being recorded in Culex (8-10). The esterases involved in both species are standard B or serine esterases according to the classification of Aldridge(11Aldridge W.N. Biochem. J. 1953; 53: 110-117Crossref PubMed Scopus (491) Google Scholar). This is a widely distributed family of enzymes, which hydrolyze carboxylester, amide, and thioester bonds in a variety of compounds. The mosquito C. quinquefasciatus is a major biting nuisance insect worldwide, and a vector of filarial and viral disease on at least two continents. Esterase banding patterns after polyacrylamide or starch gel electrophoresis of crude homogenates of this mosquito can be complex(12Callaghan A. Hemingway J. Malcolm C.A. Biochem. Genet. 1993; 31: 459-472Crossref PubMed Scopus (4) Google Scholar, 13Georghiou G.P. Pasteur N. J. Econ. Entomol. 1978; 71: 201-205Crossref PubMed Scopus (85) Google Scholar). Esterases can be classified on the basis of their biochemical, molecular, immunological, and electrophoretic characteristics, although their broad substrate specificities and multi-allelic nature make classification difficult. A classification system based on the esterase's preference for the general esterase substrates α- or β-naphthyl acetate and their relative electrophoretic mobilities has been suggested(14Raymond M. Pasteur N. Georghiou G.P. Mellon R.B. Wirth M.C. Hawley M. J. Med. Entomol. 1987; 24: 24-27Crossref PubMed Scopus (56) Google Scholar). On the basis of this classification, two predominant elevated esterase phenotypes, B1 and co-elevated A2/B2, which cause insecticide resistance, can be distinguished electrophoretically in the mosquito. The A2/B2 phenotype is by far the most common, occurring in all continents where C. quinquefasciatus is found. The elevated A2 and B2 esterases are in complete linkage disequilibrium in mosquito populations, i.e. the two esterases are co-elevated in all mosquitoes. Amplification of the B1 esterase was first shown in the Californian Tem-R strain of mosquito(9Mouches C. Pasteur N. Berge J.B. Hyrien O. Raymond M. De Saint Vincent B.R. De Silvestri M. Georghiou G.P. Science. 1986; 233: 778-780Crossref PubMed Scopus (279) Google Scholar). Immunological and molecular studies have since shown that B2 and B1 were originally alleles of the same locus and that the B2 gene is also amplified in resistant insects(10Raymond M. Beyssat-Arnaouty V. Sivasubramanian N. Mouches C. Georghiou G.P. Pasteur N. Biochem. Genet. 1989; 27: 417-423Crossref PubMed Scopus (78) Google Scholar, 15Mouches C. Magnin M. Berge J.B. De Silvestri M. Beyssat V. Pasteur N. Georghiou G.P. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 2113-2116Crossref PubMed Scopus (87) Google Scholar, 16Mouches C. Pauplin Y. Agarwal M. Lemieux L. Herzog M. Abadon M. Beyssat-Arnaouty V. Hyrien O. De Saint Vincent B.R. Georghiou G.P. Pasteur N. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 2574-2578Crossref PubMed Scopus (127) Google Scholar, 17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar). Initial reports suggested that there was no cross-reaction between the A and B esterases at either an immunological or molecular level, and it was concluded that if these genes had arisen from duplication of a common ancestral gene, this must undoubtedly have occurred a long time ago(15Mouches C. Magnin M. Berge J.B. De Silvestri M. Beyssat V. Pasteur N. Georghiou G.P. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 2113-2116Crossref PubMed Scopus (87) Google Scholar, 18Fournier D. Bride J.-M. Mouches C. Raymond M. Magnin M. Berge J.-B. Pasteur N. Georghiou G.P. Pestic. Biochem. Physiol. 1987; 27: 211-217Crossref Scopus (57) Google Scholar). However, we have recently shown that a polyclonal antiserum raised to the nondenatured A2 esterase does cross-react with the B2, although the specificity is ∼50-fold lower for B2 than A2, suggesting that the two enzymes share some common epitopes(19Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Ketterman A.J. Biochem. Soc. Trans. 1993; 22: 127SCrossref Scopus (2) Google Scholar, 20Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Pestic. Biochem. Physiol. 1995; 51 (in press)Google Scholar). We here report the original cloning and cDNA sequence of the A2 mosquito esterase from an insecticide-resistant strain of C. quinquefasciatus from Peliyagoda, Sri Lanka(21Peiris H.T.R. Hemingway J. Bull. Entomol. Res. 1993; 83: 127-132Crossref Scopus (25) Google Scholar). The cDNA sequence is compared with the other mosquito and aphid esterases that are involved in insecticide resistance. We also show that the underlying mechanism of A2 elevation in the Pel RR insecticide-resistant strain is gene amplification. The Culex esterase nomenclature is currently out of line with that used for other organisms, and there is confusion between the earlier general esterase classification (11Aldridge W.N. Biochem. J. 1953; 53: 110-117Crossref PubMed Scopus (491) Google Scholar) and that subsequently proposed for mosquitoes(14Raymond M. Pasteur N. Georghiou G.P. Mellon R.B. Wirth M.C. Hawley M. J. Med. Entomol. 1987; 24: 24-27Crossref PubMed Scopus (56) Google Scholar). Karunaratne (22Karunaratne S.H.P.P. Characterization of Multiple Variants of Carboxylesterases Which Are Involved in Insecticide Resistance in the Mosquito Culex quinquefasciatus. Ph.D. thesis, University of London. 1994; Google Scholar) has proposed that the Culex esterases should be renamed using α and β rather than the A and B of Raymond et al.(14Raymond M. Pasteur N. Georghiou G.P. Mellon R.B. Wirth M.C. Hawley M. J. Med. Entomol. 1987; 24: 24-27Crossref PubMed Scopus (56) Google Scholar), with numerical superscripts denoting sequence variants with the same electrophoretic mobility. Hence the two amplified B1s, which differ in their inferred amino acid sequence but are electrophoretically identical, become Estβ11 and Estβ12, from the Californian and Cuban strains of Culex, respectively, and the nonelevated B from the Sri Lankan Pel SS strain, again with the same electrophoretic mobility, becomes Estβ13. This system ensures that esterases can still be named initially on the basis of their electrophoretic mobilities, but distinct DNA sequence variants can also be indicated as this level of data becomes available. As this nomenclature system will be adopted in this paper, a summary of the past and proposed nomenclature for mosquito strains where esterase DNA sequence data is available is given in Table I. On the basis of this classification(22Karunaratne S.H.P.P. Characterization of Multiple Variants of Carboxylesterases Which Are Involved in Insecticide Resistance in the Mosquito Culex quinquefasciatus. Ph.D. thesis, University of London. 1994; Google Scholar), the Pel RR A2 esterase now becomes Estα21, as the first available sequence of this esterase.Table I:Suggested nomenclatures for the esterases of the mosquito Culex quinquefasciatus Open table in a new tab A heterogeneous population (Pel) of mosquito was collected from Peliyagoda, Sri Lanka, in 1986. It was selected to give an insecticide-susceptible strain, Pel SS, and a resistant strain, Pel RR (23, 24). Pel RR was 31-fold more resistant to the organophosphorus insecticide temephos than Pel SS(25Peiris H.T.R. Hemingway J. Bull. Entomol. Res. 1990; 80: 49-55Crossref Scopus (45) Google Scholar). The Pel SS strain was obtained by multiple single family selection from the Pel strain(23Amin A.M. Peiris H.T.R. Med. Vet. Entomol. 1990; 4: 269-273Crossref PubMed Scopus (22) Google Scholar). The Pel RR strain was obtained by mass selection with temephos(24Peiris H.T.R. Hemingway J. Bull. Entomol. Res. 1990; 80: 453-457Crossref Scopus (29) Google Scholar). Since then, insecticide resistance in this strain has been maintained by exposing fourth instar larvae every third generation to the LD50 concentration of temephos. The only organophosphate resistance mechanism in this strain is co-elevation of the Estα21 and Estβ21 esterases. A polyclonal antiserum was prepared by injecting a New Zealand White rabbit with purified nondenatured Estα21 from Pel RR. The esterase was administered after mixing with reconstituted Ribi adjuvant to give a final enzyme concentration of 200 μg/ml. Four injections of 1 ml of antigen were given at 2-week intervals. Each injection was split between intramuscular, intradermal, and subcutaneous sites(26Jayawardena K.G.I. Purification and Characterization of Two Carboxylesterases from Organophosphate Resistant Culex quinquefasciatus from Sri Lanka. Ph.D. thesis, University of London. 1992; Google Scholar). The resultant antiserum had a high level of specificity to Estα2 and much lower (∼50-fold) levels of sensitivity to the Estβ2 and Estβ1 esterases and the organophosphate target site acetylcholinesterase(19Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Ketterman A.J. Biochem. Soc. Trans. 1993; 22: 127SCrossref Scopus (2) Google Scholar, 20Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Pestic. Biochem. Physiol. 1995; 51 (in press)Google Scholar). This antiserum was used to screen a Pel RR cDNA expression library. The cDNA for library construction was synthesized from 5 μg of mRNA from fourth instar Pel RR larvae by a method previously used(17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar). The cDNA was blunted and size-selected (>400 base pairs) on a Sephacryl S-400 column (Pharmacia Biotech Inc.). After EcoRI linker addition (Promega) and subsequent phosphorylation, the cDNA was ligated into Lambda ZAP II (Stratagene). This vector allows plasmid containing cDNAs to be directly excised from the phage library. The bacteriophage arms were supplied digested with EcoRI and dephosphorylated. 5 μg of arms were ligated to 125 ng of cDNA and packaged with Packagene extracts (Promega). Approximately 200,000 recombinant clones were obtained. This was carried out with the Protoblot Immunoscreening System (Promega). The 200,000 unamplified recombinant clones were screened with the polyclonal anti-Estα21 esterase antisera at a dilution of 1:10,000. Positive clones were purified by successive rounds of screening and in vivo excised to give recombinant pBluescript plasmids. Sequencing was carried out with Sequenase Version 2.0 (Amersham Corp.) using universal primers complemented with the ExoIII/Mung Bean Nuclease deletion method (Stratagene) and with primers specific to the estα21 clone. This allowed the sequencing of both strands of the insert cDNA from the plasmid. A modified 5′-RACE procedure as performed previously (17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar) was used to isolate the full-length estα2 1The abbreviations used are: RACErapid amplification of cDNA endsPCRpolymerase chain reactionkbkilobase. cDNA. The reverse primer used in the PCR was 5′-ACCGTACATCTCCACTCC-3′ and was close to the 5′-end of the cDNA isolated from the cDNA library. The PCR product was subcloned into the pBluescript T vector(17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar, 27Marchuk D. Drumm M. Saulino A. Collins F.S. Nucleic Acids Res. 1991; 19: 1154Crossref PubMed Scopus (1130) Google Scholar). Two separate PCR products were sequenced in both directions. rapid amplification of cDNA ends polymerase chain reaction kilobase. A Pel RR estα21 cDNA fragment was used as a probe to determine the haplotype of the Estα esterases from the Pel RR and Pel SS strains. Genomic DNA was isolated from fourth instar larvae as described previously(17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar). 10 μg of genomic DNA was digested to completion with EcoRI and separated by gel electrophoresis through 0.8% (w/v) agarose. The DNA was transferred to charged nylon membranes (Amersham Corp.) and hybridized with a 32P-labeled Estα21 cDNA probe (specific activity > 2 × 106 cpm/μg) at 65°C for 16 h in hybridization buffer (5 × Denhardt's solution, 6 × SSC, 0.1% (w/v) SDS, 0.1% (w/v) sodium pyrophosphate, 5% (w/v) PEG 8000, and 100 μg/ml boiled, sheared herring sperm DNA). The final washes were at 65°C in 0.1 × SSC and 0.1% (w/v) SDS for 20 min. Similarity searches of protein data bases with the the estα21 cDNA sequence were undertaken using the B17049 and MPsearch programs available through NCBI (National Institutes of Health) and EMBL (Heidelberg), respectively. Family structure determination was undertaken using the Prosite section of the Motif finder program through the Motif E-mail server on Genome. Prediction of secondary structure was undertaken using the PHD program(28Rost B. Sander C. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 7558-7562Crossref PubMed Scopus (466) Google Scholar). Sequence alignments were undertaken using the MegAlign program of the LASERGENE package (DNASTAR). Similarity indexes were calculated using the FASTA algorithm, via the GeneMan program of LASERGENE. Four positive plaques were obtained from the initial screening of the 200,000 recombinant clones of the unamplified Pel RR cDNA library. The number of positives obtained from this screening suggested that the target sequence was initially present at a higher frequency than that expected for a single copy gene. The four plaques were purified, in vivo excised, and partially sequenced with M13 forward and reverse primers. The sequence for all four clones over 400 nucleotides was identical. The insert from one of the plasmids (pBlueAV.A2) was completely sequenced in both directions. The cDNA sequence had an open reading frame at its 5′-end, and terminated in a stop codon, 3′-untranslated region, and a poly(A) tail. There was no start methionine codon (AUG), so a modified 5′-RACE procedure (17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar) was used to isolate the 5′-end of the cDNA. The sequence of the two subcloned 5′-RACE PCR products analyzed was identical in both directions and also overlapped exactly with the previously sequenced partial length cDNA clone. The full-length cDNA (made up of the insert from pBlueAV.A2 and the 5′-RACE PCR product) had an open reading frame of 1623 base pairs complete with an AUG start codon and coded for a protein of 540 amino acids. This is in the expected range, since the purified monomeric protein has an estimated molecular mass of 58-67 kDa estimated by SDS-PAGE, native PAGE, and Sephacryl S200 chromatography(29Ketterman A.J. Jayawardena K.G.I. Hemingway J. Biochem. J. 1992; 287: 355-360Crossref PubMed Scopus (51) Google Scholar). Fig. 1 shows the full-length estα21 cDNA nucleotide sequence and the proposed amino acid sequence. Fig. 2 shows the predicted amino acid sequence of Estα21 and its alignment with the Culex Estβ21 (Pel RR B2) and aphid E4 amplified esterases. The Estα21 sequence, like the Estβ21, contained nine cysteine residues, but only three of these are conserved between the two esterases. The triad of precisely located active site amino acids, Ser, Glu, and His, of the serine esterase family were present in the Estα21 sequence at positions 190, 324, and 445, respectively.Figure 2:Alignment of the amino acid sequence of Culex mosquito serine Estα21 and Estβ21 esterases with the aphid E4 esterase involved in insecticide resistance. Possible glycosylation sites are underlined. Conserved amino acids in a large range of esterases (30) are in boldface.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The estα21 had a similarity index of 49.2 with estβ21, which was the highest similarity in the protein database. The top 30 alignments of estα21 undertaken with both the B17049 and MPsearch programs were all esterases or lipases. The protein was also unambiguously assigned to the carboxylesterase B-1 family of B type serine esterases using the Motif program. Possible N-linked glycosylation sites, conforming to the sequence NXT or NXS, where X is not proline are underlined in Fig. 2. There are only two possible glycosylation sites in estα21; both are very close to the active site serine and are not shared by other esterases in the alignment. Since the estα21 is not proceeded by a signal sequence, and as the purified mature protein is not retained by Con A chromatography(22Karunaratne S.H.P.P. Characterization of Multiple Variants of Carboxylesterases Which Are Involved in Insecticide Resistance in the Mosquito Culex quinquefasciatus. Ph.D. thesis, University of London. 1994; Google Scholar), it is probable that the protein is not glycosylated. The predicted secondary structure of Estα21 obtained using the PHD program (28Rost B. Sander C. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 7558-7562Crossref PubMed Scopus (466) Google Scholar) is given in Fig. 3 compared with the known consensus secondary structure of the Torpedo californica acetylcholinesterase and Geotrichum candidum lipase(30Cygler M. Schrag J.D. Sussman J.L. Harel M. Silman I. Gentry M.K. Doctor B.P. Protein Sci. 1993; 2: 366-382Crossref PubMed Scopus (539) Google Scholar). The predicted secondary structures of the Estα21 and Estβ21 proteins, determined independently, were identical with the exception of the final α-helix, which was not predicted in Estβ21. The predicted secondary structure is remarkably similar to that of the known consensus sequence for serine esterases and lipases, differing only in three α-helices. One of these, the helix between β-sheets 8 and 9 is also present in the G. candidum lipase structure but is not present in T. californica acetylcholinesterase (30). The estα21 cDNA was used as a probe for Southern blot analysis of EcoRI restriction digests of equal amounts of genomic DNA from the insecticide-susceptible (Pel SS) and resistant (Pel RR) mosquito strains. After hybridization and high stringency washing, the probe bound to a single 7.4-kb band of Pel SS DNA (Fig. 4), which demonstrates the existence of an estα gene with high homology to estα21 in this strain. A 7.5-kb band was found in the resistant strain at an equally low intensity, suggesting that the resistant strain still carries a nonamplified estα allele. The Pel RR strain contained a 5.8-kb band with a high signal intensity, which was not present in the Pel SS strain. The higher signal intensity in the resistant strain compared with the susceptible, implies that gene amplification is the underlying mechanism of the Estα21-associated resistance. Further proof of this was obtained by undertaking dot blots with the estα21 cDNA probe, using genomic DNA from four other resistant strains of C. quinquefasciatus from different geographical locations with elevated Estα2 activity. All the resistant strains gave a much higher intensity of signal than the susceptible strain (Pel SS), and the signals obtained were similar when the blots were reprobed with an estβ21 cDNA. This suggests that the amplification levels of the estα2 and estβ2 genes are similar in the resistant strains. In this paper a new Culex esterase nomenclature system (17) has been adopted, as the molecular data now accumulating is making the old system unworkable. For example, a nonamplified B esterase from a susceptible Culex strain, which has an identical electrophoretic mobility to B1, has been shown to be distinct from the B1s from TemR and MRES, which are in turn distinct from each other, and from esterase B2 at both a kinetic and amino acid level(17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar).2 2S. H. P. P. Karunaratne, K. G. I. Jayawardena, A. Vaughan, and J. Hemingway, submitted for publication. Thus on the basis of restriction fragment length polymorphism (RFLP) patterns, DNA sequences, inferred amino acid sequences, and kinetic interactions of these enzymes at least one nonamplified and two amplified "B1" esterases occur(17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar).3 3G. J. Small, S. H. P. P. Karunaratne, and J. Hemingway, submitted for publication. This level of variability of electrophoretically identical esterases makes application of the earlier classification difficult, and the problem has been compounded by electrophoretically similar esterases being given different numerical values on the basis of their distinct RFLP patterns(31Poire M. Raymond M. Pasteur N. Biochem. Genet. 1992; 30: 13-26Crossref PubMed Scopus (86) Google Scholar). The nomenclature system we have used allows preliminary assignment of the esterase by electrophoresis, with an extension when sequence data is available, while the old system means esterases may need complete reclassification once sequence or RFLP data is available. With the new classification the esterase that we have now cloned, which was previously referred to as A2, is classified as Estα21 on the basis of its electrophoretic mobility and sequence. The estα21 esterase cDNA from the insecticide-resistant Pel RR strain of the C. quinquefasciatus mosquito has now been cloned and sequenced. The high number of positive clones obtained from the initial screening of the cDNA library suggested either that there was increased transcription of the gene or that gene amplification was the underlying mechanism of increased esterase activity. Southern blot analysis showed that the resistant strain had a unique, amplified 5.8-kb EcoRI RFLP when compared with the susceptible strain, which had a single unamplified band at 7.4 kb. Amplification of the estα2 gene was also shown in other C. quinquefasciatus strains with elevated Estα2 and Estβ2 activity, using dot blot analysis of genomic DNA. Since amplification of estβ2 is already well documented(17Vaughan A. Rodriguez M. Hemingway J. Biochem. J. 1995; 305: 651-658Crossref PubMed Scopus (54) Google Scholar), the underlying genetic mechanism for organophosphorus insecticide resistance in mosquitoes with the elevated Estα2/Estβ2 activity phenotype is therefore amplification of both esterase genes. The different RFLP pattern for estα in the Pel RR and Pel SS strains demonstrates the existence of two distinct estα alleles. The amplified and nonamplified Estα and Estβ esterases have been purified and characterized physically and kinetically from a range of insecticide-resistant strains and a susceptible strain of mosquito(28Rost B. Sander C. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 7558-7562Crossref PubMed Scopus (466) Google Scholar, 32Jayawardena K.G.I. Karunaratne S.H.P.P. Ketterman A.J. Hemingway J. Bull. Entomol. Res. 1994; 84: 39-44Crossref Scopus (21) Google Scholar, 33Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Ketterman A.J. Biochem. J. 1993; 294: 575-579Crossref PubMed Scopus (75) Google Scholar, 34Ketterman A.J. Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Pestic. Biochem. Physiol. 1993; 47: 142-148Crossref Scopus (32) Google Scholar). 2S. H. P. P. Karunaratne, K. G. I. Jayawardena, A. Vaughan, and J. Hemingway, submitted for publication., 2S. H. P. P. Karunaratne, K. G. I. Jayawardena, A. Vaughan, and J. Hemingway, submitted for publication. The Estα enzymes from the susceptible (Pel SS) and resistant (Pel RR) strains are kinetically and electrophoretically distinct from each other,2 suggesting that they are different alleles, which is supported by the different RFLP patterns seen in the two strains. The amplified Estα21, Estβ12, and Estβ21 esterases all have similar sizes (∼60 kDa) and PI values(29Ketterman A.J. Jayawardena K.G.I. Hemingway J. Biochem. J. 1992; 287: 355-360Crossref PubMed Scopus (51) Google Scholar, 33Karunaratne S.H.P.P. Jayawardena K.G.I. Hemingway J. Ketterman A.J. Biochem. J. 1993; 294: 575-579Crossref PubMed Scopus (75) Google Scholar).3 We now know that all three esterases are coded for by cDNAs with the same length open reading frame. The amplified esterases E4 and FE4 in insecticide-resistant aphids are comparable in function and size with the mosquito esterases, but unlike the Culex esterases these enzymes confer resistance to both organophosphates and pyrethroids(35Devonshire A.L. Biochem. J. 1977; 167: 675-683Crossref PubMed Scopus (162) Google Scholar, 36Devonshire A.L. Moores G.D. Pestic. Biochem. Physiol. 1982; 18: 235-246Crossref Scopus (377) Google Scholar, 37Devonshire A.L. Moores G.D. Reiner E. Aldridge W.N. Hoskin F.C.G. Enzymes Hydrolysing Organophosphorus Compounds. Ellis Horwood Chichester, UK1989: 181-192Google Scholar). The E4 esterase and the Estα21 and various Estβ Culex esterases all produce resistance to the organophosphates by sequestration, although the aphid and mosquito Estα and Estβ esterases share only 21.8 and 22.9% similarity, respectively, while the similarity of the mosquito esterases is 49.2%. Hence the same kinetic properties can clearly be conferred by a large number of different sequences. This is perhaps not surprising when the predicted secondary structures of these esterases are considered. The predicted secondary structures for the Estα21 and Estβ21 esterases differ only in the final element, and the structure bears a striking resemblance to actual secondary structures already resolved for two members of the serine esterase/lipase family(30Cygler M. Schrag J.D. Sussman J.L. Harel M. Silman I. Gentry M.K. Doctor B.P. Protein Sci. 1993; 2: 366-382Crossref PubMed Scopus (539) Google Scholar). The high level of amino acid identity of Estα21 and Estβ21 suggests that they originated from a common ancestor. The two genes probably arose through gene duplication and subsequently diversified. The Culex esterase cDNAs each code for proteins of 540 amino acids, and they have a higher sequence homology with each other than for any other sequence within the data banks. The two esterases also share a number of common features. For example, many carboxylesterases contain a short hydrophobic leader sequence, which initially directs protein sorting down a secretory pathway. All sequenced human liver carboxylesterases and the aphid E4 carboxylesterase contain signal sequences of between 17 and 23 amino acids(38Field L.M. Williamson M.S. Moores G.D. Devonshire A.L. Biochem. J. 1993; 294: 569-574Crossref PubMed Scopus (88) Google Scholar, 39Kroetz D.L. McBride O.W. Gonzalez F.J. Biochemistry. 1993; 32: 11606-11617Crossref PubMed Scopus (101) Google Scholar). The mosquito estα21 sequence and the estβ21 sequence are relatively unusual in that they contain no signal sequence, suggesting that neither esterase is exported from the cell. Neither of these purified esterases bind to Con A chromatography columns(22Karunaratne S.H.P.P. Characterization of Multiple Variants of Carboxylesterases Which Are Involved in Insecticide Resistance in the Mosquito Culex quinquefasciatus. Ph.D. thesis, University of London. 1994; Google Scholar), which suggests that the esterases are not glycosylated, in contrast to the aphid and some human esterases(39Kroetz D.L. McBride O.W. Gonzalez F.J. Biochemistry. 1993; 32: 11606-11617Crossref PubMed Scopus (101) Google Scholar, 40Devonshire A.L. Searle L.M. Moores G.D. Insect Biochem. 1986; 16: 659-665Crossref Scopus (25) Google Scholar). This is further supported by the sequence data, since there are only two possible N-linked glycosylation sites on the Estα21 esterase, which are situated only 4 and 31 amino acids from the active site serine. In many soluble carboxylesterases that are retained by the endoplasmic reticulum, the tetrapeptide KDEL (Lys-Asp-Glu-Leu) occurs at the carboxyl terminus(41Pelham H.R.B. Trends Biochem. Sci. 1990; 15: 483-486Abstract Full Text PDF PubMed Scopus (432) Google Scholar). Variants of the KDEL sequence that direct intracellular retention of proteins have since been identified, although it appears that the Glu-Leu is a major requirement(42Robbi M. Beaufay H. Biochem. Biophys. Res. Commun. 1992; 183: 836-841Crossref PubMed Scopus (16) Google Scholar). The four Culex estβ cDNAs sequenced to date all end in the sequence NDELF. However, the carboxyl terminus of Estα21 is KDKLY. Twenty-three amino acids are conserved through a series of 29 related proteins, and it was argued that these amino acids are essential for the structure (salt bridges, packing, and disulfide bridges) and function (active site) of the proteins(30Cygler M. Schrag J.D. Sussman J.L. Harel M. Silman I. Gentry M.K. Doctor B.P. Protein Sci. 1993; 2: 366-382Crossref PubMed Scopus (539) Google Scholar). Only 21 of these are conserved in the Estα21. A cysteine is one of the nonconserved residues, and it is notable that both the Estα21 and Estβ21 have nine cysteine residues, the majority of which are not conserved in any other serine esterase. The missing cysteine, at position 65 (Cys61 and Cys67 in T. californica AChE and G. candidum lipase, respectively) in the Pel RR Estα21 sequence, forms a disulfide bridge in T. californica acetylcholinesterase and G. candidum lipase with a cysteine at position 84 (Cys105 and Cys94), which is a serine in the Estα21 and various Estβ Culex esterases. In the alignment of Cygler et al.(30Cygler M. Schrag J.D. Sussman J.L. Harel M. Silman I. Gentry M.K. Doctor B.P. Protein Sci. 1993; 2: 366-382Crossref PubMed Scopus (539) Google Scholar), the only esterase in which the latter cysteine was not conserved was the Culex TemR Estβ11. The reason for the large number of cysteines in the Culex esterases is unknown, but it has been suggested that five of these may not be involved in disulfide bond formation, leaving them free to oxidize. This results in the development of evenly spaced satellite bands after native polyacrylamide gel electrophoresis of purified native esterases under buffer conditions that do not protect the thiol groups(43Jayawardena K.G.I. Hemingway J. Med. Vet. Entomol. 1995; 9Crossref PubMed Scopus (3) Google Scholar). Having shown that the mechanism of elevation of the Estα21 esterase is gene amplification, we now intend to determine the genomic structure of this esterase. The almost complete linkage disequilibrium in which these two elevated esterases occur, coupled with their sequence similarities, indicating that they arose through gene duplication, may also suggest that the two esterases are situated on the same amplification unit. This has, however, been contradicted by some classical genetic data on the inheritance patterns of these esterases,(12Callaghan A. Hemingway J. Malcolm C.A. Biochem. Genet. 1993; 31: 459-472Crossref PubMed Scopus (4) Google Scholar, 44Wirth M.C. Marquine M. Georghiou G.P. Pasteur N. J. Med. Entomol. 1990; 27: 202-206Crossref PubMed Scopus (57) Google Scholar), and the final determination of the physical location of the two esterases in relation to each other awaits further genomic studies.