Desaturases from the Spotted Fireworm Moth (Choristoneura Parallela) Shed Light on the Evolutionary Origins of Novel Moth Sex Pheromone Desaturases

Gene 342 (2004) 303–311 www.elsevier.com/locate/gene

Desaturases from the spotted fireworm moth (Choristoneura parallela) shed light on the evolutionary origins of novel moth sex pheromone desaturases

Weitian Liua, Alejandro P. Rooneyb, Bingye Xuea, Wendell L. Roelofsa,*

aDepartment of Entomology, Cornell University, Geneva, NY 14456, USA bNational Center for Agricultural Utilization Research, Agricultural Research Service, Peoria, IL 61604, USA

Received 9 April 2004; received in revised form 6 July 2004; accepted 19 August 2004

Abstract

Six acyl-CoA desaturase-encoding cDNAs from mRNA isolated from the spotted fireworm moth, Choristoneura parallela (Lepidoptera: Tortricidae) were characterized and assayed for functionality. The expression levels of these cDNAs were determined in the pheromone gland and fat body by real-time PCR and the resulting patterns are in line with results from published studies on other moth sex pheromone desaturases. The cDNAs were found to correspond to six genes. Using both biochemical and phylogenetic analyses, four of these were found to belong to previously characterized desaturase functional groups [the D10,11, the D9 (16N18) and the D9 (18N16) groups]. A desaturase highly expressed in the pheromone gland was a novel E11 desaturase that was specific to 14-carbon precursor acids. The fifth gene [CpaZ9(14–26)] was found to display a novel Z9 activity indicating that it belongs to a new D9 functional group, whereas the sixth gene was determined to be nonfunctional with respect to desaturase activity. In accordance with previous studies, we find that desaturases of the D10,11 and D14 groups, which are the fastest evolving desaturases and possess the novel pheromone biosynthetic function, are expressed primarily in the pheromone gland whereas all other desaturases, which do not possess the novel reproductive function, evolve more slowly and display the ancestral metabolic function and pattern of gene expression. D 2004 Elsevier B.V. All rights reserved.

Keywords: Phylogeny; Evolution; Yeast functional assay; Cloning; mRNA; RT-PCR

1. Introduction displayed by these enzymes (Knipple et al., 1998; Liu et al., 1999, 2002a,b; Rosenfield et al., 2001; Hao et al., Acyl-CoA desaturases are enzymes that catalyze the 2002a,b; Roelofs et al., 2002; Jeong et al., 2003). These introduction of a double bond into long-chain fatty acids. enzymes are responsible for the production of distinct, These enzymes function in regular lipid metabolism and desaturated fatty acids that form a blend of pheromones also play a role in the cellular response to cold. The unique to each species. For example, the main sex- desaturases expressed in the female sex-pheromone gland pheromone component of Choristoneura parallela, the of moth species have evolved to play the key role in the spotted fireworm moth (SFW), is (E)-11-tetradecenyl biosynthesis of female sex pheromones as a result of the acetate (E11-14:OAc) (Neal et al., 1982; Polavarapu and amazing diversity of substrate, regio- and stereospecificities Lonergan, 1998), whereas the closely related Choristo- neura rosaceana possesses a 95:5 Z/E11-14:OAc pheromone blend (Hill and Roelofs, 1979). These pheromone Abbreviations: mRNA, messenger RNA; ORF, open reading frame; blends have evolved a dual purpose to serve as chemical PCR, polymerase chain reaction; RACE, rapid amplification of cDNA ends; RT-PCR, real-time PCR; DMDS, dimethyl disulfide. beacons for mate attraction and as species recognition * Corresponding author. Tel.: +1 315 787 2321; fax: +1 315 787 2326. signals, which guarantee that interspecific hybridization E-mail address: [email protected] (W.L. Roelofs). does not occur.

In light of the complex biochemistry of moth sex- Two degenerate primers PR1 and PR2, which were pheromone desaturases, it is no surprise that the genes in described previously (Liu et al., 1999, 2002a,b; Hao et al., which they are encoded have undergone complex patterns of 2002a,b)(Table 1), were used to amplify the central regions evolution. The genes are encoded by a multigene family that of the desaturase genes from fat-body cDNA library and appears to have originated several hundred million years another two degenerate primers, PR3 and PR4 (Table 1), ago during the early evolutionary history of insects (Roelofs were designed to amplify the central regions of desaturase et al., 2002). Varying amounts of gene duplication and loss genes from the pheromone-gland cDNA library. The as a result of evolution under the birth-and-death process primers were designed from conserved cDNA sequences have produced an assortment of desaturase genes, many of encoding previously characterized moth desaturases. All the which are unique to insect lineages other than moths central regions were ligated to PCRR4-TOPO vector for (Roelofs and Rooney, 2003). Some of these genes have sequencing and all sequences were aligned with other diverged in function and have become specialized for central regions of known desaturase genes to verify their reproduction (i.e., the E11 desaturases of moths), whereas homology to known desaturase genes. Based on the others have retained an ancestral metabolic function (i.e., the Z9 desaturases common to all insects) (Roelofs and Table 1 Rooney, 2003). Primers used for RACE PCR On the basis of their phylogenetic relationships and Degenerate primers enzymatic activities, Roelofs et al. (2002) classified the PR1: 5V-ATYACHGCCGGKKMYCAYMG-3V moth desaturases into four functional groups: D9(16N18), PR2: 5V-GGRAABDYGTGRTGGWAGTT-3V D9(18N16), D10,11 and D14. Biochemical and genetic PR3: 5V-GGYATYACVGCHGGNGCWCA-3V characterization of newly discovered desaturases would PR4: 5V-TGRTARTTRTGGAABSCYTCNCC-3V Primers for CpaZ9(16N18) and -Z9a(16N18) add to the diversity of the specificities already characterized SFW-A1: 5V-GGGACTAGACCTGAGCGACCTGTA-3V and would also provide valuable information on the amino SFW-A2: 5V-ACAAACCCTACGACAAGTCCATCAAG-3V acid changes needed to generate their unique enzymatic SFW-A3: 5V-CTTGTGGGCGGCGGAGTTGACGAG-3V specificities. More importantly, such information could SFW-A4: 5V-CAGCCGATGTGGGAGAAGAAGAAGC-3V provide important insights into the evolution of novel SFW-A5: 5V-CGCGTCTAGAATGGCACCTAACGTAACGGAAGAA-3V SFW-A6: function in moth sex-pheromone desaturases. Therefore, 5V-GCGGAGCTCTTAATCATCTTTAGGGTTAATTCTTATAG-3V we have cloned six desaturase genes from SFW, including a Primers for CpaZ9(18N16) new gene that encodes a Z9 desaturase that has an unusual SFW-B1: 5V-GCAGTTTCGCGACATCCATTAAC-3V broad range of substrates from C14 to C26, and studied their SFW-B2: 5V-GTTGTTCCGAGGAAGAAGGGCTAC-3V expression patterns in vitro. SFW-B3: 5V-ACTTCTAGAATGCCTCCTCAAGGGCAATCACCTCCATCC-3V SFW-B4: 5V-TGTGAGCTCTCATTCAGTCTTTTCAGGGTTGATGATAGT-3V 2. Materials and methods Primers for CpaZ9(14–26) SFW-C1: 5V-AAAAGCTTGCCGCTCTCTCGTACCAAAG-3V 2.1. Collection of insect tissue and RNA extraction SFW-C2: 5V-GACAGGGAGGGTCAAGCTTTGGACG-3V SFW-C3: 5V-AGATCTAGAATGACGACTGGCACTTCCCTGGCACTC-3V Female SFW were obtained from pupae shipped by Dr. SFW-C4:5V-ATATCTAGAATGACGACTGGCACTTCCCTGG-3V Sridhar Polavarapu, Rutgers University, USA. Fat bodies Primers for CpaE11 and pheromone glands were carefully dissected from 2- to SFW-D1: 5V-GAAAAAGCACCCCGAAGTCAAGAA-3V 3-day-old female moths and stored at À80 8C. Poly A+ SFW-D2: 5V-GCTACAGGCCCTACGACAAAAACA-3V SFW-D3: 5V-AATGGGCTGCGCTGTTTACAAGGAATA-3V RNA (mRNA) was isolated and purified from fat bodies and SFW-D4: 5V-ACTATATTTGTGGTGCATTCTGTGG-3V pheromone glands by using a mRNA Isolation Kit SFW-D5: 5V-ATATCTAGAATGGCGCCAAATGTAGAAGATATG-3V (Ambion) according to the procedures recommended by SFW-D6: 5V-ATTGAGCTCTTATTGCAACACTTCACTAGACTCG-3V the manufacturer. SFW-D7: 5V-ATAGAGCTCAAAATGGCGCCAAATGTAGAAGATATG-3V SFW-D8: 5V-ATTTCTAGATTATTGCAACACTTCACTAGACTCG-3V 2.2. cDNA cloning Primers for CpaNF SFW-E1: 5V-GTACACCATAAATTTGCCGACACC-3V Using a GeneRacerk Kit (Invitrogen), 1 Ag of mRNA SFW-E2: 5V-TCGGACATAGATGGGGAAACAGG-3V isolated from different SFW tissues (pheromone glands or SFW-E3: 5V-CCCAGAAGTACATCGGGACGACCG-3V fat bodies) was dephosphorylated with calf intestinal SFW-E4: 5V-CTTGACTAGCTTATGTTTGAATGGAACGAC-3V SFW-E5: 5V-TATTCTAGAATGGGGCTTTCAGAGGGTGCGGAAG-3V phosphatase, decapped with tobacco acid pyrophosphatase, SFW-E6: ligated with GeneRacerk RNA oligo, and reverse-tran- 5V-AAATCTAGACTATACATTTTTAGTATACAAATATTTAACTC-3V scribed with GeneRacerk Oligo dT Primer (homologus to SFW-E7: 5V-ATAGGTACCATGGGGCTTTCAGAGGGTGCGGAAG-3V the GeneRacerk RNA Oligo). The resulting single-strand SFW-E8: cDNA was used as library (template) for gene cloning. 5V-CGAGGTACCCTATACATTTTTAGTATACAAATATTTAACTC-3V W. Liu et al. / Gene 342 (2004) 303–311 305 sequence information of the central regions, gene-specific SYBR Green I, 1.25 U Hot Start Ext Taq, 2.5 Al10Âbuffer, primers (Table 1) were designed for RACE PCR. Two full- 5 Al template DNA and water. In every run, two negative length cDNA sequences were isolated from the fat-body controls (water), eight standard cDNA (four concentrations, cDNA library and six full-length cDNA sequences were two times for each concentration), and six tissue samples isolated from the pheromone-gland cDNA library. (three for gland cDNA library, three for fat-body cDNA library) were included. The following protocol was used for 2.3. Functional assay the RT-PCR: pre-incubation of 2 min at 95 8C followed by 50 cycles with 15 s at 95 8C, 20 s at 58–65 8C (optimum Gene-specific primers (Table 1) were designed to amplify temperature used for different primers), and 20 s at 72 8C. the ORFs of all the cloned genes. All the ORFs were The reaction was heated above the melting temperature of digested and ligated with linearized YEpOLEX expression the dimers for each gene for an additional 10 s and vector as described previously (Knipple et al., 1998; Liu et fluorescence measurements were made at the end of the al., 1999; Hao et al., 2002a,b). The consensus clone of the additional extension phase. The threshold line was set at 30 recombinant plasmids was transformed to mutant yeast ole1 fluorescent units on the Y-axis and the primary curve was cells (strain L8-14C) (Stukey et al., 1989) for functional selected in the curve analysis. expression as described previously (Knipple et al., 1998; A specific primer pair (Table 2) was used to determine Liu et al., 1999; Hao et al., 2002a,b) for analysis of the the transcription level for each of the six desaturase genes of unsaturated fatty acid products by mass spectral analysis of SFW. Melting curve analyses were performed immediately the DMDS adducts. following PCR to identify specific PCR products and primer Four clones, which produced Z9 unsaturated products, dimers as the temperature slowly increased from 60 to 95 were expressed well in the YEpOLEX system, but the other 8C. The reactions were also checked with agarose gel two clones (from the pheromone-gland cDNA library) were electrophoresis to correlate product length with melting not expressed in this system. Gene-specific primers (Table peaks and to determine if primer dimers or nonspecific 1) for the uncharacterized clones were designed to amplify amplification products were formed. Analyses of data were the ORFs for construction in the pYES2 expression system accomplished using the Smart Cycler Software (Cepheid). (Invitrogen). The pYES2 vectors containing each consensus Relative copy numbers of unknown samples were deter- ORF of the remaining two clones were transformed to mined according to a standard curve equation (Table 2). Saccharomyces cerevisiae strain Elo1 competent cells (Toke and Martin, 1996) with the methods described previously 2.5. Phylogenetic analyses (Liu et al., 2002b) for the functional assay. The newly characterized nucleotide sequences for the 2.4. Quantitative RT-PCR SFW acyl-CoA desaturase genes described in this paper were compared to previously published moth and dipteran The fat-body and pheromone-gland cDNA libraries were sequences extracted from GenBank. The accession numbers constructed as described in Materials and methods. The for these sequences are given in Fig. 5. The sequences were Cepheid’s Smart Cycler System, a commercially available aligned using their deduced amino acid sequences with the RT-PCR system, was used for amplification of target computer program ClustalX (Thompson et al., 1997). The cDNAs. Reactions were carried out in a total volume of resulting alignment was subsequently checked for errors by 25 Al/tube. Each reaction master mixture contained 3 mM visual inspection. The computer program MEGA2 (Kumar MgCl2, 0.1 AM of each primer, 0.2 mM dNTP’s, 0.5 Al10Â et al., 2001) was used to construct phylogenetic trees using

Table 2 Real-time PCR primers and standard graph formulae Name of genes Real-time PCR Real-time PCR primers Formula of standard graph product size (bp) CpaZ9(16 N18) 446 5V-ACCTGTACTCGGATCCCATTCTTC-3V y=À0.353x+14.576, 5V-CGTCCGTTTCACCCTTTGCTG-3V r2=0.901 CpaZ9a(16N18) 382 5V-ACGAAATAGATAACTGTCCTGAAAG-3V y=À0.282x+11.867, 5V-ATGGTGTTGAAGATAGTAAGAATGAG-3V r2=0.9 CpaZ9(18N16) 512 5V-GTTCGGACCCTGTGCTGAGA-3V y=À0.316x+9.61, 5V-GATAGTTGCCGCCGCTTTCTG-3V r2=0.977 CpaZ9(14–26) 402 5V-TGCCGATCCTCTGTTTCCTGCTAC-3V y=À0.247x+8.16, 5V-TCCCATCTCCCGTTCTCGTTACTC-3V r2=0.995 CpaE11 395 5V-ATAGCAATCCTGTCTTGCGGTTTC-3V y=À0.252x+9.354, 5V-CTTCAAGTCGTAAGCCCAGCCAATC-3V r2=0.962 CpaNF 296 5V-TCGGACATAGATGGGGAAACAGGCCTTAT-3V y=À0.261x+9.35, 5V-GTCAACATCACCCCAGCCCCAAAGATTCCT-3V r2=0.997 306 W. Liu et al. / Gene 342 (2004) 303–311 the neighbor-joining method (Saitou and Nei, 1987) with length cDNA sequences were obtained by combining the 1500 bootstrap replicates as a measure of statistical central region with 3V- and 5V-RACE fragments that were reliability. amplified by RACE PCR. The two ORFs obtained from both libraries had deduced amino acid sequences of 352 (ORF-I) and 353 (ORF-II). The four ORFs for the pheromone-gland 3. Results genes had deduced amino acid sequences of 383 (ORF-III), 371 (ORF-IV), 334 (ORF-V), and 328 (ORF-VI). 3.1. Cloning of the desaturase genes 3.2. Functional assay of Z9 desaturases Two fragments in both the fat-body and the pheromone- gland cDNA libraries, and four additional fragments in the Consensus clones of the YEpOLEX recombinant plas- pheromone-gland cDNA library were found with the mids were transformed to desaturase-deficient mutant ole1 degenerate primers designed for desaturase genes. Full- yeast cells for functional expression. The yeast grew well

Fig. 1. Comparison of deduced amino acid sequences of the six desaturase genes identified in SFW. Residues identical to the top sequence are represented by dots (.); gaps are indicated by dashes (-). W. Liu et al. / Gene 342 (2004) 303–311 307 without addition of unsaturated fatty acids for four ORFs (I– IV). ORF-I from both fat-body and pheromone-gland tissue was found to be a Z9 desaturase and produced Z9-16:Acid in greater amounts than Z9-18:Acid (5:2, respectively) and was named CpaZ9(16N18). ORF-IV was found to be identical to ORF-I in sequence and function, except for a string encoding 19 amino acids in the 5V end, and was named CpaZ9a(16N18) (Fig. 1). ORF-II produced less Z9-16:Acid than Z9-18:Acid (1:6) (Fig. 2), similar to those of other moth species in this more rapidly evolving class of Z9 desaturases (Roelofs and Rooney, 2003), which are most commonly present in the pheromone gland, except for a number of noctuid species in which it is the main Z9 desaturase in the fat body. This one was named CpaZ9(18N16).

Fig. 3. GC/MS analyses of fatty acid methyl esters (DMDS adducts) produced in the pYES2 functional expression system using elo1 cells with the consensus ORF clone of CpaE11.

ORF-III proved to be an unusual Z9 desaturase in that it produced a long series of Z9 unsaturated acids of chain- lengths C14, C16, C18, C20, C22, C24 and C26 (Fig. 2). The predominant products were Z9-16:Acid and Z9- 18:Acid in a 7:1 ratio. This gene was named CpaZ9(14–26).

3.3. Functional assay of E11 desaturase

The consensus clone of ORF-V in a pYES2 recombinant plasmid was transformed to elo1 yeast cells for functional expression because these cells lack the chain-elongation enzymes between C14 and C16 products (Toke and Martin, 1996). GC/MS analysis of the DMDS derivatives of the fatty acid methyl esters obtained by methanolysis of the yeast lipid extracts showed production of only E11-14:Acid with available precursors C14, C16 and C18 acids (Fig. 3). This desaturase, which is named CpaE11 and produces only E11- 14:Acid, has amino acid identity with desaturases characterized in moth species producing D11-14:Acids as follows: 92% to CroZ/E11, which produces a 7:1 mixture of the Z/ E11-14:Acids in the closely related species, C. rosaceana (Hao et al., 2002a); 80% to AveZ/E11 which produces a 3:2 mixture of the Z/E11-14:Acids in Argyrotaenia velutinana (Liu et al., 2002b); 76% to EpoE11, which produces E11- 14:Acid, E9,E11-14:Acid, and E11-16:Acid in Epiphyas postvittana (Liu et al., 2002a); and 63% to OnuZ/E11, which produces a 5:4 mixture of Z/E11-14:Acids and Z11-16:Acid Fig. 2. SIM (217 m/z) GC/MS analyses of Z9-unsaturated fatty acid methyl esters (DMDS adducts) produced in the YEpOLEX functional expression in Ostrinia nubilalis (Roelofs et al., 2002). system using ole1 cells with consensus ORF clones of (A) Cpa(18N16) and A diunsaturated product, Z9,E11-14:Acid, also is pro- (B) Cpa(14–26). duced (Fig. 3) with the yeast expression system by action of 308 W. Liu et al. / Gene 342 (2004) 303–311 the Z9 desaturase of the yeast acting on the product, E11- 14:Acid, of CpaE11. The diene was characterized by the retention time of a DMDS product containing the typical 285 ion produced by a 14-carbon methyl ester adduct of a conjugated diene, and by the isomeric specific retention time of a 4-methyl-1,2,4-triazoline-3,5-dione derivative (McElfresh and Millar, 1999) identified by its characteristic 194 ion.

3.4. Functional assay of non-functional desaturase

Attempts were made to show desaturase activity for the consensus clone of ORF-VI in both the YEpOLEX and pYES2 expression systems. Various saturated and mono- unsaturated fatty acids from C14 to C20 were added to or present in the yeast, but none resulted in an unsaturated product from the inserted ORF-VI, and so the nonfunctional protein was named CpaNF.

3.5. Expression of desaturases in fat-body and pheromone- gland tissue

Quantitative RT-PCR was conducted to amplify all six desaturase genes from the pheromone-gland and fat-body cDNA libraries generated from tissue taken from 2- to 3- day-old female moths. The data (Fig. 4) show that the key pheromone desaturase, CpaE11, is ca. 10 times more abundant in the adult pheromone gland than all the other desaturases combined. The most abundant Z9 desaturase in the fat body is CpaZ9(16N18), which produces more C16 than C18 product and has the same ORF as CpaZ9a(16N18), except the latter has additional nucleotides encoding 19 amino acids in the 5V end. The unusual Z9 desaturase, CpaZ9(14–26), which produces a range of Z9 products from C14 to C26, is found above trace levels in the pheromone Fig. 4. Quantification of desaturase genes in 2–3-day-old female SFW gland. tissues by RT-PCR.

3.6. Phylogenetic analyses stitutes a new functional group, which we propose to call bD9 (14–26)Q. The phylogenetic position of the SFW genes characterized in this study in relation to other previously characterized moth sex pheromone genes is shown in 4. Discussion Fig. 5. All of these genes, except one, cluster within the functional groups previously identified by Roelofs et al. In this study, we found six new desaturase genes in (2002) [i.e., the D9 (16N18), D9 (18N16), D10,11 and SFW. Five of these genes were found to be closely related D14 functional groups] on the basis of previous bio- to previously characterized desaturases that have been chemical and phylogenetic analyses. These five genes assigned to functional groups D9 (16N18), D9 (18N16) and cluster in the D9 (16N18), D9 (18N16), D10,11 groups. D10,11. Our findings support a previous suggestion that Accordingly, the biochemical activities of the SFW computational and phylogenetic analyses can be used to desaturases follow what would be predicted on the basis predict the functions of genes that have been assigned as of the clustering pattern observed here (Fig. 5). However, putative desaturases (Roelofs and Rooney, 2003). As such, one gene that did not cluster within the previously we suggest here that if a desaturase clusters within one of identified functional groups is the novel Z9 desaturase, the functional groups shown in Fig. 5 with 90% or greater CpaZ9(14–26), which does not appear to be closely bootstrap support, it should be considered a member of related to any of the desaturases that have been fully that group and that it is highly likely to possess desaturase characterized biochemically (Fig. 5). Therefore, it con- activity. However, we caution that biochemical character- W. Liu et al. / Gene 342 (2004) 303–311 309 ization is still important, particularly in the case of genes nonfunctional with respect to desaturase activity. This gene is falling within the D10,11 group, as they can possess highly similar to a nonfunctional desaturase in C. rosaceana enzymatic activities with slight differences in their regio- (Fig. 5), and together they form a clade that is related to the and stereospecificities. D10,11 group. It has been suggested that nonfunctional The Z9(16N18) and Z9a(16N18) sequences are identical desaturase genes might represent either recent pseudogenes with one another except that the latter has an additional 19 or genes that are epigenetically repressed or possibly down- amino acid residues (excluding the start codon) at the 5V regulated for some reason yet to be determined (Roelofs and end. Although it is possible that these could represent recent Rooney, 2003). Since it has been shown (Shanklin et al., gene duplicates, it is more likely that they represent 1997; Broadwater et al., 2002; Shanklin and Whittle, 2003) alternative splice forms. In fact, the primary Z9 gene that acyl-coA desaturases share structural similarities to (CG5887) of Drosophila melanogaster has several alter- omega-hydroxylases, the possibility that nonfunctional desa- native splice forms. However, a genomic organization study turase genes represent functional hydroxylases is an intrigu- on the C. parallela Z9(16N18) gene is ultimately required to ing hypothesis worthy of future investigation. resolve this question. Regardless, is it possible that nonfunctional desaturases In addition, there are certain desaturase genes that lack any could act as a warehouse of sorts, from which novel detectable desaturase activity in vitro. We found such a gene desaturases could evolve and produce new pheromone in SFW, which we have named CpaNF to reflect that it is blends? Such a scenario has already been identified in

Fig. 5. Phylogeny of the desaturase genes of various moth and dipteran species. The tree was reconstructed using Jones et al. (1992) amino acid distances. Numbers along branches indicate bootstrap support from 1500 replicates. The GenBank sequence accession numbers are given after the species abbreviation, which consists of the first letter of the genus name and the first two letters of the species name followed by the biochemical activity if known (Roelofs and Rooney, 2003). In the case of D. melanogaster and Anopheles gambiae, the gene names are given as they are listed in the genome databases for those species. 310 W. Liu et al. / Gene 342 (2004) 303–311

Ostrinia furnacalis, in which a non-functional desaturase Sridhar Polavarapu for supplying SFW pupae, Kathy Poole gene was shown to have evolved into the key gene in the and Callie Musto for supplying the adult moths for this production of female sex pheromones in O. furnacalis study, Sepp Kohlwein and Charles Martin for supplying the (Roelofs et al., 2002). This gene has not been characterized elo1 and ole1 mutant yeast strains, respectively, Marion in other moth species and is more closely related to O’Connor and Hongmei Jiao for technical assistance. and dipteran genes of unknown function than to other moth Joanna Warder for preparing figures. Sequence data from desaturase genes. Surprisingly, in the present study we this article have been deposited in the GenBank database found a novel D9 desaturase [CpaZ9(14–26)] that produces under accession no. AF518020 [CpaZ9(16N18)], AF518021 a series of Z9 unsaturated fatty acids, including C14, C16, [CpaZ9(16N18)a], AF518010 [CpaZ9(18N16)], AF518008 C18, C20, C22, C24 and C26 acids. Like the Ostrinia D14 [CpaZ9(14–26)], AF518014 (CpaE11) and AF518012 gene, this gene is also more closely related to certain (CpaNF). dipteran desaturases than to other moth desaturases (Fig. 5). Currently, we do not know the purpose of this gene, although it is clearly functional. It is tempting to speculate that the gene might somehow impact reproduction in C. References parallela, as it is expressed in levels above trace amounts in the pheromone gland of this species. However, evidence Broadwater, J.A., Whittle, E., Shanklin, J., 2002. Desaturation and linking this gene to sex pheromone biosynthesis has yet to hydroxylation. J. Biol. Chem. 277, 15613–15620. Hao, G., O’Connor, M., Liu, W., Roelofs, W.L., 2002a. Characterization of be found. Z/E11- and Z9-desaturases from the obliquebanded leafroller moth, This study and others (Knipple et al., 1998; Liu et al., Choristoneura rosaceana. J. Insect Sci. 2:26. insectscience.org/2.26. 1999, 2002a,b; Rosenfield et al., 2001; Hao et al., Hao, G., Liu, W., O’Connor, M., Roelofs, W.L., 2002b. Characterization 2002a,b; Roelofs et al., 2002; Jeong et al., 2003) show and functional expression of cDNAs encoding acyl-CoA Z9- and that genes in the D10,11 group are expressed primarily in Z10-desaturase of Planotortrix octo. Insect Biochem. Mol. Biol. 32, 961–966. the pheromone gland whereas all other moth desaturases Hill, A.S., Roelofs, W.L., 1979. Sex pheromone components of the are primarily expressed in the fat body. The only gene obliquebanded leafroller, Choristoneura rosaceana. J. Chem. Ecol. 5, known to be an exception is the D14 gene of O. 3–11. furnacalis, which is also expressed in the pheromone Jeong, S.E., Rosenfield, C.L., Marsella-Herrick, P., You, K.M., Knipple, gland. However, O. furnacalis lacks a functional D10,11 D.C., 2003. Multiple acyl-CoA desaturase-encoding transcripts in pheromone glands of Helicoverpa assulta, the oriental tobacco gene, so the D14 gene presumably has been recruited to budworm. Insect Biochem. Mol. Biol. 33, 609–622. replace it. Interestingly, Roelofs et al. (2002) showed that Jones, D.T., Taylor, W.R., Thornton, J.M., 1992. The rapid generation of the D14 and D10,11 group, which represent sex pher- mutation data matrices from protein sequences. Comput. Appl. Biosci. omone desaturases, evolve significantly faster at the 8, 275–282. protein level than the other desaturase groups and are less Knipple, D.C., Miller, S.J., Rosenfield, C.-L., Liu, W., Tang, J., Ma, P.W.K., Roelofs, W.L., 1998. Cloning and functional expression of a constrained by purifying selection. Thus, desaturase genes cDNA encoding a pheromone gland-specific acyl-CoA D11-desaturase possessing the novel reproductive function and pattern of of the cabbage looper moth, Trichoplusia ni. Proc. Natl. Acad. Sci. gene expression evolve faster than the desaturases that U. S. A. 95, 15287–15292. display the ancestral function and pattern of gene Kumar, S., Tamura, K., Jakobsen, I.B., Nei, M., 2001. MEGA2: expression. This indicates that changes at the amino acid molecular evolutionary genetics analysis software. Bioinformatics 17, 1244–1245. level are tied to changes in both expression pattern and Liu, W., Ma, P.W.K., Marsella-Herrick, P., Rosenfield, C.-L., Knipple, function in moth desaturases. D.C., Roelofs, W.L., 1999. Cloning and functional expression of a While we do not know for certain if the moth sex cDNA encoding a metabolic acyl-CoA D9-desaturase of the cabbage pheromone desaturases evolved completely free of selective looper moth, Trichoplusia ni. Insect Biochem. Mol. Biol. 29, 435–443. constraints subsequent to their duplication, they certainly Liu, W., Jiao, H., O’Connor, M., Roelofs, W.L., 2002a. Gene characterized for membrane desaturase that produces (E)-11 isomers of mono- and are less constrained and evolve faster than other moth diunsaturated fatty acids. Proc. Natl. Acad. Sci. U. S. A. 99, 620–624. desaturases that display the ancestral metabolic function Liu, W., Jiao, H., O’Connor, M., Roelofs, W.L., 2002b. Moth desaturase (Roelofs and Rooney, 2003). These observations are characterized that produces both Z and E isomers of D11-tetradecenoic consistent with expectations under Ohno’s (1970) hypoth- acids. Insect Biochem. Mol. Biol. 32, 1489–1495. esis that duplicate genes experiencing relaxed selection McElfresh, J.S., Millar, J.S., 1999. Sex pheromone of the common sheep moth, Hemileuca eglanterina, from the San Gabriel mountains of pressure often acquire novel functions that become fixed in California. J. Chem. Ecol. 25, 687–709. the genome. Neal Jr., J.W., Klun, J.A., Bierl-Leonhardt, B.A., Schwarz, M., 1982. Female sex pheromone of Choristoneura parallela (Lepidoptera: Tortricidae). Environ. Entomol. 11, 893–896. Acknowledgements Ohno, S., 1970. Evolution by Gene Duplication. Springer-Verlag. Polavarapu, S., Lonergan, G., 1998. Sex pheromone of Choristoneura parallela (Lepidoptera: Tortricidae): components and development of a This study was supported by the National Science pheromone lure for population monitoring. Environ. Entomol. 27, Foundation (NSF #IBN-9870669). The authors thank 1242–1249. W. Liu et al. / Gene 342 (2004) 303–311 311

Roelofs, W.L., Rooney, A.P., 2003. Molecular genetics and evolution of Shanklin, J., Achim, J., Schmidt, H., Fox, B.G., Munck, E., 1997. pheromone biosynthesis in Lepidoptera. Proc. Natl. Acad. Sci. U. S. A. Mfssbauer studies of alkane w-hydroxylase: evidence for a diiron 100, 9179–9184. cluster in an integral-membrane enzyme. Proc. Natl. Acad. Sci. U. S. A. Roelofs, W.L., Liu, W., Hao, G., Jiao, H., Rooney, A.P., Linn Jr., C.E., 94, 2981–2986. 2002. Evolution of moth sex pheromones via ancestral genes. Proc. Stukey, J.E., McDonough, V.M., Martin, C.E., 1989. Isolation and Natl. Acad. Sci. U. S. A. 99, 13621–13626. characterization of OLE1, a gene affecting fatty acid desaturation from Rosenfield, C.L., You, K.M., Marsella-Herrick, P., Roelofs, W.L., Knipple, Saccharomyces cerevisiae. J. Biol. Chem. 264, 16537–16544. D.C., 2001. Structural and functional conservation and divergence Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., among acyl-CoA desaturases of two noctuid species, the corn earworm, 1997. The CLUSTAL_X windows interface: flexible strategies for Helicoverpa zea and the cabbage looper, Trichoplusia ni. Insect multiple sequence alignment aided by quality analysis tools. Nucleic Biochem. Mol. Biol. 31, 949–964. Acids Res. 25, 4876–4882. Saitou, N., Nei, M., 1987. The neighbor-joining method: a new method for Toke, D.A, Martin, C.E., 1996. Isolation and characterization of a gene reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. affecting fatty acid elongation in Saccharomyces cerevisiae. J. Biol. Shanklin, J., Whittle, E., 2003. Evidence linking the Pseudomonas Chem. 271, 18413–18420. oleovorans alkane w-hydroxylase, an integral membrane diiron enzyme, and the fatty acid desaturase family. FEBS Lett. 545, 188–192.