Naturwissenschaften (2000) 87 : 450–454 Q Springer-Verlag 2000 SHORT COMMUNICATION Regine Gries 7 Gerhard Gries 7 Grigori Khaskin Skip King 7 Owen Olfert 7 Lori-Ann Kaminski Robert Lamb 7 Robb Bennett Sex pheromone of orange wheat blossom midge, Sitodiplosis mosellana Received: 2 March 2000 / Accepted in revised form: 18 August 2000 Abstract Pheromone extract of the female orange wheat blossom midge, Sitodiplosis mosellana (Géhin) (SM) (Diptera: Cecidomyiidae), was analyzed by coupled gas chromatographic-electroantennographic detection (GC-EAD) and GC-mass spectrometry (MS), employing fused silica columns coated with DB-5, DB210, DB-23 or SP-1000. These analyses revealed a single, EAD-active candidate pheromone which was identified as 2,7-nonanediyl dibutyrate. In experiments in wheat fields in Saskatchewan, traps baited with (2S,7S)2,7-nonanediyl dibutyrate attracted significant numbers of male SM. The presence of other stereoisomers did not adversely affect trap captures. Facile synthesis of stereoisomeric 2,7-nonanediyl dibutyrate will facilitate the development of pheromone-based monitoring or even control of SM populations. The orange wheat blossom midge, Sitodiplosis mosellana (Géhin) (SM) (Diptera: Cecidomyiidae), is found in most parts of the world where wheat is grown. Larvae feed on developing kernels, causing them to shrivel, crack, and deform. Damaged kernels are inferior in milling quality and germination capacity (Miller and R. Gries 7 G. Gries (Y) 7 G. Khaskin 7 S. King Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia V5A 1S6, Canada e-mail: gries 6sfu.ca O. Olfert 7 L.-A. Kaminski Agriculture and Agri-Food Canada, Saskatoon Research Centre, Saskatoon, Saskatchewan S7N OX2, Canada R. Lamb Agriculture and Agri-Food Canada, Cereal Research Centre, Winnipeg, Manitoba R3T 2M9, Canada R. Bennett Ministry of Forests, Tree Improvement Branch, Saanichton, British Columbia V8M 1W4, Canada Dedicated to Studiendirektor Joachim Grote in honor of his 80th birthday Halton 1961). Grain yield decreases exponentially with the increase in SM larval infestations (Olfert et al. 1985). SM larval infestations in Quebec are also correlated with the presence of wheat scab, Fusarium graminearum (Schwabe), and glume blotch, Septoria nordorum (Berk.), suggesting that SM adults vector fungal spores (Mongrain et al. 1997). SM outbreak populations cause substantial losses in revenue. For example, losses exceeded C$50 million in Manitoba and C$100 million in Saskatchewan in 1995 (Lamb 1998; Lamb et al. 1999). SM impact is alleviated mainly by insecticides (Elliott 1998a,b). Development of pheromone-based monitoring of SM populations would: (1) promote judicious rather than prophylactic use of insecticides; (2) optimize timing of insecticide treatments; and (3) save direct crop protection costs. Synthetic, micro-encapsulated pheromone may even be explored for control of SM populations. There is strong evidence that female SM produce a sex pheromone (Pivnick and Labbé 1992; Pivnick 1993). We report identification and field testing of the SM pheromone. Aliquots of 5–10 female equivalents (FE) of abdominal extract were analyzed by coupled gas chromatographic-electroantennographic detection (GC-EAD) (Arn et al. 1975), employing a Hewlett-Packard (HP) 5890 gas chromatograph fitted with a fused silica column (30 m!0.25 or 0.32 mm i.d.) coated with either DB-5, DB-210, DB-23 (J&W Scientific, Folsom, Calif.) or SP-1000 (Supelco, Bellefonte, Pa.). These analyses revealed a single, EAD-active candidate pheromone (CP) (Fig. 1A) not detectable by the flame ionization detector. Hydrogenation of pheromone gland extract (50 FE) followed by renewed GC-EAD analysis did not alter the antennal response to CP, indicating that it had no double bond(s). A very weak mass spectrum of CP in a highly concentrated extract ( 1 200 FE) revealed the fragmentation ion m/z 89, indicative of a butyl ester. However, inter-column differences in retention indices (e.g. DB-23 to DB-5 : 485; DB-5 to DB-210 : 480; DB-23 to DB-210 : 5; DB-5 to SP 1000 : 409) were about twice as large as those of monobutyl esters. 451 A E DETECT OR RESPONSE (mV) O SR+RS RR O FID: pheromone extract O CP EAD: UVD O 9 antenna 14 SS RS Chiral GC b FID 16 17 26 18 FID SR+RS RELATIVE INTENSITY (%) 300-(87+87+1) 71 100 43 C4H9O2 75 50 EAD 8 9 10 11 TIME (min) 95 125 82 55 25 c SS+RR 300 (MW)-87 B 29 min GC-EAD RETENTION TIME (min) 89 113 141 154 168 183 213 194 0 50 OH C6H5CH2Br NaH 1 D min RR ~0.3mV 15 C a SS Chiral HPLC 100 OBn 2 (C 3H7CO) 2O 10% Pd/C OH 3 4 OH 6 OCOC 3H7 H2 OH THF; –20°C 7 8 OH OH C5H5N; DMAP 9 O + OH 11 (C 3H7CO) 2O 12 11 13 OH O OCOC 3H7 S C5H5N; DMAP THF; –20°C OH OH H2O (0.55 equiv); 24h OH 2CH3MgBr; CuI OCOC 3H7 5 10 (0.4mol %); THF [O] CuI + BrMg 200 OH OBn 1. BuLi 2. O O 150 m/z 14 S OCOC 3H7 Fig. 1 (Legend see page 452) Taking into account: (1) that CP was likely a dibutyl ester with no more than nine carbon atoms in the molecule chain, and (2) that secondary tridecen-2-yl and tridecadien-2-yl acetates are known to attract cecidomyiid midges (Foster et al. 1991; Gries et al., unpub- lished results), we hypothesized that CP may be 2,7-octanediyl dibutyrate or 2,8-nonanediyl dibutyrate. The latter had retention indices on all analytical columns similar to those of CP. The correct molecular structure of CP was then approximated by synthesizing various 452 Fig. 1. A: Flame ionization detector (FID) and electroantennographic detector (EAD: male S. mosellana antenna) response to one equivalent of female S. mosellana pheromone extract. Chromatography: Hewlett Packard (HP) gas chromatograph (GC) 5890 equipped with a fused silica column (30 m!0.25 mm i.d.) coated with DB-5; splitless injection, temperature of injection port 240 7C and FID 250 7C; temperature program: 50 7C (1 min), 25 7C/min to 100 7C, then 10 7C to 280 7C. Retention indices (Van den Dool and Kratz 1963) of the candidate pheromone (CP) on columns coated with DB-5, DB-210, DB-23 and SP-1000 were 1863, 2343, 2348 and 2272, respectively. B: Mass spectrum of synthetic 2,7-nonanediyl dibutyrate (mosellin); fragmentation ions m/z 213 and 125 indicate the loss of 1 and 2 butyl ester group(s), respectively. Varian 2000 GC-mass spectrometer (MS). C: Scheme for the synthesis of 2,7-nonanediyl dibutyrate: racemic 4-pentyne-2-ol (1) was benzylated, resulting in benzyloxy acetylene (2), which was treated with butyl lithium at –70 7C, and then warmedup in the presence of butylene epoxide to give 2-O-benzyl-4-nonyne-2,7-diol (3; 55% yield). Hydrogenation of (3) in the presence of 10% Pd/C afforded 2,7-nonanediyl (4; 93% yield). Diol (4) was then esterified to 2,7-nonanediyl dibutyrate (5; mosellin; 98% yield), using butyric anhydride in the presence of pyridine and catalytic amounts of N,N-dimethylaminopyridine. D: Scheme for the synthesis of field-active 2(S),7(S)-nonanediyl dibutyrate (14): S-propylene oxide (6) was coupled at –20 7C with Grignard reagent (7) in the presence of 0.1 equivalents of a CuI catalyst. Resulting (2S)-8-octen-2-ol (8) was oxidized by m-chloroperoxybenzoic acid to the terminal epoxide (9). Hydrolytic kinetic resolution of (9) with Co(II) salen (Jacobsen’s) catalyst (R,R)-N,Nbbis(3,5 - di - tert - butylsalicylidene) - 1,2 - cyclohexanediaminocobalt(II) (10) (Tokunaga et al. 1997; Schaus et al. 1998) afforded 2(R),7(S)1,2-epoxy-7-hydroxyoctane (11), which was separated from unwanted 1,2(S),7(S)-octanetriol (12) by silica flash chromatography [hexane: ether (1 : 1), then ether as eluents]. Ring opening of (11) with excess of methylmagnesium bromide and catalytic amounts of CuI yielded 2(S),7(S)-nonanediol (13) which was esterified to 2(S),7(S)-nonanediyl dibutyrate [14 ; 77% enantiomeric excess (ee) as determined by chiral GC; 37% overall yield]. GC-MS and NMR data for (14) were consistent with those of (5). The remaining three stereoisomers of diester (5) were synthesized from (S)or (R)-propylene oxide and 5-bromo-1-pentene, again by hydrolytic kinetic resolution of terminal epoxide intermediates, with ee of end products ranging between 73 and 85%. E: Separation of stereoisomeric 2,7-nonanediyl dibutyrate by various chromatographic techniques: a High-performance liquid chromatography (HPLC): Waters LC 626 high-performance liquid chromatograph equipped with a Waters 486 variable wavelength UV visible detector (UVD) set to 200 nm, a Waters 746 Data Module and a Chiralpak AD column (250!4.6 mm i.d.); solvent system: hexane (99%) plus 2-propanol (1%) with a flow rate of 0.8 ml/min. b Gas chromatography (GC): HP5890 GC equipped with a custommade chiral fused silica column coated with a 1 : 1 mixture of heptakis-(2,6-di-O-methyl-3-O-pentyl)-ß-cyclodextrin and OV-1701 (König et al. 1992; Pietruszka et al. 1992); splitless injection, temperature of injection port and FID as in A; temperature program: 145 7C isothermal. c FID and EAD (male S. mosellana antenna) responses to synthetic stereoisomers of 2,7-nonanediyl dibutyrate; chromatography: HP5890 GC equipped with a fused silica column coated with DB-210, splitless injection, temperature of injection port and FID as in A; temperature program: 180 7C isothermal. dibutyl and diisobutyl esters, with one ester group conservatively kept in position C2, whereas the second ester group was placed in C6, C7 or C9. Synthetic 2,7-nonanediyl dibutyrate (Fig. 1C; compound 5) (here termed ‘mosellin’) and CP had identical retention and comparable EAD characteristics on all four analytical columns. Moreover, the mass spectrum of synthetic mosellin (Fig. 1B) greatly resembled the weak mass spectrum of CP, further confirming the structural assignment. GC, GC-EAD and HPLC analyses of stereoisomeric mosellin, or fractionated stereoisomers thereof, with chiral and DB-210 columns determined that the early- and/or late-eluting stereoisomer elicited a strong antennal response (Fig. 1Ec). Field experiments with HPLC-separated synthetic stereoisomers (Fig. 1Ea) were carried out in wheat fields, employing a complete randomized design. Intercept-W wing traps [IPM Technologies (Canada), Calgary, Alberta] were set up in a single line 5–10 m from, and paralleling the field’s margin. Traps were placed at 20 m intervals F 20 cm above ground and baited with Whatman No. 1 filter paper impregnated with HPLC fractionated mosellin stereoisomers (Fig. 1Ea). Experiment 1 was carried out near Saskatoon, Saskatchewan, Canada between 22 and 28 July 1999 in a field known to have a dense SM population. Experiments 2 and 3 were conducted near Neilburg (230 km northwest of Saskatoon) between 1 and 6 August 1999 in a field with moderate SM population density and SM adults emerging late in the season due to slow degree-day accumulation in this northern location. Experiment 1 revealed that significant captures of male SM (Fig. 2) required the presence of the late-eluting stereoisomer (Fig. 1Ea). Other stereoisomers did not adversely affect trap captures (Fig. 2, Experiments 1, 2). Increasing doses of stereoisomeric mosellin resulted in increasing captures of male SM (Fig. 2, Experiment 3). Following field experiments, the absolute configuration of the attractive stereoisomer was determined to be (2S,7S)-2,7-nonanediyl dibutyrate through stereospecific syntheses (Fig. 1D) and chromatographic analyses of all four individual stereoisomers. The SS-stereoisomer had chromatographic characteristics identical to the late-eluting stereoisomer of synthetic mosellin on the chiral HPLC column (Fig. 1Ea), and to the first-eluting stereoisomer on the chiral GC column (Fig. 1Eb). It also elicited the strongest antennal response. Pheromone components of two cecidomyiid midges have been reported. (2S)-(E)-10-Tridecen-2-yl acetate was identified in the Hessian fly, Mayetiola destructor (Foster et al. 1991), and 2-acetoxytridecane, (2S,11S)diacetoxytridecane and (2S,12S)-diacetoxytridecane were identified in the pea midge, Contarinia pisi (Hillbur et al. 1999). These pioneering studies provided valuable information on pheromone chemistry in cecidomyiid midges. This paper is the first to report both identification and successful field testing of a cecidomyiid pheromone. Data from elemental and NMR analyses were consistent with the proposed pheromone structure 1. Whether the di-butyrate mosellin is a representative pheromone for the genus Sitodiplosis will be1 While this manuscript was in press, field trapping data on pea midge were published in J. Chem. Ecol. 26(8) : 1941–1952 453 Experiment 1 100 a 18 a 80 16 60 NUMBER OF MALES CAPTURED (x+SE) a Experiment 3 a 40 14 ab 20 b b – – – – 0.3 – – – 12 – – – 0.3 0.3 – – 0.3 0.3 0.3 0.3 0.3 RR SR RS SS 10 a 12 8 Experiment 2 b 10 a 6 8 6 4 4 2 2 b – – – – b c b c – 1 1 – 1 – – – – – – 1 1 1 1 1 RR SR RS SS PHEROMONE STEREOISOMERS ( µg) Fig. 2 Captures of male S. mosellana in Experiments 1–3 in Intercept-W wing traps, placed in wheat fields and baited with the SS-, SR- plus RS-, or RR-stereoisomer of 2,7-nonanediyl dibutyrate (mosellin) singly and in combination. Experiment 1 : three replicates; Experiments 2, 3 : ten replicates each. The absolute configuration of HPLC-separated (Fig. 1Ea) and field-tested stereoisomers was determined after completion of field experiments through stereospecific syntheses of stereoisomers (Fig. 1D) and comparative chiral GC analyses. SR- and RS-stereoisomers could not be separated by analytical procedures (Fig. 1E) and needed fieldtesting in combination. Doses of test chemicals in Experiments 1 and 2 reflected compound availability, with larger quantities of HPLC-separated pheromone stereoisomers available for Experiment 2. Data in all three experiments were subjected to KruskalWallis analysis of variance by ranks (Zar 1984) followed by comparison of means (Student-Newman-Keuls test), P~0.05. Bars in each experiment with the same letter superscript are not significantly different, Pp 0.05. come apparent as pheromones from congeners are identified. Facile synthesis of mosellin (Fig. 1C) will facilitate the development of pheromone-based monitoring, or even control, of SM populations. Acknowledgements We thank Wittko Francke for lending a chiral GC column that separated stereoisomers of mosellin, Ian Wise and Murray Braun for shipment of S. mosellana cocoons, and an anonymous reviewer for constructive comments. The research was supported through an equipment grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) to G.G. and colleagues at Simon Fraser University. 0 c 0.02 0.2 2 20 200 DOSE ( µg) OF STEREOISOMERIC PHEROMONE References Arn H, Städler E, Rauscher S (1975) The electroantennographic detector – a selective and sensitive tool in the gas chromatographic analysis of insect pheromones. Z Naturforsch 30c:722–725 Elliott RH (1998a) Evaluation of insecticides for protection of wheat against damage by the wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae). Can Entomol 120 : 615–626 Elliott RH (1998b) Factors influencing the efficacy and economic returns of aerial sprays against the wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae). Can Entomol 120 : 941–954 Foster SP, Harris MO, Millar JG (1991) Identification of the sex pheromone of the Hessian fly, Mayetiola destructor (Say). Naturwissenschaften 78 : 130–131 Hillbur Y, Anderson P, Arn H, Bengtsson M, Löfqvist J, Biddle AJ, Smitt O, Högberg H-E, Plass E, Franke S, Francke W (1999) Identification of pheromone components of the pea midge, Contarinia pisi (Diptera: Cecidomyiidae). Naturwissenschaften 86 : 292–294 König WA, Gehrcke B, Icheln D, Evers P, Donnecke J, Wang W (1992) New selectively substituted cyclodextrins as stationary phases for the analysis of chiral constituents of essential oils. J High Resol Chromatogr 15 : 367–372 Lamb RJ (1998) Crop resistance for wheat midge in spring wheat. In: Prospects for innovation: pressures for change. Entomol Soc Manitoba Inc Pest Control, October 16–17, Abstract Lamb RJ, Wise IL, Olfert OO, Gavloski J, Barker PS (1999) Distribution and seasonal abundance of Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae) in spring wheat. Can Entomol 131 : 387–397 454 Miller BS, Halton P (1961) The damage to wheat kernels caused by the wheat blossom midge (Sitodiplosis mosellana). J Sci Food Agric 12 : 391–398 Mongrain D, Couture L, Dubuc J-P, Comeau A (1997) Occurrence of the orange wheat blossom midge (Diptera: Cecidomyiidae) in Quebec and its incidence on wheat grain microflora. Phytoprotection 78 : 17–22 Olfert OO, Mukerji MK, Doane JF (1985) Relationship between infestation levels and yield loss caused by wheat midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae), in spring wheat in Saskatchewan. Can Entomol 117 : 593–598 Pietruszka J, Hochmuth DH, Gehrcke B, Icheln D, Runge T, König WA (1992) Gas chromatographic enantioseparation of allenes. Tetrahedr Asymm 3 : 661–670 Pivnick KA (1993) Response of males to female sex pheromone in the orange wheat blossom midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae). J Chem Ecol 91 : 1677– 1689 Pivnick KA, Labbé E (1992) Emergence and calling rhythms, and mating behaviour of the orange wheat blossom midge, Sitodiplosis mosellana (Géhin) (Diptera: Cecidomyiidae). Can Entomol 124 : 501–507 Schaus SE, Branalt J, Jacobsen EN (1998) Total synthesis of Muconin by efficient assembly of chiral building blocks. J Org Chem 63 : 4876–4877 Tokunaga M, Farrow JF, Kakuichi F, Jacobsen EN (1997) Asymmetric catalysis with water: efficient kinetic resolution of terminal epoxides by means of catalytic hydrolysis. Science 277 : 936–938 Van den Dool H, Kratz PD (1963) A generalization of the retention index system including linear temperature programmed gas-liquid partition chromatography. J Chromatogr 2 : 463– 471 Zar JH (1984) Biostatistical analysis. Prentice-Hall, Englewood Cliffs, N.J.