Identification, Synthesis, And Field Testing Of

Journal of Chemical Ecology, Vol. 32, No. 1, January 2006 ( #2006) DOI: 10.1007/s10886-006-9359-6

IDENTIFICATION, SYNTHESIS, AND FIELD TESTING OF THE SEX PHEROMONE OF THE CITRUS LEAFMINER, Phyllocnistis citrella

JARDEL A. MOREIRA, J. STEVEN MCELFRESH, and JOCELYN G. MILLAR* Department of Entomology, University of California, Riverside, CA 92521, USA (Received June 9, 2005; revised August 11, 2005; accepted October 3, 2005)

Abstract—The citrus leafminer is an important vector of citrus canker in many of the major citrus production areas of the world. (7Z,11Z)-Hexadecadienal was reported as a sex attractant for this insect in the 1980s, based on trap catches during pheromone screening trials in Japan. However, attempts to reproduce this work in other areas of the world have not been successful. We report here that (7Z,11Z)-hexadecadienal is only one component of the pheromone, with the other critical component being the analogous trienal, (7Z,11Z,13E)-hexadecatrienal. Both compounds were identified in the effluvia from live female moths by coupled gas chromatography (GC)-electroantennography using nonpolar and polar GC columns, and the identifications were confirmed by comparisons of mass spectra with those of authentic standards. Stereoisomers of the two compounds, and a number of analogs, were synthesized to confirm the identifications. In field trials, neither compound alone was attractive to male moths, but blends of the two were highly attractive, with thousands of insects being caught per trial. Addition of the isomeric (7Z,11Z,13Z)-hexadecatrienal inhibited attraction to the twocomponent blend. Key Words—Sex pheromone, (7Z,11Z)-hexadecadienal, (7Z,11Z,13E)-hexadecatrienal, 5,7,11-hexadecatrienal, citrus leafminer, (7Z,11Z,13Z)-hexadecatrienal.

* To whom correspondence should be addressed. E-mail: [email protected] This paper and the preceeding paper (Leal et al.) were submitted within a few days of each other. The editors and the authors agreed that they should be published in tandem.

169 0098-0331/06/0100-0169/0 # 2006 Springer Science + Business Media, Inc.

170

MOREIRA ET AL. INTRODUCTION

The citrus leafminer (CLM), Phyllocnistis citrella Stainton (Lepidoptera: Gracillariidae), is a pest in most of the citrus-growing regions of the world (Heppner, 1993). It attacks all varieties of citrus and some related plant species, with grapefruit, tangerine, and pumello being among the most susceptible hosts (Legaspi and French, 2003). Although mature trees can tolerate some damage, growth of young trees in nurseries and newly planted orchards is reduced. Damage is caused by the larvae forming serpentine mines in the leaves, in which they are well protected from insecticide sprays, making them difficult to control. Even more important than direct damage is the potential for the adult moths to vector the highly contagious and lethal citrus canker bacterium, Xanthomonas axonopodis pv citri, where this pathogen occurs (Ando et al., 1985). The importance of this highly contagious disease to growers is reflected in the control program in Florida, USA, which calls for the destruction of all exposed citrus trees within 1900 ft of an infected one (Florida Department of Agriculture and Consumer Services, http://www.doacs.state.fl.us/pi/canker/cankerflorida.pdf ). There are currently no good methods for early detection and sampling of this insect to track the continued spread through California and other citrusgrowing regions in the United States. Often, the first signs of its presence are the characteristic mines in the leaves, by which time the infestation may be well established. Although (7Z,11Z)-hexadecadienal (7Z,11Z-16:Ald) was reported as a sex attractant for this insect from screening trials 20 yr ago (Ando et al., 1985), in subsequent trials in other parts of the world, including China, Spain, Florida, and Italy (Jacas and Pen˜a, 2002), this compound was not effective. Due to the threat of CLM to the California citrus industry, we undertook an investigation of its pheromone chemistry, and report that the pheromone of California populations consists of at least two components, the previously identified (7Z,11Z)-hexadecadienal, and the corresponding (7Z,11Z,13E)-hexadecatrienal (7Z,11Z,13E-16:Ald).

METHODS AND MATERIALS

Insects. Insects were obtained by collecting branches with infested leaves from lemon orchards in the Coachella Valley, Riverside Co., CA, USA and from Escondido, San Diego Co., CA. Branches with larvae were kept in humidified, vented plastic boxes [32  17  9 cm (L  W  H), with four 18mm vents covered with fine brass screening, or for larger quantities, 40  30  14 cm with four 38-mm screen-covered vents) until the insects pupated. Pupae were removed from cocoons and transferred, using a damp camel-hair brush if necessary, into a Petri dish lined with a damp filter paper. The sex of pupae was

PHEROMONE OF CITRUS LEAFMINER

171

determined by examining the terminal abdominal segments (Garrido and Jacas, 1996). Pupae were then transferred to individual 1-dram glass shell vials (45  15 mm, L  diam.) with a small disk of filter paper to prevent them from becoming trapped in condensation on the glass. The vials, closed with a plastic cap with a pinhole for ventilation, were held in groups in a vented, humidified plastic box. Vials were checked daily for adult emergence. Adults were fed a sugar–water solution (10 g sugar in 100 ml deionized water) absorbed on 2- to 3-mm cubes of cellulose sponge (ca. 10–20 ml/cube). Male moths were used for coupled gas chromatography-electroantennogram detection (GC-EAD) studies, and females were used for pheromone sampling (see below). Collection of Pheromone from Live Females. Pheromone was collected from female CLM by placing groups of 1–30 virgin females in aeration chambers for 1–4 nights, collecting the emitted pheromone on solid phase microextraction (SPME) fibers. The aeration chambers were constructed from Swagelok\ fittings (two 1/8 to 1/2 in. unions and one 1/2 to 1/2 in. union) connected to central 12-mm OD glass tubes (6.5 cm long). Activated-charcoal purified medical air was humidified by passage through a õ5  1 cm (L  diam.) wad of Soxhlet-extracted glass wool wetted with 1 ml of MilliQ\ purified water, then it was passed through a 5  1 cm (L  diam.) bed of 50– 200 mesh activated charcoal, and finally into the aeration chamber through a steel frit, at a flow rate of õ8–10 ml/min. An SPME fiber (100 mm polydimethylsiloxane coating, SPME portable field sampler; Supelco Inc., Bellefonte, PA, USA) was inserted into the 1-mm-diam. Teflon outlet tube at the downwind end of the chamber for collection of volatiles. Insects were added to the chamber prior to assembly, and were provided with sugar–water on a cellulose sponge cube during aerations. When there were >3 females, a 1  9 cm piece of clean, fine brass screening was provided as a perching substrate. Aerations were conducted in the laboratory with natural daylight, augmented with room fluorescent lighting during the day. Room temperatures varied from 20 to 25-C and were not precisely controlled. Coupled Gas Chromatography-Electroantennography. Female-produced volatiles collected by SPME and synthetic standards were analyzed by GCEAD on Hewlett-Packard 5890 series II gas chromatographs, with the first fitted with a DB-5 column (30 m  0.25 mm ID, 0.25 mm film, 100-C/1 min, 10-/min to 275-C for 15 min; J&W Scientific, Folsom, CA, USA) and the second with a DB-WAX column (30 m  0.25 mm ID, 0.25 mm film, 100-C/1 min, 10-/min to 240-). Loaded SPME fibers were desorbed in the injection port for 1 min prior to starting the GC. Helium was used as the carrier and makeup gas, and all injections were made splitless. The column effluent was split equally with a glass X-cross with one branch going to the flame ionization detector (FID), another to the EAG detector (EAD), and the final branch providing helium makeup gas (3 ml/min). The EAD branch passed through a heated conduit

172

MOREIRA ET AL.

(250-C) and into the side of a 15-mm ID glass tube swept with humidified medical air (770 ml/min), with the air flow directed over the antennal preparation. Signals were recorded on either matched HP 3394 integrators or with SRI model 202 PeakSimple chromatography data system running on a personal computer with PeakSimple v. 2.83 software (SRI Instruments, Torrance, CA, USA). Males (1–2 d old) used in GC-EAD analyses were not chilled or anesthetized prior to use. Males were held with fine forceps, the head removed with a scalpel, and the antennal tips cut off. The head was mounted on the ground electrode and both cut antennal tips were placed in contact with the saline-filled glass recording electrode. Field Trials. In all trials, moths were counted under a stereomicroscope to ensure correct identification (based on distinctive wing patterns) because of the insects’ small size (less than 3 mm) and the large numbers trapped. Moths were readily identified from their distinctive wing patterns. Voucher specimens, UCRC 92975-92982 (pinned) and 92983 (specimens in alcohol), have been deposited in the UC Riverside Entomology Museum. Lures consisted of 11-mm gray rubber septa (West Pharmaceutical Services Co., Lionville, PA, USA) individually labeled with a treatment number and loaded with heptane solutions (75 or 100 ml) of test blends. Field trials were conducted in September 2004, in a heavily infested lemon orchard in Escondido, CA. Green Delta traps were used [Pherocon IIID traps (Tre´ce´ Inc., Salinas, CA, USA) or Intercept D traps (IPM Tech Inc., Portland, OR, USA)]. In the first two trials, testing blend ratio and total dose, respectively, five replicate blocks containing a complete set of treatments (including a solvent control) were spaced four to seven rows apart. Treatments were randomly placed within the rows, four trees in from the end and on every other tree within a row. Traps were positioned just inside the canopy, 1.5–2 m above ground. Counts were taken 1–2 d after initial setup, and then the traps were rerandomized, with a second count made 2 d later. Traps with more than 10 moths were changed by transferring the rubber septum to a new trap, and the moths in the old trap counted. In the first trial, testing blend ratios, trap catches from the two counts were summed prior to transformation (¾x + 0.5), and then subjected to two-way analysis of variance using SigmaStat v. 1.0 (Jandel Scientific, San Rafael, CA, USA). Significantly different treatment means were distinguished using the Student–Neuman–Keuls test (SNK, a = 0.05). In the second trial, testing dose response, data could not be normalized by transformation, and so the data were analyzed by Friedman’s two-way nonparametric ANOVA on the untransformed data, followed by Bonferroni’s tests for separation of means (a = 0.05). A final trial testing the effects of 7Z,11Z,13Z-16:Ald as a possible synergist or antagonist was carried out on June 7–8 2005, using a standard blend of

PHEROMONE OF CITRUS LEAFMINER

173

7Z,11Z,13E-16Ald and 7Z,11Z-16:Ald (100:33 mg) augmented with 0 to 100 mg 7Z,11Z,13Z-16:Ald. The experiment was set up in the same orchard as used previously. After log10(x + 1) transformation to normalize the data, the transformed counts were analyzed by two-way ANOVA followed by SNK tests (a = 0.05). Synthesis of Pheromones and Related Compounds. Tetrahydrofuran (THF) was distilled from sodium/benzophenone under argon atmosphere. 1H and 13C NMR spectra were recorded with a Varian INOVA-400 (400 and 100 MHz, respectively) spectrometer, as CDCl3 solutions. Chemical shifts are expressed in ppm relative to CDCl3 (7.26 and 77.23 ppm for 1H and 13C NMR, respectively). Mass spectra were obtained with an HP 5890 GC equipped with a DB5-MS column (25 m  0.20 mm ID  0.33 mm film; J&W Scientific, Folsom, CA, USA), interfaced to an HP 5970 mass selective detector, in EI mode (70 eV) with helium carrier gas. Products were purified by flash or vacuum flash chromatography using silica gel (230–400 mesh; EM Science, Gibbstown, NJ, USA). Reactions with air- or water-sensitive reagents were carried out in dried glassware under argon atmosphere. Worked up reaction mixtures were dried over anhydrous Na2SO4 and concentrated by rotary evaporation under reduced pressure. The syntheses of the 5,7,11-hexadecatrienal isomers (Schemes 1 and 2), and the previously known (7Z,11Z)-hexadecadienal (Scheme 5) are described in detail in the online supplement to this manuscript (Electronic Supplementary Material is available for this article at http://dx.doi.org/10.1007/s10886-0069359-6 and is accessible for authorized users). Short, Nonstereospecific Preparation of 7,11,13-Hexadecatrienal Isomers. (Z)-11-Acetoxyundec-4-enal (28). (Scheme 3) A solution of KMnO4 (1.70 g, 10.7 mmol) in water (12.7 ml) was added dropwise to a stirred solution of (7Z,11E/Z)-hexadecadienyl acetate 25 (3.00 g, 10.7 mmol; Shin-Etsu Chemical Co., Tokyo, Japan) in ethanol (35 ml) at 0-C over 2.5 hr, and the resulting mixture was stirred at ambient temperature for 2 hr. The mixture was then filtered through a pad of Celite and the filtrate was concentrated. The residue was taken up in Et2O (100 ml), washed with brine, dried, and concentrated. The crude product was filtered through a plug of silica gel (hexane/EtOAc, 9:1) giving 0.71 g of diols 26 and 27. The diols (0.61 g, 1.94 mmol) were added to a solution of NaIO4 (1.45 g, 6.8 mmol) in a mixture of THF/acetone/water (8:4:13–20 ml) and stirred for 4 hr at 0-C, then diluted with ethyl acetate (50 ml), washed with water and brine, dried, and concentrated. Purification by vacuum flash chromatography (hexane/EtOAc, 95:5 as eluent) afforded 0.21 g of aldehyde 28 (10.1% yield over two steps). MS: 226 (M+, trace), 148 (6), 122 (14), 109 (3), 95 (11), 93 (14), 84 (19), 81 (27), 80 (17), 79 (19), 69 (13), 68 (28), 67 (51), 55 (35), 54 (20), 43 (100), 41 (56). (7Z,11E/Z,13Z)-Hexadecatrienyl Acetate (31). LDA (1.5 M in cyclohexane, 0.44 ml, 0.66 mmol) was added dropwise to a suspension of (Z)-2-pentenyl

174

MOREIRA ET AL.

SCHEME 1. Synthesis of (5,7,11Z)-hexadecatrienals.

triphenylphosphonium bromide 30 (0.273 g, 0.66 mmol) in THF (2.0 ml) at j10-C. The mixture was stirred for 2 hr and then cooled to j20-C, and (Z)-11acetoxyundec-4-enal 28 (0.100 g, 0.44 mmol) dissolved in THF (1.0 ml) was added dropwise. The resulting mixture was warmed to room temperature and stirred overnight, then quenched with water and poured into saturated aqueous

PHEROMONE OF CITRUS LEAFMINER

175

SCHEME 2. Synthesis of (5Z,7Z,11Z)-hexadecadienal 24.

NH4Cl solution. The organic layer was separated and the aqueous phase extracted with Et2O (3  10 ml). The combined organic layers were washed with brine, dried, and concentrated. The residue was purified by flash chromatography (hexane/ethyl acetate, 95:5) affording (7Z,11E/Z,13Z)-hexadecatrienyl

SCHEME 3. Synthesis of (7Z,11,13)-hexadecatrienals.

176

MOREIRA ET AL.

acetate 31 (0.052 g) in 42.3% yield, in a 49:51 ratio of geometric isomers. MS: 7Z,11E,13Z (minor) isomer: 278 (M+, 10), 207 (1), 149 (3), 135 (8), 119 (5), 95 (100), 81 (16), 79 (23), 67 (48), 55 (21), 43 (51), 41 (30). MS: 7Z,11Z,13Z (major) isomer: 278 (M+, 8), 149 (2), 135 (7), 121 (7), 95 (100), 81 (17), 79 (24), 67 (58), 55 (24), 43 (54), 41 (31). (7Z,11E/Z,13Z)-Hexadecatrien-1-ol (32). (7Z,11E/Z,13Z)-hexadecatrienyl acetate 31 (0.040 g, 0.14 mmol) was added to a solution of NaOH (6 M in water, 2 drops) in methanol (1.0 ml) and stirred for 4 hr at room temperature. After concentration, the residue was diluted with EtOAc, washed with water and brine, dried, and then concentrated. The residue was purified by flash chromatography (hexane/EtOAc, 5:1) affording (7Z,11E/Z,13Z)-hexadecatrien-1-ol 32 (0.030 g) in 88.2% yield. MS: 7Z,11E,13Z (minor) isomer: 236 (M+, 6), 163 (1), 149 (2), 135 (5), 121 (4), 95 (100), 81 (14), 79 (18), 67 (46), 55 (24), 41 (36). MS: 7Z,11Z,13Z (major) isomer: 236 (M+, 5), 149 (2), 135 (7), 121 (5), 95 (100), 81 (15), 79 (19), 67 (54), 55 (27), 41 (37). (7Z,11E/Z,13Z)-Hexadecatrienal (33). (7Z,11E/Z,13Z)-Hexadecatrien-1-ol ˚ molecular 32 (5.0 mg, 0.021 mmol) was oxidized with PCC and powdered 4 A sieve in CH2Cl2 as described above, yielding (7Z,11E/Z,13Z)-hexadecatrienal 33 in 56.6% yield (2.8 mg). MS: 7Z,11E,13Z (minor) isomer: 234 (M+, 3), 163 (1), 149 (2), 135 (5), 121 (2), 95 (100), 79 (16), 67 (50), 55 (25), 41 (36). MS: 7Z,11Z,13Z (major) isomer: 234 (M+, 3), 163 (1), 149 (1), 135 (3), 121 (1), 95 (100), 79 (13), 67 (44), 55 (21), 41 (34). (7Z,11E/Z,13E)-Hexadecatrienyl acetate (35). In the same manner described for the preparation of (7Z,11E/Z,13Z)-hexadecatrienyl acetate 31, the reaction of aldehyde 28 with the ylide from (E)-2-pentenyltriphenylphosphonium bromide 34 (0.273 g, 0.66 mmol) gave (7Z,11E/Z,13E)-hexadecatrienyl acetate 35 (0.071 g, 57.7% yield), in a 41:59 ratio of geometric isomers. MS: 7Z,11Z,13E (minor) isomer: 278 (M+, 7), 189 (1), 175 (1), 161 (2), 149 (2), 135 (5), 121 (4), 95 (100), 81 (13), 79 (22), 67 (47), 55 (23), 43 (47), 41 (29). MS: 7Z,11E,13E (major) isomer: 278 (M+, 7), 175 (1), 161 (1), 149 (1), 135 (4), 121 (3), 95 (100), 81 (10), 79 (13), 67 (40), 55 (18), 43 (40), 41 (23). (7Z,11E/Z,13E)-Hexadecatrien-1-ol (36). (7Z,11E/Z,13E)-Hexadecatrienyl acetate 35 (0.060 g, 0.22 mmol) was hydrolyzed as described above for acetate 31, giving (7Z,11E/Z,13E)-hexadecatrien-1-ol 36 (0.041 g, 80.6% yield). MS: 7Z,11Z,13E (minor) isomer: 236 (M+, 3), 207 (traces), 163 (1), 149 (2), 135 (4), 121 (3), 95 (100), 81 (10), 79 (16), 67 (45), 55 (24), 41 (34). MS: 7Z,11E,13E (major) isomer: 236 (M+, 4), 207 (traces), 149 (1), 135 (3), 121 (2), 95 (100), 81 (8), 79 (13), 67 (39), 55 (20), 41 (28). (7Z,11E/Z,13E)-Hexadecatrienal (37). Alcohol 36 (5.0 mg, 0.021 mmol was oxidized as described above for alcohol 32, yielding (7Z,11E/Z,13E)hexadecatrienal 37 (3.1 mg) in 62.7% yield. MS: 7Z,11Z,13E (minor) isomer: 234 (M+, 2), 149 (2), 135 (3), 121 (2), 95 (100), 79 (15), 67 (47), 55 (22), 41

PHEROMONE OF CITRUS LEAFMINER

177

SCHEME 4. Synthesis of (7Z,11Z,13E)-hexadecatrienal 49.

(32). MS: 7Z,11E,13E (major) isomer: 234 (M+, 3), 216 (traces), 149 (1), 135 (2), 121 (2), 95 (100), 79 (11), 67 (37), 55 (19), 41 (27). Preparation of (7Z,11Z,13E)-Hexadecatrienal (49). 1-(t-Butyldimethylsilyloxy)-3-bromopropane (39). (Scheme 4) 4-(Dimethylamino)pyridine (70 mg, 0.57 mmol) and triethylamine (6.42 g, 63.4 mmol) were added dropwise to a j10-C, stirred mixture of 3-bromo-1-propanol 38 (8.00 g, 57.6 mmol) and t-butyl-dimethylsilyl chloride (9.11 g, 60.4 mmol) in CH2Cl2 (65 ml). After 1 hr, the mixture was warmed to room temperature and stirred for 16 hr. Water (50 ml) was added, and the mixture then was extracted three times with Et2O (1  80, 2  50 ml). The combined organic phases were washed with 10%

SCHEME 5. Synthesis of (7Z,11Z)-hexadecadienal 56.

178

MOREIRA ET AL.

aqueous HCl (3  30 ml), saturated NaHCO3 solution (2  50 ml), and brine, then dried and concentrated. The residue was Kugelrohr distilled (oven temp. 68–72-C, 10 mmHg) giving 13.66 g (93.7% yield) of bromide 39. 1H NMR: d 0.07 (s, 6H), 0.90 (s, 9H), 1.93–2.00 (m, 2H), 3.51 (t, 2H, J = 6.4 Hz), 3.73 (t, 2H, J = 5.8 Hz). 13C NMR: d j5.17, 18.51, 26.12, 30.89, 35.77, 60.64 ppm. MS: m/z 197 (M+ j57, 48), 195 (45), 169 (61), 167 (60), 155 (5), 153 (6), 139 (100), 137 (97), 115 (65), 99 (7), 85 (8), 75 (23), 73 (28), 59 (27), 58 (16), 57 (24), 47 (19), 45 (31), 43 (20), 41 (43). The NMR spectrum matched literature values (Kerr et al., 1990). 1-(Tetrahydropyranyloxy)-oct-7-yne (41). Dihydropyran (16.00 g, 190.2 mmol) was added dropwise to a solution of 7-octyn-1-ol 40 (16.00 g, 126.8 mmol; prepared from 2-octyn-1-ol by the acetylene zipper reaction; Abrams and Shaw, 1988) and a few crystals of PTSA in CH2Cl2 (30 ml) at 0-C under argon. The reaction was allowed to warm to room temperature and stirred overnight. Most of the solvent was removed by rotary evaporation, the residue was diluted with hexane (150 ml), washed with saturated NaHCO3 solution (2  50 ml) and brine, dried, and concentrated. The resulting oil was purified by vacuum flash chromatography (hexane, then hexane/ethyl acetate, 95:5) followed by Kugelrohr distillation (bp õ94–96-C, 1.0 mmHg) yielding 25.48 g (95.6%) of pure product. 1H NMR: d 1.32–1.48 (m, 4H), 1.48–1.65 (m, 8H), 1.67–1.75 (m, 1H), 1.78–1.88 (m, 1H), 1.93 (t, 1H, J = 2.7 Hz), 2.18 (td, 2H, J = 2.7 and 6.8 Hz), 3.38 (dt, 1H, J = 9.8 and 6.4 Hz), 3.46–3.53 (m, 1H), 3.73 (dt, 1H, J = 9.8 and 6.8 Hz), 3.83–3.90 (m, 1H), 4.57 (dd, 1H, J = 4.4 and 2.7 Hz). 13C NMR: d 18.57, 19.92, 25.70, 25.99, 28.65, 28.79, 29.83, 31.00, 62.58, 67.75, 68.33, 84.90, 99.09 ppm. MS: m/z 209 (M+ j 1, 1), 125 (1), 109 (3), 101 (29), 85 (100), 67 (40), 55 (23), 43 (21), 41 (51). 1-(t-Butyldimethylsilyloxy)-11-(tetrahydropyranyloxy)-undec-4-yne (42). n-BuLi (1.6M in hexanes, 12.2 ml, 19.5 mmol) was added dropwise to a stirred solution of 1-(tetrahydropyranyloxy)-oct-7-yne 41 (3.90 g, 18.5 mmol) in dry THF (20 ml) at j40-C. The mixture was allowed to warm to j10-C and stirred for 4 hr. 1,3-Dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone (DMPU, 9.0 ml, 74.2 mmol) was added over 10 min, the resulting mixture was stirred for 1 hr, and then cooled to j40-C. A solution of 1-(t-butyldimethylsilyloxy)-3bromopropane 39 (4.70 g, 18.5 mmol) in dry THF (10 ml) was added dropwise, and the mixture was allowed to warm to room temperature over õ4 hr and then stirred overnight. The reaction was quenched with saturated NH4Cl solution (20 ml) and extracted with ethyl acetate (4  20 ml). The combined organic phases were washed with saturated NaHCO3 solution and brine, dried, and concentrated. The residue was purified by vacuum flash chromatography (hexane/ethyl acetate, 95:5) affording 6.83 g of product contaminated with unreacted 1-(t-butyldimethylsilyloxy)-3-bromopropane 39. This impure material was used directly in the next step. 1H NMR: d 0.05 (s, 6H), 0.89 (s, 9H), 1.33–

PHEROMONE OF CITRUS LEAFMINER

179

1.88 (m, 16H), 2.18 (tt, 2H, J = 7.0 and 2.3 Hz), 2.26 (tt, 2H, J = 7.0 and 2.3 Hz), 3.38 (dt, 1H, J = 9.6 and 6.6 Hz), 3.47–2.53 (m, 1H), 3.68 (t, 2H, J = 6.0 Hz), 3.73 (dt, 1H, J = 9.6 and 6.8 Hz), 3.77–3.87 (m, 1H), 4.57 (dd, 1H, J = 4.3 and 2.7 Hz). 13C NMR: d j5.12, 15.36, 18.57, 18.91, 19.90, 25.72, 26.04, 26.16, 28.93, 29.29, 29.88, 30.99, 32.36, 61.98, 62.54, 67.80, 79.90, 80.51, 99.06 ppm. MS: m/z 325 (M+ j57, 2), 241 (5), 165 (3), 159 (12), 93 (9), 85 (100), 81 (12), 79 (10), 75 (29), 73 (13), 67 (19), 57 (10), 55 (13), 43 (12), 41 (22). 11-(Tetrahydropyranyloxy)-undec-4-yn-1-ol (43). A solution of tetrabutylammonium fluoride (1.0 M in THF, 21.4 ml, 21.4 mmol) was added to a solution of 1-(t-butyldimethylsilyloxy)-11-(tetrahydropyranyloxy)-undec-4-yne 42 (6.82 g) in dry THF (35 ml), at room temperature. The mixture was stirred until the starting material had been consumed (1.5 hr), then diluted with Et2O (50 ml) and washed with water. The organic phase was separated, and the aqueous layer was extracted with Et2O (2  20 ml). The combined organic phases were washed with brine, dried, and concentrated. The residue was purified by vacuum flash chromatography (hexane/EtOAc, 4:1) to yield 3.49 g of 43 (70.1% over two steps). 1H NMR: d 1.31–1.64 (m, 10H), 1.66–1.88 (m, 4H), 1.73 (quint, 2H, J = 6.3 Hz), 2.13 (tt, 2H, J = 6.8 and 2.5 Hz), 2.27 (tt, 2H, J = 6.8 and 2.5 Hz), 3.38 (dt, 1H, J = 9.6 and 6.6 Hz), 3.46–3.54 (m, 1H), 3.73 (dt, 1H, J = 9.6 and 6.8 Hz), 3.74 (t, 2H, J = 6.0 Hz), 3.82–3.90 (m, 1H), 4.57 (dd, 1H, J = 4.5 and 2.5 Hz). 13C NMR: d 15.64, 18.88, 19.88, 25.71, 25.95, 28.84, 29.15, 29.82, 30.98, 31.80, 62.24, 62.56, 67.76, 79.61, 81.22, 99.08 ppm. MS: m/z 268 (M+, trace), 195 (3), 183 (3), 167 (3), 125 (2), 111 (3), 101 (18), 85 (100), 81 (13), 79 (15), 67 (21), 57 (14), 55 (26), 43 (19), 41 (44). Anal. calculated for C16H28O3: C, 71.60; H, 10.52. Found: C, 71.85; H, 10.59. The 1H NMR spectrum was in agreement with that reported in the literature (Yadav et al. 1988). (Z)-11-(Tetrahydropyranyloxy)undec-4-en-1-ol (44). Alkynol 43 (3.40 g, 12.67 mmol) was reduced to the corresponding alkenol with hydrogen and P2nickel catalyst as described above. The crude product was purified by vacuum flash chromatography (hexane/ethyl acetate, 9:1), yielding 3.14 g of 44 (91.8% yield). 1H NMR: d 1.24–1.42 (m, 6H), 1.44–1.64 (m, 8H), 1.64–1.73 (m, 1H), 1.73–1.85 (m, 1H), 1.96–2.07 (m, 2H), 2.07–2.14 (m, 2H), 3.36 (dt, 1H, J = 9.6 and 6.6 Hz), 3.44–3.51 (m, 1H), 3.62 (dt, 2H, J = 5.3 and 6.4 Hz), 3.70 (dt, 1H, J = 9.6 and 6.8 Hz), 3.81–3.88 (m, 1H), 4.55 (dd, 1H, J = 4.3 and 2.7 Hz), 5.30– 5.41 (m, 2H). 13C NMR: d 19.89, 23.81, 25.70, 26.31, 27.30, 29.28, 29.79, 29.90, 30.97, 32.86, 62.57, 62.76, 67.87, 99.06, 129.18, 130.85 ppm. MS: m/z 270 (M+, trace), 186 (1), 168 (1), 150 (1), 109 (3), 101 (4), 95 (7), 85 (100), 81 (11), 67 (19), 57 (10), 55 (17), 43 (12), 41 (28). The 1H NMR spectrum was in agreement with that reported in the literature (Sharma et al. 1994). (4Z)-1-Iodo-11-(tetrahydropyranyloxy)undec-4-ene (45). Iodine (2.91 g, 11.46 mmol) was added in small portions to a j40-C solution of Ph3P (3.00 g,

180

MOREIRA ET AL.

11.46 mmol), imidazole (1.56 g, 22.92 mmol), and (Z)-11-(tetrahydropyranyloxy)undec-4-en-1-ol 44 (3.10 g, 11.46 mmol) in dry THF (18 ml). The resulting mixture was warmed to 0-C and stirred for 3 hr. The mixture was diluted with Et2O (100 ml), washed successively with saturated sodium thiosulfate (100 ml), water (2  50 ml), and brine, then dried and concentrated. The residue was suspended in hexane to precipitate most of the Ph3PO and filtered. After concentration, the crude product was purified by vacuum flash chromatography (hexane/ethyl acetate, 9:1) yielding 2.15 g of 45 (57.6% yield) as colorless oil and 0.201 g of unreacted starting material. 1H NMR: d 1.22– 1.42 (m, 6H), 1.45–1.64 (m, 6H), 1.64–1.75 (m, 1H), 1.77–1.87 (m, 1H), 1.87 (quint, 2H, J = 7.0 Hz), 2.05 (q, 2H, J = 6.6 Hz), 2.13 (q, 2H, J = 7.2 Hz), 3.18 (t, 2H, J = 6.8 Hz), 3.37 (dt, 1H, J = 9.6 and 6.6 Hz), 3.46–3.52 (m, 1H), 3.72 (dt, 1H, J = 9.6 and 6.8 Hz), 3.82–3.90 (m, 1H), 4.57 (dd, 1H, J = 4.3 and 2.7 Hz), 5.28 (dtt, 1H, J = 10.7, 7.2 and 1.4 Hz), 5.42 (dtt, 1H, J = 10.7, 7.4 and 1.4 Hz). 13 C NMR: d 6.87, 19.93, 25.73, 26.37, 27.56, 28.14, 29.35, 29.88, 29.95, 31.02, 33.64, 62.57, 67.86, 99.07, 127.51, 131.92 ppm. MS: m/z 183 (1), 155 (4), 109 (4), 101 (4), 95 (10), 85 (100), 81 (13), 69 (8), 67 (17), 55 (18), 43 (9), 41 (27). Anal. calculated for C16H29IO2: C, 50.53; H, 7.69. Found: C, 50.64; H, 7.73. (Z)-11-(Tetrahydropyranyloxy)undec-4-enyl triphenylphosphonium iodide (46). A solution of Ph3P (1.66 g; 6.31 mmol) and (Z)-1-iodo-11-(tetrahydropyranyloxy)undec-4-ene 45 (2.40 g; 6.31 mmol) in toluene (7.0 ml) was refluxed for 2 d (GC analysis showed no remaining iodide), then poured with vigorous stirring into 50 ml hexane, to produce a viscous oil. The mixture was stirred for 1 hr, the solvent was replaced with fresh hexane (50 ml), and then stirred for an additional 1 hr. The solvent was removed and the remaining phosphonium salt 46 was pumped under high vacuum for 8 hr, then used without further purification. (7Z,11Z,13E)-1-(Tetrahydropyranyloxy)-hexadecatriene (47). Lithium hexamethyldisilazide (LiHMDS, 1.0 M in THF, 4.50 ml, 4.5 mmol) was added to a solution of dry DMPU (2.56 g, 20.0 mmol) in THF (12 ml) at 0-C. The resulting mixture was stirred for 30 min, then cooled to j78-C. A solution of (Z)-11-(tetrahydropyranyloxy)undec-4-enyl triphenylphosphonium iodide 46 (3.21 g, 5.00 mmol) in THF (18.0 ml) was added dropwise. After stirring for 1 hr, (E)-2-pentenal (0.560 g, 6.50 mmol) in THF (2 ml) was added dropwise. The resulting mixture was stirred for 2 hr at j78-C, then warmed to room temperature and stirred an additional 2 hr. The mixture was diluted with hexane/ Et2O (3:1–50 ml) and washed with saturated NH4Cl solution and brine, dried, and concentrated. Purification by vacuum flash chromatography (hexane/ EtOAc, 95:5) afforded a 93:7 mixture of the (7Z,11Z,13E)- and (7Z,11E,13E)trienes, respectively (0.664 g, 41.5% yield). (7Z,11Z,13E)-isomer: 1H NMR: d 1.01 (t, 3H, J = 7.6 Hz), 1.22–1.40 (m, 6H), 1.46–1.64 (m, 6H), 1.66–1.80 (m, 2H), 1.96–2.06 (m, 2H), 2.06–2.16 (m, 4H), 2.16–2.26 (m, 2H), 3.37 (dt, 1H, J = 9.6 and 6.6 Hz), 3.45–3.55 (m, 1H), 3.72 (dt, 1H, J = 9.6 and 6.8 Hz), 3.84–

PHEROMONE OF CITRUS LEAFMINER

181

3.90 (m, 1H), 4.56 (dd, 1H, J = 4.3 and 2.7 Hz), 5.30 (dt, 1H, J = 10.7 and 7.4 Hz), 5.32–5.42 (m, 2H), 5.70 (dt, 1H, J = 15.2 and 6.6 Hz), 5.95 (t, 1H, J = 10.9 Hz), 6.25–6.33 (m, 1H). 13C NMR: d 13.85, 19.92, 25.73, 26.11, 26.38, 27.45, 27.58, 28.07, 29.37, 29.88, 29.95, 31.01, 62.56, 67.87, 99.06, 124.85, 129.18, 129.21, 129.51, 130.69, 136.64 ppm. MS: m/z 320 (M+, 2), 302 (trace), 236 (1), 235 (1), 175 (1), 161 (1), 135 (2), 122 (3), 95 (35), 85 (100), 79 (15), 67 (29), 55 (20), 43 (11), 41 (30). (7Z,11Z,13E)-Hexadecatrien-1-ol (48). The THP protecting group was removed from (7Z,11Z,13E)-1-(tetrahydropyranyloxy)-hexadecatriene 47 (0.650 g, 2.02 mmol) with PTSA in methanol, as described above. The crude product was purified by vacuum flash chromatography (hexane/ethyl, acetate 5:1) followed by Kugelrohr distillation (oven temp. õ 112-C, 0.03 mmHg), yielding (7Z,11Z,13E)hexadecatrien-1-ol (0.338 g, 70.8% yield). The isomeric purity of the product was increased to 97% by recrystallization from hexane at j20-C. 1H NMR: d 1.01 (t, 3H, J = 7.4 Hz), 1.26–1.42 (m, 6H), 1.52–1.60 (m, 2H), 1.98–2.06 (m, 2H), 2.06–2.16 (m, 4H), 2.18–2.26 (m, 2H), 3.63 (t, 2H, J = 6.7 Hz), 5.30 (dt, 1H, J = 10.9 and 7.4 Hz), 5.35–5.42 (m, 2H), 5.70 (dt, 1H, J = 15.2 and 6.6 Hz), 5.96 (t, 1H, J = 10.9 Hz), 6.29 (ddq, 1H, J = 15.2, 10.9 and 1.6 Hz). 13C NMR: d 13.84, 25.86, 26.11, 27.41, 27.58, 28.06, 29.28, 29.87, 32.97, 63.25, 124.84, 129.18, 129.28, 129.49, 130.61, 136.67 ppm. MS: m/z 236 (M+, 3), 163 (1), 149 (2), 135 (4), 121 (3), 107 (3), 96 (10), 95 (100), 93 (12), 81 (10), 79 (18), 67 (48), 55 (24), 53 (9), 41 (37). Anal. calculated for C16H28O: C, 81.29; H, 11.94. Found: C, 80.11; H, 12.06. (7Z,11Z,13E)-Hexadecatrienal (49). (7Z,11Z,13E)-Hexadecatrien-1-ol 48 (18.0 mg, 0.076 mmol) was oxidized to the aldehyde with PCC and powdered 4 ˚ molecular sieve in CH2Cl2 as described above, and the crude product was A flash chromatographed (pentane/Et2O 9:1) yielding (7Z,11Z,13E)-hexadecatrienal 49 in 77.4% yield (13.8 mg). 1H NMR: d 1.01 (t, 3H, J = 7.4 Hz), 1.34–1.42 (m, 4H), 1.63 (quint, 2H, J = 7.4 Hz), 2.00–2.06 (m, 2H), 2.06–2.17 (m, 4H), 2.18–2.26 (m, 2H), 2.42 (td, 2H, J = 1.8 and 7.4 Hz), 5.30 (dt, 1H, J = 10.8 and 7.4 Hz), 5.34–5.42 (m, 2H), 5.71 (dt, 1H, J = 15.2 and 6.6 Hz), 5.96 (t, 1H, J = 10.9 Hz), 6.29 (ddq, 1H, J = 15.2, 10.9 and 1.6 Hz), 9.76 (t, 1H, J = 1.8 Hz). 13C NMR: d 13.84, 22.21, 26.11, 27.25, 27.59, 28.02, 29.00, 29.62, 44.10, 124.82, 129.23, 129.42, 129.53, 130.27, 136.71, 203.04 ppm. MS: m/z 234 (M+, 2), 216 (trace), 187 (1), 173 (1), 149 (2), 135 (3), 121 (2), 107 (2), 95 (100), 93 (11), 79 (14), 67 (45), 55 (22), 41 (32). Preparation of (7Z,11Z,13Z)-hexadecatrienal (62)(Scheme 6). 2-Pentynal (58). 2-Pentyn-1-ol 57 was oxidized with PCC, as described above. The crude product was flash chromatographed (pentane/Et2O, 9:1), and the solvent was distilled off using a Vigreux column, giving 2-pentynal 58 (0.832g) in a ca. 15% mixture with solvent. MS: m/z 82 (M+, 34), 81 (M+ j1, 57), 67 (1), 66 (7), 54 (57), 53 (100), 52 (12), 51 (32), 50 (40), 49 (17).

182

MOREIRA ET AL.

SCHEME 6. Synthesis of (7Z,11Z,13Z)-hexadecatrienal 62.

(Z)-1-(Tetrahydropyranyloxy)-hexadec-7,13-diyn-11-ene (59). The title compound was prepared by a Z-selective Wittig between the anion of 11(tetrahydropyranyloxy)-undec-4-ynyl triphenylphosphonium iodide 52 and 2pentynal 58 in THF/DMPU as described above. After workup, purification by vacuum flash chromatography (hexane/EtOAc, 95:5) gave endiyne 59 in 57.7% yield. 1H NMR: d 1.161 (t, 3H, J = 7.6 Hz), 1.26–1.59 (m, 12H), 1.61–1.69 (m, 1H), 1.72–1.82 (m, 1H), 2.08–2.17 (m, 2H), 2.17–2.26 (m, 2H), 2.33 (dq, 2H, J = 7.6 and 2.2 Hz), 2.45 (q, 2H, J = 7.2 Hz), 3.32 (dt, 1H, J = 9.6 and 6.4 Hz), 3.40–3.47 (m, 1H), 3.67 (dt, 1H, J = 9.6 and 6.8 Hz), 3.76–3.84 (m, 1H), 4.51– 4.59 (m, 1H), 5.37–5.48 (m, 1H), 5.83 (dt, 1H, J = 10.8 and 7.2 Hz). 13C NMR: d 13.43, 14.26, 18.64, 18.92, 19.90, 25.72, 26.03, 28.90, 29.24, 29.87, 29.88, 31.00, 62.54, 67.80, 76.62, 79,55, 80.90, 96.50, 99.07, 110.54, 140.69. MS: m/z 287 (1), 231 (1), 215 (2), 201 (1), 185 (3), 171 (3), 157 (9), 145 (13), 143 (16), 131 (13), 129 (21), 117 (16), 91 (39), 85 (100), 77 (40), 67 (22), 57 (19), 55 (21), 43 (18), 41 (47). (7Z,11Z,13Z)-1-(Tetrahydropyranyloxy)-hexadecatriene (60). A solution of cyclohexene (0.549 g, 6.68 mmol) in dry THF (2.0 ml) was added dropwise under argon to a solution of BH3IDMDS (solution 2.0 M in THF, 1.67 ml, 3.34 mmol) in dry THF (5.0 ml) at j10-C. The resulting mixture was stirred for 3 hr between j10 and 0-C, then allowed to warm to room temperature and stirred for 1 hr. The resulting white slurry of dicyclohexylborane was recooled to j20-C and a solution of (Z)-1-(tetrahydropyranyloxy)-hexadec-7,13-diyn-11ene 59 (0.264 g, 0.834 mmol) in dry THF (2.0 ml) was added dropwise over

PHEROMONE OF CITRUS LEAFMINER

183

20 min. The resulting mixture was slowly warmed to 0-C (ca. 2.5 hr) and then stirred for 4 hr. Glacial acetic acid (0.84 ml) was added dropwise, and the mixture was warmed to room temperature and stirred overnight. The solution was cooled to 0-C and treated with aqueous NaOH (20% by wt., 6.0 ml) followed by dropwise addition of hydrogen peroxide (30% by wt., 0.40 ml), keeping the temperature below 15-C. The mixture was warmed to room temperature and stirred for 1 hr, diluted with water (10 ml), and extracted with hexane (4  15 ml). The combined organic layers were washed with brine, dried, concentrated, and purified by vacuum flash chromatography on silica gel (hexane/EtOAc, 95:5) giving (7Z,11Z,13Z)-1-(tetrahydropyranyloxy)-hexadecatriene 60 (0.213 g) in 79.8% yield as a 92:8 mixture of the (7Z,11Z,13Z)- and (7Z,11E,13Z)-trienes, respectively. 1H NMR: d 0.98 (t, 3H, J = 7.6 Hz), 1.26– 1.43 (m, 6H), 1.46–1.65 (m, 6H), 1.66–1.74 (m, 1H), 1.76–1.80 (m, 1H), 1.96– 2.06 (m, 2H), 2.06–2.26 (m, 6H), 3.36 (dt, 1H, J = 9.6 and 6.6 Hz), 3.44–3.52 (m, 1H), 3.71 (dt, 1H, J = 9.6 and 6.8 Hz), 3.81–3.89 (m, 1H), 4.53–4.58 (m, 1H), 5.20–5.30 (m, 2H), 5.30–5.39 (m, 2H), 6.06–6.20 (m, 2H). 13C NMR: d 14.40, 19.91, 21.01, 25.73, 26.37, 27.43, 27.47, 27.83, 29.34, 29.86, 29.94, 31.00, 62.54, 67.85, 99.05, 123.17, 124.03, 129.13, 130.71, 131.47, 134.09 ppm. MS: m/z 320 (M+, 2), 236 (1), 235 (1), 175 (1), 161 (1), 135 (2), 121 (3), 95 (38), 85 (100), 79 (12), 67 (33), 55 (20), 43 (11), 41 (32). (7Z,11Z,13Z)-Hexadecatrien-1-ol (61). The THP protecting group was removed from (7Z,11Z,13Z)-1-(tetrahydropyranyloxy)-hexadecatriene 60 (0.163 g, 0.51 mmol) with PTSA in methanol as described above. The crude product was purified by vacuum flash chromatography (hexane/ethyl acetate, 9:1), yielding (7Z,11Z,13Z)-hexadecatrien-1-ol (0.096 g, 79.9% yield). The isomeric purity of the product was increased to 95% by chromatography over silica gel with 10% AgNO3 (hexane/Et2O, 9:1). 1H NMR: d 0.99 (t, 3H, J = 7.6 Hz), 1.28–1.42 (m, 6H), 1.50–1.62 (m, 2H), 1.98–2.06 (m, 2H), 2.07–2.26 (m, 6H), 3.62 (t, 2H, J = 6.6 Hz), 5.32–5.40 (m, 2H), 5.40–5.48 (m, 2H), 6.16–6.30 (m, 2H). 13C NMR: d 14.40, 21.02, 25.85, 27.39, 27.47, 27.83, 29.26, 29.85, 32.98, 63.26, 123.16, 124.03, 129.21, 130.64, 131.47, 134.14 ppm. MS: m/z 236 (M+, 7), 207 (1), 163 (1), 149 (3), 135 (8), 121 (5), 107 (5), 96 (11), 95 (100), 93 (17), 81 (19), 79 (23), 67 (65), 55 (31), 53 (13), 41 (46). (7Z,11Z,13Z)-Hexadecatrienal (62). (7Z,11Z,13Z)-Hexadecatrien-1-ol 61 (13.0 mg, 0.055 mmol) was oxidized to the aldehyde with PCC and powdered ˚ molecular sieve in CH2Cl2 as described above, and the crude product was 4A flash chromatographed (pentane/Et2O, 9:1) affording (7Z,11Z,13Z)-hexadecatrienal 62 in 71.5% yield (9.2 mg). 1H NMR: d 0.99 (t, 3H, J = 7.6 Hz), 1.28– 1.42 (m, 4H), 1.63 (quint, 2H, J = 7.4 Hz), 1.99–2.06 (m, 2H), 2.06–2.26 (m, 6H), 2.41 (td, 2H, J = 2.0 and 7.4 Hz), 5.34–5.39 (m, 2H), 5.39–5.49 (m, 2H), 6.16–6.30 (m, 2H), 9.76 (t, 1H, J = 2.0 Hz). 13C NMR: d 14.40, 21.02, 22.20, 27.23, 27.48, 27.79, 28.99, 29.61, 44.09, 123.13, 124.08, 129.46, 130.30,

184

MOREIRA ET AL.

131.39, 134.19, 203.01. MS: m/z 234 (M+, 6), 219 (trace), 187 (1), 173 (1), 149 (2), 135 (6), 121 (3), 107 (3), 95 (100), 93 (15), 79 (20), 67 (59), 55 (28), 41 (43).

RESULTS

The amounts of the pheromone components collected by SPME from overnight aerations of 1–30 female moths were usually below the detection limit of GC and GC-MS. Therefore, identification of the active compounds was primarily guided by a combination of GC-EAD, using the retention behavior of the compounds on GC columns of different polarity, and the relative responses of male moth antennae to various standards with different functional groups. Male moth antennae responded consistently to two compounds in the effluvia from females. Sporadically, there was a third, much smaller response visible when extracts were analyzed on the DB-5 column (Figure 1, Table 1). The earliest response on both columns (for DB-5, Figure 1, peak 1) matched the retention time of synthetic 7Z,11Z-16:Ald, the previously reported sex attractant

FIG. 1. Coupled gas chromatography-electroantennogram detection traces showing responses from the GC and a male moth antenna to volatiles collected from two citrus leafminer females over 2 nights on a SPME fiber. GC responses with letters were also present in blank runs. Peak 1: (7Z,11Z)-hexadecadienal; peak 2: (7Z,11Z,13E)hexadecatrienal; peak 3: tentatively identified as (7Z,11Z,13Z)-hexadecatrienal. Column: DB-5 (30 m  0.25 mm ID, 0.25 mm film), 100-C/1 min, 10-C/min to 275-C.

185

PHEROMONE OF CITRUS LEAFMINER

TABLE 1. COMPARISON OF RETENTION INDICES ON DB-5 AND DB-WAX GC COLUMNS OF INSECT-PRODUCED COMPOUNDS ELICITING ANTENNAL RESPONSES IN GC-EAD EXPERIMENTS, AND OF SYNTHETIC STANDARDS Retention indices Compound Insect-produced compounds Leafminer first response Leafminer second response Leafminer third response Synthetic standards Z11–16:Ald Z11,E13–16:Ald E11,Z13–16:Ald Z11,Z13–16:Ald Z7,Z11–16:Ald Z7,E11–16:Ald 5Z,7E,11Z-16:Ald 5E,7Z,11Z-16:Ald 5Z,7Z,11Z-16:Ald 5E,7E,11Z-16:Ald 7Z,11Z,13E-16:Ald 7Z,11Z,13Z-16:Ald 7Z,11E,13E-16:Ald 7Z,11E,13Z-16:Ald

a

DB-5

DB-WAXa

17.96 18.43 18.58

22.19 23.71 –b

18.10 18.56 18.65 18.74 17.95 17.93 18.29 18.36 18.55 18.62 18.43 18.57 18.56 18.44

21.86 23.35 23.47 23.57 22.20 22.12 23.46 23.60 23.83 23.86 23.70 23.89 23.97 23.75

Matches are shown in boldface type. a GC conditions: DB-5 (30 m  0.25 mm ID, 0.25 mm film, 100-C/1 min, 10-C/min–275-C/15 min); DB-WAX (30 m  0.25 mm ID, 0.25 mm film, 100-C/1 min, 10-C/min–240-C/10 min). Because the final temperature was reached before the tetracosane standard eluted on DB-WAX, the retention indices on this column may differ slightly from true Kovat’s indices. b Third response not seen on this column.

for this species (Ando et al., 1985). Furthermore, when challenged with synthetic standards of 7Z,11Z-16:Ald, male moth antennal responses were consistently stronger to the aldehyde than to the corresponding acetate or alcohol. The second compound, which consistently elicited the largest antennal responses, eluted about 50 and 150 retention units later than 7Z,11Z-16:Ald from the DB-5 and DB-WAX columns, respectively (Table 1). Both retention time and antennal response data ruled out the possibility of the unknown being 7Z,11Z-16:OH or 7Z,11Z-16:Ac. Retention times on both the nonpolar and polar columns were markedly longer than those of 7Z,11Z-16:Ald, suggesting the presence of a conjugated diene or triene in the unknown. The presence of a conjugated triene (e.g., a 7,9,11–16:Ald) was ruled out by examination of retention time differences of standards available from other projects (10E,12E,14E-16:Ald, Chen and Millar, 2000; 9E,11Z,13Z-16:Ald, Millar

186

MOREIRA ET AL.

et al., 1996). On the DB-5 column, these standards eluted >100 retention units later than the unknown, whereas the difference was even greater on the DBWAX column (208 and 179 retention units, respectively). Therefore, the most likely candidates for the unknown were the 5E/Z,7Z,11Z- or 7Z,11Z,13E/Z16:Ald isomers. These isomers were synthesized, and only 7Z,11Z,13E-16:Ald matched the retention times of the unknown on both columns. The identifications of both 7Z,11Z-16:Ald and 7Z,11Z,13E-16:Ald were subsequently confirmed from full-scan mass spectra of both compounds (Figure 2)

FIG. 2. EI mass spectra of (A) (7Z,11Z,13E)-hexadecatrienal and (B) (7Z,11Z)hexadecadienal.

PHEROMONE OF CITRUS LEAFMINER

187

obtained from an SPME sample collected from two females aerated over five consecutive nights. In particular, the spectrum of 7Z,11Z,13E-16:Ald exhibits a diagnostic base peak at m/z 95, corresponding to a C7H11+ ion arising from cleavage between allylic carbons 9 and 10, with the charge stabilized by the conjugated diene system. An analogous cleavage has been reported for 4,6,10–16:Ac isomers, which have a similar arrangement of double bonds (Beevor et al., 1986). The third EAD response matched the retention time of 7Z,11Z,13Z-16:Ald on the DB-5 column. However, it was not seen on the DB-WAX column, nor was it possible to obtain a mass spectrum, and so the identity of this compound could not be confirmed. Synthesis of the Pheromone Components and Analogs. The GC-EAD data indicated that the pheromone extracts contained 7Z,11Z-16:Ald, and we reasoned that the putative trienal, with a conjugated diene present, must either be a 5,7,11- or a 7,11,13-hexadecatrienal. Thus, we initially focused on short, nonstereoselective syntheses that would provide mixtures of isomers for comparisons of retention times with those of the insect-produced compound. These compounds enabled us to eliminate the 5,7,11-isomers, and confirmed that one of the 7,11,13-isomers matched the insect produced triene, so stereoselective syntheses were developed. The syntheses of both pairs of (5E/Z,7E,11Z)-hexadecatrienals 11 and (5E/ Z,7Z,11Z)-hexadecatrienals 16 started with the preparation of the key intermediate 4 (Scheme 1). Thus, (Z)-3-octen-1-ol 1 was transformed into its corresponding tosylate 2 (83% yield), which was then coupled with the terminal alkyne 3 followed by deprotection to give alcohol 4 in 31% yield in two steps. Enyne alcohol 4 was reduced to the corresponding (E)- and (Z)-allylic alcohols 7 and 12 by reaction with lithium aluminum hydride or P-2 Ni and H2, respectively. The alcohols were treated with PBr3 to give bromides 8 and 13. To introduce the third double bond into the alkyl chain, the bromides 8 and 13 were converted into the phosphonium salts 9 and 14, followed by Wittig reactions of the corresponding ylides with protected hydroxyaldehyde 6. Deprotection of the resulting trienes with PTSA in methanol gave (5E/Z,7E,11Z)-hexadecatrien-1-ol 10 (71% yield over two steps, 42:58 ratio of isomers) and (5E/Z,7Z,11Z)hexadecatrien-1-ol 15 (75% yield, 55:45 ratio of isomers). Oxidation of alcohols 10 and 15 with PCC in dichloromethane gave the desired aldehydes 11 and 16 as pairs of isomers. The synthesis of the (5E/Z,7Z,11Z)-hexadecatrienal mixture 16 gave an approximately 1:1 ratio of isomers. To determine the identities of the two isomers, one of which eluted very close to the insect-produced trienal, a stereospecific synthesis of the 5Z,7Z,11Z-isomer was carried out (Scheme 2). The key steps were the sequential alkylation of (Z)-1,2-dichloroethene 19 with the terminal alkyne 18 in the presence of CuI, PdCl2(PPh3)2, and diisopropylamine, followed by a second alkylation of the resulting chloroenyne 20 with (Z)-3octenyl magnesium bromide with Fe(acac)3 catalysis and N-methylpyrrolidi-

188

MOREIRA ET AL.

none (NMP) as cosolvent (Cahiez and Avedissian, 1998). The latter coupling step was inefficient, returning mainly unreacted starting material, but it did furnish 21 in sufficient yield (13.5%) to proceed. Dienyne 21 was stereoselectively reduced to the corresponding triene 22 (76% yield) with dicyclohexylborane, in >99% isomeric purity by GC analysis. Removal of the THP group by acid catalysis in methanol gave trien-1-ol 23, which was oxidized to (5Z, 7Z,11Z)-hexadecatrienal 24 with PCC. Although (5Z,7Z,11Z)-hexadecatrienal had retention times on both columns that were close to those of the insect-produced compound (Table 1), none of the four 5,7,11-isomers had retention times that exactly matched those of the insect-produced trienal on the two columns. Thus, we embarked on the syntheses of the other most likely isomers, the 7,11,13-trienals. Initially, four of the eight possible geometric isomers of 7,11,13-hexadecatrienal were synthesized as two pairs of isomers, that is, (7Z,11E/Z,13E)-hexadecatrienal and (7Z,11E/Z,13Z)-hexadecatrienal, starting from commercially available (7Z,11E/Z)-hexadecadienyl acetate 25 (pheromone of pink bollworm) (Scheme 3). Thus, oxidative cleavage of 25 gave aldehyde 28 (2 steps, 10.1% yield), which was reacted separately with the ylides generated from the phosphonium salts 30 and 34. These reactions yielded the mixtures of trienes (7Z,11E/Z,13Z)-31 and (7Z,11E/Z,13E)-35 in 42 and 58% yield, respectively. After hydrolysis of the acetates, the corresponding alcohols 32 and 36 were oxidized with PCC, furnishing the desired mixtures of aldehydes 33 and 37. With these mixtures in hand, the insect-produced trienal was identified as the 7Z,11Z,13E-16:Ald isomer. A more stereoselective route was developed to provide material for field trials, along with a parallel route to provide the diene component, 7Z,11Z-16:Ald, in high isomeric purity as well. The synthetic routes to 7Z,11Z,13E-16:Ald 49 and 7Z,11Z-16:Ald 56 are shown in Schemes 4 and 5. The key step in both routes was a stereoselective Wittig reaction that placed a (Z) double bond in the 11 position of the desired compounds. The synthetic route to 7Z,11Z,13E-16:Ald 49 was based on a (C8 + C3) + C5 approach. Thus, alkyne 41 (Waanders et al., 1987), prepared from 7octyn-1-ol 40 by protection of the hydroxyl group as the tetrahydropyranyl (THP) ether, was deprotonated with BuLi, followed by addition of bromide 39 to give diprotected yne-diol 42. Selective removal of the t-butyldimethylsilyl ether with tetrabutylammonium fluoride furnished monoprotected yne-diol 43 (Yadav et al., 1988) (70% over two steps). The alkyne was reduced by P-2 nickel-catalyzed reduction of 43 with H2 (Brown and Ahuja, 1973) to produce 44 (Sharma et al., 1994) in 92% yield. Alcohol 44 was converted to the corresponding iodide 45 by reaction with iodine, triphenylphosphine, and imidazole (Gna¨dig et al., 2001) (58%), followed by reaction of 45 with triphenylphosphine to give the corresponding phosphonium salt. Wittig reaction of ylide 46 derived from this salt with (E)-2-pentenal in a THF–DMPU solvent

PHEROMONE OF CITRUS LEAFMINER

189

mixture, in a modification of the Z-selective olefination method of Labelle et al. (1990), afforded triene 47 as a 93:7 mixture of the 7Z,11E/Z,13,E-trienes in 42% yield. The Labelle et al. (1990) method called for a THF–HMPA solvent mixture, whereas our results showed that DMPU was an effective and safer substitute for toxic and carcinogenic HMPA. Trienol 48 was obtained by removal of the THP protecting group with PTSA in methanol, and the isomeric purity of 48 was increased to 97:3 by recrystallization from hexane. Oxidation of 48 with pyridinium chlorochromate (PCC) then produced the desired aldehyde 49 in 77% yield. The synthesis of 7Z,11Z-16:Ald was carried out by a modification of this route (Scheme 5). Thus, alkyne 41 was treated with BuLi followed by addition of 1-chloro-3-iodopropane to give chloride 50 (Sonnet, 1974) (64% yield). This was converted to the corresponding iodide 51 by refluxing with sodium iodide in acetone (95% yield). Reaction of 51 with triphenylphosphine gave the corresponding phosphonium salt 52, which can be used as an intermediate for the syntheses of both the diene and the triene pheromone components. Wittig reaction of the ylide derived from 52 via treatment with lithium hexamethyldisilazide in THF–DMPU solvent gave the Z olefinic compound 53 (Yadav et al., 1988) in 65% yield as a 98.8Z:1.2E isomeric mixture. After removal of the THP group with PTSA in methanol, alcohol 54 (Yadav et al., 1988) was reduced with P-2 nickel and hydrogen, yielding dienol 55 in 71% yield. 7Z,11Z-16:Ald 56 was obtained by oxidation of 55 with PCC as before. The synthetic route for the preparation of 7Z,11Z,13Z-16:Ald 62 is depicted in Scheme 6. Phosphonium salt 52, previously used to prepare 7Z,11Z-16:Ald 56, was deprotonated with lithium hexamethyldisilazide in THF–DMPU, and reacted with 2-pentynal to give protected endiynol 59 (58%), which was then stereoselectively reduced to the corresponding triene 60 with dicyclohexylborane (80%), yielding a 92:8 mixture of the (7Z,11Z,13Z)- and (7Z,11E,13Z)trienes, respectively. After removal of the THP protecting group with PTSA in methanol (80%), the isomeric purity of trienol 61 was increased to 95:5 by chromatography on silica gel impregnated with 10% AgNO3. Oxidation of 61 with PCC completed the synthesis, giving aldehyde 62 in 73% yield. Field Trials. A preliminary, unreplicated field trial run for 1 d provided evidence that both 7Z,11Z-16:Ald and 7Z,11Z,13E-16:Ald were required for attraction of male moths, because traps baited with 100 mg of either component alone attracted no moths, whereas traps baited with 100:10, 100:50, or 100:100 mg ratios of 7Z,11Z,13E-16:Ald to 7Z,11Z-16:Ald caught 135, 192, and 146 moths, respectively. A subsequent, replicated field trial testing different blend ratios of the two components caught >13,000 moths, with a 3:1 ratio of the trienal and dienal being most attractive (Figure 3). A dose-response trial of the 3:1 ratio showed that trienal doses of 100–1000 mg were more attractive than lower doses of 1.0–33 mg/lure (Figure 4). Many thousands of moths were

190

MOREIRA ET AL.

FIG. 3. Citrus leafminer field trial to determine the optimum ratio of the trienal to the dienal, September 8–12, 2004. Each septum was loaded with 100 mg Z7,Z11,E13–16:Ald and variable amounts of Z7,Z11–16:Ald. Bars with different letters are significantly different (two-way ANOVA followed by Student–Neumann–Keuls test, a = 0.05, F = 180.4, df = 4, 24, P < 0.001). Total moths trapped = 13,915. Solvent controls and the 1 mg dose of Z7,Z11–16:Ald each caught 1 moth, and were not included in the statistical analysis.

FIG. 4. Effect of pheromone dose on trap catches of citrus leafminer, October 15–22, 2004. The trienal/dienal ratio was fixed at 3:1. Bars with different letters are significantly different (two-way nonparametric ANOVA followed by Bonferroni’s t-tests for means separation, a = 0.05). For treatment, F = 22.92, df = 6, 34, P < 0.001; for block effect, F = 0.0, df = 4, 34, P = 1.00. Total number of moths trapped = 13,122. Control traps caught two moths, and were not included in the statistical analysis.

191

PHEROMONE OF CITRUS LEAFMINER

FIG. 5. Citrus leafminer blend ratio trial, June 2005. The trienal/dienal ratio was fixed at 100:33 mg, and variable amounts of 7Z,11Z,13Z–16:Ald were added to this base blend. Bars surmounted with different letters are significantly different (two-way ANOVA on log10(x + 1) transformed data, followed by Student–Neumann–Keuls test, a = 0.05). For treatment, F = 44.0, df = 5, 29, P < 0.001; for block, F = 1.93, df = 4, 29, P = 0.14. Total moths trapped = 617. Control traps caught no moths, and were not included in the statistical analysis.

caught, again indicating the effectiveness of the lure. The addition of increasing amounts of 7Z,11Z,13Z-16:Ald to the two component blend of 7Z,11Z,13E16:Ald and 7Z,11Z-16:Ald (100:33) resulted in decreasing numbers of moths caught, indicating that this isomer was inhibitory at higher percentages of the blend (Figure 5).

DISCUSSION

Solvent extracts of pheromone glands were difficult to prepare because of the tiny size of these moths, and the pheromone components were not detected in these extracts by GC or GC-MS. In contrast, the SPME method of dynamic collection of pheromone from the effluvia from live female moths provided extracts in which the amounts of the pheromone components were detectable by GC and GC-MS, and these provided an accurate representation of the compounds released by the moths. However, even with overnight SPME collections made from groups of as many as 30 moths, and with the efficiency inherent in the desorption of the entire SPME sample directly into the injection port of the GC or GC-MS, we frequently could not detect the pheromone

192

MOREIRA ET AL.

components with either of these instruments. Thus, the preliminary identification of the two compounds was based solely on the combination of data from relative retention times on two GC columns of different polarities determined from EAD responses (Table 1), the relative sizes of antennal responses to different functional groups and double bond orientations, and latterly, the bioassay results. The sum total of these data presented a compelling case for the identification of the pheromone components. Confirmation of the pheromone structures was finally acquired after the field trials, when mass spectra of the two compounds were obtained from an SPME sample obtained from aeration of female moths for several sequential nights. The analytical results and the field trials conclusively demonstrate that the citrus leafminer pheromone for California populations of this moth consists of two major synergistic components, with no attraction at all to either compound alone. This explains why researchers in several areas of the world had not been able to reproduce the results of Ando et al. (1985), who reported attraction of several hundred moths of a Japanese population of this species to 7Z,11Z16:Ald as a single component. Furthermore, the marked difference between the attractiveness of 7Z,11Z-16:Ald between populations in Japan and other areas of the world suggests that the Japanese population may constitute a geographically distinct pheromone race, as has been reported for other lepidopteran species (e.g., turnip moth, Agrotis segetum, To´th et al., 1992; moths in the genus Hemileuca, McElfresh and Millar, 2002). The GC-EAD traces from the DB-5 column showed occasional weak antennal responses to another trace compound, tentatively identified as 7Z,11Z,13Z-16:Ald. However, addition of this isomer to the optimal twocomponent blend did not increase trap catches, and at higher doses, was actually inhibitory. Given that our field bioassays indicated that the two-component blend of 7Z,11Z,13E-16:Ald and 7Z,11Z-16:Ald was highly attractive to male moths, with many thousand moths being caught in each of the two bioassays in fall of 2004, it seems unlikely that this compound would be an important component of the pheromone blend. To our knowledge, the citrus leafminer is the only lepidopteran species known to use 7Z,11Z-16:Ald, although the corresponding acetate has been reported as a sex attractant or sex pheromone component for one stathmopid, eight gelechiid, and one noctuid species (Witzgall et al., 2004). This is the first report of the trienal component, 7Z,11Z,13E-16:Ald, and to our knowledge, the 7Z,11Z,13E-16:X structural motif has not been reported before from any lepidopteran pheromone or other natural source. However, this may be a reflection of the few Phyllocnistis or related leafminer species’ pheromones studied rather than a unique structural motif in nature. In summary, this new pheromone blend will find immediate use in monitoring citrus leafminer populations, particularly in areas such as California

193

PHEROMONE OF CITRUS LEAFMINER

where the insect has been introduced fairly recently, and where it is still expanding its range. Acknowledgments—We thank Mariana Krugner for technical assistance, and the Citrus Research Board of California for financial support for this project.

REFERENCES ABRAMS, S. R. and SHAW, A. C. 1988. Triple bond isomerizations: 2-decyn-1-ol to 9-decyn-1-ol. Org. Synth. 66:127–131. ANDO, T., TAGUCHI, K. Y., UCHIYAMA, M., UJIYE, T., and KUROKO, H. 1985. (7Z,11Z)-7,11Hexadecadienal: sex attractant of the citrus leafminer moth, Phyllocnistis citrella Stainton (Lepidoptera: Phyllocnistidae). Agric. Biol. Chem. 49:3633–3635. BEEVOR, P. S., CORK, A., HALL, D. R., NESBITT, B. F., DAY, R. K., and MUMFORD, J. D. 1986. Components of female sex pheromone of cocoa pod borer moth, Conopomorpha cramerella. J. Chem. Ecol. 12:1–23. BROWN, C. A. and AHUJA, V. K. 1973. Catalytic hydrogenation. VI. The reaction of sodium borohydride with nickel salts in ethanol solution. P-2 nickel, a highly convenient, new, selective hydrogenation catalyst with great sensitivity to substrate structure. J. Org. Chem. 38:2226–2230. CAHIEZ, G. and AVEDISSIAN, H. 1998. Highly stereo- and chemoselective iron-catalyzed alkenylation of organomagnesium compounds. Synthesis 1998:1199–1205. CHEN, X. and MILLAR, J. G. 2000. Preparative scale syntheses of isomerically pure (10E,12E,14Z)and (10E,12E,14E)-hexadeca-10,12,14-trienals, sex pheromone components of Manduca sexta. Synthesis 2000:113–118. GARRIDO, A. and JACAS, J. 1996. Differences in the morphology of male and female pupae of Phyllocnistis citrella (Lepidoptera: Gracillariidae). Fla. Entomol. 79:603–606. GNA¨DIG, S., BERDEAUX, O., LOREAU, O., NOE¨L, J. P., and SE´BE´DIO, J. L. 2001. Synthesis of (6Z,9Z,11E)-octadecatrienoic and (8Z,11Z,13E)-eicosatrienoic acids and their [1-14C]radiolabeled analogs. Chem. Phys. Lipids 112:121–135. HEPPNER, J. B. 1993. Citrus Leafminer. Entomology Circular #359. Florida Dept. of Agriculture and Consumer Services, Gainesville, FL. JACAS, J. A. and PEN˜A, J. E. 2002. Calling behavior of two different field populations of Phyllocnistis citrella (Lepidoptera: Gracillariidae): effect of age and photoperiod. Fla. Entomol. 85:378–381. KERR, D. E., KISSINGER, L. F., and SHOYAB, M. 1990. Cyclohexane diester analogs of phorbol ester as potential activators of protein kinase C. J. Med. Chem. 33:1958–1962. LABELLE, M., FALGUEYRET, J. P., RIENDEAU, D., and ROKACH, J. 1990. Synthesis of two analogues of arachidonic acid and their reactions with 12-lipoxygenase. Tetrahedron 46: 6301–6310. LEGASPI, J. C. and FRENCH, J. V. 2003. The citrus leafminer and its natural enemies. http:// primera.tamu.edu/kcchome/pubs/leafminer.htm. MCELFRESH, J. S. and MILLAR, J. G. 2002. Geographic variation in the pheromone system of the saturniid moth Hemileuca eglanterina. Ecology 82:3505–3518. MILLAR, J. G., KNUDSON, A. E., MCELFRESH, J. S., GRIES, R., GRIES, G., and DAVIS, J. H. 1996. Sex attractant pheromone of the pecan nut casebearer (Lepidoptera: Pyralidue). Bioorg. Med. Chem. 4:331–339.

194

MOREIRA ET AL.

SHARMA, M. L., CHAND, T., and VERMA, S. 1994. Synthesis of (Z)-7-hexadecen-1-yl acetate and (Z)-7-hexadecenal. Proc. Indian Acad. Sci., Chem. Sci. 106:45–48. SONNET, P. E. 1974. A practical synthesis of the sex pheromone of the pink bollworm. J. Org. Chem. 39:3793–3794. TO´TH, M., LOFSTEDT, C., BLAIR, B. W., CABELLO, T., FARAG, A. I., HANSSON, B. S., KOVALEV, B. G., MAINI, S., NESTEROV, E. A., PAJOR, I., SAZONOV, A. P., SHAMSHEV, I. V., SUBCHEV, M., and SZO¨CS, G. 1992. Attraction of male turnip moths Agrotis segetum (Lepidoptera: Noctuidae) to sex pheromone components and their mixtures at 11 sites in Europe, Asia, and Africa. J. Chem. Ecol. 18:1337–1347. WAANDERS, P. P., THIJS, L., and ZWANENBURG, B. 1987. Stereocontrolled total synthesis of the macrocyclic lactone (j)-aspicilin. Tetrahedron Lett. 28:2409–2412. WITZGALL, P., LINDBLOM, T., BENGTSSON, M. and TO´TH, M. 2004. The Pherolist. http:// www.pherolist.slu.se. YADAV, J. S., UPENDER, V., SHEKHARAM, T., and REDDY, E. R. 1988. A practical synthesis of sex pheromones of pink bollworm: gossyplure. Indian J. Chem., Sect. B 27B:1012–1014.