(Z)-4-Tridecenal, a pheromonally active air oxidation product from a series of (Z,Z)-9,13 dienes inMacrocentrus grandii Goidanich (Hymenoptera: Braconidae)

Journal of Chemical Ecology, 1Iol. 18, No. 11, 1992

(Z)-4-TRIDECENAL, A PHEROMONALLY ACTIVE AIR OXIDATION PRODUCT FROM A SERIES OF (Z,Z)-9,13 DIENES IN Macrocentrus grandii GOIDANICH (HYMENOPTERA: BRACONIDAE)

P A U L D. SWEDENBORG* and R I C H A R D L. JONES

Department of Entomology University of Minnesota St. Paul, Minnesota 55108 (Received December 12, 1991; accepted June 17, 1992) Abstract--(Z)4-Tridecenal was identifiedas a sex pheromone componentof Macrocentrus grandii, a larval parasitoid of the European corn borer. The aldehyde was found to be a common air oxidation product from a series of (Z,Z)-9,13-dienes of 27-33, 35, 37, 39, and 41 carbonatoms in female wasps. Synthetic (Z)-4-tridecenaland (Z,Z)-9,13-heptacosdieneeach elicited flight initiation,upwind anemotaxis, casting, landingon the source, walking, wing fanning, and copulatory attempts by male wasps in a wind-tunnel.Field studies demonstrated that both compounds are synergizedby a more polar component. Synthetic(E)-4-tridecenalwas not attractive. Key Words--Pheromone, hydrocarbon, diene, (Z)-4-tridecenal, oxidation, Hymenoptera, Braconidae, Macrocentrus grandii, parasitoid.

INTRODUCTION

Macrocentrus grandii Goidanich is a larval parasitoid of the European corn borer, Ostrinia nubilalis (Hfibner). In a previous report we described evidence of a female-produced multicomponent sex pheromone for M. grandii (Swedenborg and Jones, 1992). These components occurred in the hexane, 5 % ether in he• (5E), and 50 % ether in hexane (50E) fractions eluted from Florisil. In this paper we report the isolation, identification, and synthesis of two pheromone components originating from the hexane and 5E fractions. In addition, *To whom correspondenceshould be addressed. 1913 0098 -0331/92/l 100-1913506,50/0 9 1992 Plenum Publishing Corporation

1914

SWEDENBORG AND JONES

we report evidence for a second pheromone component occurring in the 5E fraction. METHODS AND MATERIALS

Wasps, Extracts, and Florisil Fractions. M. grandii and O. nubilalis were reared according to Ding et al. (1989) and Guthrie et al. (1971), respectively. Handling, extraction, and initial purification of the extracts using Florisil were as described by Swedenborg and Jones (1992). Voucher specimens have been placed in the University of Minnesota insect museum. Chromatography. Open-column argentation chromatography was conducted using a 9-in. Pasteur capillary pipet packed with 5 cm of 20% silver nitrate on silica gel (60/200 mesh). The column (void volume ca. 1 ml) was eluted first with 2 ml hexane, followed by 2 ml of 10% ether in hexane. The sample was the original Florisil hexane fraction (Swedenborg and Jones, 1992). Similar chromatography was conducted using Florisil (2.5 % water by weight). This column was eluted consecutively with aliquots of hexane then 2.5, 5.0, 7.5, 10.0, and 50.0% ether in hexane, ether, acetone, and methanol. The samples were the products of diene oxidation and chemical syntheses. All fractions were capped under nitrogen and stored at - 7 0 ~ until used for bioassay or analysis. High-pressure liquid chromatography (I-IPLC) was done isocratically with a Waters Associates M6000A pump and a 11401 differential refractometer detector. Columns used were a Serva 250-ram x 4.6-mm-ID Servachrome silica gel column coated in situ with 20% silver nitrate, as described by Heath and Sonnet (1980), and an Altex 300-ram x 8-mm-ID t~-Sphemgel 50 ,~, size-exclusion column. Samples on the Serva Column were eluted with 25 % toluene in hexane (hydrocarbon analysis) or 100% toluene (aldehyde analysis) at 2.0 ml/min. Samples on the Altex column (aldehyde analysis) were eluted with 25 % ether in hexane at t.0 ml/min. Gas chromatography (GC) was performed with Hewlett Packard 5830A and 5890A instruments, both equipped with flame ionization detectors. The 5830A was modified to contain an effluent splitter and thermal gradient collector (similar to that described by Brownlee and Silverstein, 1968). Columns used were: glass columns, 1.9 m x 2 mm or 4 mm ID packed with 3% Dexsil 300 on 80/100 Gas Chrom Q for hydrocarbon and aldehyde analysis, collection, and hydrocarbon equivalent chain length (ECL) determination; a bonded fused silica column, Supelco Supelcowax 10, 30 m x 0.53 mm ID, for aldehyde analysis and collection; a bonded fused silica capillary column, J & W DB-1, 30 m x 0.32 mm ID, for hydrocarbon and aldehyde analysis; and a fused silica capillary column, Quadrex SLP-Silar-10C 50 m x 0.25 mm ID, for aldehyde double bond configuration. Key GC temperature programs are noted with results.

(Z)~-TRIDECENAL

1915

Spectroscopy. Coupled gas chromatography-mass spectroscopy (GC-MS) was carried out with either an LKB-9000 instrument using a 0.6-m x 2-mm-ID 3% Dexsil 300 column and electron ionization energy of 20 eV or 70 eV, or a Kratos MS25 instrument using a J & W DB-1 15-m x 0.53-tumID column. The Fourier-transform infrared (FT-IR) spectrum was taken with a Nicolet 170SX spectrometer. The IR spectrum was run with ca. 2 mg of the HPLCpurified diene material held between two KBr 25-mm x 4-mm disks. Bands were reported in wave numbers (cm-1). Chemical Tests. Ozonolysis was done as described by Beroza and Bierl (1967). Samples (0.1-2 #g of diene material) were ozonized in 20 #1 of carbon disulfide for 2 min at - 7 0 ~ The ozonides were then reduced with triphenylphosphine to aldehydes. Subsequently, ozonized samples were run on the GC or GC-MS to identify reaction products. For molecular sieve tests, samples (1-2/zg of diene material) were evaporated under nitrogen then resolubilized in isooctane containing activated 5 ,~ molecular sieve pellets to separate branched from straight-chain compounds (Hutchins and Martin, 1968). Control samples included the dienes but no molecular sieve pellets. Twenty-four hours later the sample solutions were analyzed by GC for changes in diene concentration. For diene oxidation tests, six samples (20/~g each) of natural and synthetic (Z,Z)-9,13-heptacosadiene, after purification by Florisil, HPLC, and GC, were solubilized in hexane and placed on 6-cm watch-glass plates. Three replicates of each source were placed on the lab bench under fluorescent ceiling lights and three were placed in an ultra-low freezer - 7 0 ~ After 24 h, the watch-glass plates were each rinsed with 1 ml hexane. Each replicate was then chromatographed through a Florisil column (to separate dienes from diene oxidation products) followed by GC or GC-MS to look at reaction products and bioassayed for activity. In addition, a similar oxidation was performed with 200/xg of the entire diene series. Synthesis of (Z,Z)-9,13-Heptacosadiene. Diene synthesis was guided by the procedure of Sonnet (1974) utilizing a Wittig condensation with the cosolvent, dimethyl sulfoxide. A procedural change involved the preparation of the alkyltriphenylphosphonium salt from 1,4-dibromobutane, which, upon simultaneous addition of nonanal and tetradecanal, provided for double-bond formation on each end of the four-carbon precursor (Bartelt et al., 1986). Foliowing synthesis, the dienes were separated from polar by-products on Florisil and 9,13heptacosadiene was separated from the nonpolar symmetrical by-products of 9,13-docosadiene and 14,18-dotriacontadiene using gas chromatography. HPLC, using a silver nitrate-coated silica gel column, eluted with 25 % toluene in hexane, was used to provide geometrical purity of the synthetic dienes (Bartelt et al., 1982). All starting chemicals were obtained from Aldrich Chemical Co. Inc., Milwaukee, Wisconsin, and used without further purification.

1916

SWEDENBORGAND JONES

Synthesis of (Z)-4- Tridecenal and (E)-4- Tridecenal. (Z)-4-Tridecenyl acetate and (E)-4-tridecenyl acetate were purchased from Bedoukian Research, Inc., Danbury, Connecticut. The isomers were base hydrolyzed to their respective alcohols using KOH in 95 % aqueous ethanol. Following hydrolysis, the ethanol was distilled off, and the sample was diluted with methylene chloride and dried over 4 A molecular sieves. Without further purification, the alcohols were oxidized with chromium trioxide and pyridinium chloride to provide (Z)4-tridecenal and (E)-4-tridecenal (Svirskaya et al., 1980). Following synthesis, the aldehydes were purified using Florisil, HPLC, and GC. Laboratory and Field Bioassay. Bioassays were conducted as described by Swedenborg and Jones (1992). For the laboratory bioassay a wind-tunnel was utilized with treatments on 4-cm watch-glass plates at the upwind end of the tunnel. About 150-200 males were used for a bioassay. Male behaviors in the bioassay included flight initiation, upwind anemotaxis, casting, landing on the source, walking, wing fanning, and mating attempts between males. Visits at the source were usually brief, and the bioassay was quantified by recording the total number of males landing on each plate during a 1-min period. Open column silver nitrate fractions were tested at 1/3 female equivalent (FE), HPLC fractions were tested at 5 FE, and GC fractions were tested at 1/3 FE (HPLC purified dienes) or 5 FE (diene oxidation products and HPLC purified 5E). All synthetic and female-derived dienes were tested at a 100-ng dosage (the total diene material extracted from one female using hexane is ca. 9 /~g), and all sources of 4-tridecenal were tested at a 17-pg dosage [the total (Z)-4-tridecenal extracted from one female using hexane is ca. 50 pg]. For pairwise comparisons, data were analyzed by a one-sided paired t test. Chemical positions were exchanged for each replicate. Three-way comparisons were considered a randomized complete block design with each bioassay being a block. Data were transformed (~/x + 1 ) to stabilize variance and tested for homogeneity with Levene's test. An analysis of variance (ANOVA) was then applied (degrees of freedom for error = 4) with mean separations by Fisher's (protected) least significant difference (PLSD), c~ _< 0.05 (Steel and Torrie, 1980). In two separate field tests ca. 1200 and 800 laboratory-reared males were released into a test plot of hybrid field corn. Traps (Pherocon 1C) were secured to the corn stalk 1.0-1.5 m above ground. Chemical extract preparations were placed on 4-cm watch-glass plates centered inside the trap bottom. A Hercon polymer-bonded slow-release wafer formulation (0.15 mg isomer/cm z) of (Z)4-tridecenal and (E)-4-tridecenal was used at target release dosages of 50 ng/hr each for test 1 (release rates not verified). The 50E fraction was applied at 15 FE for test 1 and 20 FE for test 2. Synthetic (Z,Z)-9,13-heptacosadiene was applied at 20 FE or 180 tzg diene for test 2. Each test was initiated near midday and run for 24 hr. Wasps were released immediately after the treatments were applied. Field data were transformed (~/x + 1 ) to stabilize variance and tested

(Z)~4-TRIDECENAL

1917

for homogeneity with Levene's test. A n A N O V A was applied with mean separations by PLSD, c~ __ 0.05.

RESULTS

Female-Derived Florisil Hexane Fraction. Bioassays of the Florisil hexane fraction were consistently active in laboratory and field studies (Swedenborg and Jones, 1992). This fraction (475 FE) was concentrated to 100/zl and chromatographed on a pipet column of 20 % silver nitrate on silica gel to separate paraffins from olefins. The active fraction was the 10% ether in hexane fraction, which contained the olefins (Table 1; test 1). The olefin fraction was comparable in activity to the original Florisil hexane fraction (Table 1; test 2), and activity was not improved by addition of the hexane- or paraffin-containing fraction (Table 1; test 3). Subsequently, the olefin fraction was chromatographed using HPLC on a silver nitrate-coated silica gel column eluted with 25 % toluene in hexane to separate monoenes from dienes (Figure 1). Bioassay activity coincided with fractions 5-9. GC and/or GC-MS analysis of the HPLC fractions identified the

TABLE 1. LANDING RESPONSES OF MALE M. grandii TO "SILVER NITRATE" OPEN COLUMN FRACTIONS OF VIRGIN FEMALES IN LABORATORY WIND-TUNNEL

Treatmenta

Meanb

Test 1 10% ether in hexane Hexane Control

35.0a 0.3b 0.3b

Test 2 10% ether in hexane Original Florisil hexane Control

35.0a 29.3a 0.0b

Test 3 10% ether in hexane Hexane + 10% ether in hexane Control

35.3a 39.3a 0.0b

aThe pamfiin and olefin portions of the original Florisil hexane fraction are represented by the hexane and 10% ether in hexane treatments, respectively. bThree replicationswith each replicate using a new group of 150-200 untested males 1-5 days otd; counts represent the mean number of males landing on each plate during a l-rain test; all fractions were bioassayedat 1/3 FE; in each test columnmeans followedby differentletters are significantly different (PLSD, c~ _< 0.05).

1918

SWEDENBORGAND JONES

SOLVENT FRONT

and MONENES

0

DIENES

INJECT

FRACTION

l

3

0

4

I

5

5

I

6

I

7

|

8

10

I

9

10

15

I 20

ELUTION V O L U M E (ml 25% T O L U E N E IN H E X A N E ) DATA: FRACTION

1

2

3

4

5

RESPONSE

0.0

0.0

0,0

0.5

18.4"*

6 3 1 , 3 *~

7 16.5"*

8 19.7""

9 4.6""

1 0 1.2

FIG. 1. HPLC chromatogram using a Serva ServaChrom silica gel column coated in situ with silver nitrate. Sample represents 425 FE from the olefin fraction. Fractions were collected every 2 mL The data represent the mean landing response of male M. grandii to each HPLC fraction in a laboratory wind-tunnel during a l-rain interval. Each treatment plate was paired with a solvent-only control; control plates paired with HPLC fractions 1 and 5-10 hand zero landings, plates 2-4 had 0.2 mean landings. 150-200 males 1-5 days old were tested; five replications with each replicate representing a new group of untested males; all HPLC fractions were bioassayed at 5 FE; * and ** imply significant differences from controls at the c~ _< 0.05 and e~