Journal of Chemical Ecology, VoL 16, No. 12, 1990

MALE PHEROMONE OF SWIFT MOTH, Hepialus hecta L. (LEPIDOPTERA: HEPIALIDAE)

S T E F A N S C H U L Z , 1 W I T T K O F R A N C K E , 1'6 W I L F R I E D A. K O N I G , 1

V O L K E R S C H U R I G , 2 K E N J I M O R I , 3 R O L F K I T T M A N N , 4 and

D I E T R I C H S C H N E I D E R 5

t lnstitut far Organische Chemie der Universitat Hamburg Martin-Luther-King-Platz 6

D-2000 Hamburg 13, Germany

2Universitiit Tabingen, lnstitut far Organische Chemie Auf der Morgenstelle 18

D-7400 Tabingen, Germany

3Department of Agricultural Chemistry University of Tokyo Tokyo 113, Japan

4Fakult~it far Biologie, Universitiit Konstanz Universiti~tsstr. 10

D-7750 Konstanz 1, Germany

5Max-Planck-lnstitut far Verhaltensphysiologie D-8130 Seewiesen (Starnberg), Germany

(Received March 5 1990; accepted July 30, 1990)

Abstraet--(R)-6-Ethyl-2-methyl-2,3-dihydro-4H-pyran-4-one, (1R,3S,5R)-3- ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1]non-7-ene, and (1R,3S,5R)-3- ethyl- 1,8-dimethyl-2,9-dioxabicyclo[3.3.1 ]non-7-en-6-one represent the main components in the male pheromone of the swift moth, Hepialus hecta. The amounts of the three components were 40, 5, and 5/zg per male, respectively. Structure elucidation of the compounds was based on spectroscopic data as compared to synthetic reference samples. The absolute configurations were determined by gas chromatography on chiral stationary phases; optically active samples served as reference compounds. Electrophysiological and behavioral experiments with natural material and synthetic samples clearly showed the three heterocyclic compounds to act as pheromones. (E,E)-a-Famesene rep- resents the main component of the scent secretion of male Hepialus humuli.

Key Words--Hepialus, Lepidoptera, Hepialidae, swift moth, pheromone, dihydropyrone, 2,9-dioxabicyclo[3.3.1] nonane, behavior, electrophysiol- ogy.

6To whom correspondence should be addressed.

3511

0098-0331/90/1200-3511506.00/0 �9 1990 Plenum Publishing Corporation

3512 ScnuLz ET AL.

INTRODUCTION

Swift moths--Hepialidae--are a primitive group of Lepidoptera. Males of a number of species in this family possess conspicuous tibial scent organs on their hind legs. In H. hecta, a common species in middle Europe, these organs and their impressive fruity odor have been known since the last century (Bertkau, 1882). The tibiae of the hind legs are enlarged, pear-shaped, and bear a great number of scent scales, while the tarsus is reduced and without function. During their striking pendular flights, the males erect these scales and emit the odor, which for a long time was believed to attract the females (for a review of the biology of Hepialus see Mallet, 1984), but until now, no behavioral or electro- physiological proof of the function of this secretion as a male sexual attractant has been reported.

The elements of male androconial scent organs in the Lepidoptera (Boppr6, 1984; Boppr6 and Schneider, 1989) are the hair-forming trichogen cells, which in this case are glandular cells as well. The secretory products then are trans- mitted on the large surface of the scent scales from which the odor is released.

We now report on structure elucidation and synthesis of the volatile com- ponents that are emitted form the scent organ. In addition, we report on the essence of electrophysiological and behavioral activity, which proves the func- tion of the identified components as pheromones. Details of these biological functions will be published elsewhere.

We also describe the identification of volatile components of similar glands of Hepialus humuli, a larger European relative of H. hecta, showing similar sexual behavior.

METHODS AND MATERIALS

Biological Material. Males of H. hecta (270) and H. humuli (50) were collected during their calling flights at locations in Bavaria, Germany during 1983 and 1984. The hindlegs were removed, immediately stored under n-pen- tane and kept at - 1 8 ~ until work-up. In addition, some samples of H. hecta were stored in vials without solvent, and the liquid secretion was collected with a glass capillary inserted directly into the tibia subsequently analyzed by gas chromatography.

Gas Chromatography. Extracts and crude secretions were analyzed by GC on a Carlo Erba Fractovap 2101 AC gas chromatograph equipped with an FID. A 50-m CP-Sil-8-CB fused silica column (0.32 mm ID, 0.25/~m film thickness) was used, which was programmed from 60~ to 280~ at 3~ Prepara- tive gas chromatography was performed as described earlier (Sinnwell et al., 1985).

Gas chromatographic separations on optically active stationary phases were

SWIFT MOTH PHEROMONE 3513

performed on fused silica capillaries coated with XE-60-L-valine-(S)-o~-phen- ylethylamide (PEA) (50 m, 0.25 mm ID, programmed from 100~ at 1.5~ min) (Krnig et al., 1981) or with Ni(II)-bis(heptafluorobutanoyl-(1R, 5S)-pinan- 4-onate) in OV-1 (50 m, 0.25 mm ID, 80~ isothermal) (Schurig and Wistuba, 1984).

Mass Spectrometry. For mass spectrometric investigation, a GC-MS sys- tem consisting of a Carlo Erba Fractovap 2100 gas chromatograph coupled to a Varian MAT 311 A mass spectrometer was used, which was equipped with the same fused silica column as described above. GC operating conditions were identical.

Synthesis. Racemic 2,6-dialkyl-2,3-dihydro-4H-pyran-4-ones 1, 2, and 3 (Figures 1 and 5 below) were prepared by acylation of/3-ketoesters with c~,/3- unsaturated acylhalides followed by cyclization, saponification, and decarbox- ylation as described in the literature (Gelin and Gelin, 1968). Enantiomerically pure (R)- and (S)-6-ethyl-2-methyl-2,3-dihydro-4H-pyran-4-one was prepared as reported by Moil and Kisida (1986).

Racemic (1R*,3S*,5R*)-3-ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1] non-7-en-9-one 4 was synthesized according to the strategy used by DeShong et al. (1985) for the synthesis of tirandamycine analogs (Scheme 1). 2,3-Di- methylfuran (Rice and Dyer, 1975) was metallated with n-butyl lithium and treated with 3-(2-tetrahydropyranyloxy) pentanal at -78~ The resulting crude trisubstituted furan was oxidized with bromine and treated with methanol to yield the corresponding bisacetal, which upon acidic treatment yielded ketone 4 in 20% yield. The product was purified by distillation (bp 112-125~ and column chromatography (silica gel, methylenechloride-diethyl ether 6:1). Thioacetalization of 4 was accomplished by reaction with 1,3-propanedithiol in the presence of boron trifluoride etherate. The crude thioacetal was desulfurized to yield (1R*,3S*,5R*)-3-ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1]non-7- ene, 7, by heating with potassium hydroxide and hydrazine in triethylenglycol (Georgian et al., 1959). Desulfurization with Raney-Ni produced only mixtures of 7 and the corresponding saturated acetal. Pure 7 was isolated by preparative gas chromatography.

1) Br 2 (CH~SH)2 NH2H4 2) MeOH BF3*Et20 KOH, t~ .3) H*

SCHEME 1. Synthesis of the bicyclic acetals 4 and 7.

3514 SCHtJLZ ET AL.

Pure enantiomers of 4 and 7 were prepared as described previously (Moil and Kisida, 1986). Microreactions of crude extracts or isolated compounds used for structure elucidation were performed according to Francke et al., (1989).

Behavioral Experiments. Investigations were performed in the natural hab- itat, semiwet meadows, during 1984 until 1988 in southern Germany. In June and July, mating behavior, lasting about half an hour at the time around sunset, could be observed. Artificial scent sources consisted of 1 mg of the pure racemic main component 1 on cotton plugs, which were suspended on plant branches. For comparison, freshly cut scent legs and extracts thereof were used. In control experiments, untreated cotton plugs were exposed. The behavior of males and females was observed and recorded. When testing the reactions of females to the exposed scent sources, all free males were removed from the test site prior to the experiments. Full details of the behavior will be published elsewhere.

Electrophysiology. Electroantennograms and single-cell recordings were performed directly after capture of the animals on the same day. Because of the short, thick antennae and the small number and size of the sensillae, electro- physiological measurements proved to be difficult. To carry out the recording, the animal was tethered and its antennae fixed with tungsten wire. For anten- nogram recordings, electrodes filled with hemolymph Ringer solution (Kais- sling and Thorson, 1980) were inserted into the tip and the base of the antenna or attached by means of a drop of dialyzed electrode paste (electrode cream of ECG, Hellige). An identical electrode inserted into the tip of the antenna was used as reference for the single-cell recordings. The other glass microelectrode was filled with receptor lymph Ringer and inserted into the base of single sen- sillum. Stimulation was performed with 1-sec air puffs through glass cartridges with odor sources (scent legs or I cm 2 filter paper with the odorants) and directed onto the antennae.

RESULTS

Chemical Analysis. Comparison between extracts of the whole scent leg of H. hecta and pure liquid secretions revealed no difference in composition except for additional common fatty acids in extracts of the scent leg. For further investigations, scent leg extracts were used for convenience. A gas chromato- gram of an extract is shown in Figure 1. The main components 1, 2, 4, and 7 were identified by isolation, chemical microreactions, mass spectrometry, and [1HI- and [13C]NMR analysis as described earlier (Francke et al., 1985; Sinn- well et al., 1985). Racemic reference samples of compounds 1, 2, 4, and 7 showed the same gas chromatographic retention times and the same mass and NMR spectra as the natural compounds. Structure elucidation of the minor com- ponents 5, 6, 8, and 9 were based on the results of microreactions and their

S W I F T M O T H P H E R O M O N E 3 5 1 5

4 7

rain

x

I I I I [ I I I I I I I 8LO 2 0

J I

FIG. 1. Gas chromatogram of scent gland secretion of H. hecta (numbers according to the text). 1, (R)-6-ethyl-2-methyl-2,3-dihydro-4H-pyran-4-one; 2, (R)-2,6-diethyl-2,3- dihydro-4H-pyran-4-one; 4, (1R,3S,5R)-3-ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1] non-7-en-6-one; 5, 1,3,8-trimethyl-2,9-dioxabicyclo[3.3.1]non-7-en-6-one; 6, 1,3- diethyl-8-methyl-2,9-dioxabicyclo[3.3.11non-7-en-6-one; 7, (1R,3S,5R)-3-ethyl-l,8- dimethyl-2,9-dioxabicyclo[3.3.1]non-7-ene; 8, 1,3,8-trimethyl-2,9-dioxabicyclo[3.3.1] non-7-ene; 9, 1,3-diethyl-8-methyl-2,9-dioxabicyclo[3.3.1]non-7-ene; x, degradation products.

characteristic mass spectra, which were closely related to those of 4 and 7. In addition to these major bicyclic acetals 4 and 7, four minor compounds could be identified, the structures of which were found to be represented by the 2,9- dioxabicyclo[3.3.1]nonane system showing the same substitution pattern as

3516 SCHULZ ET AL.

compared to the major components. Two pairs of compounds could be attrib- uted to structures 4 and 7, respectively: compounds 5 and 8 show a methyl group at positions 1 and 3 while compounds 6 and 9 show ethyl groups at these positions.

Determinations of the absolute configurations of compounds 1, 2, 4, and 7 were carried out by gas chromatography using two different chiral stationary phases. The enantiomers of 1, 2 and 4 could be separated on a XE-60-L-Val- (S)-o~-PEA phase, whereas T was not resolved (Figure 2). On the other hand, 7 could be well resolved by metal complexation gas chromatography on a Ni(II)- bis(heptafluorobutanoyl-(1R,5S)-pinan-4-onate) (Schurig and Wistuba, 1984) phase while 1, 2, and 4 did not elute, most probably due to complexation to the metal (Figure 3).

Gas chromatographic investigations of the hindleg extracts proved the nat- ural products to be of high enantiomeric purity. Upon comparison with synthetic reference samples, 1 was shown to have 6-R-configuration and thus is (R)-6- ethyl-2-methyl-2,3-dihydro-4H-pyran-4-one. Since spiking of the natural extract with racemic 2 enhanced the second-eluting enantiomer, which in the case of 1 was attributed to the 6-R-enantiomer, we concluded natural 2 to be (R)-2,6- diethyl-2,3-dihydro-4H-pyran-4-one. Upon comparison with synthetic refer- ence samples, both natural 4 and 7 proved to show 3-S-configuration, and thus are (1R,3S,5R)-3-ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1]non-7-ene and (1R,3S,5R)-3-ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1]non-7-en-6-one. The amounts of 1, 4, and 7 per individual were about 40, 5, and 5/zg, respectively. Interestingly, the closely related H. californicus was reported to produce an

~ o

, I , I , --~ 30rain 30 min 30min 0

FIG. 2. Gas chromatographic separation of enantiomers of 1, 2, and 4 [50-m fused silica capillary coated with XE-60-L-valine-(S)-ol-phenylethylamide, 0.25 mm ID, pro- grammed from 100~ at 1.5 ~ (A) Synthetic mixture of enantiomers; (B) natural sample; (C) mixture of A and B.

SWIFT M O T H P H E R O M O N E 3517

(3R,5S)

min 3"0 0 ra in 3o 0 ra in 30

Fro. 3. Gas chromatographic separation of enantiomers of 7 [50-m fused silica capillary coated with Ni(II)-bis(heptafluorobutanoyl-(lR,5S)-pinan-4-onate) in OV-1, 0.25 mm ID, 80~ isothermal]. (A) Natural sample, (B) synthetic 3R enantiomer, (C) mixture of A and B.

isomer of 1, (R)-2-ethyl-6-methyl-2,3-dihydro-4H-pyran-4-one, 3, as the only volatile compound in their scent legs (Kubo et al., 1985). In contrast to H. hecta, no bicyclic acetals were found in the scent glands (Kubo, personal com- munication).

In addition to H. hecta, the secretion of the scent legs of H. humuli was investigated. Gas chromatographic analysis of the extracts showed only a single main component, which upon GCMS and comparison of retention times could be identified as (E,E)-c~-farnesene, a widespread naturally occurring sesquiter- pene, known from plants and insects. Small amounts of (E)-~-famesene and (E,Z)- or (Z,E)-c~-farnesene also were found in the extracts (possibly due to decomposition or rearrangement of farnesene in the injection port of the gas chromatograph). No compounds related to 1, 4, or 7 could be detected in the extracts.

Field Tests. The behavior of male and female H. hecta in response to artifical and natural odor sources was observed in the field. Females were attracted when passing the plume of the male odor during their cruising flight through the habitat. Then they flew upwind to the odor source and landed directly on or close to it. As part of the courtship behavior (Turner, 1976), after landing, a wing fluttering of the females was performed close to the odor source. On 20 evenings when scent sources were tested, more than 50 females showed

3518 SCHULZ ET AL.

the described behavior. These reactions of the females could be observed with all odor source that were tested in the field: freshly cut scent legs, or cotton plugs, which were treated with n-pentane extracts of the scent glands or with 1 mg of racemic main component 1. Control cotton plugs showed no attractancy. Nevertheless, the attractancy of odor sources was restricted to a few meters, even with high concentrations.

Male response to the above odor sources is more variable. Only occasion- ally did they fly upwind to the source and touch it. Usually they choose a place at some distance (0.5-2 m) downwind to the odor source and started their pen- dular calling flight (n > 100).

Electrophysiology. Twelve males and 15 females were tested. Electroan- tennograms showed significant responses of male and female antennae to both enantiomers of compounds 1, 4, and 7 (see Figure 4A). The absolute ampli- tudes (mV) of control and odor responses varied with the animals and also slowly changed during the experiments. Nevertheless, relations of response amplitudes remained relatively constant between control and odor stimuli. Due to the short antennae, only small antennogram amplitudes between 2 and 5 mV were recorded even in response to strong odor stimuli, while the control response

a c 1 c 4 c 7

5 s

b

/

1 ,0 -1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

FIG. 4. Electroantennogram recordings. (a) Response of H. hecta females to control (c), and 1000 ~g of compounds 1, 4, and 7 (lower trace). Stimulus duration 1 sec (recorded with an anemometer, upper trace). (b) Dose-response curve from 12 males and 15 females to racemic 1. Response amplitudes are calculated in ratio to control reactions (= 1). Standard deviations are indicated (n > 15).

SWIFT M O T H P H E R O M O N E

1 2 3

3519

4 5 6

7 8 9

FIG. 5. Absolute configurations of identified compounds.

was only two to three times smaller. A dose-response curve for the male and female antennae to racemic compound 1 is given in Figure 4. Amplitudes are given in relation to the control amplitude set to 1.

The threshold amount for all compounds was between 1 and 10 txg, while saturation was reached at 1000/xg or by whole scent legs as well as their extracts. In general, male and female antennae responded similarly to the stimuli.

Single-cell recordings of female H. hecta showed the existence of sensory cells with different sensitivities and reaction patterns to the artificial pheromone components 1, 4, and 7. Some cells showed their best reactivity to component 1 and its optical antipode and were not excited by the components 4 and 7, while the latter induced significant reactions in other types of sense cells. Details on the complex response pattern will be published elsewhere.

D I S C U S S I O N

The results of chemical analysis and behavioral and electrophysiological investigations revealed that pure enantiomers (R)-6-ethyl-2-methyl-2,3-dihy- dro-4H-pyran-4-one, ( 1 R, 3S, 5R) - 3 -ethyl- 1,8-dimethyl-2,9 -dioxabicyclo [3.3.1] non-7-ene, and (1R,3S,5R)-3-ethyl-l,8-dimethyl-2,9-dioxabicyclo[3.3.1]non-7- en-6-one make up the sex pheromone of male H. hecta. These compounds are not reported from other Lepidoptera. Both acetals belong to a new class of biologically active insect volatiles.

3520 SCHULZ ET AL.

Electroantennograms and single-cell recordings demonstrate that all three components are biologically active. Pure compound 1, as well as the whole bouquet, elicits the same sequences of the female mating behavior, with initial attraction and wing fluttering. Nevertheless, the additional components might play a role during the final part of the mating behavior, i.e., between landing of the male and the completion of copulation, or during the ("agonistic"?) behavioral interactions between males (Turner, 1988).

From the biogenetic point of view, dihydropyrone 1 of H. hecta is closely related to the dihydropyrone 3 of H. caIifornicus. Both acetogenins should have the same biogenetic precursor, possibly a triply oxidized C8 unit, which is dif- ferently cyclicized in the two species, yielding the respective dihydropyrones.

While 4 and 7 may share a biosynthetic precursor, there is no obvious biogenetic relation to 1. In contrast to 1, both compounds possess a branched carbon skeleton, which points to a mixed acetate-propionate biosynthesis. This trialkyl-2,9-dioxabicyclo[3.3.1]nonane system is found in nature only seldom, so that a specific informative function of these molecules is more than likely.

In contrast to the rather stereotypical attractance of male moths by their female pheromones, male lepidopteran pheromones show a variety of functions (Boppr6, 1984). Attraction of females--as now found in Hepialus--is usual (see Wunderer et al., 1985). Like other male pheromone systems (Schneider, 1984), the gland of male H. hecta contains an amount of pheromone much higher than found in female calling systems. Nevertheless, the range of attractancy in this calling system is restricted to several meters only.

Acknowledgments--Financial support by the DFG and the Founds der Chemischen Industrie is gratefully acknowledged.

REFERENCES

BERTKAU, P. 1882. Ueber den Duftapparat von Hepialus hectus L. Arch. Naturgesch. 17:223-224. BOPPRI~, M. 1984. Chemically medicated interactions between butterflies, pp. 259-275 in R.I.

Vane-Wright, and P.R. Ackery (eds.). The Biology of Butterflies. Academic Press, London. BOPPRg, M., and SCHNEIDER, D. 1989. The biology of Creatonotos (Lepidoptera: Arctiidae) with

special reference to the androconial system. Zool. J. Linn. Soc. 96:339-356. DESHONG, S., RAMESH, S., ELANGO, V., and PEREZ, J.J. 1985. Total synthesis of (+)-trianda-

mycin A. J. Am Chem. Soc. 107:5219-5224. FRANCKE, W., MACKENROTH, W., SCHRODER, W., SCHULTZ, S., TENGO, J., ENGEL, E., ENGELS,

W., KITTMANN, R., and SCHNEIDER, D. 1985. Identification of cyclic enolethers from insects: Alkyldihydropyranes from bees and alkyldihydro-4H-pyran-4-ones from male moth. Natur- wissenschafien 40c: 145-147.

FRANCKE, W., SCHULZ, S., S1NNWELL, A., KONIG, W. A., and ROlSrN, Y. 1989. Epoxytetrahy- droedulan, a new terpenoid from the hairpencils of Euploea (Lep. : Danainae) butterflies. Lie- bigs Ann. Chem. 1989:1195-1201.

SWIFT MOTH PHEROMONE 3521

GELIN, S., and GELIN, R. 1968. Syntheses et propri~tds de dihydropyrones-4 et de 3-dic~tones ~thyl6niques. Bull. Soc. Chim. Fr. pp. 288-298.

GEORGIAN, V., HARRISSON, R., and GUBISCH, N. 1959. A new chemical desulfurization method. J. Am. Chem. Soc. 81:5834-5835.

KAISSLING, K.E., and THORSON, J. 1980. pp. 261-282, in D.B. Satelle, et al. (eds.) Receptors for Neurotransmitters, Hormones and Pheromones in Insects. Elsevier/North Holland Biomedical Press.

KONIG, W.A., BENECKE, I., and BRETTING, H. 1981. Gaschromatographische Trennung enantiom- erer Kohlenhydrate an einer neuen chiralen station~iren Phase. Angew. Chem. 93:688-690.

KUBO, I., MATSEMOTO, T., and WAGNER, D.L. 1985. Isolation and structure of hepialone; prin- ciple component from male sex scales of (H. californicus (Lepidoptera). Tetrahedron Lett. 26:563-566.

MALLET, J. 1984. Sex roles in the ghost moth Hepialus humuli (L.) and a review of mating in the Hepialidae (Lepidoptera). Zool. J. Linn. Soc. 79:67-82.

MORI, K., and K1SIDA, H. 1986. Synthesis of both enantiomers of the heterocyclic pheromones isolated from the male swift moth H. hecta L. Tetrahedron 42:5281-5290.

RICE, K.C. and DYER, J.R. 1975. A practical synthesis of 2,3-dimethylfuran and an efficient ster- eoselective preparation of (Z)-3-methyl-2-penten-l,4-diot. J. Heterocycl. Chem. 12:1325- 1326.

SCHNEIDER, D. 1984. Pheromone biology of the Lepidoptera: overview, some recent findings and some generalizations, pp. 301-313, in L. Bolis, R.D. Keynes, and S.H.P. Maddrell (eds.). Comparative Physiology of Sensory Systems. Cambridge University Press, Cambridge, U.K.

SCHURIG, V., and WISTUBA, D. 1984. Analytical enantiomer separation of aliphatic diols as boron- ates and acetals by complexation gas chromatography. Tetrahedron Lett. 25:5633-5636.

SINNWELL, V., SCHULZ, S., FRANCKE, W., K1TTMANN, R., and SCHNEIDER, D. 1985. Identification of pheromones from the male swift moth Hepialus hecta L. Tetrahedron Lett. 26:1707-1710.

TURNER, J.R.G. 1976. Sexual behaviour: female swift moth is not the aggressive partner. Anim. Behav. 24:188-190.

TURNER, J.R.G. 1988. Sex, leks and fechts in swift moths Hepialus (Lepidoptera, Hepialidae): Evidence for the hot shot moth. Entomologist 102:90-95.

WUNDERER, H., HANSEN, K., BELL, T.W., SCHNEIDER, D., and MEINWALD, J. 1985. Sex phero- mones of two Asian moths (Creatonotos tansiens, C. gangis, Lepidoptera--Arctiidae): Behav- ior, morphology, chemistry and electrophysiology. Exp. Biol. 46:11-27.