Agric. Biol. Chem., 54 (5), 1265-1270, 1990 1265

Oviposition Deterrent of a -feeding , xuthus, from a Non-host Rutaceous Plant, Ritsuo ,Nishida, Takao Ohsugi,1 Hiroshi Fukami and Shuhei Nakajima* Pesticide Research Institute, Faculty of Agriculture, Kyoto University, Kyoto 606 Japan * Department of Agricultural Science, OkayamaUniversity, Tsushima, Okayama 700, Japan Received December 4, 1989

Females of the Rutaceae-feeding swallowtail butterfly, Papilio xuthus, do not oviposit on the rutaceous plant, Orixa japonica. A methanolic extract of O. japonica leaves was found to contain contact-chemical factors which deterred oviposition of the butterfly. Oneof the oviposition deterrent factors was isolated from a butanol soluble fraction, and determined to be quercetin 3-0-(2g-/?-d- xylopyranosylrutinoside).

A large number of swallowtail butterfly mixing it with an extract of 0. japonica. We species feed exclusively on members of the describe here the isolation and identification plant family, Rutaceae, and the females lay of an oviposition deterrent contained in the eggs with great precision on their host plant leaves of O. japonica. leaves, detecting specific chemicals through their tarsal chemoreceptors.1'2) Papilio xuthus Results is one of the Rutaceae-feeding swallowtails, and the females have been stimulated to ovi- Gravid females of P. xuthus could be stimu- posit by contact with a piece of filter paper lated to lay eggs on a piece of filter paper treated with a methanolic extract of rutaceous treated with a methanolic extract of Citrus plants such as Citrus unshiu, C. natsudaidai, unshiu at a dose of0.03g ofleafequivalent per Poncirus trifoliata, .1 ^ filter paper (g.l.e./f.p.).3'4) The females, how- The oviposition stimulants of P. xuthus have ever, did not respond to a filter paper treated been isolated from the leaves of C. unshiu, and with a methanolic extract of Orixa japonica characterized as a mixture of 10 components, (0.1 g.l.e./f.p.). Moreover, a mixture of the including flavonoid glycosides, tryptamines Citrus extract (0.03 g.l.e./f.p.) and the Orixa and other polar substances.3~6) extract (0.1 g.l.e./f.p.) was strongly rejected by Although P. xuthus feeds on various kinds the females, which clearly indicated the pres- of rutaceous plants, the females do not ovi- ence of an oviposition deterrent(s) in the posit on the rutaceous plant, Orixa japonica. leaves of O. japonica. This suggested that leaves of O. japonica con- A methanolic extract of O. japonica was tained an oviposition deterrent(s), because the fractionated into ether, ethyl acetate, butanol oviposition stimulant activity of a methanolic and water layers by solvent extraction. The extract of C. unshiu was markedly decreased by oviposition deterrent activity was monitor- f Present address: Department of Chemistry, WakayamaMedical College, 651 Hironishi, Wakayama649-63, Japan. 1266 R. Nishida et al.

Table I. 13C-NMRAssignments of Compound Y, Rutin and Methyl /?-d-Xylopyranoside in d6 -DlMETHYLSULFOXIDE

~.. Compound _ . Me C-Position ^ Rutin , . , Y xyloside

Quercetin 2 1 56.29a 1 56.51a 3 132.94 133.27 Fig. 1. Separation Procedure and Oviposition Deterrent 4 177.25 177.31 Activity of Orixajaponica Fractions against Papilio xuthus 5 161.20 161.17 Females. 6 98.21 98.97 dose, 0.1 g.l.e. of Orixa fraction+0.03 g.l.e. of Citrus 7 163.77 163.98 extract. 8 93.39 93.51 9 155.75a 156.35a 10 103.91 103.91 ed by applying each fraction (0.1 g.l.e./f.p.) T 121.24b 121.13b 2' 115.17 115.17 together with a standard methanolic extract 3' 144.65 144.65 of C. unshiu (0.03 g.l.e./f.p.) by the filter pa- 4' 148.22 148.33 per bioassay described in the experimental 5' 116.04 116.20 section. The deterrent activity was split be- 6' 121.78b 121.51b tween the butanol and water layers (Fig. 1). Glucosyl 1 98.49 101. 14 The butanol layer was separated into three 2 81.48 74.00 fractions by a C18-reverse phase column as 3 76.60c 76.38c 4 70.48d 70.53d shownin Fig. 1, and the activity was recover- 5 75.79c 75.84c ed from fraction B. Fraction B was rechro- 6 66.42 66.91 matographed on the same C18-column, elut- Rhamnosyl 1 1 00.44 1 00.66 ing with a mixture of methanol, water and 2 70.2Id 70.32d acetic acid by gradually increasing the con- 3 69.56d 69.94d centration of methanol. The deterrent activity 4 71.78 71.78 was found mainly in the 35% methanol eluate, 5 68.15 68.15 although several other fractions also exhibit- 6 17.55 17.66

ed the activity to some extent. A deterrent Xylosyl 1 104.23 104.56 compoundY, tentatively named here, was iso- 2 73.62 73.08 lated from 35%eluate by means of high-per- 3 75.77c 76.39 formance liquid chromatography (HPLC) as 4 69.29 69.45 described in the experimental section. 5 65.39 65.50 The molecular weight of compoundY was S-Values with letters in the same column are inter- determined to be 742, since a prominent peak changable. was observed at mjz 743 (M++H) in its secondary ion mass spectrum. Uponmild hy- drolysis, compound Y produced rutin and coincided very well with those of methyl /?-d- xylose. UV spectra of compound Y under xylopyranoside. Since the 2G-carbon signal various condition exhibited very similar absorp- ofY was found atd 81.48, while that ofrutin tion maximato those of rutin (see Experimen- was observed at S 74.00, the xylose moiety tal section), suggesting Y to be a quercetin 3- was suggested to be attached at carbon 2 of <9-glycoside. Table I lists the 13C-NMR as- the glucose. signments of Yin comparison with those of In order to determine the xylosyl linkage, rutin and methyl jS-D-xylopyranoside. Five ex- compoundYwas converted to its peracetyl cess signals arising from the pentosyl moiety derivative, and the acylation shifts were exam- were revealed in the spectrum of Y, which ined in its ^-NMRspectrum compared with Oviposition Deterrent of Papilio xuthus 1267

Table II. ^-NMRSpectral Data of CompoundY Dodecaacetate and Rutin Decaacetate (Chemical Shifts, Multiplicities and J-Values in Hz) CompoundYdodecaacetate Rutin decaacetate Moiety Coupling (/) Coupling (/)

Quercetin 7.26 d (2.5) 7.28 d (2.5) 6.80 d (2.5) 6.81 d (2.5) 7.81 d (2.0) 7.88 d (2.0). 7.33 d (8.7) 7.31 d (8.6) 7.90 dd (8.7, 2.0) 7.93 dd (8.6, 2.0)

Glucosyl 5.87 d (7.8) 5.40 d (7.8) 3.86 dd (7.8, 9.6) 5.15 dd (7.8, 9.8) 5.25 t (-9.5) 5.25 t (-9.5) 5.00 t (-9.5) 4.92 t (-9.5) 3.61 ddd (9.5, 4.0, 3.0) 3.54 ddd (9.5, 5.6, 3.1) 3.30 dd (ll.0, 4.0) 3.24 dd (ll,0, 5.6) 3.66 dd (ll.0, 3.0) 3.49 dd (ll.0, 3.1) 1 Rhamnosyl 4.52 d (1.5) 4.49 bs 2 5.08 dd (3.4, 1.5) 5.06 bd (-3.4) 3 5.05 dd (10.0, 3.4) 5.05 dd (9.0, 3.4) 4 4.88 t (10.0) 4.92 t (-9.5) 5 3.60 dq (10.0, 6.3) 3.62 dq (9.5, 6.3) 6 0.89 d (6.3) 1.03 d (6.3) 1 4.79 d (5.5) 2 4.81 dd (7.0, 5.5) 3 5.10 t (7.0) 4 4.86 ddd (7.0, 6.0, 4.0) 5a 3.22 dd (13.0, 6.0) 5b 4.ll dd (13.0, 4.0)

2.40 2.41 2.32 2.32 2.31 2.31 2.28 2.27 2.08 2.12 2.04 2.06 2.03 2.00 2.02 2.00 2.01 1.93 2.00 1.91 1.90 1.82 that of peracetyl rutin. The assignments (Table proton signals between Y dodecaacetate (3 II) were given with the help of two- 3.86) and rutin decaacetate (3 5.15), which dimensional NMRanalyses (H-H COSYand clearly verified the xylosyl moiety to be at- C-HCOSYspectra). Most of the signals of tached to the 2-position of the glucose. The compoundY dodecaacetate gave values con- glucosyl anomeric proton of Y dodecaacetate sistent with the corresponding signals of rutin also showed a down field shift (zl<5 =0.47 ppm). decaacetate. Only one extreme difference The coupling constant of the xylosyl ano- (AS=1.29ppm) was observed with the 2G- meric proton of Y dodecaacetate (7=5.5 Hz) 1268 R. Nishida et al.

Table III. Oviposition Deterrent Activity of an Orixa Extract and CompoundY Di scussion Sample (dose/filter paper) Inhibition The oviposition deterrent of Papilio xuthus contained in the leaves of Orixa japonica was Citrus extract 0% foundto be composedofa complexmixture of Citrus extract+ Orixa extract (0. 1 g.l.e.) 100% Citrus extract+compound Y (300 fig) 85% several compounds, in which a flavonoid tri- Citrus extract + rutin (300 fig) 10% glycoside Y was verified here as one of the major components. It is of great interest that a Citrus extract dose: 0.03 g.l.e., vV=20. flavonoid glycoside plays such a deterrent role, because some of the flavonoid glycosides have been characterized instead as positive ovipo- sition stimulants for Papilio xuthus 4) and two other closely related papilionid butterflies, P. protenor9) and P. polyxenes.10) The deterrent compound Y is a xylosyl derivative of rutin, while rutin has been characterized as one of the oviposition stimulant ingredients for P. xuthus.4) Rutin in a common components amongthe Rutaceae as well as in other plant families,U) and a large quantity (>500 jug/g.l.e.) was present in the leaves of O.japon- ica. In the course of an ecolutionary histo- was found to be very close to that of methyl ry between and plants, O. japonica may /?-D-xylopyranoside triacetate (J= 5.0 Hz).7) have developed such a flavonoid glycoside CompoundY gave a negative molecular ro- with an entirely opposite effect by simply tation ([(/>]D -240°) (cf. rutin: []D +64°; modifying the kairomonal substance with a methyl /?-D-xylopyranoside: [0]D - 126°). By pentose unit. Although our knowledge of the applying Klyne's rule to the new glycoside as distribution of both stimulant and deterrent used in the case of quercetin 3-0-(2g-/?-d- components in plants is incomplete, simul- apiosylrutinoside),8) the xylose was suggested taneous occurrenceof stimulants and the ab- to possess a /?-D-pyranosyl form. Combining sence of oviposition deterrents seems to ac- with these data, compoundY was verified as count for the narrow specificity in host rec- quercetin 3-<9-(2G-/?-D-xylopyranosylrutino- ognition amongst these butterflies. side). Cardenolides have recently been character- The oviposition deterrent activity of com- ized as the oviposition deterrents of the cab- pound Y is shown in Table III together with bage white butterflies, Pieris brassicae12) and that ofa crude extract of O.japonica and rutin. P. rapae.13) In this case also, the deterrent P. xuthus females were significantly deterred chemicals appeared to be perceived by contact by compound Y at a 300jig dose (correspond- chemoreception. It was clearly observed that ing to the content in approximately 2 g.l.e. of P. xuthus females stopped drumming, and O. japonica extract). Rutin, an oviposition quickly flew away from the chemical source as stimulant of P. xuthus^ did not exhibit signif- soon as they touched the filter paper treated icant deterrent activity at the same dose. It has with O. japonica extracts. In contrast, they been suggested that O. japonica leaves contain often kept drumming for more than several several other compounds which deter ovipo- seconds on an untreated blank filter paper. sition of P. xuthus. Isolation of other ovipo- This behavior implies that the*chemicals ac- sition deterrent factors in both the butanol tually repelled the , rather than masked and water layers is nowin progress. the activity of the coexisting oviposition stimu- Oviposition Deterrent of Papilio xuthus 1269 lants. It has not yet been determined if the test. Females giving no response to the standard meth- Orixa leaves contain the same oviposition anolic extract of C. unshiu were also rejected. Thus the stimulant ingredients as were found in C. oviposition deterrent effect was evaluated by the per- centage offemales giving no response to the test samples in unshiu. On the other hand, O. japonica is spite of the presence of the C. unshiu extract (0.03 g.l.e.). knownto serve as a major host plant for some The bioassay was performed during daylight hours at Rutaceae-feeding swallowtail butterflies such 24±2°C. as Papilio macilentus and P., Manor. A sys- Extraction andfractionation of O. japonica leaves. Fresh tematic study of the host preferences amongst leaves of Orixa japonica (1.5 kg) were collected from the these closely related species mayprovide clues campus of Kyoto University in July-1987, and extracted for understanding the coevolution of the phy- with methanol (5 1). Evaporation of the extract gave a tophagous insects with the stimulant and de- viscous brown oil (48.0g), which was dissolved in water terrent chemicals present in their host plants. (1.3 1), and successively washed with ether (0.7 1 x 3), ethyl acetate (0.71x3) and 1-butanol (0.71x 3). The yields of Experimental the ether, ethyl acetate, butanol and water layers were 14.3g, 0.5g, 4.4g and 29.1g, respectively. The butanol Instruments. Optical rotation was measured with a layer was chromatographed on a reverse-phase column JASCO ORDModel J-5 spectropolarimeter, and UV (80g of ODS-W micro-bead silica gel 5D, 100-200 mesh, spectra were measured with a Shimadzu UV-360 recording Fuji-Davison Chemical Ltd., 170mm x 35mm i.d.), elut- spectrophotometer and a Beckman DU-64 spectropho- ing with mixtures of water and acetic acid (100 : 1, 500ml, tometer. Secondary ion mass spectrum (SIMS) was re- fraction A), methanol, water and acetic acid (40:60: 1, corded with a Hitachi M-80 mass spectrometer at 8kV 500ml, fraction B), and methanol and acetic acid (100 : 1, (primary beam gas, xenon; matrix, glycerol; stainless steel 500ml, fraction C). Fraction B was found to be re- sample stage). 1H- and 13C-NMRspectra were measured sponsible for deterring oviposition (see Fig. 1). Fraction B with JEOL JNM FX-90Q (90 MHz) and Varian VXR-500 was finely chromatographed on the same reverse-phase (500MHz) instruments. The letters s, d, t, q and m column, eluting successively with solvent mixtures (1 1 represent singlet, doublet, triplet, quartet and multiplet, each; every solution containing 1 %of acetic acid) of 15% respectively. methanol, 20% methanol, 25% methanol and 30% meth- Bioassay. Papilio xuthus larvae were raised on leaves anol. The subsequent 35% methanol eluate (500ml, of Poncirus trifoliata or Citrus unshiu. Female but- 0.98 g yield) exhibited oviposition deterrent activity, al- terflies were hand-paired within 2 days of emergence, though some other fractions also showedweak activity. fed with a 10%sucrose solution daily, and used for the Portions of the 35%methanol eluate were chromato- bioassay when 3-14 days old. Test samples were applied to graphed on a reverse-phase column (^Bondasphere 5ji a piece of filter paper (10cm2) by cutting a round paper C18-100A, 150mmx l9mm i.d., Waters Associates), (Toyo filter paper, qualitative no. 2, 9 cm in diameter) into eluting with a mixture of acetonitrile, water and acetic 6 equal fan-shaped pieces. Sample solutions (100-200 ^1) acid (20:80: 1) at 7ml/min. The deterrent compound Y of given concentrations were applied to the filter paper, was isolated at tR 8.8min (detection at I 254nm). The allowed to dry and misted with distilled water immediately content of compound Y in the butanol layer (1.5 kg.l.e.) before the bioassay. Doses are expressed either as gram was approximately 120mg. Compound Y. [a]23 -32.2° leaf equivalent per 10cm2 of filter paper (g.l.e./f.p.) or as (c=1.08, methanol). UV Amax nm (e): 256 (8200), 266 the absolute amount applied to the filter paper (^g/f.p.). shoulder (sh) (7100), 294 sh (3300), 356 (6800) (methanol); Each female butterfly was introduced into a test chamber 276 (8800), 326 (3400), 408 (7300) (methanol+NaOCH3); (50x 50x 50cm3), and placed in turn on the test paper, 274 (9500), 304 sh (2700), 334 sh (1900), 430 (7600) contact with it being maintained as much as possible (methanol+A1C13); 268 (7000), 300 sh (2800), 358 (4300), during 30sec. The females responded immediately to a 398 (4700) (methanol+AICI3+HCI); 273 (8900), 294 standard methanolic extract of C. unshiu at a dose of0.03 sh (3300), 322 sh (3900), 384 (5300) (methanol+ g.l.e./f.p., curling their abdomens and eventually laying NaOCOCH3). SIMS m/z (%): 743 (71), 765 (5), 781 (2). eggs.3) However, the females did not show any oviposi- ^-NMR (J6-dimethylsulfoxide) 5: 12.62 (1H, broad), tional response to a filter paper treated with a mixture of 7.58 (1H, br.d, J=9), 7.55 (1H, br.s), 6.83 (1H, d, J=9), both the C. unshiu extract (0.03 g.l.e.) and O. japonica 6.37 (1H, br.s), 6.18 (1H, br.s), 5.60 (1H, m), 4.60 (1H, extract (0.1 g.l.e.). Therefore, the oviposition deterrent m), 4.37 (1H, br.s), 1.00 (3H, br.d, J=6). 13C-NMR: see activity was monitored by adding each Onxa-fraction to Table I. the standard methanolic extract of C. unshiu. Anyfemales giving a positive ovipositional response either to a moist- Hydrolysis of compound Y. CompoundY (20mg) was ened blank filter paper or to the standard mixture (C. dissolved in I n hydrochloric acid (0.8ml), and held at unshiu extract + O. japonica extract) were rejected from the 60°C for 1 hr. After removing the yellow precipitate of 1270 R. Nishida et al. quercetin, the supernatant was passed through a Sep-pak Methyl /?-D-xylose triacetate: MS (70eV) m/z (%): 290 C18 cartridge (Waters Associates), eluting with an ad- (M+, 3), 259(M+-OCH3, 100), 231 (M+-OCOCH3, 6), ditional 2ml of water, and then with 40% methanol in 199 (52), 171 (29), 173 (43), 157 (49), 139(61), 97 (41), 69 water (4ml). Rutin was recovered from the 40%methanol (32), 44(71). !H-NMRS(CDC13): 5.10(1H, t, 7=8.5Hz), eluate under the sameHPLCconditions as those used for 4.92 (1H, double double d, 7-5.0, 8.5, 8.7Hz), 4.88 (1H, isolating compound Y (fR= 14.5min, yield=2.5mg). UV double d, 7=6.8, 8.7Hz), 4.37 (1H, d, 7=6.8Hz), 4.10 Xmaxnm: 256, 295 sh, 357 (methanol). Portions of the water (1H, double d, 7-8.7, ll.9Hz), 3.47 (3H, s), 3.34 (1H, layer were separated by a partition-phase column (PA-03, double d, 7=5.0, ll.9Hz), 2.06 (3H, s), 2.05 (3H, s), 2.04 250mmx 4.6mm i.d., YMC-pack), eluting with 67% ace- (3H, s). 13C-NMR 5 (CDC13): 170.05, 169.84, 169.46, tonitrile in water (l ml/min), and xylose was isolated at 101.63, 71.62, 70.86, 69.02, 62.03, 56.56, 20.69 (2C), 20.66. ^=7.3min. The !H-NMR spectrum (in D2O) coincided with that of D-xylose (Wako Pure Chemical Industries Acknowledgments. Wethank Dr. T. Ueno and Mr. S. Ltd.). Kokuboof our institute for assistance with this work. We also thank Professor P. Feeny of Cornell University, and Preparation of the peracetyl derivatives. CompoundY Dr. K. Honda of Seisho Biological Laboratory for dis- was dissolved in a mixture of pyridine (1 ml) and acetic cussions and encouragement. anhydride (1 ml), and the solution maintained overnight at roomtemperature. The reaction mixture was dissolved in References ethyl acetate, and washed successively with saturated NaHCO3,1 n HC1and saturated NaCl. The organic layer 1) R. Nishida, Botyu-Kagaku, 42, 133 (1977). was dried over anhydrous Na2SO4and purified through a 2) T. Ichinose and H. Honda, Appl. Ent. Z00L, 13, 103 short silica gel column (5 g, Wako Gel C-200), eluting with (1978). 40%ethyl acetate in benzene, to yield a yellow oil of the 3) T. Ohsugi, R. Nishida and H. Fukami, Agric. Biol. acetate (13 mg). Chem., 49, 1897 (1985). CompoundY dodecaacetate: !H-NMR,see Table II. 4) R. Nishida, T. Ohsugi, S. Kokubo and H. Fukami, 13C-NMR d (CDC13): 173.32, 171.32 (2C), 171.25, 171.06, Experientia, 43, 342 (1987). 170.96, 170.88, 170.80, 170.45, 170.26, 169.19, 169.04, 5) R. Nishida, T. Ohsugi and H. Fukami, Agric. Biol. 168.84, 157.67, 155.03, 153.96, 151.42, 145.29, 143.13, Chem., accepted. 137.57, 129.93, 128.58, 125.44, 124.71, 116.25, 114.50, 6) T. Ohsugi, R. Nishida and H. Fukami, Appl. 109.92, 100.98, 99.ll, 98.65, 78.61, 75.87, 73.59, 71.91, Entomol. Zool., received. 71.32, 70.82, 70.63, 70.53, 70.13, 69.92, 67.47, 67.00, 62.28, 7) P. L. Durette and D. Horton, Carbohyd. Res., 18, 403 22.37, 22.29, 22.01 (3C), 21.95, 21.87 (2C), 21.84, 21.82, (1971). 21.76, 21.71, 18.06. 8) I. Rappaportt, D. Giacopello, A. M. Seldes, M. C. In the samemanner, rutin decaacetate and methyl p-D- Blanco and V. Deulofeu, Phytochemistry, 16, 1115 xylopyranoside triacetate were prepared from rutin (Wako (1977). Pure Chemical Industries Ltd.) and methyl /?-D-xylo- 9) K. Honda, 7. Chem. Ecoi, 12, 1999 (1986). pyranoside (Aldrich Chemical Co.), respectively. 10) P. Feeny, K. Sachdev, L. Rosenberry and M. Carter, Rutin decaacetate: 1H-NMR,see Table II. 13C-NMRS Phytochemistry, 27, 3439 (1988). (CDCI3): 173.06, 171.28, 171.21, 171.03 (2C), 170.87, ll) J. B. Harborne and C. A. Williams, in "The 170.72, 170.39, 169.19, 169.00, 168.76, 157.72, 158.78, Flavonoids," ed. by J. B. Harborne, T. J. Mabry and 155.06, 151.28, 145.21, 142.88, 138.05, 129.70, 128.35, H. Mabry, Academic Press, New York, 1975, p, 376. 125.80, 124.55, 116.18, 114.56, 110.16, 100.71, 98.85, 12) M. Rothschild, H. Alborn, G. Stenhagen and L. M. 73.89, 73.65, 72.50, 72.00, 70.60, 70.47, 70.12, 68.06, 67.45, Schoonhoven, Phytochemistry, 27, 101 (1988). 22.30, 22.22, 21.97, 21.84 (2C), 21.81 (3C), 21.78, 21.76, 13) J. A. A. Renwick, C. D. Radke and K. Sachdev- 18.30. Gupta, 7. Chem. Ecol, 15, 2161 (1989).