transformation of plant pyrrolizidine alkaloids into novel insect alkaloids by arctiid moths...

Biochemical Systemadcs and Ecology, Vol. 18, No. 7/8, pp. 549-554, 1990. 0305-1978/90 $3.00 + 0.00 Printed in Great Britain. © 1991 Pergamon Press plc.

Transformation of Plant Pyrrolizidine Alkaloids into Novel Insect Alkaloids by Arctiid Moths (Lepidoptera)

THOMAS HARTMANN,*§ ANDREAS BILLER,* LUDGER WlTTE,* LUDGER ERNST1" and MICHAEL BOPPRI~;

*lnstitut f(~r Pharmazeutische Biologie, Technische Universit~t Braunschweig, Mendelssohnstrasse 1, D-3300 Braunschweig, F.R.G.;

tNMR-Laboratorium der Chemischen Institute, Technische Universit&t Braunschweig, Hagenring 30, D-3300 Braunschweig, F.R.G.;

~;Forstzoologisches Institut der Universit~it Freiburg i.Br., Fohrenb~hl 27, D-7801 Stegen-Wittental, F.R.G.

Key Word Index--Creatonotos transiens; Arctiidae; Lepidoptera; pyrrolizidine alkaloids (insect specific); creatonotine; isocreatonotine; callimorphine; structure elucidation; biosynthesis.

Abstract--Two new pyrrolizidine alkaloids (PAs) were isolated from adults of Creatonotos transiens, the larvae of which had ingested retronecine or ester alkaloids, The structures were elucidated by GC-MS and 1H and 1~C NMR spectroscopy as 09-(2 - hydroxy-3-methylpentanoyl)-retronecine (creatonotine) and its isomer OT-(2-hydroxy-3-methylpentanoyl)-retronecine (isocreatonotine). The 2-hydroxy-3-methylpentanoic acid obtained by hydrolysis of the creatonotines is the 2S,3S-stereoisomer. The creatonotines as well as accompanying trace amounts of callimorphine were shown to be synthesized by both sexes from dietary retronecine by esterification. If Creatonotos larvae had been fed with Gynura scandens as PA source, the insect PAs accounted for about 75% of total PAs isolated from adults, indicating that Creatonotos degraded plant PAs and subsequently re-esterified the resulting retronecine. This work provides definitive proof that an arctiid moth is able to hydrolyse plant ester alkaloids and re-esterify the resulting necine base with acids which are intermediates of insect metabolism. PAs isolated from Creatonotoswere exclusively present in the form of their N-oxides.

Introduction Pyrrolizidine alkaloids (PAs) represent a class of typical plant secondary compounds [1] and have received much attention in the study of insect-- plant relationships [2]. So far, some 200 structures have been isolated from a great variety of plant species, particularly from Senecio (Aster- aceae), Heliotropium (Boraginaceae) and Crotala- ria (Fabaceae) [1]. Although PA producing plants are usually avoided by herbivores, a number of insect species have evolved adaptations not only to cope with these compounds but also to utilize them for their own benefit: the insects store PAs obtained from plants and gain protection from predators, and in addition to the use of PAs for defence, males of various Lepidoptera synthesize sex pheromones from PAs of plant origin [2-5 and refs therein].

Apart from studies on pheromone biosynthesis [5, 6 and refs therein] the metabolism of PAs by insects has received little attention to

§Author to whom correspondence should be addressed.

(Received 13 June 1990)

date. However, in addition to unmodified PAs sequestered from larval host plants, callimorphine (5), a PA not known from plant sources, has been isolated from several arctiid moths; Tyria jacobaeae [7-9], 'Callimorpha dominula [7], Arctia caja [8], Creatonotos transiens [10] and Gnophaela latipennis [11]. Recently, we have observed that Tyriajacobaeae possesses the ability to esterify [14C]retronecine yielding radioactively labelled 5 [12]. In the course of studies aimed to elucidate details of quantitative aspects as well as on the specificity of PA storage in Creatonotos moths, the formation of insect-specific PAs was found to be a more general phenomenon [13]. Here we demonstrate the moths' ability to hydrolyse ingested plant PAs and to synthesize novel structures by esterification of the resulting necine with necic acids of insect origin.

Results Identification of insect pyrrolizidine alkaloids GC analysis of PAs extracted from adult Creato- notos transiens, the larvae of which had ingested

549

550 T. HARTMANN ETAL.

retronecine (1) as the only source of alkaloids, revealed 1 and several derivatives, all as N- oxides (cf. Fig. 1). Two major compounds (6, 4), which by GC-MS analysis had been identified as isomeric 09 and O7-esters of 1 proved to be novel natural products. Compounds 6: R1=1985; MS (m/z(%)]=M+269(1), 225(5), 155(3), 138(98), 94(43), 93(100), 80(16), 67(9), 53(7), 45(12), 41 (15). Compound 4: R1=1960; MS [m/z(%)]=M+269(4), 251 (28), 140(16), 138(65), 137(27), 124(20), 120(25), (111)(76), 108(16), 106(52), 94(33), 80(100), 68(13), 45(16), 41 (19).

The mass spectrum of 6 is typical of 1,2- dehydropyrrolizidine alkaloids esterified at C-9 and with a free hydroxyl at C-7, whereas 4 repre- sents the respective alkaloid esterified at C-7 and with a free hydroxyl at C-9 [14]. 1H and 13C NMR spectroscopy was necessary to elucidate the structures of the necic acids (C603H13). Approxi- mately 1 mg of a purified mixture of 6 and 4 was analysed. Integration of suitable 1H NMR signals showed the presence of two compounds, 6 and 4 in a molar ratio of 7:3. Signal overlap was present in many regions of the spectrum, yet a two dimensional homonuclear shift-correlated spectrum (1H, 1H-COSY) [15] allowed the deter- mination of the spin coupling network and the identification of all 1H chemical shifts (Table 1). The spin coupling constants could not be com- pletely analysed, but this was not necessary for the structure elucidation. In spite of the small amount of material, the signals of the major compound could be unambiguously identified in the 13C NMR spectrum (Table 1). The 1H NMR chemical shifts and signal multiplicities clearly identified the acid component of 6 and 4 as

100

~ 5o

/6

5 ~ S

1\ Z,~ I / I 2,3F ,F I

200 600 1000 scan FIG. 1. GC SEPARATION OF ALKALOIDS EXTRACTED FROM A MALE CREATONOTOS MOTH WHICH AS LARVA HAD RECEIVED RETRO- NECINE (GC-MS; total-ion chromatogram), l~retronecine; 2,3=PAs, M + 255 and 237, respectively; 4=PA, M + 269 (isocreatonotine); 5=callimorphine; 8=PA, M + 269 (creatonotine); S=hyoscyamine (internal standard); F=fatty acid methyl esters.

TABLE 1. ~H AND ~3C NMR DATA OF COMPOUNDS 6 AND 4 (CDCI3)*

Compound B Compound 4

Position (~H 1" ~c ~H:~

1 - - 132.7 - - 2 5.84 124.8 5.70 3 4.31, 3.62 60.7§ 4.10, 3.45 5 3.76, 3.05 54.1 3.55, 2.76 6 =2.17, =2.17 35.9 2.44, 2.13 7 4.53 70.0 5.50 8 4.67 78.9 4.56 9 4.93, 4.73 60.6§ 4.20, 4.20 1 ' - - 174 .3 - -

2' 4.14 75.2 4.01 3' 1.84 39.0 1.74 4' 1.46-1.19 23.9 1.46-1.19 5' 0.91 11.7 0.90 6' 1.0O 15.4 0.96

*When the compounds are stored in CDCI 3 solution for several days, chemical shift changes occur which are probably caused by partial hydrochloride formation.

1-Coupling contants: J(9a,9b)=13.1; J(2',3')~4.2; J(3',4')=7; J(3',6')=6.9, J(4",5')=7.5 Hz.

$Coupling constants: J(2',3")=4.1 ; J(3',4')=7; J(3",6")=6.9, J(4",5")=7,6 Hz.

§Interchangeable assignments [for numbering cf. Fig. 2 (6)].

2-hydroxy-3-methylpentanoic acid. The protons on C-9 (~)H=4.93 and 4.73) in 6 are strongly deshielded relative to the corresponding protons in 1 (~)H=4.33 and 4.12; our data). Similarly, C-9 in 6 (~c=60.6 or 60.7) is deshielded relative to 1 (6c=58.9 in CDCI3; our data and ref. [16]). In 4, on the other hand, H-7 is very much deshielded ((3R=5.50) relative to 1 (6H=4.32). This proves, in addition to MS analysis, 6 to be the O9-ester and 4 to be the O7-ester.

According to GC-MS and NMR data the necine base could be either 1 or its 7S-isomer heliotridine. To answer this question the insect PAs, were hydrolysed, silylated with MSTFA and the resulting TMS-derivatives were analysed by GC-MS in comparison to reference compounds. Compound 1 and heliotridine displayed identical MS fragmentation (not shown) but could easily be distinguished by their RIs; compound 1 accounted for 99% of total necine base (Table 2). Thus, 6 is Og-(2-hydroxy-3-methylpentanoyl) - retronecine (named creatonotine) and 4 is 07- (2-hydroxy-3-methylpentanoyl)-retronecine (isocreatonotine) (cf. Fig. 2).

In order to elucidate the stereochemistry of the chiral carbons of the 2-hydroxy-3-methylpentanoyl moiety, the acid fraction obtained by

TRANSFORMATION OF PLANT PYRROLIZIDINE ALKALOIDS 551

TABLE 2. GC-MS ANALYSIS OF TMS-DERIVATIVES OF NECINE BASES AND NECIC ACIDS OBTAINED BY ALKALINE HYDROLYSIS OF PA EXTRACTS FROM CREATONOTOS PREVIOUSLY FED WITH RETRONE- CINE

Relative abundance TMS-derivatives RI (%)

Necine base Retronecine 1602 99 Heliotridine 1638 1

Necic acid 2-Hydroxy-3-methylpentanoic acid 1248 93 2-Hydroxy-3-methylbutanoic acid 1168 5 2-Hydroxy-2-methylbutanoic acid * 1140 2 Tiglic acid 1000 tr¢ Senecioic acid 995 tr

*Resulting from hydrolysis of callimorphine. Ctr=trace amounts.

hydrolysis of a PA extract was esterified with Mosher's acid and analysed by comparison with the Mosher's esters of authentic 2S,3S- and 2R,3R-enantiomers of 2-hydroxy-3-methylpentanoic acid. The Mosher's esters of the two enantiomers were clearly separated by GC analysis: 2R,3R-stereoisomer, R1=1762; 2S,3S- stereoisomer, R1=1785. The respective Mosher's ester of the necic acid obtained by hydrolysis of 6 showed a RI of 1785, thus indicating a 2S,3S- stereochemistry of the necic acid (Fig. 2). Other stereoisomers (2S,3R or 2R,3S) should not be present since the methyl ester of the necic acid showed a single GC peak when separated on a Carbowax column with a retention index indenti- cal to that of the authentic 2S,3S-stereoisomer.

In addition to the creatonotines, 5 (cf. Figs 1, 2) has been found. It was identified by GC-MS (RI 1965; M + 297) by comparison with an authentic sample isolated from pupae of Tyria jacobaeae. Furthermore, among the hydrolysis products of PA extracts 2-hydroxy-2-methylbutanoic acid was detected as the TMS-derivative (Table 2). This acid is formed during hydrolysis from 2-acetoxy-2-methylbutanoic acid, the acid moiety of 5. Quantification of 5 is limited due to partial overlapping of the signals of 4 and 5 in GC analyses (Fig. 1). From GC-MS measurements as well as quantification of the acid moieties of PAs (Table 2), 5 accounts for less than 5% of total insect PAs. The peaks around scan 724 (2,3 in Fig. 1) indicate the presence of trace amounts of tigloyl- and 2-hydroxy-3-methylbutanoyl-retronecines (07- and O9-esters), respectively. The

TABLE 3. TRANSFORMATION OF RETRONECINE (1) INTO CREATONO- TINES (6, 4) BY CREATONOTOS TRANSlENS. Retronecine (6.5 limol/larva) was fed to last instar larvae; moths were extracted 24 h after emerging from pupae.

Amount Concentration

p, mol animal ' Percentage of fed p, mol (g ft. wt) -1

Males (n=6) Creatonotine 6 0.75±0.24 11.5 Isocreetonotine 4 0.35±0.09 5.3 Total 1.10±0.33 16.8

Females (n=6) Creatonotine 6 0.68±0.30 10.4 Isocreatonotine 4 0.24±0.13 3.7 Total 0.92±0.43 14.1

3.59+1.14 1.68±0.41 5.27±1.55

2.46±1.10 O.86±0.45 3.32±1.55

H3Ch 0 .0H I X oH

[1] [~]

0

0 0 " ~ C H - I/ i ~ s OH I01" Iv "CH3 ,,~ o / O . j . ~ C H 3

. . . . . ~f i i ~ T 6'

5 3 [ 5 ] [ 6 ]

FIG. 2. STRUCTURES OF RETRONECINE (1) AND THE INSECT PAs CALLIMORPHINE (5), CREATONOTINE (6) AND ISOCREATONOTINE (4). The 2-hydroxy-3-methylpentanoyl moiety of the creatonotines has 2S,3S-configuration.

552 T. HARTMANN ETAL.

respective necic acids were identified as TMS- gynuramine could not be detected in the PA derivatives in the PA-hydrolysates (Table 2). extracts from insects.

Origin of the necic acids Analyses of Creatonotos without access to PAs did not even show traces of derivatives of 1. Analyses of the organic acid fraction, however, revealed the presence of a number of acids, particularly 2-hydroxy acids. Among these, small quantities of 2-hydroxy-3-methylpentanoic acid were detected and identified by GC-MS comparison with authentic material. Thus, the necic acid of 4 and 6 is probably a product of insect metabolism.

Transformation of plant pyrrolizidine alkaloids into the creatonotines About 14-17% of dietary 1 was transformed into 4 and 6 by both sexes of Creatonotos (Table 3). In order to see whether the insects are able to produce the creatonotines if they had taken up ester alkaloids instead of the pure necine base, larvae were fed with Gynura scandens which contains gynuramine and O-acetylgynuramine as major PAs [17]. GC analyses of extracts of adult moths revealed 4, 6 and 9 as the major PAs (Fig. 3), the creatonotines accounting for 60-75% of the total PAs. In addition, components of plant origin, i.e. senecionine (7), integerrimine (8) and an unknown alkaloid "M + 351" (10), and insect PAs such as 5 and the two isomeric 2-hydroxy-3- methylbutanoyl-retronecines (2 and 3 in Fig. 3), have been found in minor amounts. O-Acetyl-

100

~ 5 0

16

F rs,l~ 2,~ . , ' i l , I~A.

50O

9" 11°"1 7.,~ A,8 . , 1000 scan 1500

FIG. 3. GC SEPARATION OF ALKALOIDS EXTRACTED FROM A MALE CREATONOTOS MOTH WHICH AS LARVA HAD BEEN FED WITH GYNURA SCANDENS (GC-MS; total-ion chromatogram), Insect PAs: 2=OT-(2-hydroxy-3-methylbutanoyl)-retronecine (M ÷ 255; RI 1865); 3=Og-(2-hydroxy-3-methylbutanoyl)-retronecine (M ÷ 255; RI 1885); 4=isocreatonotine and trace amounts of callimorphine; 6=creatonotine; Plant PAs: 7=senecionine (M ~ 335; RI 2290); 8=integerrimine (M + 335; RI 2290); 9=gynuramine (M * 351; RI 2525); 10=unknown PA (M + 351; RI 2695); S=heliotrine (internal standard); F~fatty acid methyl esters.

Discussion Creatonotos moths are able to esterify retronecine (1) resulting in creatonotine (6), isocreatonotine (4) and trace amounts of callimorphine (5). Compounds 4 and 6 represent new PAs unknown from plant sources. For the first time it is thus demonstrated that insects are capable of synthesizing their own PAs by "partial biosynthesis", i.e. 1, which definitely is of plant origin, is esterified with necic acids which the insects apparently synthesize de novo.

Previously, 5 has been found in arctiid moths which as larvae had ingested PAs of quite different structures from various host plants [7-11]. Recently, labelled 5 could be isolated from pupae of Tyria jacobaeae which as larvae received [l"C]retronecine [12], indicating the ability of Tyria to esterify the necine base. This observation and the results presented above prove that 5 is formed by the insects through re-esterification of 1 as suggested by L'Empereur et al. [11]; however, the necic acid moiety is of insect and not of plant origin.

Both male and female Creatonotos transiens esterify dietary 1 and there is no significant quantitative difference between the sexes. If Creatonotos larvae had ingested complex ester alkaloids, again a large proportion of total PAs found in the adult insects would occur as creatonotines. Thus, Creatonotos is not only able to re- esterify 1 but also to produce 1 by hydrolysing plant ester alkaloids. Interestingly, feeding experiments with PAs of different structure show great quantitative differences in the capability of the insects to utilize plant PAs as precursors for creatonotines. Monocrotaline, for instance, is an excellent precursor whereas heliotrine is a poor one. Furthermore, sex differences exist in the ability to transform plant PAs into the creatonotines. Although both sexes esterify 1 with the same efficiency (cf. Table 3), the ability to transform PAs into the creatonines is expressed much more in males [13], i.e. males seem to be adapted to degrade ester alkaloids into 1. This might be related to the biosynthesis of the male pheromone (hydroxydanaidal) which also pro- ceeds via degradation of ester alkaloids [18]. Fur- ther quantitative studies involving various PA

TRANSFORMATION OF PLANT PYRROLIZIDINE ALKALOIDS 553

plants and pure PAs are in progress to clarify these phenomena.

In any case, two parallels between PA metabolism in plants and insects should be emphasized. (1) Plants synthesize PAs as N- oxides and these are the specific forms of alkaloid translocation and accumulation [19]. The same seems to be true for Creatonotos: both the sequestered and the synthesized PAs are present as N-oxides. Creatonotos possesses the ability to N-oxidase PAs. Specific carrier-mediated uptake systems for PA N-oxides exist in the plant cell [20] as well as in the midgut of Creatonotos [21 ]. (2) In plants, necic acids are preferably derived from the amino acid isoleucine [22], which is likely to be true also for insect-derived necic acids found in the creatonotines as well as in 5.

Experimental Insect material and feeding. Our culture of Creatonotos transiens L. (Lepidoptera: Arctiidae) originates from females collected in Sumatra (Indonesia). It is maintained on a semi- artificial diet [23] which is free of PAs. Since Creatonotos larvae are pharmacophagous with respect to PAs and liberally consume glass-fibre discs impregnated with the alkaloids [3 and M. Boppr~ unpublished data], quantitative feeding experiments with pure PAs could be performed. Glass-fibre discs (Whatman GF/B 2.1) were impregnated with 1 mg each of I solubilized in MeOH. After evaporation of the solvent the discs were offered to larvae which were kept individually. In order to standardize the physiological state during which PAs are taken up, the discs were offered as the first meal after the moult to the last instar. After complete ingestion of the disc, i.e. after 2-12 h, larvae were again fed with the semi-artificial diet. In qualitative feeding experiments with ester alkaloids, leaves of Gynura scandens O. Hoffm. (Asteraceae) from our glass- house culture were fed to last instar larvae in addition to artificial diet.

Alkaloid extraction [24]. For chemical studies, moths which had been killed by freezing 24 h after emerging from pupae and stored at --25 to --70°C were analysed. Each specimen was macerated with sand and 6 ml of acidic MeOH (1% HCI) in a mortar for 10 min. After centrifugation the supernatant was divided into two aliquots which were evaporated under reduced pressure. One aliquot was redissolved in 3 ml of dilute NH4OH and directly applied to an Extrelut column (3 ml; Merck). PAs were eluted with 30 ml of CH2CI 2. After evaporation the residue was redissolved in 0.1-0.5 ml of MeOH. This extract provided the fraction of tertiary PAs. To obtain the fraction of total PAs (tertiary PAs and PA N-oxides), the remain- ing aliquot was redissolved in 2 ml of 0.1 M H2SO 4 and mixed with Zn dust in excess. The mixture was stirred at room temperature for 4 h, then the solution was made alkaline and treated as described above.

Alkaloid analyses. PAs were separated and evaluated quantitatively by capillary GC (Perkin Elmer, Sigma 2B) on a quartz column (WCOT, 30mx0.25 mm; DB-1, J&W scientific CA) [25]. Conditions: injector 250°C, temp. prog. 150-300=C,

6"C/min; split ratio 1:20; inj.vol. 1-2 Id; carrier gas He 0.7 bar; detection: flame ionization (FID) and nitrogen (PND) detectors. Atropine and heliotrine were used as internal standards.

GC-MS. A Carlo Erba Mega 5160 gas chromatograph equipped with a quartz column (30 m×0.32 mm) specified as given above, was directly coupled to a quadrupole mass spectrometer Finnigan MAT 4515. Conditions: injector 250°C; temp. prog. 70, 100 or 150-300°C, 6"C/min; split ratio 1:20; carrier gas He 0.5 bar. Retention indices (RI) were calculated from co-chromatographed hydrocarbon standards according to ref. [26].

NMR. 1H and lzC NMR spectra were measured on a Bruker AM 400 spectrometer at 400.1 and 100.6 MHz, respectively. The solvent was CDCI3; chemical shifts were referenced to tetra- methylsilane. About I mg PA in 0.5 ml of solvent was analysed. The two-dimensional ~H, 1H-COSY spectrum was run in the 90 °, 90°-mode with a relaxation delay of 0.4 s. The digital resolution of the data matrix was 2.1 Hz/point in both dimensions.

TMS-derivatives. Extracts of retronecine-fed Creatonotos were hydrolysed in 10% NaOH at 80°C for 30 rain. Necine bases were extracted three times with diethylether, then the aqueous phase was acidified with 1 N HCI and extracted three times with ethylacetate yielding the necic acids. The extracts were dried under N 2 and silylated with MSTFA (N-methyI-N- trimethylsilyl-trifluoro acetamide) at 80°C for 15 rain. The TMS- derivatives of both fractions were analysed by GC-MS as given above.

Stereochemistry of the necic acid. Authentic L- and m2- hydroxy-3-methylpentanoic acids from Sigma represent the 2S,3S and 2R,3R-enantiomers (Sigma, personal communica- tion). These reference compounds as well as the necic acid fraction, obtained by hydrolysis of PA extracts as described for the preparation of the TMS-derivatives, were esterified with R- (--)-c¢-methoxy-~-trifluoromethylphenylacetic acid chloride (Mosher's acid chloride) in CH2CI 2 in the presence of trace amounts of pyridine. Esterification with Mosher's acid intro- duces a third chiral centre into the molecules, thus allowing separation of the enantiomeric acids by GC.

Chemicals. Retronecine (1) was prepared from monocrotaline (Aldrich-Chemie GmbH, Steinheim, F.R.G,) by hydrolysis (10% NaOH, 2 h at 100°C). The crude hydrolysate was purified via Extrelute [24] and a silica gel column (eluent, CH2CI2). The final product was crystallized from acetone. R-(--)- c( -Methoxy-c( -trifluoromethylphenylacetic acid chloride (Mosher's acid chloride) was obtained from Fluka, Buchs (Switzerland), trans-2-methylbutenoic acid (tiglic acid) and 3-methyl-2-butenoic acid (senecioic acid) from Merck, Darm- stadt (F.R.G.), and L-2-hydroxy-3-methylpentanoic acid (2S,3S- stereoisorner) and o-2-hydroxy-3-methylpentanoic acid (2R,3R-stereoisomer) from Sigma Chemie GmbH, Deisenhofen (F.R.G.).

Acknowledgements--This work was supported by grants of the Deutsche Forschungsgemeinschaft (Ha 641/12-1 and Bo 664/4-1) and of the Fonds der Chemischen Industrie to T. H. We thank Anita Kiesel, Claudine Theuring and Claudia Wahren- burg for skilful technical assistance.

References 1. Mattocks, A, R. (1986) Chemistry and Toxicology of Pyrro-

lizidine Alkaloids. Academic Press, London. 2. Boppre, M. (1990) J. Chem. Ecol. 16, 165.

554 T. HARTMANN ETAL

3. Boppre, M. (1986) Naturwissenschaften 73, 17. 4. Schneider, D. (1986) in Perspectives in Chemoreception

and Behavior (Chapman, R. F., Bernays, E. A. and Stoffo- lano, J. G., eds), p. 123. Springer, Berlin.

5. Eisner, T. and Meinwald, J. (1987) in Pheromone Bio- chemistry (Prestwich, G. D. and Blomquist, G. J., eds), p. 251. Academic Press, Orlando, Florida.

6. Boppr~, M. and Schneider, D. (1989) Zoo/. J. Linn. Soc. 96, 339.

7. Edgar, J. A., Culvenor, C. C. J., Cockrum, P. A., Smith, L. W. and Rothschild, M. (1980) Tetrahedron Left. 21, 1383.

8. Aplin, R. T., Benn, M. H. and Rothschild, M. (1968) Nature 219, 747.

9. Aplin, R. T. and Rothschild, M. (1972) in Toxins of Animal andPlant Origin (de Vries, A. and Kochva, E., eds), p. 579. Gordon & Breach, London.

10. Wink, M., Schneider, D. and Witte, L. (1988) Z Naturf. 43c, 737.

11. L'Empereur, K. M., Li, Y. and Stermitz, E R. (1989) J. Nat. Prod. 52, 360.

12. Ehmke, A., Witte, L., Billet, A. and Hartmann, T. (1990) Z. Naturf. (in press).

13. Boppr6, M., Biller, A., Witte, L. and Hartmann, T. (1990) (in preparation).

14. Pedersen, E. and Larsen, E. (1970) Org. Mass Spectr. 4, 249. 15. Bax, A., Freeman, R. and Morris, G. (1981) J. Magn. Reson.

42, 164. 16. Molyneaux, R. J., Roitman, J. N., Benson, M. and Lundin,

R. E. (1982) Phytochemistry21, 439. 17. Wiedenfeld, H. (1982) Phytochemistry21, 2767. 18. Schulz, S., Francke, W., Boppr~, M., Eisner, T. and Mein-

wald, J. (1990) Proc. Natn. Acad. Sc~ U.S.A. (in preparation).

19. Ha~mann, T., Ehmke, A., Eilert, U., v. Borstel, K. and Theuring, C. (1989) Planta 177, 98.

20. Ehmke, A., v. Borstel, K. and Hartmann, T. (1988) Planta 176, 83.

21. Wink, M. and Schneider, D. (1988) Naturwissenschaften75, 524.

22. Cahill, R., Crout, D. H. G., Mitchell, M. B. and ML~ller, U.S. (1980) J. Chem. Soc. Chem. Commun. 419.

23. Bergomaz, R. and Boppre, M. (1986) J. Lep. Soc. 40, 131. 24. Hartmann, T. and Toppel, G. (1987) Phytochemistry 26,

1639. 25. Toppel, G., Witte, L., Riebesehl, B., v. Borstel, K. and Hart-

mann, 1". (1987) Plant Cell. Rep. 6, 466. 26. Wehrli, A. and Kovats, E. (1955) Helv. Chim. Acta 42, 2709.

transformation of plant pyrrolizidine alkaloids into novel insect alkaloids by arctiid moths...

Documents