Organ Specific Storage of Dietary Pyrrolizidine Alkaloids in the Arctiid transiens A. Egelhaaf, K. Cölln, B. Schmitz, M. Buck Zoologisches Institut der Universität, Im Weyertal 119, D-5000 Köln 41, Bundesrepublik Deutschland M. Wink* Pharmazeutisches Institut der Universität, Saarstraße 21, D-6500 Mainz, Bundesrepublik Deutschland D. Schneider Max-Planck-Institut für Verhaltensphysiologie, D-8130 Seewiesen/Starnberg, Bundesrepublik Deutschland Z. Naturforsch. 45c, 115-120(1990); received September 25, 1989 Creatonotos, Arctiidae, Pyrrolizidine Alkaloids, Heliotrine Metabolism, Storage Larvae of the arctiid moth obtained each 5 mg of heliotrine, a pyrroli­ zidine alkaloid, via an artificial diet. 7 S-Heliotrine is converted into its enantiomer, 7 Ä-heliotrine, and some minor metabolites, such as callimorphine. IS- and 7 /?-heliotrine are present in the predominantly (more than 97%) as their N-oxides. The distribution of heliotrine in the organs and tissues of larvae, prepupae, pupae and imagines was analyzed by capillary gas-liquid chromatography. A large proportion of the alkaloid is stored in the integu­ ment of all developmental stages, where it probably serves as a chemical defence compound against predators. Female imagines had transferred substantial amounts of heliotrine to their ovaries and subsequently to their eggs; males partly directed it to their pheromone biosyn­ thesis.

Introduction and -dissipating organ, the corema [6, 7]. The The East-Asian arctiid moth Creatonotos tran­ quantity of this morphogenetic effect is directly de­ siens, is polyphagous and thus also feeds on a pendent upon the dosis, but independent of the number of plants which contain noxious secondary temporal spreading of the feeding program. All metabolites. Whereas alkaloids of the tropane-, this clearly indicated already that the larvae quinolizidine-, or purine-type are not resorbed but sequester all or some of the ingested PA. To date eliminated with the faeces, pyrrolizidine alkaloids we do not know, when exactly the sensitive period (PA) are selectively taken up and processed by for the corema induction begins (some time in the these [1-3]. The resorption of PA seems to late final larva?); but we found that this period ter­ be catalyzed by specific carrier proteins [4], In all minates during the second prepupal day [9, 10], In stages and both sexes, PA fed specimens appeared the process of the growth of the corema anlage, the to be protected from several predators [5]. ingested PA are a sufficient (yet perhaps not indis­ In the males, PA further serve as modulators for pensable [8]) factor in addition to the pupation the development of an abdominal scent-producing hormone, the ecdysone [9, 10]. The scent-hair bearing corema dissipates 7 R-hy- droxydanaidal, a pheromone which derived in all Abbreviations: GLC, gas-liquid chromatography; MS, our laboratory experiments from dietary PA mass spectrometry; El, electron impact; Cl, chemical [6 - 12]. ionization; PA, pyrrolizidine alkaloids; L7, last larval PA thus have three functions in this specialized instar. herbivore: they act (as in some other ; * New address: Institut für Pharmazeutische Biologie, see [1, 5, 13]) as defence substances and as phero­ Universität Heidelberg, Im Neuenheimer Feld 364, D-6900 Heidelberg, Bundesrepublik Deutschland. mone precursors, but only in Creatonotos can they Reprint requests to A. Egelhaaf, D. Schneider or also act as a morphogen for the male pheromone M. Wink. gland. Verlag der Zeitschrift für Naturforschung, D-7400 Tübingen The aim of this study was to follow the path of a 0341-0382/90/0100-0115 $01.30/0 pure PA, which last instar larvae ingested. We ana- 116 A. Egelhaaf et al. ■ Storage of Pyrrolizidine Alkaloids in an Arctiid Moth

Table I. Distribution of pyrrolizidine alkaloids in Creatonotos transiens. Larvae obtained 5 mg 75-heliotrine each. Alkaloids were extracted from tissue preparations and analyzed by capillary GLC. Alkaloid content/tissue is given as |ig PA (IS- and 7 /?-heliotrine) including PA-N-oxides which were reduced prior to extraction. Data are means of at least 2 . - = Tissue not analyzed/not available; Haem. = haemolymph; Fat b. = fat body; Int. = integument: Sex o. = sex organ; Cor. = corema; Rest = not defined tissues; Faec. = faeces; Exuv. = exuviae; % rec. = % PA recovered excluding faeces and exuviae.

Developmental Alkaloid content (fig/tissue or organ) Stage Sex Haem. Fat b . Int. Wings Guts Sex o. Eggs Cor. Rest Faec. Exuv . % rec.

Larvae M 26 45 547 132 1 .0 _ _ 98 308 1.7 17 F 28 13 343 24 0 . 8 -- 11 1091 1 . 0 8 Prepupae(1d) M 37 19 520 - 13 0 . 2 -- 18 -- 1 2 F 2 1 61 876 - 8 0.7 -- 27 - - 2 0 Prepupae ^2d) M 34 108 509 - 109 2.5 -- 147 -- 16 F 18 35 473 - 37 7.8 -- 80 -- 14 Pupae(1 d) M 71 - 699 - 41 2 . 2 -- 264 -- 2 2 F 15 - 198 - 5 --- 170 -- 8 Pupae (5d) M 17 6 664 - 39 6 -- 309 -- 2 0 F 8 18 557 - 1 0 --- 204 -- 16 Pupae ( 8 d) M 25 - 329 - 3 0 . 8 - 129 459 - 4 19 F 3 - 166 ----- 90 - 3 5 Imago M -- 568 27 2 7 - 2 1 --- 1 2 F - - 76 1 0 2 500* 17+ - 387 -- 2 0

* In 3 females the ovary could be isolated, for the other animals the ovary fraction was included in the “rest” section. + Data underrepresent the real value since only a fraction of the eggs was recovered. See [ 8 ] for the latest measurements.

lyzed the major tissues through the subsequent de­ Alkaloid analysis velopmental stages, beginning with the wander- Organs were homogenized in 500 )^1 0.5 m HC1 raupe, a stage which does no longer feed. in Eppendorf vials and stored at -20 °C until fur­ ther processing. In order to reduce PA-N-oxides Materials and Methods the homogenates were treated with zinc powder Animals for at least 3 h. Alkaloids were extracted by liquid- solid extraction using Chem elut columns (ICT, A laboratory population of Creatonotos tran­ Analytichem; [2, 15]). Crude alkaloid extracts were siens (Walker) was maintained on a semi-artificial analyzed by capillary GLC on a Perkin Elmer in­ diet [14]. The origin (, Indonesia) of this labo­ strument (GLC 8500) equipped with flame ioniza­ ratory population is important because of the un­ tion and nitrogen specific detectors. GLC condi­ clear systematic status of this species and its sub­ tions: 30 mxO.l mm DB-5 column (J&W; ICT species from different localities. L7 (last instar) Frankfurt); oven: 170 °C, 2 min isothermal, then larvae were each fed 5 mg of the pure (commercial­ to 300 °C at 30 °C/min; detectors: 320 °C; injector: ly available) PA 7S'-heliotrine. This dosis was of­ 250 °C, split injection (1:20); carrier gas: helium, fered in the diet. At given intervals, wanderraupen 90 kPa. - GLC and GLC-MS (EI, Cl) was per­ (late L7) or later developmental stages (prepupae, formed as in a previous study [15]. pupae, imagines) were dissected to obtain the fol­ lowing organs or tissues: haemolymph, fat body, Results and Discussion gut, integument, ovaries, eggs, testes, coremata, wings, exuviae and faeces. The comparatively Identification o f heliotrine and its metabolites large variation of the PA amounts recovered from Heliotrine containing diet is readily accepted by the different tissues (in particular from the “rest” the larvae [3, 5], Alkaloid extracts obtained from fraction, see Table I) appears to be mainly a natu­ whole larvae, pupae and imagines contained the ral phenomenon and only to a minor degree indic­ original 7 S-heliotrine and metabolites derived ative of technical difficulties to obtain “clean” from it [15]. Up to 20% of the heliotrine and samples of fresh tissue for the chemical analysis. its PA metabolites could be recovered from the A. Egelhaaf Egelhaaf A. Table I). Other experimental details as in Tables I and III. Iand Tables in as details experimental Other I). Table For females, the amounts found in the ovaries were omitted from this illustration but attributed to the rest fraction fraction rest the to f o attributed sexes but both f o illustration this stages from omitted were developmental ovaries and the in tissues found different in amounts the heliotrine females, of For Distribution 1. Fig. ALKALOID CONTENT (%) 40 - 60 80 20 - F G C HFIGSCRHFIGSCR G C R W C S G I F H t l ■ al. et trg ofProiiie laod i a Acid oth M Arctiid an in Alkaloids Pyrrolizidine f o Storage IMAGINES wings corema gonads gut fatbody integument haemolymph rest UA (8d) PUPAE

o

¥ Creatonotos transiens. Creatonotos (cf. 17

118 A. Egelhaaf et al. ■ Storage of Pyrrolizidine Alkaloids in an Arctiid Moth animals and up to a further 20% from the faeces alkaloidal compounds by detoxification processes (Table I). The low PA recovery rate from the and thus escaped our detection process? Experi­ animals is in part explained by other experiments ments in progress support the latter explanation [2] where feeding of more than 3 mg PA per [17]. larva seemed to overcharge the intestinal uptake A major chemical change is apparent in the PA mechanism with PA excretion as a consequence. extracts, namely a conversion of the “free” alka­ The fate of the “’missing’' portion (up to 60% of loid into its more hydrophilic and less cell-mem- the original PA) needs to be studied and we ask: brane permeable N-oxides (Table II). This occurs was the heliotrine converted to alkaloids which are rather soon after the PA feeding [17], The site of either not extractable or not sufficiently volatile this change is still unknown but might be the gut for GLC studies or was it mainly degraded to non- wall or the haemocoel. A similar prevalence of PA-N-oxides has earlier been reported of the arc­ tiid Thyriajacobeae: after feeding on the PA plants Table II. Occurrence o f PA-N-oxides in Creatonotos Senecio doria and Crotalaria retusa, 71—87% of its transiens. A larva which had obtained 5 mg 7 S-helio- trine (free base) at stage L7 was analyzed at the wander- PA were oxidized [16]. raupen stage. The was freezed in liquid nitrogen Another important chemical change is an inver­ and then cut into segments of equal size. 4 of these seg­ sion of the stereochemistry at C7 of the hetero­ ments (N o. 2, 4, 6 , 8 ) were extracted directly with cyclic moiety of the 7 S-heliotrine molecule, which C12CH 2 to obtained the free PA. 5 segments (No. 1, 3, 5, 7, 9) were reduced with zinc powder prior to solvent ex­ leads to the enantiomeric 7 /^-heliotrine in both traction; they contain free PA plus PA-N-oxides. sexes [15]: 60-80% of the PA recorded from males (pupae and imagines) and 40-70% from females Segment No. Alkaloid content (|j.g heliotrine) “free PA” “free PA + PA-N-oxide” could be contributed to this derivative of heliotrine (Table III). This 7 /^-metabolite - which we called _ 1 28.34 7/^-heliotrine before [15] - is of lower abundance 2 1 . 0 2 - 3 - 28.09 in faeces and larval exuviae than in other tissues. 4 0.75 - The conversion is presumably the necessary step in 5 - 69.09 the biosynthesis of 7 /?-hydroxydanaidal in males, 6 3.04 - and it is likely that the higher conversion rate in 7 - 88.07 8 0.23 - males (Table III) can be contributed to it. 9 - 62.00 Minor heliotrine metabolites were callimor- phine and others which have not been identified Mean ± s.d. 1.26 ± 1.23 55.12 ± 26.34 yet.

Table III. Formation of 7/^-heliotrine in Creatonotos transiens. Larvae were fed with 5 mg 7 S-heliotrine. Alkaloid extracts from tissue preparations were analyzed by capillary GLC and the percentage of 7 /?-heliotrine determined (total alkaloid/tissue = 100%). Abbreviations s. Table I.

Developmental Content of 7 /^-heliotrine (% of total PA) Stage Sex Haem. I at b . Int. Wings Guts Sex o. Eggs Cor. Rest Faec Exuv.

Larvae (L7) M 59 6 8 53 _ 61 56 _ _ 6 8 28 47 F 42 58 28 - 46 --- 69 5 30 Prepupae (1 d) M 47 54 48 - 37 6 8 -- 54 -- F 50 60 39 - 55 --- 57 -- Prepupae (2d) M 62 58 52 - 61 55 -- 53 -- F 34 28 30 - 38 38 -- 35 -- Pupae(1 d) M 78 - 82 - 82 71 -- 78 -- F 61 - 69 - 71 --- 71 -- Pupae (5d) M 6 8 71 72 - 83 82 -- 70 -- F 49 53 53 - 55 --- 53 -- Pupae ( 8 d) M 78 - 74 - 80 -- 77 75 - 53 F 69 - 69 - - --- 58 - 59 Imagines M -- 79 77 67 67 - 83 81 -- F - 61 52 57 55 73 - 59 - - A. Egelhaaf et al. ■ Storage of Pyrrolizidine Alkaloids in an Arctiid Moth 119

Tissue distribution o f he/iotrine ites (and also of biosynthesized defensive com­ Larvae (wanderraupen, L7 which stopped feed­ pounds) in the integument has been reported and ing), prepupae (days 1 and 2), pupae (days 1, 5 and might be a general phenomenon. Examples are 8) and imagines were dissected as far as possible cardenolides in danaine butterflies [18], in a ctenu- into distinct organs or tissues and their alkaloid chid moth [2], in the milkweed bug O n copeltu s [19], contents and alkaloid patterns analyzed by capil­ and a chrysomelid beetle [20]; cyanogenid glyco­ lary GLC (Tables I, III). Fig. 1 allows an overview sides in zygaenid [21]; salicylaldehydes in of the PA pattern: the major site of PA storage in chrysomelid beetles [22]; PA in another such beetle all developmental stages is the integument. Hae- [23]; flavonoids in a butterfly [24], It is worth to molymph, gut and fat body contain some PA, but note that PA storage in danaine butterflies was - lose most of it after pupation. In some female spec­ in contrast to our findings with Creatonotos - re­ imens, the ovaries had a large PA content and ported to be restricted to the inner body and not some egg batches contained PA (but not in such found in the integument [18]. An interesting and large amounts as we found it recently in a Philip­ yet unsolved problem is, where and how the PA pines strain of this moth [8, 17], Some of the PA are stored in the integument of Creatonotos, storage in the sex organs may be “hidden” in the whether extra- or intracellularly, or even in com­ larger rest fraction in later developmentals stages partments of specialized cells as in plants [25], Re­ (Fig. 1). PA are practically not excreted with larval spective investigations are under way in our labo­ or pupal exuviae. ratories. In male pupae of day 8 (one day before emer­ gence of the moth), about 13% of the recovered al­ kaloid was found in the corema (77% as /?-isomer; Acknowledgements Tables I, III). In the subsequent adult stage, the This study was supported by grants of the Deut­ PA content of the corema decreased with a parallel sche Forschungsgemeinschaft, the Max-Planck- increase in the content of 7 /?-hydroxydanaidal Gesellschaft, and the Fonds der Chemischen Indu­ (45 ng/corema). This suggests that part of strie. We are grateful to Mss. E. Roth, U. Schade, 7 /?-heliotrine is converted into the pheromone. H. Söchting-Mayr and M. Weyerer for technical The storage of plant derived secondary metabol­ assistance. 1 2 0 A. Egelhaaf et al. ■ Storage of Pyrrolizidine Alkaloids in an Arctiid Moth

[1] D. Schneider, in: Perspectives in chemoreception [15] M. Wink, D. Schneider, and L. Witte, Z. Natur- and behavior (R. F. Chapman, E. A. Bernays, and forsch. 43c, 737(1988). J. G. Stoffolano, eds.), p. 123, Springer, Berlin, Hei­ [16] A. R. Mattocks, Xenobiotica 1, 451 (1971). delberg, New York 1987. [17] E. von Nickisch-Rosenegk, D. Schneider, and M. [2] M. Wink and D. Schneider, J. Comp. Physiol. B Wink, unpublished. (submitted). [18] J. A. Parsons, J. Physiol. (London) 178, 290 (1965); [3] M. Boppre and D. Schneider, Zool. J. Linnean Soc. M. Rothschild and T. Reichstein, Nova Acta Leo­ 96,339(1989). poldina, Suppl. 7, 507 (1976); L. P. Brower, p. 119. [4] M. Wink and D. Schneider, Naturwissenschaften in quotation [13]; L. P Brower and S. C. Glazier, 75, 524(1988). Science 188, 19 (1975); L. P. Brower, J. N. Seiber. [5] M. Boppre, Naturwissenschaften 73, 17 (1986). C. J. Nelson, S. P. Lynch, and P. M. Tuskes, J. [6 ] D. Schneider, M. Boppre, J. Zweig, S. B. Horsley, Chem. Ecol. 8 , 579(1982). T. W. Bell, J. Meinwald, K. Hansen, and E. W. [19] G. G. E. Scudder, L. V. Moore, and M. B. Isman, J. Diehl, Science 215, 1264 (1982). Chem. Ecol. 12, 1171 (1986). [7] M. Boppre and D. Schneider, J. Comp. Physiol. A [20] J. M. Pasteels and D. Daloze, Science 197, 70 (1977); 57, 569(1985). D. Daloze and J. M. Pasteels, J. Chem. Ecol.5, 63 [8 ] M. Wink, E. von Nickisch-Rosenegk, and D. (1979). Schneider, Proceedings of the 7th Internat. Sympos. [21] S. Franzl and C. M. Naumann, Tiss. Cell 17, 267 Insect Plant Relationships, Budapest 1989, in press. (1985); S. Franzl, C. M. Naumann, and A. Nahr- [9] D. Schneider and A. Egelhaaf, in: Endocrinological stedt, Zoomorphology 108, 183 (1988). frontiers in physiological insect ecology (F. Sehnal, [22] J. M. Pasteels, M. Rowell-Rahier, J. C. Braekman. A. Zabza, and D. L. Denlinger, eds.), p. 37, Techni­ and A. Dupont, Physiol. Entomol. 8 , 307 (1983); M. cal Univ. Press, W roclaw/Poland 1988. Rowell-Rahier and J. M. Pasteels, J. Chem. Ecol. [10] B. Schmitz, M. Buck, A. Egelhaaf, and D. Schnei­ 12, 1189(1986). der, Roux’s Arch. Dev. Biol. 198, 1 (1989). [23] J. M. Pasteels, M. Rowell-Rahier, T. Randoux, J. C. [11] H. Wunderer, K. Hansen, T. W. Bell, D. Schneider, Braekman, and D. Daloze, Entomol. Exp. Appl. 49, and J. Meinwald, Exp. Biol. 46, 11 (1986). 55(1988). [12] T. W. Bell, M. Boppre, D. Schneider, and J. Mein­ [24] A. Wilson, J. Chem. Ecol. 12,49 (1986). wald, Experientia 40, 713(1984). [25] M. Wink, in: Cell culture and somatic cell genetics [13] M. Boppre, in: The biology o f butterflies (R. I. of plants (I. Vasil and F. Constabel, eds.), p. 17. Vane-Wright and P. R. Ackery, eds.), p. 259, Aca­ Academic Press, New York 1987. demic Press, London, New York 1984. [14] R. Bergomaz and M. Boppre, J. Lepidopt. Soc. 40, 385(1986).