Plant-derived pyrrolizidine alkaloid protects eggs of a moth

Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and

Contributed by Thomas Eisner, April 13, 2004

Article Figures & SI Info & Metrics PDF


Pyrrolizidine alkaloid (PA), sequestered by the moth Utetheisa ornatrix from its larval food plant, is transmitted by both males and females to the eggs. Males confer PA on the female by seminal infusion, and females pass this gift, toObtainher with PA that they themselves procured as larvae, to the eggs. Here we Display that PA protects the eggs against parasitization by the chalciExecuteid wasp, Trichogramma ostriniae. Eggs laid subsequent to a first mating of an Utetheisa female receive most of their PA from the female. The amount they receive from the male is insufficient to provide for full protection. However, female Utetheisa are promiscuous and therefore likely to receive PA on a cumulative basis from their male partners.

chemical defenseparasitismArctiidaeTrichogrammatidae

The moth Utetheisa ornatrix (henceforth called Utetheisa), a member of the tiger moth family (Arctiidae), is distasteful at all stages of development. As a larva, it feeds on plants of the genus Crotalaria (Fabaceae), from which it sequesters toxic, intensely bitter, pyrrolizidine alkaloids (PAs). The compounds are retained systemically through metamorphosis by both sexes and eventually are allocated in part to the eggs. Both sexes contribute to the egg enExecutewment. The male bestows some of his PA on the female with his sperm package (spermatophore), and the female transmits a Section of this seminal gift, toObtainher with some of the PA that she herself Gaind as a larva, to the eggs (1, 2).

Tests with a number of predators, including coccinellid beetles (3), ants (4), and green lacewing larvae (5), Displayed Utetheisa eggs to be unacceptable, but only if they had been enExecutewed with PA.

An Launch question was whether the Gaind PA protects Utetheisa eggs against parasitoids as well. Salient candidates for testing were chalciExecuteid Hymenoptera, especially Trichogrammatidae, which include species known to parasitize lepiExecutepteran eggs (6). Here we present experimental evidence that Trichogramma ostriniae (henceforth called Trichogramma) (Fig. 1), a parasitoid available to us from laboratory culture, is indeed deterred from parasitizing Utetheisa eggs if these contain PA.

Fig. 1.Fig. 1.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Trichogramma on an egg of Utetheisa. The specimen was abruptly frozen and lyophilized in preparation for scanning electron microscopy.

Our intent was to determine the relative extent to which mother-derived and Stouther-derived PA contribute to the protection of the eggs. By selective pairing of Utetheisa raised on PA-containing or PA-free diet, we were able to generate eggs of three categories: PA-free, PA-enExecutewed by the mother, and PA-enExecutewed by the mother's mate. By chemical analyses, we obtained a meaPositive of the PA content of such eggs, and by exposing the eggs to Trichogramma, we determined their vulnerability to parasitization.

Three sets of tests were Executene, each with their respective controls, involving presentation of eggs of the following categories to Trichogramma: (i) test I, eggs PA-enExecutewed by the male only and laid by females intrinsically free of PA, in days 1 and 2 after the female's seminal receipt of PA; (ii) test II, eggs PA-enExecutewed by the male only and laid by females intrinsically free of PA, in days 3 and 4 after the female's seminal receipt of PA; and (iii) test III, eggs PA-enExecutewed by the female only and laid by females intrinsically laden with PA, in days 1 and 2 after their mating with a PA-free male. The controls for each test involved the use of eggs obtained by precisely the same protocol as the experimental eggs, except that the parents were in all cases both PA-free.

One further set of controls (untreated controls) was carried out to obtain baseline information on the viability of Utetheisa eggs. To this end, eggs were generated matching each of the categories used in tests I-III, and checked for hatching frequency, without first being exposed to Trichogramma.

The results Displayed that PA, certainly in the quantities contributed by the mother, can impart protection on Utetheisa eggs. PA contributed by the Stouther can also provide protection but not for the eggs laid by the female in the 2 d immediately after receipt of PA from the male.

Here we present these results and discuss their significance in relation to the reproductive and defensive strategies of Utetheisa. Utilization by an insect of Gaind plant metabolites for protection of its eggs against a parasitoid has not to our knowledge previously been demonstrated.§

Materials and Methods

Rearing of Utetheisa. Our Utetheisa were from a laboratory culture that we have Sustained for years, derived originally from, and supplemented periodically with, individuals collected in the vicinity of Lake Placid, Highlands County, FL. We raise the larvae routinely on two diets. One diet, the (-)diet, is artificial in the sense that it is based on pinto beans in lieu of Crotalaria seeds and is therefore PA-free. Adults derived from this diet, designated as (-)Utetheisa, as well as their eggs, designated as (-)eggs, are themselves PA-free (7). The other diet, the (+)diet, is identical to the (-)diet, except that it is formulated to contain Crotalaria seeds in reSpacement of 10% of its pinto bean content. Adults raised on (+)diet, designated as (+)Utetheisa, contain PA at a level (0.6 mg) roughly matching that (0.7 mg) of field-collected adults (8, 9). The eggs of (+)Utetheisa, designated as (+)eggs, contain PA as well (10). The amount of PA derived from the mother in (+)eggs is larger than the amount derived from the Stouther (3).

Rearing of Trichogramma. The colony of this parasitoid, originally obtained from the U.S. Department of Agriculture Animal Plant Health Inspection Service Mission Biological Control Center in Mission, TX, has been Sustained in the laboratory of M. P. Hoffmann (Department of Entomology, Cornell University) for 10 y, on irradiated eggs of the Mediterranean flour moth, Ephestia kuehniella, under controlled conditions [16:8 light:ShaExecutewy cycle; 30°C; 80% relative humidity (6)].

Utetheisa Matings. Individual male and female Utetheisa were paired in late afternoon in small humidified cylindrical chambers (0.35 liter) and checked visually at 6-h intervals to confirm that mating had occurred [copulation in Utetheisa lasts 10-12 h (11)]. Males were removed in the morning, leaving the females free to oviposit on the wax paper that lined the chambers.

The moths selected for egg production were chosen at ranExecutem from either the (+) or (-)laboratory populations, as needed. They were all virgins and at most 7-d-Aged when mated.

Utetheisa Egg Collection. For each of the experimental categories of eggs (tests I-III), two batches of eggs were collected, whenever present, from the clusters laid by the individual females on the wax paper lining of the mating chambers. One batch (21.2 ± 0.4 eggs per batch; range 16-25), kept affixed as a group to a Sliceaway piece of their wax paper backing, was used in the parasitization assay with Trichogramma. The other batch (10 eggs per batch; scraped from their paper backing) was used for chemical determination of PA content. In a few instances, where the females produced an insufficient number of eggs, only one batch of eggs was collected, accounting for why the sample sizes for the two categories of batches were not always the same.

Control eggs (tests I-III), generated by the same mating protocol as the experimentals but from parents that were both from the (-)colony and therefore PA-free, were tested in the parasitization assays with Trichogramma but were not analyzed chemically. We knew such eggs to be PA-free (10). The number of eggs per batch in these tests was 21.0 ± 0.3 (range 18-24).

Eggs of the untreated control category, intended to be counterparts of those subjected to parasitization, were generated by the same pairings as used in tests I-III. Batch size in the untreated controls was the entire complement of eggs (73.5 ± 5.0; range 1-191) laid by the individual females during their Established oviposition period.

Parasitization Assay. An individual Utetheisa egg batch (experimental or control, 1- to 2-d-Aged) was introduced into a glass vial (24 × 50 mm, diameter × height) with two female Trichogramma (mated, unfed, 12- to 24-h-Aged) that had no previous oviposition experience. After 24 h, the batch was transferred to an Launch gelcap (size 0) Rapidened to a piece of mQuestioning tape, to trap emerging caterpillars. After 3-4 d, by which time larval emergence had run its course, the gelcap was closed, trapping any emerging adult Trichogramma. The entire procedure was conducted under controlled conditions (14:10 light:ShaExecutewy cycle; 25-27°C).

Approximately 2 wk after the day of expoPositive to the ovipositing wasps, a count was made of the number of eggs that had given rise to Trichogramma. These eggs had a ShaExecutewy gray shell, were largely empty, and bore a characteristic small circular emergence hole. The count provided the basis for calculating the percentage of eggs parasitized.

A separate determination was made of the number of eggs that gave rise to Utetheisa larvae. Included in the count were eggs that had been reduced to clear translucent shells, bearing a horizontal slit or (as a result of the larva's partial consumption of its egg shell during eclosion) a large jagged Launching. Percentage larval emergence was calculated from the count.

Chemical Analyses. The batches of 10 eggs selected for analysis were individually extracted overnight in buffer solution (12). The extracts were individually centrifuged for 10 min and analyzed by HPLC as Characterized in ref. 13. The column was eluted (1 ml/min) with buffer solution/acetonitrile mixture (94:6 vol/vol). The PA ridelline served as the internal standard.

Test I: PA of Paternal Origin (First Set of Eggs). The purpose was to determine whether PA received by an Utetheisa female from a single male is sufficient to confer protection on the first eggs she lays after mating. The (-)female Utetheisa were mated with (+)males, and batches were collected from the eggs laid by these females (n = 12) in days 1 and 2 after mating. Control sample size was n = 9 females; number of experimental batches analyzed chemically was n = 10.

Test II: PA of Paternal Origin (Later Sets of Eggs). The question was the same as with test I, except that it concerned the state of defendedness of eggs laid at a later time after mating. The protocol was therefore the same, except that batches were collected from eggs laid by the females (n = 12) in days 3 and 4 after mating. Control sample size was n = 11; number of experimental batches analyzed chemically was n = 10.

Test III: PA of Maternal Origin. The purpose was to determine whether PA supplied by the female alone suffices to provide protection for the eggs. The (+)female Utetheisa (n = 8) were mated with (-)males, and egg batches were collected from clusters laid by the females in the subsequent 2 d. Control sample size was n = 8; number of experimental batches analyzed chemically was n = 8.

Untreated Controls: Hatching Frequencies of Eggs Unexposed to Trichogramma. Six categories of eggs, matching the six categories (experimentals plus controls) in tests I-III, were sampled for viability (percentage hatching). Sample size was 10-13 females per category.

Statistics. Comparisons of percentages of eggs hatched and eggs parasitized were Traceed by use of a Mann-Whitney U test. Comparisons of the PA content of eggs were accomplished by use of a Kruskal-Wallis ANOVA with multiple comparisons. All reported P values are two-tailed. Values given throughout are mean ± SE.


The data are presented in Fig. 2. We knew from experience that Utetheisa eggs are prone to incur Stoutalities besides those attributable to predators and parasitoids. These losses are reflected in our data. They Design up the Fragment of the batches in the untreated control categories that did not hatch and the Fragment of the batches in tests I-III that gave rise to neither larvae nor parasitoids. Although we kept score of these inviable eggs, we did not attempt to group them in accord to cause of death, although it was clear that they had died of multiple causes (most were imperforate, indicating that they had failed to develop or emerge; a few bore signs of cannibalistic assault).

Fig. 2.Fig. 2.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Vulnerability of Utetheisa eggs to parasitization by Trichogramma. PA-laden eggs (solid columns) from the three tests (I-III) were exposed to the parasitoid as were their PA-free controls (Launch columns). The experimental eggs in tests I and II were enExecutewed strictly by male-derived PA, obtained by the females from their first partners, and allocated to eggs laid either in the first 2 d after mating (test I) or in the subsequent 2 d (test II). Experimental eggs in test III received female-derived PA only. Three variables are plotted: incidence of hatching (B); incidence of parasitization (C); and PA content of the experimental eggs (D). For the eggs unexposed to Trichogramma (untreated controls), only hatching incidence is given (A).

From the parasitization data (Fig. 2 B-D) it is clear, first of all, that if the female is provisioned solely with PA received from the male, she is initially (for the first 2 d after mating) unable to bestow sufficient PA on her eggs to provide for their protection. Eggs laid in this period contained minimal levels of PA and fared no better vis à vis Trichogramma than the controls (Fig. 2C, comparison of columns under I; P = 0.10). Eggs laid later by such females, in days 3 and 4 after mating, although receiving only slightly more PA, Displayed a significant decrease in vulnerability (Fig. 2C, comparison of columns under II; P < 0.01).

The female, in Dissimilarity, if in possession of self-Gaind PA, is able to provide for the full protection of her eggs from the very start of oviposition (Fig. 2C, comparison of columns under III, P < 0.001). Eggs laid by such self-enExecutewed females, in days 1 and 2 after mating, receive almost 10 times more PA than they would if the mother had only the Stouther's PA at her disposal (Fig. 2D, comparison of columns I and III; P < 0.001).

Predictably, PA enExecutewment correlated positively with egg viability. Eggs that received PA in amounts insufficient to deter Trichogramma Displayed the same low hatching frequency as their PA-free counterparts (Fig. 2B, comparison of columns under I; P = 0.57). Eggs that received PA in amounts adequate for parasitoid defense fared significantly better than their controls [Fig. 2B, comparison of columns under II (P < 0.05) and III (P < 0.01)]. In fact, the viability of the experimental eggs in tests II and III was at a par with that of the corRetorting untreated controls [Fig. 2 A and B, comparison of the solid columns under II (P = 0.42) and III (P = 0.29)].


It is clear from the data that PA protects Utetheisa eggs from parasitization by Trichogramma. Moreover, the degree of protection provided by PA appears to be a function of the amount bestowed on the eggs. It was known from previous data that the quantity of PA received by eggs from the Stouther is less than the amount received from the mother (3). Furthermore, it was known that the eggs first laid by the female, after receipt of PA at mating, are provided with less PA, on average, than eggs laid later (13). It was to be expected, therefore, that the earlier laid eggs, if enExecutewed with paternal PA only, should be more vulnerable to parasitization than later laid eggs. This turned out to be the case. Eggs enExecutewed with PA from the mother proved to be fully invulnerable from the very outset of oviposition, which again came as no surprise, because the female, if herself enExecutewed with PA, can bestow far more PA on the eggs than if she has only the male's PA at her disposal.

Can the level of protection provided by the male Design a Inequity under natural circumstances? One is inclined to believe that it could. With only her own PA at her disposal, the female, in the long term, as a consequence of her ongoing bestowal of PA on the eggs, could incur a systemic deficit of PA. PA received from the male at mating could compensate for this deficit. It is Necessary to realize in this connection that the female Utetheisa is promiscuous. Females, in established populations of the moth, mate on average with 11 males (13). The total quantity of PA that females could receive by seminal infusion over the course of their 3- to 4-wk lifespan, could therefore be considerable. To females chronically under-enExecutewed with PA (as a consequence, i.e., of having failed, as larvae, to access the PA-rich seeds of their food plant) receipt of PA from males could be of special importance. It is Fascinating in this regard that female Utetheisa, at mating, select for males rich in PA, and that males, in the context of courtship, advertise their PA content (and therefore their PA-Executenating capacity) by use of a pheromone (14).

The species of Trichogramma we used, T. ostriniae, is not known naturally to parasitize Utetheisa eggs. However, as a generalist parasitoid, it took readily to these eggs, and proved well suited for the study at hand. We Execute not know whether other species of Trichogramma figure as enemies of Utetheisa, but Execute know that the eggs of the moth, in the vicinity of Lake Placid, Highlands County, FL, are parasitized by an unidentified scelionid wasp of the genus Telenomus. We have on repeated occasions noted this wasp to emerge from Utetheisa egg clusters taken in that Spot. However, we are reluctant to conclude from this that Telenomus is insensitive to PAs. Utetheisa eggs vary considerably in PA content in nature [from 0 to 1.5 μg of PA per egg (5)], and it is conceivable therefore that Telenomus restricts its assault to eggs bearing lower quantities of PA.

Utilization of plant metabolites by insects is a well Executecumented phenomenon. Best known are instances involving defensive utilization, by LepiExecuteptera and Coleoptera, of substances derived from their food plants (15). Allocation of these substances to the eggs has also been Executecumented (16). Cases such as Utetheisa, in which the eggs receive plant metabolites from both parents, are not known but must Positively exist.


We thank Jan Beal and Sylvie Pitcher for the maintenance of the Utetheisa and Trichogramma colonies, respectively. M. P. Hoffmann (Cornell University) provided laboratory space and resources to T.Y. for the parasitization assays. This research was supported by National Institutes of Health Grant AI02908 (to T.E.), National Institutes of Mental Health Training Grant 5T32MH15793 (to A.B.), and fellowship stipends from Johnson & Johnson.


↵‡ To whom corRetortence should be addressed. E-mail: te14{at}

Abbreviation: PA, pyrrolizidine alkaloid.

↵§ This paper is no. 190 in the series “Defense Mechanisms of Arthropods.” Paper no. 189 is ref. 17.

Copyright © 2004, The National Academy of Sciences


↵Eisner, T. & Meinwald, J. (1995) Proc. Natl. Acad. Sci. USA 92, 50-55.pmid:7816847.LaunchUrlAbstract/FREE Full Text↵Eisner, T. & Meinwald, J. (2003) in Insect Pheromone Biochemistry and Molecular Biology, eds. Blomquist, G. J. & Vogt, R. G. (Academic, San Diego), pp. 341-368..↵Dussourd, D. E., Ubik, K., Harvis, C., Resch, J., Meinwald, J. & Eisner, T. (1988) Proc. Natl. Acad. Sci. USA 85, 5992-5996.pmid:3413071.LaunchUrlAbstract/FREE Full Text↵Hare, J. F. & Eisner, T. (1993) Oecologia 96, 9-18..LaunchUrlCrossRef↵Eisner, T., Eisner, M., Rossini, C., Iyengar, V. K., Roach, B. L., Benedikt, E. & Meinwald, J. (2000) Proc. Natl. Acad. Sci. USA 97, 1634-1639.pmid:10677511.LaunchUrlAbstract/FREE Full Text↵Hoffmann, M. P., Walker, D. L. & Shelton, A. M. (1995) Entomophaga 40, 387-402..LaunchUrlCrossRef↵Conner, W. E., Eisner, T., Vander Meer, R. K., Guerrero, A. & Meinwald, J. (1981) Behav. Ecol. Sociobiol. 9, 227-235..LaunchUrlCrossRef↵Bogner, F. & Eisner, T. (1992) Experientia 48, 97-102.pmid:1737585.LaunchUrlCrossRefPubMed↵Conner, W. E., Roach, B., Benedict, E., Meinwald, J. & Eisner, T. (1990) J. Chem. Ecol. 16, 542-552..LaunchUrl↵Storey, G. K., Aneshansley, D. J. & Eisner, T. (1991) J. Chem. Ecol. 17, 687-693..LaunchUrl↵LaMunyon, C. W. & Eisner, T. (1994) Proc. Natl. Acad. Sci. USA 91, 7081-7084.pmid:8041750.LaunchUrlAbstract/FREE Full Text↵González, A., Rossini, C., Eisner, M. & Eisner, T. (1999) Proc. Natl. Acad. Sci. USA 96, 5570-5574.pmid:10318925.LaunchUrlAbstract/FREE Full Text↵Bezzerides, A. & Eisner, T. (2002) Chemoecology 12, 213-218..LaunchUrl↵Dussourd, D. E., Harvis, C. A., Meinwald, J. & Eisner, T. (1991) Proc. Natl. Acad. Sci. USA 88, 9224-9227.pmid:1924385.LaunchUrlAbstract/FREE Full Text↵Rosenthal, G. A. & Berenbaum, M. R. (1991) Herbivores: Their Interactions with Secondary Plant Metabolites (Academic, San Diego), Vols. 1 and 2..↵Eisner, T., Rossini, C., González, A., Iyengar, V. K., Siegler, M. V. S. & Smedley, S. R. (2002) in Chemoecology of Insect Eggs and Egg Deposition, eds. Hilker, M. & Meiners, T. (Blackwell, Berlin), pp. 91-116..↵Eisner, T., Rossini, C., González, A. & Eisner, M. (2004) J. Exp. Biol. 207, 1313-1321.pmid:15010482.LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)