- Email: [email protected]
Caste-specific differences in ecdysteroid titers in early larval stages of the bumblebee Bombus terrestris
Journal of Insect Physiology 46 (2000) 1433–1439 www.elsevier.com/locate/jinsphys
Caste-specific differences in ecdysteroid titers in early larval stages of the bumblebee Bombus terrestris Klaus Hartfelder a b
a,*
, Jonathan Cnaani
b, c
, Abraham Hefetz
b
Departamento de Biologia, FFCLRP–Universidade de Sa˜o Paulo, 14040-901 Ribeirao Preto, SP, Brazil G.S. Wise Faculty of Life Sciences, Department of Zoology, Tel Aviv University, 69978 Tel Aviv, Israel c Carl Hayden Bee Research Center, USDA-ARS, 2000 E. Allen Road, Tucson, AZ 85719-1596, USA Received 15 November 1999; received in revised form 30 March 2000; accepted 8 April 2000
Abstract Mounting evidence implicates ecdysteroids in queen–worker differentiation during the last larval instars of highly social insects. In the present study, we analyzed ecdysteroid titers in queen and worker larvae of the bumblebee Bombus terrestris from the second to the early fourth instar. B. terrestris is of particular interest because caste is already determined in the second instar, presumably by a pheromonal signal emitted by the egg-laying queen. Caste differences in the adults, however, are only expressed at the physiological and not at the morphological level, except for the distinctly larger size of the queen. In the second and third instar, ecdysteroid titers in queen larvae were generally higher than those of workers. These early caste-specific differences, however, were abolished in the fourth instar. In the early fourth instar we could detect two small ecdysteroid peaks, with the one preceding the cocoon-spinning phase presenting the characteristics of a pupal commitment peak. The synchrony of caste differences in ecdysteroid and juvenile hormone titers suggests a synergistic action of these hormones in caste determination. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Ecdysteroids; Radioimmunoassay; Juvenile hormone; Caste development; Bumblebee
1. Introduction Reproductive division of labor or monopolization of reproduction by one or few individuals is a core characteristic of social insect colonies. In the highly eusocial insects this is accompanied by the evolution of phenotypic differences between the highly fertile queens and the functionally or factually sterile workers. As caste development, with rare exceptions, is not based on genetic differences between these morphs, the phenotypic differences between queens and workers have to be accomplished by flexible developmental programs. The observed flexibility in preimaginal development is a prime example of phenotypic plasticity (West-Eberhard, 1989) and has been explored in many of the highly social Hymenoptera. Most bumblebees have an annual cycle of colony
* Corresponding author: Tel.: +55-16-602-3077; fax: +55-16-6336482. E-mail address: [email protected] (K. Hartfelder).
development. An overwintered, fertilized queen selects a suitable nest site and rears a first round of worker brood. These aid the queen in rearing several more brood cycles. In Bombus terrestris, queens start to lay haploid, male-producing eggs at some point in colony social development (switch point), and shortly thereafter some of the workers challenge the queen’s reproductive dominance (competition point) and start to lay haploid eggs themselves (Duchateau and Velthuis, 1988). This is also the time when new queens are reared from diploid eggs laid by the queen, and soon after, the colony begins to disintegrate. These virgin queens are considerably larger than their worker sisters but do not exhibit any further morphological differences. Physiologically, however, virgin queens are clearly distinct from workers, as they are attractive to males, and capable of overwintering and forming a new colony. There is a high correlation between the onset of the competition point and queen rearing, suggesting that social signals affect larval commitment to the queen pathway (Cnaani et al., 2000). The observed incipient caste differentiation in bumblebees is attractive for studying initial steps of
0022-1910/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 2 - 1 9 1 0 ( 0 0 ) 0 0 0 6 7 - 6
1434
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
developmental regulation. Wheeler (1986) has proposed a general model according to which differential feeding of young larvae by adult workers generates a “nutritional switch” that subsequently resets growth parameters in endogenous developmental programs. This nutritional switch can consist of a quantitative and/or qualitative change in the diet with which the larvae are being fed. The bumblebees can be subdivided into two groups with respect to their modes of larval feeding. In the “pocket makers”, as exemplified by Bombus hypnorum, larvae of a single egg batch feed on a common supply of pollen before being individualized at the end of the larval feeding period. In the “pollen storer” species, B. terrestris, the adult workers provide larvae from each egg batch not only with a common pollen supply, but also feed them individually even while they are still reared together in a common wax cell. The presence of a dominant B. terrestris queen has been shown to inhibit larvae from developing into queens during the early phases of the colony cycle (Ro¨seler, 1970; Ro¨seler and Ro¨seler, 1974). A queen pheromone, which has not yet been chemically identified, has been proposed to suppress queen development in larvae during the early phases of the colony life-cycle. Its presence supposedly biases a female larva to enter the worker developmental pathway and inhibits it from becoming a queen. Studies on developmental rates and feeding frequencies in the larval phases of B. terrestris development have not provided evidence for caste-related differences in these parameters (Cnaani et al., 1997; Ribeiro et al., 1999; Ribeiro, 1999; Pereboom, 2000), and induction of physiological caste differences by trophic stimuli can thus apparently be ruled out in this bumblebee species. In practically all social insects, the environmental signals inducing caste development need to be transformed into endogenous signals to result in phenotypic differences. In the primitively eusocial bumblebees, hormonal regulation of caste development has been investigated in two species, in the pocket maker B. hypnorum and in the pollen storer B. terrestris, concentrating mainly on the prepupal phases. Simultaneous determination of juvenile hormone (JH) and ecdysteroid titers by highly specific radioimmunoassays revealed pronounced queen– worker differences in B. hypnorum, whereas in B. terrestris prospective queens and workers exhibited only a temporal shift in the prepupal titer maxima (Strambi et al., 1984). When early larval stages were included in more recent investigations on hormonal regulation of caste differentiation in B. terrestris, strong caste differences in corpora allata activity became evident (Cnaani et al., 1997). These early differences in JH synthesis have also been confirmed by detailed analyses of JH titers in second to early fourth instar larvae (Cnaani, 1997). JH is clearly a key player in caste development and regulation of fertility in social insects (Wheeler, 1986; Hartfelder and Engels, 1998). Since the tight integration
of caste differentiation into metamorphosis suggests that JH exerts this role in concert with ecdysteroids, we investigated in the present study the ecdysteroid titer of prospective queens and workers in the early larval instars. We were able to show that the marked caste differences in ecdysteroid titers observed in these early instars occur in synchrony with JH titer differences. These findings strongly suggest that the two hormones act synergistically in caste development of B. terrrestris.
2. Materials and methods 2.1. Bees Newly established colonies of B. terrestris were obtained from Bio-Bee Sde-Eliyahu Industry, Israel, a few days after the first workers in each colony had emerged. Colonies were maintained in nesting boxes (18×27×12 cm) with a cardboard bottom in an environmental chamber at a temperature of 29±2°C. Sugar syrup (Bee Happy, obtained from Bio-Bee Sde-Eliyahu Industry, Israel) and pollen that was freshly collected by honey bees were supplied to all colonies ad libitum. Development of the colonies was checked daily to monitor the “switch point” and the “competition point” in the colony cycle, as described by Duchateau and Velthuis (1988) and Cnaani et al. (1997). 2.2. Larval development and hemolymph sampling Based on previously established growth parameters (Cnaani et al., 1997), larvae were classified by determining head capsule diameter and weight. The head capsule diameter permitted identification of the instar. Each instar was further divided into five substages according to progressively increasing weight. To do so, logistic regression was used to calculate the weight at which 50% of the queen or worker larvae molted from one instar to the next. These minimal weight limits between two subsequent molts were then divided into five equalweight classes (Table 1). Larvae destined to become workers were obtained from young colonies well before the “competition point”. Prospective queen larvae were obtained from queenless colonies (Ro¨seler, 1970; Cnaani, 1997). We used only larvae that hatched after the queen had been removed. Hemolymph was collected from individual larvae, starting with the early second instar (L2-1) and continuing through to the early spinning-stage in the last, fourth instar (L4-5). Owing to their small size, it was not possible to obtain hemolymph from first instar larvae in sufficient quantities to be included in this study. We terminated sampling at the spinning stage of the fourth instar, because for the subsequent, prepupal stage hormone tit-
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
Table 1 Classification of Bombus terrestris queen and worker larvae. Larval instars were distinguished by head capsule diameter. Each instar was then subdivided into five weight classes (weight in mg) Instar L2-1 L2-2 L2-3 L2-4 L2-5 L3-1 L3-2 L3-3 L3-4 L3-5 L4-1 L4-2 L4-3 L4-4 L4-5
Queen 6.63–15.44 15.45–24.26 24.27–33.07 33.08–41.89 41.90–50.71 50.72–86.27 86.28–121.82 121.83–157.38 157.39–192.93 192.93–228.50 228.51–456.79 456.80–685.09 685.10–913.39 913.40–1141.69 1141.70–1370.00
Worker 6.51–11.40 11.41–16.30 16.31–21.10 21.11–26.10 26.11–31.00 31.01–40.47 40.48–49.95 49.96–59.43 59.44–68.91 68.92–78.39 78.39–132.71 132.72–187.04 187.04–241.35 241.36–295.67 295.68–350.00
ers have already been thoroughly investigated by Strambi et al. (1984). Hemolymph was collected by means of microcapillaries from an incision made into the dorsal vessel.
1435
1985; Feldlaufer and Hartfelder, 1997; Bloch et al., 2000). In total, 139 hemolymph samples from queen and 106 from worker larvae (2–23 individual samples for each phase, as well as pools of hemolymph for early second instar larvae) were analyzed in this study. 2.4. Data analysis Ecdysteroid titers of unknown samples were calculated by log linear regression analysis of standard curve doses (log pg 20E) on logit binding values. The titer values were then plotted against larval developmental stages, or against larval weight. To extract differences specifically related to caste from the raw titer data, which reflect modulations related not only to caste but also to molting cycles, we calculated mean values for the total dataset of each developmental stage (queens plus workers). In a second step we then subtracted this overall mean from the caste-specific mean of each phase. This allowed for a clearer representation of titer differences specifically related to caste development.
3. Results 3.1. Developmental profiles of ecdysteroid titers
2.3. Ecdysone radioimmunoassay Hemolymph volumes from individual samples were quantified in microcapillaries before adding a 100-fold volume of methanol (⫺20°C) to precipitate hemolymph proteins. As only small quantities of hemolymph could be obtained from second instar larvae (L2-2, L2-3, and L2-4 workers), we pooled individual samples obtained for these phases. To extract ecdysteroids, the hemolymph–methanol mixture was thoroughly vortexed and kept at 4°C for at least 1 h before centrifugation (5000g, 10 min). Aliquots of the supernatant were transferred to 1.5 ml Eppendorf vials and the solvent was removed by centrifugation at reduced pressure. Ecdysteroids extracted from the hemolymph samples were quantified by radioimmunoassay (RIA). The antiserum employed in this study had been raised in rabbits injected with a hemisuccinate derivative of ecdysone conjugated at C-22 (Bollenbacher et al., 1983; Warren and Gilbert, 1986); in this RIA it was used at a final concentration of 0.03%. [23,24-3H(N)]ecdysone (NEN, specific activity 102 Ci/mmol) served as labeled ligand. Since standard curves were established using 20-hydroxyecdysone (20E) (Simes, Milan) as nonradioactive ligand, all results are expressed as 20E equivalents. Under the micro-RIA conditions of this assay, the standard curves covered the range of 25–5000 pg. Assay conditions and cross-reactivity factors for quantification of ecdysteroids from honey bee and bumblebee hemolymph have previously been described (Feldlaufer et al.,
The ecdysteroid titers of B. terrestris queen and worker larvae show a typical cyclic profile in the early instars. Peak values are reached approximately midway through the second and third instar (Fig. 1). The highest values of around 100 pg 20E equivalents/µl hemolymph were detected in second instar larvae. In the third instar, maximal titer levels are already considerably lower, thus indicating a continuous decline in subsequent instars. Superimposed on this general developmental profile are caste-specific differences in the ecdysteroid titers. In most of the stages analyzed in this study, queen larvae exhibit higher mean ecdysteroid levels than worker larvae. The differences are most pronounced and statistically significant (Mann–Whitney signed-rank test, P⬍0.05) in the substages of the second (2.3 and 2.5) and third instars (3.2, 3.3, and 3.4). For the fourth larval instar, we determined the ecdysteroid titers of premetamorphic larvae only. In the last substage of the feeding phase (L4-5), the larvae had already started to spin their cocoons while still being fed. Subsequently, they defecate and enter the prepupal stage, for which ecdysteroid titers have already been reported in detail (Strambi et al., 1984). We were able to show that ecdysteroid titers do not exceed basal levels during the feeding phase of the last instar. These low ecdysteroid titers in the last instar are in clear contrast with the earlier instars. When replotting ecdysteroid titers for early fourth instar larvae on a high-resolution scale, and by larval weight instead of the five substages,
1436
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
Fig. 1. Ecdysteroid titers in the feeding phase of Bombus terrestris queen and worker larvae. The developmental profile covers the beginning of the second (L2), third (L3), and first half of the last larval instar (L4). For comparison of equivalent physiological stages, each of these instars was subdivided into five substages, according to progressive weight increase in the two larval types. Titer values are expressed as pg 20-hydroxyecdysone equivalents per µl hemolymph (means±sem). Numbers indicate sample size analyzed for each stage. In second instar larvae, sample numbers marked by an asterisk represent hemolymph samples pooled from several individuals. In the fourth instar, the number of samples for queen larvae precedes that of worker larvae.
we could observe two small ecdysteroid peaks in both queen and worker larvae (Fig. 2). In both castes, the first peak occurs shortly after the molt to the fourth instar. The second peak, which is even smaller than the first
peak, precedes the onset of cocoon-spinning activities by the larvae, and thus can be interpreted as an entry point to the first metamorphic molt. Such timing is characteristic of a commitment peak in holometabolous insects (Riddiford 1978, 1994), the occurrence of which has not been previously described in aculeate hymenopterans.
3.2. Queen–worker differences—synchrony of ecdysteroids with JH
Fig. 2. Ecdysteroid titer in fourth instar queen and worker larvae of Bombus terrestris plotted against larval weight. Even though generally at low levels, the ecdysteroid titer shows coordinated fluctuations in both castes, with an early peak shortly after ecdysis and a second peak preceding the cocoon spinning stage.
A parallel investigation, using the same staging method, also revealed caste-specific differences in JH synthesis and JH titer of B. terrestris larvae (Cnaani, 1997; Cnaani et al., 1997). When we calculated the deviation between the respective titer values for queens and workers, a surprising synchrony became evident in the caste-specific modulation of JH and ecdysteroid titers (Fig. 3). Both hormones exhibit maximal caste differences during the second and particularly during the third instar, with practically zero difference in the feeding stage of the fourth instar. This observed synchrony in JH and ecdysteroid titer differences in B. terrestris larvae suggests a synergistic mode of action for these morphogenetic hormones in bumblebee caste development.
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
Fig. 3. Synchrony in JH and ecdysteroid titer differences in queen and worker larvae of Bombus terrestris. Mean values for each larval stage were normalized (for details see data analysis in Section 2) to depict the deviation from the overall mean for all females, represented as baselines in these graphs. This representation, therefore, illustrates relative titer differences for queens and workers, respectively. Castespecific differences in hormone titers are pronounced in the second (L2) and even more so in the third larval instar (L3). JH titers in these instars were determined by Cnaani (1997).
4. Discussion Bombus terrestris is now the first species of social Hymenoptera for which detailed ecdysteroid measurements are also available for the early larval stages. An earlier study on hormonal regulation of caste development (Strambi et al., 1984) focused on the last larval instar, considering the differentiation of caste-specific characters as an integral element of metamorphosis. There has been only one investigation to date on hormone titers in the early larval instars of bees. This exceptional study was carried out on JH contents in queens and workers of the honey bee, Apis mellifera, starting with embryonic stages and covering the entire preimaginal period until emergence of the adults (Rembold, 1987; Rembold et al., 1992). Ecdysteroid titer fluctuations in the early larval stages and their relation to caste development in honey bees, however, have not yet been subject to investigation. Our present data on ecdysteroid titers in bumblebee larvae provide evidence that not only JH but also molting hormones may play a very early role in caste development. When combining our results with those obtained by Strambi et al. (1984), the cyclicity in ecdysteroid titer profiles for the larval instars becomes apparent. Interestingly, the highest titer values were observed in second
1437
instar larvae. Furthermore, we could detect two small ecdysteroid peaks in the feeding phase of the last larval instar. Such minor peaks have been observed in other holometabolous insects as well (for review see Riddiford, 1994), but have not yet been described for aculeate hymenopterans. By analogy to timing and function in Manduca sexta, at least the small ecdysteroid peak preceding the cocoon-spinning phase in bumblebee larvae is reminiscent of the commitment peak of M. sexta, and could thus be involved in developmental reprogramming in bee larvae as well. The apparent caste-specific timing of the putative commitment peak may even explain the differences in timing observed for the prepupal ecdysteroid peak in queens and workers of B. terrestris (Strambi et al., 1984). The parallel analyses on JH and ecdysteroid titers in bumblebee larvae raise two questions of a general nature with regard to the hormonal regulation of caste development. The first concerns the temporal profile of JH and ecdysteroid titers and possible regulatory implications. The synchrony observed for the differences in JH and ecdysteroid titers between queen and worker larvae, especially during the early larval stages, is quite astounding, considering that relative titer differences are much more relevant than absolute titer values to trigger stageand tissue-specific hormone response cascades. We can therefore also expect to detect in the early larval stages of honey bees and bumblebees an interendocrine regulation of prothoracic gland activity by JH, similar to that documented for last instar larvae of M. sexta (Bollenbacher, 1988) and also last-instar honey bee larvae (Rachinsky and Engels, 1995). The second question raised by these results concerns possible evolutionary implications of endocrine regulation of caste development in social bees. More precisely, the question is how the differences in hormone titers shown for the early larval stages of B. terrestris result in physiological caste differences, and how the endocrine differences observed in the highly eusocial honey bees and stingless bees elicit morphological caste differentiation in addition to physiological differences already observed in the primitively social bumblebees. Considering the rather small database of only four bee species, for which the endocrinology of caste development has been sufficiently well studied so far, this question can only be answered tentatively. The endocrine program of caste development in B. terrestris presents remarkable quantitative differences in the early larval stages (Cnaani, 1997; Cnaani et al., 1997; and this study), but in the prepupal period it exhibits only differences in timing and not in titer levels (Strambi et al., 1984). The hormonal differences in the second and third larval instars of B. terrestris can be interpreted as a consequence of the suppression of queen development by the dominant egg-laying queen during the early phases of the colony cycle (Ro¨seler, 1970). Even though
1438
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
some effects attributed to a putative bumblebee queen pheromone have recently been called into question (Bloch and Hefetz, 1999), the presence of the queen during a critical phase in the second larval instar has been singled out as the prime inhibitory signal for queen development in B. terrestris larvae. The endocrine system appears to show a strong response to such a signal. With progressing larval development, however, the caste differences in JH and ecdysteroid titers then seem to diminish. In contrast, the endocrine system in the highly eusocial honey bees and stingless bees exhibits phases of differential activity not only in the early larval instars but also during the metamorphic stages (Rembold, 1987; Rachinsky et al., 1990; Hartfelder and Rembold, 1991; Rembold et al., 1992), resulting in morphological caste phenotypes. The transition in socioevolution from physiological queen–worker differences to morphologically differentiated caste phenotypes may have as a developmental basis differential responses of different tissues to castespecifically-modulated hormone titers. This involves differences in the timing of competence periods, as well as in hormone response thresholds. Caste-specificallymodulated hormone titers in the early larval stages could primarily affect physiological parameters associated with reproduction, a response exemplified by B. terrestris. According to this model, physiologically and morphologically distinct castes should require caste-specificallymodulated hormone titers in the early larval stages, as well as in the metamorphic phases, as is indeed the case in the honey bee (Rembold, 1987, Rembold et al., 1992). Such a distinction between physiologically and morphologically differentiated castes obviously only refers to the final result of caste development, while in their mode of action, JH and ecdysteroid may well exhibit a mix of effects during the different developmental stages. In fact, recent studies on differential gene expression (Evans and Wheeler, 1999) and on programmed cell death (Schmidt Capella and Hartfelder, 1998) in honey bee caste development indicate that JH acts in both directions, regulating the expression of metabolic enzymes, as well as cell cycle and cell death programs to generate physiological and morphological caste differences. The major challenge will now definitely be to decipher tissue-specific responses to morphogenetic hormones via such molecular and cell biology approaches to understand caste development in social insects.
Acknowledgements We thank Nomi Paz for proofreading the manuscript, and an anonymous referee for critical comments. Financial support was provided by the DFG (Ha 1625/3-1) and a CAPES/DAAD visiting-professor fellowship to K.H.
References Bloch, G., Hefetz, A., 1999. Reevaluation of the role of mandibular glands in regulation of reproduction in bumblebee colonies. Journal of Chemical Ecology 25, 881–896. Bloch, G., Hefetz, A., Hartfelder, K., 2000. Ecdysteroid titer, ovary status, and dominance in adult worker and queen bumble bees (Bombus terrestris). Journal of Insect Physiology 46, 1033–1040. Bollenbacher, W.E., 1988. The interendocrine regulation of larval– pupal development in the tobacco hornworm, Manduca sexta—a model. Journal of Insect Physiology 34, 941–947. Bollenbacher, W.E., O’Brien, M.A., Katahira, E.J., Gilbert, L.I., 1983. A kinetic-analysis of the action of the insect prothoracicotropic hormone. Molecular and Cellular Endocrinology 32, 27–46. Cnaani, J., 1997. Larval development and caste differentiation in the bumble bee Bombus terrestris. PhD thesis, Tel Aviv University. Cnaani, J., Borst, D.W., Huang, Z.Y., Robinson, G.E., Hefetz, A., 1997. Caste determination in Bombus terrestris: differences in development and rates of JH biosynthesis between queen and worker larvae. Journal of Insect Physiology 43, 373–381. Cnaani, J., Robinson, G.E., Bloch, G., Borst, D.W., Hefetz, A., 2000. The effect of queen–worker conflict on caste determination in the bumble bee Bombus terrestris. Behavioral Ecology and Sociobiology 47, 346–352. Duchateau, M.J., Velthuis, H.H.W., 1988. Development and reproductive strategies in Bombus terrestris colonies. Behaviour 107, 186–207. Evans, J.D., Wheeler, D.E., 1999. Differential gene expression between developing queens and workers in the honey bee, Apis mellifera. Proceedings of the National Academy of Science USA 96, 575–580. Feldlaufer, M.F., Hartfelder, K., 1997. Relationship of the neutral sterols and ecdysteroids of the parasitic mite, Varroa jacobsoni to those of the honey bee, Apis mellifera. Journal of Insect Physiology 43, 541–545. Feldlaufer, M.F., Herbert, E.W., Svoboda, J.A., Thompson, M.J., Lusby, W.R., 1985. Makisterone A—the major ecdysteroid from the pupa of the honey bee, Apis mellifera. Insect Biochemistry 15, 597–600. Hartfelder, K., Engels, W., 1998. Social insect polymorphism: Hormonal regulation of plasticity in development and reproduction in the honeybee. Current Topics in Developmental Biology 40, 45–77. Hartfelder, K., Rembold, H., 1991. Caste-specific modulation of juvenile hormone. III. Content and ecdysteroid titer in postembryonic development of the stingless bee, Scaptotrigona postica depilis. Journal of Comparative Physiology B—Biochemical, Systemic and Environmental Physiology 160, 617–620. Pereboom, Z., 2000. The composition of larval food and the significance of exocrine secretions in the bumblebee Bombus terrestris. Insectes Sociaux 47, 11–20. Rachinsky, A., Engels, W., 1995. Caste development in honey bees (Apis mellifera): Juvenile hormone turns on ecdysteroids. Naturwissenschaften 82, 378–379. Rachinsky, A., Strambi, C., Strambi, A., Hartfelder, K., 1990. Caste and metamorphosis: Hemolymph titers of juvenile hormone and ecdysteroids in last instar honey bee larvae. General and Comparative Endocrinology 79, 31–38. Rembold, H., 1987. Caste-specific modulation of juvenile hormone titers in Apis mellifera. Insect Biochemistry 17, 1003–1006. Rembold, H., Czoppelt, C., Gru¨ne, M., Lackner, B., Pfeffer, J., Woker, E., 1992. Juvenile hormone titers during honey bee embryogenesis and metamorphosis. In: Mauchamp, B., Couillaud, F., Baehr, J.C. (Eds.), Insect Juvenile Hormone Research. INRA, Paris, pp. 37–43. Ribeiro, M., 1999. Long duration feedings and caste differentiation in Bombus terrestris larvae. Insectes Sociaux 46, 315–322. Ribeiro, M., Velthuis, H.H.W., Duchateaux, M.J., van der Tweel, I.,
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
1999. Feeding frequency and caste differentiation in Bombus terrestris larvae. Insectes Sociaux 46, 306–314. Riddiford, L., 1978. Ecdysone-induced changes in cellular commitment of the epidermis of the tobacco hornworm, Manduca sexta, at the initiation of metamorphosis. General and Comparative Endocrinology 34, 438–446. Riddiford, L., 1994. Cellular and molecular actions of juvenile hormone. I. General considerations and premetamorphic actions. Advances in Insect Physiology 24, 213–274. Ro¨seler, P.-F., 1970. Unterschiede in der Kastendetermination zwischen den Hummelarten Bombus hypnorum und Bombus terrestris. Zeitschrift fu¨r Naturwissenschaften B 25, 543–548. Ro¨seler, P.-F., Ro¨seler, I., 1974. Morphologische und physiologische Differenzierung der Kasten bei den Hummelarten Bombus hypnorum (L.) und Bombus terrestris (L.). Zoologische Jahrbu¨cher Abteilung Allgemeine Zoologie und Physiologie der Tiere 78, 175–198. Schmidt Capella, I.C., Hartfelder, K., 1998. Juvenile hormone effect
1439
on DNA synthesis and apoptosis in caste-specific differentiation of the larval honey bee (Apis mellifera L.) ovary. Journal of Insect Physiology 44, 385–391. Strambi, A., Strambi, C., Ro¨seler, P.F., Ro¨seler, I., 1984. Simultaneous determination of juvenile hormone and ecdysteroid titers in the hemolymph of bumblebee prepupae (Bombus hypnorum and B. terrestris). General and Comparative Endocrinology 55, 83–88. Warren, J.T., Gilbert, L.I., 1986. Ecdysone metabolism and distribution during the pupal–adult development of Manduca sexta. Insect Biochemistry 16, 65–82. West-Eberhard, M.J., 1989. Phenotypic plasticity and the origins of diversity. Annual Reviews of Ecology and Systematics 20, 249– 278. Wheeler, D.E., 1986. Developmental and physiological determinants of caste in social Hymenoptera: evolutionary implications. American Naturalist 128, 13–24.
Caste-specific differences in ecdysteroid titers in early larval stages of the bumblebee Bombus terrestris Klaus Hartfelder a b
a,*
, Jonathan Cnaani
b, c
, Abraham Hefetz
b
Departamento de Biologia, FFCLRP–Universidade de Sa˜o Paulo, 14040-901 Ribeirao Preto, SP, Brazil G.S. Wise Faculty of Life Sciences, Department of Zoology, Tel Aviv University, 69978 Tel Aviv, Israel c Carl Hayden Bee Research Center, USDA-ARS, 2000 E. Allen Road, Tucson, AZ 85719-1596, USA Received 15 November 1999; received in revised form 30 March 2000; accepted 8 April 2000
Abstract Mounting evidence implicates ecdysteroids in queen–worker differentiation during the last larval instars of highly social insects. In the present study, we analyzed ecdysteroid titers in queen and worker larvae of the bumblebee Bombus terrestris from the second to the early fourth instar. B. terrestris is of particular interest because caste is already determined in the second instar, presumably by a pheromonal signal emitted by the egg-laying queen. Caste differences in the adults, however, are only expressed at the physiological and not at the morphological level, except for the distinctly larger size of the queen. In the second and third instar, ecdysteroid titers in queen larvae were generally higher than those of workers. These early caste-specific differences, however, were abolished in the fourth instar. In the early fourth instar we could detect two small ecdysteroid peaks, with the one preceding the cocoon-spinning phase presenting the characteristics of a pupal commitment peak. The synchrony of caste differences in ecdysteroid and juvenile hormone titers suggests a synergistic action of these hormones in caste determination. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Ecdysteroids; Radioimmunoassay; Juvenile hormone; Caste development; Bumblebee
1. Introduction Reproductive division of labor or monopolization of reproduction by one or few individuals is a core characteristic of social insect colonies. In the highly eusocial insects this is accompanied by the evolution of phenotypic differences between the highly fertile queens and the functionally or factually sterile workers. As caste development, with rare exceptions, is not based on genetic differences between these morphs, the phenotypic differences between queens and workers have to be accomplished by flexible developmental programs. The observed flexibility in preimaginal development is a prime example of phenotypic plasticity (West-Eberhard, 1989) and has been explored in many of the highly social Hymenoptera. Most bumblebees have an annual cycle of colony
* Corresponding author: Tel.: +55-16-602-3077; fax: +55-16-6336482. E-mail address: [email protected] (K. Hartfelder).
development. An overwintered, fertilized queen selects a suitable nest site and rears a first round of worker brood. These aid the queen in rearing several more brood cycles. In Bombus terrestris, queens start to lay haploid, male-producing eggs at some point in colony social development (switch point), and shortly thereafter some of the workers challenge the queen’s reproductive dominance (competition point) and start to lay haploid eggs themselves (Duchateau and Velthuis, 1988). This is also the time when new queens are reared from diploid eggs laid by the queen, and soon after, the colony begins to disintegrate. These virgin queens are considerably larger than their worker sisters but do not exhibit any further morphological differences. Physiologically, however, virgin queens are clearly distinct from workers, as they are attractive to males, and capable of overwintering and forming a new colony. There is a high correlation between the onset of the competition point and queen rearing, suggesting that social signals affect larval commitment to the queen pathway (Cnaani et al., 2000). The observed incipient caste differentiation in bumblebees is attractive for studying initial steps of
0022-1910/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 2 - 1 9 1 0 ( 0 0 ) 0 0 0 6 7 - 6
1434
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
developmental regulation. Wheeler (1986) has proposed a general model according to which differential feeding of young larvae by adult workers generates a “nutritional switch” that subsequently resets growth parameters in endogenous developmental programs. This nutritional switch can consist of a quantitative and/or qualitative change in the diet with which the larvae are being fed. The bumblebees can be subdivided into two groups with respect to their modes of larval feeding. In the “pocket makers”, as exemplified by Bombus hypnorum, larvae of a single egg batch feed on a common supply of pollen before being individualized at the end of the larval feeding period. In the “pollen storer” species, B. terrestris, the adult workers provide larvae from each egg batch not only with a common pollen supply, but also feed them individually even while they are still reared together in a common wax cell. The presence of a dominant B. terrestris queen has been shown to inhibit larvae from developing into queens during the early phases of the colony cycle (Ro¨seler, 1970; Ro¨seler and Ro¨seler, 1974). A queen pheromone, which has not yet been chemically identified, has been proposed to suppress queen development in larvae during the early phases of the colony life-cycle. Its presence supposedly biases a female larva to enter the worker developmental pathway and inhibits it from becoming a queen. Studies on developmental rates and feeding frequencies in the larval phases of B. terrestris development have not provided evidence for caste-related differences in these parameters (Cnaani et al., 1997; Ribeiro et al., 1999; Ribeiro, 1999; Pereboom, 2000), and induction of physiological caste differences by trophic stimuli can thus apparently be ruled out in this bumblebee species. In practically all social insects, the environmental signals inducing caste development need to be transformed into endogenous signals to result in phenotypic differences. In the primitively eusocial bumblebees, hormonal regulation of caste development has been investigated in two species, in the pocket maker B. hypnorum and in the pollen storer B. terrestris, concentrating mainly on the prepupal phases. Simultaneous determination of juvenile hormone (JH) and ecdysteroid titers by highly specific radioimmunoassays revealed pronounced queen– worker differences in B. hypnorum, whereas in B. terrestris prospective queens and workers exhibited only a temporal shift in the prepupal titer maxima (Strambi et al., 1984). When early larval stages were included in more recent investigations on hormonal regulation of caste differentiation in B. terrestris, strong caste differences in corpora allata activity became evident (Cnaani et al., 1997). These early differences in JH synthesis have also been confirmed by detailed analyses of JH titers in second to early fourth instar larvae (Cnaani, 1997). JH is clearly a key player in caste development and regulation of fertility in social insects (Wheeler, 1986; Hartfelder and Engels, 1998). Since the tight integration
of caste differentiation into metamorphosis suggests that JH exerts this role in concert with ecdysteroids, we investigated in the present study the ecdysteroid titer of prospective queens and workers in the early larval instars. We were able to show that the marked caste differences in ecdysteroid titers observed in these early instars occur in synchrony with JH titer differences. These findings strongly suggest that the two hormones act synergistically in caste development of B. terrrestris.
2. Materials and methods 2.1. Bees Newly established colonies of B. terrestris were obtained from Bio-Bee Sde-Eliyahu Industry, Israel, a few days after the first workers in each colony had emerged. Colonies were maintained in nesting boxes (18×27×12 cm) with a cardboard bottom in an environmental chamber at a temperature of 29±2°C. Sugar syrup (Bee Happy, obtained from Bio-Bee Sde-Eliyahu Industry, Israel) and pollen that was freshly collected by honey bees were supplied to all colonies ad libitum. Development of the colonies was checked daily to monitor the “switch point” and the “competition point” in the colony cycle, as described by Duchateau and Velthuis (1988) and Cnaani et al. (1997). 2.2. Larval development and hemolymph sampling Based on previously established growth parameters (Cnaani et al., 1997), larvae were classified by determining head capsule diameter and weight. The head capsule diameter permitted identification of the instar. Each instar was further divided into five substages according to progressively increasing weight. To do so, logistic regression was used to calculate the weight at which 50% of the queen or worker larvae molted from one instar to the next. These minimal weight limits between two subsequent molts were then divided into five equalweight classes (Table 1). Larvae destined to become workers were obtained from young colonies well before the “competition point”. Prospective queen larvae were obtained from queenless colonies (Ro¨seler, 1970; Cnaani, 1997). We used only larvae that hatched after the queen had been removed. Hemolymph was collected from individual larvae, starting with the early second instar (L2-1) and continuing through to the early spinning-stage in the last, fourth instar (L4-5). Owing to their small size, it was not possible to obtain hemolymph from first instar larvae in sufficient quantities to be included in this study. We terminated sampling at the spinning stage of the fourth instar, because for the subsequent, prepupal stage hormone tit-
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
Table 1 Classification of Bombus terrestris queen and worker larvae. Larval instars were distinguished by head capsule diameter. Each instar was then subdivided into five weight classes (weight in mg) Instar L2-1 L2-2 L2-3 L2-4 L2-5 L3-1 L3-2 L3-3 L3-4 L3-5 L4-1 L4-2 L4-3 L4-4 L4-5
Queen 6.63–15.44 15.45–24.26 24.27–33.07 33.08–41.89 41.90–50.71 50.72–86.27 86.28–121.82 121.83–157.38 157.39–192.93 192.93–228.50 228.51–456.79 456.80–685.09 685.10–913.39 913.40–1141.69 1141.70–1370.00
Worker 6.51–11.40 11.41–16.30 16.31–21.10 21.11–26.10 26.11–31.00 31.01–40.47 40.48–49.95 49.96–59.43 59.44–68.91 68.92–78.39 78.39–132.71 132.72–187.04 187.04–241.35 241.36–295.67 295.68–350.00
ers have already been thoroughly investigated by Strambi et al. (1984). Hemolymph was collected by means of microcapillaries from an incision made into the dorsal vessel.
1435
1985; Feldlaufer and Hartfelder, 1997; Bloch et al., 2000). In total, 139 hemolymph samples from queen and 106 from worker larvae (2–23 individual samples for each phase, as well as pools of hemolymph for early second instar larvae) were analyzed in this study. 2.4. Data analysis Ecdysteroid titers of unknown samples were calculated by log linear regression analysis of standard curve doses (log pg 20E) on logit binding values. The titer values were then plotted against larval developmental stages, or against larval weight. To extract differences specifically related to caste from the raw titer data, which reflect modulations related not only to caste but also to molting cycles, we calculated mean values for the total dataset of each developmental stage (queens plus workers). In a second step we then subtracted this overall mean from the caste-specific mean of each phase. This allowed for a clearer representation of titer differences specifically related to caste development.
3. Results 3.1. Developmental profiles of ecdysteroid titers
2.3. Ecdysone radioimmunoassay Hemolymph volumes from individual samples were quantified in microcapillaries before adding a 100-fold volume of methanol (⫺20°C) to precipitate hemolymph proteins. As only small quantities of hemolymph could be obtained from second instar larvae (L2-2, L2-3, and L2-4 workers), we pooled individual samples obtained for these phases. To extract ecdysteroids, the hemolymph–methanol mixture was thoroughly vortexed and kept at 4°C for at least 1 h before centrifugation (5000g, 10 min). Aliquots of the supernatant were transferred to 1.5 ml Eppendorf vials and the solvent was removed by centrifugation at reduced pressure. Ecdysteroids extracted from the hemolymph samples were quantified by radioimmunoassay (RIA). The antiserum employed in this study had been raised in rabbits injected with a hemisuccinate derivative of ecdysone conjugated at C-22 (Bollenbacher et al., 1983; Warren and Gilbert, 1986); in this RIA it was used at a final concentration of 0.03%. [23,24-3H(N)]ecdysone (NEN, specific activity 102 Ci/mmol) served as labeled ligand. Since standard curves were established using 20-hydroxyecdysone (20E) (Simes, Milan) as nonradioactive ligand, all results are expressed as 20E equivalents. Under the micro-RIA conditions of this assay, the standard curves covered the range of 25–5000 pg. Assay conditions and cross-reactivity factors for quantification of ecdysteroids from honey bee and bumblebee hemolymph have previously been described (Feldlaufer et al.,
The ecdysteroid titers of B. terrestris queen and worker larvae show a typical cyclic profile in the early instars. Peak values are reached approximately midway through the second and third instar (Fig. 1). The highest values of around 100 pg 20E equivalents/µl hemolymph were detected in second instar larvae. In the third instar, maximal titer levels are already considerably lower, thus indicating a continuous decline in subsequent instars. Superimposed on this general developmental profile are caste-specific differences in the ecdysteroid titers. In most of the stages analyzed in this study, queen larvae exhibit higher mean ecdysteroid levels than worker larvae. The differences are most pronounced and statistically significant (Mann–Whitney signed-rank test, P⬍0.05) in the substages of the second (2.3 and 2.5) and third instars (3.2, 3.3, and 3.4). For the fourth larval instar, we determined the ecdysteroid titers of premetamorphic larvae only. In the last substage of the feeding phase (L4-5), the larvae had already started to spin their cocoons while still being fed. Subsequently, they defecate and enter the prepupal stage, for which ecdysteroid titers have already been reported in detail (Strambi et al., 1984). We were able to show that ecdysteroid titers do not exceed basal levels during the feeding phase of the last instar. These low ecdysteroid titers in the last instar are in clear contrast with the earlier instars. When replotting ecdysteroid titers for early fourth instar larvae on a high-resolution scale, and by larval weight instead of the five substages,
1436
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
Fig. 1. Ecdysteroid titers in the feeding phase of Bombus terrestris queen and worker larvae. The developmental profile covers the beginning of the second (L2), third (L3), and first half of the last larval instar (L4). For comparison of equivalent physiological stages, each of these instars was subdivided into five substages, according to progressive weight increase in the two larval types. Titer values are expressed as pg 20-hydroxyecdysone equivalents per µl hemolymph (means±sem). Numbers indicate sample size analyzed for each stage. In second instar larvae, sample numbers marked by an asterisk represent hemolymph samples pooled from several individuals. In the fourth instar, the number of samples for queen larvae precedes that of worker larvae.
we could observe two small ecdysteroid peaks in both queen and worker larvae (Fig. 2). In both castes, the first peak occurs shortly after the molt to the fourth instar. The second peak, which is even smaller than the first
peak, precedes the onset of cocoon-spinning activities by the larvae, and thus can be interpreted as an entry point to the first metamorphic molt. Such timing is characteristic of a commitment peak in holometabolous insects (Riddiford 1978, 1994), the occurrence of which has not been previously described in aculeate hymenopterans.
3.2. Queen–worker differences—synchrony of ecdysteroids with JH
Fig. 2. Ecdysteroid titer in fourth instar queen and worker larvae of Bombus terrestris plotted against larval weight. Even though generally at low levels, the ecdysteroid titer shows coordinated fluctuations in both castes, with an early peak shortly after ecdysis and a second peak preceding the cocoon spinning stage.
A parallel investigation, using the same staging method, also revealed caste-specific differences in JH synthesis and JH titer of B. terrestris larvae (Cnaani, 1997; Cnaani et al., 1997). When we calculated the deviation between the respective titer values for queens and workers, a surprising synchrony became evident in the caste-specific modulation of JH and ecdysteroid titers (Fig. 3). Both hormones exhibit maximal caste differences during the second and particularly during the third instar, with practically zero difference in the feeding stage of the fourth instar. This observed synchrony in JH and ecdysteroid titer differences in B. terrestris larvae suggests a synergistic mode of action for these morphogenetic hormones in bumblebee caste development.
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
Fig. 3. Synchrony in JH and ecdysteroid titer differences in queen and worker larvae of Bombus terrestris. Mean values for each larval stage were normalized (for details see data analysis in Section 2) to depict the deviation from the overall mean for all females, represented as baselines in these graphs. This representation, therefore, illustrates relative titer differences for queens and workers, respectively. Castespecific differences in hormone titers are pronounced in the second (L2) and even more so in the third larval instar (L3). JH titers in these instars were determined by Cnaani (1997).
4. Discussion Bombus terrestris is now the first species of social Hymenoptera for which detailed ecdysteroid measurements are also available for the early larval stages. An earlier study on hormonal regulation of caste development (Strambi et al., 1984) focused on the last larval instar, considering the differentiation of caste-specific characters as an integral element of metamorphosis. There has been only one investigation to date on hormone titers in the early larval instars of bees. This exceptional study was carried out on JH contents in queens and workers of the honey bee, Apis mellifera, starting with embryonic stages and covering the entire preimaginal period until emergence of the adults (Rembold, 1987; Rembold et al., 1992). Ecdysteroid titer fluctuations in the early larval stages and their relation to caste development in honey bees, however, have not yet been subject to investigation. Our present data on ecdysteroid titers in bumblebee larvae provide evidence that not only JH but also molting hormones may play a very early role in caste development. When combining our results with those obtained by Strambi et al. (1984), the cyclicity in ecdysteroid titer profiles for the larval instars becomes apparent. Interestingly, the highest titer values were observed in second
1437
instar larvae. Furthermore, we could detect two small ecdysteroid peaks in the feeding phase of the last larval instar. Such minor peaks have been observed in other holometabolous insects as well (for review see Riddiford, 1994), but have not yet been described for aculeate hymenopterans. By analogy to timing and function in Manduca sexta, at least the small ecdysteroid peak preceding the cocoon-spinning phase in bumblebee larvae is reminiscent of the commitment peak of M. sexta, and could thus be involved in developmental reprogramming in bee larvae as well. The apparent caste-specific timing of the putative commitment peak may even explain the differences in timing observed for the prepupal ecdysteroid peak in queens and workers of B. terrestris (Strambi et al., 1984). The parallel analyses on JH and ecdysteroid titers in bumblebee larvae raise two questions of a general nature with regard to the hormonal regulation of caste development. The first concerns the temporal profile of JH and ecdysteroid titers and possible regulatory implications. The synchrony observed for the differences in JH and ecdysteroid titers between queen and worker larvae, especially during the early larval stages, is quite astounding, considering that relative titer differences are much more relevant than absolute titer values to trigger stageand tissue-specific hormone response cascades. We can therefore also expect to detect in the early larval stages of honey bees and bumblebees an interendocrine regulation of prothoracic gland activity by JH, similar to that documented for last instar larvae of M. sexta (Bollenbacher, 1988) and also last-instar honey bee larvae (Rachinsky and Engels, 1995). The second question raised by these results concerns possible evolutionary implications of endocrine regulation of caste development in social bees. More precisely, the question is how the differences in hormone titers shown for the early larval stages of B. terrestris result in physiological caste differences, and how the endocrine differences observed in the highly eusocial honey bees and stingless bees elicit morphological caste differentiation in addition to physiological differences already observed in the primitively social bumblebees. Considering the rather small database of only four bee species, for which the endocrinology of caste development has been sufficiently well studied so far, this question can only be answered tentatively. The endocrine program of caste development in B. terrestris presents remarkable quantitative differences in the early larval stages (Cnaani, 1997; Cnaani et al., 1997; and this study), but in the prepupal period it exhibits only differences in timing and not in titer levels (Strambi et al., 1984). The hormonal differences in the second and third larval instars of B. terrestris can be interpreted as a consequence of the suppression of queen development by the dominant egg-laying queen during the early phases of the colony cycle (Ro¨seler, 1970). Even though
1438
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
some effects attributed to a putative bumblebee queen pheromone have recently been called into question (Bloch and Hefetz, 1999), the presence of the queen during a critical phase in the second larval instar has been singled out as the prime inhibitory signal for queen development in B. terrestris larvae. The endocrine system appears to show a strong response to such a signal. With progressing larval development, however, the caste differences in JH and ecdysteroid titers then seem to diminish. In contrast, the endocrine system in the highly eusocial honey bees and stingless bees exhibits phases of differential activity not only in the early larval instars but also during the metamorphic stages (Rembold, 1987; Rachinsky et al., 1990; Hartfelder and Rembold, 1991; Rembold et al., 1992), resulting in morphological caste phenotypes. The transition in socioevolution from physiological queen–worker differences to morphologically differentiated caste phenotypes may have as a developmental basis differential responses of different tissues to castespecifically-modulated hormone titers. This involves differences in the timing of competence periods, as well as in hormone response thresholds. Caste-specificallymodulated hormone titers in the early larval stages could primarily affect physiological parameters associated with reproduction, a response exemplified by B. terrestris. According to this model, physiologically and morphologically distinct castes should require caste-specificallymodulated hormone titers in the early larval stages, as well as in the metamorphic phases, as is indeed the case in the honey bee (Rembold, 1987, Rembold et al., 1992). Such a distinction between physiologically and morphologically differentiated castes obviously only refers to the final result of caste development, while in their mode of action, JH and ecdysteroid may well exhibit a mix of effects during the different developmental stages. In fact, recent studies on differential gene expression (Evans and Wheeler, 1999) and on programmed cell death (Schmidt Capella and Hartfelder, 1998) in honey bee caste development indicate that JH acts in both directions, regulating the expression of metabolic enzymes, as well as cell cycle and cell death programs to generate physiological and morphological caste differences. The major challenge will now definitely be to decipher tissue-specific responses to morphogenetic hormones via such molecular and cell biology approaches to understand caste development in social insects.
Acknowledgements We thank Nomi Paz for proofreading the manuscript, and an anonymous referee for critical comments. Financial support was provided by the DFG (Ha 1625/3-1) and a CAPES/DAAD visiting-professor fellowship to K.H.
References Bloch, G., Hefetz, A., 1999. Reevaluation of the role of mandibular glands in regulation of reproduction in bumblebee colonies. Journal of Chemical Ecology 25, 881–896. Bloch, G., Hefetz, A., Hartfelder, K., 2000. Ecdysteroid titer, ovary status, and dominance in adult worker and queen bumble bees (Bombus terrestris). Journal of Insect Physiology 46, 1033–1040. Bollenbacher, W.E., 1988. The interendocrine regulation of larval– pupal development in the tobacco hornworm, Manduca sexta—a model. Journal of Insect Physiology 34, 941–947. Bollenbacher, W.E., O’Brien, M.A., Katahira, E.J., Gilbert, L.I., 1983. A kinetic-analysis of the action of the insect prothoracicotropic hormone. Molecular and Cellular Endocrinology 32, 27–46. Cnaani, J., 1997. Larval development and caste differentiation in the bumble bee Bombus terrestris. PhD thesis, Tel Aviv University. Cnaani, J., Borst, D.W., Huang, Z.Y., Robinson, G.E., Hefetz, A., 1997. Caste determination in Bombus terrestris: differences in development and rates of JH biosynthesis between queen and worker larvae. Journal of Insect Physiology 43, 373–381. Cnaani, J., Robinson, G.E., Bloch, G., Borst, D.W., Hefetz, A., 2000. The effect of queen–worker conflict on caste determination in the bumble bee Bombus terrestris. Behavioral Ecology and Sociobiology 47, 346–352. Duchateau, M.J., Velthuis, H.H.W., 1988. Development and reproductive strategies in Bombus terrestris colonies. Behaviour 107, 186–207. Evans, J.D., Wheeler, D.E., 1999. Differential gene expression between developing queens and workers in the honey bee, Apis mellifera. Proceedings of the National Academy of Science USA 96, 575–580. Feldlaufer, M.F., Hartfelder, K., 1997. Relationship of the neutral sterols and ecdysteroids of the parasitic mite, Varroa jacobsoni to those of the honey bee, Apis mellifera. Journal of Insect Physiology 43, 541–545. Feldlaufer, M.F., Herbert, E.W., Svoboda, J.A., Thompson, M.J., Lusby, W.R., 1985. Makisterone A—the major ecdysteroid from the pupa of the honey bee, Apis mellifera. Insect Biochemistry 15, 597–600. Hartfelder, K., Engels, W., 1998. Social insect polymorphism: Hormonal regulation of plasticity in development and reproduction in the honeybee. Current Topics in Developmental Biology 40, 45–77. Hartfelder, K., Rembold, H., 1991. Caste-specific modulation of juvenile hormone. III. Content and ecdysteroid titer in postembryonic development of the stingless bee, Scaptotrigona postica depilis. Journal of Comparative Physiology B—Biochemical, Systemic and Environmental Physiology 160, 617–620. Pereboom, Z., 2000. The composition of larval food and the significance of exocrine secretions in the bumblebee Bombus terrestris. Insectes Sociaux 47, 11–20. Rachinsky, A., Engels, W., 1995. Caste development in honey bees (Apis mellifera): Juvenile hormone turns on ecdysteroids. Naturwissenschaften 82, 378–379. Rachinsky, A., Strambi, C., Strambi, A., Hartfelder, K., 1990. Caste and metamorphosis: Hemolymph titers of juvenile hormone and ecdysteroids in last instar honey bee larvae. General and Comparative Endocrinology 79, 31–38. Rembold, H., 1987. Caste-specific modulation of juvenile hormone titers in Apis mellifera. Insect Biochemistry 17, 1003–1006. Rembold, H., Czoppelt, C., Gru¨ne, M., Lackner, B., Pfeffer, J., Woker, E., 1992. Juvenile hormone titers during honey bee embryogenesis and metamorphosis. In: Mauchamp, B., Couillaud, F., Baehr, J.C. (Eds.), Insect Juvenile Hormone Research. INRA, Paris, pp. 37–43. Ribeiro, M., 1999. Long duration feedings and caste differentiation in Bombus terrestris larvae. Insectes Sociaux 46, 315–322. Ribeiro, M., Velthuis, H.H.W., Duchateaux, M.J., van der Tweel, I.,
K. Hartfelder et al. / Journal of Insect Physiology 46 (2000) 1433–1439
1999. Feeding frequency and caste differentiation in Bombus terrestris larvae. Insectes Sociaux 46, 306–314. Riddiford, L., 1978. Ecdysone-induced changes in cellular commitment of the epidermis of the tobacco hornworm, Manduca sexta, at the initiation of metamorphosis. General and Comparative Endocrinology 34, 438–446. Riddiford, L., 1994. Cellular and molecular actions of juvenile hormone. I. General considerations and premetamorphic actions. Advances in Insect Physiology 24, 213–274. Ro¨seler, P.-F., 1970. Unterschiede in der Kastendetermination zwischen den Hummelarten Bombus hypnorum und Bombus terrestris. Zeitschrift fu¨r Naturwissenschaften B 25, 543–548. Ro¨seler, P.-F., Ro¨seler, I., 1974. Morphologische und physiologische Differenzierung der Kasten bei den Hummelarten Bombus hypnorum (L.) und Bombus terrestris (L.). Zoologische Jahrbu¨cher Abteilung Allgemeine Zoologie und Physiologie der Tiere 78, 175–198. Schmidt Capella, I.C., Hartfelder, K., 1998. Juvenile hormone effect
1439
on DNA synthesis and apoptosis in caste-specific differentiation of the larval honey bee (Apis mellifera L.) ovary. Journal of Insect Physiology 44, 385–391. Strambi, A., Strambi, C., Ro¨seler, P.F., Ro¨seler, I., 1984. Simultaneous determination of juvenile hormone and ecdysteroid titers in the hemolymph of bumblebee prepupae (Bombus hypnorum and B. terrestris). General and Comparative Endocrinology 55, 83–88. Warren, J.T., Gilbert, L.I., 1986. Ecdysone metabolism and distribution during the pupal–adult development of Manduca sexta. Insect Biochemistry 16, 65–82. West-Eberhard, M.J., 1989. Phenotypic plasticity and the origins of diversity. Annual Reviews of Ecology and Systematics 20, 249– 278. Wheeler, D.E., 1986. Developmental and physiological determinants of caste in social Hymenoptera: evolutionary implications. American Naturalist 128, 13–24.