Germ cell cluster formation and cell death are alternatives in caste-specific differentiation of the larval honey bee ovary

Summary

Caste-specific differentiation of the female honey bee gonad takes place in the fifth larval instar. In queen larvae most ovarioles exhibit almost simultaneous formation of numerous germ cell clusters within the first 20 h after the last larval molt. Ultrastructurally distinctive fusomal cytoplasm connects these cystocytes. Germ cell differentiation is accompanied by morphological changes in somatic components of the ovarioles, the follicle and the terminal filament cells. Subsequently, queen ovarioles elongate and differentiate basal stalks that coalesce in a basal calyx. A second round of mitotic activity was found to occur in the late prepupal and early pupal queen ovary. This round may elevate germ cell numbers composing each cluster to levels observed in follicles of adult honey bee queens. In contrast, germ cell cluster formation does not occur in most of the 120–160 ovarioles of the larval worker ovary, but instead many cells in such ovarioles show signs of impending degeneration, such as large autophagic bodies. DNA extracted from worker ovaries did not reveal nucleosomal laddering, and ultrastructurally, chromatin in germ cell nuclei appeared intact. In the 4–7 surviving ovarioles of the small worker ovary, germ cell clusters were found with ultrastructural characteristics identical to those in queen ovarioles. The temporal window during which divergence in developmental pathways of the larval ovaries initiates shortly after the last larval molt coincides with caste-specific differences in juvenile hormone titer which have long been considered critical to caste-specific morphogenesis.     

Regulation of oogenesis in honey bee workers via programed cell death

Reproductive division of labour characterises eusociality. Currently little is known about the mechanisms that underlie the ‘sterility’ of the worker caste, but queen pheromone plays a major role in regulating the reproductive state. Here we investigate oogenesis in the young adult honey bee worker ovary in the presence of queen pheromone and in its absence. When queen pheromone is absent, workers can activate their ovaries and have well-developed follicles. When queen pheromone is present, even though workers have non-activated ovaries, they continually produce oocytes which are aborted at an early stage. Therefore, irrespective of the presence of the queen, the young adult worker ovary contains oocytes. By this means young workers retain reproductive plasticity. The degeneration of the germ cells in the ovarioles of workers in the presence of queen pheromone has the morphological hallmarks of programmed cell death. Therefore the mechanistic basis of ‘worker sterility’ relies in part on the regulation of oogenesis via programmed cell death. Our results suggest that honey bees have co-opted a highly conserved checkpoint at mid-oogenesis to regulate the fertility of the worker caste.     

Juvenile-hormone-dependent interaction of actin and spectrin is crucial for polymorphic differentiation of the larval honey bee ovary

Programmed cell death in the worker ovary of Apis mellifera reduces the number of ovarioles during metamorphosis from 150-200 primordia to less than 10. In contrast, practically all ovarioles in the ovary of queens survive to the adult stage. The correct formation and persistence of polyfusomes has been suggested as a critical factor for ovariole survival. We have analyzed the developmental dynamics of F-actin and alpha-spectrin in fusomes of queen and worker larvae, and in juvenile-hormone-treated worker larvae. Small fusomes containing actin and spectrin can be detected in the ovaries of fourth instar larvae in both castes. After molting to the fifth instar, the actin-spectrin association persists in the enlarged fusomes of queen ovarioles. In workers, actin dissociates from the fusomal and cortical alpha-spectrin. Coinciding with the appearance of apoptosis markers, large agglomerates of actin are detectable in worker ovarioles. Treatment of fourth-instar worker larvae with juvenile hormone rescues ovarioles from apoptosis and maintains the actin-spectrin association. Juvenile-hormone-dependent actin-spectrin interaction is thus one of the earliest steps in the differentiation of a polymorphic ovary. Plasticity in ovariole numbers as a result of hormone-dependent fusome formation may be a more widespread phenomenon in insects, extending beyond caste polymorphism in highly eusocial Hymenoptera.     

Worker laying in leafcutter ant Acromyrmex subterraneus brunneus （Formicidae, Attini）

We studied the process of offspring production in queenless colonies of Acromyrmex subterraneus brunneus, and particularly evaluated the ovary development of workers as a function of their age. For this, subcolonies were set up and evaluated at different periods of isolation from the queen （2, 4 and 6 months）, besides individually labeled age groups. The subcolonies were assessed according to offspring production and ovaries containing oocytes or not. The evaluations showed worker oviposition and development of males originating from worker-laid eggs. At 2 months＇ absence of the queen, eggs and larvae were found, with eggs in a higher proportion than larvae. After 4 months, the proportion of eggs had reduced while larvae had increased, and a pupa was found in one subcolony. At 6 months, besides a higher share of larvae, one pupa and one adult male were found. Dissection of workers revealed ovaries containing oocytes during the periods of evaluation. Only a group of medium-sized and large workers, 23.3%, 20.9% and 37.5% of the population from each period assessed in queenless subcolonies respectively, presented developed oocytes in the ovary. The same was observed in colonies with a queen, with 17.6%, 19.6% and 7.8% of the group of dissected workers from each time period, respectively. With respect to worker age, we observed by dissection of the ovary, that the greatest percentage of individuals with ovarioles containing oocytes occurred at 45 days （6 weeks） up to 90 days （12 weeks）. These results probably are associated with the workers reproduction and the laying of trophic and reproductive eggs in colonies with and without a queen; these eggs have distinct functions in each situation.     

Effect of honey bee queen mating condition on worker ovary activation

The presence of the honey bee queen reduces worker ovary activation. When the queen is healthy and fecund, this is interpreted as an adaptive response as workers can gain fitness from helping the queen raise additional offspring, their sisters. However, when the queen is absent, workers activate their ovaries and lay unfertilized eggs that become males. Queen pheromones are recognised as a factor affecting worker ovary activation. Recent work has shown that queen mandibular pheromone composition changes with queen mating condition and workers show different behavioural responses to pheromone extracts from these queens. Here, we tested whether workers reared in colonies with queens of different mating condition varied in level of ovary activation. We also examined the changes in the chemical composition of the queen mandibular glands to determine if the pheromone blend varied among the queens. We found that the workers activated their ovaries when queens were unmated and had lower ovary activation when raised with mated queens, suggesting that workers detect and respond adaptively to queens of differing mating status. Moreover, variation in queen mandibular gland’s chemical composition correlated with the levels of worker ovary activation. Although correlative, this evidence suggests that queen pheromone may act as a signal of queen mating condition for workers, in response to which they alter their level of ovary activation.     

Sex-specific developmental profiles of juvenile hormone synthesis in honey bee larvae

Summary Juvenile hormone synthesis in drone larvae of the honey bee was measured by an in vitro radiochemical assay. The developmental profile of corpora allata activity in male larvae showed considerable differences from queen larvae, the presumptive reproductive females, and was comparable to workers, the sterile female morph. Drone and worker larvae, however, differed drastically in the regulation of juvenile hormone biosynthesis, as revealed by the addition of farnesoic acid to the culture medium. This precursor stimulated juvenile hormone synthesis of drone glands nearly eightfold, whereas in worker larvae it is known to lead to an accumulation of methyl farnesoate. The sex-specific differences in endocrine activity indicate a role for juvenile hormone in the expression of genetically determined sexually dimorphic characters during metamorphosis, a role not currently accounted for in models describing endocrine regulation of insect development. Correspondence to: K. Hartfelder     

蜜蜂级型分化相关生理因子研究进展

蜜蜂Apis melliferaL.是典型的社会性昆虫,蜂群内蜂王和工蜂的级型分化现象吸引了众多研究者的高度关注。蜜蜂级型分化是一个极为复杂的生长发育调节过程,许多重要生理因子参与其中,包括保幼激素、蜕皮激素、胰岛素/胰岛素类似物信号通路等。本文概述了蜜蜂级型分化相关生理因子研究进展:保幼激素决定着蜜蜂级型发育的轨迹并能阻止卵巢发育过程中细胞的程序性死亡;蜕皮激素协同保幼激素发挥调节作用;蜜蜂蜂王与工蜂的躯体大小及相关器官大小的级型特异性差异与胰岛素/胰岛素类似物信号通路密切相关。此外,本文还介绍了蜜蜂级型分化模型,该模型系统描述了上述各生理因子对级型分化的综合作用。     

Representational Difference Analysis (RDA) reveals differential expression of conserved as well as novel genes during caste-specific development of the honey bee (Apis mellifera L.) ovary

In highly eusocial insects, such as the honey bee, Apis mellifera, the reproductive bias has become embedded in morphological caste differences. These are most expressively denoted in ovary size, with adult queens having large ovaries consisting of 150-200 ovarioles each, while workers typically have only 1-20 ovarioles per ovary. This morphological differentiation is a result of hormonal signals triggered by the diet change in the third larval instar, which eventually generate caste-specific gene expression patterns. To reveal these we produced differential gene expression libraries by Representational Difference Analysis (RDA) for queen and worker ovaries in a developmental stage when cell death is a prominent feature in the ovarioles of workers, whereas all ovarioles are maintained and extend in length in queens. In the queen library, 48% of the gene set represented homologs of known Drosophila genes, whereas in the worker ovary, the largest set (59%) were ESTs evidencing novel genes, not even computationally predicted in the honey bee genome. Differential expression was confirmed by quantitative RT-PCR for a selected gene set, denoting major differences for two queen and two worker library genes. These included two unpredicted genes located in chromosome 11 (Group11.35 and Group11.31, respectively) possibly representing long non-coding RNAs. Being candidates as modulators of ovary development, their expression and functional analysis should be a focal point for future studies.     

Effect of juvenile hormone treatment on caste differentiation in the honeybee, Apis mellifera

Synthetic juvenile hormone (methyl trans,trans,cis-10-epoxy-7-ethyl-3,11-dimethyl-2,6-tridecadienoate, 1 μg/μl acetone per animal) (JH) was topically applied to 2- to 3-day-old worker honeybee larvae in the hive. Eighty per cent of the hormone-treated larvae were removed from their brood cell before pupation. Only 1 out of 42 adults showed characteristics of an intercaste. Fifty per cent of the control larvae (1 μl acetone) developed to adults, all of which were workers.After topical application of JH and feeding on royal jelly under in vitro conditions, the rate of survival is high (up to 85 per cent adults), but up to 67 per cent of queens and 44 per cent of workers exhibit eye malformations with characteristics of somatic mutation. Formation of a more solid web by the spinning larvae, shortening of the diapause by 1 to 2 days, and unusual shapes of mandibles, legs, and abdomen are a consequence of hormone treatment. The effects are less marked after application of 0·1 instead of 1 μg hormone or after its addition to the food (2 μg/g royal jelly). Treatment of the 2- to 3-day-old worker larvae and subsequent rearing on royal jelly is followed by a shift in caste differentiation from queens and workers to intercastes. In no case, are more queens developed after juvenile hormone treatment.Queen bee determinator, partially purified from royal jelly, induces a concentration-dependent shift from workers to queen differentiation. A threefold increase in the natural determinator concentration of royal jelly results in an almost exclusive (98 per cent) queen formation from 2- to 3-day-old worker larvae. In contrast to this direct effect, the influence of JH is explained as an indirect morphogenetic effect not directly coupled with honeybee caste differentiation.     

The critical period for caste determination in Bombus terrestris and its juvenile hormone correlates

The critical period for caste determination and its juvenile hormone (JH III) correlates were studied in Bombus terrestris. Larvae of known age and instar were taken from young colonies, in which they would have been reared as workers, and placed into groups of queenless workers. Under these conditions the critical age for caste determination was 5 days, during the second instar. Endocrine correlates of caste determination were obtained by determining profiles of juvenile hormone titer and juvenile hormone biosynthesis, measured by chiral-specific radioimmunoassay and the in vitro radiochemical assay, respectively. By the middle of the second instar prospective queen larvae had significantly higher rates of juvenile hormone biosynthesis and juvenile hormone titer than prospective worker larvae. Based on the coincidence of timing of both the critical period and the appearance of caste-specific juvenile hormone titer, we suggest that juvenile hormone plays a role in the mechanisms that control caste determination in B. terrestris.     

Anarchistic queen honey bees have normal queen mandibular pheromones

Summary. Anarchistic honey bees are a line developed by recurrent selection in which workers frequently lay eggs. In unselected colonies, workers refrain from reproduction in response to pheromonal signals that indicate the presence of brood and a queen. We show that queen type (anarchistic or wild type) has no effect on rates of ovary activation of anarchistic or wild type workers. In addition, we show that an important component of the queens signalling system, the queen mandibular gland pheromone, is similar in wild type and anarchistic queens. Anarchistic larvae do not inhibit worker ovary development to the same degree as wild type larvae, however all colonies in this experiment contained only wild type larvae. Anarchistic workers had greater rates of ovary activation than wild type workers in colonies headed by either queen type. We therefore conclude that there must be differences in the transmission or reception of queen pheromones, or worker sensitivity to these compounds. These results clearly demonstrate that anarchy is a complex syndrome, not simply the result of reduced pheromone production by anarchist queens and larvae.Received 23 December 2003; revised 14 May 2004; accepted 1 June 2004.     

Behavior and molecular physiology of nurses of worker and queen larvae in honey bees (Apis mellifera)

In a honey bee colony, worker bees rear a new queen by providing her with a larger cell in which to develop and a large amount of richer food (royal jelly). Royal jelly and worker jelly (fed to developing worker larvae) differ in terms of sugar, vitamin, protein and nucleotide composition. Here we examined whether workers attending queen and worker larvae are separate specialized sub-castes of the nurse bees. We collected nurse bees attending queen larvae (AQL) and worker larvae (AWL) and compared gene expression profiles of hypopharyngeal gland tissues, using Solexa/Illumina digital gene expression tag profiling (DGE). Significant differences in gene expression were found that included a disproportionate number of genes involved in glandular secretion and royal jelly synthesis. However behavioral observations showed that these were not two entirely distinct populations. Nurse workers were observed attending both worker larvae and queen larvae, and there was no evidence of a specialized group of workers that preferentially or exclusively attended developing queens. Nevertheless, AQL attended larvae more frequently compared to AWL, suggesting that nurses sampled attending queen larvae may have been the most active nurses. This study serves as another example of the relationship between differences in gene expression and behavioral specialisation in honey bees.     

Larval juvenile hormone treatment affects pre-adult development,but not adult age at onset of foraging in worker honey bees (Apis mellifera)

Previous research has shown that juvenile hormone (JH) titers increase as adult worker honey bees age and treatments with JH, JH analogs and JH mimics induce precocious foraging. Larvae from genotypes exhibiting faster adult behavioral development had significantly higher levels of juvenile hormone during the 2nd and 3rd larval instar. It is known that highly increased JH during this period causes the totipotent female larvae to differentiate into a queen. We treated third instar larvae with JH to test the hypothesis that this time period may be a developmental critical period for organizational effects of JH on brain and behavior also in the worker caste, such that JH treatment at a lower level than required to produce queens will speed adult behavioral development in workers. Larval JH treatment did not influence adult worker behavioral development. However, it made pre-adult development more queen-like in two ways: treated larvae were capped sooner by adult bees, and emerged from pupation earlier. These results suggest that some aspects of honey bee behavioral development may be relatively insensitive to pre-adult perturbation. These results also suggest JH titer may be connected to cues perceived by the adult bees indicating larval readiness for pupation resulting in adult bee cell capping behavior.     

Queen pheromones affecting the production of queen-like secretion in workers

The honeybee queen pheromones promote both worker sterility and worker-like pheromone composition; in their absence workers become fertile and express the queen pheromones. Which of the queen pheromones regulate worker pheromone expression and how, is still elusive. Here we investigated how two queen pheromones, the mandibular and Dufour’s, singly or combined, affect worker ovarian activation and occurrence of queen-like Dufour’s esters. Although queen mandibular pheromone (QMP) alone, or combined with Dufour’s secretion, inhibited to some extent worker reproduction, neither was as effective as the queen. The effect of the queen pheromones on worker pheromone expression was limited to workers with developed ovaries. Here too, QMP and Dufour’s combined had the greatest inhibitory effect. In contrast, treatment with Dufour’s alone resulted in augmentation of esters in the workers. This is another demonstration that a pheromone emitted by one individual affects the rates of its production in another individual. Ester production was tightly coupled to ovarian development. However fertile workers from queenright or QMP-treated colonies had significantly higher amounts of esters in their Dufour’s gland than untreated queenless colonies. The fact that the queen or QMP exert greater suppression on signal production than on ovary activation, suggests disparate regulatory pathways, and presents a challenging ultimate as well as proximate questions.     

Queen-signal modulation of worker pheromonal composition in honeybees

Worker sterility in honeybees is neither absolute nor irreversible. Whether under queen or worker control, it is likely to be mediated by pheromones. Queen-specific pheromones are not exclusive to queens; workers with activated ovaries also produce them. The association between ovarian activation and queen-like pheromone occurrence suggests the latter as providing a reliable signal of reproductive ability. In this study we investigated the effect of queen pheromones on ovary development and occurrence of queen-like esters in workers' Dufour's gland. Workers separated from the queenright compartment by a double mesh behaved like queenless workers, activating their ovaries and expressing a queen-like Dufour's gland secretion, confirming that the pheromones regulating both systems are non-volatile. Workers with developed ovaries produced significantly more secretion than sterile workers, which we attribute primarily to increased ester production. Workers separated from the queenright compartment by a single mesh displayed a delayed ovarian development, which we attribute to interrupted transfer of the non-volatile pheromone between compartments. We suggest that worker expression of queen-like characters reflects a queen-worker arms race; and that Dufour's gland secretion may provide a reliable signal for ovarian activation. The associative nature between ovary development and Dufour's gland ester production remains elusive.     

The occurrence of vitellogenin in workers and queens of Apis mellifica and the possibility of its transmission to the queen

Immuno-diffusion tests show that worker and queen haemolymph contains a protein fraction which does not occur in the haemolymph of drones. Its immunological and electrophoretical properties are identical with those of the main soluble fraction in the ovaries of queens in oviposition. It therefore is a vitellogenin.The titre of this vitellogenin in the haemolymph of 0- to 28-day-old workers was determined by rocket-immunoelectrophoresis. It attains a maximum on day 12. Its changes seem to be positively correlated with the volume of the corpora allata during the first 12 days of adult life.The hypopharyngeal, mandibular, and salivary glands and the content of the honey stomach of workers were immunologically examined. Vitellogenin could not be found in these organs nor in worker or royal jelly. It is also absent from the digestive tract of queens.¹⁴C-labelled amino acids were injected into 5-day-old workers. Later the uptake of radioactive proteins by the queen was examined. Autoradiography of immuno-diffusion plates showed that within 72 hr active material passed from the injected workers into the eggs laid by the queen. The soluble proteins were extracted from the ovaries and the thorax of the queens and their radioactivity determined. The ratio of ovary to thorax radioactivity of queens directly injected was significantly different from that of queens kept with injected workers.Several proteins of the homogenized hypopharyngeal glands of workers showed precipitation reactions with the antiserum against homogenate of queen ovaries. This together with the results of the tracer experiments indicates that the proteins of the worker hypopharyngeal glands may be precursors of queen yolk components.