Refine
Has Fulltext
- yes (18) (remove)
Is part of the Bibliography
- yes (18)
Year of publication
Document Type
- Journal article (11)
- Doctoral Thesis (7)
Keywords
- honeybee (18) (remove)
Institute
- Theodor-Boveri-Institut für Biowissenschaften (18) (remove)
Background
The honeybee (Apis mellifera) represents a model organism for social insects displaying behavioral plasticity. This is reflected by an age-dependent task allocation. The most protruding tasks are performed by young nurse bees and older forager bees that take care of the brood inside the hive and collect food from outside the hive, respectively. The molecular mechanism leading to the transition from nurse bees to foragers is currently under intense research. Circular RNAs, however, were not considered in this context so far. As of today, this group of non-coding RNAs was only known to exist in two other insects, Drosophila melanogaster and Bombyx mori. Here we complement the state of circular RNA research with the first characterization in a social insect.
Results
We identified numerous circular RNAs in the brain of A. mellifera nurse bees and forager bees using RNA-Seq with exonuclease enrichment. Presence and circularity were verified for the most abundant representatives. Back-splicing in honeybee occurs further towards the end of transcripts and in transcripts with a high number of exons. The occurrence of circularized exons is correlated with length and CpG-content of their flanking introns. The latter coincides with increased DNA-methylation in the respective loci. For two prominent circular RNAs the abundance in worker bee brains was quantified in TaqMan assays. In line with previous findings of circular RNAs in Drosophila, circAmrsmep2 accumulates with increasing age of the insect. In contrast, the levels of circAmrad appear age-independent and correlate with the bee's task. Its parental gene is related to amnesia-resistant memory.
Conclusions
We provide the first characterization of circRNAs in a social insect. Many of the RNAs identified here show homologies to circular RNAs found in Drosophila and Bombyx, indicating that circular RNAs are a common feature among insects. We find that exon circularization is correlated to DNA-methylation at the flanking introns. The levels of circAmrad suggest a task-dependent abundance that is decoupled from age. Moreover, a GO term analysis shows an enrichment of task-related functions. We conclude that circular RNAs could be relevant for task allocation in honeybee and should be investigated further in this context.
In the eusocial insect honeybee (Apis mellifera), many sterile worker bees live together with a reproductive queen in a colony. All tasks of the colony are performed by the workers, undergoing age-dependent division of labor. Beginning as hive bees, they take on tasks inside the hive such as cleaning or the producing of larval food, later developing into foragers. With that, the perception of sweetness plays a crucial role for all honeybees whether they are sitting on the honey stores in the hive or foraging for food. Their ability to sense sweetness is undoubtedly necessary to develop and evaluate food sources. Many of the behavioral decisions in honeybees are based on sugar perception, either on an individual level for ingestion, or for social behavior such as the impulse to collect or process nectar. In this context, honeybees show a complex spectrum of abilities to perceive sweetness on many levels. They are able to perceive at least seven types of sugars and decide to collect them for the colony. Further, they seem to distinguish between these sugars or at least show clear preferences when collecting them. Additionally, the perception of sugar is not rigid in honeybees. For instance, their responsiveness towards sugar changes during the transition from in-hive bees (e.g. nurses) to foraging and is linked to the division of labor. Other direct or immediate factors changing responsiveness to sugars are stress, starvation or underlying factors, such as genotype.
Interestingly, the complexity in their sugar perception is in stark contrast to the fact that honeybees seem to have only three predicted sugar receptors.
In this work, we were able to characterize the three known sugar receptors (AmGr1, AmGr2 and AmGr3) of the honeybee fully and comprehensively in oocytes (Manuscript II, Chapter 3 and Manuscript III, Chapter 4). We could show that AmGr1 is a broad sugar receptor reacting to sucrose, glucose, maltose, melezitose and trehalose (which is the honeybees’ main blood sugar), but not fructose. AmGr2 acts as its co-receptor altering AmGr1’s specificity, AmGr3 is a specific fructose receptor and we proved the heterodimerization of all receptors. With my studies, I was able to reproduce and compare the ligand specificity of the sugar receptors in vivo by generating receptor mutants with CRISPR/Cas9. With this thesis, I was able to define AmGr1 and AmGr3 as the honeybees’ basis receptors already capable to detect all sugars of its known taste spectrum.
In the expression analysis of my doctoral thesis (Manuscript I, Chapter 2) I demonstrated that both basis receptors are expressed in the antennae and the brain of nurse bees and foragers. This thesis assumes that AmGr3 (like the Drosophila homologue) functions as a sensor for fructose, which might be the satiety signal, while AmGr1 can sense trehalose as the main blood sugar in the brain. Both receptors show a reduced expression in the brain of foragers when compared with nurse bees. These results may reflect the higher concentrated diet of nurse bees in the hive. The higher number of receptors in the brain may allow nurse bees to perceive hunger earlier and to consume the food their sitting on. Forager bees have to be more persistent to hunger, when they are foraging, and food is not so accessible. The findings of reduced expression of the fructose receptor AmGr3 in the antennae of nurse bees are congruent with my other result that nurse bees are also less responsive to fructose at the antennae when compared to foragers (Manuscript I, Chapter 2). This is possible, since nurse bees sit more likely on ripe honey which contains not only higher levels of sugars but also monosaccharides (such as fructose), while foragers have to evaluate less-concentrated nectar.
My investigations of the expression of AmGr1 in the antennae of honeybees found no differences between nurse bees and foragers, although foragers are more responsive to the respective sugar sucrose (Manuscript I, Chapter 2). Considering my finding that AmGr2 is the co-receptor of AmGr1, it can be assumed that AmGr1 and the mediated sucrose taste might not be directly controlled by its expression, but indirectly by its co-receptor. My thesis therefore clearly shows that sugar perception is associated with division of labor in honeybees and appears to be directly or indirectly regulated via expression.
The comparison with a characterization study using other bee breeds and thus an alternative protein sequence of AmGr1 shows that co-expression of different AmGr1 versions with AmGr2 alters the sugar response differently. Therefore, this thesis provides first important indications that alternative splicing could also represent an important regulatory mechanism for sugar perception in honeybees.
Further, I found out that the bitter compound quinine lowers the reward quality in learning experiments for honeybees (Manuscript IV, Chapter 5). So far, no bitter receptor has been found in the genome of honeybees and this thesis strongly assumes that bitter substances such as quinine inhibit sugar receptors in honeybees. With this finding, my work includes other molecules as possible regulatory mechanism in the honeybee sugar perception as well. We showed that the inhibitory effect is lower for fructose compared to sucrose. Considering that sugar signals might be processed as differently attractive in honeybees, this thesis concludes that the sugar receptor inhibition via quinine in honeybees might depend on the receptor (or its co-receptor), is concentration-dependent and based on the salience or attractiveness and concentration of the sugar present.
With my thesis, I was able to expand the knowledge on honeybee’s sugar perception and formulate a complex, comprehensive overview. Thereby, I demonstrated the multidimensional mechanism that regulates the sugar receptors and thus the sugar perception of honeybees. With this work, I defined AmGr1 and AmGr3 as the basis of sugar perception and enlarged these components to the co-receptor AmGr2 and the possible splice variants of AmGr1. I further demonstrated how those sugar receptor components function, interact and that they are clearly involved in the division of labor in honeybees. In summary, my thesis describes the mechanisms that enable honeybees to perceive sugar in a complex way, even though they inhere a limited number of sugar receptors. My data strongly suggest that honeybees overall might not only differentiate sugars and their diet by their general sweetness (as expected with only one main sugar receptor). The found sugar receptor mechanisms and their interplay further suggest that honeybees might be able to discriminate directly between monosaccharides and disaccharides or sugar molecules and with that their diet (honey and nectar).
In dieser Arbeit untersuche ich das Verhalten von Arbeiterbienen beim Brutwärmen, die Wärmeübertragung von den Bienen auf die gedeckelte Brut, die thermophysikalischen Eigenschaften des Brutnests und spezielle Aspekte des Brutnestaufbaus, die für dieses Thema relevant sind und bisher nicht untersucht wurden. Meine Arbeit umfasst Verhaltensbeobachtungen und thermografische Messungen an individuellen Bienen, die Simulation des Heizverhaltens von Arbeiterinnen und das Messen der Temperaturänderungen in der Wabe, die Messung der thermophysikalischen Eigenschaften der Brutwabe und der Zellwände (Wärmeleitfähigkeit und Durchlässigkeit für Wärmestrahlung), die Auswertung von Brutzelltemperaturen als Ergebnis des Verhaltens von Arbeiterbienen, die Analyse der Anzahl und der räumlichen Verteilung von Brutlücken (Auswertung in 2-D und 3-D bezüglich beider Wabenseiten) und die Entwicklung spezifischer Computersoftware, die zur Erarbeitung dieser Ergebnisse unverzichtbar ist. Ein wichtiges Ergebnis dieser Arbeit ist die Entdeckung und Beschreibung eines bemerkenswerten, bislang unbekannten Verhaltens der Honigbiene: Die Aufrechterhaltung hoher Thoraxtemperaturen (TTh) bei Langzeitbesuchen in offenen Zellen („Lücken“) die verstreut in der gedeckelten Brutfläche vorkommen. Hier zeige ich, dass die Aufrechterhaltung der hohen TTh nicht auf den Zellinhalt (z. B. offene Brut) bezogen ist - in den meisten Fällen waren die besuchten Zellen ohnehin leer - sondern auf die direkt benachbarte gedeckelte Brut, mit der diese Zellen über gemeinsame Zellwände in Kontakt stehen. Dieses Verhalten liefert eine Erklärung für Langzeitzellbesuche von sehr langer Dauer ohne erkennbare Aktivität, die in früheren Arbeiten beschrieben aber nicht völlig verstanden wurden, und es rehabilitiert die scheinbar „faulen“ Bienen im Zellinnern. Diesem Verhalten kommt eine große Bedeutung für das Brutwärmen zu, da sich der aufgeheizte Thorax tief in der Wabe (fast an der Mittelwand) befindet wo der Wärmeverlust an die Luft minimiert ist und von wo bis zu 6 umliegende Puppenzellen gleichzeitig gewärmt werden können. Im Vergleich zum Brutwärmeverhalten an der Wabenoberfläche (Andrücken des Thorax an die Brutdeckel), wo nur 1 oder Teile von 3 Brutdeckeln mit dem Thorax in Berührung stehen, ist das Wärmen im Zellinnern mit derselben TTh bis zu 2,6-fach effizienter. Die Messung der thermophysikalischen Eigenschaften der Brutwabe und die Simulation des Brutwärmeverhaltens unter kontrollierten Bedingungen zeigen, dass sich die Wabe langsam aufwärmt und eher ein lokal begrenztes Wärmen als eine rasche Wärmeausbreitung über eine große Fläche begünstigt. Der Einflussbereich eines einzelnen Zellbesuchers hängt von seiner TTh und der Dauer des Zellbesuchs ab. Anstiege der Bruttemperatur in bis zu 3 Zellen Abstand zum Zellbesucher sind nachweisbar. Das hier beschriebene Brutwärmeverhalten im Innern von Lücken (offenen Zellen) bietet nicht nur neue Einsichten in das Bienenverhalten. Es ermöglicht auch eine Neubewertung der Lücken und ihrer Nützlichkeit für die Bienen. Eine von mir entwickelte Computersoftware („CombUse 2.0“) ermöglicht es, das Vorkommen und die räumliche Verteilung von Lücken mit hoher Genauigkeit auf der Ebene einzelner Zellen zu erfassen und auszuwerten. Die räumliche Verteilung der Lücken in der gedeckelten Brutfläche zeigt, dass schon bei geringen Lückenhäufigkeiten von ca. 4 bis 10 %, die in gesunden Kolonien normal sind, eine überraschend große Zahl gedeckelter Brutzellen (88 % bis 99 %, wenn die dreidimensionale Verteilung berücksichtigt wird) im Einflussbereich von Brut wärmenden Zellbesuchern sind. Obwohl das Brutwärmeverhalten im Zellinnern schwer zu entdecken und zu beobachten ist, führen die in dieser Arbeit präsentierten Daten zu dem Schluss, dass es sich dabei um einen wichtigen Bestandteil der Nestklimatisierung bei Honigbienen handelt.