Expression of AmGR10 of the Gustatory Receptor Family in Honey Bee Is Correlated with Nursing Behavior

We investigated the association between the expression of a gene encoding gustatory receptor (G10) and division of labor in the honey bee, Apis mellifera. Among 10 GR genes encoding proteins 15% ~ 99% amino acid identity in the honey bee, we found that AmGR10 with 99% identity is involved in nursing or brood care. Expression of AmGR10 was restricted to organs of the hypopharyngeal gland, brain, and ovary in the nurse bee phase. Members of an extended nursing caste under natural conditions continued to express this gene. RNAi knockdown of AmGR10 accelerated the transition to foraging. Our findings demonstrate that this one gene has profound effects on the division of labor associated with the development and physiology of honeybee society.


Introduction
Genetic and epigenetic studies can be fruitful in revealing the social structures of animal species. Diverse social organisms such as insects, fishes, voles, and humans have been used to explore the relations between genes, brains, and social behavior [1][2][3]. Conserved genes are crucial to understanding a wide range of characteristics from social cognition to clinical disorders; for example, FoxP1 and FoxP2 expression patterns in human fetal brain are similar to those in the songbird [4,5]. Although some genes influence complex behaviors via pleiotrophic effects, manipulation of a single gene or its orthologs can have remarkable effects on behavior [1]; for example, disruption of a single copy of FoxP2 in mice causes modest development delay and a significant alteration in ultrasonic vocalization [6]. Behavioral transitions in Drosophila melanogaster food searching behavior from rover to sitter [7] and in adult worker honey bees (Apis mellifera) from hive tasks to foraging [8] are caused by expression of the same foraging gene, for.
The honey bee is an important model organism in studies of nursing or brooding behavior. During the first 2-3 weeks of their adult lives, worker bees perform different tasks in the hive, including nursing or caring for larvae and the queen. After that, for remaining 5-7 weeks of their lives, they become foragers, collecting pollen and nectar [9][10][11]. This transition in behavior involves changes in the physiological processes of hormones and neurochemicals and in the expression of thousands of genes [12][13][14].
Although exposure to brood pheromone can delay onset of foraging and regulate the expression of genes in the brain [15], and vitellogenin influences social foraging specialization [16], a single nursing or brood care-related gene which causes an adverse effect to the foraging gene, may also be involved in the division of labor. Therefore, we looked for nursing-related genes that influence nursing behavior, in contrast with foraging behavior.

Bees
European honeybees were maintained in the hives 1 and 2 at Iwate University and in the hive 3 at the Tropical Biosphere Research Center, University of the Ryukyus. Newly emerged workers were marked with a small spot of enamel paint on the thorax for later identification [9]. Marked nurse bees (seen with their heads in larval cells) aged 0 to 14 days after adult emergence were collected in the hive (hive 1). Marked forager bees (carrying pollen or nectar) aged 19 to 29 days were collected on their return to the hive (hive 2) [9,16]. An extended-duration nursing caste in 29-day-old adults within the hive (hive 2) was created under natural conditions during October to early November of 2007 [11]. We marked them and also used as the extended-duration nursing caste.

Total RNA extraction and mRNA purification
To investigate the influence of gene action on behavior, we chose the hypopharyngeal gland (HPG), which initially synthesizes and secretes royal jelly, and later shrinks in foragers and synthesizes digestion enzymes, indicating a transition in an organ-level trait that reflects the division of labor [17]. To examine organ-specific expression, we also analyzed ovaries, brains, midguts, mandibular glands, and salivary glands.
Bees were anesthetized in insect saline (0.75% NaCl) on ice, and the organs were dissected out a binocular microscope. Total RNAs from bulked organs were prepared with Isogen (Nippon Gene) and frozen in liquid nitrogen and stored at -80°C until use. Polyadenylated RNA was separated with the aid of oligotex-dT30-coated magnetic beads (TaKaRa Bio), according to the manufacturer's instructions.
Differential display and subcloning, sequencing, construction of cDNA library, and PCR These materials in experiments 1 and 2 were performed as described in supporting information.
Life-stage-specific RT-PCR of AmGR10 in organs To examine stage-specific expression during from egg to adult forager and the organ-specific distribution of AmGR10 transcripts, we collected randomly 300 eggs within 24 h after oviposition, and 50 larvae and 10 pupae within the hive (hive 2) between June and October of 2007 and 2008. Total RNAs from these samples were frozen in liquid nitrogen and stored at -80°C until use. We subjected all RNA samples to first-strand reverse transcription. The reverse-transcribed cDNA samples were amplified with AmGR10 primers. Expression was normalized to that of the constitutively expressed β-actin gene of A. mellifera, and was confirmed by RT-PCR [18].

Preparation of double-stranded RNA
To synthesize double-stranded RNA (dsRNA) for RNA interference, we used a PCR-template method. Forward and reverse primer sequences to amplify a 513-base-pair region were selected from the AmGR10 nucleotide sequence and from A. mellifera putative GR 10 from the NCBI Honey Bee Genome Resources. A T7-promoter sequence (TAATACGACTCACTCAC TATAGGGAGA) was attached to the 5 0 end of each primer to facilitate in vitro transcription of sense and antisense RNAs simultaneously. The following primers were used: forward, primer-5 0 -TAATACGACTCACTATAGGGAGACCACATAGAACTCTCTAAGGC-3 0 ; reverse, primer-5 0 -TAATACGACTCACTATAGGGAGACCACAGTAAGGATCACCAAG-3 0 . PCR was performed with 100 ng plasmid DNA template, 25 pmol of each T7-linked primer, 8 mM MgCl 2 , 10×PCR buffer, all four deoxynucleotides at 5 mM, and 2.5 U of Taq DNA polymerase in a 50 μL PCR reaction. PCR was performed as for cloning. Amplified products were purified by gel extraction with a Mag Extractor-PCR & Gel Clean up kit (Toyobo) and used as templates for in vitro transcription for dsRNA synthesis in a Megascript RNAi kit (Ambion). Transcription products were treated as instructed by the user manual and re-suspended in nuclease-free water.

RNA interference (RNAi)
Seven-day-old nurse bees were divided into three groups. The first group was injected through the neck membrane with 8.2 μg/2.0 μL AmGR10 dsRNA (ds AmGR10 group, n = 100 total in 4 independent experiments), as described previously [19]. As a handling control, the second group (sham group, n = 80 total in 4 independent experiments) received a 2.0 μL of nucleasefree water [19,20]. The third group were left untreated (n = 80 total in 4 independent experiments). All bees were tagged with paint to identify the treatment. Immediately after the injections, the bees were returned to the same hive (hive 1-2010, hive 2-2011 and 2012, hive 3-2009), which held approximately 5,000-10,000 worker bees of all age classes and an egg-laying queen. Following the transition to foraging (as indicated by the presence of marked bees outside of the hives), they no longer engaged in any within-hive-tasks [9]. The bees were observed daily for a total of 4.5 h (10:00-12:00 and 13:00-15:30) over 3 consecutive days.

Data analysis
We compared performance within an experimental group by Wilcoxon's signed-rank test, and between groups by Mann-Whitney U-test.

Results and Discussion
First we compared the expression of HPG mRNAs in nurse and forager bees. One candidate band sequence that was strongly expressed in nurse bees was used in a BLAST search of Gen-Bank and the NCBI Honey Bee Genome (S1 The insect GR family was identified in the Drosophila melanogaster genome and named for its expression in gustatory organs such as the mouthparts [23]. So far, candidates for 68 GRs encoded by 60 genes have been identified in D. melanogaster [24] and candidates for 76 GRs encoded by 52 genes have been found in Anopheles gambiae [25]. AmGR1 in the honeybee antennae was recently shown to function as a sweet receptor that responds to some sugars, and AmGR2 in the same organ may act as a co-receptor [26], but Robertson and Wanner [21] revealed a total of 10 genes encoding proteins with 15% to 99% amino acid identity to each other and proposed that these 10 GRs are very limited in terms of gustatory mechanism and capacity. Here, we propose that the GR10 protein is primarily involved in nursing or broodcaring behavior in honey bees. We monitored the expression of AmGR10 in the HPGs of 7-day-old nurse bees and 29-dayold foraging bees by using RT-PCR (Fig 1A). AmGR10 was highly expressed in the HPGs of the nurse bees, but not of the foraging bees (Fig 1B). We then compared the expression of AmGR10 in total RNAs prepared from newly laid eggs and in larvae and pupae collected randomly. AmGR10 was not expressed in the eggs or larvae, but it was expressed at low levels in the pupae (Fig 1B). It was expressed in the HPGs of nurse bees from age 1 to 14 days, but not in foraging bees from age 19 to 29 days (Fig 1C). These expression profiles reflect the division of labor and physiological processes between nurse and foraging workers [1,9,10,12].
Next, we created an artificially extended nurse state, in which the HPGs are hypertrophied and remain active [11]. AmGR10 remained expressed even in 29-day-old nurse bees, unlike in foragers (Fig 2A). In nurse bees aged 7 days, we found AmGR10 expression not only in the HPGs, but also in the brain and ovary (Fig 2B). Our results provide new evidence for a nursespecific gene linked to the HPG, brain, and reproductive organs.
If AmGR10 is fundamentally involved in worker behavior, its function should be detectable in the transition to foraging. To investigate this possibility, we generated AmGR10-knockdown bees by injecting ds RNA into 7-day-old nurse bees. These workers had significantly lower levels of AmGR10 mRNA in the HPG than controls (Fig 2C). Monitoring of bee movements showed that this knockdown of AmGR10 activity caused earlier nurse-to-forager transition (Fig 2D, S1 Video). Although the collective activities of foraging workers remain to be demonstrated [12], our data strongly support the notion that the AmGR10 influences nursing behavior.
Complex relations between crucial genes and reversible DNA methylation have challenged our molecular understanding of the division of labor in honey bees [1,3,27], but little is known about how such genes or epigenetic changes can explain the complex pathways that determine the division of labor. Examination of the nine genes encoding GRs in A. mellifera by qRT-PCR has revealed that the expression of seven of them is enriched in gustatory organs such as the labial palps and the glossa, and AmGr7 is expressed at high levels in the heads (although AmGR10 was not investigated) [21]. Our finding of AmGR10 as crucial to nursing or brood-caring behavior in the hive will help to explain how the GR protein family in social insects mediates the synthesis of royal jelly, as well as how it contributes to behavior in the hive. Although the ligand-receptor binding of the AmGR10 product remains to be analyzed, our results suggest that AmGR10 is important in the organization of honeybee societies.
In the invertebrate model system D. melanogaster, recent research has uncovered the fact that GR43a is expressed in a group of neurons in the posterior superior lateral protocerebrum; it is both necessary and sufficient to sense hemolymph fructose, and promote feeding in hungry flies but suppress feeding in satiated flies [28,29]. Trehalose and glucose are the main hemolymph sugars in many insects, but fructose also is a main sugar in honey bee [30,31]. This unique role of Drosophila GR43a in the sensing of fructose in the diet and the hemolymph may provide new insights into the mechanisms of the division of labor in honey bee and the nutrient receptor function of AmGR10 in the HPG, brain, and ovary. Supporting Information S1 Fig. Differential expression patterns of genes by gel electrophoresis. Total RNA samples from 4-to 7-day-old nurse bees and foragers collected randomly outside of the hive were prescreened according to the Seegene user manual, and some differentially expressed genes were found. (A) In experiment 1, bands in columns GP 21, 23, 32, and 34 are differentially expressed genes between nurse bees and foragers (arrows). These bands were reamplified and directly sequenced (see Materials and Methods). Genes 1, 2, and 7 showed no significant similarity, but gene 6 showed high similarity to a royal-jelly-related milk protein (pRJP57-1) of A. mellifera (DOCX) S1 Video. Effects of the injection of RNA interference on foraging onset. Seven-day-old nurse bees were injected through the neck membane with ds RNA (513-bp; ethanol-precipitated, suspended in nuclease-free water at 8.2 μg/2 μL) or with 2 μL of nuclease-free water (handling control), or were not treated (mock). The bees were observed daily for a total of 4.5 h (10:00-12:00 and 13:00-15:30) for 3 consecutive days, and the emergence of newly transitioned forager bees (tagged in white paint on the thorax), was recorded by video. (AVI)