Differential Expression of Major Royal Jelly Proteins in the Hypopharyngeal Glands of the Honeybee Apis mellifera upon Bacterial Ingestion

Simple Summary Transgenerational immune priming (TGIP) to elicit social immunity in the honeybee Apis mellifera has two axes: the first is the ingested pathogen fragments–vitellogenin (Vg)–queen’s ovary axis for the developing embryo, and the second is the ingested pathogen fragments–Vg–nurse’s hypopharyngeal gland axis for the queen and young larvae through royal jelly. However, the dynamics of the expression of the major royal jelly proteins (MRJPs) in the hypopharyngeal glands of A. mellifera nurse bees after bacterial ingestion must be determined to improve our understanding of the second axis of TGIP. In this study, we investigated the expression patterns of MRJPs 1–7 and defensin-1 in the hypopharyngeal glands and Vg in the fat body of nurse bees fed with live or heat-killed Paenibacillus larvae over 12 h or 24 h by using northern blot analysis. We found that the expression of MRJPs and defensin-1 in the hypopharyngeal glands and Vg in the fat body was significantly induced in nurse bees upon bacterial ingestion, indicating that the differential expression patterns of MRJPs, defensin-1, and Vg were dependent on the bacterial status and timing of bacterial ingestion. We also found that antimicrobial peptide (AMP) genes showed induced expression in young larvae upon bacterial ingestion. In summary, our findings indicate that MRJPs in the hypopharyngeal glands are upregulated along with Vg in the fat body of nurse bees upon bacterial ingestion, providing novel insights into the ingested pathogen fragments–Vg–nurse’s hypopharyngeal gland axis for TGIP. Abstract Honeybee vitellogenin (Vg) transports pathogen fragments from the gut to the hypopharyngeal glands and is also used by nurse bees to synthesize royal jelly (RJ), which serves as a vehicle for transferring pathogen fragments to the queen and young larvae. The proteomic profile of RJ from bacterial-challenged and control colonies was compared using mass spectrometry; however, the expression changes of major royal jelly proteins (MRJPs) in hypopharyngeal glands of the honeybee Apis mellifera in response to bacterial ingestion is not well-characterized. In this study, we investigated the expression patterns of Vg in the fat body and MRJPs 1–7 in the hypopharyngeal glands of nurse bees after feeding them live or heat-killed Paenibacillus larvae. The expression levels of MRJPs and defensin-1 in the hypopharyngeal glands were upregulated along with Vg in the fat body of nurse bees fed with live or heat-killed P. larvae over 12 h or 24 h. We observed that the expression patterns of MRJPs and defensin-1 in the hypopharyngeal glands and Vg in the fat body of nurse bees upon bacterial ingestion were differentially expressed depending on the bacterial status and the time since bacterial ingestion. In addition, the AMP genes had increased expression in young larvae fed heat-killed P. larvae. Thus, our findings indicate that bacterial ingestion upregulates the transcriptional expression of MRJPs in the hypopharyngeal glands as well as Vg in the fat body of A. mellifera nurse bees.

Vitellogenin (Vg), an egg-yolk precursor in insects, is expressed in the fat body, secreted into the hemolymph, and then taken up by developing oocytes [21,22]. The fat body is the primary site of synthesis of Vg and antimicrobial peptides (AMPs) [23,24]. In addition to reproduction, Vg plays a role in immunity, life span, and antioxidation [25][26][27][28][29]. In honeybees, Vg is also taken up in the hypopharyngeal glands of nurses and is used as an amino acid donor for the synthesis of RJ [30,31]. In addition to the delivery of pathogen fragments into the queen's ovaries by Vg [27], honeybee Vg transports pathogen fragments from the gut to the hypopharyngeal glands, which then results in the incorporation of pathogen fragments into RJ and the subsequent transfer of pathogen fragments to the queen and young larvae [32,33]. Earlier studies have shown that Vg is a transporter in transferring ingested pathogen fragments to the queen's ovaries and nurse's hypopharyngeal glands for transgenerational immunity in A. mellifera [27,32,33].
Transgenerational immune priming (TGIP) is attracting considerable attention as a means for forming social immunity in honeybees [27,33,34]. In the delivery for TGIP in A. mellifera, Vg plays an important role in the transport of ingested pathogen fragments to the queen's ovaries or the nurse's hypopharyngeal glands [27,32], and RJ acts as a vehicle in transferring pathogen fragments between nestmates [33]. Most recently, a study involving mass spectrometry reported changes of proteomic profile, including MRJPs 1-7 and defensin-1, in RJ from bacterial-challenged and control colonies [33]. However, the gene expression pattern of Vg in the fat body and MRJPs 1-7 in the hypopharyngeal glands of A. mellifera nurse bees upon live or heat-killed bacterial challenge remains unclear. Furthermore, the differential expression of MRJPs 1-7 and Vg in nurse bees fed with live or heat-killed bacteria is yet to be reported, and it is unclear whether the expression pattern of MRJPs in hypopharyngeal glands is similar to that of Vg in the fat body of nurse bees upon bacterial ingestion. Therefore, the gene expression pattern of MRJPs in the hypopharyngeal glands and Vg in the fat body of nurse bees must be determined to understand the response of MRJPs 1-7 in the hypopharyngeal glands of A. mellifera nurse bees to bacterial ingestion.
In this study, we performed studies on the transcriptional expression patterns of MRJPs 1-7 in the hypopharyngeal glands and Vg in the fat body of nurse bees fed with live or heat-killed Paenibacillus larvae. Here, we report the differential expression of MRJPs in the hypopharyngeal glands of A. mellifera nurse bees upon bacterial ingestion.

Honeybees
The honeybees (Apis mellifera) used in this experiment were reared in an apiary at Dong-A University. Newly emerged workers (1-day-old) were marked on the thorax with a spot of paint marker (Monami Co., Ltd., Seoul, Korea) and introduced into a colony. The marked worker bees were collected from the colony when they were six days old.

Feeding Experiment
Six-day-old A. mellifera nurse bees were placed in cages (11.3 × 7 × 4.3 cm) with 40% sucrose solution and bee bread. American foulbrood P. larvae [35] was cultured in Luria-Bertani (LB) medium [36]. The heat treatment of P. larvae was autoclaved twice at 121 • C for 15 min. The resulting non-viable P. larvae cells were confirmed by culture on LB agar plates. The dose of P. larvae used in this study was determined based on previous studies [36,37]. The nurse bees were fed ad libitum a diet without or with live or heat-killed P. larvae (1 × 10 5 cells per nurse bee) over 12 h or 24 h (n = 7 for each treatment). The cages were incubated in an incubator at 34 • C with 80% humidity. After incubation over 12 h or 24 h, A. mellifera nurse bees were dissected on ice under a stereo-microscope (Zeiss, Jena, Germany). Tissue samples (fat body and hypopharyngeal glands) from A. mellifera nurse bees were individually collected and washed twice with phosphate-buffered saline (PBS; 140 mM NaCl, 27 mM KCl, 8 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 , pH 7.4). For the bacterial ingestion to young larvae of A. mellifera, eggs were collected from the same colony. The dietary composition and amount of larval diet were prepared according to the in vitro rearing protocol of honeybee workers [38]. The young larvae (1 day old), which receive RJ from nurse bees during the first day of the larval stage [39], were supplied with a larval diet (royal jelly 44.25%, glucose 5.30%, fructose 5.30%, yeast extract 0.90%, and water 44.25%) [38] without or with heat-killed P. larvae (1 × 10 5 cells per larva) in 96-well plates. After feeding in an incubator at 34 • C with 80% humidity over 24 h or 48 h, the young larvae were individually collected (n = 9 for each treatment).

RNA Extraction
Total RNA was directly extracted from the fat body and hypopharyngeal glands that were freshly dissected from A. mellifera nurse bees by using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), as per the manufacturer's protocol. Total RNA from young larvae of A. mellifera was extracted from the whole body.

Statistical Analysis
To determine the statistical significance at p-values < 0.05, the data were analyzed using one-way ANOVA with post hoc Tukey's test (Statistical software SPSS version 22.0, IBM Inc., Chicago, IL, USA).

Expression Profile of Vg and Abaecin in the Fat Body of Nurse Bees
Gene expression of Vg and abaecin in the fat body of nurse bees fed with live or heatkilled P. larvae over 12 h or 24 h was examined ( Figure 1A). Gene expression levels of Vg increased substantially in the fat body of nurse bees fed with live or heat-killed P. larvae over 12 h or 24 h ( Figure 1B). Ingestion of live P. larvae over 12 h had a higher impact on Vg expression compared to heat-killed P. larvae, whereas bacterial ingestion over 24 h showed a similar increase in Vg expression in the fat body of nurse bees fed with live and heat-killed P. larvae. Abaecin expression was highly upregulated in feeding with live P. larvae over 12 h and showed a significant increase in feeding with heat-killed P. larvae over 24 h ( Figure 1B).
Insects 2022, 13, x FOR PEER REVIEW increased substantially in the fat body of nurse bees fed with live or heat-killed P. over 12 h or 24 h ( Figure 1B). Ingestion of live P. larvae over 12 h had a higher imp Vg expression compared to heat-killed P. larvae, whereas bacterial ingestion ove showed a similar increase in Vg expression in the fat body of nurse bees fed with liv heat-killed P. larvae. Abaecin expression was highly upregulated in feeding with l larvae over 12 h and showed a significant increase in feeding with heat-killed P. larva 24 h ( Figure 1B). . β-Actin was used as an internal control to depict the total RNA loading amount. C, un controls; LP, feeding with live P. larvae; DP, feeding with heat-killed P. larvae. 12 h and 24 h rep feeding time. (B) The relative levels of Vg and abaecin mRNAs represent the average band de of the target genes normalized to the expression levels of the control (12 h). The bars repres mean ± SD from three measurements. A one-way ANOVA test was used to determine the sign difference (p < 0.05) with different lowercase letters (a-d).

Expression Profile of MRJPs and Defensin-1 in Hypopharyngeal Glands of Nurse Bees
The gene expression of MRJPs (MRJP1 through MRJP 7) and defensin-1 in the pharyngeal glands of nurse bees fed with live or heat-killed P. larvae over 12 h or 24 investigated (Figure 2A). We found that the expression patterns of MRJPs and defe in the hypopharyngeal glands of nurse bees were highly upregulated in feeding wi or heat-killed P. larvae over 12 h or 24 h (Figures 2B and S1). MRJPs and defensin-1 popharyngeal glands of nurse bees fed with live P. larvae over 12 h showed a high pression compared to heat-killed P. larvae. However, bacterial ingestion over 24 h de strated more impact on the expression of MRJPs and defensin-1 in feeding with heat-P. larvae compared to live P. larvae ( Figure 2B). In the result of expression profile of M no clear difference in MRJP 7 expression between bacterial ingestion and control w served. Defensin-1 expression pattern was similar to the expression of MRJPs. . β-Actin was used as an internal control to depict the total RNA loading amount. C, untreated controls; LP, feeding with live P. larvae; DP, feeding with heat-killed P. larvae. 12 h and 24 h represent feeding time. (B) The relative levels of Vg and abaecin mRNAs represent the average band densities of the target genes normalized to the expression levels of the control (12 h). The bars represent the mean ± SD from three measurements. A one-way ANOVA test was used to determine the significant difference (p < 0.05) with different lowercase letters (a-d).

Expression Profile of MRJPs and Defensin-1 in Hypopharyngeal Glands of Nurse Bees
The gene expression of MRJPs (MRJP1 through MRJP 7) and defensin-1 in the hypopharyngeal glands of nurse bees fed with live or heat-killed P. larvae over 12 h or 24 h was investigated (Figure 2A). We found that the expression patterns of MRJPs and defensin-1 in the hypopharyngeal glands of nurse bees were highly upregulated in feeding with live or heat-killed P. larvae over 12 h or 24 h ( Figure 2B). MRJPs and defensin-1 in hypopharyngeal glands of nurse bees fed with live P. larvae over 12 h showed a higher expression compared to heat-killed P. larvae. However, bacterial ingestion over 24 h demonstrated more impact on the expression of MRJPs and defensin-1 in feeding with heat-killed P. larvae compared to live P. larvae ( Figure 2B). In the result of expression profile of MRJPs, no clear difference in MRJP 7 expression between bacterial ingestion and control was observed. Defensin-1 expression pattern was similar to the expression of MRJPs. . β-Actin was used as an internal control to depict the total RNA loading amount. C, untreated controls; LP, feeding with live P. larvae; DP, feeding with heat-killed P. larvae. 12 h and 24 h represent feeding time. (B) The relative levels of MRJPs 1-7 and defensin-1 mRNAs represent the average band densities of the target genes normalized to the expression levels of the controls (12 h). The bars represent the mean ± SD from three measurements. A one-way ANOVA test was used to determine the significant difference (p < 0.05) with different lowercase letters (a-d).

Expression Profile of Defensin-1, Hymenoptaecin, and Abaecin in Whole Body of Young Larvae
Gene expression of AMP (defensin-1, hymenoptaecin, and abaecin) in the whole body of young larvae fed with heat-killed P. larvae over 24 h or 48 h was examined ( Figure 3A). The gene expression patterns of defensin-1, hymenoptaecin, and abaecin showed a tendency of significant differences between heat-killed P. larvae ingested and control young larvae: AMP expression showed a significant increase in defensin-1 in feeding over 24 h, hymenoptaecin in feeding over 48 h, and abaecin in feeding over both 24 h and 48 h. . β-Actin was used as an internal control to depict the total RNA loading amount. C, untreated controls; LP, feeding with live P. larvae; DP, feeding with heat-killed P. larvae. 12 h and 24 h represent feeding time. (B) The relative levels of MRJPs 1-7 and defensin-1 mRNAs represent the average band densities of the target genes normalized to the expression levels of the controls (12 h). The bars represent the mean ± SD from three measurements. A one-way ANOVA test was used to determine the significant difference (p < 0.05) with different lowercase letters (a-d).

Expression Profile of Defensin-1, Hymenoptaecin, and Abaecin in Whole Body of Young Larvae
Gene expression of AMP (defensin-1, hymenoptaecin, and abaecin) in the whole body of young larvae fed with heat-killed P. larvae over 24 h or 48 h was examined ( Figure 3A). The gene expression patterns of defensin-1, hymenoptaecin, and abaecin showed a tendency of significant differences between heat-killed P. larvae ingested and control young larvae: AMP expression showed a significant increase in defensin-1 in feeding over 24 h, hymenoptaecin in feeding over 48 h, and abaecin in feeding over both 24 h and 48 h.  Figure 3. Expression profile of defensin-1, hymenoptaecin, and abaecin in the whole body of A. mellifera young larvae fed with heat-killed P. larvae. (A) The expression of defensin-1, hymenoptaecin, and abaecin, as ascertained by northern blot analysis (n = 9). β-Actin was used as an internal control to depict the total RNA loading amount. C, untreated controls; DP, feeding with heat-killed P. larvae. 24 h and 48 h represent feeding time. (B) The relative levels of defensin-1, hymenoptaecin, and abaecin mRNAs represent the average band densities of the target genes normalized to the expression levels of the control (24 h). The bars represent the mean ± SD from three measurements. A one-way ANOVA test was used to determine the significant difference (p < 0.05) with different lowercase letters (a-d).

Discussion
TGIP seems to be an effective way to elicit social immunity in honeybees [27,[32][33][34]. The model representing TGIP in A. mellifera [27,32,33] is known to have two axes: one is the ingested pathogen fragments-Vg-queen's ovary axis to elicit an immune response in the developing embryo, and the other is the ingested pathogen fragments-Vg-nurse's hypopharyngeal gland axis to elicit an immune response in the queen and young larvae through RJ. In these axes, Vg and RJ play a role as transporters [27,32,33]. As Vg, which is transferred through ingested bacterial fragments to the hypopharyngeal glands of nurse bees [32,33], is used as an amino acid donor for synthesis of RJ [30,31] and MRJPs are major protein components of RJ [2,3], we focused on the expression patterns of MRJPs in the hypopharyngeal glands of nurse bees upon bacterial ingestion.
As Vg showed induced expression in the fat body of bees against bacterial challenge [29] and was bound to bacterial pathogens, which are transferred by Vg to the queen's ovaries or nurse's hypopharyngeal glands [27,32], first, we determined the expression pattern of Vg in nurse bees upon bacterial ingestion. We found that Vg showed relatively high induction in the fat body of nurse bees fed with live P. larvae over 12 h, whereas Vg expression in feeding over 24 h increased in a similar manner between live and heat-killed P. larvae. These results revealed that Vg is highly upregulated along with abaecin upon bacterial ingestion, which indicates that Vg is involved in the innate immunity of A. mellifera nurse bees. The expression of defensin-1, hymenoptaecin, and abaecin, as ascertained by northern blot analysis (n = 9). β-Actin was used as an internal control to depict the total RNA loading amount. C, untreated controls; DP, feeding with heat-killed P. larvae. 24 h and 48 h represent feeding time. (B) The relative levels of defensin-1, hymenoptaecin, and abaecin mRNAs represent the average band densities of the target genes normalized to the expression levels of the control (24 h). The bars represent the mean ± SD from three measurements. A one-way ANOVA test was used to determine the significant difference (p < 0.05) with different lowercase letters (a-d).

Discussion
TGIP seems to be an effective way to elicit social immunity in honeybees [27,[32][33][34]. The model representing TGIP in A. mellifera [27,32,33] is known to have two axes: One is the ingested pathogen fragments-Vg-queen's ovary axis to elicit an immune response in the developing embryo, and the other is the ingested pathogen fragments-Vg-nurse's hypopharyngeal gland axis to elicit an immune response in the queen and young larvae through RJ. In these axes, Vg and RJ play a role as transporters [27,32,33]. As Vg, which is transferred through ingested bacterial fragments to the hypopharyngeal glands of nurse bees [32,33], is used as an amino acid donor for synthesis of RJ [30,31] and MRJPs are major protein components of RJ [2,3], we focused on the expression patterns of MRJPs in the hypopharyngeal glands of nurse bees upon bacterial ingestion.
As Vg showed induced expression in the fat body of bees against bacterial challenge [29] and was bound to bacterial pathogens, which are transferred by Vg to the queen's ovaries or nurse's hypopharyngeal glands [27,32], first, we determined the expression pattern of Vg in nurse bees upon bacterial ingestion. We found that Vg showed relatively high induction in the fat body of nurse bees fed with live P. larvae over 12 h, whereas Vg expression in feeding over 24 h increased in a similar manner between live and heat-killed P. larvae. These results revealed that Vg is highly upregulated along with abaecin upon bacterial ingestion, which indicates that Vg is involved in the innate immunity of A. mellifera nurse bees.
A recent study applying mass spectrometry reported that the relative protein abundance in RJ from bacterial-challenged and control colonies exhibited minimal changes, and defensin-1 showed modest increases in RJ from pathogen-diet colonies compared to controls [33]. Another study showed that defensin-1 is not induced in honeybee adults after a colony-level P. larvae infection [40]. In this study, our results point to changes in the gene expression patterns of MRJPs and defensin-1 in the hypopharyngeal glands of nurse bees in response to bacterial ingestion. Notably, the expression profiles of MRJPs 1-6 and defensin-1 in the hypopharyngeal glands of nurse bees were highly upregulated after feeding with live P. larvae over 12 h and feeding with heat-killed P. larvae over 24 h. Furthermore, the expression patterns of MRJPs in the hypopharyngeal glands and Vg in the fat body of nurse bees indicated that Vg and MRJPs were differentially expressed, depending on the exposure to live or heat-killed P. larvae. Thus, our results suggest that the ingestion of live P. larvae rapidly induced the expression of MRJPs, AMPs, and Vg in nurse bees compared to the ingestion of heat-killed P. larvae, whereas the ingestion of heat-killed P. larvae seemed to have a relatively long-lasting effect on the regulation of the expression of MRJPs, AMPs, and Vg. Considering that ingested bacteria are transferred by Vg to the hypopharyngeal glands of nurse bees, and then they are transported into RJ [32,33], it is likely that upon bacterial ingestion, the increased level of Vg expression in the fat body leads to increased MRJPs and defensin-1 in the hypopharyngeal glands of nurse bees. This finding suggests an association between the expression patterns of MRJPs and defensin-1 in the hypopharyngeal glands and Vg in the fat body of nurse bees, depending on the bacterial status and the time since bacterial ingestion. Considering that the gene expression of abaecin, defensin-1, and hymenoptaecin in bumblebees showed high levels beginning at 8 h until 12 h post injection of live bacteria, and then decreased slightly 24 h post injection [41], our results confirmed that the differential expression level of MRJPs and defensin-1 in hypopharyngeal glands of nurse bees fed with live or heat-killed P. larvae over 12 h or 24 h was dependent on the bacterial status. We have recently demonstrated that MRJPs bind to the cell walls of bacteria and exhibit antimicrobial activities at different levels [18]. Thus, our findings suggest that the increased expression levels of MRJPs and defensin-1 in the hypopharyngeal glands of nurse bees upon bacterial ingestion represent an immune response of MRJPs in RJ of A. mellifera. However, in the present study, we were unable to observe a clear difference in the MRJP 7 expression profile between the bacterial ingestion group and the control. Additional studies will be required to determine the role of upregulated MRJPs in the hypopharyngeal glands of nurse bees after bacterial ingestion.
The second axis in TGIP is that the ingested bacteria are transferred by Vg to the hypopharyngeal glands of nurse bees and transported into RJ, which then elicits an immune response in the queen and young larvae through RJ [32,33]. We examined the gene expression patterns of AMPs in A. mellifera young larvae fed with heat-killed P. larvae and found that AMPs are differentially induced in young larvae upon bacterial ingestion: defensin-1 in feeding over 24 h, hymenoptaecin in feeding over 48 h, and abaecin in feeding over 24 h and 48 h. The results of this study indicate that although AMPs, especially in defensin-1, are induced at a modest level, AMPs show differential expression in young larvae of A. mellifera in response to bacterial ingestion. Moreover, the expression profile of AMPs in young larvae indicates that bacterial ingestion during the first day of the larval stage has immune priming effects on the second (24 h) and third (48 h) days of A. mellifera young larval stage.

Conclusions
Our findings provide the first evidence that MRJPs in the hypopharyngeal glands of nurse bees fed with live or heat-killed P. larvae were differentially expressed depending on bacterial status and time of bacterial ingestion. The present study indicates that the expression of MRJPs and defensin-1 in hypopharyngeal glands shows a pattern similar to the Vg expression in the fat body of nurse bees upon bacterial ingestion. These findings show the differential expression of MRJPs in the hypopharyngeal glands of nurse bees in response to bacterial ingestion, providing novel insights into MRJPs to better understand the ingested pathogen fragments-Vg-nurse's hypopharyngeal gland axis for TGIP.