1 Introduction

A normal honey bee colony is usually made up of three types of individuals, those called diploid females, which include queens and worker bees, and those called haploid male drones (Woyke 1971). Queens play key roles in bee colonies, not only because of their ability to lay many eggs but also because they can control the common activities of the bee colony by secreting pheromones, which is crucial for the development of the colony. Although both queens and worker bees develop from fertilized eggs, the destiny of the larvae that selectively develop into queens or worker bees is dependent on the quantity and duration of royal jelly intake (Cristino et al. 2010). If worker larvae are transplanted into queen cells within 3 days after egg hatch and fed royal jelly in the next couple of days, they can develop into queens with mature ovaries. This phenomenon is known as caste determination in social insects (Rangel et al. 2013; He et al.  2017; Yin et al. 2018). It is widely applied in commercial queen breeding practices by artificially transplanting larvae from the worker bee rearing environment into the queen bee rearing environment (Shivani and Lovleen 2021). However, growing evidence shows that the quality of queens raised from transplanted younger larvae is not as good as those that develop from the egg stage (Hatch et al. 1999; He et al. 2017).

Studies have shown that queens raised from transplanted worker larvae are smaller in size, lighter in weight, smaller in seminal vesicles and have fewer ovarioles than queens raised from natural or transplanted eggs (Woyke 1971; He et al. 2017; Wei et al. 2019). Furthermore, the differences became increasingly pronounced as the age of the transplanted larvae increased, and the gene expression pattern also varied with the queen rearing method (He et al. 2017; Yin et al. 2018). In fact, concerns about the long-term consequences of commercial breeding of bee populations are not new. In 1923, Rudolf Steiner predicted that bees would become extinct within 100 years due to the gradual reduction in commercial queen bee breeding (Engelsdorp et al. 2010; Tarpy et al. 2011). It is well known that high-quality virgin queens typically exhibit the following characteristics: a longer (Yi et al. 2021) and a heavier body (Kahya et al. 2008; Tarpy et al. 2011), a larger thorax (Tarpy et al. 2012), and larger ovaries with more ovarioles (Kahya et al. 2008; Niño et al. 2013) than other queen bees. Hence, these traits could be considered for evaluating the quality of queens. Moreover, queens with these high-quality traits usually exhibit abundant expression levels of developmental and reproductive related genes, including Hexamerin 110 (Hex110) (Martins et al. 2010), Vitellogenin (Vg) (Koywiwattrakul and Sittipraneed 2009), and Transferrin (Trf) (Kucharski and Maleszka 2003; Koywiwattrakul and Sittipraneed 2009). Hence, these traits could also be considered for evaluating the quality of queens.

Apis cerana, the Eastern honey bee, is a honey bee species native to South, Southeast, and East Asia. It is widely kept in mountain areas in China as a honey producer. Ascribe to its several outstanding biological merits, such as resisting the mite, Varroa destructor; adapting to extreme weather conditions; and collecting nectar from scattered floral resources (Park et al. 2015), the amount of bee hives has been doubled in recent years with the implementation of poverty eradication and rural revitalization in China, accounting for more than 6 million colonies. It plays important roles in farmers’ income improving ecological construction and environmental protection in rural areas. Colony increasing is mainly depended on the reproductive swarming in most apiary. The practice of queen breeding and rearing is not as popular and skillful as in western bees, Apis mellifera. Meanwhile, few studies have measured queen morphology and the expression of reproduction-related genes in larvae of Eastern honey bees of different ages. Therefore, here, we explored the morphology of queens reared from transplanted worker larvae (Apis cerana) of different ages (60 h E, 1 L, 2 L, and 3 L) and quantified the differences in gene expression among adult queens of different rearing types of Eastern honey bees.

2 Materials and methods

2.1 Honey bee

Honey bee (Apis cerana) samples were collected from the apiary of Cangyuan County, Yunnan Province, China, from April to May 2023. The experiments were repeated five times by using five different colonies, each with four frames, similar brood pattern, and sufficient food storage for brood nursery. Those colonies were rewarded with sugar syrup (approximately 100 mL each) daily to motivate the nurse bee to nurture the grafted young larvae. Additionally, one colony of seven frames with excellent egg-laying status as egg and small larvae supplier and rewarded with sugar syrup (approximately 200 mL each) throughout the experiment.

2.2 Queen rearing

In this study, we compared four groups of queens: queens raised from eggs transplanted from worker cells to queen cells 60 h (E) after egg laying and queens raised from worker larvae transplanted to queen cells 4 (1 L), 5 (2 L), and 6 (3 L) days after egg laying. Queens from E, 1 L, 2 L, and 3 L were obtained from the same mother queen.

Virgin queens were obtained according to the queen rearing strategy summarized in Figure 1. On April 23rd, an empty worker comb was inserted into the colony headed by a naturally mated queen for her to lay eggs for 8 h (marked every 8 h), but normal egg laying began on April 25th. We marked a total of three egg-laying events, with 83 eggs (marked A) laid from 7:00 to 15:00 on April 25th, 121 eggs (marked B) laid from 15:00 to 23:00 on April 25th, and 97 eggs (marked C) laid from 23:00 on April 25th to 7:00 on April 26th. In addition, to ensure consistency in the development time of eggs and larvae, sampling was started 4 h after egg laying. Specifically, 45 larvae (A) were transferred to queen cells at 11:00 on April 31st (3 L), 90 larvae (B) were transferred to queen cells at 19:00 on April 29th (1 L) and 19:00 on April 30th (2 L), and 45 eggs (C) were transferred to queen cells at 15:00 on April 28th (E). After the incubation time, E, 1 L, 2 L, and 3 L were alternately transferred to the same queen cells to ensure uniform distribution. Thus, four types of daughter queen groups were established. The queen cells were placed (middle of the four frames) in racks in five queenless honey bee colonies to be tended by workers, fed royal jelly, and reared as queens.

Figure 1.
figure 1

Flow chart of queen rearing.

Full size image

2.3 Morphological measurements

These queens were placed in a dark incubator (33 °C, 80% relative humidity), and the queen cells were numbered when they had been completely capped. The capped queen cells were checked every 2 h for queen emergence on the 15th day after egg laying and hourly after the first queen emerged. The newly emerged queens were taken for morphological measurements. The birth weight and thorax weight were measured using an analytical balance (BAS124S; Sartorius, Gottingen, Germany). A Vernier caliper was applied to measure the length, caliber, and maximum diameter of queen cells, and the wing length, wing width, head width, thorax length, and thorax width were measured using ImageView in a dissecting microscope (HY-950S; Leica, Wetzlar, Germany) according to the manufacturer’s instructions.

2.4 Ovary and fat body gene expression analysis

Ovary and fat body of four queens from each group were dissected to measure the expression patterns of genes related with queen development. The dissected ovaries and fat bodies were directly stored at − 80 °C for further total RNA extraction.

2.5 RNA extraction and cDNA synthesis

Total RNA was extracted using a TransZol Up Plus RNA kit (TransGen Biotech, Beijing, China) according to the manufacturer’s protocol. The integrity of RNA was assessed by 1.2% denaturing gel electrophoresis, and the concentration of all the samples was checked using NanoDrop One (Thermo Fisher Scientific Inc., USA). RNA with high purity and integrity was used to synthesize cDNA by using a reverse transcription kit (Takara, Tokyo, Japan). RNA was preserved in a − 80 °C freezer, and the reverse transcripts were preserved at − 20 °C.

Primers were designed with Primer 5.0 software and synthesized at GENEWIZ (Suzhou, China) (Table I). The RT-qPCR was accomplished with a PCR instrument (ETC811; Eastwin Scientific Equipment Inc., Suzhou, China) by using an RT-qPCR kit (SYBR® Premix Ex Taq™ II, Takara, Tokyo, Japan). Each reaction had three technical replicates, and β-actin was selected as the internal reference gene.

Table I Sequences of primers used in quantitative PCR
Full size table

2.6 Ovariole number of virgin queens

Five queens from five different colonies of each group were obtained for ovaries dissection and ovariole counting based on the previous published methods (Raulino-Domanski et al. 2019) with some modifications. Briefly, the ovaries were dissected under a dissecting microscope (HY-950S; Leica, Wetzlar, Germany), then preserved in a cocktail containing 95% alcohol, formalin, and water (70:5:25) for 7 days making the ovaries harden and leathery (Carreck et al. 2013). For ovariole counting, the right ovary was separated and transferred to a slide; then, the slide was mounted with cover slip and photographed in a digital camera coupled to a binocular light microscope (MZ16 Leica, Wetzlar, Germany). The images were processed and the ovarioles were counted manually.

2.7 Data analysis

Statistical analyses were performed using SPSS Statistics version 26, and all differences among the four groups were determined by one-way ANOVA; Tukey’s test was used to determine if there were any differences among different groups, where P < 0.05 was considered significantly different.

The gene expression level was calculated using the 2−(△△Ct) method (Ganger et al. 2017) based on the consistency of primer amplification efficiency and the normal distribution of PCR data.

3 Results

3.1 The E-reared queens had the largest head width and birth weight

The head width of queens was significantly greater in E queens than in 1 L, 2 L, and 3 L queens, while in 1 L queens, it was significantly greater than that in 2 L and 3 L queens (Figure 2a, P < 0.05). Moreover, the average birth weight from E-reared queens was 172.79 mg, which was significantly heavier than queens from the 1 L (157.72 mg), 2 L (152.78 mg), and 3 L (151.93 mg) groups, and that of 1 L queens was significantly heavier than those of 2 L and 3 L queens (Figure 2b, P < 0.05).

Figure 2.
figure 2

The head width (a) and birth weight (b) of the four queen groups (E, 1 L, 2 L, and 3 L). Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by performing ANOVA followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05), and the same character indicates no difference (P > 0.05).

Full size image

3.2 The E-reared queens had the largest thorax size and weight

The queens from the E group had remarkably wider and longer thoraxes than queens from the 1 L, 2 L, and 3 L groups (Figure 3a, P < 0.05). Moreover, the thorax weight of queens from the E group was also significantly higher than that of 1 L, 2 L, and 3 L queens and was significantly higher in 1 L queens than in 2 L and 3 L queens (Figure 3b, P < 0.5).

Figure 3.
figure 3

Thorax length, width (a), and weight (b) of newly emerged queens from E, 1 L, 2 L, and 3 L. Each bar shows the mean ± SE of thorax length or width. Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by performing ANOVA followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05), and the same character indicates no difference (P > 0.05).

Full size image

3.3 The E-reared queens had the largest wing length and wing width

The wing lengths of E and 1 L queens were significantly greater than those of 2 L and 3 L queens, while there was no statistical significance in these values between E and 1 L queens. Additionally, the wing width of queens from the E group was the widest of the four groups, and its values were significantly greater than wing width of queens from the 1 L, 2 L, and 3 L groups (Figure 4, P < 0.05).

Figure 4.
figure 4

The wing length and wing width of the four queen groups (E, 1 L, 2 L, and 3 L). Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by performing ANOVA followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05), and the same character indicates no difference (P > 0.05).

Full size image

3.4 The E-reared queens exhibited high expression levels of developmental and reproductive related genes

The relative expression levels of Hex110, Trf1, and Vg in ovaries (Figure 5a, P < 0.05) and abdominal fat bodies (Figure 5b, P < 0.05) showed the similar expression pattern, which were all decreased with the increasing of the brood age used for queen rearing, respectively. Queens raised form E and 1 L were much higher than those from 2 and 3 L, respectively (Figure 5, P < 0.05).

Figure 5.
figure 5

The relative expression level analysis of Hex110, Trf1, and Vg of the four queen groups (E, 1 L, 2 L, and 3 L). Samples were dissected from fat bodies (a) and ovaries (b) of the same queen group. Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by ANOVA performing followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05).

Full size image

3.5 The E-reared queens had the greatest number of ovarian tubes

The number of ovarioles in the right ovary was notably decreased with the increased age of the larva transferred to the queen cell (Figure 6, P < 0.05). The queens from the E group possessed the greatest number of ovarioles.

Figure 6.
figure 6

Number of right ovarioles of reared queens from the four groups (E, 1 L, 2 L, and 3 L). Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by ANOVA performing followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05), and the same character indicates no difference (P > 0.05).

Full size image

3.6 The 3 L-reared queens had the highest acceptance rate of bee broods

Egg and larvae transferred to the queen cells were not always accepted by the bees, and the average acceptance rate was significantly increased with the age of the brood transferred to the queen cells. No marked difference was found between the acceptance of 2 L and 3 L groups, which showed the highest rates, with all acceptance rates being 42.22% (E), 57.78% (1 L), 82.22% (2 L), and 82.22% (3 L), respectively (Figure 7, P < 0.05). However, the acceptance rate did not represent the final number of queens emerged, which was ranged from 17 to 34 and represented from 8.11 to 18.92%.

Figure 7.
figure 7

Acceptance rate of queens from E, 1 L, 2 L, and 3 L. Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by ANOVA performing followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05), and the same character indicates no difference (P > 0.05).

Full size image

3.7 Queen cells of E-reared queens had the greatest length, caliber, and maximum diameter

The length and caliber of queen cells (Figure 8a, P < 0.05) of E-reared queens were the diameter of queen cells (Figure 8b, P < 0.05) of E and 1 L queens which was significantly higher than that of 2 L and 3 L, while there was no statistical significance between these values for E and 1 L queen cells.

Figure 8.
figure 8

Queen cell caliber and length (a) and maximum diameter (b) of the four queen groups (E, 1 L, 2 L, and 3 L). Each bar shows the mean ± SEM. The sample size of each queen group is shown in each bar. Data were analyzed by ANOVA performing followed by Tukey’s test. Different characters on the top of bars represent significant differences (P < 0.05), and the same character indicates no difference (P > 0.05).

Full size image

4 Discussion

Artificial queen rearing is widely used in commercial beekeeping practices. The quality of the queen greatly influences the growth, productivity, and survival of honey bee colonies. A poor-quality queen is always considered the top culprit for colony loss or failure (Nazzi et al. 2012). Evaluating the traits of virgin queens might be used as a step forward to assess the quality of mature queens. In this study, we explored the transplantation of eggs and 1-, 2-, and 3-day-old larvae of Apis cerana into queen cells for queen rearing. Physical traits, including the birth weight, length and width of the thorax and wing, number of ovarioles, and expression pattern of reproduction and development-related genes of virgin queens, were investigated. The relatively outstanding morphological parameters and high expression levels of reproductive genes in abdominal fat cells indicate that virgin queens artificially reared from fertilized eggs may have higher fecundity potential than those reared from honey bee larvae, which coincides with the results obtained from studies on the western bees (Apis mellifera) (Özbakır 2021; Yi et al. 2021; Yu et al. 2023).

Queen body weight could be considered one of the most informative indicators of queen quality (Amiri et al. 2017). It is known that adults are usually selected to be larger (Kingsolver and Pfennig 2004). However, the weight of queen bees could be influenced by genetic and environmental factors, such as the age of the grafted brood, season, and strength of the starter hives (Gencer et al. 2000). In this study, all eggs and larvae were come from the same colony, and the starter hives were managed with similar colony strength, brood pattern, and food storage. Therefore, the influence of genetic and hive environments could be excluded. Thus, the variation in weight could be affected by rearing methods. The birth weight of queens reared from eggs was heavier than that of queens reared from larvae in this investigation, which is in line with a study on Apis mellifera (Yi et al. 2021; Yu et al. 2023). The queen larvae are provided with royal jelly throughout their whole larval stage, while the worker larvae are fed with far less worker jelly only on the first 3 days during the larval period. Furthermore, the transferred worker larvae only obtain royal jelly after larval grafting, while the transferred eggs are exclusively and adequately supplied with royal jelly after egg hatching. The nutritional difference between worker jelly and royal jelly should be the main cause of body weight diversity (Kucharski et al. 2008; Wang et al. 2016). Moreover, queen body weight was found to be significantly correlated with her fitness and colony productivity, not only positively associated with the acceptance rate of newly introduced queens (Moretto et al. 2004) but also positively correlated with reproductive organs, such as ovaries and number of ovarioles, the diameter of the spermatheca, and the number of stored spermatozoa (Tarpy et al. 2011; Collins and Pettis 2012). Therefore, queens raised from transplanted eggs might have better quality than others.

The wing length and width, thorax length and width, and head width of queen bees were significantly affected by the age of the grafted larvae. The thoracic segment, to which the legs and wings are attached, is a tagma specialized for the locomotory function of insect. For virgin queens, the wings and thorax are useful for the mate-selection process, and this process takes place outside the hive (Minarti et al. 2022). Although a larger wing length does not always suggest greater flight ability, wing metrics are usually considered a key element in flying performance evaluation (Boecking and Spivak 1999). It was reported that the thorax width of virgin queens was positively correlated with both stored sperm number and mating frequency. Queens with larger thorax presumably indicate powerful flight muscles that move the wings, which are predisposed to fly greater distances and for longer durations on their mating flights and therefore mate with a greater number of drones. On the other hand, a larger thorax width is a good proxy for the volume of the spermatheca, which would enable queens to store more sperm (Hatch et al. 1999; Tarpy et al. 2011). The larger thorax and wings of E-reared queens might have greater egg-laying potential than those of queens reared from grafted larvae. In addition, the head width of honey bees is correlated with the volume of the brain (Gronenberg and Couvillon 2010) and the mushroom bodies (Mare et al. 2005). Bees with larger heads showed better cognitive performance (Monchanin et al. 2021). A larger head size of queens might be helpful for their localization and hive return after mating flight.

The queen bee ovaries, consisting of a bundle of ovarioles, are the main organs involved in the production of eggs in mated and egg-laying queens. The queens raised from fertilized eggs with more ovarioles presumably have better egg-laying abilities because the number of ovarioles has been reported to be a phenotypic indicator of the reproductive potential of queen honey bees (Chuda-Mickiewicz and Samborski 2015). The ovaries of virgin queens are underdeveloped, but development takes place soon after mating, and physiological changes and distinct expression patterns of reproduction-related genes are also associated with this development (Kahya et al. 2008; Niño et al. 2011). Hex110, Vg, and Trf are three crucial types of proteins in honey bee, especially for the functionality of ovary (Koywiwattrakul and Sittipraneed 2009). Vg is a precursor for the synthesis of yolk proteins and primarily synthesized in the abdominal fat body and acts as an antioxidant to promote longevity in queen bees (Corona et al. 2007). Those proteins can also be transported from the fat bodies or ovarial cells into oocytes through a hormonally and neurally controlled process named vitellogenesis (Kocher et al. 2010; Kodrík et al. 2023). Insect hexamerins (hex 70a, hex 70b, hex 70c, and hex 110), known as storage proteins, are massively synthesized by fat body. The detected transcriptional profile of hex 110 in developing ovaries and testes and the highly transcribed hex 110 in the ovaries of egg-laying queens demonstrate its function in reproduction besides protein storage (Martins et al. 2010; Martins and Bitondi 2016). The upregulation of Hex110, Trf1, and Vg in fat body of ovary-developing worker bees suggests that these genes might be correlated with the development of honey bee reproductive system and reproductive activities (Martins et al. 2010; Yu et al. 2022). Although the higher expression level of Vg in queen ovary might be due to the presence of surrounding fat cells during sample preparation (Corona et al. 2007), the relative higher expression level of Vg, Hex110, and Trf1 in abdominal fat bodies (Fig. 5b) indicates that queens raised from younger brood are highly likely to possess higher fecundity potential. As such, the expression level of these three genes may serve as effective proxies for queen quality estimation.

To obtain a good queen cell acceptance rate is the first and most important step for commercial queen rearing practices. Therefore, the effect of transplanted brood age on the queen cell acceptance rate was investigated in this study. Our finding was similar to that reported by Okuyan and Akyol (2018) for Apis mellifera in which the highest acceptance rate was found for 2 L and 3 L groups, followed by 1 L and eggs (Okuyan and Akyol 2018). Honey bee eggs are very fragile, and transplanting eggs is technically difficult. Therefore, it is difficult to avoid causing unnoticeable damage to eggs during grafting. These defective eggs are inevitably removed when they are returned to the hive after egg transplantation. It can be inferred that the lower queen cell acceptance rate of the transplanted eggs is influenced by the excellent hygienic behavior of Eastern bees, in which worker bees are responsible of detecting and removing the injured, diseased, or dead brood from the colony, thus blocking the further spread of some diseases or eliminating the potential threats within the colony (Boecking and Spivak 1999). Although larval transplanting is much easier and has a higher acceptance rate for queen breeders, epigenomic analyses have revealed that commercial rearing practices may affect queen development via epigenetic processes, and the adult phenotype of queens reared from larvae is partially intercaste and more worker like with lower queen quality than naturally reared queens (He et al. 2017). Therefore, rearing queens from eggs is a good remedy to yield a better outcome for queen performance and colony function.

The size of queen cells has been documented to affect queen-worker differentiation through DNA methylation in honey bees (Apis mellifera) (Shi et al. 2011). It was noteworthy in this study that the queen cells were larger for queens reared from eggs than queens reared from larvae, which is consistent with the previous report in Apis mellifera, which demonstrates that the developmental and nutritional environment of queens reared from transplanted larvae is not the same as queens reared from transplanted eggs (He et al. 2017). Queen cell size affects queens through increased development space and more abundant food; the larger the queen cells are, the more royal jelly is given, and the better the reared queen is (Kucharski et al. 2008). Although it is still unclear and indescribable why and how honey bee workers can construct different-sized queen cells based on the same-sized cell foundation during queen rearing, using larger queen cells has been shown to be a good choice that will increase the quality of reared queens.

In conclusion, we evaluated the influence of queen rearing methods on the queen morphological characteristics and reproductive gene expression. Our results indicate that rearing queens from fertilized eggs is a promising option for beekeepers and queen breeders to yield a better outcome for queen performance and colony function than that of from larvae.