The Juvenile-Hormone-Responsive Factor AmKr-h1 Regulates Caste Differentiation in Honey Bees

Honey bees are typical model organisms for the study of caste differentiation, and the juvenile hormone (JH) is a crucial link in the regulatory network of caste differentiation in honey bees. To investigate the mechanism of JH-mediated caste differentiation, we analyzed the effect of the JH response gene AmKr-h1 on this process. We observed that AmKr-h1 expression levels were significantly higher in queen larvae than in worker larvae at the 48 h, 84 h, and 120 h larval stages, and were regulated by JH. Inhibiting AmKr-h1 expression in honey bee larvae using RNAi could lead to the development of larvae toward workers. We also analyzed the transcriptome changes in honey bee larvae after AmKr-h1 RNAi and identified 191 differentially expressed genes (DEGs) and 682 differentially expressed alternative splicing events (DEASEs); of these, many were related to honey bee caste differentiation. Our results indicate that AmKr-h1 regulates caste differentiation in honey bees by acting as a JH-responsive gene.


Introduction
In honey bee colonies, the queen and worker are diploids, developing from fertilized eggs and, although they share the same genetic material, they differ significantly in morphological characteristics, behavior, and longevity [1].Currently, nutrition [2], epigenetic modifications [3], and hormones [4] are considered to be important factors in regulating the differentiation of queen-worker castes.Of them, juvenile hormone (JH) is known to play a key role in this process [5].
JH is a pleiotropic hormone secreted by the insect corpora allata (CA) and released into the hemolymph [6].It interacts antagonistically with 20-hydroxyecdysone (20E) to maintain larval traits in insects, where 20E induces molting and metamorphosis, and JH prevents the metamorphosis caused by 20E.During the last larva instar, the JH titer drops sharply, resulting in the pupation of completely metamorphosed insects and the emergence of incompletely metamorphosed insect adults [7][8][9][10].In the honey bee, JH is a central regulator of the non-genetic diversity of social castes.Its physiological role in the honey bee is not only limited to maintaining the juvenile state but also has a significant ability to regulate growth and development, caste differentiation, and affects longevity [11,12].Queen larvae have larger corpora allata and significantly higher JH titres in their hemolymph compared to worker larvae, especially at the critical point of caste differentiation, where the JH titre threshold determines the developmental trajectory of the larvae (development into a queen or worker).This suggests that caste differentiation is accompanied by strict regulation of JH synthesis and metabolism [5].In addition, JH also regulates the synthesis of vitellogenin (Vg), which promotes ovarian development in honey bees, resulting in higher levels of Vg and more ovarian tubes in queens [13,14].
Biomolecules 2023, 13, 1657 2 of 13 Krüppel homolog 1 (Kr-h1), a C2H2 zinc finger (Znf) transcription factor, is an early responder gene for JH that can be activated by the JH receptors Met and Gce in Drosophila [9].It plays a crucial role in the JH signaling pathway by regulating the expression of its downstream target genes.In many insects, the JH response elements (JHREs) in the Kr-h1 promoter have been identified as a canonical E-box (CACGTG) or a C-box (CACGCG) [15][16][17], which are the typical binding elements for bHLH-PAS proteins and are also essential for JH-induced gene expression via Met [18,19].The Kr-h1 is highly expressed in the larval stage in insects and decreases in the pupal stage, which is consistent with changes in JH titres in haemolymph [20][21][22].The main function of Kr-h1 is to repress the metamorphosis of insects by transmitting JH signaling.It also regulates insects' yolk production and egg maturation [23][24][25], but the regulatory effect of Kr-h1 on yolk production varies across species.For example, down-regulation of Kr-h1 in Locusta migratoria reduced the mRNA expression level of Vg by an average of 95%, which significantly inhibited oocyte maturation and ovarian development [26].However, in Tribolium castaneum the low expression of Kr-h1 could only down-regulate Vg by about 30% [27].In addition, Kr-h1 is also involved in the development of the nervous system of Drosophila larvae [28] and foraging behavior of honey bees [29].
In the JH signaling pathway, Kr-h1 is downstream of the JH receptors Met and Gce and can be up-regulated by JH [30,31].Due to the antagonistic effect of JH and 20E, AmKr-h1 can inhibit 20E-induced metamorphosis by transducing JH signaling to suppress the expression of the major 20E response genes during the larval stage when JH titres are high [7,32].At the terminal instar of insect larvae, the JH titre decreases dramatically and low-expressing AmKr-h1 is unable to override high 20E signaling, thus, 20E-induced metamorphosis occurs [9].
Since AmKr-h1 is regulated by JH, we speculate that AmKr-h1 may be the key regulator in the queen-worker caste differentiation.Therefore, in the present study, we analyzed the function of AmKr-h1 in the caste differentiation in honey bees using RNAi, and analyzed transcriptome changes in the bee larvae after AmKr-h1 RNAi.

Insects
The honey bee colonies (Apis mellifera ligustica) used in this study were bred at the Honeybee Research Institute of Jiangxi Agricultural University (28.46 N, 115.49E), Nanchang, Jiangxi, China.

Sequence Analysis of AmKr-h1 Gene
The exon-intron structure was analyzed by mapping the mRNA sequence (GenBank accession no.: NM_001242470.1,NM_001011566.1) of the AmKr-h1 to the A. mellifera genome at the NCBI website from the A. mellifera AmKr-h1 gene deposited in the GenBank database (https://www.ncbi.nlm.nih.gov/genbank/,accessed on 20 March 2022).The structural domains of the AmKr-h1 protein were analyzed at the SMART online website (http:// smart.embl-heidelberg.de/,accessed on 20 March 2022).Multiple sequence alignments of amino acid sequences downloaded from NCBI were performed using the CLUSTALX v1.81 software [33].

Expression Analysis Using qRT-PCR
A queen was restricted on an empty comb to lay eggs for six hours.After the eggs hatched into larvae, some of the larvae were transferred to queen cells, while the remaining larvae continued to be bred in worker cells.Six biological replicates of both queen larvae and worker larvae were sampled at 48 h, 84 h, and 120 h after hatching of the eggs for qRT-PCR analysis.During sampling, the larvae were washed three times with double-distilled water to remove any royal jelly, and then were stored in liquid nitrogen for further use.
Total RNA was extracted from the collected samples using the TrizolUp Kit (Transgen, ER501-01, Beijing, China), and the RNA concentration of each sample was measured using a spectrophotometer.A total of 1 µg of RNA was taken from each sample and reverse transcribed into cDNA using M-MLV reverse transcriptase (TakaRa, RR047A, Dalian, China).The Gapdh (GenBank accession no.: NM_001014994.1)gene was used as the internal reference.
The qRT-PCR primer sequences of each gene were designed using the Primer Premier 5.0 software based on their mRNA sequences (Table S1).The qRT-PCR reaction system was as follows: 5.0 µL of SYBR GREEN, 0.2 µL of ROX correction fluid, 0.4 µL of forward primer (0.0125 nmol/µL), 0.4 µL of reverse primer (0.0125 nmol/µL), 1.0 µL of cDNA (150 µg/µL), and 3.0 µL of ddH2O.The qRT-PCR amplification conditions were as follows: 95 • C for 10 min; followed by 40 cycles of 95 • C for 15 s, Tm for 1 min.Four technical replicates were set up for each biological replicate.The data were analyzed using the 2 −∆∆CT method [34] and t-tests were performed using the SPSS 26.0 software to analyze the differences in gene expression.

JH Treatment
The queen of an A. mellifera colony was controlled on an empty comb to lay eggs for 6 h.When the eggs hatched, the larvae were transferred to a 24-well culture plate containing 300 µL of artificially prepared food (6% fructose, 6% glucose, 1% yeast, 37% distilled water, and 50% royal jelly) per well.The larvae were randomly divided into three groups: to the first group (JH), 1 µL of juvenile hormone III (JH III) (APExBIO, purity ≥ 65.00%) solution diluted with ethanol (10 µg/µL) was dripped on the back; the second group (ET) received 1 µL of ethanol on the back as a control; the third group (NO) was normal larvae that received no treatment.Then, the larvae were moved to an incubator at 34 • C and 85% humidity.For the next 2 days, the JH group and ethanol group continued to be treated with JH III or ethanol three times per day.Meanwhile, each group received 150 µL of artificial foods twice daily.
Twenty-four hours after treatment completion, the larvae in each group were collected for mRNA expression analysis using qRT-PCR.For each treatment group, seven biological replicates with two larvae per replicate were sampled.The samples were washed three times with double-distilled water and immediately snap-frozen in liquid nitrogen, and then stored in liquid nitrogen for RNA extraction.The qRT-PCR protocol detailed in Section 2.3 was followed.The differences in the expression of AmKr-h1 between the JH and two control groups were analyzed by ANOVA using SPSS 26.0.
Once again, a queen of an A. mellifera colony was limited to an empty comb to lay eggs for 6 h.After hatching of the eggs, the one-day-old larvae were transferred to a 24-well culture plate containing 300 µL of artificially prepared food in each well.The plates were kept in an incubator at 34 • C and 85% humidity.Then, the larvae were fed twice per day with 150 µL of artificial food.When they reached 60 h, the larvae were randomly divided into two groups and injected with the AmKr-h1 siRNA solution (solubilizing siRNA with RNase-free water; the concentration was 2 µg/µL) and the NT siRNA solution (solubilizing Con with RNase-free water; the concentration was 2 µg/µL), respectively.The NT group was set as the control group.Each larva was injected with 400 ng of siRNA and then provided with 300 µL of artificial food and kept in an incubator (34 • C, 85% humidity).When the larvae were 3 days old, they were transferred to a new plate and fed twice per day with 200 µL of artificial food until pupation.
At 48 h after manual injection of siRNA, ten surviving larvae were selected and set as five biological replicates (each containing two larvae) for detecting the efficiency of RNAi by qRT-PCR.Total RNA extraction, cDNA synthesis, and qRT-PCR were performed as described above.The remaining larvae were kept in the incubator until emergence.The morphological traits of 35 newly emerged bees from the Con group and 35 bees from the AmKr-h1 RNAi group were measured as follows.Firstly, bees were frozen to death on ice.Then, the body weight of each bee in each group was weighed with an electronic balance, and the length of the body, proboscis, and forewing of each bee was measured by a stereomicroscope (Guiguang, GL-99TI, Guilin, China) with the tiny creature morphological measurement and data analysis system of the Optec OPTPro software (cnoptec, version: x86, Chongqing, China).All data were analyzed by t-test in SPSS 26.0.

cDNA Library Construction and Sequencing
Total RNA was extracted from both the RNAi group and Con group samples (each sample containing two larvae).After measuring the quality and concentration of the RNA samples, the mRNA molecules were enriched using Oligo (dT) magnetic beads, and then fragmentation buffer was added to break them randomly into small fragments.Using these small fragments as templates, double-stranded cDNA was synthesized by reverse transcriptase.Then, cDNA fragments were purified, end-repaired, and poly(A) tails and Illumina sequencing adaptors were added.The cDNA fragments were isolated using AMPure XPbeads and were PCR-amplified, the PCR products were purified again to build a cDNA sequencing library.Finally, the cDNA libraries were sequenced using the Illumina high-throughput sequencing platform.
The raw sequences (raw reads) obtained from sequencing were filtered with the fastp software by removing adaptors, N (no base information could be determined) ratio > 10%, low-quality (the number of bases with quality value Q ≤ 10 accounted for more than 50% of the whole read), and trimmed sequences < 500 bp in length, to obtain clean sequences (clean reads).Clean reads were compared with the A. mellifera reference genome sequences (https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/003/254/395/GCF_003254395.2_Amel_HAv3.1/,accessed on 19 January 2022).

Screening of DEGs and DEASEs
The expression level of each gene was calculated by counting the number of reads matched to each gene with the Cufflinks v2.2.1 software.Differentially expressed genes were screened by DESeq using |log2(fold change)| > 2 and p-value < 0.05 as the criteria.Then, a GO and KEGG enrichment analysis was performed with the DEGs.Differentially expressed alternative splicing events were analyzed using the rMATS v4.1.2software with FDR < 0.05 as a screening criterion.The reliability of the RNA-seq results of nine DEGs was verified using qRT-PCR, as mentioned above.

AmKr-h1 Encodes a Zinc Finger Protein
The mRNA sequence of A. mellifera AmKr-h1 is 1756 bp long, containing three exons and encoding a protein of 500 aa, containing eight C2H2 zinc finger (ZnF_C2H2) domains (Figure 1A).The AmKr-h1 protein is highly conserved in the ZnF_C2H2 domain region compared with those of Aedes aegypti, T. castaneum, Drosophila melanogaster, and Bombyx mori (Figure 1B).

AmKr-h1 Has Higher Expression in Queen Larvae and Is Regulated by JH
The relative mRNA expression levels of AmKr-h1 between queen larvae and worker larvae at 48 h, 84 h, and 120 h after the hatching of eggs were measured using quantitative real-time polymerase chain reaction (qRT-PCR).The results showed that the expression levels of the AmKr-h1 gene in the queen larvae were significantly higher than in the worker bee larvae at all the three time points (t-test, p < 0.05) (Figure 2A).

AmKr-h1 Has Higher Expression in Queen Larvae and Is Regulated by JH
The relative mRNA expression levels of AmKr-h1 between queen larvae and worker larvae at 48 h, 84 h, and 120 h after the hatching of eggs were measured using quantitative real-time polymerase chain reaction (qRT-PCR).The results showed that the expression levels of the AmKr-h1 gene in the queen larvae were significantly higher than in the worker bee larvae at all the three time points (t-test, p < 0.05) (Figure 2A).
We investigated the effect of JH on the expression of AmKr-h1.After treatment of the larvae with JH, the expression level of AmKr-h1 in the JH group was significantly higher than that in the ethanol group and the untreated group (ANOVA, p < 0.05), and there was no significant difference between the ethanol group and the no treatment group (ANOVA, p > 0.05) (Figure 2B).We investigated the effect of JH on the expression of AmKr-h1.After treatment of the larvae with JH, the expression level of AmKr-h1 in the JH group was significantly higher than that in the ethanol group and the untreated group (ANOVA, p < 0.05), and there was no significant difference between the ethanol group and the no treatment group (ANOVA, p > 0.05) (Figure 2B).

AmKr-h1 Affects Castes Differentiation of Queen-Worker
The expression changes in AmKr-h1 after RNAi were analyzed using qRT-PCR, and the results showed that the expression levels of AmKr-h1 in the RNAi group were significantly lower than in the Con group (t-test, p < 0.05) (Figure 3A).This indicated that our RNAi experiment was effective.

AmKr-h1 Affects Castes Differentiation of Queen-Worker
The expression changes in AmKr-h1 after RNAi were analyzed using qRT-PCR, and the results showed that the expression levels of AmKr-h1 in the RNAi group were significantly lower than in the Con group (t-test, p < 0.05) (Figure 3A).This indicated that our RNAi experiment was effective.

AmKr-h1 Affects Castes Differentiation of Queen-Worker
The expression changes in AmKr-h1 after RNAi were analyzed using qRT-PCR, and the results showed that the expression levels of AmKr-h1 in the RNAi group were significantly lower than in the Con group (t-test, p < 0.05) (Figure 3A).This indicated that our RNAi experiment was effective.The morphological indexes of the newly emerged adult bees in the RNAi and the Con groups were measured.After knocking down the expression of AmKr-h1, the emergence weight, body length, and forewing length of the newly emerged bees in the RNAi group decreased significantly compared with the Con group (t-test, p < 0.05), while the proboscis length was significantly longer than in the Con group (t-test, p < 0.05) (Figure 3B).The bees in the RNAi group were significantly smaller than those in the Con group (Figure 3C).This indicates that reduced AmKr-h1 expression can lead to the differentiation of larvae into workers.

Summary of Transcriptome Sequencing Data
The number of raw reads obtained after sequencing for each biological replicate in the RNAi and Con groups ranged from 45,302,028 to 62,997,770, and the number of clean reads remaining after filtering low-quality reads and ambiguous reads ranged from 43,925,842 to 60,493,146.The percentages of Q20 and Q30 bases in each sample were higher than 97.16% and 92.36%, respectively.The GC content of clean reads in each sample was between 40.66 and 41.41%.The number of reads matched to the honey bee reference genome ranged from 41,673,370 to 57,479,542, with the proportion ranging from 89.45% to 95.28%, and the uniquely matched reads ranged from 40,781,360 to 56,370,089, with the proportion ranging from 87.76% to 93.83% (Table S2).
The correlation coefficients between biological replicates in the RNAi and Con groups were from 0.9388 to 0.9972, indicating high reliability between biological replicates (Table S3).

DEGs between the RNAi and Con Groups
A total of 191 DEGs were identified between the RNAi and Con groups.Among them, 85 genes were up-regulated and 106 genes were down-regulated in the RNAi group compared with the Con group (Figure 4A, Table S4).Nine DEGs were selected for verifying the reliability of the RNA-seq results using qRT-PCR, including LOC552286, Catalase (Cat, LOC443552), Phosphoenolpyruvate carboxykinase 2 (Pepck 2, LOC412843), LOC107965199, LOC727618, LOC100576458, LOC100577163, Cathepsin D (CathD, LOC409341), and LOC409143.The results indicated that all of them were significantly differentially expressed between the RNAi group and the control group, while six of these genes were down-regulated and three genes up-regulated in the RNAi group.These results are consistent with the RNA-seq results, indicating that the transcriptome sequencing results are reliable (Figure 5).The KEGG pathway enrichment analysis of the DEGs showed that 16 KEGG pathways were significantly enriched (p-value < 0.05), including "longevity-regulating pathway-multiple species", "biotin metabolism", "vitamin digestion and absorption", "linoleic acid metabolism", "terpenoid backbone biosynthesis", "alpha-linolenic acid metabolism", "protein processing in endoplasmic reticulum", "steroid hormone biosynthesis", "arachidonic acid metabolism", "ether lipid metabolism", "lysine biosynthesis", "influenza A", "glycerophospholipid metabolism", "AMPK signaling pathway", "tryptophan metabolism", and "biosynthesis of unsaturated fatty acids".Among these, the "longevity-regulating pathway-multiple species", "steroid hormone biosynthesis", and "AMPK signaling pathway" were associated with honey bee caste differentiation (Figure 4B, Table S5), including a total of eleven DEGs (Figure 4C), and the other pathways were mainly related to metabolism.Nine DEGs were selected for verifying the reliability of the RNA-seq results using qRT-PCR, including LOC552286, Catalase (Cat, LOC443552), Phosphoenolpyruvate carboxykinase 2 (Pepck 2, LOC412843), LOC107965199, LOC727618, LOC100576458, LOC100577163, Cathepsin D (CathD, LOC409341), and LOC409143.The results indicated that all of them were significantly differentially expressed between the RNAi group and the control group, while six of these genes were down-regulated and three genes up-regulated in the RNAi group.These results are consistent with the RNA-seq results, indicating that the transcriptome sequencing results are reliable (Figure 5).Nine DEGs were selected for verifying the reliability of the RNA-seq results using qRT-PCR, including LOC552286, Catalase (Cat, LOC443552), Phosphoenolpyruvate carboxykinase 2 (Pepck 2, LOC412843), LOC107965199, LOC727618, LOC100576458, LOC100577163, Cathepsin D (CathD, LOC409341), and LOC409143.The results indicated that all of them were significantly differentially expressed between the RNAi group and the control group, while six of these genes were down-regulated and three genes up-regulated in the RNAi group.These results are consistent with the RNA-seq results, indicating that the transcriptome sequencing results are reliable (Figure 5).

DEASEs between the RNAi and Control Groups
In this study, five alternative splicing forms between the RNAi and Con groups were analyzed, including skipped exon (SE), alternative 5′ splice site (A5SS), alternative 3′ splice 5. Verification of DEGs using qRT-PCR.Quantitative RT-PCR verification of nine DEGs between the RNAi and control (Con) groups.Asterisk (*) represents significant difference at p < 0.05.

DEASEs between the RNAi and Control Groups
In this study, five alternative splicing forms between the RNAi and Con groups were analyzed, including skipped exon (SE), alternative 5 splice site (A5SS), alternative 3 splice site (A3SS), mutually exclusive exons (MXE), and retained intron (RI), and a total of 42,174 and 44,051 splicing forms were detected in the three biological replicates from the RNAi and Con groups, respectively, of which skipped exon (SE) was the most splicing type.In addition, there were 682 DEASEs related to 495 genes between the RNAi and Con groups, including 323 SE, 121 A5SS, 115 A3SS, 52 MXE, and 71 RI (Figure 6A), related to 262, 113, 101, 41, and 49 genes, respectively (Figure 6A, Table S6).Of the genes related to these DEASEs, HR3 (LOC408586), Hex110 (LOC551648), and Br-C (LOC552255) were involved in honey bee caste differentiation.

Discussion
In the present study, we found that RNAi of the AmKr-h1 gene resulted in the development of honey bee larvae into workers.In honey bees, JH is the principal factor influencing caste differentiation.AmKr-h1 is the response gene downstream of the JH receptors Met and Gce, which are early response genes to JH [30,31].Data from our results suggest that JH regulates the caste differentiation in honey bees by acting on the AmKr-h1 gene.
The Cat gene encodes a catalase important to the cellular antioxidant system.It can guide the conversion of oxygen radicals in cells into hydrogen peroxide, with lower oxidative activity, and then be broken down, thus reducing the degree of cellular oxidation

Discussion
In the present study, we found that RNAi of the AmKr-h1 gene resulted in the development of honey bee larvae into workers.In honey bees, JH is the principal factor influencing caste differentiation.AmKr-h1 is the response gene downstream of the JH receptors Met and Gce, which are early response genes to JH [30,31].Data from our results suggest that JH regulates the caste differentiation in honey bees by acting on the AmKr-h1 gene.
The Cat gene encodes a catalase important to the cellular antioxidant system.It can guide the conversion of oxygen radicals in cells into hydrogen peroxide, with lower oxidative activity, and then be broken down, thus reducing the degree of cellular oxidation [35].Studies have shown that Cat is significantly expressed at a higher level in queen larvae than in worker larvae, suggesting that the queen is more efficient in removing reactive oxygen species (ROS) and has a stronger antioxidant capacity than workers [36].In our results, the expression of the Cat gene was down-regulated after AmKr-h1 RNAi, suggesting that AmKr-h1 may influence the antioxidant capacity of honey bees by regulating the Cat gene.
The CathD gene encodes a lysosomal aspartate protease.In insects, lysosome-mediated organelle degradation is particularly important in liposomal remodeling during the vitellogenesis cycle of female larvae [37,38].In the fat bodies of insects with periodic egg maturation, lysosomes play a key role in terminating vitellogenesis by selectively degrading the secretory machinery involved in the mass production of yolk protein precursors [39].During the termination of vitellogenesis in A. aegypti, the number of lysosomes in the fat body increases significantly [40].In the present study, the expression of the CathD gene was up-regulated after RNAi of the AmKr-h1 gene, indicating that AmKr-h1 can suppress the expression of CathD, which leads to a decrease in aspartate protease activity and, thus, promotes honey bee ovarian development.
The Pepck2 encodes a phosphoenolpyruvate carboxykinase (PEPCK) that is closely associated with pyruvate homeostasis [41].In Helicoverpa armigera, high levels of pyruvate promote growth and development, and low levels of pyruvate lead to reduced metabolic activity [42].In the present study, the expression of Pepck2 was significantly down-regulated after RNAi of the AmKr-h1 gene, indicating that the AmKr-h1 gene may affect the growth and development of honey bees by regulating Pepck2.
Alternative splicing has a significant effect on caste differentiation [43].Among the genes associated with DEASEs, we found three genes associated with honey bee caste differentiation: HR3, Hex110, and Br-C.The HR3 gene encodes for a 20E-induced transcription factor that acts as a developmental switch in 20E-regulated insect development and metamorphosis [44].In A. aegypti, inhibition of HR3 gene expression by RNAi resulted in diminutive ovaries and a significantly reduced number of eggs [45].Hexamerins are mainly involved in the dynamic changes in amino acid storage and utilization that occur during insect development and they can also function as JH-binding proteins [46][47][48].In Reticulitermes flavipes, inhibition of hexamerins' expression promotes the differentiation of larvae into workers [49].The Hex110 gene is highly expressed in the ovaries of the queens and is associated with the reproductive function in honey bees [50].Br-C, a 20E response gene, is essential for maintaining the ecdysteroid titer [51].During larval-pupal metamorphosis, Ecdysone signaling induces programmed cell death by inducing Br-C gene expression, eliminating larval tissues, and promoting adult tissue formation [52].RNAi of the Br-C gene in Drosophila larvae at 120 h leads to a delay in pupal development [53].In the present study, the expression levels of the HR3 and Hex110 isoforms were down-regulated and the Br-C isoform was up-regulated after inhibiting AmKr-h1 expression, indicating that AmKr-h1 may influence the caste differentiation in honey bees by positively regulating the expression levels of the HR3 and Hex110 isoforms while negatively regulating the Br-C isoform.
Through a KEGG pathway enrichment analysis of DEGs and DEASE-related genes, we found that several important pathways related to caste differentiation in honey bees were significantly enriched, such as the longevity-regulating pathway-multiple species, AMPK signaling pathway, MAPK signaling pathway-fly, and Wnt signaling pathway.This suggests that the JH signaling pathway has a cross-talk with the Wnt signaling pathway.It suggests that AmKr-h1, upon receiving juvenile hormone signals, may induce differentiation of queen and worker bees by acting on these critical pathways, but further experimental verification is needed.

Conclusions
In summary, our results indicate that JH regulates caste differentiation in honey bees by acting on AmKr-h1.Moreover, we identified a large number of differentially expressed genes after AmKr-h1 knockdown, and future studies should focus on elucidating the specific pathways downstream of AmKr-h1 that affect phenotypic differentiation between honey bee castes.

Figure 1 .
Figure 1.The intron/exon structure and amino acid sequence comparison of the AmKr-h1 gene of insects.(A) The intron/exon structure and predicted protein of AmKr-h1 gene.In the upper part, boxes, lines, and black boxes represent exons, introns, and coding regions, respectively; (B) Amino acid sequence comparison of the C2H2 zinc finger domains between the Kr-h1 protein of A. mellifera and other species.

Figure 1 .
Figure 1.The intron/exon structure and amino acid sequence comparison of the AmKr-h1 gene of insects.(A) The intron/exon structure and predicted protein of AmKr-h1 gene.In the upper part, boxes, lines, and black boxes represent exons, introns, and coding regions, respectively; (B) Amino acid sequence comparison of the C 2 H 2 zinc finger domains between the Kr-h1 protein of A. mellifera and other species.

Figure 2 .
Figure 2. Differences in expression of AmKr-h1 between A. mellifera queen larvae and worker larvae as well as expression changes after JH treatment.(A) Relative mRNA expression levels of AmKr-h1 in queen larvae and worker larvae at 48 h, 84 h, and 120 h of the larval stage.Different lowercase letters represent significant differences at p < 0.05; the same in (B).(B) Relative mRNA expression levels of AmKr-h1 in larvae of Apis mellifera after JH treatment.JH: the group treated with juvenile hormone III; ET: the group treated with ethanol; NO: the group without treatment.

Figure 3 .
Figure 3.Effect of AmKr-h1 on honey bee caste differentiation.(A) Expression change of AmKr-h1 in larvae of A. mellifera after RNAi.Different lowercase letters represent a significant difference at p < 0.05; the same in (B).(B) Morphological changes in newly emerged bees after AmKr-h1 RNAi.(C)

Figure 2 .
Figure 2. Differences in expression of AmKr-h1 between A. mellifera queen larvae and worker larvae as well as expression changes after JH treatment.(A) Relative mRNA expression levels of AmKr-h1 in queen larvae and worker larvae at 48 h, 84 h, and 120 h of the larval stage.Different lowercase letters represent significant differences at p < 0.05; the same in (B).(B) Relative mRNA expression levels of AmKr-h1 in larvae of Apis mellifera after JH treatment.JH: the group treated with juvenile hormone III; ET: the group treated with ethanol; NO: the group without treatment.

Figure 2 .
Figure 2. Differences in expression of AmKr-h1 between A. mellifera queen larvae and worker larvae as well as expression changes after JH treatment.(A) Relative mRNA expression levels of AmKr-h1 in queen larvae and worker larvae at 48 h, 84 h, and 120 h of the larval stage.Different lowercase letters represent significant differences at p < 0.05; the same in (B).(B) Relative mRNA expression levels of AmKr-h1 in larvae of Apis mellifera after JH treatment.JH: the group treated with juvenile hormone III; ET: the group treated with ethanol; NO: the group without treatment.

Figure 3 .
Figure 3.Effect of AmKr-h1 on honey bee caste differentiation.(A) Expression change of AmKr-h1 in larvae of A. mellifera after RNAi.Different lowercase letters represent a significant difference at p < 0.05; the same in (B).(B) Morphological changes in newly emerged bees after AmKr-h1 RNAi.(C)

Figure 3 .
Figure 3.Effect of AmKr-h1 on honey bee caste differentiation.(A) Expression change of AmKr-h1 in larvae of A. mellifera after RNAi.Different lowercase letters represent a significant difference at p < 0.05; the same in (B).(B) Morphological changes in newly emerged bees after AmKr-h1 RNAi.(C) An example of morphological change in newly emerged bees after AmKr-h1 RNAi."Con" represents control group.

Figure 4 .
Figure 4. DEGs between the RNAi and Con groups.(A) Up-and down-regulated DEGs in the RNAi group compared to the control group in the larval stage of A. mellifera.Both the RNAi group and control group contain three biological replicates.(B) The significantly enriched KEGG pathways of the DEGs identified between the RNAi group and the control group.Pathways marked with a box are those associated with caste differentiation.(C) The DEGs involved in honey bee caste differentiation."Con" represents control group.

Figure 4 .
Figure 4. DEGs between the RNAi and Con groups.(A) Up-and down-regulated DEGs in the RNAi group compared to the control group in the larval stage of A. mellifera.Both the RNAi group and control group contain three biological replicates.(B) The significantly enriched KEGG pathways of the DEGs identified between the RNAi group and the control group.Pathways marked with a box are those associated with caste differentiation.(C) The DEGs involved in honey bee caste differentiation."Con" represents control group.
group compared to the control group in the larval stage of A. mellifera.Both the RNAi group and control group contain three biological replicates.(B) The significantly enriched KEGG pathways of the DEGs identified between the RNAi group and the control group.Pathways marked with a box are those associated with caste differentiation.(C) The DEGs involved in honey bee caste differentiation."Con" represents control group.

Figure 5 .
Figure 5. Verification of DEGs using qRT-PCR.Quantitative RT-PCR verification of nine DEGs between the RNAi and control (Con) groups.Asterisk (*) represents significant difference at p < 0.05.

Figure 6 .
Figure 6.DEASEs between the RNAi and Con groups of honey bee.(A) Differential alternative splicing events between the RNAi group and control (Con) group.(B) The significant enriched KEGG pathways between the RNAi group and control (Con) group.The pathways marked with a box are associated with honey bee caste differentiation.

Figure 6 .
Figure 6.DEASEs between the RNAi and Con groups of honey bee.(A) Differential alternative splicing events between the RNAi group and control (Con) group.(B) The significant enriched KEGG pathways between the RNAi group and control (Con) group.The pathways marked with a box are associated with honey bee caste differentiation.