Abstract
Among hymenopteran insects, aculeate species such as bees, ants, and wasps have enlarged and morphologically elaborate mushroom bodies (MBs), a higher-order brain center in the insect, implying their relationship with the advanced behavioral traits of aculeate species. The molecular bases leading to the acquisition of complicated MB functions, however, remains unclear. We previously reported the constitutive and MB-preferential expression of an ecdysone-signaling related transcription factor, Mblk-1/E93, in the honey bee brain. Here, we searched for target genes of Mblk-1 in the worker honey bee MBs using chromatin immunoprecipitation sequence analyses and found that Mblk-1 targets several genes involved in synaptic plasticity, learning, and memory abilities. We also demonstrated that Mblk-1 expression is self-regulated via Mblk-1-binding sites, which are located upstream of Mblk-1. Furthermore, we showed that the number of the Mblk-1-binding motif located upstream of Mblk-1 homologs increased associated with evolution of hymenopteran insects. Our findings suggest that Mblk-1, which has been focused on as a developmental gene transiently induced by ecdysone, has acquired a novel expression pattern to play a role in synaptic plasticity in honey bee MBs, raising a possibility that molecular evolution of Mblk-1 may have partly contributed to the elaboration of MB function in insects.
Similar content being viewed by others
Introduction
In insects, mushroom bodies (MBs), a higher-order center in insect brains, play important roles in olfactory associative learning1, spatial learning2, sensory integration3, and social behavior4. In Hymenoptera, aculeata species such as bees, ants, and wasps have enlarged and morphologically elaborated MBs compared to the most primitive hymenopteran insects like sawflies, suggesting a relationship between MB elaboration and the advanced behavioral traits of these insects5,6. The molecular and neural bases underlying how aculeate MBs acquired elaborated functions, however, remain elusive. Aculeate hymenopteran species show advanced social behaviors7,8. For example, European honey bee (Apis mellifera L.) workers perform the ‘waggle dance’ to inform their nestmates of the location of food sources that they memorize during foraging flights9. Therefore, honey bees are useful model insects for understanding neural mechanisms underlying sociality, learning ability10,11 and brain evolution in insects.
The honey bee MBs comprise 4 types of interneurons termed Kenyon Cells (KCs): class I large-, middle-, small-type KCs, and class II KCs, which are distinguished on the basis of their location, the size of their somata, and their gene expression profiles12. Among these KC subtypes, large-type KCs (lKCs) are characteristic in that they preferentially express genes involved in Ca2+-signaling, including Ca2+/calmodulin-dependent protein kinase II (CaMKII)13,14. Since Ca2+-signaling pathways play an important role in synaptic plasticity, the primary mechanism of learning and memory functions in many animal species15, it is plausible that lKCs are involved in the learning and memory abilities that underlie honey bee social behaviors.
In addition to the genes involved in Ca2+-signaling, a sequence-specific transcription factor, termed mushroom body large-type Kenyon-cell specific protein-1 (Mblk-1)/ecdysone-induced protein 93F (E93) (hereafter termed as Mblk-1) is preferentially expressed in lKCs16,17,18. Mblk-1 is widely conserved in animal species from invertebrates to vertebrates19. Especially in insects, E93, the Mblk-1 homolog, is upregulated by the molting hormone ecdysone and ecdysone receptor and triggers programmed cell death of larval tissues during metamorphosis20,21,22,23,24. In Drosophila melanogaster, E93 (DmE93) is expressed in MB neuroblasts during later pupal stages where it activates autophagy to eliminate the neuroblasts24. In the honey bee, the preferential expression of Mblk-1 in lKCs suggests its important role in the function of lKCs, but the function of Mblk-1 as a transcription factor and its target genes in honey bee lKCs are unknown.
We previously used an affinity-purified antibody raised against a partial recombinant Mblk-1 protein to examine Mblk-1 protein expression in the honey bee brain18. In the present study, we performed chromatin immunoprecipitation sequencing (ChIP-seq) using this antibody to identify Mblk-1 target genes in honey bee brains. We found that Mblk-1 regulates genes related to synaptic plasticity, such as CaMKII25,26,27 and Mblk-1 itself, suggesting that Mblk-1 has acquired a new expression mechanism and roles related to synaptic plasticity in the honey bee MBs. We propose that acquisition of unconventional functions of Mblk-1 in lKCs might have partly contributed to the evolution of learning and memory abilities and social behaviors in aculeate Hymenoptera.
Methods
Insects
European honey bee colonies maintained at The University of Tokyo (Hongo Campus) were used. Some colonies were also purchased from a local dealer (Kumagaya Honeybee Farm, Saitama, Japan). For ChIP-seq analysis using adult worker samples, we randomly collected workers from the hives irrespective of age and/or role. For ChIP-seq analysis using pupal worker samples, we selected and used worker pupae with brown or purple eyes (stages: P3-P5). Sawflies (Athalia rosae) were a gift from Mr. Takayoshi Kuwabara. Adult sawflies within 10 days after emergence were used for the experiment.
Analysis of genes expressed in adult worker MBs using RNA-seq data
The RNA-seq data (low-value reward and no-dance) was obtained from GEO accession: GSM374796728. The value of 2012_309_TTAGGC_L007_Aligned.sortedByCoord.out.bam in GSM3747967_2012_309_TTAGGC_L007_Aligned.sortedByCoord.out.bam.counts.txt was divided by each gene length, and the top 40 genes with the highest divided values were displayed.
Chromatin preparation
80 MBs dissected from adult workers or 60 whole brains dissected from worker pupae were used per single chromatin immunoprecipitation sequence (ChIP-seq) experiment. The ChIP-seq experiments were conducted using 2 biological replicates for the adult and pupal samples. The MBs or whole brains were dissected under binocular microscopy from cooled and anesthetized honey bees and homogenized with plastic pestles in 200 µl of buffered insect saline (20 mM Tris–HCl [pH 7.4], 130 mM NaCl, 5 mM KCl, 1 mM CaCl2, 10 mM HEPES buffer, protease inhibitor cocktail [Roche], and PhosSTOP [Roche]). The cells were fixed with formaldehyde (1% final conc.) for 12 min at room temperature. The fixing reaction was quenched with 2.5 M glycine (125 mM final conc.) for 10 min on ice. The cells were washed 3 times with 150 µl of buffered insect saline and centrifuged (600×g, 5 min, 4 °C). The supernatant was removed, and the cells were suspended in 300 µl of buffered insect saline. The cells were crushed for 25 s using zirconium beads (TOMY) and Micro Smash MS-100 (TOMY). After centrifugation, the supernatant was discarded and 200 µl of nuclei lysis solution (10% glycerol, 0.5% NP-40, 2 mM EDTA, 50 mM Tris–HCl [pH 7.4], and protease inhibitor cocktail [Roche]) were added. The sample was incubated on ice for 30 min. After centrifugation, the supernatant was discarded and 500 µl of Micrococcal Nuclease buffer (50 mM Tris–HCl [pH 7.4] and 5 mM CaCl2) was added. The sample was kept on ice for 10 min and centrifuged. After the supernatant was discarded, the pellet was resuspended in 140 µl of Micrococcal Nuclease buffer containing protease inhibitor cocktail and 0.2 µl of Micrococcal Nuclease (2000 gel units/µl, New England BioLabs). The sample was incubated at 37 °C for 12 min and digestion was stopped by 500 mM EDTA (total volume: 150 µl). After adding 300 µl of 1.5 × ChIP dilution buffer (1 × ChIP dilution buffer: 50 mM Tris–HCl [pH 7.4], 167 mM NaCl, 1.1% Triton X-100, 0.11% SDS), the sample was sonicated (output 10%, 12 s at 60 s intervals, 3 times) using a SONIFIER250 (Branson). After sonication, the sample was centrifuged, and the supernatant was collected for immunoprecipitation.
Chromatin immunoprecipitation (ChIP)
The sample containing the fragmented chromatin and anti-Mblk-1 antibody-attached Dynabeads Protein A (Invitrogen) were incubated with rotation for 45 min at 4 °C. The Dynabeads complex was washed twice with RIPA buffer (50 mM Tris–HCl [pH 7.4], 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% Triton X-100), and 3 times with Phosphate-Buffered Saline (PBS). The washed Dynabeads complex was resuspended in PBS. The solution containing the beads was transferred to a new clean tube. The tube was placed on magnet and the supernatant was removed. The anti-Mblk-1 antibody and protein complex was eluted by elution buffer (1% SDS, 50 mM Tris–HCl [pH 7.4], 8.5 mM EDTA). The tube was placed on magnet and the supernatant was transferred to a new tube. After adding TE buffer and 10 mg/ml RNase A, the sample was incubated for 1 h at 50 °C, and then 20 mg/ml proteinase K was added, and the crosslink was reversed by incubation for 12 h at 50 °C. The DNA was purified and processed to sequencing preparation. The sequencing library was prepared according to the Illumina protocol and sequenced on an Illumina Hiseq2500.
ChIP-seq data processing
Sequence reads were aligned to the reference genome (Amel_HAv3.1)29 using Bowtie2 with default parameters30. Peak calling was performed using MACS231 and the input DNA sample as a control. After peak calling, we visualized the ChIP-seq data using the Integrative Genomics Viewer (IGV)32. The ChIP-seq experiments were performed twice using 2 biological replicates. The peak signals detected in the 2 replicates were considered reproducible if they were located within 1 kb in the genome. Subsequent analysis was conducted using the signals that were reproducible.
GO enrichment analysis
Taking advantage of the well-characterized GO annotations for Drosophila melanogaster genes, we first assigned a Drosophila melanogaster ortholog for each Apis mellifera protein by reciprocal BLAST searches between all Apis mellifera protein sequences (Amel_HAv3.1) and all Drosophila melanogaster protein sequences (UP000000803 in UniProt). The Drosophila melanogaster protein with the reciprocal best hit was considered the ortholog for each Apis mellifera protein. Genes unique to Apis mellifera and/or genes with no corresponding Drosophila melanogaster homologs were not used for further analysis. We performed GO enrichment analysis using the assigned Drosophila melanogaster homologs and Metascape33.
Motif analysis
Sequences of ± 150 bp (300 bp) from the centers of the reproducible signals with a peak calling q-value threshold of 0.001 in the analysis of adult MBs were obtained. These sequences were inputted into MEME-ChIP (Version 5.2.0), which discovers novel DNA-binding motifs with MEME and DREME34. As the motif width options, -minw was set to 6 and -maxw was set to 17. No control sequences were used for MEME. The positive sequences were shuffled to create the control set for DREME.
Luciferase assay
We obtained the coding sequence of Mblk-1 from Mblk-1/pPac-PL35 and inserted it into a multiple cloning region of an expression vector, pHEK293 Ultra Expression Vector II (TaKaRa). Reporter vectors were constructed by inserting 6xGAGA, 6xGAGA with a 4-bp spacer sequence, 6xATTTTG, 6xGAAATTTT, 6xATTTTGG, and an Mblk-1-binding region sequence (NC_037652.1:7063408-7063595, 188 bp), respectively, into the upstream region of minimal promoter of pGL4.23 [luc2/minp]. For the Mblk-1-binding region sequence, we used the overlap region of the 2 Mblk-1-binding sequences (NC_037652.1:7063147-7063709 and NC_037652.1:7063400-7063629) obtained from the 2 independent ChIP-seq experiments. pRL-TK was used as an internal control vector to normalize transfection efficiency. The Mblk-1 expression vector or pHEK 293 Ultra Expression Vector II without the Mblk-1 coding sequence as a negative control, pHEK293 Enhancer Vector, one of the reporter vectors (pGL4.23 with the motif sequences), and the internal control vector (pRL-TK) were transfected to HEK293T cells (RCB2202) from RIKEN BioResource Center using Lipofectamine LTX Reagent with PLUS Reagent (Invitrogen). The pHEK293 Enhancer Vector was used to enhance the expression of Mblk-1 in pHEK293 Ultra Expression Vector II. We measured Photinus pyralis (firefly) and Renilla reniformis luciferase activities 40–48 h after transfection with a Dual-Luciferase Reporter Assay System (Promega) and 2030 ARVO X5 (PerkinElmer). Photinus pyralis (firefly) luciferase activity was first normalized with the co-transfected internal control luciferase activity and then values of normalized luciferase activity of the control assay using the pHEK 293 Ultra Expression Vector II without the Mblk-1 coding sequence were subtracted to exclude effects other than those of Mblk-1 expression. The obtained value is shown as relative luminescence on the vertical axis of the graphs. We compared the values from vectors containing candidate sequences with that of the control vector (pGL4.23) to determine how the sequences affect the expression of luciferase.
Single-cell RNA-seq data analysis
We obtained single-cell RNA-seq data from the previously reported GEO accession ID GSE13078536. Using Seurat Version 4.1.137, we selected cells for clustering under the following conditions (min.cells = 3 and min.features = 200 and nCount_RNA > 1000 and nFeature_RNA > 500). After SCTransform normalization, we performed cell clustering (PC = 30 and resolution = 0.5). Doublet cells were detected using DoubletFinder (doubletFinder_v3(PCs = 1:30, pN = 0.25, pK = 0.01, nExp = nExp_poi.adj, reuse.pANN = FALSE, sct = TRUE))38. Cells determined to be doublets were removed and marker genes for each cluster were found by FindAllMarkers. Marker genes with an adjusted p-value under 0.05 were used for further analyses.
RNA probe synthesis and in situ hybridization
Total RNA was extracted from 3 honey bee adult worker brains using TRIzol Reagent (Invitrogen). After removing the genomic DNA with the gDNA Eraser from the PrimeScript RT reagent Kit with gDNA Eraser (Perfect Real Time), the extracted RNA was reverse-transcribed into cDNA using the same kit. The cDNA was partially amplified by PCR using Ex Taq (TaKaRa) with primers specific for each gene. The PCR products were cloned using pGEM-T Easy Vector Systems (Promega). The primer sequences used to synthesize the in situ hybridization RNA probes are listed in Supplementary Table 1. The DNA templates for in vitro transcription were amplified by PCR with M13 forward and reverse primers. The digoxigenin (DIG)-labeled sense and antisense RNA probes were synthesized using the amplified DNA template, a DIG RNA Labeling Mix (Roche), SP6 RNA polymerase (Roche), and T7 RNA polymerase (Roche). For in situ hybridization, dissected whole brains were embedded in Tissue-Tek OCT compound (Sakura Finetek Japan) without fixation and immediately frozen and stored at − 80 °C until use. The frozen whole brains were sliced into 10-µm thick sections. In situ hybridization was performed essentially as described previously39. Images were captured with a light microscope BX-50 (Olympus).
Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)
Total RNA was extracted from each tissue (brain, head without brain, thorax, and abdomen) using TRIzol Reagent (Invitrogen). cDNA synthesis was performed using a PrimeScript RT reagent Kit with a gDNA Eraser (Perfect Real Time; TaKaRa). qRT-PCR analysis was performed using TB Green Premix Ex Taq II (Tli RNaseH plus; Takara) and Light Cycler 480 Instrument II (Roche Life Science, Indianapolis, IN, USA). The following primers were used for gene amplification: Mblk-1 homolog (LOC105691346), 5′-GCAACAGCAACAGAGAGAAAGG-3′ and 5′-CGTTCCCGTACCGTCATAC-3′; 60S ribosomal protein L32 (LOC105692622), 5′-ACAGAGTTCGTAGGCGCTTC-3′ and 5′-GCATCATCAAGACCTCCAAC-3′. The PCR conditions were as follows: denaturation at 95 °C for 30 s; 45 cycles of PCR (95 °C for 5 s, 61 °C for 15 s, 72 °C for 20 s), melting, 95 °C for 5 s, 65 °C for 60 s, then 95 °C (Ramp Rate: 0.11 °C/s) and cooling, 50 °C for 30 s.
Analysis of the GAGA motifs upstream of Mblk-1 homologs
Upstream sequences within 2 kb or 10 kb of the transcription start site of each Mblk-1 homolog were obtained from NCBI. The numbers of GAGA or GAG sequences in the upstream region of each Mblk-1 homolog were counted, allowing for overlap. For example, a ‘GAGA’ sequence was counted as one GAGA and one GAG, and a ‘GAGAG’ sequence was counted as one GAGA and two GAGs. The assemblies used were as follows: Apis mellifera, Amel_HAv3.129; Harpegnathos saltator, Hsal_v8.540; Solenopsis invicta, UNIL_Sinv_3.0; Vespa mandarinia, V.mandarinia_Nanaimo_p1.0; Fopius arisanus, ASM80636v141; Nasonia vitripennis, Nvit_psr_1.142; and Athalia rosae, Aros_2.043.
Statistical analysis
Statistical analysis was conducted as indicated in the text and figure legends. Where required P values were adjusted using Bonferroni or Benjamini–Hochberg correction. In Fig. 3D, Tukey–Kramer’s tests after one-way ANOVA were performed (n = 3). F values, P values, and degrees of freedom were 8.09, 0.00831 and 3 (right panel) and 99.3, 2.52 × 10–5 and 2 (left panel), respectively. Student’s t test used in Fig. 6B was two sided (n = 4). In Fig. 6D, one-way ANOVA was performed. F value was 2.09, P value was 0.180 and degrees of freedom was 3 (n = 3).
Results
Identification of Mblk-1 target gene candidates in the worker MBs by ChIP-seq analysis
Mblk-1 is expressed in a lKC-preferential manner in adult worker MBs (Fig. 1A)16,18. When we searched for highly expressed genes in the honey bee MBs using previously reported RNA-seq data28, we found that Mblk-1 was the 38th most highly expressed gene and the most highly expressed among transcription factors in MBs (Fig. 1B), implying its important role in the adult honey bee MBs.
Analysis of genes highly expressed in adult worker MBs. (A) Schematic drawings of the head, brain, and MBs of the honey bee. The MBs are a paired structure in the insect brain, and the somata of the 3 class I KC subtypes are localized inside of each calyx, whereas those of class II KCs are localized at the outer surface of the MB calyces. lKCs, large-type KCs shown in green; mKCs, middle-type KCs shown in red; sKCs, small-type KCs shown in blue. (B) List of the top 40 genes highly expressed in adult worker MBs. Gene ID and gene description are provided for each gene.
We previously prepared an antibody that binds specifically to Mblk-1 protein (hereafter termed anti-Mblk-1 antibody) (Fig. S1A)18. We first tested whether the anti-Mblk-1 antibody could be used for ChIP-seq analysis. In the immunoprecipitated sample, as well as in the input and supernatant samples after immunoprecipitation using worker MB lysate, full-length Mblk-1 protein was mainly detected as a 250-kDa band (Fig. S1B). The molecular weight of Mblk-1 is estimated as 175 kDa based on the protein sequence, but our previous study confirmed that the full-length Mblk-1 protein migrated at the position around 250-kDa in SDS-PAGE analysis18. Therefore, the immunoprecipitation results confirmed that the anti-Mblk-1 antibody can be used for immunoprecipitation.
We performed ChIP-seq analysis using homogenates of adult worker MBs and the anti-Mblk-1 antibody. A total of 593 genes were identified as target gene candidates of Mblk-1, in which Mblk-1-binding signals were located within 10 kb upstream or downstream of the corresponding genes (Table 1, Supplementary Data 1, and 3) when using a peak calling q-value threshold of 0.005 and the input DNA sample as a control. Schematic diagrams of some peak signals and their surrounding Mblk-1 target gene candidates are shown in Fig. 2A–F. The signal with the lowest q-value was in the third intron of an uncharacterized non-coding RNA gene (Gene 1 in Table 1 and Fig. 2A). Genes related to neural function, such as suppressor of lurcher protein 1 (Gene 2–1 in Table 1 and Fig. 2B), fasciclin-3 (Gene 4 in Table 1 and Fig. 2C), potassium voltage-gated channel subfamily KQT member 1 (Gene 9 in Table 1 and Fig. 2D), and CaMKII (Gene 17-1 in Table 1 and Fig. 2F) were also identified as Mblk-1 target gene candidates. Moreover, among the Mblk-1 target gene candidates, some ecdysone-signaling pathway-related genes such as ecdysone-induced protein 75 (E75) and Ultraspiracle (Usp) were identified. This is consistent with the fact that E75 is a direct regulatory target of DmE9321. Usp was listed in Table 1, and the peak signal was detected in the 5′ untranslated region of this gene (Fig. 2E).
Mblk-1-binding signals and their surrounding genomic regions. The waveforms in each panel show the accumulation of sequence reads in the corresponding genomic region when using samples immunoprecipitated by the anti-Mblk-1 antibody (upper lane), and input samples (lower lane, as negative control). The Mblk-1-binding signals are boxed with black lines in the waveform lanes. Bars and boxes under the waveforms indicate exons and introns, respectively, annotated within the corresponding genomic regions. The gene ID and gene name of each gene are shown below the gene structure. (A–F) Panels correspond to (A) Gene Nos. 1, (B) Gene Nos. 2, (C) Gene Nos. 4, (D) Gene Nos. 9, (E) Gene Nos. 14, and (F) Gene Nos. 17 listed in Table 1, respectively. Note that there are sometimes overlapping genes for each genomic region, for example, LOC726948, which is annotated as suppressor of lurcher protein 1, overlaps with LOC102655522 and LOC102655477.
Three distinct DNA binding motifs of Mblk-1
To search for Mblk-1-binding motifs, we next performed motif analysis using sequences corresponding to the Mblk-1-binding regions and MEME-ChIP34. Three consensus sequences were identified by DREME, which discovers short and ungapped motifs: GA-rich sequence (GAGA or GAG, hereafter termed a GAGA motif), GAAATTTT, and AT(C)TTTGTA (Fig. 3A). A combination of these motifs was also found by MEME, which discovers relatively long motifs (Fig. 3B). In a previous study, we identified an Mblk-1 preferred binding sequence named MBE (Mblk-1-binding element, 5′-AGGTAGAGATCGATCGATAGGG-3′) by using an in vitro binding site selection method17. The motif we found in this study is consistent with MBE, because MBE has a GAGA sequence. In addition, the ATTTTC(T)GG motif is reported as a binding motif of DmE9344, and it is also reported that Bombyx mori E93 (BmE93) binds to GAGA-containing motifs45. These previous findings support the quality of our data and suggest that Mblk-1-binding motifs are conserved in insects.
Mblk-1-binding motif analysis and luciferase assay. (A) Three Mblk-1-binding motifs identified with DREME. (B) Two Mblk-1-binding motifs identified with MEME. (C) pGL4.23 reporter vector construction used for the luciferase assay. (D) The results of the luciferase assay using reporter vectors containing motif 1: 6xGAGA or 6xGAGA with 4-bp spacer sequence (left panel) and motif 2: 6xGAAATTTT or motif 3: 6xATTTTGG sequence (right panel). Data represent means ± SD (n = 3). Significant differences are indicated by asterisks (*p < 0.05, **p < 0.01, ***p < 0.005, Tukey–Kramer test).
We next performed a luciferase assay to examine whether Mblk-1 recognizes these 3 binding motifs and affects the transcription of a reporter gene. Human HEK293T cells were transfected with an Mblk-1 expression vector; a Renilla luciferase reporter vector; and 5 types of pGL4.23 firefly luciferase reporter vectors, each of which contains 6xGAGA, 6xGAGA with 4-bp spacer sequences, 6xATTTTG, 6xGAAATTTT, or 6xATTTTGG upstream of the minimal promoter, respectively (Fig. 3C). Normalized firefly luciferase activity increased about 3.0-fold and 6.6-fold when co-transfected with pGL4.23 vectors containing 6xGAGA or 6xGAGA with 4-bp spacer sequences, respectively, compared with that when co-transfected with a pGL4.23 vector containing no binding motifs (Fig. 3D). This finding indicates that Mblk-1 recognizes GAGA motifs and upregulates the transcription of genes with these motifs. Normalized firefly luciferase activity, however, significantly decreased by approximately 60% when co-transfected with pGL4.23 vectors containing 6xGAAATTTT compared with that when co-transfected with a pGL4.23 vector containing no binding motifs (Fig. 3D). This finding suggests that Mblk-1 recognizes GAAATTTT motifs to downregulate the transcription of genes having this motif in contrast to those having GAGA motifs. In addition, no significant change in firefly luciferase activity was observed when co-transfected with a pGL4.23 vector containing 6xATTTTG or 6xATTTTGG (Fig. 3D), although the absence of an effect on the luciferase activity does not necessarily exclude the possibility that Mblk-1 binds to these motifs. Taken together, our results indicated that Mblk-1 affects gene transcription through GAGA or GAAATTTT motifs.
Identification of genes upregulated in lKCs as candidates for Mblk-1 target genes
Given that Mblk-1 upregulates its target genes in a lKC-preferential manner in the honey bee MBs, Mblk-1 target gene candidates are expected to be expressed preferentially in lKCs. To find genes expressed preferentially in lKCs among the identified Mblk-1 target gene candidates, we used recently reported honey bee MB single-cell RNA-seq data36. When the MB cells were classified into 10 clusters (Fig. 4A), high expression of PLCe (LOC408804) was observed in clusters 1–8, but not in clusters 0 or 9 (Fig. S2A). We considered clusters 1–8 with high PLCe expression to be KCs, because we previously reported high expression of this gene in all KCs (the entire MBs)39. Of these 8 clusters, preferential Mblk-1 expression was observed in clusters 1, 2, 6, and 7 (Fig. 4B). In addition to Mblk-1, genes indicated as preferentially expressed in lKCs by in situ hybridization, such as disks large homolog 5 (LOC410178)39 and Ryr (LOC408680)46, were highly expressed at least in 1 of the 4 clusters (clusters 1, 2, 6, and/or 7) (Fig. S2B,C and Supplementary Data 4). Therefore, we regarded these 4 clusters as lKCs. Preferential expression of middle-type Kenyon cell-preferential arrestin-related protein (LOC725542), which is reported to be expressed in mKCs47, was observed in clusters 0, 3, and 8 (Fig. S2D and Supplementary Data 4). This finding suggests that clusters 3 and 8 corresponded to mKCs. Although the expression level of PLCe in cluster 0 was lower than that in other clusters, the expression level of the pan-neuronal marker ELAV-like protein 2 (LOC410689) was high (Fig. S2E), suggesting that cells in this cluster are non-KCs (possibly optic lobe neurons) that contaminated the MB samples. In cluster 4, the most preferentially expressed gene was POU domain, class 6, transcription factor 2 (LOC408435) (Fig. 4C and Supplementary Data 4), but no known KC-subtype marker genes were suggested and therefore the known KCs to which this cluster corresponds remain unclear. Among the 8 clusters corresponding to KCs, Tk (tachykinin) appeared to be preferentially expressed in cluster 5 (Fig. 4C). This gene is expressed in both sKCs and lKCs, with greater expression in sKCs than in lKCs48. In the violin plot, Tk expression was highest in cluster 5 (Fig. S2F), suggesting that cluster 5 corresponds to sKCs. Genes specifically expressed in Class II KCs have not yet been identified by in situ hybridization analyses12, making it difficult to identify which clusters correspond to Class II KCs. Cluster 9 likely comprises doublet cells, because more than half of the cells in this cluster were determined to be doublets by DoubletFinder (Fig. S3).
Single-cell RNA-seq data analysis using adult MBs and identification of Mblk-1 target gene candidates preferentially expressed in lKCs. (A) UMAP visualization of single-cell RNA-seq data obtained from adult honey bee MBs by Traniello et al.36. Clusters 1, 2, 6, and 7 correspond to lKCs. (B) Violin plots of gene expression: Mblk-1, CaMKII, and pumilio homolog 2. (C) Heatmap of expression of the top 5 cluster marker genes in each cluster. Positive values (shown in red or yellow) indicate that the gene is more highly expressed in the cluster indicated than in other clusters. (D) Ratios of genes suggested to be marker genes for 3 of the 4 clusters that correspond to lKCs (light blue), those for 2 clusters (blue), and those for 1 cluster (dark blue), respectively. Numbers in parentheses represent the number of Mblk-1 target gene candidates. (E–G) In situ hybridization results for the 3 marker genes. (E) Ca2+/calmodulin-dependent protein kinase II (F) pumilio homolog 2 (G) CUGBP Elav-like family member 4. The upper panels show results using anti-sense RNA probes, whereas the lower panels show results using sense RNA probes. Schematic drawing of the expression pattern of each gene in the MB is shown at the bottom. MB, mushroom bodies; OL, optic lobes; AL, antennal lobes; lKCs, large-type KCs. Scale bar: 250 μm.
We identified 75 genes as possible Mblk-1 target genes, including Mblk-1 itself, with higher expression in at least one of clusters 1, 2, 6, and 7 than in the other clusters (Supplementary Data 5). Of these 75 genes, 2 genes (CaMKII and pumilio homolog 2) were suggested to be marker genes for 3 of the 4 clusters comprising lKCs (Fig. 4D and Supplementary Data 4); 19 genes, including CUGBP Elav-like family member 4, were marker genes for 2 of the 4 clusters; and 53 genes, including suppressor of lurcher protein 1 and Usp, were marker genes for 1 of the 4 clusters (Fig. 4D). In situ hybridization analysis of these genes was performed to examine whether they are preferentially expressed in lKCs. CaMKII was confirmed to be expressed preferentially in lKCs (Fig. 4E), consistent with a previous report13. In addition, we found preferential expression of pumilio homolog 2 in lKCs (Fig. 4F). CUGBP Elav-like family member 4 was expressed in the whole MBs, but not in other brain regions (Fig. 4G), consistent with the result of scRNA-seq analysis showing that CUGBP Elav-like family member 4 was also a marker gene for clusters other than the 4 clusters, which correspond to lKCs (Supplementary Data 4). Schematic diagrams of the peak signals located around CUGBP Elav-like family member 4 and pumilio homolog 2 are shown in Fig. S4.
Comparison of Mblk-1 target gene candidate profiles between pupal and adult worker honey bee brains
Mblk-1 expression level is higher in pupal brains than adult brains18. In addition, DmE93 is expressed transiently in pupal MB neuroblasts to activate autophagy during metamorphosis24. Thus, we next examined whether Mblk-1 has distinct target genes between adult and pupal worker brains in the honey bee. To this end, we performed ChIP-seq analysis using pupal worker brain homogenates and the anti-Mblk-1 antibody. A total of 263 genes were identified as pupal Mblk-1 target gene candidates, for which Mblk-1-binding signals are located within 10 kb upstream or downstream of the corresponding genes when using a peak calling q-value threshold of 0.005 and the input DNA sample as a control. Approximately half of the pupal Mblk-1 target gene candidates (138/263, 52%) were detected specifically in the ChIP-seq analysis using pupal brains (Fig. 5A, Supplementary Data 2 and 3). The Mblk-1-binding signals with the 20 lowest q-values (top 20 signals) and genes that have these signals within 10 kb upstream or downstream are shown in Fig. 5B. Among the 20 pupal target gene candidates shown in Fig. 5B, 8 genes were pupal-specific (8/20, 40%). For almost half of the 20 signals (9/20, 45%), no genes were localized within 10 kb upstream or downstream of the signals. For example, a neural developmental gene, Down syndrome cell adhesion molecule (Dscam) shown in Fig. 5B was not present among the adult target gene candidates, suggesting that this gene is a pupal-specific target. The peak signal was found in the first intron of Dscam, and only the GAGA motif was found in the Mblk-1-binding region. In addition to Dscam, several developmental genes, including Ultrabithorax and Tyrosine-protein kinase transmembrane receptor Ror, were specifically detected in the pupal analysis (Supplementary Data 2).
Identification of Mblk-1 target gene candidates in pupal brains. (A) Venn diagram showing the numbers of Mblk-1 target gene candidates identified as pupal- or adult-specific and common target genes. (B) List of the top 20 ChIP-seq peak signals identified from pupal worker brain homogenates and Mblk-1 target gene candidates located within ± 10 kb of the corresponding signals. The Gene ID, gene description, signal fold enrichment, − log10 (q-value), and signal position relative to the genes are shown for each ChIP-seq peak signal. Dashes indicate the absence of genes located within ± 10 kb of the corresponding peak signal. Note that side numbers (e.g., 3-1 and 3-2) indicate that multiple genes are located within ± 10 kb of the corresponding peak (e.g., peak 3).
We performed GO enrichment analysis for adult-specific, pupal-specific, or common target gene candidates by Metascape33. The target gene candidates used for the GO enrichment analysis are listed in Supplementary Data 6. In our analysis of the adult-specific target gene candidates, generation of neurons (GO:0048699) and open tracheal system development (GO:0007424) were listed as the top 2 (Fig. S5). On the other hand, in our analysis of the pupal-specific target gene candidates, compound eye development (GO:0048749) and urogenital system development (GO:0001655) were suggested as the top 2 (Fig. S6). We also performed GO enrichment analysis using the common target gene candidates and locomotion (GO:0040011) was suggested as the top GO term (Fig. S7).
Self-regulation of Mblk-1 expression in honey bee MBs
Interestingly, two distinct peak signals were found 2 kb and 25 kb upstream of the Mblk-1 gene, respectively (Fig. 6A). All these signals contained GAGA motifs, which induced the expression of the downstream luciferase gene in the reporter assay (Fig. 3D). From this result, we expected that Mblk-1 induces its own expression via these GAGA motifs. To confirm this, we performed a luciferase assay with a pGL4.23 firefly luciferase reporter vector that contains the sequence corresponding to the peak signal located 25 kb upstream of Mblk-1 gene (Fig. 6A). We observed an approximately 4.9-fold increase in firefly luciferase activity when using the pGL4.23 vector containing the peak signal sequence compared with that when using the pGL4.23 vector containing no binding motifs (Fig. 6B). Therefore, Mblk-1 is suggested to recognize this GAGA-containing region as an enhancer to upregulate its own expression. It is possible that once Mblk-1 is expressed, it can upregulate its own expression, which leads to its constitutive expression in adult honey bee lKCs.
Mblk-1 upregulates its own expression via GAGA motif-containing regions. (A) Two ChIP-seq peak signals were identified upstream of the Mblk-1 gene (outlined in black). Construction of the pGL4.23 reporter vector used in the luciferase assay is indicated below the Mblk-1 gene structure. The transcription start site of the Mblk-1 gene is indicated by a blue arrow. (B) Results of the luciferase assay using a reporter vector containing the Mblk-1-binding region sequence and a control vector. Data represent means ± SD (n = 4). Significant difference is indicated by 3 asterisks (p < 0.005, Student’s t test). (C) Phylogenetic tree showing the relationship among hymenopteran species and Drosophila melanogaster. (D) Comparison of the relative expression of the Athalia rosae Mblk-1 homolog with respect to RpL32 among various adult body parts (head without brain, brain, thorax, and abdomen, n = 3). Black circles represent values from 3 experiments and yellow circles represent their average values. Because expression of the Mblk-1 homolog in brain samples was below the detection threshold, brain Cp values were set to 40 in a total of 45 PCR cycles and the relative expression levels were calculated. (E) Relationship between the number of GAGA and GAG sequences localized within 2 kb upstream of Mblk-1 homologs and expression of Mblk-1 homologs in adult MBs of various hymenopteran insect species. Yellow and gray bars indicate the number of GAGA and GAG sequences, respectively. Table below shows whether Mblk-1 homolog is expressed in the adult MBs in each species.
Expression analysis of Mblk-1 homolog in the sawfly
In the honey bee, Mblk-1 is almost selectively expressed in the brain among various adult tissues18. In addition, in the ant Harpegnathos saltator workers and gamergates, the expression of the Mblk-1 homolog in the brain is observed49. We next tested whether the constitutive expression of Mblk-1 homologs in the adult brain is conserved even in other hymenopteran insect species (Fig. 6C). We performed qRT-PCR analysis to examine the expression of the Mblk-1 homolog (LOC105691346) in various adult body parts of the sawfly (Athalia rosae), which is a solitary, phytophagous, and primitive hymenopteran insect species43. Although significant expression of the sawfly Mblk-1 homolog was detected in the head without the brain, thorax, and abdomen, the expression in the brain was below the detection threshold (crossing point Cp: 40) (Table S2). The relative expression levels of the Mblk-1 homolog did not differ significantly among body parts when the brain Cp values were set to 40 (Fig. 6D). These results indicate that the expression level of Mblk-1 homolog is very low in the sawfly brain, and that brain-specific expression, which is observed in the honey bee, is not conserved in the sawfly.
Comparative analysis of the upstream sequences of Mblk-1 homologs
Finally, we investigated how accumulation of the GAGA motif (GAGA and GAG) related to the expression of Mblk-1 homologs in the brains of hymenopteran insects. We counted the number of GAGA motifs located within 2 kb and 10 kb from the transcription start site of each Mblk-1 homolog (Fig. 6E and Fig. S8). In all 4 species of Aculeata, the number of GAG sequences within 2 kb upstream was greater than 100, and the number of GAGA sequences was greater than 30. On the other hand, in Athalia rosae, in which no Mblk-1 homolog expression was observed in adult brains, the number of GAG sequences was 41. In addition, to determine the evolutional stage in hymenopteran insects at which the accumulation of GAGA motifs was acquired, we investigated the numbers of GAGA motifs in the 2 parasitoid wasps, Nasonia vitripennis and Fopius arisanus. The numbers of GAGA motifs in Nasonia vitripennis were comparable to those in Aculeata, while the numbers in Fopius arisanus were closer to those in Athalia rosae. The same trend was observed in the numbers of GAGA motifs within upstream of 10 kb (Fig. S8).
Discussion
Mblk-1 targets genes related to synaptic plasticity and ecdysone-signaling in adult honey bee MBs
In the present study, we newly identified genes related to neural function, such as CaMKII and pumilio homolog 2 as Mblk-1 target gene candidates in the worker honey bee MBs by ChIP-seq analyses. CaMKII is involved in synaptic functions25,26,27. In addition, Drosophila melanogaster pumilio has a role in synaptic plasticity and learning ability50. Therefore, it is likely that Mblk-1 is involved in neural function, such as learning and memory abilities through transcriptional regulation of these synaptic plasticity-related genes.
Several ecdysone-related genes such as Usp were also identified as Mblk-1 target gene candidates. In forager bees (> 21 days), the expression of Usp is stronger in lKCs than in the inner compact KCs, which correspond to mKCs and sKCs51. It is likely that Usp is regulated differently between lKCs and inner compact KCs, and its expression in lKCs is controlled by Mblk-1. It is plausible that Usp has some neural functions that differ from its developmental functions like Mblk-1. This is consistent with reports that ecdysone and ecdysone-related genes are involved in neural functions, including learning and memory abilities, in Drosophila melanogaster52.
The honeybee Mblk-1 has three DNA binding motifs that are shared with DmE93 or BmE93
We identified three Mblk-1-binding motifs (GAGA, GAAATTTT, and AT(C)TTTGTA; Fig. 3A,B). The luciferase assay suggested that the GAGA and GAAATTTT motifs have different functions in transcriptional regulation by Mblk-1. GAGA and ATTTTGG motifs are suggested to be binding motifs for BmE93 and DmE93, respectively44,45, indicating that these two Mblk-1-binding motifs are conserved in insects. To the best of our knowledge, however, only AWTTTYGG and HTTTBGG, both of which are like ATTTTGG, are reported as DmE93-binding motifs in Fly Factor Survey and previous research using ChIP-seq analysis44. It is thus likely that Mblk-1 possesses a binding motif (the GAGA motif) that is not observed in DmE93. In the luciferase assay experiments, the expression of a reporter gene was upregulated when using the reporter vector containing the 6xGAGA sequence (Fig. 3D). Therefore, it is probable that Mblk-1 recognizes the GAGA motifs and upregulates the expression of genes around them in lKCs.
Distinct functions of Mblk-1 between pupal and adult brains
We identified pupal-specific Mblk-1 target gene candidates (Supplementary Data 2), which include some neural developmental genes. Approximately half of the target gene candidates for pupal brains were not identified as Mblk-1 target gene candidates for adult MBs (Fig. 5A), suggesting that Mblk-1 alters its role by changing its target genes as honey bees develop from pupae to adults. There are two possible explanations for this change. One is that target genes of Mblk-1 that functions in regions other than MBs, such as OLs and ALs where Mblk-1 is expressed in early pupal stage39, were included since we used pupal whole brains for the ChIP-seq analysis. The other possible explanation is that Mblk-1 regulates different genes when it is phosphorylated. Mblk-1 is phosphorylated by mitogen-activated protein kinase35 in a pupae-specific manner18. The anti-Mblk-1 antibody was designed to recognize both non-phosphorylated and phosphorylated Mblk-118. It is thus possible that Mblk-1 changes its binding regions and target genes depending on phosphorylation.
The GO enrichment analysis of adult-specific target gene candidates suggested GO terms related to development, such as generation of neurons (GO:0048699) and open tracheal system development (GO:0007424) (Fig. S5), although the genes used in the analysis were obtained from the ChIP-seq experiment using adult MBs. This finding indicates that Mblk-1 in adult MBs regulates the expression of genes involved in development as well as those related to synaptic plasticity. It is likely that the functions of these developmental genes in adult MBs are different from the ones in developing pupae. Similarly, in the GO enrichment analysis of the pupal-specific target gene candidates, GO terms related to development, such as compound eye development (GO:0048749) and urogenital system development (GO:0001655), were suggested (Fig. S6). Because the genes used in the analysis were obtained from the ChIP-seq experiment using pupal brains, it is reasonable to assume that these genes are involved in neural development. In conclusion from the GO enrichment analyses, many of the Mblk-1 target gene candidates are related to development as expected from their conventional role, but some have neural functions in adult MBs.
Self-regulation: a possible model for constitutive expression of Mblk-1 in adult honey bee MBs
Two distinct peak signals were found upstream of the Mblk-1 gene, and the GAGA motif (GAGA or GAG) is enriched in those signal regions. The results of the luciferase assay indicated that the DNA sequence corresponding to one of the peak signals was recognized by Mblk-1 and the expression of its downstream reporter gene was upregulated (Fig. 6B). Generally, in insect metamorphosis, E93 is transiently induced by ecdysone and ecdysone receptors20. On the other hand, Mblk-1 is constitutively expressed in honey bee lKCs18, whereas ecdysone receptor (EcR) is mainly expressed in sKCs in the honey bee MBs53. Therefore, together with the result of the luciferase assay, it is likely that the constitutive expression of Mblk-1 in lKCs is, at least partially, accounted for by self-regulation instead of ecdysone-dependent regulation. Mblk-1 expression seems to be initiated in the pupal stage, because its expression is observed in early pupae (P2), but not in the larval brain39. It is likely that Mblk-1 is first induced in the pupal brain under the control of ecdysone-EcR-Usp ternary complex, as suggested by earlier reports20,21. We hypothesize that Mblk-1, once activated, induces its own expression autonomously by self-regulation through the pupal to adult stages.
The above luciferase assay finding also suggests that the accumulation of the GAGA motif sequences upstream of Mblk-1 is important in the self-induction of Mblk-1. The numbers of the GAGA motif sequences in species in which the constitutive expression of Mblk-1 homologs in adult brains are observed (Apis mellifera and Harpegnathos saltator49) were much greater than those in Athalia rosae. Consistent with this, the qRT-PCR analysis using sawfly body parts indicated that the Mblk-1 homolog is not selectively expressed in the brains of adult sawflies (Fig. 6D). For Nasonia vitripennis, the GAGA motifs are accumulated upstream of the Mblk-1 homolog, although we have not analyzed whether their Mblk-1 homologs are preferentially expressed in the adult brain. If the GAGA motif accumulation and the constitutive expression of Mblk-1 homologs in brains has been acquired in parasitoid wasps, the role of Mblk-1 homologs related to neural functions, including learning and memory abilities, might have been a platform for the regulation of parasitic behaviors and prerequisite for further advanced behaviors of aculeate species.
Recently, Davie et al.54 reported the results of scRNA-seq using the adult Drosophila brain. When we investigated the expression of DmE93 using the data by Davie et al.54, we found that DmE93 was identified as a cluster marker in 25 among 87 clusters, and 3 of the former 25 clusters correspond to KCs. Therefore, unlike Mblk-1, DmE93 is expressed in broad brain regions, including the whole MBs, in the fly brain. In addition, from regulons suggested in the SCope, DmE93 is not likely to target CaMKII or pum in the fly brain. From the expression pattern of DmE93, it seems that the constitutive expression of E93 in MBs itself is not directly related to the MB function specific to the honey bee. Our present findings suggest that both the autonomous and preferential expression of Mblk-1 in lKCs and its altered binding-specificity are important for the function of Mblk-1 in lKCs in honey bees.
In summary, although Mblk-1/E93 is known as a transcription factor required for metamorphosis in insects21,22,23,24, the findings of the present study indicate that the roles of Mblk-1 have expanded to regulate synaptic plasticity. By acquiring new expression mechanisms, binding motifs, and target genes during the evolution of hymenopteran insects, Mblk-1 may contribute to neural development in pupae and the advanced learning and memory abilities in adults. Our study provides an example of molecular evolution that could contribute to the acquisition of novel MB function in insects. Similar mechanisms may be at work in the brains of animals in general, and our findings could contribute to a better understanding of the evolution of brains and behaviors.
Data availability
ChIP-seq data in this study were deposited in the GEO with the accession number GSE173409. The link for data access is as follows: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE173409. We declare that all data supporting the findings of this study are available within the article and its supplementary information files.
References
McGuire, S. E., Le, P. T. & Davis, R. L. The role of Drosophila mushroom body signaling in olfactory memory. Science 293, 1330–1333 (2001).
Mizunami, M., Weibrecht, J. M. & Strausfeld, N. J. Mushroom bodies of the cockroach: Their participation in place memory. J. Comp. Neurol. 402, 520–537 (1998).
Schildberger, K. Multimodal interneurons in the cricket brain: Properties of identified extrinsic mushroom body cells. J. Comp. Physiol. 154, 71–79 (1984).
Ben-Shahar, Y., Robichon, A., Sokolowski, M. B. & Robinson, G. E. Influence of gene action across different time scales on behavior. Science 296, 741–744 (2002).
Farris, S. M. & Schulmeister, S. Parasitoidism, not sociality, is associated with the evolution of elaborate mushroom bodies in the brains of hymenopteran insects. Proc. Biol. Sci. 278, 940–951 (2011).
Oya, S., Kohno, H., Kainoh, Y., Ono, M. & Kubo, T. Increased complexity of mushroom body Kenyon cell subtypes in the brain is associated with behavioral evolution in hymenopteran insects. Sci. Rep. 7, 13785 (2017).
Winston, M. L. The Biology of the Honeybee (Harvard University Press, 1987).
Seeley, T. D. The Wisdom of the Hive: The Social Physiology of the Honey Bee Colonies (Harvard University Press, 1995).
Von Frisch, K. The Dance Language and Orientation of Bees (Harvard University Press, 1993).
Menzel, R., Leboulle, G. & Eisenhardt, D. Small brains, bright minds. Cell 124, 237–239 (2006).
Giurfa, M. Behavioral and neural analysis of associative learning in the honeybee: A taste from the magic well. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 193, 801–824 (2007).
Kaneko, K., Suenami, S. & Kubo, T. Gene expression profiles and neural activities of Kenyon cell subtypes in the honeybee brain: Identification of novel ‘middle-type’ Kenyon cells. Zool. Lett. 2, 14 (2016).
Kamikouchi, A., Takeuchi, H., Sawata, M., Natori, S. & Kubo, T. Concentrated expression of Ca2+/calmodulin-dependent protein kinase II and protein kinase C in the mushroom bodies of the brain of the honeybee Apis mellifera L. J. Comp. Neurol. 417, 501–510 (2000).
Kamikouchi, A. et al. Preferential expression of the gene for a putative inositol 1,4,5-trisphosphate receptor homologue in the mushroom bodies of the brain of the worker honeybee Apis mellifera L. Biochem. Biophys. Res. Commun. 242, 181–186 (1998).
Neves, G., Cooke, S. F. & Bliss, T. V. Synaptic plasticity, memory and the hippocampus: A neural network approach to causality. Nat. Rev. Neurosci. 9, 65–75 (2008).
Takeuchi, H. et al. Identification of a novel gene, Mblk-1, that encodes a putative transcription factor expressed preferentially in the large-type Kenyon cells of the honeybee brain. Insect Mol. Biol. 10, 487–494 (2001).
Park, J. M., Kunieda, T., Takeuchi, H. & Kubo, T. DNA-binding properties of Mblk-1, a putative transcription factor from the honeybee. Biochem. Biophys. Res. Commun. 291, 23–28 (2002).
Kumagai, H. et al. Developmental stage-specific distribution and phosphorylation of Mblk-1, a transcription factor involved in ecdysteroid-signaling in the honey bee brain. Sci. Rep. 10, 8735 (2020).
Takayanagi-Kiya, S., Kiya, T., Kunieda, T. & Kubo, T. Mblk-1 transcription factor family: Its roles in various animals and regulation by NOL4 splice variants in mammals. Int. J. Mol. Sci. 18, 246 (2017).
Baehrecke, E. H. & Thummel, C. S. The Drosophila E93 gene from the 93F early puff displays stage- and tissue-specific regulation by 20-hydroxyecdysone. Dev. Biol. 171, 85–97 (1995).
Lee, C. Y. et al. E93 directs steroid-triggered programmed cell death in Drosophila. Mol. Cell 6, 433–443 (2000).
Lee, C. Y. & Baehrecke, E. H. Steroid regulation of autophagic programmed cell death during development. Development 128, 1443–1445 (2001).
Ureña, E., Manjón, C., Franch-Marro, X. & Martín, D. Transcription factor E93 specifies adult metamorphosis in hemimetabolous and holometabolous insects. Proc. Natl. Acad. Sci. U.S.A. 111, 7024–7029 (2014).
Pahl, M. C., Doyle, S. E. & Siegrist, S. E. E93 integrates neuroblast intrinsic state with developmental time to terminate MB neurogenesis via autophagy. Curr. Biol. 29, 750-762.e3 (2019).
Shonesy, B. C., Jalan-Sakrikar, N., Cavener, V. S. & Colbran, R. J. CaMKII: A molecular substrate for synaptic plasticity and memory. Prog. Mol. Biol. Transl. Sci. 122, 61–87 (2014).
Matsumoto, Y. et al. cGMP-gated channels, calmodulin, AC, and CamKII required for late, but not early, long-term memory formation in the honeybee. Learn. Mem. 21, 272–286 (2014).
Scholl, C., Kübert, N., Muenz, T. S. & Rössler, W. CaMKII knockdown affects both early and late phases of olfactory long-term memory in the honeybee. J. Exp. Biol. 218, 3788–3796 (2015).
McNeill, M. S., Kapheim, K. M., Brockmann, A., McGill, T. A. & Robinson, G. E. Brain regions and molecular pathways responding to food reward type and value in honey bees. Genes Brain Behav. 15, 305–317 (2016).
Wallberg, A. et al. A hybrid de novo genome assembly of the honeybee, Apis mellifera, with chromosome-length scaffolds. BMC Genom. 20, 275 (2019).
Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).
Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008).
Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).
Zhou, Y. et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat. Commun. 10, 1523 (2019).
Machanick, P. & Bailey, T. L. MEME-ChIP: Motif analysis of large DNA datasets. Bioinformatics 27, 1696–1697 (2011).
Park, J. M., Kunieda, T. & Kubo, T. The activity of Mblk-1, a mushroom body-selective transcription factor from the honeybee, is modulated by the ras/MAPK pathway. J. Biol. Chem. 278, 18689–18694 (2003).
Traniello, I. M. et al. Meta-analysis of honey bee neurogenomic response links Deformed wing virus type A to precocious behavioral maturation. Sci. Rep. 10, 3101 (2020).
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573-3587.e29 (2021).
McGinnis, C. S., Murrow, L. M. & Gartner, Z. J. DoubletFinder: Doublet detection in single-cell RNA sequencing data using artificial nearest neighbors. Cell Syst. 8, 329-337.e4 (2019).
Suenami, S. et al. Analysis of the differentiation of Kenyon cell subtypes using three mushroom body-preferential genes during metamorphosis in the honeybee (Apis mellifera L.). PLoS One 11, e0157841 (2016).
Shields, E. J., Sheng, L., Weiner, A. K., Garcia, B. A. & Bonasio, R. High-quality genome assemblies reveal long non-coding RNAs expressed in ant brains. Cell Rep. 23, 3078–3090 (2018).
Geib, S. M., Liang, G. H., Murphy, T. D. & Sim, S. B. Whole genome sequencing of the braconid parasitoid wasp Fopius arisanus, an important biocontrol agent of pest tepritid fruit flies. G3 (Bethesda) 7, 2407–2411 (2017).
Dalla Benetta, E. et al. Genome elimination mediated by gene expression from a selfish chromosome. Sci. Adv. 6, eaaz9808 (2020).
Oeyen, J. P. et al. Sawfly genomes reveal evolutionary acquisitions that fostered the mega-radiation of parasitoid and eusocial hymenoptera. Genome Biol. Evol. 12, 1099–1188 (2020).
Uyehara, C. M. et al. Hormone-dependent control of developmental timing through regulation of chromatin accessibility. Genes Dev. 31, 862–875 (2017).
Liu, X. et al. 20-Hydroxyecdysone (20E) primary response gene E93 modulates 20E signaling to promote bombyx larval-pupal metamorphosis. J. Biol. Chem. 290, 27370–27383 (2015).
Uno, Y., Fujiyuki, T., Morioka, M. & Kubo, T. Mushroom body-preferential expression of proteins/genes involved in endoplasmic reticulum Ca2+-transport in the worker honeybee (Apis mellifera L.) brain. Insect Mol. Biol. 22, 52–61 (2013).
Kaneko, K. et al. Novel middle-type Kenyon cells in the honeybee brain revealed by area-preferential gene expression analysis. PLoS One 8, e71732 (2013).
Takeuchi, H. et al. Prepro-tachykinin gene expression in the brain of the honeybee Apis mellifera. Cell Tissue Res. 316, 281–293 (2004).
Sheng, L. et al. Social reprogramming in ants induces longevity-associated glia remodeling. Sci. Adv. 6, eaba9869 (2020).
Dubnau, J. et al. The staufen/pumilio pathway is involved in Drosophila long-term memory. Curr. Biol. 13, 286–296 (2003).
Velarde, R. A., Robinson, G. E. & Fahrbach, S. E. Nuclear receptors of the honey bee: Annotation and expression in the adult brain. Insect Mol. Biol. 15, 583–595 (2006).
Ishimoto, H., Sakai, T. & Kitamoto, T. Ecdysone signaling regulates the formation of long-term courtship memory in adult Drosophila melanogaster. Proc. Natl. Acad. Sci. U.S.A. 106, 6381–6386 (2009).
Takeuchi, H., Paul, R. K., Matsuzaka, E. & Kubo, T. EcR-A expression in the brain and ovary of the honeybee (Apis mellifera L.). Zool. Sci. 24, 596–603 (2007).
Davie, K. et al. A single-cell transcriptome atlas of the aging Drosophila brain. Cell 174, 982–998.e20 (2018).
Acknowledgements
We are grateful to Mr. Takayoshi Kuwabara for his help in using sawflies.
Funding
This work was supported by Grant-in-Aid for Scientific Research(B) (JSPS KAKENHI Grant Number 20H03300) from Japan Society for the Promotion of Science.
Author information
Authors and Affiliations
Contributions
Y.M. designed and performed all experiments except for NGS preparation and sequencing. T.K.T. performed NGS preparation and sequencing. T.Kunieda designed and performed the identification of the Drosophila melanogaster orthologs for all Apis mellifera genes. All analyzed and discussed the data. Y.M. prepared the figures. Y.M., H.K. and T.Kubo drafted the manuscript and all wrote and approved the paper. T.Kakutani and T.Kubo supervised the study.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Matsumura, Y., To, T.K., Kunieda, T. et al. Mblk-1/E93, an ecdysone related-transcription factor, targets synaptic plasticity-related genes in the honey bee mushroom bodies. Sci Rep 12, 21367 (2022). https://doi.org/10.1038/s41598-022-23329-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-022-23329-z
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
