The European honey bee (Apis mellifera L.) is a model organism for investigating the molecular and neural bases underlying social behaviors and advanced brain functions (Kamikouchi et al., 1998; Kucharski et al., 1998; Toma et al., 2000; Ben-Shahar et al., 2002; Whitfield et al., 2009; Suenami et al., 2018). Mushroom bodies (MBs) are a higher-order center of memory, learning, and sensory integration of the insect brain (Hammer, 1993; Heisenberg, 1998; Menzel and Giurfa, 2001), and honey bee MBs have been used as a model to study adult neuronal plasticity (Withers et al., 1993; Fahrbach et al., 1998; Groh and Rössler, 2020). In the honey bee, MBs are composed of four subtypes of interneurons, termed Kenyon cells (KCs): Class I (large-, middle-, and small-type) KCs and Class II KCs, which are distinguished based on their size and localization of their somata in the MBs, and gene expression profiles (Figure 1) (Mobbs, 1982; Strausfeld, 2002; Fahrbach, 2006; Kaneko et al., 2013; Kaneko et al., 2016; Suenami et al., 2018). Large-type KCs have been suggested to function in learning and memory based on their gene expression profiles (Kamikouchi et al., 1998; Kamikouchi et al., 2000; Takeuchi et al., 2002; Uno et al., 2012; Suenami et al., 2018), whereas small- and a part of middle-type KCs have been suggested to be related to information processing during foraging behavior based on neural activity mapping using immediate early genes (IEGs) (Kiya et al., 2007; Ugajin et al., 2013).

One of the prominent molecular characteristics of honey bee MBs is that the genes involved in ecdysone signaling are expressed preferentially in the MBs of the adult honey bee brain (Truman et al., 1994; Baehrecke, 1996; Strausfeld, 2002; Kayukawa et al., 2017; Liu et al., 2018; Suenami et al., 2018). In holometabolous insects, metamorphosis is coordinated by two insect hormones, 20-hydroxyecdysone (20E) and juvenile hormone (JH) (Baehrecke and Thummel, 1995; Buszczak and Segraves, 2000; Lee et al., 2000; Lee and Baehrecke, 2001; Truman, 2019). Although 20E induces larval-larval molting in the presence of JH, it induces larval-pupal and pulpal-adult molts in the absence of JH. 20E binds to ecdysone receptor (EcR) and its co-factor, ultraspiracle (USP) and induces a group of genes encoding transcription factors that function downstream of EcR during metamorphosis: e.g., Mushroom body large-type Kenyon cell-preferential protein-1 (Mblk-1)/E93, BR-C, E74, and E75 (Takeuchi et al., 2001; Paul et al., 2005; Paul et al., 2006; Takeuchi et al., 2007). In the honey bee, the hemolymph titer of JH increases with the division of labor of workers from nurse bees to foragers and influences the timing of the behavioral transition of workers (Robinson et al., 1991; Jassim et al., 2000). In addition, both 20E and JH modulate the expression of genes involved in ecdysone signaling in the MBs of newly emerged workers (Velarde et al., 2009). On the other hand, ecdysone signaling regulates behavior and brain function even in adult Drosophila. For example, 20E and EcR are required for the formation of long-term courtship memory (Ishimoto et al., 2009). Overexpression or knockdown of E75 in clock neurons disrupts circadian rhythms (Kumar et al., 2014). Therefore, the reiterative use of ecdysone signaling in the adult brain is not specific to honey bees. However, honey bee MBs are characteristic in that the genes for ecdysone signaling-related transcription factors are expressed in both temporally and spatially regulated (that is, KC-subtype-preferential) manner (Figure 1). Thus, it is plausible that ecdysone signaling plays an important role in the functional specialization of each KC subtype and behavioral control.

Recent studies have reported that both Mblk-1 and EcR target unique genes in the adult honey bee brain compared to those during metamorphosis (Matsumura et al., 2022; Iino et al., 2023). Matsumura et al. (2022) reported that, Mblk-1 targets two genes involved in synaptic plasticity, which underlies learning and memory abilities in the animal brain, such as Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), besides many neural-related genes in the worker honey bee brain (Matsumura et al., 2022). On the other hand, Iino et al. (2023) reported that some EcR target genes are upregulated in the forager brain, some of which are implicated in the repression of metabolic processes (Iino et al., 2023). These findings suggest that ecdysone signaling-related transcription factors are reiteratively used to regulate brain function in the adult honey bee brain. This minireview summarizes the recent findings regarding the functions of ecdysone signaling-related transcription factors in the adult honey bee brain and discusses their possible roles in acquiring unique KC subtype functions during the evolution of nidifying Aculeata species including honey bees.

Among the genes for ecdysone-related transcription factors expressed in the MBs of the adult honey bee, Mblk-1 was originally identified as a gene that is expressed preferentially in the large-type KCs in the worker honey bee brain (Figure 1) (Takeuchi et al., 2001). Subsequent studies have shown that Mblk-1 encodes a transcription factor (Park et al., 2002; Park et al., 2003) and is constitutively and preferentially expressed in the large-type KCs in the worker brain from the pupal to adult stages (Suenami et al., 2016; Kumagai et al., 2020). Recently, Matsumura et al. (2022) conducted chromatin immunoprecipitaion-sequencing (ChIP-seq) analysis to search for candidate target genes for Mblk-1 using the MBs of adult worker brains (Matsumura et al., 2022). Among the many neural-related genes identified as Mblk-1 target gene candidates, two synaptic plasticity-related genes, CaMKII and pumilio homolog 2 (pum), were confirmed to be expressed preferentially in large-type KCs, just like Mblk-1, suggesting that Mblk-1 upregulates their expression levels in large-type KCs (Figure 2A) (Kamikouchi et al., 2000; Pasch et al., 2011; Kaneko et al., 2013). Mblk-1 binds to Mblk-1-binding elements located in introns of these synaptic plasticity-related target genes. CaMKII functions in synaptic plasticity, which is the molecular basis for learning and memory ability (Shonesy et al., 2014), and is required for long-term memory formation, even in honey bees (Matsumoto et al., 2014; Scholl et al., 2015). pum is involved in learning and memory via synaptic plasticity in Drosophila (Dubnau et al., 2003). Therefore, it is plausible that Mblk-1 functions in learning and memory via the transactivation of these synaptic plasticity-related genes in large-type KCs in the adult honey bee brain (Figure 2A).

In addition to neural-related genes, two ecdysone signaling-related genes, USP and E75, were identified as Mblk-1 candidate target genes in the brains of adult workers (Matsumura et al., 2022). However, the expression patterns of USP and E75 are different from that of Mblk-1; USP is expressed in the entire MBs in 1-day-old workers but it is expressed more strongly in large-type KCs in foragers than in 1-day-old workers (Velarde et al., 2009), and E75 is expressed in the entire MBs in the brain of workers (Figure 1) (Paul et al., 2006). Therefore, it is plausible that Mblk-1 contributes to the large-type KC preferential expression of USP in foragers, at least partly, and that transcription factors other than Mblk-1 also regulate its expression in the entire MBs in 1-day-old workers, as well as the expression of E75 in adult honey bee workers. Considering that the E75 and USP expression levels in the MBs of newly emerged workers are affected by exogenous JH treatment (Velarde et al., 2009), and considering that the Mblk-1 expression levels in the brain are higher in foragers than that in nurse bees (Liu et al., 2022), it might be that the expression levels of Mblk-1, and thus its downstream genes, are also regulated by JH signaling.

Matsumura et al. (2022) also compared profiles of Mblk-1 target gene candidates in pupal and adult worker brains (Matsumura et al., 2022). Approximately half of pupal Mblk-1 target gene candidates, including some developmental genes, such as Dscam, Ubx, and Ror are specifically detected in pupal brains but not in adult worker brains. These findings suggest that Mblk-1 alters its target genes in the brain between the pupal and adult stages (Figure 2A).

Liu et al. (2022) recently reported that Mblk-1 expressed in the worker brain is related to sugar responsiveness in foragers via the regulation of the expression of a gustatory receptor gene (AmGR1) (Liu et al., 2022). However, it seems unclear at present whether Mblk-1 expression in large-type KCs is directly related to the regulation of sugar responsiveness and whether Mblk-1 directly targets AmGR1. For example, the expression of Mblk-1 in brain regions apart from MBs, such as suboesophageal zone, a primary gustatory center, might underlie this functionality.

In adult honey bees, EcR is expressed in the ovaries of queens, a reproductive caste, and involved in oogenesis but not in the abdomens of workers, a non-reproductive caste (Takeuchi et al., 2007; Uno et al., 2012). However, EcR is also expressed in the adult brains, especially strongly in small-type KCs in the MBs, of both castes (Takeuchi et al., 2007), suggesting that EcR is related not only to oogenesis but also to brain functions in adult honey bees. Velarde et al. (2009) reported that the EcR expression levels in MBs are higher in newly emerged workers than in foragers (Figure 2B) (Velarde et al., 2009). Notably, the increase in the volume of MB calyces, where KCs extend their dendrites, takes place in newly emerged workers in an experience-independent manner (Fahrbach et al., 1998), suggesting that the high EcR expression levels might be related to this enlargement. Singe et al. (2018) and Iino et al. (2022) reported that EcR is induced in the forager brain by foraging flight (Figure 2B) (Singh et al., 2018; Iino et al., 2020). Therefore, it is plausible that the elevated EcR expression levels are also associated with the foraging experience-dependent increase in the volume of MB calyces (Withers et al., 1995; Farris et al., 2001). So far, small-type KCs have been suggested to function in sensory information processing and/or cognitive processes associated with foraging such as navigation, orientation, and learning and memory because kakusei, HR38, and Egr, all of which are IEGs (Yamazaki et al., 2006; Kiya et al., 2007; Fujita et al., 2013; Ugajin et al., 2013), are induced in small-type KCs during foraging flight in worker honey bees (Yamazaki et al., 2006; Kiya et al., 2007; Kaneko et al., 2013; Ugajin et al., 2013). Therefore, it is possible that EcR also functions in the information processing and/or the regulation of physiological states during foraging in adult honey bee brains.

Recently, Iino et al. (2023) conducted ChIP-seq analysis using the whole brains of nurse bees and foragers to investigate the function of EcR in the adult honey bee brain (Iino et al., 2023). Most EcR target gene candidates are common to both nurse bees and foragers and include genes for canonical ecdysone signaling, such as Hr4, E75, and Kr-h1, whereas several genes known to be targets of EcR, including Mblk-1, are not detected. GO enrichment analysis of EcR target gene candidates also detected functions related to canonical ecdysone signaling, such as post-embryonic development, oogenesis, and neurogenesis. However, genes related to the regulation of metabolism were also enriched, suggesting a novel function of EcR in adult honey bee brains. Additionally, RNA-seq analysis revealed that several EcR target gene candidates detected in the brains of foragers, including those for ecdysone signaling-related transcription factors, such as E75, E78C, and Myc, are upregulated in the brains of foragers after foraging compared with those of nurse bees and foragers before foraging (Iino et al., 2023). Some of these genes are involved in the regulation of lipid levels and metabolism in Drosophila and mice (Inagaki et al., 2009; Su and Peng, 2020; Praggastis et al., 2021). Since foragers require a large amount of energy for the vibration of flight muscles and brain neural activity during foraging flight (Stabentheiner and Kovac, 2016; Rittschof et al., 2018), by regulating the transcription of the identified target metabolism genes, honey bee EcR may reduce energy consumption in the brain and/or generate the energy needed for foraging behaviors through lipolysis. Since EcR expression is also detected in brain regions other than MBs (Iino et al., 2020), the function of EcR in the regulation of metabolism suggested by Iino et al. (2023) may be related not only to small-type KCs but also to the other brain regions (Iino et al., 2023). It is noteworthy that E75 was detected as a target gene candidate not only for Mblk-1 but also for EcR, suggesting that E75 expression is differentially regulated in each KC subtype in the adult honey bee brain (Matsumura et al., 2022; Iino et al., 2023). This may partly explain the previous finding that E75 is expressed in all KC subtypes (Paul et al., 2006). Finally, the target genes and thus the functions of EcR induced in the MBs of newly emerged workers in a foraging experience-independent manner remain unknown.

In the adult honey bee MB, the expression patterns of EcR and its conventional downstream gene Mblk-1 are different. EcR is expressed preferentially in small-type KCs, whereas Mblk-1 is expressed preferentially in large-type KCs (Figure 1). Matsumura et al. (2022) reported that Mblk-1-binding regions containing GA-rich sequences are located upstream of Mblk-1 and that Mblk-1 can transactivate a reporter gene through one of these Mblk-1 binding regions in vitro (Matsumura et al., 2022). In contrast, Mblk-1 has not been identified as a candidate EcR target gene in the adult honey bee brain (Iino et al., 2023). Therefore, the constitutive expression of Mblk-1 in large-type KCs may be accounted for by its self-regulation, independent of ecdysone and EcR, although the initial large-type KC preferential expression of Mblk-1 may be induced by ecdysone and EcR at the pupal stage (Figure 2A).

Recent studies shed light on the possible evolution of KC subtypes in the MBs in Hymenoptera. The number of KC subtypes has been proposed to have increased from one in solitary sawflies, basal hymenopteran species (Symphyta), to two in parasitic wasps, and then to three in nidifying Aculeata species associated with behavioral evolution in Hymenoptera (Oya et al., 2017). Recently, Kuwabara et al. (2023) conducted a comparative transcriptome analysis to propose a model for evolutionary dynamics, in which the number of subtypes increased from one ancestral, multifunctional KC type through functional segregation and divergence (specialization) (Kuwabara et al., 2023). They also suggested that CaMKII expression and its functions in long-term memory are located in the entire MBs in the turnip sawfly, which are composed of a single ancestral-like KC subtype, whereas they are localized to large-type KCs in honey bees. Considering that Mblk-1 expression is restricted to some KC populations in the MBs of the brains of two ant species, another Aculeate species (Sheng et al., 2020; Li et al., 2022), whereas it is not detected in the brain of turnip sawflies (Matsumura et al., 2022), large-type KC preferential Mblk-1 expression in adult brains has likely been acquired in Aculeata (Matsumura et al., 2022). Therefore, Mblk-1 may have contributed, at least in part, to the functional specialization of large-type KCs, such as advanced learning and memory abilities, by elevating CaMKII expression and upregulating other synaptic plasticity-related genes in Aculeata MBs (Figure 2C). Regarding EcR, small-type KC preferential expression and upregulation after foraging flight in the adult brain are also observed in the bumble bee Bombus ignitus, a close honey bee relative (Iino et al., 2020), suggesting a conserved function of EcR in the adult brains at least in Apidae. EcR may have contributed to the functional segregation or specialization of small-type KCs by regulating downstream gene expression, although the EcR expression patterns in the adult brains of sawflies and other Aculeata species need to be clarified in the future.

Finally, it is necessary to substantiate whether the genes identified as target gene candidates for Mblk-1 and EcR (Matsumura et al., 2022; Iino et al., 2023) are up- or down-regulated by these transcription factors, for example, by examining changes in expression levels of these target gene candidates in the adult worker bee brain in which Mblk-1/EcR is knocked down in the future research.

In this Minireview, we mainly focused on the possible functions of Mblk-1 and EcR in worker honey bee brains. Some questions have arisen regarding the mechanisms and functions of the induction of these ecdysone signaling-related transcription factors in the brains of adult honey bees and their relationship to the acquisition of behavioral traits characteristic of Aculeata species, including honey bees.

The first is the relationship between the transcriptional regulation by these ecdysone-related transcription factors and the hemolymph 20E titer in adult worker honey bees. Considering that the hemolymph ecdysteroid titer in workers transiently rises on 3-day after the emergence (Hartfelder et al., 2002; Amdam et al., 2004) and the EcR expression levels are affected by exogenous 20E treatment in newly emerged workers (Velarde et al., 2009), EcR might be activated directly by hemolymph 20E at least in newly emerged workers. However, Yamazaki et al. (2011) suggested that ecdysone is synthesized in the brains of adult workers (Yamazaki et al., 2011). Therefore, EcR might also be activated directly by 20E synthesized in the worker brain. In contrast, Matsumura et al. (2022) proposed that Mblk-1 may be induced autonomously by self-regulation independent of ecdysone (Matsumura et al., 2022). This could be testified by examining whether the knockdown of genes involved in the ecdysteroid synthesis alters the Mblk-1/E93 expression levels.

Then, why are ecdysone signaling-related transcription factors reiteratively used in the worker honey bee brain, among many other transcription factors? Mblk-1/E93 acts as an “adult specifier” and BR-C as a “pupal specifier” in the holometabolous insect (Zhou and Riddiford, 2002; Konopova and Jindra, 2008; Ureña et al., 2014). EcR, Mblk-1, and BR-C, all of which are key factors in metamorphosis, are preferentially expressed in honey bee MBs. These ecdysone-related transcription factors regulate the expression of many downstream genes in a tissue-dependent manner during metamorphosis (Lee et al., 2000; Liu et al., 2015; Uyehara and McKay, 2019). Therefore, the reiterative use of ecdysone-related transcription factors might have caused drastic changes in gene expression and contributed to conferring the unique characteristic to each KC subtype during the evolution of the hymenopteran insect.

Finally, what are the possible roles of ecdysone signaling in the acquisition of behavioral traits of honey bees? Both learning and memory abilities, in which Mblk-1 is proposed to be involved, and the metabolic regulation in brains, in which EcR is proposed to be involved, seem to be behavioral and physiological traits especially important for nidifying Aculeate species, including honey bees (Figure 2C). After foraging, adult Aculeata species need to return to their nest to feed their brood (Capaldi et al., 2000; Goulson and Stout, 2001; Gathmann and Tscharntke, 2002; Klein et al., 2004; Woodgate et al., 2016); therefore, they are assumed to have excellent foraging ability and highly advanced learning and memory abilities (Pyke, 1978; Dyer, 1996; Menzel and Giurfa, 2001; Hendricsma et al., 2019). The reiterative use of ecdysone-related transcription factors may have contributed to increasing the functional complexities of MBs by conferring unique gene expression profiles to each KC subtype, leading to the acquisition of behavioral and physiological regulatory mechanisms specific to Aculeata, including honey bees (Figure 2C). It seems that the analysis of the functional diversification of ecdysone-related transcription factors in hymenopteran species remains to be an attractive subject in the search for molecular mechanisms underlying the honey bee behaviors and the driving force behind their evolution. Several other transcription factors are involved in ecdysone signaling that are reiteratively used in the adult honey bee brain besides Mblk-1 and EcR. Therefore, by comparing the expression patterns of these transcription factors and identifying their target genes, it may be possible to test the hypothesis that ecdysone signaling reiteratively used in the adult honey bee brain may be involved in behavioral regulation and the acquisition of brain function specific to Aculeata, including honey bees.

All authors contributed to the article and approved the submitted version. YM and HK created the figures.