Control of the insect metamorphic transition by ecdysteroid production and secretion

Abstract

Ecdysteroids are a class of steroid hormones that controls molting and metamorphic transitions in Ecdysozoan species including insects, in which ecdysteroid biosynthesis and its regulation have been extensively studied. Insect ecdysteroids are produced from dietary sterols by a series of reduction-oxidation reactions in the prothoracic gland and in Drosophila they are released into the hemolymph via vesicle-mediated secretion at the time of metamorphosis. To initiate precisely controlled ecdysteroid pulses, the prothoracic gland functions as a central node integrating both intrinsic and extrinsic signals to control ecdysteroid biosynthesis and secretion. In this review, we outline recent progress in the characterization of ecdysone biosynthesis and steroid trafficking pathways and the discoveries of novel factors regulating prothoracic gland function.

Introduction

Steroid hormones, a group of systemic signaling molecules that are produced from cholesterol and thus share the steroid backbone, comprise an important class of messengers and exert a myriad of physiological functions in metazoans. A well-known example is to promote sexual maturation of animals. In mammals, sex hormones such as estradiol and testosterone circulate and act in multiple organs to promote the transition of a juvenile individual to a sexually reproductive adult, a process known as puberty [1]. Another well-characterized example exists in insects, where ecdysteroids control metamorphic transitions. In both mammals and holometabolous insects, the sexual maturation process is controlled by a neuroendocrine pathway in which a steroid hormone functions as the final output signal. In mammals, the pathway operates within the hypothalamic-pituitary-gonadal (HPG) axis [1] and utilizes a cascade of neuropeptide signals that begin with Kisspeptin stimulating hypothalamic release of gonadotropin-releasing hormone (GnRH) which in turn promotes systemic release of luteinizing hormone (LH) and follicle stimulating hormone (FSH) from the pituitary gland to increase gonadal production of the sex steroids [1,2]. These steroids bind to nuclear hormone receptors to form various types of transcriptional regulatory complexes that execute the developmental programs associated with sexual maturation and fertility.

In insects, a pair of neurons in each brain lobe produce prothoracicotropic hormone (PTTH), which stimulates ecdysone production in the prothoracic gland (PG), a major endocrine organ of the larva. A recent study in the fruit fly Drosophila melanogaster demonstrated that the PTTH producing neurons can be activated by a presynaptic Allastatin A (AstA) signal [3 **]. Surprisingly, AstA and the AstA receptor exhibit high levels of homology to the mammalian Kisspeptin and cognate receptor GPR54 which activate the HPG axis to initiate puberty [2]. Ecdysone also binds to a related nuclear hormone receptor (EcR/Usp) further highlighting the conserved aspects of the mechanism that underlies the juvenile to adult transition in both insects and man.

In insects, the effect of ecdysteroid on metamorphosis control has been studied in a plethora of species [4*]. In the pre-genomic/genetic era, Lepidopteran species such as tobacco hornworm Manduca sexta and silkworm Bombyx mori were commonly used as model organisms due to their large size and rapid life cycle which facilitated physiological and biochemical studies. More recently, Drosophila has become a predominant model system for these inquiries because of its unparalleled genetic toolkit [5]. This review will briefly summarize the current understanding of ecdysteroid synthesis and its regulation. First, we will outline the ecdysteroid biosynthetic pathway. Then, we will discuss recent progress on the signal inputs and intracellular regulatory mechanisms controlling ecdysone synthesis. Finally, we will highlight newly emerging evidence for vesicle mediated ecdysone secretion.

The ecdysteroid biosynthetic pathway: glimmer in the Black Box

Insects utilize dietary sterols as precursors for ecdysteroid biosynthesis [6]. Dietary cholesterol, found in laboratory food, is converted to ecdysone (E) through a series of oxygen additions to the steroid ring. Plant and yeast sterols are likewise converted to makisterone A and 24(28)-dehydromakisterone A, respectively, which differ from E at carbon 24. While all three ecdysteroids can support development, cholesterol is the preferential ecdysteroid precursor [6,7].

Despite some variations across insect species, the core ecdysteroid biosynthetic pathway appears to be very similar, as demonstrated by conservation of the principle biosynthetic enzymes and reaction intermediates (Fig. 1) [8]. Briefly, dehydrogenation of cholesterol at carbons 7 and 8 by the Reiske oxygenase Neverland forms 7-dehydrocholesterol (7dC) [9,10]. Subsequently, 7dC is modified by the “Black box” reactions, so named because the intermediates and enzymes responsible remain unclear. The output from the black box is the intermediate 5β-ketodiol which is then subject to sequential hydroxylation on carbons 25, 22 and 2 to produce ecdysone [11–14]. After secretion from the PG and import into peripheral tissues, E is hydroxylated on carbon 20 by Shade to produce the active ecdysteroid 20-hydroxyecdysone (20E) [15]. In some species, such as Manduca, the pathway is slightly different in that the first post-Black Box compound is a 5β-diketol and the major PG-secreted ecdysteroid is 3-dehydroecdysone (3dE). 3dE is then processed into E in hemolymph [8].

Figure 1.

The ecdysteroid biosynthetic pathway (see text for details). Light yellow circles mark the modification sites in each reaction. The Black Box reactions are marked by light grey background. Red question marks indicate the catalyzing enzymes are unknown. “A” and “B” presents the two alternative ways in which Δ⁴-diketol is transformed into 5β-ketodiol. The post-Black Box reactions in Drosophila and Manduca are marked by light green and light orange background, respectively.

To date, the Black Box reactions have not been fully elucidated. However, recent observations have provided important insights. Key intermediates such as 3-oxo-steroids and Δ⁴-diketol have been confirmed in the Black Box [16–18]. The results unambiguously show that 7dC is first oxidized at carbon 3 to form 3-oxo-7dC, and the unstable 3-oxo-7dC is then isomerized into the more stable 3-oxo-Δ^4,7C. Feeding Δ⁴-diketol rescues neverland mutants, suggesting 3-oxo-Δ^4,7C is converted to Δ⁴-diketol by hydroxylation of carbon 14 and oxidation of carbon 6, though the intermediates remain a mystery [17,18]. The Δ⁴-diketol is in turn reduced at carbon 5 and carbon 3. Rescue experiments suggest the order of these reactions may be flexible [17,18]. In Drosophila, the resulting 5β-ketodiol is the final Black Box product. In Manduca, the carbon 3 reduction is initially skipped, and 5β-diketol is subjected to the terminal hydroxylations to form 3dE.

Several enzymes have been identified and associated with the Black Box reactions including Non-molting glossy/Shroud [19], Spooky/Spookier/Spookiest [20–22], Cyp6t3 [23] and Cyp6u1 [24] although the exact reactions catalyzed by each remain to be determined. Downstream of 5β ketodiol, the enzymes encoded by phantom [11,12], disembodied [13] shadow [14] and shade [15], all cytochrome P450 monooxygenases, act sequentially to add OH groups to the 25, 22, 2 and 20 carbons, respectively, to produce the final active 20-E hormone. These genes have been historically classified as “Halloween genes” due to the characteristically “empty, and ghost-like” embryonic cuticles formed by zygotic loss-of-function mutants [25]. As in mammals, steroids are also important for fertility, and follicle cell produced 20E has been shown to be required for both oogenesis and ovulation [26]. Curiously. the putative Black box enzyme Spookier substitutes for Spook during 20E production in the follicle cells [21].

Extracellular signals: A network of ever-increasing complexity

As the major endocrine organ that produces ecdysone, the PG functions as a central node to integrate diverse physiological and environmental signals and converts them into E pulses that trigger molting and the metamorphic transition (Fig. 2). In this section we focus on the extracellular input signals that regulate ecdysteroidogenesis in the PG.

Figure 2.

Extracellular signals regulating ecdysteroid synthesis in the PG (see text for details).

Prothoracicotropic hormone (PTTH)

The existence of a brain-derived ecdysteroidogenic factor was proposed almost a century ago [27]. The first neuropeptide identified in the brain was prothoraciotropic hormone (PTTH), whose activity has been demonstrated in many insect species [28]. PTTH is produced by specific neuroendocrine cells and reaches the PG either via the hemolymph (Manduca) [29] or through direct neural innervation (Drosophila) [30]. Once at the PG, PTTH binds with the receptor tyrosine kinase (RTK) Torso and activates the Ras/Raf/Erk pathway [31]. How Ras/Raf/Erk regulates ecdysone biosynthesis is not fully understood, but several transcription factors including hormone receptor 4 (Hr4) and Pointed (Ptn) are involved in the pathway [23,32**].

Despite the demonstration that purified PTTH can stimulate PG glands to make ecdysone [33], recent studies using ptth mutants indicate that it is not strictly required for metamorphosis. In Bombyx, many ptth null mutant larvae manage to metamorphose at either the L4 or L5 stages [34*]. In Drosophila, ablation of PTTH producing neurons severely delays metamorphosis [30], but does not eliminate it, and ptth null mutants cause only a modest change in metamorphic timing [35*]. Recent studies have found three additional RTK receptors including epidermal growth factor receptor (Egfr), anaplastic lymphoma kinase (Alk), and PDGF- and VEGF-receptor related (Pvr) to be expressed in the PG and able to regulate E synthesis via the Ras/Raf/Erk pathway [32**,36]. Interestingly, the Alk ligand Jelly belly and the Pvr ligand Pvf3 are expressed in the PTTH producing neurons [36]. This may explain some of the phenotypic differences between the ptth null mutant and neuron ablation. In contrast, the EGFR ligands Spitz and Vein are expressed in the PG itself and act in an autocrine manner, which may explain why even PTTH neuron-ablated larvae are still able to undergo metamorphosis (see below) [32**].

Insulin/insulin-like growth factors

Another RTK mediated signal is produced by the insulin/insulin-like (IIS) growth factor ligand family. Systemically, the IIS pathway functions in various organs to couple nutrition with the overall growth and development of the animal [37]. In the PG, the insulin receptor (InR) and its downstream signal transduction components PI3K/Akt and target of rapamycin (TOR) are indispensable for PG tissue growth and ecdysone synthesis in both Drosophila and Bombyx [38–42]. However, the observations in Manduca appear controversial. Feeding TOR inhibitor Rapamycin to Manduca compromises PG growth, reduces ecdysone synthesis and delays metamorphosis [43,44]. However, treatment with PI3K inhibitors LY294002 and wortmannin does not attenuate E production [45]. Species differences are also evident upon injection of the insulin-like peptide Bombyxin [46], which triggers ecdysone synthesis and secretion in Bombyx [39,47] but not in Manduca [45,48]. Further investigation is needed to elucidate the differences in mechanisms that underlie these observations.

Neuropeptides and serotonin

Other than RTKs, G-protein coupled receptors (GPCRs) have also been proposed to mediate ecdysteroidogenic signals, since typical second messengers such as Ca²⁺ and cAMP are crucial factors regulating ecdysteroid synthesis in Lepidoptera [28]. In Bombyx, a series of neuropeptides have been shown to exert either prothoracicostatic or prothoracicotropic effects through GPCR receptors in the PG. The prothoracicostatic factors include prothoracicostatic peptides (PTSPs), Bommo-myosuppressin (BMS) and FMRFamide-related peptide (BRFa) , while the prothoracicotropic molecules include FXPRL-amide peptide, Orcokinin and pigment dispersing factor (PDF) [49,50]. Some of these neuropeptides are released into hemolymph, while others such as BRFa and Orcokinin are delivered to the PG by direct neural innervation [49]. Despite the intense study of Drosophila neuropeptides [51], neuropeptide F is the only confirmed factor (other than PTTH and Ilps) shown to regulate ecdysteroidogenesis in the Drosophila PG [52]. Besides peptidergic neurons, a subset of serotoninergic neurons also regulate PG function in Drosophila [53].

Juvenile hormone and ecdysone

Juvenile hormone (JH) and ecdysone are the two key hormones that together coordinate molting and metamorphic activities. The systemic regulation of ecdysone synthesis by these two hormones involves multiple organs and complicated inter-organ communication, which cannot be fully discussed here (for review, see [54]). Instead, we will briefly introduce how the PG directly responds to each hormone.

JH is the best-known anti-metamorphic hormone which prevents larvae/nymphs from undergoing precocious metamorphosis. Accordingly, it is often referred to as the status quo hormone (for review, see [55] and this issue of Curr Opin). Although early evidence reported that JH suppressed ecdysteroidogenesis in Bombyx PGs [56], the direct regulation of PG cell function by JH was not verified until the JH receptors Methoprene-tolerant (Met)/Germ cell-expressed (Gce) were identified in Drosophila and subsequently in other insect models [55,57]. In Drosophila, PG specific knockdown of Met/gce causes precocious ecdysone production and accelerated pupariation, clearly demonstrating that the JH signal directly antagonizes ecdysteroid synthesis in the PG [58].

20E itself is also thought to exhibit both positive and negative feedback on its own production. Such a mechanism helps create a hormone pulse wherein a small amount of 20E stimulates its own synthesis through an EcR/Usp mediated positive feedback loop. At a certain hormone level, a negative feedback loop kicks in to bring the ecdysone level back to baseline thereby helping create a pulse. Additional mechanisms such as steroid hormone inactivation [59] also contribute to hormone removal and pulse control. Positive and negative feedback control in steroid hormone level modulation has been noted in both Lepidoptera and Drosophila [54,60] , and Drosophila genetic analysis has implicated both the ecdysone receptor (EcR) and the early responsive transcription factor Broad (Br) [60] in feedback control. Moreover, it appears that two isoforms of Br, Br-Z1 and Br-Z4, differentially mediate positive and negative feedback, respectively [60].

Transforming growth factor β

The TGFβ family is comprised of two branches of factors, the bone morphogenetic proteins (BMPs) and the Activins. Both pathways have been studied in Drosophila PGs and function in opposite ways. Loss of dSmad2/Smox, the primary transcriptional transducer of the Activin pathway, compromises E synthesis and causes a severe metamorphic delay. Although the identity and the source of the ligand that activates the pathway is not yet clear [61], Activin signaling appears to act as a competence pathway since it is required for normal expression of Torso and InR [61]. In contrast, BMP signaling functions as a gatekeeper to suppress precocious metamorphosis. The BMP ligand Decapentaplegic (Dpp) is released from imaginal discs during early L3 stage, which results in phosphorylation of Mad by the Thickvein (Tkv)-Punt/Wishful thinking (Wit) receptors in the PG to suppresses ecdysone synthesis. When larvae grow beyond a certain size, Dpp “leakage/release” from imaginal discs terminates, allowing resumption of hormone synthesis by the PG [62**].

Hedgehog

Hedgehog (Hh) signaling is a highly conserved pathway that controls embryonic patterning and adult tissue homeostasis from Drosophila to mammals [63]. As a canonical morphogen and paracrine factor, secreted Hh diffuses within tissues such as imaginal discs to regulate differentiation and patterning. However, recent studies also identified Hh circulating in the hemolymph of Drosophila larvae, which it acts in an endocrine manner to regulate larval body growth and developmental timing [64]. The source of the circulating Hh is the midgut, and it signals to both the PG and fat body. The expression and secretion of Hh in the midgut is upregulated during starvation. Activation of Hh signaling in the PG suppresses expression of Halloween genes and ecdysone biosynthesis, while in fat body it promotes neutral lipid mobilization during starvation. These activities enable Hh to play an important role during starvation by delaying pupariation and mobilizing nutrient stores to support the survival of animals. Under fed conditions, however, the combinatory effects of Hh on slowing down both body growth and developmental timing results in normal sized pupae irrespective of gain or loss of circulating Hh [64].

Autocrine factors

The existence of uncharacterized autocrine factor(s) regulating ecdysteroid synthesis was first observed in Bombyx [65,66]. Recently, several autocrine factors have been identified in Drosophila, including β3-octopamine [67], the Egfr ligands Vein, and Spitz [32] and the Pvr ligands Pvf2, Pvf3 [36]. The β3-octopamine-induced pathway functions upstream of both IIS and PTTH signaling [67], while the Egfr and Pvr ligands activate Ras/Raf/Erk directly [32**,36]. Intriguingly, the expression of Vein and Spitz in the PG are induced by E feeding [32**], indicating a role of Egfr signaling as part of the ecdysone synthesis positive feedback loop described above.

Cell-autonomous factors: power of the -omics

In response to input signals, the PG cells undergo many intracellular modifications that help stimulate ecdysteroid production. Recent studies have uncovered an increasing number of intracellular factors that are essential to PG function, including transcription factors controlling Halloween gene expression [68,69*], iron metabolism regulators required for production of biosynthetic enzyme cofactors such as heme [70], cholesterol trafficking regulators [71,72], circadian cycle factors [73], glutathione [72,74], and nitric oxide (NO) [75] production. Among these factors, most were identified in Drosophila by PG-specific gene knockdown experiments. At the genome-wide level, PG-knockdown of an astonishing 1,906 genes (out of 12,504 total) cause various levels of developmental defects [71]. Although many of these “ecdysteroidogenic” genes have known or speculated functions relevant to hormone production, or are required for maintenance of basic cell functions, there are many whose potential role(s) in ecdysone synthesis and trafficking remain to be investigated.

In addition to traditional genetic screening, advancements in transcriptome and proteome profiling techniques and bioinformatics also provide powerful insights into the biological activities of the PG. To date, at least four RNA-seq [24,76–78], one microarray [76] and one proteomic dataset [78] have been obtained using the Drosophila ring gland or Bombyx PG tissues. Moreover, the samples have been prepared from tissues subject to different sorts of manipulations (e.g. early vs. late stage larvae [76], pre- vs. post-PTTH treatment [76,78], providing a better coverage of different biological scenarios. Analyses of these datasets have uncovered membrane receptor distributions [78], cholesterol trafficking components [71,79] and cytochrome P450 enzyme profiles [24,76] in the PG. Further mining of these data may lead to additional discoveries concerning regulation of E synthesis in the PG.

Steroid trafficking: an old question yields a new perspective

In addition to the enzymology, steroid hormone biosynthesis also appears to utilize complex trafficking mechanisms to move steroid molecules into, out of, and within cells. As strict cholesterol auxotrophs, insects fully rely on dietary sterol for viability. Upon ingestion, cholesterol is transported to the PG in lipophorins (insect lipoprotein) [80,81] and imported into PG cells using scavenger [82] or lipophorin receptors [83]. Subsequently, cholesterol translocates into the ER through the endo-lysosome system, during which cholesterol binding proteins such as Neimann-Pick complex 1 (NPC1), NPC2 and Start1 play essential roles [84]. Additional cholesterol trafficking factors influencing hormone production have been described in recent reports, including the fatty acid elongase Sit, the glutathione S-transferase Nopperabo [72,85], the deadenylase CCR4-NOT complex [86], and the autophagic machinery [87*,88*], but the mechanisms involved are not fully understood. During E biosynthesis the first step involving conversion of cholesterol to 7dC occurs in the ER. From there, the Black Box reactions continue to live up to their name with respect to the subcellular compartments in which they act. The microsomal localization of Spok indicates that they likely occur in the ER [21]. The Black Box product is further hydroxylated into 2,22-dideoxyecdysone in the ER and then into ecdysone within mitochondria [8]. Thus, ecdysone biosynthesis appears to involve shuttling of steroid intermediates between ER and mitochondria perhaps at mitochondrial associated membrane contact sites (MAMs). However, the mechanism and potential steroid shuttling proteins involved in this process are still unknown as are inferred hemolymphatic steroid carrier proteins [89].

Until recently, the discussion of steroid trafficking pathways stopped at the secretion point. It was, and still is, widely assumed that steroid hormones can freely diffuse across the plasma membrane due to the molecules’ inherently lipophilic nature. Recently, this idea was challenged by the demonstration that in Drosophila, E exits PG cells via vesicle-mediated secretion [90]. In this scenario, in the presence of ATP, ecdysone is pumped into secretory vesicles by Atet, a specific member of the ABC transporter family. At the time of ecdysone secretion, the vesicles are released into the hemolymph in response to a Ca²⁺ signal [90]. Following this discovery, an ecdysone importer (EcI) was also identified, which is expressed in peripheral tissues and mediates the entry of E into responding cells from the hemolymph [91**]. These discoveries demonstrate that ecdysone secretion is distinct from its synthesis, and that both secretion and uptake could potentially serve as novel regulatory nodes. Whether similar exporters and importers are employed by mammals for proper steroid trafficking and tissue response remains to be determined.

Future outlook

As efforts devoted to understanding the biology of ecdysteroids and their regulation during metamorphosis continue, many issues (re)emerge as the result of continuously accumulating knowledge. Old mysteries such as the chemistry and enzymatic steps of the Black Box remain to be fully elucidated, and new inquiries concerning the mechanisms of steroid trafficking, ecdysteroid secretion and its uptake are needed. As the number of signals regulating PG function rise, the cross talk between these signals and the mechanism(s) for integrating them into an appropriate temporal E production profile will require significant additional intellectual and experimental inquires. Studies utilizing other insect orders and species, besides Drosophila, are also essential to understand both the similarities and differences in how E production and release are regulated relative to the ecology of each species.

Figure 3.

Steroid trafficking during ecdysteroidogenesis in PG cells (see text for details). Dashed line indicates that the steroid trafficking process between the two cellular compartments are not fully understood.