MicroRNAs regulate the sesquiterpenoid hormonal pathway in Drosophila and other arthropods

Abstract

Arthropods comprise the majority of all described animal species, and understanding their evolution is a central question in biology. Their developmental processes are under the precise control of distinct hormonal regulators, including the sesquiterpenoids juvenile hormone (JH) and methyl farnesoate. The control of the synthesis and mode of action of these hormones played important roles in the evolution of arthropods and their adaptation to diverse habitats. However, the precise roles of non-coding RNAs, such as microRNAs (miRNAs), controlling arthropod hormonal pathways are unknown. Here, we investigated the miRNA regulation of the expression of the juvenile hormone acid methyltransferase gene (JHAMT), which encodes a rate-determining sesquiterpenoid biosynthetic enzyme. Loss of function of the miRNA bantam in the fly Drosophila melanogaster increased JHAMT expression, while overexpression of the bantam repressed JHAMT expression and resulted in pupal lethality. The male genital organs of the pupae were malformed, and exogenous sesquiterpenoid application partially rescued the genital deformities. The role of the bantam in the regulation of sesquiterpenoid biosynthesis was validated by transcriptomic, qPCR and hormone titre (JHB3 and JH III) analyses. In addition, we found a conserved set of miRNAs that interacted with JHAMT, and the sesquiterpenoid receptor methoprene-tolerant (Met) in different arthropod lineages, including insects (fly, mosquito and beetle), crustaceans (water flea and shrimp), myriapod (centipede) and chelicerate (horseshoe crab). This suggests that these miRNAs might have conserved roles in the post-transcriptional regulation of genes in sesquiterpenoid pathways across the Panarthropoda. Some of the identified lineage-specific miRNAs are potential targets for the development of new strategies in aquaculture and agricultural pest control.

1. Introduction

Fluctuating titres of sesquiterpenoids and 20-hydroxyecdysone (20E) orchestrate the timing of organism-wide developmental transitions. The controls of the levels of these hormones were probably critical to the success of some groups of arthropods during evolution [1–4]. Animals produce either steroid or sesquiterpenoid hormones by way of the mevalonate/isoprene biosynthetic pathway, which diverges into two branches after the formation of farnesyl diphosphate. Arthropods have lost the cholesterol-synthetic branch present in vertebrates; but instead, an alternate pathway has led to the biosynthesis of sesquiterpenoids, including farnesoic acid (FA), methyl farnesoate (MF) and juvenile hormone (JH) [5,6]. In some groups of arthropods, sesquiterpenoids are involved in the control of metamorphic modes, moulting, reproduction, behaviours and morphogenesis [1]. Although the regulatory roles of sesquiterpenoids have only been demonstrated in insects and crustaceans, an analysis of various arthropod genomes established the conservation of enzymes for biosynthesis and degradation of MF/JH, as well as proteins that are part of their signalling pathways [7]. Among these conserved genes are JH acid methyltransferase (JHAMT), a rate-limiting enzyme in MF and JH biosynthesis [8,9], as well as the MF and JH receptor, methoprene-tolerant (Met), a member of the basic Helix-Loop-Helix Per-Arnt-Sim (bHLH-PAS) family of transcription factors [10,11]. MicroRNAs (miRNAs) have recently been identified as regulators of downstream factors in JH signalling in insects; a 20E-induced miRNA, miR-1890, controls the expression of a JH-induced serine protease in the mosquito midgut [12], whereas miR-2 facilitates metamorphosis in cockroaches by downregulating the JH-dependent transcription factor Krüppel homolog 1 (Kr-h1) [13]. By contrast, miRNAs have not been implicated in the regulation of genes from the sesquiterpenoid biosynthetic pathways. Hence, this study aimed to systematically examine the role of miRNAs in the regulation of key genes in the sesquiterpenoid biosynthetic and signalling pathways (JHAMT and Met) across different arthropods, and shed light on their roles in hormonal regulation and evolution of this successful phylum.

2. Material and methods

(a). Prediction of microRNA candidates targeting juvenile hormone acid methyltransferase and methoprene-tolerant/germ cell-expressed

MiRNAs of Anopheles gambiae, Drosophila melanogaster, Tribolium castaneum and Daphnia pulex were obtained from miRBase [14]. Neocaridina denticulata [15], Strigamia maritima [16] and Tachypleus tridentatus [17] miRNAs were identified from their genomes using BlastN [18], and later confirmed by using Centroidfold [19]. These miRNAs were used to predict their interactions with the 3′untranslated regions (UTRs) of JHAMT and methoprene-tolerant/germ cell-expressed (Met/Gce) using miRanda and RNAhybrid with at least 6-mer of seed complementarity [20,21].

(b). Luciferase reporter construction, mutagenesis, dual luciferase reporter assays and labelled microRNA pull-down assays

Genomic DNA was used to amplify miRNA stem-loops with 100–300 bp sequences flanking each side of the genes from seven selected arthropods: An. gambiae [22], Dr. melanogaster (flybase.org), Tr. castaneum (beetlebase.org), Da. pulex [23], N. denticulata [15], S. maritima [16] and Ta. tridentatus [17]. The 3′UTRs of the JHAMT and Met/Gce genes of An. gambiae, Dr. melanogaster, Tr. castaneum and Da. pulex were amplified from their respective genomic DNA (primer information is shown in the electronic supplementary material, table S3). In the three species where JHAMT and Met sequences have not been validated by transcriptome studies (N. denticulata, S. maritima and Ta. tridentatus), RNA was extracted using Trizol (Invitrogen), and 3′RACE was completed (3'-Full RACE Core Set, Takara). Sequencing was carried out to reveal their 3′UTRs. Amplicons of miRNAs and 3′UTRs were subcloned into pAC5.1 (Invitrogen) and psicheck-2 (Promega) vectors, respectively. Mutations of the predicted miRNA-binding sites (electronic supplementary material, S1, mutated sites are in bold red) were carried out using a QuikChange II site-directed mutagenesis kit (Stratagene) (primer information is shown in the electronic supplementary material, table S3). All constructs were sequenced to confirm their identities. Drosophila S2 cells (DRSC) were kept at 23°C in Schneider Drosophila medium (Life Technologies) with 10% (v/v) heat-inactivated fetal bovine serum (Gibco, Life Technologies) and 1 : 100 penicillin-streptomycin (Gibco, Life Technologies). The psicheck-2-sensor (100 ng) and pAC5.1-miRNA (200 ng) were co-transfected into Drosophila S2 cells using Effectene (Qiagen). Luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega), and a Tecan Infinite M200 luminometer. Assays were performed at 48 h post-transfection, and repeated in at least three biological replicates with Renilla : firefly luciferase activity ratios averaged over three technical replicates each [24]. The Renilla : firefly luciferase activity ratios were normalized using as control S2 cells that were transfected with the respective psicheck-2-sensor alone. Statistical significances were assessed with t-tests. For the labelled miRNA pull-down (LAMP) assays [25], RNA extracts of Dr. melanogaster 3rd instar larvae were pulled down using biotin-labelled mature miRNA DNA oligos and Dynabeads^®M-280 Streptavidin (Invitrogen™), followed by RT-PCR to confirm the interactions of JHAMT mRNA and individual miRNAs.

(c). Fly cultures, mutants construction, TaqMan™ microRNA assays and hormonal rescue

To prepare the ‘gain of function' (GOF) construct of the Dr. melanogaster miRNA bantam, the corresponding stem-loop with a 500 bp flanking sequence was amplified and cloned into the GAL4-inducible vector pUAST (primer information is provided in the electronic supplementary material, table S3) [26]. The construct was sequenced prior to its injection into Dr. melanogaster w¹¹¹⁸ embryos. Flies were screened and crossed to generate stable homozygous transformants. A bantam deficient line, as well as various GAL4 drivers, including GAL4-Aug21, GAL4-Act5C, GAL4-Kr-h1 and GAL4-ptc, were obtained from the Bloomington Stock Center. The UAS-JHAMT lines were kindly provided by Ryusuke Niwa and Tetsuro Shinoda [27]. All flies were maintained on standard yeast-cornmeal-agar medium at 25°C. Males and virgin females from each fly line were randomly collected for crossings. For each crossing of GAL4 and UAS fly lines, three random males and three random virgin females were used, and reciprocal crosses were carried out. At least three separate crossings were performed for each GAL4 and UAS pair. To assess the expression of bantam in GOF and loss of function (LOF) mutant flies, TaqMan™ miRNA assays (Applied Biosystems™) were completed on an ABI 7500 PCR instrument, and expression levels were normalized relative to U6 titres. Hormone treatments (rescue experiments) of GAL4-ptc>bantam mutants involved topical application of 10⁻⁵ M juvenile hormone III (JH-III, Sigma) or MF (Echelon) dissolved in ethanol, and directly applied to a mixture of first, second and third instar flies.

(d). Transcriptome sequencing and real-time PCR

Two biological replicates of RNA were extracted from pupariating 3rd instar larvae of GAL4-ptc>bantam, GAL4-ptc>JHAMT and GAL4-ptc (control) lines. Libraries were constructed using the TruSeq stranded mRNA LT sample prep kit, and sequenced on an Illumina HiSeq2500 platform (Macrogen Inc). Raw reads were filtered using Trimmomatic [28] and mapped to Flybase v. 6.14 using TopHat and Cufflinks [29]. Differential gene expressions were evaluated using Cuffdiff [29]. RNAs from the respective crosses were reverse-transcribed into cDNA using the iScript™ cDNA synthesis Kit (BioRad). Real-time PCR reactions were conducted in three biological replicates using the CFX96 Touch™ Real-Time PCR Detection System (BioRad), with a cycle of denaturation at 95°C for 3 min followed by 40 cycles of 95°C/10 s, 55°C/10 s and 72°C/15 s. PCRs were run with half iTaq^TM Universal SYBR^® Green Supermix (BioRad) and 0.2 µM of each primer pair (primer information is listed in the electronic supplementary material, table S3).

(e). Hormonal titre measurements

JH III was purchased from Toronto Research Chemicals (Toronto, ON, Canada). Citronellol, HPLC-grade n-hexane and acetonitrile were purchased from Sigma-Aldrich (St. Louis, MO). The hormonal titre of JHB3 and JH III in the pupariating 3rd instar larvae of GAL4-ptc>bantam and GAL4-ptc (control) were measured by adapting a protocol from a previous study [30]. In brief, sixty 3rd instar larvae from each group were collected, washed with distilled water and placed in stainless steel grinding jars containing balls of Retsch MM400. The grinding jars were chilled in liquid nitrogen for 30 min, followed by homogenization with a frequency of 20 Hz for 30 s. The homogenates were then immediately transferred to 10 ml glass centrifuge tubes containing 200 µl acetonitrile, 200 µl 0.9% sodium chloride solution and 20 ng of citronellol as an internal standard, and subjected to ultrasonic treatment for 1 min, vortexed and further extracted twice with 200 µl hexane. The n-hexane phases (upper layer) were removed and transferred to new glass vials. The amounts of JHB3 and JH III were determined by gas chromatography tandem mass spectrometry as previously described [30].

3. Results and discussion

(a). The microRNA bantam targets juvenile hormone acid methyltransferase and regulates sesquiterpenoid pathways in Drosophila melanogaster

(i). Bantam interacts with juvenile hormone acid methyltransferase in vitro

The software miRanda [21] and RNAhybrid [20] were used for the in silico prediction of miRNAs that could potentially bind to the JHAMT 3′UTR of Dr. melanogaster (electronic supplementary material, dataset). A dual-luciferase reporter assay was used to validate the predicted interactions between miRNAs and the 3′UTR of the fly JHAMT [24]. In brief, vectors constitutively expressing miRNA candidate precursors were co-transfected into Drosophila S2 cells, along with vectors that express luciferase fused to either the JHAMT native 3′UTRs or 3′UTRs that had predicted binding sites mutated [24]. MiRNAs bantam, miR-252 and miR-304 repressed luciferase activity specifically through interactions with the native JHAMT 3′UTR-binding sites (electronic supplementary material, figure S1A), while repression of luciferase activity was abolished when the 3′ UTR-binding sites for each of these three miRNAs were mutated (figure 1a). To confirm the interactions between these three miRNAs and JHAMT, a LAMP assay was performed [25]. RNAs extracted from Dr. melanogaster 3rd instar larvae were mixed with biotinylated labelled oligonucleotides coding bantam, miR-252 and miR-304 mature sequences, and the pull-down of JHAMT transcripts was revealed by RT-PCR (figure 1b). These data clearly validated that bantam, miR-252 and miR-304 can bind the JHAMT mRNA, even in the presence of other native transcripts in cell extracts.

Figure 1.

MiRNAs interactions with JHAMT transcripts in Dr. melanogaster triggered JH-dependent phenotypes. (a) Renilla : firefly luciferase activity in Drosophila S2 cells transfected with the constructs expressing miRNAs and psi-check-2-sensors containing either a wild-type JHAMT 3′UTRs (WT, white bars) or a binding-site mutated JHAMT 3′UTR (Mut, grey bars). Values represent the mean ± s.e.m., *p < 0.05, **p < 0.005, ***p < 0.001. (b) Schematic diagram showing the miRNA pull-down assay. The gel analysis showed that bantam, miR-252 and miR-304 were bound to JHAMT mRNAs from Dr. melanogaster larvae extracts (M: DNA marker; lanes 1–5: bantam, miR-252, miR-304, positive control, negative control). (c–h) Misexpression of JHAMT and bantam resulted in different phenotypes. GAL4-ptc>JHAMT mutants were viable but sterile, and showed misorientation of male genitalia, while GAL4-ptc>bantam mutants were lethal at the pupal stage, and the dead pupae showed misorientation of male genitalia. Male genitalia of wild-type w1118 (c), ptc>JHAMT (e) and ptc>bantam (g); (d,f,h), magnification of the genital part of (c,e,g), respectively; arrows indicate: anal plate (AP), clasper (CL), lateral plate (LP) and penis (PE).

(ii). Bantam interacts with juvenile hormone acid methyltransferase and regulates sesquiterpenoids in vivo

To assess whether these miRNAs could interact with JHAMT in vivo, we generated transgenic Dr. melanogaster lines that created GOF of the desired miRNAs, and evaluated their effects on JH-dependent phenotypes. We used GAL4-Aug21 lines to construct GOF mutants that expressed miRNAs specifically in the corpora allata (CA). GAL4-Aug21>bantam flies were lethal at the pupal stage (electronic supplementary material, figure S3 and table S1), while no obvious phenotypes were observed for GAL4-Aug21>miR-252 and GAL4-Aug21>miR-304 mutants. Although the timing of death varied, pupal lethality was also observed when bantam was overexpressed using GAL4 drivers with more ubiquitous expression, such as ptc, Act5C or Kr-h1 promoters; while GAL4-Aug21>bantam resulted in dead pharate adults, GAL4-Kr-h1>bantam led to death during the prepupal or early pupal stages (electronic supplementary material, table S1 and figure S3). Moreover, examination of dead pupae revealed male genital rotation defects in GAL4-ptc>bantam flies (figure 1g,h). A previous study reported that ubiquitous overexpression of JHAMT, driven by the Act5C promoter, led to pupal death and male genital organ malformation [27]. Therefore, we generated GAL4-ptc>JHAMT mutants with ubiquitous overexpression of JHAMT, and observed that they could indeed develop into adults, but had misoriented male genital organs and were sterile (figure 1e,f). These experiments clearly confirmed the GAL4-ptc>bantam and GAL4-ptc>JHAMT mutants exhibited different phenotypes, but both mutants resulted in male genital defects. To address the issue of cell autonomic effects of bantam, and demonstrate that bantam can specifically downregulate JHAMT expression in the ring gland, we used qPCR to assess the expression of JHAMT in samples of ‘brains' (brain complex + ring glands) of 3rd instar larvae before settlement. Relative JHAMT expression in ‘brains' of GAL4-ptc>bantam was significantly lower than in ‘brains' of w1118 controls and GAL4-ptc>JHAMT flies (electronic supplementary material, figure S4A). In addition, specific in vivo interactions between bantam and JHAMT were validated using the dual-luciferase assays, which revealed that the luciferase activity in S2 cells, which endogenously express bantam, transfected with a wild-type JHAMT 3′UTR was significantly lower than those transfected with JHAMT 3′UTR with a mutated bantam-binding site (electronic supplementary material, figure S4b). Furthermore, the specific interactions between bantam and bantam-binding site regions of the JHAMT 3′UTR were validated when using a pull-down assay employing biotin-labelled five different regions of JHAMT 3′UTR; only the two regions containing bantam-binding sites were able to pull down bantam (electronic supplementary material, figure S4c,d).

Remarkably, during larva–pupal transitions in Drosophila, either the absence or excess of JH results in abnormal phenotypes, including pupa lethality [31,32]. Complete genetic ablation of the CA using GAL4-Aug21>Grim results in JH deficiency and pupal lethality [33]. Surprisingly, JHAMT null mutants are viable and have normal levels of MF [34]; because the CA of Drosophila synthesizes and releases three sesquiterpenoids: MF, juvenile hormone III bisepoxide (JHB3) and JH III; only the simultaneous decrease in titres of the three sesquiterpenoids results in pupal lethality [34]. In addition, Drosophila has two JH receptors, Met and Gce. Interestingly, Abdul et al. [35] observed identical phenotypes in null mutants of Met and Gce larvae, and proposed that Met and Gce are paralogous genes which code for proteins that heterodimerize to regulate the cross-talk between the 20E and JH signalling pathways. Met-Gce double mutants die during the larval–pupal transition, exhibiting lethal and defective phenotypes similar to JH-deficient animals [35,36]. The coexistence of three JHs and two JH receptors complicates the analysis of Drosophila JH pathways. In our studies, in vitro validations and phenotypic changes suggest that bantam might modulate sesquiterpenoid synthesis via JHAMT regulation; therefore, additional experiments were completed to test this hypothesis. To assess the role of bantam in regulation of sesquiterpenoid levels in Drosophila, we measured the titres of JHB3 and JH III on the whole body of L3 of the GOF flies (GAL4-ptc>bantam), having both reduced JHAMT and JH titres. In control groups (GAL4-ptc), JHB3 and JH III titres were 17.74 ± 7.74 fmol larva⁻¹ (10.61 ± 1.64 pmol g⁻¹) and 3.41 ± 1.53 fmol larva⁻¹ (2.04 ± 0.92 pmol g⁻¹), respectively, while the whole-body titres of JHB3 and JH III in GAL4-ptc>bantam mutants decreased by approximately 37% and 76%, respectively (figure 2a–d). The complete absence of expression of bantam in a bantam-LOF deficiency mutant line was confirmed by miRNA taqman assays (figure 2e). To confirm bantam targets, the expression levels of two critical genes involved in JH biosynthesis or action (JHAMT and Gce) were studied in both L3 larvae of bantam-GOF and bantam-LOF mutants. JHAMT was significantly downregulated in the GOF line (GAL4-ptc>bantam) and upregulated in the bantam-LOF flies, confirming the in vivo regulation of JHAMT by bantam (figure 2f). In addition, the JH receptor Gce had a similar opposite misexpression in both mutants (figure 2g). The premise that bantam-GOF phenotypes are mediated by a decrease in JH titres was reinforced by experiments where topical application of JH III or MF to GAL4-ptc>bantam larvae partially rescued the genital defect phenotypes (figure 2h). On the other hand, our results exposed intriguing antagonistic roles of bantam in repressing JH biosynthesis and activating JH signalling by modulating the receptor gene expression.

Figure 2.

MiRNA bantam modulates JH pathway genes expression in Dr. melanogaster. (a) JHB3 titre per wet mass, (b) JHB3 titre per larva, (c) JH III titre per wet mass and (d) JH III titre per larva. (e) Expression of bantam in L3 of bantam-GOF (ptc>bantam) and bantam-LOF mutants. (f,g) Relative gene expression of JHAMT and Gce in L3 of bantam and JHAMT GOF mutants. Values represent mean ± s.e.m., *p < 0.05, **p < 0.005, ***p < 0.001. The statistical significances were assessed with t-tests. (h) Application of JH III and MF on GAL4-ptc>bantam larvae partially rescued the male genital defects. n, number of observed flies.

(iii). Bantam modulates ecdysteroid pathways

To characterize the global gene expression changes triggered by bantam dysregulations, we compared the transcriptomes of two biological replicates of RNA extracted from L3 larvae of the GOF lines GAL4-ptc>bantam and GAL4-ptc>JHAMT, as well as a control line (GAL4-ptc). The expression of genes involved in sesquiterpenoid biosynthesis, degradation and signalling, as well as genes from additional regulatory pathways were compared (electronic supplementary material, table S2). In GAL4-ptc>JHAMT flies, global gene expression was similar to controls (less than two-fold changes), except for the upregulation of JHAMT and juvenile hormone esterase duplication (JHEDUP). In GAL4-ptc>bantam, several genes involved in sesquiterpenoid pathways were differentially expressed (upregulated: JHEDUP and Gce; downregulated: juvenile hormone diol kinase and allatostatin C receptor1 (electronic supplementary material, table S2). Considering that topical application of JH III and MF could only partially rescue male genital defects, and knowing that sesquiterpenoids and ecdysteroids ‘cross-talk' to regulate Drosophila physiology, the expression levels of genes involved in ecdysteroid synthesis and signalling were also investigated. In GAL4-ptc>bantam mutants, we detected the downregulation of several ecdysteroid biosynthetic pathway genes, including Neverland (Nvd), CYP302A1 and CYP314A1, as well as critical ecdysteroid signalling genes such as ecdysone-induced protein 75B (electronic supplementary material, table S2). Validating experiments using qPCR confirmed that Nvd, CYP306A1, CYP302A1, CYP314A1 and Broad (Br-C) were significantly repressed in GAL4-ptc>bantam mutants (figure 3a–d,f), but not the ecdysone receptor (EcR) (figure 3e). Our results confirmed previous reports describing that bantam suppresses ecdysteroid production in Drosophila [37]. These experiments also suggest that the lethal phenotypes observed in GAL4-ptc>bantam flies could also be mediated by modulation of ecdysteroid pathways genes, while misexpression of ecdysteroid pathway genes might not play a major role in the pupal lethality observed in GAL4-ptc>JHAMT mutants.

Figure 3.

MiRNA bantam regulates ecdysteroids and left/right asymmetry pathway genes in Dr. melanogaster. (a–h) Real-time PCR validation of differentially expressed genes identified in the transcriptome analyses (electronic supplementary material, table S2). Relative gene expression of L3 larvae of (a) Nvd, (b) CYP306A1, (c) CYP302A1, (d) CYP314A1, (e) EcR, (f) Br-C, (g) Abd-B, and (h) MyoID. Values represent mean ± s.e.m., *p < 0.05, **p < 0.005, ***p < 0.001. The statistical significances were assessed with t-tests.

Overexpression of either bantam or JHAMT resulted in male genital rotation defects in the left/right (L/R) asymmetry. Previous studies have demonstrated that the Hox gene Abdominal-B (Abd-B) and MyosinID (MyoID) play important roles in Drosophila L/R asymmetry determination [38,39]. Transcriptomic and qPCR data revealed that Abd-B was significantly downregulated, while MyoID was significantly upregulated in GAL4-ptc>bantam flies, but not in GAL4-ptc>JHAMT flies (figure 3g,h). In summary, the modulation of expression profiles observed in sesquiterpenoid pathway genes, together with previous reports describing the effects of ectopic JH treatment on pupae [40], suggest that the genital misorientation phenotype of JHAMT overexpression mutants might be solely a consequence of JH modulation. On the other hand, the modulation of expression profiles of both sesquiterpenoid and ecdysteroid pathway genes, the misorientation phenotype of male genital organs observed in GAL4-ptc>bantam mutants, and reports on ecdysone and JH interacting to control genitalia rotation [41], suggest that bantam-resulting phenotypes are the consequence of a modulation of both ecdysteroid and sesquiterpenoid pathways.

(b). MicroRNAs targeting juvenile hormone acid methyltransferase and methoprene-tolerant in other arthropods

Although the roles of sesquiterpenoids controlling development and reproduction have been demonstrated only in insects and crustaceans, JHAMT and Met have orthologues in other arthropod groups [7]. We therefore explored whether bantam interactions with JHAMT were also conserved in other arthropods. The software miRanda and RNAhybrid were used to predict all miRNAs that could potentially bind to JHAMT and Met in six arthropod genomes (electronic supplementary material, dataset). The predicted interactions were validated using the dual-luciferase reporter assay previously described. Functional binding sites for miRNAs on 3′UTRs of JHAMT and Met were functionally tested in two additional insects (the mosquito An. gambiae and the beetle Tr. castaneum), two crustaceans (the water flea Da. pulex and the shrimp N. denticulata), a myriapod (the centipede S. maritima) and a chelicerate (the horseshoe crab Ta. tridentatus) (figure 4; electronic supplementary material, figures S1 and S2). In An. gambiae, miR-278 bound to the JHAMT 3′UTR and repressed luciferase expression (figure 4a). The beetle Tr. castaneum JHAMT, as described in Dr. melanogaster, presented functional binding sites for bantam, miR-252 and miR-304; and in addition, interacted with let-7 and miR-92b (figure 4b). In crustaceans, bantam, miR-92 and miR-252 interacted with the JHAMT 3′UTRs of Da. pulex and N. denticulata (figure 4c,d); whereas in the centipede S. maritima, let-7, miR-34, miR-252 and miR-278 repressed luciferase activity by targeting JHAMT 3′UTRs (figure 4e–g). In the chelicerate Ta. tridentatus, bantam, let-7, miR-34, miR-92 and miR-278 interacted with the JHAMT 3′UTR (figure 4h). All these results suggest that a conserved set of miRNAs could modulate JHAMT expression in arthropods.

Figure 4.

Dual luciferase reporter assays validating miRNA regulation of JHAMT and Met/Gce in different arthropods. The relative Renilla : firefly luciferase activity in Drosophila S2 cells was measured for wild-type (WT, white bars) and mutant (Mut, grey bars) 3′UTRs of (a–h) JHAMT and (i–o) Met/Gce. Error bars represent mean ± s.e.m., *p < 0.05, **p < 0.005, ***p < 0.001. (p) Conserved set of miRNAs regulating the JHAMT and Met/Gce in different arthropods. An enlarged figure is included in the electronic supplementary material, figure S5. ‘n.a' indicates that none of the miRNAs with putative regulatory functions on Met/Gce in other arthropods interacted with Met in the centipede.

Binding sites for orthologues of the sesquiterpenoid receptors (Met and Gce) were predicted and functionally tested in the same seven arthropods. In Dr. melanogaster, let-7, miR-8, miR-14, miR-34, miR-278 and miR-304 decreased luciferase activity by binding to the 3′UTR-binding sites of Gce mRNA (figure 4i). The roles of miR-8, miR-14, miR-34 and miR-278 in regulating Met orthologues were conserved in An. gambiae (figure 4j). In Da. pulex, Met was regulated by bantam and miR-278 (figure 4l), while conserved miR-8, miR-34 and miR-278 functional binding sites were identified on N. denticulata Met 3′UTR (figure 4m). Horseshoe crabs have a duplicated genome [17], and we identified two Met orthologues in Ta. tridentatus. While Met-1 3′UTR sequence was regulated by bantam, let-7, miR-8, miR-34 and miR-252; Met-2 3′UTR only revealed binding sites for interactions with bantam and let-7 (figure 4n,o). The 3′UTR of the beetle Tr. castaneum Met appears to be unique, interacting solely with miR-92b (figure 4k). Several miRNAs interacted with 3′UTRs of both JHAMT and Met. The chelicerate Ta. tridentatus had three miRNAs that were able to bind to both JHAMT and Met (bantam, let-7 and miR-34) (figure 4h,n,o). Other arthropods presented lineage-specific miRNA with the ability to bind transcripts of JHAMT and Met (figure 4p). Our studies could not detect any 3′UTR- miRNA interaction with Met in the centipede S. maritima (electronic supplementary material, figure S2g).

Some of these miRNAs have been previously described as modulators of other hormonal pathways or physiological processes, including miR-14 regulating ecdysteroid signalling and insulin production in Drosophila [42,43], miR-34 modulating Drosophila lifespan, innate immunity and ecdysteroid signalling [44,45], and let-7 controlling metamorphosis of silkworm and cockroach [46,47]. Some of the newly discovered miRNA target interactions described here will require further investigation in vivo to reveal their mechanistic details. Nevertheless, the set of conserved miRNAs that we have identified provides an important starting point for future analyses. We have identified a basic set of miRNAs that are conserved, with dynamic gains and losses in multiple arthropod lineages. This situation mirrors previous studies [48–51], and the precise reconstruction of the complete ancestral set of miRNAs interacting with genes of sesquiterpenoid pathways will be improved with the analysis of new sequenced arthropod genomes.

Hormonal pathways are major targets for the development of new approaches for agricultural pest management, disease vector control and improvements in aquaculture productivity. The study of lineage-specific miRNAs that regulate sesquiterpenoid pathways could provide leads for identifying targets for designing new, specific and affordable strategies suitable for arthropod control.

Data accessibility

The datasets supporting the conclusions of this article are included within the article and its additional files.

Authors' contribution

J.H.L.H. conceived and supervised the study. Z.Q., W.N., Z.-P.K., Y.-Y.Z. and A.C.K. carried out the experimental works. K.W.S. obtained the RNA from centipedes. Z.Q., W.G.B., W.N., F.G.N., Z.-P.K., M.A., S.S.T. and J.H.L.H. analysed the data. Z.Q., W.G.B., W.N., K.W.S., F.G.N., Z.-P.K., Y.-Y.Z., H.Y.E.C., T.F.C., K.H.C., H.M.L., M.A., S.S.T. and J.H.L.H. wrote and approved the manuscript.

Competing interests

The authors have no conflict of interests to declare.

Funding

Z.Q. was supported by a PhD studentship from The Chinese University of Hong Kong. This research was supported by the Hong Kong Research Grant Council ECS grant 24100015 and a CUHK Faculty of Science Strategic Development Scheme (J.H.L.H.). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.