Identification of 20-Hydroxyecdysone Late-Response Genes in the Chitin Biosynthesis Pathway

Background 20-hydroxyecdysone (20E) and its receptor complex ecdysone receptor (EcR) and ultraspiracle (USP) play a crucial role in controlling development, metamorphosis, reproduction and diapause. The ligand-receptor complex 20E-EcR/USP directly activates a small set of early-response genes and a much larger set of late-response genes. However, ecdysone-responsive genes have not been previously characterized in the context of insect chitin biosynthesis. Principal Findings Here, we show that injection-based RNA interference (RNAi) directed towards a common region of the two isoforms of SeEcR in a lepidopteron insect Spodoptera exigua was effective, with phenotypes including a high mortality prior to pupation and developmental defects. After gene specific RNAi, chitin contents in the cuticle of an abnormal larva significantly decreased. The expression levels of five genes in the chitin biosynthesis pathway, SeTre-1, SeG6PI, SeUAP, SeCHSA and SeCHSB, were significantly reduced, while there was no difference in the expression of SeTre-2 prior to 72 hr after injection of EcR dsRNA. Meanwhile, injection of 20E in vivo induced the expression of the five genes mentioned above. Moreover, the SeTre-1, SeG6PI, SeUAP and SeCHSB genes showed late responses to the hormone and the induction of SeTre-1, SeG6PI, SeUAP and SeCHSB genes by 20E were able to be inhibited by the protein synthesis inhibitor cycloheximide in vitro indicating these genes are 20E late-response genes. Conclusions We conclude that SeTre-1, SeG6PI, SeUAP and SeCHSB in the chitin biosynthesis pathway are 20E late-response genes and 20E and its specific receptors plays a key role in the regulation of chitin biosynthesis via inducing their expression.


Introduction
Throughout the insect life cycle, the steroid hormone 20hydroxyecdysone (20E) coordinates multiple developmental events by eliciting a complex genetic program via a heterodimeric nuclear receptor composed of the ecdysone receptor (EcR) and ultraspiracle proteins (USP). Evidenced in a study of fruit fly Drosophila melanogaster by Ashburner et al. first revealed that a part of this developmental program consists of a genetic cascade in which the ligand-receptor complex 20E-EcR/USP directly activates the expression of a very small number of early-response genes. The products of the early-response genes in turn trigger the expression of a much larger set of late-response genes, so called secondaryresponse genes [1,2,3]. During the last decade, the EcR, USP and a large number of ecdysone-responsive genes from the two gene categories proposed in the Ashburner model have been characterized in D. melanogaster and several other insect species [4,5,6]. All of those studies published over the past decade have provided powerful evidence in support of the Ashburner model, meantime taken beyond that model which presented us with diverting new directions for our understanding of ecdysone signaling [7,8,9,10].
Most of the previous studies are extensively emphasize DNA puffs induced by ecdysone, whereas a few specific DNA puff genes are studied for a better understanding of the mechanisms underlying the ecdysone response on molecular level. The roles and functions of many early-response genes (eg. E74, E75 and broad-complex) have been well studied [11,12], while much less is known about the late-response genes (L63, L71 and L82) in insects. L63 has homology to the cyclin dependent kinase protein family and is required for Drosophila development [13]; L71 genes encode a set of polypeptides and provide an antimicrobial defense during metamorphosis [14]. L82 mutations displayed developmental delay and eclosion lethal phenotypes in Drosophila [15]. The late genes play direct or indirect roles and also a distinct role in controlling the appropriate biological response to hormone, including development, metamorphosis, reproduction and diapause. As outlined above, most of the studies regarding the ecdysone-responsive genes in insects have focused on the regulatory genes at the top of the ecdysone-elicit genetic hierarchy.
Chitin is the major polysaccharide layed in the cuticle, peritrophic matrix, tracheae and muscle attachment points as a characteristic constituent of insects and other arthropods. The chitin biosynthesis pathway begins with glycogen and trehalose, and consists of a patchwork of at least eight key enzymes ( Fig. 1) [16]. Importantly, the process of chitin biosynthesis and degradation is strictly coordinated within the cycle of molts and behaves as an ecdysone-induced response [17,18,19]. Many studies have demonstrated the participation of 20E-EcR/USP complex in regulation of gene expression; however, the ecdysoneresponsive genes involved in the process of chitin biosynthesis are still largely unclear.
Here we cloned two EcR isoforms, SeEcR-A and SeEcR-B1, from the beet armyworm, Spodoptera exigua, a wide-spread, destructive and polyphagous noctuid lepidopteron pest. Consequently, we confirmed that the injection-based RNAi of SeEcR leads to a delay in developmental duration, reduced food intake, kinds of defect phenotypes in pupae formation and adults eclosion, and also chitin content reduction in the cuticle of abnormal larvae. The effects of RNAi on the target gene were proved to be gene-specific and effective, with the efficiency lasting for 108 hr. The results after the injection of dsRNA for EcR or 20E in vivo and the midgut culture experiments in vitro revealed that the regulation of EcR remarkably affects the expression of five genes (SeTre-1, SeG6PI, SeUAP, SeCHSA and SeCHSB) in the chitin biosynthesis pathway, and genes encoding SeTre-1, SeG6PI ,SeUAP and SeCHSB were identified as 20E late-response genes.

Isolation and sequence analysis of EcR-A and EcR-B1 cDNAs
Cloning of EcR cDNAs from S. exigua was accomplished by RT-PCR using degenerate primers designed on the basis of conserved motifs in the DNA binding domain (DBD) and ligand binding domain (LBD) of insect EcR sequences. To subclone, cDNA from the first instar larvae was used as a template to obtain a 527-bp PCR fragment of the shared common domain, and two full-length cDNAs were obtained with a combination of 39-RACE and 59-RACE methodologies. Database BLAST search revealed that these cDNAs encoded S. exigua homologs of the nuclear receptor EcR, which contain a common carboxy-terminal region with a DNA-binding and ligand-binding domains but unique aminotermini of 113 and 187 amino acids ( Fig. 2A). The two cDNAs were sequenced with an open reading frame of 1546 and 1767 nucleotides, encoding a protein of 514 and 588 amino acids with a calculated molecular weight of 57.4 and 66.0 kDa, respectively.
Hence, on the basis of the similarity in their A/B domains when aligned with other insect EcR members, one cloned cDNA encodes an EcR-A isoform with a conserved 20 amino acid sequence (specific A-box) at the C-terminal of the A/B domain of all isoform-A receptors, and the other one is a SeEcR-B1 isoform. The deduced amino acid sequences of the two isoforms showed high similarity to the reported insect EcR sequences. SeEcR-A (GenBank accession number: GU296540) is 90% identical to SeEcR-B1 (GenBank accession number: EU426551).
By quantitative real-time PCR, the developmental expression patterns of SeEcR-A and SeEcR-B1 were investigated in all developmental stages including the first day of egg to the last day of adult. The results showed that both SeEcR-A and SeEcR-B1 were continuously expressed through the whole life cycle of S.exigua in correlation with the ecdysteroid pulse. (Fig. 2B).

The efficiency of RNAi towards the two EcR isoforms
To analyze whether it was possible to reduce SeEcR levels by RNAi in vivo, a 546 bp fragment containing the coding region shared by the two EcR isoforms, named as dsEcRcom was used for dsRNA synthesis ( Fig. 2A). qRT-PCR results showed that SeEcR mRNA levels clearly decreased in the dsEcRcom-injected insects as compared with those in the dsGFP-injected control (Fig. 3D). At 12 hours post-treatment, dsEcRcom injection resulted in an approximately 3-fold knock-down of the SeEcR transcript level. Moreover, in western blot analysis result, the protein level changes of the target genes showed that both SeEcR-A and SeEcR-B1 were not influenced at 12 hours after dsRNA injection, but decreased immune signals are detected in response to the dsEcRcom injection at 36 hours post-injection (Fig. 3E). Additionally, the silencing effect on the target gene lasted for as long as 108 hr, suggesting that the microinjection of dsRNA was effective at silencing the EcR genes in S. exigua.
We observed severe developmental abnormity and phenotypic defects caused by the dramatic changes in the mRNA level of SeEcR by SeEcR-specific RNAi in S. exigua. Quite remarkably, following dsEcRcom injection, the average survival rates were 81.11%, 30.00%, 15.56%, and 5.56% in the fifth-instar larvae, prepupae, pupae and adults, respectively, and were significantly lower than those in the three controls (Fig. 3A). A sharp decline in survival rate was observed between larva-prepupa metamorphosis after RNAi. In the dsEcRcom-injected insect, pupation was abnormal and the pupa-adult metamorphasis could not complete, with the pupal exocuticle partially shelled, the adult in pupal cuticle arrested entrapped or the wing brokenly extended. Although the severity of the phenotype defects varied somewhat, two typical severe defect phenotypes, ''half pupation'' larvae and ''abnormal eclosion'' adults, were observed ( Fig. 4B, C, H, I). These ''half pupation'' larvae, which were not observed in the controls, always failed to complete pupation and died within one or two days. Moreover, the dsEcRcom-injected group also had evidently longer developmental durations in the larva, prepupa and pupa phase (Fig. 3B). In contrast, no significant lethality differences were found among the three controls. In addition, many individuals injected with dsEcRcom shrank in body size and had reduced food-intake (Fig. 3C).

Chitin levels in the epidermis and midgut/peritrophic membrane after RNAi
To address the impact of lower SeEcR transcript levels on the chitin metabolism, the chitin contents in the epidermis and the midgut/peritrophic membrane (PM) were determined. From 12 hr to 36 hr post injection of dsRNA, the chitin contents in both the midgut/PM and the epidermis decreased in both treatments of dsGFP and dsEcRcom (Table 1). At 12 hr post injection, no significant difference on chitin levels was found between treatments of dsGFP and dsEcRcom. At 36 hr post injection, however, the chitin content of the midgut/PM in the dsGFP treatment group was relatively lower than that in the dsEcRcom treatment group. Since chitin degradation starts in the midgut/PM in this stage [18,20], the reduction in SeEcR levels may result in slow chitin degradation carried out in the midgut/ PM. While in the epidermis at 36 hr post injection, both degradation of larvae epidermis chitin and generation of pupae cuticle chitin happened in prepupal stage. Lower levels of SeECR may have different effects on the degradation and the generation of the chitin, which may result in relatively higher chitin content in the dsGFP treatment. These results were also in correlation with the longer developmental duration observed in this study.

Effects of SeEcR RNAi on the expression of key genes in the chitin biosynthesis pathway
The chitin analysis results (Table 1) suggested that SeEcR expression affected the chitin contents in the epidermis and the peritrophic membrane. Thus, six genes involved in the chitin biosynthesis pathway were examined to ascertain their responsiveness to ecdysone receptor expression using qRT-PCR. The developmental expression profiles of SeEcR-A and SeEcR-B1 mRNA. Total RNAs were prepared from individuals of S. exigua collected at all developmental stages from the first day of eggs to the last day of adults. The levels of the mRNA were detected using qRT-PCR. The mRNA level of b-Acitn was used as a control to normalize for loading. The data represent the mean values 6 SE of three replicates. The age of the insects is indicated in: E1, the first day of eggs; I1, the first day of the first instar larvae; pre, prepupal period; P1, the first day of pupae; A1, the first day of adults. doi:10.1371/journal.pone.0014058.g002 Concurrently, at 12 hr after dsEcRcom injection, the expressions of SeTre-1 (EU427311), SeCHSA (DQ062153) and SeCHSB (EU622827) were dramatically down-regulated in the dsEcRcominjected larvae and reduced to approximately half of the mRNA levels in the control insects. Coinciding with the low expression level of SeEcR, the significantly suppressed effect on these three genes lasted for 108 hr (Fig. 5A, E and F). The SeCHSA and SeCHSB genes seem to share a similar down-regulation pattern. Meanwhile, SeTre-2, SeG6PI and SeUAP had extraordinary responses to SeEcR RNAi. The transcriptional abundance of SeG6PI (unpublished) was repressed at 12 hr after dsEcRcom injection, with a small rise at 36 hr, and then remained low until 108 hr post-injection of dsEcRcom (Fig. 5C). Quite differently, SeUAP (FJ380203) expression was slightly affected at the first testing point, and the repressed transcription level of SeUAP was only observed at 24 hr and 72 hr post-injection with a detectable but subtle increase in expression at 108 hr after dsEcRcom injection ( Fig. 5D). Conversely, hardly any expressional variation of SeTre-2 (EU106080) was observed before 72hr post-injection of dsEcRcom (Fig. 5B). The qRT-PCR results revealed different transcript level changes in the six genes involved in chitin biosynthesis, which suggests that the effects of EcR on the six genes varied somewhat.

Influence of 20E on expression of SeEcR and key genes in the pathway
After 20E injection, the relative mRNA expressions of SeEcR and the six genes in the various groups of S. exigua larvae were assayed using qRT-PCR. The results showed that there was a burst in SeEcR transcription that increased nearly 4-fold when compared to the mRNA level in control insects at 4 hr after 20E injection, while there was no difference in the relative abundance at 12 hr and 36 hr after 20E injection as compared with the control (Fig. 6). At 12 hr post-injection of 20E, the mRNA expressions of SeTre-1, SeG6PI and SeCHSB showed about 3-fold represents no significant difference among dsGFP, buffer and CK groups but the survival rates in all three control groups are significantly different from that in the dsEcRcom-injection group (labeled with a lowercase ''b'' at the same developmental stage). Different letters at the same developmental stage in Fig. 3B and Fig. 3C among different treatments indicate significant differences of either developmental duration or foodintake (p,0.05, Duncan multiple comparison test, SPSS). For (D), the expression level of SeEcR and Seb-actin transcripts were detected using qRT-PCR. By definition, the mRNA level of the target gene was normalized to Seb-actin in each group; and the relative expression level of the target gene was measured as the level in the dsEcRcom-injection group divided by the level in the dsGFP-injection group. Data are shown as mean 6 SE from three independent experiments. An asterisk indicates significant differences in the expression level between the treated and control groups measured at the same time (p,0.05, T test). For (E), the protein levels of target genes were detected using Western blot. A total of 80 ug of proteins from each individual treated with dsGFP or dsEcRcom was applied to each lane and separated by 10% SDS-polyacrylamide gel electrophoresis. Lanes 1 and 3: 36 and 12 hours post-dsGFP injection; Lanes 2 and 4: 36 and 12 hours post-dsEcRcom injection, respectively. The results are representative of four replicates. doi:10.1371/journal.pone.0014058.g003 up-regulation, SeCHSA transcript was up-regulated over 4-fold and SeUAP mRNA level was increased nearly 7-fold. Thereafter, a significantly stimulated up-regulation of the SeCHSA and SeCHSB genes were observed at 36 hr after 20E injection, synchronously coinciding with distinct expression of the SeTre-1, SeG6PI and SeUAP genes whose variation of transcript abundance were barely detectable at this detection point. On the contrary, treatment with 20E resulted in a delay in changes of the SeTre-2 expression before 36 hr post-injection of 20E. Hence, the expression of the five genes (excluding SeTre-2) continued to increase in the 20E injection group, suggesting that these five genes are probably among the ecdysone-responsive genes.

Response of genes in the pathway to ecdysone or cycloheximide in vitro
To further determine the hierarchy position in the EcR/USP-20E-mediated genetic regulatory network, the transcript accumulation patterns of all genes (except SeCHSA, a non-midgut gene) in the pathway were investigated in cultured midguts in vitro.
Incubation of the midguts in the presence of both 10 26 M 20E and 10 23 M cycloheximide (Chx,the protein synthesis inhibitor) in the medium slightly or hardly induced the following four genes: SeTre-1, SeG6PI, SeUAP and SeCHSB, a pattern quite similar to that in incubation with the same dose of Chx in the hormone-free medium as determined at p,0.05 using T test (Data not shown). As opposed to two treatments mentioned above, the four mRNA induction profiles in the cultured tissues incubated with 10 26 M 20E were siginificantly different at various times points (Fig. 7). When analyzing larval midguts cultured in the presence of 20E, the induced expression peaks of SeTre-1 and SeCHSB were detected 4 hr later, and those of SeG6PI and SeUAP could be sensed 8 hr later (Fig. 7A, B, C, D). In this in vivo midgut culture experiment, the data demonstrated that the expression of these four genes could be induced by 20E but inhibited by Chx.

Characteristics of S. exigua EcR isoforms and their role in developmental processes
We cloned and characterized two EcR isoforms (Fig. 2). Analysis of the deduced amino acid sequences indicated that they were two isoforms of EcR, a member of the steroid hormone receptor superfamily, homologous to the Drosophila EcR (Fig. 8). The fact that the two isoforms have distinctive N-terminal A/B domain and share a highly conserved amino acid sequence in all other domains at the C-termini suggests they are splice variants generated from the same EcR gene transcript. Ever since the Drosophila EcR gene was cloned and the three EcR isoforms (EcR-A, EcR-B1 and EcR-B2) were characterized [21], two isoforms of EcR (EcR-A and EcR-B1) were only reported in a few lepidopteron species, including Bombyx mori, Choristoneura fumiferana and Chilo suppressalis [22,23,24]. More evidence is needed to confirm whether three isoforms actually exist in lepidopteron insects.
The regulation of EcR on insect development has been well demonstrated by mutational analysis and by using inducible expression of double stranded RNA (dsRNA) of EcR isoforms in D. melanogaster, which resulted in a block in neuronal remodeling and ecdysone responses during larval molting and metamorphosis [25,26,27], developmental arrest during embryogenesis and failure of pupation [28]. The results of delayed development duration and reduced food-intake, which were presented herein by using RNAi in vivo during the last larvae instar of S. exigua, indicated that SeEcR is necessary for appropriate developmental behavior of the insect. The effect of EcR deficiency on survival rate varied in different insects. D.melanogaster carrying mutations that inactivate all three isoforms of EcR are embryonically lethal [29], and EcR RNAi studies in Tribolium castaneum showed nearly 100% lethality rate in injected larvae [30]. In our study, as high as a 95% mortality prior to eclosion was observed (Fig. 3A). In addition, the SeEcR knockdown larvae show defective phenotypes, including the typical ''half pupation'' phenotype (Fig. 4E), which resembles misshaped structures that were observed after the ingestion of dsRNA for SeCHSA [31].
The effect of RNAi on the target gene SeEcR in this study was effective (Fig. 3). To confirm that the observed RNAi effect is not due to an off-target effect, a second different SeEcR dsRNA fragment (362 bp), designed from the DBD (DNA binding domain) domain and the D (hinge) domain of two EcR isoforms, was used for injection. Similar results were observed in survival rates (such as sharp decline during the larvae to prepupae period) and in the typical phenotypes (e.g. half pupation and abnormal eclosion) (Data not shown). This proved that the RNAi of SeEcR is gene-specific.  The five genes (except SeTre-2) in the chitin biosynthesis pathway are regulated by SeEcR EcR acts as an ecdysteroid-inducible transcription factor by directly interacting with USP via the DNA-response elements in the promoters of target genes [32,33]. Apparently, the EcR plays a vital role in the transcription of multifarious genes [34,35]. However, the role of the EcR in controlling the transcription of the genes responsible for the key enzymes involved in the chitin biosynthesis processes remains enigmatic. Measurement of the transcript levels following dsEcRcom injection allowed us to identify distinct gene-specific responses of SeTre-1, SeTre-2, SeG6PI, SeUAP, SeCHSA and SeCHSB.
Most likely, SeEcR continuously regulated the transcription of SeTre-1, SeCHSA and SeCHSB (Fig. 5A, E and F). In the bamboo borer Omphisa fuscidentalis, expression of EcR-A and EcR-B1 was followed by a decrease in trehalose concentration and an increase in trehalase activity [36]. The CHSA and CHSB mRNA levels could be up-regulated in a stage-specific manner during the onset of metamorphosis in D. melanogaster [18]. Whereas, the change of SeG6PI and SeUAP transcripts in S. exigua were more complicated ( Fig. 5C and D). The undisturbed mRNA levels of SeG6PI at 24 hr and an unexpectedly increased expression of SeUAP at 108 hr after dsEcRcom injection led to the hypothesis that catalyzed products of G6PI or UAP might be potential feedback inhibitory substances in the pathway. It is not suprising that the role of UAP and G6PI might be far more complex than we expected [37]. Both our studies (Figs. 5B and 6) and those of Tatun et al. revealed 20E and EcR hardly had effect on a transmembrane enzyme gene (Tre-2) at early detection point, which substantially dued to the uninvolvement of Tre-2 in the dynamic changes in the hemolymph trehalose concentration that occurred during the larval-pupal transformation [38]. In short, the loss of SeEcR results in not only reduced abundance of the five genes transcripts responsible for the enzymatic reactions in chitin biosynthesis pathway, but also Figure 5. Changes in mRNA expression levels for key genes in the pathway following SeEcR RNAi. The expression levels of six transcripts (SeTre-1, SeTre-2, SeG6PI, SeUAP, SeCHSA and SeCHSB) were detected by qRT-PCR at five detection points after dsEcRcom injection. Data are represented as mean 6 SE from three independent experiments, which were analyzed in the same way as in Fig. 3. *: a significant difference between the dsEcRcom injection group and the dsGFP injection group (p,0.05, T test); NS: no significant difference between the two groups. doi:10.1371/journal.pone.0014058.g005  obviously suppressed chitin metabolism (Table 1) in a stage-and tissue-specific response to insect hormone during insect metamorphosis.
The four genes in the pathway are ecdysone-regulated late genes Ecdysone has been shown to negatively regulate transcription of pupal cuticle protein genes in imaginal disks and is required for apical extracellular matrix formation and epithelial morphogenesis in D. melanogaster [39]. In an in vitro system of lepidopteron imaginal wing discs, ecdysone had been reported to stimulate the incorporation of N-acetyl-D-glucosamine into glycoproteins, and upregulate the expression of adult cuticular genes in cell lines derived from Tenebrio molitor [40]. This data provides important clues in regards to an intriguing extension of the ecdysonemediated genetic regulatory network, and the question as to whether the five genes in the chitin biosynthesis pathway are among the ecdysone-induced genes arises.
Here in our study, significant up-regulation of SeCHSA and SeCHSB was detectable at 4 hr and durative to 36 hr after 20E injection into the larvae of S. exigua, and high levels of SeTre-1, SeG6PI and SeUAP expression were detected at 12 hr post-injection of 20E. This implies that the expression of the five genes are induced by 20E (Fig. 7). In agreement with our present results, previous in vivo studies had demonstrated that hemolymph trehalose concentration was sensitive to 20E level in Omphisa fuscidentalis [41]. In vitro up-regulation of CHSA and CHSB was due to an indirect interaction between the ecdysone receptor complex and ecdysone-responsive elements within the CHSA and CHSB promoters in D. melanogaster and these two genes act as members of the late genes [18]. The expressiong of the mmy gene encoding UDP-Nacetylglucosamine pyrophosphorylase (UAP) in Drosophila could be altered by the moulting hormone 20-Hydroxyecdysone [42]. Discriminatively, SeTre-2 seems insensitive to 20E, which was more similar to the ecdysone-independent genes. Taken together, we speculate that the five genes SeTre-1, SeTre-2, SeG6PI, SeUAP, SeCHSA and SeCHSB excluding SeTre-2 are among the ecdysoneinduced genes hierarchies (Fig. 9).
Different 20E-responsive genes seem to be differentially regulated by ecdysone. This has been experimentally demonstrated by the in vitro culture experiments in larval salivary glands of D.melanogaster [2,8,9]. When there is a continuouse high level of 20E, early genes are activated with a quick response, reach their maximum expression and then regress within 4 hours. Due to the inhibition by cycloheximide (Chx) on their induced expression and a delayed reaction when incubating with ecdysone, the 20E lateresponse genes could be classified at a lower hierarchy position [1,5,43,44]. In the in vitro cultured midguts, the four genes (SeTre-1, SeG6PI, SeUAP and SeCHSB) showed a slow peak in activation by 20E at least 4 hr after incubation, but no continuous increase in their mRNA expression was detected during the 12 hours when cultured in the presence of both Chx and 20E (Fig. 7). This response of gene transcription to 20E and Chx in this manner is characteristic of the ''late'' genes analogous to genes of the ecdysone regulatory hierarchy as described by Ashburner and his colleagues in D. melanogaster (Fig. 9).
Further studies to directly identify the relationship between the expression of the chitin biosynthesis interrelated genes and the hormone governing the stages of insect metamorphosis is currently in progress in our laboratory, and the regulation of other 20E early-response genes is required to identify in order to be verify for this connection.

Insects
S.exigua were reared in the laboratory at 25uC62uC and 75%65% r.h. (relative humidity) on a 14L:10D photoperiod using an artificial diet [45]. All dissections and tissue sampling of specimens were carried out in ice-cold saline and stored at 280uC for later use.

RNA isolation and cDNA synthesis
Total RNA was extracted from individual specimens of S. exigua using Trizol reagent (Invitrogen, USA). Total RNA (5 mg) was used for synthesizing cDNA with the AMV reverse transcriptase XL (Takara, Japan) with the following reaction conditions: 37uC for 10 min, 42uC for 1 h, 99uC for 5 min, and 5uC for 5 min [46].

Cloning and analysis of SeEcR cDNAs
The ABD SMART RACE cDNA amplification kit (BD Bioscience Clontech, CA, USA) was used to obtain the full-length SeEcR cDNA. To clone the S. exigua EcR gene, degenerate primers based on the conserved amino acids sequences of ''PRQQEE'' (forward primer, SeEcR-F) and ''QPSXE'' (reverse primer, SeEcR-R) were used to obtain a common fragment of EcR ( Table 2). The touch-down PCR amplification reactions were carried out as follows: 94uC for 5 min, 5 cycles of 94uC for 40 s, 48uC for 40 s and 72uC for 40 s, 25 cycles of 94uC for 40 s, 54uC for 40 s and 72uC for 40 s, and with an extension step at 72uC for 10 min at the end. The amplified product was then separated on an agarose gel and purified using the Gel Extraction Kit (OMEGA, USA). The amplified fragment was subcloned into the pMD-18T vector (Takara, Japan) and sequenced. To get a longer cDNA sequence, we performed 39-RACE and 59-RACE using nested gene-specific primers for SeEcR (Table 2) and anchor primers supplied in a SMART RACE cDNA Amplification Kit (Clontech).
On the National Center for Biotechnology Information (NCBI) website, the sequence of the two EcR cDNAs were compared with the other EcR sequences deposited in the GenBank using the ''BLAST-X'' tool. Amino acid sequences were deduced from the corresponding cDNA sequences using the translation tool at the ExPASy Proteomics website (http://expasy.org/tools/dna.html). A protein sequence analysis tool from the ExPASy Proteomics website (http://expasy.org/) was used to predict molecular weight. Multiple sequence alignments of the deduced amino acid sequences were made using Multiple Alignment software (http://www.ebi.ac. uk/clustalw/index.html) [47]. Phylogenic and evolutionary analyses

dsRNA synthesis
A 546 bp fragment from the EcR common region ( Fig. 2A) and a 714 bp fragment from GFP (ACY56286) were PCR amplified. dsRNA synthesis was performed using the T7 RiboMAX TM Express RNAi System Kit (Promega, USA) according to the manufacturer's instructions. Primers for EcR and GFP were designed by adding the T7 polymerase promoter sequence to their 59ends ( Table 2). The dsRNA was resuspended in an appropriate amount of nuclease-free water (DEPC water). The purified dsRNAs were quantified by spectroscopy and examined by agarose gel electrophoresis to ensure their integrity.

RNA interference bioassays and sampling
In vivo RNAi in S. exigua larvae was performed as previously described [48]. The 2nd day of fifth-instar (the last instar) larvae were anaesthetized on ice for 2-3 min and used for the injection experiment. Five micrograms of dsRNA dissolved in 5 ml of DEPC water were injected into larvae between the second and the third abdominal segment using 10 ml microliter syringes (Hamilton), and the injection point was sealed immediately with wax. Both the dsRNA of GFP and an equivalent volume of buffer were used as controls.
For phenotype observation, feeding and developmental duration experiments, each group contained 30 individual larvae in each set of repetition and three replicates were used. Observation was performed every 12 hours after injection. The following three controls were included: the dsGFP group (an equivalent amount of GFP dsRNA), the buffer group (the same volume of DEPC water) and the CK group (control check with no treatment). The artificial diet pellets were cut into appropriate analogous sizes. All diets were replaced daily and each piece of food was weighed on an accurate electronic scale to measure the food intake of each individual insect. The results are expressed as the mean6standard errors (SE) of three independent replicates (n = 3), each representing 30 fifth-instar larvae. The statistical significance of data from the survival rates, food-intake and developmental duration were determined by one-way analysis of variance (ANOVA) and analyzed by Duncan multiple comparison test. Percentage data were arcsine square root transformed prior to ANOVA. For chitin content assay, each group included 50 individuals and 8 abnormal larvae were chosen for analysis (n = 8). The cuticle or midgut of individual misshapen specimens was disesscted as the chitin analysis sample. For the target gene detection experiment, each group included 60 individuals with three replicates, and the SeEcR mRNA levels of randomly chosen individual samples were measured at 12 hr, 24 hr, 36 hr, 72 hr and 108 hr after the injection of SeEcR dsRNA. For the pathway gene detection assay, each treatment included 60 individuals with three replicates and three misshapen samples were selected at 12 hr, 24 hr, 36 hr, 72 hr and 108 hr post-injection for the independent detection.

Western blot analysis
Protein samples were made by homogenizing 1 larva per 150 ml of PBS (20 mM sodium phosphate and 130 mM NaCl, pH 7.4) plus protease inhibitor cocktail Phenylmethyl sulfonylfluoride (PMSF). After freeze thawing for three times, centrifuged at 12,000 g, 4uC for 30 min. The protein content in supernatant was determined by BCA kit, and the supernatant containing equivalent protein was mixed with 56loading buffer for SDS polyacrylamide gel electrophoresis and boiled for 10 minutes, After centrifugation at 12,000 g for 5 min, protein samples were cooled to room temperature before loading on a 10% polyacrylamide SDS gel. The process of western blotting were modified from the method previously described [49]. A purified primary antibody (anti-SeEcRcom antibody, 1:750 dilution) and an lgG goat anti-rabbit antibody conjugated with HRP was used for secondary antibody (BOSTER, 1:5000 dilution).

In vivo ecdysteroid injection
For the in vivo ecdysteroid injection experiment, 20E (Sigma, St. Louis, MO, USA) was dissolved in 5 ml of DEPC water at doses of 2 mg per insect, and each group included 30 individuals with six replicates. Three individual larvae were randomly chosen for the experiment 4 hr, 12 hr and 36 hr after injection.

In vitro midgut culture experiments
The 2nd day of fifth-instar larvae were dissected in saline. The midguts together with the peritrophic membranes were ligated tightly using cotton thread at the anterior and posterior ends, and then cut to the ligatures. After the isolated midguts were rinsed four times with medium, they were independently cultured in Grace's Insect Tissue Culture Medium at 25uC62uC for 12 hr.
Four groups were included in this culture experiments: the CK control group (control check group with no addictive in culture medium); the 20E group (added 20E hormone to final concentration of 10 26 M); the Chx group (added cycloheximide to a final concentration of 10 23 M); the 20E+Chx group (added 20E to final concentration 10 26 M and cycloheximide to 10 23 M). Each group included 10 individual tissues with three replicates. Samples for mRNA expression analysis were selected randomly at 0 hr, 4 hr, 8 hr and 12 hr after incubation.

Quantitative real-time RT-PCR
The relative mRNA expression of genes was assessed by quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) by using the SYBR Green based detection system (SYBR Premix Ex Taq, Takara, Japan) with Light-Cycler480 system (Roche, Germany). This assay allows for precise quantification of weakly expressed genes. The oligonucleotides used in qRT-PCR are indicated in Table 2. Specificity of the reaction was checked by analysis of the melting curve of the final amplified product. The housekeeping gene and target gene from each sample were run in parallel on the same PCR plate. All of the qRT-PCR reactions were performed using the following common program: pre-incubation at 95uC for 10 s, followed by 45 cycles of denaturation at 95uC for 15 s, annealing at 60uC for 20 s, and elongation at 72uC for 15 s. At the end of each cycle, a fluorescence reading were used to determined the extent of amplification. Standard curves were obtained using a 10-fold serial dilution of pooled cDNA from all stages.
As described above, cDNA was synthesized from specimens of S. exigua from experimental group and the GFP dsRNA group at every detection point using the AMV reverse transcriptase XL (Takara, Japan). The qRT-PCR measurement was performed in triplicate and normalized to an internal control for each sample (the S. exigua b-actin gene AY507963, Table 2). The relative expression level of the target gene was calculated by the mRNA level in the treated group divided by the level in the control group.

Chitin analysis
A modified analysis procedure was used in the chitin content measurement experiment as previously described [50]. For chitinase digestion, each pellet was resuspended in 250 ml of McIlvaine's buffer, and Streptomyces plicatus chitinase (Sigma, 5 mg/ ml in PBS, 50 mg for cuticle and 25 mg for midgut) was added and incubated for 72 h at 37uC to hydolyse chitin to GlcNAc. Standard curves were generated for data processing.