Introduction

Ecdysteroids are crucial to the growth, reproduction, development, regeneration and molting of crustaceans1,2. The nuclear heterodimeric complex ecdysone receptor (EcR)/retinoid X receptor (RXR) mediates ecdysteroid activity in crustaceans3. Upon activation by ecdysteroids, the heterodimer complex induces the sequential transcription of protein-coding genes that ultimately direct molting, metamorphosis, and growth4,5. EcR is a member of the nuclear receptor (NR) superfamily and is a ligand-inducible nuclear transcription factor6. Like other NR members in crustaceans and insects, EcR has a consensus domain architecture that comprises the A/B, C, D and E domains. Among insects, the N-terminal A/B is the least conserved domain and is associated with transcriptional activation. The C domain, also called the DNA-binding domain (DBD), is highly conserved. The D region is a flexible region hinge that mediate nuclear localization and subunit pairing. The E domain contain, also called ligand binding domain (LBD), is moderately conserved and contains a hydrophobic pocket for ligands. Moreover, the E domain is involved in receptor dimerization. In addition to these common domains, a distinct and highly diversified F domain is present in some insects7,8,9.

Alternative splicing is an important mechanism of post-translational modifications, that regulates gene expression and increases gene diversity10. A single coding sequence can produce many gene products with different function by alternative splicing, that greatly enrich the biological genetic information11,12. Alternative splicing is tightly regulated in a manner that is specific to cell or developmental stage and is a common biological process in human and other organisms13. In plants, alternative splicing is pervasive in different developmental stages and environmental conditions14. Alternative splicing controls the expression of key sex-determining genes in insects15. More than 90% of human genes are alternatively spliced, thus implicating alternative splicing in diverse biological processes16. Alternative spliced EcR isoforms have been identified and studied in both insects and crustaceans6,17. Previous studies have shown that in the Drosophila melanogaster, three different functional EcR isoforms EcR-A, EcR-B1 and EcR-B2 harbor the same DBD and LBD regions, and differ only in their N-terminal regions18. Each isoform dominates in different target tissues at different developmental stages18,19. Different EcR isoforms control cell-specific responses during nervous system remodeling in metamorphosis: for example, EcR-B is required for larval molting and neuron remodeling during metamorphosis20. Two EcR isoforms have different functions during the development of the epidermis and wings of the Tobacco Hornworm, Manduca sexta 21. Two EcR isoforms are expressed in stage- and cell-specific manners during midgut remodeling in the yellow fever mosquito, Aedes aegypti 22. Alternative EcR isoforms have been identified in other insect species with specific functions in different tissues and developmental stages23,24,25,26,27,28,29,30.

The crustaceans EcR gene has similar functions to insects, but has different numbers and structures of isoforms6. Three EcR isoforms from the mud crab, Scylla paramamosain, were identified with a variable LBD domain31. Three isoforms from the water flea Daphnia magna exhibit different temporal expression patterns during molting period32. An 18-amino-acid insertion/deletion and a 49-amino-acid substitution have been identified in the coding region of MnEcR from the freshwater prawn Macrobrachium nipponense. These sequence modifications result in splice variants MnEcR-L1, -L2, -S1 and -S2 33. Two isoforms have been isolated from the American lobster, Homarus americanus 34 and four from the blue crab, Callinectes sapidu 35. However, most studies on crustaceans have only presented the number of alternative EcR isoforms. The complete structure of EcR, the specific function of each isoform, and the formation of alternative isoforms are less understood in crustaceans. In addition, some of the conserved domains in insects are reportedly different from those of crustaceans6. Whether insects and crustacean possess the same evolutionary pattern for EcR gene is also unknown.

The Chinese mitten crab, Eriocheir sinensis, is an economically important farmed species in East Asia and is an extensively influential species in Europe and Northern America36,37. E. sinensis presents the representative biological characteristics of crustaceans, such as molting, metamorphosis, regeneration, development, and reproduction38,39,40,41. Elucidating EcR structure and its alternative isoforms will help us decipher the functional roles of EcR in E. sinensis and expand our knowledge of the crustaceans EcR. However, studies that are related to the structure and function of the alternative isoforms in E. sinensis are sparse41. In the current study, we obtained the whole EcR gene sequence, clearly deciphered the EcR gene structure, and identified four EcR alternative isoforms from E. sinensis. To thoroughly understand the biological function of the different isoforms in specific molting stages and tissues, semi-quantitative, quantitative gene expression analyses, and RNA interference experiments were conducted during molting and different tissues of E. sinensis.

Results

Characterization of the EcR gene structure

The EcR gene was 45,481 bp long based on the assembled E. sinensis genome from previous study. The gene included 9 exons and 8 introns (accession number KY303919) (Fig. 1, Supplementary Table S2). To verify the accuracy of the assembled EcR gene, partial EcR DNA sequences were amplified and sequenced by an ABI 3730 sequencer. ORFs were amplified to identify the four alternatively spliced isoforms (Es-EcR-1 to Es-EcR-4). The ORF lengths were: Es-EcR-1 (1638 bp, accession number KY303915), Es-EcR-2 (1626 bp, accession number KY303916), Es-EcR-3 (1545 bp, accession number KY303917), and Es-EcR-4 (1638 bp, accession number KY303918) (Fig. 1, Supplementary Fig. S1).

Figure 1
figure 1

Gene structure of the four E. sinensis EcR isoforms. Different colored boxes in the upper figure indicate the different exons (exon 1 to exon 9) of the EcR gene. The black underline in the bottom figure denotes the intron boundary. EcR gene is located in scaffold5630 of assembled E. sinensis genome (http://gigadb.org/dataset/100186)62.

Full size image

Alternative splicing patterns of the EcR gene

The E. sinensis EcR gene had three alternative splicing patterns. The Es-EcR-2 isoform belonged to the alternative 5′splice site (5′AE). Compared with Es-EcR-1, a 12 bp sequence on the 3′ ends of exon 4 was absent in Es-EcR-2 (Fig. 1). Es-EcR-3 lost the entirety of exon 5 and 12 bp from exon 4. This pattern was similar to that in Es-EcR-2. Thus, Es-EcR-3 belonged to the 5′AE and exon skipping (SKIP). Although Es-EcR-4 was as long as Es-EcR-1, but the difference was that Es-EcR-4 lost the whole exon 7 (150 bp in length) compared with Es-EcR-1 to Es-EcR-3, and a same length intron (150 bp) retained as the exon 7 for Es-EcR-4. This result indicated that Es-EcR-4 belonged to exon skipping and intron retention (SKIP and IR) (Fig. 1).

Functional domain prediction for the four isoforms

Sequences alignment revealed that all four EcR isoforms possessed the four characteristic functional domains of NRs (A/B, C, D, and E domains) (Fig. 2A and B). The A/B and C domains were conserved across the four isoforms, whereas, the lost 12 bp of Es-EcR-2, as well as the lost 12 bp and the whole exon of Es-EcR-3 were located in the D hinge domain. Es-EcR-4 retained introns in the LBD (E domain). The four alternative isoforms had distinctly different predicted protein structure, which may regulate specific developmental processes or may have different function in E. sinensis (Fig. 2C).

Figure 2
figure 2

Amino acid sequence alignment, protein domain structure, and predicted protein structure of the four EcR isoforms. (A) Amino acid sequence alignment of the four EcR isoforms. Blue and red boxs indicate the missing sequences of Es-EcR-2 and Es-EcR-3. The purple box denotes the variation in Es-EcR-4. (B) Pattern diagrams of the domain structures of the four EcR isoforms. Variations are identified in the D and E regions. (C region indicates DBD, D region indicates the hinge region between the C and E regions, and E region indicates LBD). (C) Predicted protein structure of the four EcR isoforms. A purple circle indicates the location of the four missing amino acids in Es-EcR-2. A yellow circle signifies the site of missing amino acids; a green circle denotes the location of amino acids variation.

Full size image

Evolutionary analysis of the EcR gene

The structures of EcR genes from 11 crustaceans and 20 insects were observed in this study. The number of alternatively spliced isoforms varied among insects and crustaceans: 10 out of 11 crustaceans harbor more than two isoforms (and most have four isoforms). Moreover, only 3 out of 20 insects had three isoforms and 17 harbored only two isoforms. The variable domains all occurred in the A/B domain of the studied insects. Variable domains occurred frequently in the D and E domains in crustaceans, and 10 out of 11 crustaceans had variable D/E domains (Table 1). We constructed a ML phylogenetic tree with the longest isoforms for each studied species. The tree revealed two independent branches that clearly discriminated between crustaceans and insects (Fig. 3).

Table 1 Number of EcR isoforms and variable regions in different species.
Full size table
Figure 3
figure 3

ML phylogenetic tree based on EcR sequences. The tree was generated by MEGA 5.0 with 1,000 bootstraps.

Full size image

Functional study of splicing isoforms

Es-EcR-2 to Es-EcR-4 were all expressed during the different molting stages, with the lowest expression level during postmolt stage (PoM) and the highest during premolt stage (PrM) (Fig. 4A). However, no significant difference in the expression of Es-EcR-2 was identified between intermolt (InM) and PrM stages. Es-EcR-4 expression levels were not statistically different among molting stages. Es-EcR-3 expression was nine times higher (fold change) during PrM than during InM (Fig. 4A). Semiquantitative PCR also indicated that Es-EcR-3 predominated during PrM (Fig. 4B). Western blot results for Es-EcR-3 were consistent with qRT-PCR and semiquantitative PCR results. Es-EcR-3 was significantly and highly expressed during PrM (Fig. 4C). The expression level of Es-EcR-3 was significantly decreased after RNA interference (Fig. 4D) and in vivo experiment indicated 3 days was prolonged during molting in siRNA group compared with negative control siRNA group (P < 0.05) (Fig. 4E and Supplementary Table S3).

Figure 4
figure 4

Expression analysis of the four EcR isoforms during molting and analysis of siRNA on Es-EcR-3 isoform. (A) qRT-PCR analysis for the relative expression levels of Es-EcR-2, Es-EcR-3, and Es-EcR-4, normalized to the geometric mean of three control genes (β-actin, S27 and VATB). Error bars represent standard error. (B) Semiquantitative PCR for the expression levels of Es-EcR-1, Es-EcR-2, and Es-EcR-3. (C) Western blot analysis for Es-EcR-3, and β-actin was as reference protein. PoM: postmolt, InM: intermolt, PrM: premolt. (D) Relative expression level of Es-EcR-3 among negative control siRNA (NC), no injection (Normal), and siRNA groups, normalized to the geometric mean of three control genes (β-actin, S27 and VATB). (E) Lasting days for molting between negative control siRNA (NC) and siRNA groups.

Full size image

qRT-PCR analysis showed that Es-EcR-2, Es-EcR-3 and Es-EcR-4 were expressed in all different tissues from adult crabs, with the highest expression in the hepatopancreas. Es-EcR-2 and Es-EcR-4 expression levels in the ovary were significantly higher than those in other tissues except in the hepatopancreas. Es-EcR-3 was expressed at relatively lower levels in the ovary and was not significantly differentially expressed in the ovary and testes (Fig. 5A). Semiquantitative PCR results revealed that the different EcR isoforms had different tissue expression patterns. Es-EcR-1 was expressed at low levels in the hepatopancreas, eyestalks and stomach, but was expressed at relatively high levels in the muscle, heart, thoracic ganglia, and testis. EsEcR-2 was expressed in all the studied tissues but was highly expressed in the ovary and hepatopancreas (Fig. 5B). The expression level of Es-EcR-2 was increased gradually during the process of ovarian maturation. The expression level of Es-EcR-2 was significantly higher in primary oocyte niche and primary oocyte growth stages than oogonium stage (Fig. 5C).

Figure 5
figure 5

Expression analysis of the four EcR isoforms in different tissues. (A) qRT-PCR analysis of the relative expression levels of Es-EcR-2, Es-EcR-3, and Es-EcR-4 in different tissues, normalized to the geometric mean of 3 control genes (β-actin, S27 and VATB). Error bars represent standard error. Groups marked by different lowercase letters represented significant difference at p < 0.05. (B) Semiquantitative PCR results for the expression levels of Es-EcR-1, Es-EcR-2, and Es-EcR-3 in the different tissues. (C) qRT-PCR analysis of the relative expression levels of Es-EcR-2 during different ovary developmental stages. Groups marked by different lowercase letters represented significant difference at p < 0.05. Hp: hepatopancreas, E: eyestalks, S: stomach, M: muscle, G: gill, H: heart, TG: thoracic ganglia O: ovary, and T: testis.

Full size image

Discussion

In the present study, we first cloned four alternatively spliced EcR isoforms from E. sinensis. Based on the assembled draft genome of E. sinensis, we identified and evaluated the complete gene structures and preliminary functions of the four EcR isoforms. The four isoforms were similar to those in other crustaceans, such as Callinectes sapidus and Marsupenaeus japonicas. Most crustaceans have more than two EcR isoforms (Table 1)33,35,42 and most insects harbor two EcR isoforms (Table 1)17,37. These findings indicate that the crustacean EcR gene has a different selection pressures and evolutionary trajectories compared to insects (Table 1, Fig. 3).

As a NR, the insect EcR has specific structural characteristics: its DBD and LBD are highly conserved but the A/B domain is highly variable in insects. The conserved DBD and LBD region demonstrates the importance roles of the binding sites, such as D. melanogaster 43,44,45,46, Tribolium castaneum 13, Apis mellifera 17, and Chilo suppressalis 23. However, in crustaceans, EcR isoforms have a different type of variation in the hinge region (D region) and LBD (E region) (Table 1, Fig. 2). Alternative E region isoforms are unusual among insect NRs, but are common in crustaceans (Table 1). In this study, we confirmed that this pattern exists in E. sinensis and may be a specific evolutionary pattern in crustaceans (Figs 1, 2, and Table 1). Our results indicated that the four EcR isoforms resulted from the alternative splicing of a single-gene locus and that changeable variant sites likely characterize EcR isoforms in crustaceans.

Insects and crustaceans have EcR isoforms with specific functions. We found that the four isoforms have fluctuating expression patterns during different molting stages and in the different tissues. Previous studies have reported that the functional specificity of isoforms is linked to their discrete physiological functions47,48,49,50. Similar to the expression of ecdysone-responsive genes, the expression of EcR fluctuates and is essential in the regulation of molting in crustaceans38. In our study, Es-EcR-3 was significantly upregulated during PrM. However, Es-EcR-2 and Es-EcR-4 were not differentially expressed during PrM and InM. Furhtermore, knockdown of Es-EcR-3 caused significantly molting delayed. The findings strongly indicate that Es-EcR-3 is a vital molting-regulated isoform in E. sinensis. Our Western blot results confirmed this hypothesis. As a NR, EcR activates gene transcription by binding to specific hormone response elements in the promoters of target genes. The most preferred response element for the EcR/USP (EcR/RXR) heterodimer is a palindrome element51,52. Moreover, the DBD and hinge regions are essential for EcR/USP heterodimerization on palindrome elements53. The Es-EcR-3 lost 12 bp and the entirety of exon located in the D hinge domain, which may affect EcR and RXR heterodimerization and prompt a specific function in molting (Fig. 2B and C). Hinge isoforms are common among NRs. The D domain is important in maintaining the integrity of functional NR structures. Moreover, different hinge isoforms may have different functions54. Similar to Es-EcR-2 and Es-EcR-3, which both have variations in the D hinge domain, all hinge isoforms showed specific functions.

Different isoforms are involved in stage- and tissue-specific regulation. EcR-B1 is predominantly expressed in salivary glands and midgut in Drosophila 55. MnEcR was highly expressed in the hepatopancreas and gills among the ten different examined tissues, and different isoforms predominate in the testis and ovary of the freshwater prawn, Macrobrachium nipponense 33. CrcEcR is expressed in all tissues of the brown shrimp, Crangon crangon, and has high expression levels in the ovary, and low in muscle. However, truncated CrcEcR isoforms are detected only in the ovary56. In our study, Es-EcR-2 expression levels were higher in the ovary than in other tissues, except hepatopancreas (Fig. 5). Es-EcR-2 was also predominatly expressed during InM (Fig. 4), which involves energy storage and specific developmental processes39. This result signifies that Es-EcR-2 is a developmentally related isoform, and may play essential roles in the growth and the regulation of ovary development57. Our qRT-PCR results from different periods of ovary stage also indicated Es-EcR-2 may participate in the course of ovary development. Further studies must be conducted to elucidate the specific functions of Es-EcR-1 and Es-EcR-4.

Several alternative splicing patterns have been reported in previous studies58. 5′AE and exon skipping are common splicing patterns. However, an IR splicing pattern in the LBD (E region) domain was identified in this study. Previous studies also showed that this intron retention pattern is also present in the mud crab, Scylla paramamosain 57. A study on mosquitoes indicated that IR may cause transcript degradation by translation-dependent nonsense-mediated mRNA decay. This process is a potential regulatory mechanism for gene expression in numerous organisms59. Therefore, IR in the Es-EcR-4 gene of E. sinensis may be strongly associated with the gene’s specific physiological function. However, we could not fully elucidate this function in the present study.

In conclusion, this paper is the first to report on the four EcR isoforms in E. sinensis. Our results indicated that the EcR genes of crustaceans had different evolutionary trajectories from those of insects. And findings implicate Es-EcR-3 in the regulation of molting. Moreover, our findings implicate Es-EcR-2 in the regulation of growth and development, and may participate in ovarian development in E. sinensis. Further work must be conducted to interpret the specific function of Es-EcR-4.

Materials and Methods

Animal and tissue collection

Healthy E. sinensis adults were collected from the Nanhui Experiment Station of Shanghai Ocean University (Shanghai, China). Sampling procedures complied with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of Shanghai Ocean University on the care and use of animals for scientific purposes. The experimental protocols were approved by the IACUC of SHOU. Three groups of crabs were sampled under different experimental conditions: (1) Hepatopancraese were sampled at the premolt (PrM), intermolt (InM) and postmolt (PoM) stages39, (2) Nine different tissues, including the heart, gill, muscle, hepatopancreas, stomach, thoracic ganglia, ovary, testis, and eyestalk were sampled from each adult crab at the InM stage, (3) Ovary tissues of crabs in July, August, September, October, and November in 2016 were collected, ovary tissues collected from July, August and September, and October and November are considered to be in oogonium stage, primary oocyte niche stage, and primary oocyte growth stage, respectively, according to previous study60. Six biological replicate samples were selected per experimental condition. All samples were snap-frozen in liquid nitrogen and subsequently stored at −80 °C before DNA and RNA isolation.

DNA and RNA isolation and cDNA synthesis

DNA was extracted from the leg muscle of each sample using the saturated sodium chloride method61. Total RNA was extracted using an AxyPrep Multisource Total RNA MiniPrep Kit (AxyGen, 09113KD1) and purified with RNase-free DNase I (Tiangen, Beijing, China) to remove DNA contaminates from each sample. The quality of the extracted DNA and RNA was evaluated on 1.0% agarose gels that were stained with ethidium bromide. Reverse-transcription-PCR was performed to synthesize complementary DNA (cDNA) using the PrimeSript™ RT reagent Kit (TaKaRa, Dalian, China) in accordance with the manufacturer’s instructions.

Sequence amplification of EcR gene, and of alternatively spliced isoforms and prediction of protein structure

In order to obtain the EcR gene sequence from the recently assembled E. sinensis, we used the EcR mRNA sequence, which was assembled in our previous study, as “query” to BLAST against the assembled E. sinensis genome, then we identified the region and extracted the sequences of EcR gene on the assembled E. sinensis genome62. To amplify and confirm the accuracy of the DNA sequence of EcR, primers were designed based on different exons (Table S1). PCR was performed using an Eppendorf Thermal Cycler (Berlin, Germany). The expected PCR products were sequenced with ABI3730 sequencer (Sangon, Shanghai, China). Sequenced sequences were assembled by Cap3 software and the EcR gene structure was predicted by Fgenesh software and manually corrected based on the assembled E. sinensis genome63.

Two alternatively spliced EcR isoforms were identified in our previous study39. To further confirm and identify novel alternatively spliced EcR isoforms, primers were designed to amplify the whole EcR open reading frame (ORF) (Supplementary Table S1). cDNA from different tissues and molting stages were mixed and used as a template. PCR was performed using an Eppendorf Thermal Cycler (Berlin, Germany) with a 50 μL reaction mixture that contained 2 U DNA polymerase (Tiangen products, Shanghai, China), 5 μL PCR buffer, 2 μL template cDNA (50 ng/μL), 2 μL dNTPs (0.4 mM), 4 μL primers (0.2 μM each), and 35 μM distilled water. PCR conditions were as follows: 94 °C for 5 min; followed by 30 cycles of 94 °C for 30 s, 58 °C for 30 s, and 72 °C for 2 min; and final extension at 72 °C for 10 min. Expected PCR products were ligated to PMD19-T vector and transformed into competent DH5α cells. At least 10 positive clones per product were sequenced by ABI3730 sequencer (Sangon, Shanghai, China). Sequences were manually edited by Bioedit software to remove vector sequences and then aligned using CLUSTAL W64,65. The alternatively spliced EcR isoforms were discovered based on EcR gene sequence alignment. The EcR gene structure of different isoforms was described by Fancygene software66.

The three-dimensional structures of the four isoforms were modelled with the online automated Build Homology Model program at Phyre2 using the crystal structure of the liganded hrxr-alpha/hlxr-beta heterodimer (PDB accession code c4nqaI) as the template (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id = index).

Evolutionary analysis of the EcR gene among insects and crustaceans

EcR gene sequences from insects and crustaceans were downloaded from NCBI database and previously published papers (Table 1). The number of alternatively spliced isoforms and varied domains were recorded in accordance with published papers. A phylogenetic tree was constructed using Maximum Likelihood (ML) methods with Gamma distributed with Invariant sites (G + I) model implemented in MEGA5 software. The phylogenetic tree was constructed based on available EcR sequences from insects and crustaceans with 1,000 bootstraps67.

Quantitative reverse transcription PCR (qRT-PCR) analysis

qRT-PCR was used to quantify the expression level of each EcR isoform in different tissues, different molting stages and ovary from different developmental stages. The isoform-specific qPCR primer pairs for EcR (Es-EcR-2, Es-EcR-3, and Es-EcR-4) were designed based on the different sequences of the isoforms (Supplementary Table S1, Supplementary Fig. S1). Given that no workable qRT-PCR primer could differentiate Es-EcR-1 from other isoforms (Es-EcR-2 to Es-EcR-4), Es-EcR-1 transcript levels could not be determined by qRT-PCR. Semiquantitative PCR was simultaneously conducted for Es-EcR-1, Es-EcR-2, and Es-EcR-3. The isoforms were distinguished based on PCR lengths (Supplementary Table S1). The housekeeping genes, beta-action (β-actin), ribosomal S27 fusion protein (s27), and vacuolar ATP synthase subunit B (vatb) were used as internal control for qRT-PCR. qRT-PCR was performed with a 25 μL reaction mixture that contained 12.5 μL SYBR Green Premix Ex Taq (Takara, Japan), 1 μL each primer (10 μM), 2 μL diluted cDNA and 8.5 μL dd H2O. The PCR procedure used the following program: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s, and 60 °C for 30 s. Temperature increased by 0.5 °C/5 s from 60 °C to 95 °C for the melting curve with 30 s elapse time per cycle. The relative expression was estimated using the 2−∆∆Ct method with InM in different molting stage, heart in different tissue, and ovary in November stage as calibration control68, the results were present as the fold-change relative to InM, heart, and ovary in November, respectively. Statistical significance (P < 0.05) was determined using Student t-tests with SPSS 17.0.

Es-EcR-3 polyclonal antibody construction and Western blot analysis

Es-EcR-3 polyclonal antibody was made according to prokaryotic expression, recombinant protein induction, and polyclonal antibody production procedure. First, the whole Es-EcR-3 ORF was amplified and inserted into the prokaryotic expression vector, pET-28a; then the expression vector was sent to HuaAn Biotechnology Company (Hangzhou, China) for further recombinant protein induction and polyclonal antibody production. For Western blot analysis, 100 μg total proteins were analyzed in 10% SDS–PAGE (10% acrylamide/bisacrylamide; 350 mM Tris-HCl (pH 8.8); 0.1% SDS; 0.1% ammonium persulfate; TEMED). After electrophoresis, the proteins were transferred onto a polyvinylidene fluoride membrane. The protein blots were blocked with 5% nonfat milk in TBST for 2 h at room temperature (25 °C). Then the membranes were incubated overnight at 4 °C with polyclonal antibodies against Es-EcR-3 (1:1000, made by Hangzhou HuaAn Biotechnology Company) and a control β-actin (1:5000), which was bought from CWBIO Company (Beijing, China). The secondary antibody was anti-rabbit IgG at a dilution of 1:5000. Immunoreactivity was detected with an enhanced chemiluminescence detection kit in accordance with the manufacturer’s instructions (Bio-Rad, Shanghai).

RNA interference analysis

siRNA was designed based on the specific region of Es-EcR-3 isoform, and was synthesized by GenePharma Biotech Company (Shanghai, China), which consisted of 21 nt sense and antisense oligonucleotides with 2′ O-Methyl oligo modification. Both siRNA and negative control siRNA were designed (Table S1). In order to test the efficiency of designed siRNA, six crab individuals were injected with siRNA (4 μl, 1 μg/μl), six crab individuals were injected with negative control siRNA (4 μl, 1 μg/ul), respectively, and six crab individuals without injection were used as experimental control. qRT-PCR was conducted on Es-EcR-3 according to method above (negative control siRNA group was used as calibration control). In order to study the biological function of Es-EcR-3 on molting, 30 crab individuals (1 ± 0.1 g) were collected immediately after their molting at the same day. Therefore, all the collected 30 crab individuals were at the same molting stage. The 30 crab individuals were randomly divided into two groups, siRNA and negative control siRNA (NC) groups, and were cultured in the same condition. When the crab individuals were at intermolt (InM) stage, 4 μl siRNA (1 μg/μl) and 4 μl negative control siRNA (1 μg/μl) were injected into siRNA and NC groups, respectively, every four days. The molting time was observed and recorded.