All mud crabs with good vitality were purchased from the Jimei market, Xiamen city, Fujian province. The eyestalk of immature female crabs weighing about 100 g was used to isolate nucleoproteins. Female crabs weighing about 250 g were temporarily kept in the seawater field of Jimei University for a week before the RNA interference experiment. After the experiment, the eyestalks, gonads, and hepatopancreas of the crabs were sampled and frozen immediately in liquid nitrogen, then stored at −80°C for total RNA extraction. All animal experiments were conducted following the regulations of the Guide for Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of Jimei University.

In order to detect whether there is a specific binding site for Oct4 or Sox9 at the core regulatory region of SpVih, EMSA experiments were performed. Briefly, according to the manufacturer's protocol for the LightShift Chemiluminescent EMSA Kit (Thermo Scientific, Waltham, MA, USA), biotin-labeled and unlabeled double-stranded DNA probes containing intact core recognition element (CRE) for Oct4 and Sox9 were prepared. For competitive analysis, the mutant probe was synthesized. Nuclear proteins were extracted from the eyestalk of immature female crabs using NE-PER Nuclear and Cytoplasmic Extraction Reagent (Thermo Scientific). In the competition experiment, we set up a 100-fold amount of mutation probe group and a 10- and 100-fold amount of unlabeled probe group, respectively. Supershift group added Oct4 or Sox9 antibodies (Abcam, Cambridge, UK, ab19857 and ab3697). Protein-DNA complexes were separated on 6% polyacrylamide gels in 1× TBE via electrophoresis at 100 V for 1 h and transferred onto nylon membranes. Chemiluminescence detection was carried out according to the Chemiluminescent Nucleic Acid Detection Module kit (Thermo Scientific).

For overexpression studies, the SpVih core promoter fragment was amplified by PCR, and then cloned into the Kpn I-Xho I site of the pGL3-basic vector. The full-length ORF of Oct4 and Sox9 was cloned into the pcDNA3.1 vector, respectively. All plasmid constructs were verified by sequencing analysis. The primers with the restriction enzyme cutting sites were listed in Table 1.

HEK293T cell line was cultured in the DMEM medium supplemented with 10% FBS (fetal bovine serum, Gibco, Waltham, USA) and 1% (v/v) antibiotic (penicillin/streptomycin, Gibco, Waltham, USA) at 37°C in a humidified incubator with 5% CO₂. The medium was changed upon the state of cells.

Transfection of plasmids in HEK 293FT cells were performed using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to manufacturer's instructions. pRL-TK plasmid (10 ng) was used as an internal control. The transfection reagent and plasmids (Oct4 and Sox9 at 50 ng, 100 and 200 ng, respectively, or 25 ng Oct4 + 25 ng Sox9, 50 ng Oct4 + 50 ng Sox9, and 100 ng Oct4 + 100 ng Sox9) were mixed and co-transfected into the HEK293T cell lines. At 3–6 h post-transfection, the transfection mixture was replaced with DMEM containing 10% FBS. The pEGFP-N1 plasmid was employed as a positive control. After continuing culture for 24 h, the cells were harvested. The firefly and renilla luciferase activities were measured by the Dual-Luciferase Reporter Assay System (Promega, Madison, USA) according to the manufacturer's protocol. The promoter activity was assessed by the average of firefly luciferase activity normalized as a ratio to the renilla luciferase activity. Each experiment was conducted for three independent samples in triplicate.

In in vivo experiments, the 286 and 275 bp-length ORF of Oct4 and Sox9, respectively, were chosen as the templates for synthesizing gene-specific dsRNA. Enhanced green fluorescent protein (EGFP) gene was used as a negative control group, and its 229 bp-length ORF was selected as a template for synthesizing EGFP dsRNA. These dsRNAs were synthesized from the linearized template by using the T7 RiboMAXTM Express RNAi System (Promega). These dsRNAs were uniformly diluted to 1 μg/μL, and each crab was injected with 10 μL of dsRNA on the base of each eyestalk. The blank control group was injected with the same volume of stroke-physiological saline solution (SPSS). After continuous injection for 3 days, the eyestalks, gonads, and hepatopancreas were dissected on the fourth day and placed in RNA later and then transferred to −20°C overnight for RNA extraction.

Total RNA was extracted from different tissues using Total RNA Extraction Kit (Promega, Shanghai, China) according to the manufacturer's protocol. Total RNA quality was assessed by agarose gel electrophoresis and Nanodrop 2000 (Thermo Scientific). The complementary DNA (cDNA) was synthesized in a 20 μL reaction system including 1 μg total RNA (previously treated with DNase I), 2 μL random primers (10 mM), 4 μL 5× First-strand Buffer, 1 μL dNTP mix (10 mM), and 1 μL M-MLV reverse transcriptase (200 U/μL) (Promega, Shanghai, China). The synthesized cDNA was diluted and stored at −20°C until use.

qRT-PCR was carried out in a LightCycler480 Roche Realtime Thermal Cycler (Mannheim, Baden-wurttemberg, Germany) in accordance with the manual with a 20 μL reaction volume containing 9 μL of 1:100 diluted original cDNA, 10 μL SYBR Green Master Mix (Promega, USA), 0.5 μL of the forward primer and 0.5 μL of reverse primer (10 mM). Primer sequences are shown in Table 1. 18s rRNA was used as the reference gene. The cycling conditions for PCR reaction were set as follows: 1 min at 95°C, followed by 40 cycles at 95°C for 15 s, 59°C for 1 min. Quantitative measurements were performed using the ΔΔCt method. Each sample had at least 4 biological replicates. Statistical analysis was performed by one-way analysis of variance (one-way ANOVA) using Statistical Package for the Social Sciences 20.0 (SPSS 20.0) software (IBM Corporation, New York, NY, USA). The statistically significant differences were shown at p < 0.05; the most significant differences were shown at p < 0.01.

According to the full-length cDNA sequences of Oct4 and Sox9, two primer pairs were designed to amplify the sequences that encode their corresponding fragments, respectively (Table 1). The amplified fragment of Oct4 was digested by BamHI and XhoI, and then inserted into a pGEX-4T-1 vector. Meanwhile, the Sox9 fragment was ligated into a pET30a expression vector after being cut by NdeI and HindIII. Two recombinant expression vectors, namely pGEX-4T-Oct4, pET30a-Sox9, were transformed into competent Escherichia coli Rosetta (DE3) host cells. Isopropyl-b-D-thio-galactoside (IPTG) was added to induce protein expression. The recombinant Oct4 protein with a glutathione S-transferase (GST) tag was purified using glutathione Sepharose 4B chromatography (Detai Biologics, Nanjing, China) according to the manufacturer's instructions. Sox9 protein with an His-tag was harvested by His Bind resin chromatography (Detai Biologics) according to the manufacturer's instructions.

Approximately 15 μg of purified recombinant Oct4 was incubated with glutathione sepharose for 1 h at 4°C and then washed with binding buffer (1× PBS, pH = 7.4). Afterward, ~15 μg of purified recombinant Sox9 was added and then incubated for 1 h at 4°C. After being washed thoroughly with binding buffer (1× PBS, pH = 7.4), the proteins were eluted with elution buffer (30 mM GSH, 50 mM Tris, 0.1% Triton X-100) and then finally analyzed by SDS-PAGE and Western blot. GST protein was used as the negative control in this assay.

To determine whether the transcription factor Oct4/Sox9 can recognize and bind to the corresponding site of SpVih promoter, the EMSA experiment was conducted after incubation with eyestalk nuclear extract and probe. The results showed that the wild-type probe SpTF-wt (Figure 1A) could combine with the eyestalk nuclear extract to form a DNA-protein complex (Figure 1B, lane 3, arrow marked). The formation of this complex can be competitively bound by cold probe (wild type unlabeled SpTF-wt). The binding band became shallow when it combined with 10-fold cold probe. When the amount of cold probe was 100-fold, the DNA-protein complex binding band is entirely undetectable (Figure 1B, lane 5), indicating that this complex is specific to SpTF-wt. When 100-fold unlabeled SpTF-mut (mutated the binding site of Oct4 and Sox9) and SpTF-wt coexisted, the complex formation was not affected, indicating that the mutated probe SpTF-mut did not participate in competitive binding.

To confirm the Oct4 and Sox9 binding to the promoter of SpVih, EMSA supershift was carried out using both biotin-labeled and unlabeled DNA probes containing intact CRE for Oct4 and Sox9. EMSA supershift with crab eyestalk nuclear extract demonstrated a DNA-protein complex formation with a biotin-labeled probe containing Oct4 and Sox9 motifs (Figure 1C, lane 2). Further, Oct4 and Sox9 antibody's addition resulted in the supershift of protein–DNA complex (Figure 1C lanes 3–4, arrow marked).

Different concentrations of Oct4 and Sox9 overexpression plasmids were co-transfected with the SpVih core promoter-reporter plasmid SpVih-4 into the HEK293T cell line, respectively. PGL3-basic was used as a negative control, while pRL-TK was used as the reference vector. The cells at 24 h after transfection were collected to detect reporter gene carrier activity. The results showed that the addition of either 50 ng Oct4 overexpressed plasmid or 100 ng Sox9 overexpressed plasmid could significantly promote the activity of SpVih, and the SpVih activity increased with the increment of the overexpressed plasmid concentration (Figures 2A,B). When HEK293T cells were co-transfected with two overexpression plasmids and SpVih fragment reporter plasmid at the same time, the activity of SpVih increased significantly. It showed a dose-effect with the concentration of transcription factors (Figure 2C). Compared with a single transcription factor overexpression plasmid group, co-transfection with two overexpression plasmids can significantly upregulate the activity of SpVih (Figure 2D).

To further verify the regulation of SpVih by Oct4 and Sox9 in cell experiments, we designed an in vivo RNA interference experiment. After 3 consecutive days of dsOct4 and dsSox9 injections, respectively, compared with the control group, the relative expression of Oct4 or Sox9 in the eyestalk of the experimental groups was significantly reduced (Figures 3A,B), indicating that the experiment successfully interfered with the expression of Oct4 or Sox9 in the crabs. Compared with the blank control group and the negative control group, the expression of SpVih of the eyestalk in the experimental group injected with dsOct4 or dsSox9 was significantly decreased (Figure 3C), indicating that interference with Oct4 or Sox9 can inhibit SpVih expression in the crab.

Vitellogenin (Vtg) is the precursor of Vitellin (Vn or Vt) and is the main component in the yolk of oviparous animals. VIH in crustaceans can inhibit the production of Vtg and regulate ovarian development and maturation. The qRT-PCR experiment was conducted to investigate the expression of SpVih in the eyestalk and the expression of SpVtg in ovary and hepatopancreas after the injection with dsOct4 and dsSox9. The experimental results showed that, compared with the blank control group and the negative control group, when dsOct4 and dsSox9 were injected into the eyestalk of the crab, the expression of SpVih in the eyestalk decreased significantly while the relative expression of SpVtg in the ovary and hepatopancreas had a significant rise (Figure 4).

Oct4 was expressed as a soluble protein after IPTG induction. It was conveniently purified by glutathione sepharose 4B chromatography. The purified protein comprised of an ~44 kDa Oct4 protein and a nearly 26 kDa GST tag expressed by the plasmid pGEX4T1, which was roughly consistent with the size (about 70 kDa) of the unique band in the purified protein lane (Figure 5A). Sox9 was purified by His-bind resin chromatography. Because it contained an additional His Tag, the Mw values of the purified Sox9 was larger than its theoretical 53 kDa, which was approximately consistent with the SDS-PAGE result (Figure 5B).

The binding activity of Oct4 with Sox9 was examined by a pull-down assay. The results are shown in Figure 6A. In the SDS-PAGE electrophoresis detection, lane 3 is the control group with GST instead of pET30a-Sox9, and only two protein bands are GST and GST-Oct4; and the lane 4 is the experimental group with pET30a-Sox9, three protein bands are GST, GST-Oct4, and His-Sox9. GST-Oct4 and His-Sox9 were detected by anti-his polyclonal antibody in Western blot. The eluent showed that bands of the same size as His-Sox9 were detected in the experimental group in which pET30a-Sox9 was added (Figure 6B). Both results illustrated that pGEX-4T-Oct4 can interact with pET30a-Sox9 in vitro, while the negative control GST cannot bind to pET30a-Sox9.