As shown in Table 1 , three isonitrogenous and isolipidic experimental diets were formulated to contain approximately 50% crude protein and 10% crude lipid. The basal diet used casein (vitamin free), gelatine, and crystalline amino acid as the primary protein sources, while fish oil, soya bean oil, and soya bean lecithin were used as dietary lipid sources. VD₃ (V8070, Solarbio, China) was added to the basal diet to provide graded concentrations of 0 (D₀), 1000 (D₁), and 4000 (D₂) IU/kg VD₃. The measured values of VD₃ concentration were determined by the high-performance liquid chromatography (HPLC) method as described previously (33), which were 32 (D₀), 1012.60 (D₁), and 3978.2 (D₂) IU/kg. Diets were extruded with an experimental single-screw feed mill (Yihe, China) in the form of 3 mm diameter pellets and dried for 12 h in a ventilated oven at 50°C. All the experimental diets were stored at −20°C until use.

Disease-free juvenile turbots were purchased from a local farm in Weihai, Shandong Province, China. Before the start of the feeding trial, the fish were maintained for two weeks to acclimate to the experimental conditions. A commercial diet (Surgreen, China) was fed during the acclimation, which is specially formulated for the nutritional requirements of the turbot. The fish were then fasted for 24 h and weighed. A total of 270 fish (12.17 g initial body weight) were randomly assigned to 9 fiberglass tanks (300 L, 30 fish per tank) connected to an indoor flow-through water system. Triplicate tanks of fish were fed with each experimental feed to apparent satiation twice daily for 12 weeks. During the feeding period, water temperature ranged from 15°C to 18°C; salinity 30-33‰; and dissolved oxygen higher than 7.0 mg/L.

Three fish were selected randomly from each tank for the sampling. All fish were anesthetized with eugenol (1:10000, Shanghai Reagent Corp, China) before handling. The intestinal tissue samples for the analysis of enzymes activities, DNA laddering, gene expression, and western blot were frozen in liquid nitrogen, and the others for the analysis of histology were fixed in Bouin’s fixative solution.

According to our previous study (34), the isolation and culture of primary intestinal epithelial cells were performed. Briefly, the starved turbots were anesthetized with eugenol (1:10000, Shanghai Reagent Corp, China), wiped, and disinfected with 75% ethanol. Then the intestine tissue was dissected and rinsed repeatedly with the solution of Hank’s Balanced Salt Solution (HBSS) containing penicillin-streptomycin (Thermo Fisher Scientific, USA). About 1 mm³ of intestinal tissues were dissociated with collagenase and dispase (Thermo Fisher Scientific, USA) for 15 to 20 min. The enzyme solution was washed several times with L-15 medium (L5520, Sigma, USA), and the supernatants were collected and centrifuged for 5 min at 1000 rpm to obtain intestinal epithelial cells suspended in the supernatants.

The intestinal epithelial cells were seeded into 6-well plates at a density of 1.0×10⁶ cells/mL with a modified L-15 medium (L5520, Sigma, USA) containing 5% fetal bovine serum (FBS, Biological Industries, Israel), 10 ng/mL epidermal growth factor (Sigma, USA), 0.2% insulin transferrin-selenium-sodium pyruvate (Thermo Fisher Scientific, USA), antifungal, and antibacterial substances at 23°C in an incubator (CO₂-free). After the cells adhered to the well for 24 h, the cells were treated with fresh medium together with 0, 1, 10, and 50 nM 1,25(OH)₂D₃ (HY-10002, Med Chem Express, USA). After 1,25(OH)₂D₃ treatments for 24 h, the cell lysates were subjected to RNA and protein extraction, while the cell climbing slides were fixed with 4% paraformaldehyde for TUNEL assay and MAP1LC3B immunofluorescence.

Three pairs of VDR-specific siRNAs and scrambled siRNA for negative control (NC siRNA) set up through BLOCK-iT™ RNAi Express are shown in Table 2 and Figure S , and synthesized by T7 RNAi Transcription Kit (TR102, Vazyme, China) according to the instructions. Seven groups of 9 turbots acclimating to an indoor flow-through water system were injected with VDR-specific siRNAs (1 μg/g) and NC siRNA (1 μg/g), and the intestinal tissues were drawn to analyze the gene expression of VDRA&B gene expression after 24 h. Three turbots were mixed together to make a sample. And the most effective siRNA of VDRA&B was selected for the formal experiment.

The formal RNA interference (RNAi) experiment lasted for two days. The body weight of the fish decided the dose of siRNA (1 μg/g). Five groups of 9 turbots were injected with equal volume (100 μl) of PBS, PBS, NC siRNA, VDRA siRNA, and VDRB siRNA on the first day, respectively. After 24 h, they continued to be injected with the same contents as the first day. In addition to the first group still injected with PBS, the last four groups were injected with lipopolysaccharides (LPS) at the rate of 2.5 μg/g body weight. They were defined as PBS, LPS, NC siRNA+LPS, VDRA siRNA+LPS, and VDRB siRNA+LPS group. The turbots were anesthetized by eugenol (1:10000, Shanghai Reagent Corp, China) and had their intestinal tissue taken for qRT-PCR and western blot.

Standard methods (AOAC, 1995) were used for analyzing experimental diets. Moisture content was determined gravimetrically to constant weight in an oven at 105°C. Crude protein was determined by the Kjeldahl method using Kjeltec 2300 (Foss, Denmark) using boric acid to trap released ammonia. Crude lipid was determined by petroleum ether extraction using Soxhlet Extraction System B-811 (Buchi, Switzerland). Ash was determined by combustion at 550°C.

After being fixed in Bouin’s fixative solution for 24 h, the intestinal samples were transferred to 70% ethanol and embedded in paraffin after dehydration. Paraffin sections of 5 μm were cut and stained with hematoxylin and eosin (H&E, G1120, Solarbio, China) according to the manufacturer’s protocol. The slides were examined under a light microscope (BX43F, Olympus, Japan) equipped with a digital microscope camera (DP72, Olympus, Japan) for image acquisition.

For the analysis of intestinal alkaline phosphatase (ALP), acid phosphatase (ACP), and lysozyme (LYZ) activities, the intestinal samples were homogenized in ice-cold physiological saline solution (1:9) and centrifuged at 2500 g for 10 min at 4°C. The relevant intestinal enzyme activities were then determined with supernatants, whose protein concentrations were measured by the bicinchoninic acid (BCA) protein analysis kit (P0012, Beyotime, China). The activities of ALP (A059-2), ACP (A060-2), and LYZ (A050-1) were quantified using the commercial kits according to the manufacturer’s protocol (Jiancheng Bioengineering Institute, China).

Briefly, 5 mg intestinal tissues were homogenized in liquid nitrogen and then mixed with 0.5 ml DNA extraction lysis buffer (pH 8.0) containing 50 mM Tris-HCl (Solarbio, China), 25 mM EDTA (Solarbio, China), 100 mM NaCl (Macklin, China), 1% Triton X-100 (Solarbio, China), and 0.5 mg/ml Protease K (Solarbio, China). The lysates were incubated overnight at 50°C with gentle shaking and then mixed with an equal volume of phenol-chloroform-isoamyl alcohol (25:24:1) (Solarbio, China). After being precipitated with 5 M ammonium acetate (Solarbio, China) and absolute ethanol (Sangon, China), the DNA precipitations were washed three times with 0.6 ml 70% ethanol and air-dried at room temperature (RT). The DNA precipitations were dissolved with Tris-EDTA buffer (Solarbio, China) containing 100 µg/ml RNase A (Solarbio, China) and incubated at 30°C for 30 min. Approximately 5 µg of DNA samples were electrophoresed (100 V) on a 2% agarose gel and were visualized with Gel-Red (TransGen Biotech, China) and recorded under UV light with an Odyssey Infrared Imaging System (Li-Cor Bioscience, USA). Trans2K^® Plus DNA Marker (BM111, TransGen, China) of nuclei acid was used as the reference. The evaluation of DNA fragmentation was determined according to Yuan et al. (35). The density of the 180 to 200 bp DNA band was quantified using ImageJ software (National Institutes of Health, USA), and the relative density was normalized to the D₀ group.

TMR (red) Tunel Cell Apoptosis Detection Kit (G1502, Servicebio, China) was used to detect positive apoptotic nuclei of cell climbing slides. The cell climbing slides were dried and then incubated in the permeabilizing work solution for 20 min at RT. Rinse with PBS solution three times, each for 5 min. After the slices were slightly dried, the equilibration buffer was added to the tissues and incubated for 10 min at RT. After washing with PBS solution three times (5 min per time), the TUNEL reaction solution mixture (TDT enzyme, dUTP, and buffer at 1:5:50 ratio) was added to objective tissue placed in a flat wet box, incubated for 2 h at 37°C. 4’,6-diamidino-2-phenylindole (DAPI) was used to stain the nucleus and washed out three times with PBS in a rocker device.

Immunofluorescence was used for detecting MAP1LC3B in climbing slides. The cell membrane rupture was conducted as mentioned above, and then the slides were incubated in 5% Bovine Serum Albumin (BSA) for 30 min at RT. MAP1LC3B antibody (AF5225, Beyotime, China) was added to the sections, and the slides were incubated overnight at 4°C. After being washed with PBS solution three times (5 min per time), the tissues were covered with the secondary antibody and incubated for 50 min at RT. The nucleus was stained by DAPI and washed three times with PBS.

The sections were observed under a fluorescence microscope, and images were collected. (DAPI UV excitation wavelength 330-380 nm, emission wavelength 420 nm, blue light emission; CY3 excitation wavelength 510-561 nm, emission wavelength 590 nm, red light emission).

RNA was extracted from the intestinal samples with Trizol Reagent (Takara, Japan). The integrity of RNA was detected by electrophoresis on 1% denaturing agarose gel, and the concentration was detected with a Nano Drop^®2000 spectrophotometer (Thermo Fisher Scientific, USA). A total of 1 μg RNA was reversely transcribed to cDNA with Evo M-MLV Mix Kit with gDNA Clean for qPCR [#AG11728, Accurate Biotechnology (Hunan) Co., Ltd., China].

The gene expression of VD₃ receptor A and B (VDRA&B), interleukin 1 beta (IL1B), interleukin 8 (IL8), tumor necrosis factor (TNF), and interferon gamma (IFNG), apoptosis-associated speck-like protein containing a caspase recruitment domain (ASC), transforming growth factor beta 1 (TGFB1), and interleukin-10 (IL 10), B cell lymphoma 2 (BCL2), caspase 3 (CASP3), microtubule-associated protein 1 light chain 3 beta (MAP1LC3B), and autophagy-related 16 like 1 (ATG16L1) was tested in the present study. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and (RPSD) RNA polymerase II subunit D were used as the housekeeping genes. Specific primers for target genes and housekeeping genes ( Table 3 ) were synthesized by Sangon (China), and the application efficiency was then assessed. Quantitative PCR was conducted in an ABIPRISM 7500 Instrument (Applied Biosystems, USA) with ChamQ Universal SYBR qPCR Master Mix (#Q711, Vazyme, China). The relative gene expression of genes was calculated using the 2^–ΔΔCT method (36).

The intestinal tissues were dissolved with RIPA lysis buffer (Solarbio, China) with the protease and phosphatase inhibitor (Roche, Switzerland). After centrifuging at 12000 g for 20 min at 4°C, the supernatants of the lysates were collected as the total proteins. The nuclear proteins of intestinal tissue were extracted using NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific, USA), as the manufacturer’s protocol showed. And then, the protein concentrations of the samples mentioned above were measured by the bicinchoninic acid (BCA) protein analysis kit (P0012, Beyotime, China). The standardized samples were mixed with Omni-Easy™ Protein Sample Loading Buffer (EpiZyme, China). A total of 20 µg of protein was loaded and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The proteins in the gel were transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, USA) for 1 h at 70 V, followed by membrane blocking at RT for 2 h with 5% non-fat milk (Sangon, China) in tris-buffered saline with Tween 20 (TBST). The membrane was incubated with primary antibodies overnight at 4°C and then washed three times for 5 min each with TBST. Next, the membrane was incubated with horseradish peroxide (HRP)-conjugated secondary antibody (A0208, Beyotime, China) dissolved with 1% non-fat milk in the TBST for 1 h at RT. After washing with TBST 3 times for 5 min, the membrane was developed with enhanced chemiluminescence (Vazyme, China) according to the manufacturer’s directions. The blots were recorded with an Odyssey Infrared Imaging System (Li-Cor Bioscience, USA). The following antibodies were used: antibodies against ASC (WL02462, Wanleibio, China), NF-κB p65 (8242, CST, USA), phos-IκBα ^{ser32 ser36} (2859, CST, USA), IκBα (WL01936, Wanleibio, China), Cleaved-caspase 3 (WL02117, Wanleibio, China), BCL2 (WL01556, Wanleibio, China), MAP1LC3B (AF5225, Beyotime, China), ATG16L1 (AF6252, Beyotime, China), Lamin B (AF1408, Beyotime, China), GAPDH (AB-P-R001, GoodHere, China). All the band intensities were quantified using ImageJ software (National Institutes of Health, USA). Respectively, the densities of total protein bands and nuclear proteins were normalized to GAPDH or Lamin B, which served as internal controls.

Data were analyzed by one-way analysis of variance (ANOVA) using IBM SPSS Statistics for WINDOWS Version 22.0. Tukey’s test was used to compare the means among individual treatments. Differences were regarded as significant when P < 0.05, and the results are presented as means ± standard deviation (SD).