Data on the growth performance and morphological parameters of grass carp are presented in Table 1. After 8 weeks, the SGR and WGR of grass carp in the two groups fed the high-fat diet signifcantly increased compared to those in the control group (P < 0.05). The grass carp in the HF+B. subtilis group showed the highest WGR (P < 0.05). Furthermore, the HF group and HF+B. subtilis group had a higher FI and lower FCR compared to the control group (P < 0.05). Lastly, no signifcant diference was observed in SR, CF, HSI, and VSI among all the experimental groups (P > 0.05).
Item Con HF HF + B. subtilis SR 100.00 ± 0.00 100.00 ± 0.00 100.00 ± 0.00 IBW 50.39 ±1.69 50.15 ± 1.44 50.19 ± 1.89 FBW 83.78 ±1.92a 92.73 ± 3.22b 95.42 ± 1.85b WGR 66.30±2.14a 84.90 ± 1.80b 90.19 ± 3.56c SGR 0.91 ± 0.03a 1.10 ± 0.02b 1.15 ± 0.04b FI 69.27 ± 3.75a 77.65 ± 2.94b 79.32 ± 1.9b FCR 2.08 ± 0.14b 1.82 ± 0.05a 1.75 ± 0.04a CF 1.82 ± 0.12 1.93 ± 0.11 1.87 ± 0.14 HSI 1.92 ±0.22 1.96 ± 0.27 1.90 ± 0.16 VSI 9.1 ± 0.51 8.95 ± 0.92 9.03 ±0.73 Values are the mean ±SE (n= 3 tanks on growth performance and feed intake, and n = 6 on morphological parameters). Values within the same row with diferent letters are signifcantly diferent (P < 0.05)
SR survival rate (%)=100×(fnal fsh number)/(initial fsh number), IBW (g · fsh-1) initial mean body weight, FBW (g · fsh-1) fnal mean body weight, WGR weight gain rate (%)= (fnal mean body weight-initial mean body weight)/initial mean body weight, SGR special growth rate (% d-1) =100×(ln (fnal mean body weight)- ln (initial mean body weight))/days, FI (g · fsh-1) feed intake, FCR feed conversion rate =feed intake/weight gain, CF condition factor= 100%×(body weight, g)/(body length, cm)3, HSI hepatosomatic index = 100%×(liver weight)/(body weight), VSI viscera index =100% × (viscera weight)/(body weight)
Table 1. Efects of the diferent kinds of feed on growth performance, feed intake and morphological parameters
Results of oil-red O staining are presented in Fig. 1a. Lipids are red colored and nuclei are blue colored after staining with oil-red O. The images show that the amount and volume of lipid droplets stained by oil-red O were higher in the HF group than those in the Con group, but lower than the HF + B. subtilis group. The relative areas of lipid droplets stained with oil-red O (Fig. 1b) were consistent with the results of hepatic lipid accumulation. As shown in Fig. 1c, the hepatic lipid content of the HF + B. subtilis group was between the control group and HF group, and was significantly different from the control group and HF group (P < 0.05).
Figure 1. Efect of B. subtilis on hepatic lipid accumulation in grass carp: (a) efect of the B. subtilis diet and high-fat diet on histochemistry (oil-red O staining) (original magnifcation×400, bars 50 μm); (b) the relative areas of the lipid droplets in oil-red O stained grass carp were analyzed by Image-Pro Plus 6.0; (c) efect of the B. subtilis diet and high-fat diet on the lipid content in the liver of grass carp. Values are the mean± SE (n=6). Diferent letters indicate signifcant diferences among groups (P < 0.05)
As shown in Table 2, analysis of key indicators for lipid metabolism showed that the serum CHO and LDL-C levels in the HF group significantly increased compared to the control group (P < 0.05). However, in the HF + B. subtilis group, the serum CHO and LDL-C levels decreased to nearly the same level as the control group (P > 0.05). Moreover, fish fed the high-fat diet with B. subtilis had lower serum AST than fish fed the high-fat diet (P < 0.05). There was no significant difference in the serum TG, HDL-C content, ALT, TP, and Alb among the control group, HF group, and HF + B. subtilis group (P > 0.05).
Item Con HF HF + B. subtilis CHO (mmol·L-1) 5.94 ± 0.26a 6.71 ± 0.22b 6.08 ± 0.19ab TG (mmol·L-1) 5.57 ±0.34 5.68 ± 0.24 5.58 ± 0.24 HDL-Cmmol·L-1) 2.23 ±0.07 2.50 ± 0.15 2.26 ± 0.11 LDL-C(mmol·L-1) 2.63 ± 0.06a 2.98±0.13b 2.60 ± 0.10a TP (g·L-1) 23.52 ± 0.78 22.61 ± 0.97 24.17 ± 0.83 Alb (g·L-1) 15.45 ± 0.51 14.73 ± 0.55 14.18 ± 0.49 AST (U·L-1) 61.48 ± 2.24a 79.81 ± 4.26b 65.46 ± 2.97a ALT (U·L-1) 7.09 ±0.27 7.48 ± 0.22 7.31 ± 0.18 Values are the mean ± SE (n=6). Values within the same row with diferent letters are signifcantly diferent (P < 0.05)
CHO cholesterol, TG plasma total triglyceride, HDL-C high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, TP total protein, Alb albumin, AST aspartate aminotransferase, ALT alanine aminotransferase
Table 2. Efect of B. subtilis on serum biochemistry indicators
As shown in Fig. 2, the groups displayed diferences in relative mRNA expression of lipid metabolism genes and transcription factors. The L-FABP was signifcantly up-regulated in the HF group (P < 0.05) with no signifcant diference in the HF + B. subtilis group compared to the control group and HF group (P > 0.05), respectively. The FAS, ACCα, SCD, LPL, CPTIα1a, SREBP-1c, PPARγ and PPARα were signifcantly down-regulated after fsh were fed the highfat diet for 8 weeks (P < 0.05). Moreover, the expression of FAS was signifcantly down-regulated and the expression of CPTIα1a was signifcantly up-regulated in the HF + B. subtilis group compared with the HF group (P < 0.05). The expression of LPL and PPARα in the HF + B. subtilis group was regulated to nearly the same level as the control group (P > 0.05).
As shown in Fig. 3, in the HF + B. subtilis group, GSH showed a signifcant increase compared to the HF group (P < 0.05), and the T-AOC higher than the control group (P < 0.05). H2O2 and MDA levels in HF + B. subtilis group were lower than in the HF group (P < 0.05) with no signifcant diference in the control group.
Efect of B. subtilis on growth performance and morphological parameters
Efect of B. subtilis on hepatic lipid accumulation in grass carp
Efect of B. subtilis on serum biochemistry indicators of grass carp
Efect of B. subtilis on the expression of genes associated with lipid metabolism
Efect of B. subtilis on the MDA and H2O2 contents and activities of antioxidant enzymes
The bacteria used in this study were isolated from the gut of grass carp. The isolated bacterial strain was identifed as B. subtilis Ch9 and stored in the Laboratory of Aquatic Animal Medicine, Fisheries College of Huazhong Agricultural University. B. subtilis was inoculated onto a Luria–Bertani (LB) agar plate, then incubated for 24 h at 37 ℃. A single clone was selected and inoculated into LB broth and cultured in a shaker at 37 ℃ for 3 days. Bacterial cells were harvested by centrifugation at 3500 rpm for 15 min. The supernatant was discarded. The pellet was re-suspended in sterile phosphate-bufered saline (PBS) to remove metabolic waste from the bacterial solution. It was then diluted.
Three diets were prepared and used in this study control (basal) diet (4% lipids of the dry matter), high-fat diet (8% lipids of the dry matter), and high-fat diet supplemented with B. subtilis (8% lipids of the dry matter). The composition and chemical analyses of the three diets are presented in Table 3. All ingredients were purchased from Hubei Haida Feed Co., Ltd. (Wuhan, China). After ingredients were thoroughly mixed (including B. subtilis solution), pellets with a diameter of 2 mm were produced by a granulator in 30 min. After the pellets were air dried, they were stored in a freezer at - 20 ℃ until use. Before use, the survival of B. subtilis was 1 × 107 CFU g-1 as determined by the plate counting method.
Experimental diets Basal diet High-fat diet High-fat diet + B. subtilis Ingredients (g·kg-1) Fish meal 80 80 80 Soybean meal 240 240 240 Rapeseed meal 340 340 340 Wheat flour 250 250 250 Soybean oil 6 46 46 Ca(H2PO4)2 20 20 20 Vitamin premixa 1 1 1 Mineral premixb 3 3 3 NaCl 2 2 2 Choline chloride 2 2 2 Sodium carboxymethyl cellulose 56 16 16 Total 1000 1000 1000 Proximate composition Crude protein (% DM) 29.83 29.66 29.61 Crude lipid (% DM) 4.39 8.27 8.29 Moisture (% DM) 10.16 9.95 9.93 Ash (% DM) 8.32 8.12 8.17 aVitamin premix (mg·kg-1): vitamin A 6500 IU, vitamin D3 4500 IU, vitamin C 120 mg, vitamin E 25 mg, vitamin K3 5 mg, vitamin B1 12.5 mg, vitamin B2 12.5 mg, vitamin B6 15.0 mg, vitamin B12 0.025 mg, niacinamide 50 mg, pantothenate 40 mg, inositol 75 mg, folic acid 2.5 mg, biotin 0.08 mg
bMineral premix (mg·kg-1): NaCl 1.0, MgSO4 15.0, NaH2PO4·2H2O 25.0, AlCl3·6H2O 0.06, KH2PO4 32.0, Ca(H2PO4)2·H2O 20.0, C6H5FeO7·10H2O 2.5, CaC6H10CaO6·5H2O 3.5, ZnSO4·7H2O 0.353, MnSO4·4H2O 0.162, CuSO4·5H2O 0.031, CoCl2·6H2O 0.001, KIO3·6H2O, 0.003, cellulose, 0.39
Table 3. Compositions of diets
Grass carp were obtained from a commercial freshwater fsh farm (Bai Rong Aquaculture Co., Ltd., Hubei Province) and reared in circular polyester tanks. After 2 weeks, 135 healthy grass carp (50.24 ± 1.38) g were randomly divided into one of the three groups: control (fed a basal diet), HF (fed a high-fat diet), and HF + B. subtilis (fed a high-fat diet supplemented with B. subtilis), with each group in three replicate tanks with a capacity of 300 L (15 fsh per tank). All grass carp were fed to apparent satiation twice daily (08:30 and 16:30) for 8 weeks. During the entire experimental period, one-third of the water was replaced daily. The recirculating water temperature was maintained at (25 ± 1) ℃, pH at 7.5 ± 0.3, dissolved oxygen at (7 ± 0.45) mg·L-1, ammonia at (0.015 ± 0.002) mg·L-1 and nitrate at (0.05 ± 0.008) mg·L-1.
At the end of the 8-week feeding trial, approximately 24 h after the last feeding, the survival rate and body weight gain of fsh in every tank were calculated. A total of 24 fsh per group were then randomly selected and euthanized (MS-222, 10 mg·L-1). Six fsh were measured for their individual body length and body weight to calculate the condition factor (CF). Blood samples were collected from the caudal vein before dissection to measure serum biochemical indicators. The fsh were then dissected on ice. The liver weight and visceral mass weight were measured to calculate the hepatosomatic index (HSI) and viscera index (VSI). Six fsh were immediately removed using sterile forceps, frozen in liquid nitrogen and stored at - 80 ℃ (not longer than 2 weeks) for total RNA extraction and antioxidant activities, six fsh for determination of liver lipid content, and another six fsh for histochemical observation.
Live samples were freeze dried for 24 h at - 50 ℃ to obtain the dried liver sample. The crude lipid content of individual samples was determined using Soxhlet extraction.
Histochemical analysis was performed as described by Song et al. (2016) with slight modifcations. Briefy, histochemical observations were carried out by oil-red O staining. First, the liver samples were fxed with paraformaldehyde, dehydrated with sucrose, embedded in optimal cutting temperature compound (OCT), sliced (thickness 8 μm) using a cryostat microtome, and stained with oil-red O. Photos were then taken under a microscope (400 ×). Ten felds of view were randomly selected from each sample to calculate the relative area of lipid droplets in oil-red-O stained liver tissue.
The CHO, TG, HDL-C, LDL-C, Alb, TP, AST, and ALT contents were determined using an automatic biochemical analyser (Selectra-xl, the Netherlands) at the Fisheries College of Huazhong Agricultural University (Wuhan, Hubei, China). All kits were purchased from Biosino Bio-Technology and Science, Inc.
TriPure Reagent Kit (Aidlab, RN0102) was used according to its instructions to extract RNA from the liver tissue of grass carp. The quality and quantity of RNA were then assessed via agarose gel (1%) electrophoresis and spectrophotometric (A260:280 nm ratio) analysis, respectively. RNA was reverse transcribed into cDNA using a TRUE script 1st Strand cDNA Synthesis Kit with gDNA Eraser (Aidlab, PC5402). β-Actin is a commonly used reference gene to normalize cDNA loading in grass carp in our laboratory (Kong et al. 2017a, b; Tang et al. 2018) and was used in this study. Specifc primers were designed according to the published sequences of grass carp and are presented in Table 4. Real-time quantitative PCR was performed on a Roche LightCycler 480 real-time PCR instrument. The total volume of the reaction was 20 μL and included 2 μL cDNA, 0.5 μL each primer, 7 μL aqueous diethylpyrocarbonate (DEPC), and 10 μL 2 × SYBR Green qPCR Mix (Aidlab, PC5902). The relative mRNA expression levels were calculated using the 2-ΔΔCt method.
Gene GenBank accession no. Primers Sequence (5' to 3') Annealing tempera- ture L-FABP EU220990.1 L-FABP-F GGGAAAACCATCACTAACTC 58 ℃ L-FABP-R TCAGGGTCTCAACCATCTC FAS HM802556.1 FAS -F GTCCACAGGGTGTCGTTC 58 ℃ FAS -R GAGGTCTTGGGCTCTTTATT ACCα GU908475 ACCα-F AGTATCGCAGTGGCATCA 58 ℃ ACCα-R TGTCCCCTTTGTTTTCCT SCD AJ243835 SCD-F GTTTGTGCCCTGGTTCTT 58 ℃ SCD-R GGGGTTAATGGTGCTGTC LPL FJ436077 LPL-F AGCCCTGTATGAACGAGA 58 ℃ LPL-R CACATCCTTGCCCACTAG CPTIα1a KJ816747 CPTIα1a-F TTTACGACGGACGGTTGC 58 ℃ CPTIα1a-R GCTTGTTCTTCCCACGACT SREBP-1c GU339498 SREBP-1c-F GGATTGAGGTGAGCCGACAT 58 ℃ SREBP-1c-R TGAGGAAAGCCATTGACTACATT PPARγ GQ220296 PPARy-F AATGCACCTTTCGTTATCC 58 ℃ PPARy-R GAGCGTCACTTGGTCGTTC PPARα FJ595500 PPARa-F TGTCAATACTGCCGTTTCC 58 ℃ PPARa-R GACTGGTGCTCCTCTTTCC β-actin M25013 β-actin-F CCTTCTTGGGTATGGAGTCTTG T β-actin-R AGAGTATTTACGCTCAGGTGGG L-FABP liver-type fatty acid-binding protein, FAS fatty acid synthase, ACCα acetyl-CoA carboxylase alpha, SCD stearoyl-CoA desaturase, LPL lipoprotein lipase, CPTIα1a carnitine palmitoyl transferases, SREBP-1c sterol-regulatory element binding proteins, PPARγ peroxisome proliferator activated receptor gamma, PPARα peroxisome proliferator activated receptor alpha, β-actin beta-actin
Table 4. Primers used for real-time PCR analysis
The total antioxidant capacity (T-AOC), the malondialdehyde (MDA) and hydrogen peroxide (H2O2) concentration, the superoxide dismutase (SOD) and catalase (CAT) activity, and glutathione (GSH) levels were determined using commercial kits provided by Nanjing Jiancheng Bioengineering Institute (China). Kits were used according to manufacturer instructions.
One-way analysis of variance was used to analyze the data. All results are expressed as the mean ± SE (standard error of the mean). Multiple comparisons were performed using the Duncan multiple range test among the groups. Statistical signifcance was P < 0.05. All statistical analyses were performed using SPSS 22.0.