All experimental protocols were approved by the University of Sichuan Agricultural Animal Care Advisory Committee (Ya'an, China). The experimental diets were varied with regard to the level of vitamin E. The basal diet (Table I) was formulated according to the Nutrient Requirements of Fish from the NRC National Research Council (24) and was supplemented with 0, 50, 100 or 1,000 mg/kg Dl-all-rac-α-tocopherol acetate (Sanyou Institute of Special Additives, Chengdu, China). A study utilizing the dose of 1,000 mg/kg has been described previously (25). For diet preparation, the ingredients were ground into fine powder and screened through a mesh (0.38 mm). All ingredients were then thoroughly mixed and made into a 4 mm-diameter soft pellet diet using a pellet feed mill. After natural air drying in the dark, the pellets were preserved in a freezer at −20°C. In addition, to assess the proximate composition of the diets, crude protein, crude fat and energy were determined using a Kjeldahl apparatus (nitrogen ×6.25; Xi'an HEB Biotechnology Co., Ltd., Xi'an, China), extraction with petroleum ether by a Soxhlet apparatus (Xi'an HEB Biotechnology Co., Ltd.) followed by ash incineration at 600°C for 3 h and bomb calorimetry, respectively. The composition of the catfish's diet is shown in Table I.

A total of 900 healthy channel catfish (Ictalurus punctatus, Rafinesque 1818; weight, 5.20±0.15 g; age, 2 years old) were purchased from a fish farm (Sichuan Fisheries Science and Technology Development Co., Ltd., Xinjin, China), where they were originally hatched and reared. They were randomly divided into four groups (225 fish per group) according to the vitamin E content in the experimental diet. For each group, fish were equally divided among three circular tanks with a volume of 20 l (total number of tanks, 12), leading to a density of 75 fish per tank.

Prior to starting the feeding trial, fish were acclimatized for two weeks on a basal diet that was not supplemented with vitamin E and then fed one of the four experimental diets by hand three times daily (07:30, 13:30 and 19:30 h) to approximate satiation for 15 weeks based on percentage body weight. Feed consumption was monitored at each feeding. The feeding rate was adjusted once or twice a week based on the feeding activity of fish. The tanks were configured as a flow-through culture system and the incoming water was filtered with aeration. The average water temperature was 21±2°C with a pH of 6.8–7.5, and the amount of dissolved oxygen was 8–10 mg/l. Fish were reared in a 12 h light/dark cycle.

All of the fish in each tank were counted and weighed to monitor growth rates at the end of the feeding trial. The parameters indicating the growth performance of catfish in each group, including average weight gain, average daily growth, feed coefficient and protein efficiency ratio, were calculated following standard protocols. The formulas were as follows: i) Average weight gain (g/fish) = average final weight (g) - average initial weight (g); ii) Average daily growth (g/fish/day) = average weight gain (g/fish)/duration of experiment (days); iii) Protein efficiency ratio = wet weight gain (g)/protein intake (g); iv) Feed coefficient (%) = dry feed fed (g)/wet weight gain (g); and v) Survival (%) = (final number of fish/initial number of fish) × 100%.

After recording of body measurements, 45 fish were collected from each group. The hepatopancreas and gut of fish were removed, frozen in liquid nitrogen and then preserved at −80°C for further analysis. For digestive enzyme analyses, tissue samples were homogenized in ice-cold physiological saline and the supernatant was retained after centrifugation at 3,200 × g for 20 min. The activities of lipase were evaluated using phenolphthalein indicator and the activities of protease were determined by the Folin-phenol reagent method. In addition, the alkaline phosphatase (AKP) assay kit (Nanjing Jiancheng, Nanjing, Jiangsu, China) were used to evaluate the AKP activities, and Na⁺/K⁺-ATPase activities were assessed using a Na⁺/K⁺-ATPase assay kit (Nanjing Jiancheng, Nanjing, China).

Another 45 fish collected from each group at the end of the feeding trial were dissected and the guts were removed. The gut weight and length of the fish from each group was recorded. Moreover, the gut weight index and gut length index were calculated as follows: i) Gut weight index (%) = [average gut weight (g)/average final body weight (g)] × 100%; and ii) Gut length index (%) = [average gut length (cm)/average final body length (cm)] × 100%.

After recording of body measurements, another 20 fish were collected from each group and 0.5 cm of each segment of intestine was sampled on a chilled cutting board according to a previous description (26). Samples were fixed in Bouin solution for 12 h, embedded in paraffin (Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) and cut into 5-µm sections using a Leica RM2128 microtome (Leica Microsystems, Wetzlar, Germany). A proportion of the sections from each segment of the intestine was stained with hematoxylin and eosin (H&E) and observed under an Olympus BH-2 microscope (Olympus, Tokyo, Japan). The height of intestinal folds and the thickness of the mucous membrane were evaluated by analyzing more than ten different views of each sample of intestine segment using Spot software (Version 4.0.6; Diagnostic Instruments, Sterling Heights, MI, USA).

The other part of the tissue sections was used for immunohistochemical analysis for detecting the distribution and expression of somatostatin. After washing with PBS three times, sections were rehydrated and incubated in 3% H₂O₂ for 10 min. Subsequently, sections were incubated with rabbit anti-somatostatin primary antibody (BA0124; Wuhan Boster Biological Technology, Ltd, Wuhan, Hubei, China) at a dilution of 1:100 at 4°C for 17 h, followed by incubation with the biotin-conjugated goat anti-rabbit secondary antibody (BA1003; Wuhan Boster Biological Technology, Ltd, Wuhan, Hubei, China). Finally, sections were stained using a Streptavidin-Biotin Peroxidase Complex kit (Wuhan Boster Biological Technology, Ltd, Wuhan, Hubei, China) and visualized using diaminobenzidine as a chromogen. Sections incubated with PBS instead of primary antibody were considered as a negative control. The slices were observed and images were captured using a digital camera (Moticam 2500, Motic, Hong Kong, China) attached to a Nikon microscope (Nikon Eclipse 50i, Nikon, Tokyo, Japan). The expression of somatostatin among groups was evaluated by determining the number of somatostatin-positive cells. Ten different views per section of each segment of intestine (three sections for each segment intestine from each group) were obtained to analyze the expression and distribution of somatostatin.

Values are expressed as the mean ± standard deviation. Significant differences among groups were determined by one-way analysis of variance followed by Duncan's method. P<0.05 was considered to indicate a statistically significant difference between groups. All statistical analyses were performed with SPSS 19.0 software (International Business Machines, Inc., Armonk, NY, USA).

Average weight gain, protein efficiency ratio, feed coefficient as well as survival were calculated and presented in Table II. Dietary supplementation with vitamin E affected catfish survival and fish with vitamin E supplementation at different concentrations showed a higher survival rate than those without. In addition, the fish receiving diets supplemented with vitamin E had significantly higher growth parameters, including average weight gain, average daily growth as well as protein efficiency ratio than those fed with the basal diet (P<0.05) and the highest values occurred in fish receiving a diet supplemented with vitamin E at 100 mg/kg. The feed coefficient is used to evaluate the feed quality with a lower the feed coefficient indicating a higher feed quality. As shown in Table II, dietary supplementation with 50 and 100 mg/kg vitamin E had an obviously lower feed coefficient than that with 1,000 mg/kg vitamin E or without vitamin E (P<0.05).

Digestive enzyme activity in the hepatopancreas and each gut segment of fish supplemented with different concentrations of vitamin E was shown in Fig. 1. In the hepatopancreas and all segments of the fish gut for 50 and 100 mg/kg vitamin E groups, protease activity was significantly increased compared to that without vitamin E supplementation (P<0.05) and that in 1,000 mg/kg group (P<0.05). However, there was no significant difference in the 1,000 mg/kg group compared with that of the 0 mg/kg group. Similarly, in the hepatopancreas and all segments of the fish gut for 50 and 100 mg/kg groups, lipase activity was significantly increased compared to that without vitamin E supplementation (P<0.05) and that in 1,000 mg/kg group (P<0.05). Additionally, there was no significant difference in the hepatopancreas, foregut and hindgut of fish in 1,000 mg/kg group compared with the 0 mg/kg group, except in the mid-gut (P<0.05). AKP activity was significantly increased in the hepatopancreas, foregut and mid-gut of fish in 50, 100 and 1,000 mg/kg groups compared to that without vitamin E supplementation (P<0.05). Na⁺/K⁺-ATPase activity was significantly increased in the hepatopancreas, foregut and hindgut of fish in 50, 100 and 1,000 mg/kg groups compared to that without vitamin E supplementation (P<0.05). However, there was no significant difference in the mid-gut of fish in 50, 100 and 1,000 mg/kg groups compared with the 0 mg/kg group (P>0.05). In addition, the concentration of supplemented vitamin E influenced the digestive enzymes activity, since fish supplemented with vitamin E at 50 and 100 mg/kg showed obviously higher activities of protease, lipase and AKP than that without vitamin E (P<0.05). Although the fish in 1,000 mg/kg group also exhibited increased activities, the values were lower than those in the 50 and 100 mg/kg groups.

The results of gut weight, gut weight index, gut length as well as gut length index of fish supplemented with different concentrations of vitamin E are shown in Table III. Fish fed diets supplemented with vitamin E had a greater gut length and gut weight compared to those without (P<0.05). The gut weight index and gut length index of fish with vitamin E supplementation were also markedly higher than those of fish without vitamin E supplementation (P<0.05).

The intestinal fold height and of mucous membrane thickness were observed under a microscope after H&E staining and the statistical results are shown in Figs. 2 and 3. The intestinal folds in fish fed a diet without vitamin E showed an irregular morphology and they were small, thin and loosely arranged throughout, while they were much larger and more tightly arranged in fish fed a diet supplemented with vitamin E. Statistical analysis indicated that fished fed diets containing vitamin E had significantly larger intestinal fold height as well as a thicker mucous membrane (P<0.05).

Somatostatin-positive cells were observed in the foregut, mid-gut as well as hindgut. The distribution of somatostatin-positive cells in the foregut of fish fed a diet supplemented with different amounts of vitamin E is shown in Fig. 4A-D. Somatostatin-positive cells (brown-yellow granules) with various cellular morphologies were mainly located at the bottom of the intestinal epithelial cell layer and some of them were found between cells of the mucous epithelium. The distribution of somatostatin-positive cells among the groups in the mid- and hindgut was similar to that in the foregut (data not shown). In addition, the expression of somatostatin was evaluated by quantifying the number of somatostatin-positive cells in ten different views per section and the results for each segment of the intestine are shown in Fig. 4E. The number of somatostatin-positive cells was significantly higher in the foregut of fish with vitamin E supplementation at different concentrations compared with that in the foregut of fish without supplementation (P<0.05). Moreover, dietary supplementation of vitamin E at 100 and 50 mg/kg resulted in an obviously larger number of somatostatin-positive cells in the foregut and mid-gut of fish than that at 1,000 mg/kg (P<0.05).