Experimental diets, fish trial and sampling

Purified glycinin was kindly provided by China Agricultural University (patent no. 200410029589·4, China). The ingredients and nutrient content of the experimental diets are shown in Table 1. Fish meal, gelatin and casein were used as the dietary protein sources. An 80 g glycinin/kg diet was used in this study to investigate the toxic effects and the corresponding mechanisms according to the following concerns. It is well-known that glycinin accounts for approximately 40 % of the total soyabean proteins⁽ Reference Utsumi, Matsumura and Mori ^{25 )}. Our previous study observed that high levels (347, 517·8 and 685·8 g/kg diet) of de-hulled SBM (protein content=47 %) significantly impaired growth and seriously disrupted the intestinal integrity in juvenile Jian carp⁽ Reference Zhang, Zhou and Liu ^{26 )}. In addition, a dosage of 12·0 g Gln/kg used to block the negative effects of glycinin was proven to be optimal for the intestinal health of Jian carp in our previous study⁽ Reference Lin and Zhou ^{23 )}. All diets were made isonitrogenous. Briefly, glycinin was made isonitrogenous, and amino acids were balanced with the addition of reduced amounts of casein and modulated amino acids and compensated with appropriate amounts of maize starch according to a previous study⁽ Reference Sun, Li and Li ^{27 )}, whereas Gln was made isonitrogenous with the addition of reduced amounts of glycine and compensated with appropriate amounts of maize starch according to our previous study⁽ Reference Lin and Zhou ^{23 )}. In the diets, lysine, methionine, threonine, riboflavin, pantothenic acid, thiamine, pyridoxine, inositol, Fe and Zn were properly added to meet the nutrient requirements of juvenile Jian carp according to our previous laboratory studies⁽ Reference Jiang, Feng and Liu ^{28 ,} Reference Jiang, Liu and Hu ^{29 )}. The levels of other nutrients were designed to meet the requirements of common carp (Cyprinus carpio L.) according to the NRC (2011). The diets were stored at −20°C until feeding, as in our previous study⁽ Reference Zhang, Guo and Feng ^{24 )}.

Table 1 Ingredients and nutrient contents of the diets

* Glycinin was made isonitrogenous, and amino acids were balanced with the addition of reduced amounts of casein and modulated amino acids and compensated with appropriate amounts of maize starch according to a previous study⁽ Reference Sun, Li and Li ^{27 )}.

† Glutamine was made isonitrogenous with the addition of reduced amounts of glycine and compensated with appropriate amounts of maize starch according to our previous study⁽ Reference Lin and Zhou ^{23 )}.

‡ Per kg of mineral premix (g/kg): FeSO₄·7H₂O (19·7 % Fe), 69·695 g; CuSO₄·5H₂O (25·0 % Cu), 1·201 g; ZnSO₄·7H₂O (22·5 % Zn), 21·640 g; MnSO₄·H₂O (31·8 % Mn), 4·089 g; KI (3·8 % I), 2·895 g; NaSeO₃ (1·0 % Se), 2·500 g. All ingredients were diluted with CaCO₃ to 1 kg.

§ Per kg of vitamin premix (g/kg): retinyl acetate (172 g/kg vitamin A), 0·800 g; cholecalciferol (12·5 g/kg vitamin D), 0·480 g; d, l-α-tocopherol acetate (50 %), 20·000 g; menadione (50 %), 0·200 g; cyanocobalamin (10 %), 0·010 g; d-biotin (20 %), 0·500 g; folic acid (96 %), 0·521 g; thiamin nitrate (98 %), 0·104 g; ascorhyl acetate (92 %), 7·247 g; niacin (98 %), 2·857 g; meso-inositol (98 %), 52·857 g; calcium-d-pantothenate (98 %) 2·511 g; riboflavine (80 %), 0·625 g; pyridoxine hydrochloride (98 %), 0·755 g. All ingredients were diluted with maize starch to 1 kg.

|| Crude protein and crude lipid contents were determined according to the method of AOAC (1998). Lysine, Met+Cys, available P, n-3 and n-6 fatty acids contents were calculated according to NRC (1993).

The juvenile Jian carps (Cyprinus carpio var. Jian) used in this study were purchased from a local hatchery. Before the trial, the fish were acclimatised to the experimental environment (the water temperature and pH were 23 (sd 1) and 7·0 (sd 0·3)°C, respectively; dissolved oxygen was higher than 5 mg/l; the aquaria were supplied with flow-through water at a rate of 1·2 l/min; and the water was drained through biofilters to remove solid substances and reduce ammonia concentration) for 4 weeks. After adapting, 450 fish with a mean initial weight of 5·37 (sd 0·02) g were randomly assigned to each of nine experimental aquaria (90 length×30 width×40 cm height). In the feeding trial, each diet was fed to a triplicate of fish six times per d for the first 4 weeks and four times per d for the 5th and 6th week, a feeding rhythm that was established in our laboratory previously⁽ Reference Xiao, Feng and Liu ^{30 )}. The fish were fed to apparent satiation. After 30 min of feeding, uneaten feed was removed by siphoning, and it was dried and weighed later to calculate feed intake.

At the beginning and end of the trial, fish in each aquarium were counted and weighed. After that, fish were anaesthetised in a benzocaine bath (50 mg/l) 12 h after the last feeding according to the method of Bohne and our previous study⁽ Reference Zhang, Guo and Feng ^{24 ,} Reference Bohne, Hamre and Arukwe ^{31 )}. After killing the fish, their intestines were quickly removed on ice, rinsed in cold physiological saline (0·9 % NaCl solution, pH=7·2), weighed, measured and frozen in liquid nitrogen and stored at −80°C until they were analysed according to the methods described in previous studies⁽ Reference Zhang, Guo and Feng ^{24 ,} Reference Qu, Wang and Feng ^{32 )}. The intestines of another six fish from each aquarium were sampled and fixed with formalin (10 %), sectioned and stained with haematoxylin–eosin stain for analysis of height of intestinal folds according to the method in our previous study⁽ Reference Zhang, Guo and Feng ^{24 )}.

Biochemical analysis

According to the method in a previous study⁽ Reference Qu, Wang and Feng ^{32 )}, with slight modifications, the intestine samples were homogenised on ice in ten volumes (w/v) of ice-cold physiological saline and centrifuged at 6000 g for 20 min at 4°C, and then the supernatant was used for biochemical analysis.

Intestinal function indexes, such as AKP, Na⁺/K⁺-ATPase, γ-glutamyl transpeptidase (γ-GT) and creatine kinase (CK), were determined according to the procedure described by Bessey et al. ⁽ Reference Bessey, Lowry and Brook ^{33 )}, Weng et al. ⁽ Reference Weng, Chiang and Gong ^{34 )}, Bauermeister et al. ⁽ Reference Bauermeister, Lewendon and Ramage ^{35 )} and Tanzer & Gilvarg⁽ Reference Tanzer and Gilvarg ^{36 )}, respectively. In the assay enzymes, 1 U of activity was considered as the amount of enzyme required to release 1 μmol of product/h, except γ-GT, which calculated as per min.

ROS were measured according to the method described by Gornicka et al. ⁽ Reference Gornicka, Fettig and Eguchi ^{37 )} with a slight modification – intestinal tissue homogenates were used in this study instead of adipose tissue homogenates in the referenced study. The values of ROS are expressed as the multiple of the levels of the control group according to the method described by a previous study⁽ Reference Xu, Zhou and Zhang ^{38 )}. Lipid peroxidation was analysed in terms of malondialdehyde (MDA) equivalents using the thiobarbituric acid (TBA) reaction as a previous study described⁽ Reference Qu, Wang and Feng ^{32 )}. In brief, samples were mixed with TCA and centrifuged. Then, TBA was added to the supernatant. The mixture was heated in water at 95°C for 40 min. MDA forms a red adduct with TBA, which has an absorbance of 532 nm. The protein carbonyl (PC) residue content was determined as previously described⁽ Reference Baltacıoğlu, Akalın and Alver ^{39 )}, using the 2,4-dinitrophenylhydrazine reagent. The carbonyl content was calculated from the peak absorbance at 340 nm, using an absorption coefficient of 22 000/M/cm. The protein contents of the intestines were determined according to a previous study⁽ Reference Bradford ^{40 )}, which were used to calculate the following parameters involved. MDA and PC are all denoted by nmol/mg protein.

The total superoxide dismutase (SOD) and glutathione peroxidase (GPx) activities were assayed as a previous study described⁽ Reference Zhang, Zhu and Cai ^{41 )}. Briefly, for SOD, the reaction mixture contained 50 mm-phosphate buffer (pH=7·8), 1·08 mm-diethylenetriaminepentaacetic acid (DETAPAC), 0·06 mm-nitro blue tetrazolium, 0·16 mm-xanthine solution and 30 μl of samples. After the addition of 0·19 U/ml of xanthine oxidase, the absorbance change at 550 nm was monitored. For GPx, the reaction mixture consisted of tissue homogenates, 40 μl of 0·25 mm-hydrogen peroxide (H₂O₂), 10 mm-sodium phosphate buffer (pH=7·0), 0·5 mm-glutathione (GSH) and 1·25 mm-NaN₃ in a total volume of 1 ml. After 3 min intervals, 0·5 ml of dithiobisnitrobenzoic acid was added. A yellow product formed as GSH reacts with dithiobisnitrobenzoic acid, which was monitored at 412 nm. The catalase (CAT) activities were measured as in our previous study⁽ Reference Jiang, Feng and Liu ^{42 )}. The assay mixture consisted of 100 mm-KPO₄ buffer (pH=7·0), 10 mm-H₂O₂ and 50 μl of intestinal homogenates in a total volume of 1 ml. The decrease of H₂O₂ was monitored by measuring the absorbance at 240 nm. Glutathione-S-transferase (GST) activity was measured by monitoring the formation of the adduct of GSH and 1-chloro–2,4-dinitrobenzene (CDNB)⁽ Reference Lushchak, Lushchak and Mota ^{43 )}. The reduced GSH was determined using the method described in a previously described study⁽ Reference Vardi, Parlakpinar and Ozturk ^{44 )}. An adduct between GSH and CDNB was monitored at 340 nm. Together, with the exception of SOD, in which 1 U of activity was considered a half of inhibition of absorbance in comparison with tube lacking enzyme according to the method described by Datkhile et al. ⁽ Reference Datkhile, Mukhopadhyaya and Dongre ^{45 )}, in the assay enzymes 1 U of activity was considered as the amount of enzyme required to transform 1 μmol of substrate/min.

RNA extraction and quantitative real-time PCR analysis

In this study, total RNA was extracted according the procedure of our previous study⁽ Reference Jiang, Liu and Hu ^{29 )}. Briefly, total RNA of the intestines was extracted using RNAiso Plus (D9108B; Takara Biotechnology, Dalian Co. Ltd), according to the manufacturer’s instructions. RNA quantity and quality were assessed by electrophoresis on 1 % agarose gels and spectrophotometric analysis (A260:280 nm ratio). Subsequently, cDNA was synthesised using a PrimeScript^TM RT reagent Kit, according to the manufacturer’s instructions. Briefly, oligo dT primers were used to reverse transcribe respective RNAs in the presence of PrimeScript^TM RT enzyme Mix I, 5×PrimeScript^TM buffer, Random 6 mers and RNase-free dH₂O at 37°C for 15 min, following inactivation at 85°C for 5 s.

The specific primers for the genes were designed with Primer Premier Software (Premier Biosoft International) based on the carp sequences (Table 2). According to the results of our preliminary experiment concerning the evaluation of internal control genes, β-actin was used as a reference gene to normalise cDNA loading. The quantitative RT-PCR assays were conducted using the relative standard curve method described by Wang et al. ⁽ Reference Wang and Gallagher ^{46 )}. Standard curves were generated for the target genes and the endogenous control gene β-actin (Table 2) based on 10-fold serial dilutions. Gene expression quantities were normalised against β-actin, and ratios for the treated samples were calculated via comparisons with the expressions in control animals.

Table 2 Real-time PCR primer sequences

CAT, catalase; GPx, glutathione peroxidase; GR, glutathione reducase; GSH, glutathione; GST, glutathione-S-transferase; PKC, protein kinase C; SOD, superoxide dismutase; TGF-β2, transformed growth factor-β2; TOR, target of rapamycin.

Statistical analysis

All results were expressed as the means and standard deviations. Data were subjected to one-way ANOVA. When a significant difference was observed (P<0·05), Tukey’s test of significance was used to resolve the difference. All statistical analyses were performed using the SPSS 13.0 for Windows (SPSS Inc.).

Effects of glycinin on the growth performances and intestinal growth and function of fish

The fish growth performance and intestinal growth and function results are presented in Table 3. The results indicated that compared with the control group treatment with glycinin alone significantly decreased the final body weight (FBW), specific growth ratio (SGR), feed intake (FI), intestinal weight (IW), intestinal protein content (IPC), intestinal length (IL) and fold heights in the PI and MI of fish (P<0·05). Upon co-treatment with Gln, the FBW, SGR, FI, IW, IPC, IL and fold heights in the PI and DI significantly increased compared with those of fish exposed to glycinin alone (P<0·05), and the FBW, SGR and IPC values and fold heights in the PI and DI completely recovered to become equivalent to those of the control group. However, no significant differences were found in the feed efficiency, survival rate and intestinal length index and intestinal somatic index among the treatments (P>0·05).

Table 3 Growth performance and intestinal growth of juvenile Jian carp (Cyprinus crpio var. Jian) exposed to dietary glycinin for 42 d (Mean values and standard deviations)

IBW, initial body weight; FBW, final body weight; SGR, specific growth ratio; FI, feed intake; FE, feed efficiency; IW, intestinal weight; IPC, intestinal protein content; IL, intestinal length; ILI, intestinal length index; ISI, intestinal somatic index; PI, proximal intestine; MI, mid intestine; DI, distal intestine.

SGR=100×(ln final weight−ln initial weight)/number of days.

FE=100×(weight gain (g)/feed intake (g)).

ILI=100×(intestine length (cm)/total body length (cm)).

ISI=100×(wet intestine weight (g)/wet body weight (g)).

IPC=100×(intestine protein (g)/wet intesitine weight (g)).

^a,b,c Mean values within a row with unlike superscript letters are significantly different (P<0·05).

* Fifty fish in each group.

Intestinal function indexes demonstrated that dietary glycinin exposure alone depressed the activities of creatine kinase and AKP in the PI, MI and DI; the activity of Na⁺ and K⁺-ATPase in the MI; and the activity of γ-GT in the PI of fish (P<0·05). When co-treated with Gln, the activity of AKP in the PI, MI and DI and the γ-GT activity in the PI were significantly increased when compared with those of fish exposed to glycinin alone (P<0·05), and even those values completely recovered to be equivalent to the control. However, glycinin alone caused increases in the γ-GT activity in the DI of fish, whereas that co-treated with Gln had no significant effects on this index. No significant differences were found in the Na⁺, K⁺-ATPase activity in the PI and DI, as well as γ-GT activities in the MI among the treatments (P>0·05).

Effects of glycinin on the oxidative status and antioxidant-related parameters in the intestines of fish

The oxidative status and antioxidative enzyme activities in the intestines of fish are presented in Fig. 1 and Table 4. With the administration of glycinin alone, ROS, MDA and PC contents in the intestines of fish increased significantly (P<0·05), and co-administration of glycinin and Gln caused the ROS and MDA content to recover to the levels similar to the control levels. However, there was no significant change in the intestinal PC content in fish between the glycinin alone group and the group with glycinin plus Gln (P>0·05). In addition, the activities of SOD, CAT, GST and GR in the intestines of fish were decreased in fish treated with glycinin compared with the unexposed controls (P<0·05). Except for intestinal GST, those activities reduced by glycinin exposure recovered with the co-treatment of Gln to be equivalent to the control values (P>0·05). GPx activity in the co-treatment with glycinin and Gln group is higher than that in the glycinin alone group. GSH was not significantly different between the control group and the glycinin alone group, whereas it significantly increased with the co-treatment of glycinin and Gln in the intestines of fish (P<0·05).

Fig. 1 Effects of different treatments on the reactive oxygen species (ROS) production in the intestine of Jian carp. Values are means of six replicates, with standard deviations represented by vertical bars. ^a,b Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Table 4 Oxidative status and antioxidant abilities in the intestine of juvenile Jian carp exposed to dietary glycinin for 42 d (Mean values and standard deviations of six replicates)

MDA, malondialdehyde (nmol/mg protein); PC, protein carbonyl content (nmol/mg protein); GSH, glutathione content (mg/g protein); SOD, superoxide dismutase (U/mg protein); CAT, catalase (U/mg protein); GST, glutathione-S-transferase (U/mg protein); GPx, glutathione peroxidase (U/mg protein); GR, glutathione reductase (U/g protein).

^a,b Mean values within a row with unlike superscript letters are significantly different (P<0·05).

Effects of glycinin on the mRNA levels of MnSOD, CuZnSOD, catalase, four isoforms of glutathione peroxidases, nine isoforms of glutathione-S-transferases and glutathione reducase genes in the intestines of fish

To investigate the effects of glycinin on the transcription of antioxidant enzymes in the intestines of fish, the mRNA levels of CuZnSOD, MnSOD, CAT, four isoforms of GPx, nine isoforms of GST and GR in the intestine of fish were determined (Fig. 2). The results indicated that dietary glycinin exposure significantly increased the MnSOD, CuZnSOD, GPx1b and GPx4a mRNA levels, but decreased the mRNA levels of GST α, GST μ, GST π, GST ρ, GST θ, GST κ, mGST1, mGST2 and mGST3. The GPx1a, GPx4b, CAT and GR mRNA levels were not changed by dietary glycinin exposure in the intestines of fish. The highest mRNA levels of MnSOD, CuZnSOD, CAT, GPx1a, GPx1b and GR, and the lowest mRNA levels of GST θ and mGST1 were observed in the intestines of fish fed the glycinin plus Gln diet (P<0·05). However, no significant difference in the mRNA expression of GPx4a, GPx4b, GST α, GST μ, GST π, GST ρ and GST κ were observed between the glycinin alone group and the glycinin plus Gln group.

Fig. 2 Effects of different treatments on the mRNA levels of CuZnSOD, MnSOD (A), catalase (CAT) (B), four isoforms of glutathione peroxidase (GPx) (C), nine isoforms of glutathione-S-transferase (GST) (D, E) and glutathione reducase (GR) (F) in the intestine of Jian carp. Values are means of six replicates, with standard deviations represented by vertical bars. ^a,b,c Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Effects of glycinin on the antioxidant-relative signalling factors in the intestines of fish

The effects of glycinin on the mRNA levels of Nrf2 signalling-related factors in the intestines of fish are presented in Fig. 3. The results indicated that glycinin exposure alone significantly decreased mRNA levels of Nrf2, Kelch-like ECH-associated protein-1a (Keap1a) and Keap1b in the intestines of fish compared with the unexposed control group (P<0·05). Diets supplemented with glycinin and Gln did not change the Nrf2 and Keap1b mRNA levels (P>0·05), but they significantly decreased the Keap1a mRNA levels (P<0·05).

Fig. 3 Effects of different treatments on the mRNA levels of Nrf2 and Keap1 in the intestine of Jian carp. Values are means of six replicates, with standard deviations represented by vertical bars. ^a,b,c Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

The effects of glycinin on the mRNA levels of eleven isoforms of protein kinase C (PKC) are presented in Fig. 4. The results indicated that glycinin significantly decreased the mRNA levels of conventional PKC (α, β1, β2 and γ). However, glycinin has different effects on the novel PKC and atypical PKC. It decreased the mRNA levels of PKC δ, η1, θ and ι but did not have significant effects on the PKC η2, ε and ζ mRNA levels. In addition, interestingly, except for PKC β2, the decreased mRNA levels of conventional PKC (including PKC α, β1 and γ) and atypical PKC (including ζ and ι) in fish fed the glycinin alone diet were partially or completely reversed, whereas the novel PKC (including PKC δ, ε, η1 and θ) except PKC η2 in the intestines of fish were not changed in fish with glycinin plus Gln.

Fig. 4 Effects of different treatments on the mRNA levels of eleven isoforms of protein kinase C (PKC) in the intestine of Jian carp. Values are means of six replicates, with standard deviations represented by vertical bars. ^a,b,c Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Effects of glycinin on the tight junction protein transcript abundance in the intestines of fish

Fig. 5 presents the relative levels of occludin, claudin-3c, claudin-11 and claudin-7 mRNA in the PI, MI and DI of fish exposed to dietary glycinin. Compared with the control fish, glycinin significantly down-regulated the expression of occludin in all intestinal segments and claudin-3c in the MI and DI (P<0·05) and up-regulated the expression of claudin-3c in the PI, claudin-11 in all intestinal segments and claudin-7 in the DI (P<0·05), whereas there was no significant change in the expression of claudin-7 in the PI and MI (P>0·05). Co-administration of glycinin and Gln demonstrated significant increases in the mRNA levels of occludin in the MI and significant decreases of the mRNA levels of claudin-3c in the PI, claudin-11 in all intestinal segments and claudin-7 in the DI. However, no significant changes in the mRNA levels of occludin in the PI and DI, claudin-3c in the MI and DI and claudin-7 in the PI and MI were observed when comparing the glycinin plus Gln group with the group of glycinin exposed alone.

Fig. 5 Effects of different treatments on the mRNA levels of occludin, claudin-3c, claudin-11 (A, B, C) and claudin-7 (D) in the proximal intestine (PI), mid intestine (MI) and distal intestine (DI) of Jian carp. Values are means of six replicates, with standard deviations represented by vertical bars. ^a,b,c Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Effects of glycinin on the relative mRNA levels of cytokines in the intestines of fish

As shown in Fig. 6, glycinin significantly down-regulated the mRNA levels of IL-1 β in the DI, TNF- α in the PI and MI and IL-10 in the PI and DI, and transformed growth factor-β2 (TGF-β2) in the MI (P<0·05) and up-regulated the mRNA levels of IL-1 β in the PI and MI, and TNF- α in the DI (P<0·05), whereas it caused no significant change in the mRNA levels of IL-10 in the MI and TGF- β2 in the PI and DI (P>0·05). Co-administration of glycinin and Gln resulted in significant reductions in the mRNA levels of IL-1 β in the PI, IL-10 in the MI and TGF-β2 in the PI (P<0·05) and significant increases in the mRNA levels of IL-1 β in the MI and DI, TNF- α in the PI and DI and TGF- β2 in the DI (P<0·05), whereas it resulted in no significant change in mRNA levels of TNF- α and TGF- β2 in the MI and IL-10 in the PI and DI as compared with the group of glycinin (P>0·05).

Fig. 6 Effects of different treatments on the mRNA levels of IL-1 β, TNF- α, IL-10 and transformed growth factor-β2 (TGF- β2) in (A) the proximal intestine (PI), (B) mid intestine (MI) and (C) distal intestine (DI) of Jian carp. Values are means of six replicates with standard deviations represented by vertical bars. ^a,b,c Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Effects of glycinin on the mRNA levels of the caspase-3, 8 and 9 in the intestines of fish

The mRNA levels of caspase-3, caspase-8 and caspase-9 in the PI, MI and DI are presented in Fig. 7. Glycinin significantly increases the mRNA levels of caspase-3 in the DI and caspase-8 and caspase-9 in the DI (P<0·05) and reduced the mRNA levels of caspase-8 and caspase-9 in the PI and MI of Jian carp (P<0·05). Co-administration of glycinin and Gln resulted in significant decreases in the mRNA levels of caspase-3, caspase-8 and caspase-9 in the DI (P<0·05), and increases in the mRNA levels of caspase-9 in the PI and MI (P<0·05), whereas it resulted in no significant change in the mRNA levels of caspase-3 and caspase-8 in the PI and MI of Jian carp when compared with the glycinin alone group (P>0·05).

Fig. 7 Effects of different treatments on the mRNA levels of caspase-3, caspase-8 and caspase-9 in (A) the proximal intestine (PI), (B) mid intestine (MI) and (C) distal intestine (DI) of Jian carp. Values are means of six replicates with standard deviations represented by vertical bars. ^a,b,c Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Effects of glycinin on the mRNA levels of target of rapamycin and 4E-BP in the intestines of fish

The mRNA levels of TOR and 4E-BP genes in the PI, MI and DI of fish are presented in Fig. 8. Significant reductions in TOR mRNA levels caused by glycinin alone were observed in the MI and DI, and the reduction in the DI of fish was completely blocked by Gln supplementation (P<0·05). However, Gln failed to significantly block the reduction of TOR by glycinin exposure in the MI of fish. Furthermore, there were no changes in the TOR mRNA levels in the PI of fish among the three groups. As for 4E-BP, the results indicated that glycinin increased the mRNA levels in the DI (P<0·05), whereas it caused no significant change in the PI and MI of fish (P>0·05). Upon treatment with Gln, the 4E-BP mRNA levels recovered to the control levels in the DI (P>0·05) and increased to higher than the control levels in the PI of fish (P<0·05). In addition, no significant effect was observed in the MI of fish supplemented with glycinin and Gln (P>0·05).

Fig. 8 Effects of different treatments on the mRNA levels of target of rapamycin (TOR) and 4E-BP in (A) the proximal intestine (PI), (B) mid intestine (MI) and (C) distal intestine (DI) of Jian carp. Values are means of six replicates with standard deviations represented by vertical bars. ^a,b Mean values with unlike letters were significantly different (P<0·05). Gln, glutamine; □, control; ■, glycinin; , glycinin+Gln.

Glycinin-depressed fish growth performance and impaired intestinal growth and function of fish

In terms of growth performance, the present study found that dietary glycinin decreased the fish growth rate and impaired feed intake, which was also observed in piglets⁽ Reference Zhao, Qin and Sun ^{5 )}. However, the survival rate was not influenced by the dietary treatments. Although no reports exist concerning the effects of glycinin on the survival rate in fish, similar findings were observed in our previous study with β-conglycinin, another allergen in soyabean protein⁽ Reference Zhang, Guo and Feng ^{24 )}. In addition, fish growth is strongly related to intestinal growth and function⁽ Reference Wen, Feng and Jiang ^{8 )}. To our knowledge, intestinal brush border enzymes, such as AKP, Na⁺, K⁺-ATPase, γ-GT and CK, are considered to reflect intestinal function in fish⁽ Reference Chen, Feng and Kuang ^{49 )}. The present study demonstrated that dietary glycinin exposure alone decreased the IW, IL and IPC, and reduced the fold heights and creatine kinase, Na⁺, K⁺-ATPase, AKP and γ-GT activities in some intestinal segments of fish, suggesting that glycinin could impair the growth and function of the intestines in fish. Similar results of decreased jejunum villus height caused by dietary glycinin exposure were found in rats⁽ Reference Ma, He and Sun ^{50 )}. However, those results do not agree with a recent study that found that glycinin had no significant effects on the fold heights and AKP activities in the mid gut and hind gut of juvenile turbot Scophthalmus maximus ⁽ Reference Gu, Bai and Xu ^{51 )}. There are at least three reasons for those differences. First, it may be partially because of the different doses. This study administered an 80 g glycinin/kg diet, which was higher than that in the juvenile turbot study (60 g glycinin/kg diet)⁽ Reference Gu, Bai and Xu ^{51 )}. Second, in this study, the fish were exposed to dietary glycinin for 6 weeks, which was longer than the exposure in the juvenile turbot study (4 weeks)⁽ Reference Gu, Bai and Xu ^{51 )}. Third, turbot is a carnivorous fish with a stomach⁽ Reference Gu, Bai and Xu ^{51 )}, whereas this study used carp, which is an omnivorous fish without a stomach. Furthermore, fish intestinal growth and function are associated with the structural integrity⁽ Reference Zhang, Guo and Feng ^{24 )}. To our knowledge, oxidative damage has been linked to the disruption of intestinal barrier integrity, cell injury and dysfunction⁽ Reference Wen, Feng and Jiang ^{8 )}. Thus, we next investigated whether glycinin-depressed intestinal growth and function are related to the oxidative damage and the potential mechanism in the intestines of fish.

Glycinin-induced oxidative damage and its potential mechanism in the intestines of fish

Glycinin impaired intestinal tight junctions and modulated cytokines and apoptosis signalling, which is partially related to changes in target of rapamycin signalling in the intestines of fish

Several lines of evidences support a barrier-forming role for occludin, claudin-3c and claudin-11 in fish⁽ Reference Chasiotis and Kelly ^{70 ,} Reference Chasiotis, Kolosov and Kelly ^{71 )}, whereas claudin-7 appears to possess pore-forming characteristics⁽ Reference Krause, Winkler and Piehl ^{72 )}. In the present study, dietary glycinin exposure significantly decreased the mRNA levels of the barrier-forming TJ, occludin and claudin-3c in the MI and DI, and increased the mRNA levels of the pore-forming TJ, claudin-7 in the DI, whereas it did not change the mRNA levels of claudin-7 in the PI and MI of fish. However, in the PI, although dietary glycinin exposure decreased occludin mRNA levels, it also increased claudin-3c mRNA levels. Meanwhile, we observed that dietary glycinin exposure induced adaptive increases in the mRNA levels of claudin-11 in the PI, MI and DI of fish. These results indicated that glycinin exposure could impair some of the TJ components, the seriousness of which followed the order DI>MI>PI. These results might partially explain a previous report that SBM-induced pathohistological changes primarily occurred in the DI of fish⁽ Reference Krogdahl, Bakke McKellep and Baeverfjord ^{73 )}. The reasons for serious glycinin-induced impairment of the TJ in the DI>MI>PI are largely unknown but might be partially associated with the apoptosis and the inflammatory cytokines. It was reported that the intestinal TJ could be destroyed by apoptosis and pro-inflammatory cytokines⁽ Reference Capaldo and Nusrat ^{18 ,} Reference Bruewer, Luegering and Kucharzik ^{68 )}. Thus, we next investigated the effects of glycinin on the apoptosis signalling and the cytokines in the intestines of fish.

It was reported that pro-inflammatory cytokines disrupt epithelial barrier function by apoptosis-independent mechanisms⁽ Reference Bruewer, Luegering and Kucharzik ^{68 )}. In general, caspase-3 is the major executioner caspase in the caspase-dependent pathway⁽ Reference Gao, Xu and Qiao ^{74 )}. Meanwhile, apoptotic pathways are mainly classified into the intrinsic (mitochondrial) pathway and the extrinsic (death ligand) pathway, and these classifications could be regulated by caspase-9 and caspase-8, respectively⁽ Reference Sharifi, Eslami and Larijani ^{75 )}. In this study, dietary glycinin exposure increased the mRNA levels of all of the apoptosis signalling, including caspase-3, caspase-8 and caspase-9 in the DI of juvenile fish, whereas it decreased caspase-8 and caspase-9 mRNA levels in the PI and MI of fish. These results indicated that glycinin-induced apoptosis signalling primarily occurred in the DI of fish, which followed a similar pattern to the effects of glycinin on intestinal TJ, indicating a possible relationship between glycinin-induced disruption of intestinal TJ and apoptosis signalling in fish. Although the patterns of apoptosis signalling are similar to the TJ patterns, the cytokines observed disorder patterns. For instances, glycinin exposure caused pro-inflammatory cytokine IL-1 β mRNA levels to increase and TNF- α mRNA levels to decrease in the PI and MI, whereas the IL-1 β mRNA levels decreased and the TNF-α mRNA levels increased in the DI of fish. For anti-inflammatory cytokines, glycinin exposure decreased only IL-10 mRNA levels in the PI and DI, and decreased only the TGF- β2 mRNA levels in the MI of fish. The reasons for these disorder patterns for cytokines are largely unknown but might be partially related to the different exposure time course in the PI, MI and DI of fish during the process when fish intake and digest feed. Studies have shown that the expression of three pro-inflammatory cytokine genes (IL-1 β, IL-8 and TNF- α) in the liver of juvenile salmons differed in the exposure timing and magnitude⁽ Reference Fast, Johnson and Jones ^{76 )}.

The TOR signalling pathway is a well-known pathway to regulate translation initiation, the limiting step in protein synthesis in animals⁽ Reference Tsukumo, Laplante and Parsyan ^{77 )}. Recently, it has been observed that the inhibition of TOR signalling promoted cell apoptosis and induced pro-inflammatory cytokine production⁽ Reference Zeng, Wang and Shi ^{21 ,} Reference Hu, Zhang and Feng ^{78 )}. However, the up-regulation of 4E-BP could inhibit TOR signalling⁽ Reference Tain, Mortiboys and Tao ^{79 )}. This study observed that dietary glycinin exposure alone decreased the TOR mRNA levels in the MI and DI and increased the 4E-BP mRNA levels in the DI, whereas it did not change the TOR and 4E-BP mRNA levels in the PI of fish. These observations indicated that glycinin exposure could inhibit TOR signalling in the MI and DI, more seriously in the DI than in the MI of fish. These effects followed similar patterns to the effects on the TJ and apoptosis signalling, suggesting that glycinin-induced TJ damage and apoptosis signalling were partially related to the inhibition of TOR signalling. However, a study from our laboratory observed that fish exposed to another dietary anti-nutrient factor, β-conglycinin, also caused the down-regulation of TOR signalling, but this change primarily occurred in the PI and MI⁽ Reference Zhang, Guo and Feng ^{24 )}. Considering those two studies together might partially explain how high doses of SBM impaired the intestinal structural integrity and caused poor growth of the fish observed in a previous study⁽ Reference Krogdahl, Bakke McKellep and Baeverfjord ^{73 )}.

Protective effects of glutamine against glycinin-induced negative effects on intestinal health and the growth performance of fish

Gln serves as a major fuel for intestinal epithelial cells in mammals⁽ Reference Windmueller ^{80 )}. Our previous study has demonstrated that dietary Gln supplementation could improve growth performance and the intestinal structure and function in Jian carp⁽ Reference Zhang, Guo and Feng ^{24 )}. According to these findings, we investigated the potential protection of Gln against glycinin toxicity in the present study. Interestingly, compared with the glycinin alone groups, for fish administered Gln with glycinin, the performance parameters (FBW, SGR and FI), intestinal growth (IW, IL, IPC and folds heights) and functional (activities of CK, Na⁺, K⁺-ATPase, AKP and γ-GT) indicators, as well as intestinal structural integrity indexes (ROS and MDA content, activities of SOD, CAT, GPx, GST and GR) were partially or completely close to or equal to those in the control group, demonstrating that Gln might mitigate the negative effects on fish growth and intestinal health caused by glycinin. Similar results of Gln against another dietary anti-nutrient factor, β-conglycinin, induced intestinal oxidative damage in juvenile Jian carp, as observed in our previous study⁽ Reference Zhang, Guo and Feng ^{24 )}. Those positive effects might be partially because of the fact that Gln has protective and reparative roles against oxidative damage. Our previous studies observed that co-treatment or post-treatment with Gln could protect or repair fish enterocytes from oxidative damage caused by H₂O₂ ⁽ Reference Chen, Zhou and Feng ^{9 ,} Reference Hu, Feng and Jiang ^{81 )}. Thus, because Gln was observed to mitigate the negative influences of glycinin or β-conglycinin observed in our previous study⁽ Reference Zhang, Guo and Feng ^{24 )}, it is reasonable for us to recommend supplementation with Gln when high levels of SBM (glycinin and β-conglycinin) are used in fish diets.

Conclusion

In this study, we report five primary, novel and interesting results. (1) Dietary glycinin exposure could depress the growth performance of fish, which might be partially related to the dysfunction of the intestines, which are the digestive/absorptive organs in stomach-less fish. (2) The dietary glycinin-induced poor growth and dysfunction of the intestines might be partially associated with the intestinal cellular oxidative damage and disruption of the cell–cell TJ. (3) The glycinin-induced intestinal cellular oxidative damage was, at least in part, related to the impaired enzymatic antioxidant ability (decreased SOD, CAT, GST and GR activities) and disturbed antioxidant enzyme gene expression, inducing adaptive increases in MnSOD, CuZnSOD, CAT, GPx1b and GPx4a mRNA levels, and decreasing the mRNA levels of nine GST isoforms. Meanwhile, dietary glycinin exposure also disturbed the antioxidant-related signalling factor mRNA levels, including Nrf2, Keap1a, Keap1b and eleven isoforms of PKC. (4) Dietary glycinin exposure-induced disruption of the cell–cell TJ was partially related to the modulation of cytokines, apoptosis and TOR signalling in the following order of seriousness DI>MI>PI. (5) Because Gln was observed to mitigate the negative influences of glycinin, it is reasonable for us to recommend supplementation with Gln when high levels of SBM (glycinin) are utilised in fish diets.