In general, carnivorous fish are characterised by a limited ability to utilise starch as energy source and thus are considered as glucose-intolerant(1). Long-term intake of excess dietary starch can induce hyperglycaemia, hepatic glycogen and lipid accumulation, and chronic inflammation response, as well as reduce the immune functions and antioxidant capabilities, which may lead to metabolic liver diseases (MLD)(Reference Hemre, Mommsen and Krogdahl2–Reference Lin, Shi and Mu5). Therefore, reducing the starch inclusion in carnivorous fish feed has become a trend to ensure its growth performance and health for many aquatic enterprises. However, the LS inclusion has made a challenge for the processing of floating feeds and product durability, which increases the energy consumption of processing and feed costs(Reference Sørensen, Morken and Kosanovic6).
Until now, numerous strategies have been used to improve the starch utilisation in fish, such as nutritional programming, genetic selection, exogenous enzymatic additives, macronutrient balance and the modified gut microbiome affected by tailored diets(Reference Enes, Panserat and Kaushik7–Reference Kamalam, Medale and Panserat10). Several studies showed that carnivorous fish own an adaptive mechanism in response to the fluctuation of dietary starch levels within a certain range in American eel (Anguilla rostrata)(Reference García-Gallego, Bazoco and Akharbach11), European sea bass (Dicentrarchus labrax)(Reference Enes, Panserat and Kaushik12), rainbow trout (Oncorhynchus mykiss)(Reference Kamalam, Medale and Kaushik13,Reference Geurden, Mennigen and Plagnes-Juan14) , White sturgeon (Acipenser transmontanus)(Reference Hung, Groff and Lutes15), Siberian sturgeon (Acipenser baerii)(Reference Gong, Xue and Wang9) and Japanese flounder (Paralichthys olivaceus)(Reference Yang, Deng and Pan16,Reference Yang, Deng and Pan17) . Generally, fish can improve hepatic glycogen synthesis and glycolysis and reduce gluconeogenesis to alleviate the stress induced by intake of excess starch diets(Reference Polakof, Panserat and Soengas3,Reference Enes, Panserat and Kaushik12) . Moreover, excess starch diets provide NADPH or carbon backbones for de novo lipogenesis and even induce an increase of lipogenic enzymatic activity and lipid accumulation in the liver(Reference Yang, Deng and Pan16,Reference Talukdar, Kumar and Varghese18) . Adaptation of hepatic enzymes to dietary starch levels has been reported consistently for fish species(Reference Ekmann, Dalsgaard and Holm19–Reference Zhou, Ge and Niu21). Moreover, Fu and Xie(Reference Fu and Xie22) found that southern catfish (Silurus meridionalis) could oxidise unwanted assimilated starch to accelerated energy expenditure by increasing oxygen consumption and fasting metabolic rate during starch overfeeding. Potential flexibility in glucose utilisation also existed in Japanese flounder in the absence of energy(Reference Yang, Deng and Pan16). The above results indicated that carnivorous fish had certain adaptability to high starch (HS).
Largemouth bass (Micropterus salmoides) is a commercially carnivorous species for freshwater culture with a high economic value in China(Reference Tidwell, Coyle and Bright23). MLD is prevalently induced by HS diet in largemouth bass(1). Previous studies demonstrated the dietary starch level should be less than 10 % to ensure growth performance and liver health(Reference Lin, Shi and Mu5,Reference Ma, Mou and Pu24,Reference Zhang, Xie and Wei25) , so the starch levels of commercial feed formulations were maintained less than 10 %. To our knowledge, information regarding the changes in intermediate metabolism and molecular adaptation of largemouth bass to HS diet has not been reported. Thus, the objective of the present study was to ascertain if largemouth bass owned certain adaptability to HS. The results of this study would provide insight into glucolipid and energy metabolism in largemouth bass in response to HS diet and shed light on the adaptative mechanism of HS in carnivorous fish.
Material and methods
During the whole experimental period, the fish were maintained in compliance with the Laboratory Animal Welfare Guidelines of China (General Administration of Quality Supervision, Inspection, and Quarantine of the People’s Republic of China, Standardization Administration of China, GB/T 35892-2018).
Growth trial and sample collection
Two isonitrogenous and isolipidic experimental diets with 9·06 % (LS) and 13·56 % (HS) starch were prepared, respectively. All ingredients were grinded and passed a 180-μm siever. Stuffs of each diet were well mixed with 13–15 % water for 20 min by a mixer (CH-100, The New Standard Powder Machinery Manufacturing Co., Ltd), then was processed into 3-mm diameter floating pellets under the following extrusion condition: feeding section (90°C/5 s), compression section (130°C/5 s) and metering section (150°C/4 s) using a Twin-screwed extruder (EXT50A, YANGGONG MACHINE, China). All diets were air-dried at room temperature and stored at −20°C until use. The diets formulations and analysed chemical compositions are shown in Table 1.
Formulation and composition of experimental diets (%)
LS, low starch; HS, high starch.
Mineral premix (mg/kg diets): CuSO4·5H2O 10; FeSO4·H2O 300; ZnSO4·H2O 200; MnSO4·H2O 100; KI (10 %) 80; CoCl2·6H2O (10 %Co) 5; Na2SeO3 (10 % Se) 10; MgSO4·5H2O 2000; NaCl 100; zeolite 4995; antioxidant 200.
* Fishmeal and fish oil were purchased form Triple Nine Fish Protein Co. Ltd.
† Tapioca starch and wheat flour were purchased from Beijing Nankou Flour Mill.
‡ Wheat gluten meal was purchased from Guanxian Xinrui Industrial Co., Ltd.
§ Cottonseed protein concentrate was purchased from Xinjiang Jinlan plant protein Co., Ltd.
|| Soyabean protein concentrate and soyabean oil were purchased from Yihai Kerry Investment Co. Ltd.
¶ Soyabean meal was purchased from Qingdao Bohai Biotechnology Co., Ltd.
** Vitamin premix (mg/kg diets): vitamin A 20; vitamin D3 10; vitamin K3 20; vitamin E 400; vitamin B1 10; vitamin B2 15; vitamin B6 15; vitamin B12 (1 %) 8; ascorbic acid (35 %) 1000; calcium pantothenate 40; niacinaminde 100; inositol 200; biotin (2 %) 2; folic acid 10; corn gluten meal 150; choline chloride (50 %) 4000.
†† Starch content was estimated based on the starch content of tapioca starch (72 % starch) and wheat flour (60 % starch).
Largemouth bass was obtained from the commercial Aquafarm. Before the formal feeding trial, fish were acclimatised and fed LS diet with a rate equalling 2 % of wet body weight per d for 2 weeks (2w). Fish (initial body weight was 47·60 ± 0·20 g) were distributed into sixteen cylindrical plastic tanks (capacity: 256 l) with twenty-five fish per tank after 24-h starvation, and each diet was randomly assigned to eight tanks. Fish were fed to apparent satiation twice daily at 08:00 h and 17:00 h. During the experiment, the water temperature was maintained at 21–25°C, pH = 7·2–8·0, dissolved oxygen > 6·0 mg/l and ammonia-N < 0·3 mg/l.
Fish were anesthetised with 200 mg/l of MS-222 (Sigma) at the end of the 2w or 8 weeks (8w) before sampling. Individual body weight, body length and liver weight of five fish in each tank after starvation for 24 h were recorded to calculate condition factor and hepatosomatic index. Blood was rapidly sampled from the caudal vein, centrifuged (4000 g, 10 min, 4°C) to obtain plasma for the analysis of haematological parameters at postprandial 3 h and 24 h. Liver, anterior intestine and muscle samples at postprandial 24 h were dissected and then immediately frozen in liquid N2 and kept at −80°C for mRNA isolation and tissue homogenate analysis until used. Three liver samples near the bile duct in each tank were fixed in 4 % paraformaldehyde (P1110, Solarbio) for histology determination. Three liver samples near the bile duct in each treatment were fixed in 2·5 % glutaraldehyde (P1126, Solarbio) for ultrastructure analysis. Livers from another three fish in each tank were pooled into ziplock bags and then stored at −20°C for the assay of crude lipid.
Chemical compositions analysis of diets
The crude protein, crude lipid, crude ash, moisture and gross energy contents of experimental diets were analysed according to standard methods as previously described materials and methods section(Reference Yu, Zhang and Chen26,Reference Zhang, Chen and Liang27) .
Plasma and hepatic homogenate parameters
Plasma alanine aminotransferase (C009–2–1), aspartate aminotransferase (C010–2–1), total protein (A045–2–2), albumin (H127–1–2), glucose (361 500), TAG (A110–2–1), total cholesterol (TC, A111–2), alkaline phosphatase (A059–2–2), total bile acid (TBA, E003–2–1), LDL-cholesterol (A113–2), HDL-cholesterol (A112–2) and hepatic TAG, TC, TBA, HDL-cholesterol, glycogen (A043), superoxide dismutase (SOD, A001–3–2), catalase (CAT, A007–1–1), malondialdehyde (MDA, A003–1–2) and intestinal amylase (C016–1–1) were determined by commercial assay kits (Nanjing Jiancheng Co.) following the protocols. The glucagon (MM-3294801), insulin (MM-190901), cyclic-AMP (MM-3259102) and reactive oxygen species (ROS, MM-091101) were determined by the commercial kits (Jiangsu Meimian industrial Co. Ltd). Plasma and hepatic NEFA (KRB0081) were measured by assay kit of Wako Pure Chemical Industries, Ltd (Wako).
Hepatic histopathological and immunofluorescence examination
Liver samples were fixed, dehydrated, embedded, stained for haematoxylin and eosin, periodic acid Schiff (PAS) or sirius red and observed by light microscopy (DM2500, Leica) according to methods of our laboratory as previously described(Reference Yu, Zhang and Chen26,Reference Zhang, Chen and Liang27) . The results of haematoxylin –eosin and PAS staining were observed by light microscopy (DM2500, Leica). The sample treatment and transmission electron microscopy observation were conducted according to the methods described by Lu et al.(Reference Lu, Ma and Wang28). The immunofluorescence test for cleaved caspase 3 were following the previously described(Reference Yu, Zhang and Chen26,Reference Zhang, Chen and Liang27) . Anti-cleaved caspase 3 (ab13847, Abcam) was used as the primary antibody. Alexa Flour 555 antibody (A21428, Life Technologies) was used as the secondary antibody. The fluorescent signal was captured using Zeiss LSM700.
Quantitative real-time PCR
Total RNA extraction and cDNA synthesis were carried out as described previously(Reference Yu, Zhang and Chen26,Reference Zhang, Chen and Liang27) . The quantitative PCR analysis was performed using a CFX96TM Real-Time System (Bio-Rad) using iTaqTM Universal SYBR® Green Supermix (1 725 121, Bio-Rad). Each sample was run in triplicate and analysed using the 2–ΔΔCt. Elongation factor 1α, EF1α (GenBank accession no. KT827794), was used as an endogenous reference gene. The primer sequences are shown in Table 2.
Primer sequences used in this study
EF1α, elongation factor-1α; GK, glucokinase; PK, pyruvate kinase; PCK, phosphoenolpyruvate carboxykinase cytosolic; G6Pase, glucose-6-phosphatase catalytic subunit; GLUT2, glucose transporter type 2; SGLT1, sodium/glucose cotransporter 1; ACC1, acetyl-CoA carboxylase 1; FASN, fatty acid synthase; LPIN1, phosphatidate phosphatase1; GPAT4, glycerol-3-phosphate acyltransferase 4; DGAT1, diacylglycerol O-acyltransferase 1; ATGL, adipose triglyceride lipase; HSL, hormone-sensitive lipase; HADH, hydroxyacyl-CoA dehydrogenase; ACADM, acyl-CoA dehydrogenase medium chain; CPT1α, carnitine palmitoyltransferase 1α; PPARα, peroxisome proliferator-activated receptor α; LPL, lipoprotein lipase; PDH, pyruvate dehydrogenase E1 subunit (α or β); HMGCR, 3-hydroxy-3-methylglutaryl-CoA reductase; CYP7A1, cytochrome P450 family 7 subfamily a member 1; FXR, farnesoid X-activated receptor; CS, citrate synthase; IDH3a, isocitrate dehydrogenase 3 (NAD+) alpha; SDH, Succinate Dehydrogenase Complex Iron Sulfur Subunit; CREB, cAMP-responsive element binding protein; TGFβ1, transforming growth factor β1.
Statistical analysis
All statistical procedures were performed with the aid of the SPSS software version 22.0 for Windows (IBM Inc.). After the homogeneity of the variances was tested, all data means were analysed. An independent t test was used to compare the differences between the two groups. Two-way ANOVA was used to analyse the significant differences among treatment means based on postprandial times, starch levels and their interactions for plasma glucose, insulin and glucagon levels. Statistical significance was determined at P < 0·05. All results were presented as standard error of the mean, and the graphics were drawn by GraphPad Prism Software version 7.0.
Results
Growth performance and morphometric parameters
The results of growth performance and morphometric parameters are presented in Table 3. The SR of largemouth bass in all groups was above 99 %. No significant differences were observed on final body weight, specific growth rate, feed conversion ratio, feeding rate and condition factor between the LS and HS groups at 2w. However, the HS diet improved hepatosomatic index significantly at 2w. During 8w, the final body weight and specific growth rate significantly decreased in the HS group, but the feed conversion ratio, feeding rate, condition factor and hepatosomatic index were not significantly affected.
Effects of HS diet on the growth performance and morphometric parameters of largemouth bass at 2w and 8w (Mean values with their standard errors of the mean, n 4)
HS, high starch; LS, low starch; IBW, initial mean weight; FBW, final body weight; SR, survival rate; SGR, specific growth rate; FCR, feed conversion ratio; FR, feeding rate; CF, condition factor; HIS, hepatosomatic index.
* IBW (g/fish): n 4.
† FBW (g/fish): n 4.
‡ SR (%) = 100 × final fish number/initial fish number, n 4.
§ SGR (%/d) = 100 × (Ln (Wf/Wi))/d, n 4.
|| FCR = feed intake/(Wf + Wd-Wi), n 4.
¶ FR (% bw/d) = 100 × feed intake/[(Wf + Wi + Wd)/2]/d n 4.
** CF (g/cm3) = 100 × (body weight, g)/(body length, cm)3, n 20.
†† HSI (%) = 100 × liver weight/whole body weight, n 20. Wf is the final total weight, Wd is the total weight of dead fish, Wi is the initial total weight.
Haematological parameters and hepatic antioxidant responses
The haematological parameters and hepatic antioxidant responses of largemouth bass are presented in Table 4, 5 and 6. The plasma glucose levels after postprandial 24 h at 2w or 8w in the LS or HS group were significantly lower than that postprandial 3 h, and the glucagon levels were significantly higher than that postprandial 3 h, but no significant differences were observed in insulin levels. HS diet significantly increased plasma glucose levels at 3 h postprandial at 8w compared with the LS diet (Table 4). No significant differences were detected in plasma total protein, albumin, IgM, alanine aminotransferase and aspartate aminotransferase between LS and HS groups at postprandial 24 h at 2w or 8w. Plasma alkaline phosphatase and TBA contents enhanced significantly in the HS group at postprandial 24 h at 2w, but no significant differences were observed at 8w (Table 5). The levels of plasma ROS, hepatic ROS and MDA in the HS group were markedly higher than LS group at 2w, whereas the opposite was true for the SOD and CAT. At 8w, no significant differences were observed in plasma ROS, hepatic ROS and MDA levels between two groups (Table 6).
Effects of HS diet on plasma glucose, insulin and glucagon of largemouth bass at postprandial 3 h or 24 h at 2w and 8w (Mean values with their standard errors of the mean, n 8)
HS, high starch; LS, low starch.
Effects of HS diet on plasma immune and hepatic function parameters of largemouth bass at postprandial 24 h at 2w and 8w (Mean values with their standard errors of the mean, n 8)
HS, high starch; LS, low starch; TP, total protein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AKP, alkaline phosphatase; TBA, total bile acid.
Effects of HS diet on antioxidant responses of largemouth bass at postprandial 24 h at 2w and 8w (Mean values with their standard errors of the mean, n 8)
HS, high starch; LS, low starch; ROS, reactive oxygen species; MDA, malondialdehyde; SOD, superoxide dismutase; CAT, catalase.
Hepatic glucose metabolism, intestinal glucose transporter and amylase activity
At 2w, the mRNA levels of both glycolysis (GK (glucokinase) and PK (pyruvate kinase)) and gluconeogenesis-related genes (PCK (phosphoenolpyruvate carboxykinase cytosolic) and G6Pase (glucose-6-phosphatase catalytic subunit)) were significantly up-regulated in the HS group, but no significant differences were observed at 8w (Fig. 1(a) and (b)). Meanwhile, enzyme activities of hepatic GK, G6Pase and PCK in the HS group at 2w were also significantly increased, but no significant differences were observed at 8w (Fig. 1(d)–(f)). The mRNA levels of pyruvate dehydrogenase genes (PDHA and PDHB) related to aerobic oxidation were significantly up-regulated at 2w, while they did not demonstrate any change at 8w (Fig. 1(c)).
Effects of HS diet on hepatic glucose metabolism, intestinal glucose transporter and amylase activity of largemouth bass at postprandial 24 h at 2w and 8w. (a) Transcriptional levels of hepatic GK and PK. (b) Transcriptional levels of hepatic PCK and G6Pase. (c) Transcriptional levels of pyruvate dehydrogenase-related genes (PDHA and PDHB). (d) Hepatic GK activity. (e) Hepatic PCK activity. (f) Hepatic G6Pase activity. (g) Transcriptional levels of intestinal GLUT2 and SGLT1. (h) Intestinal amylase activity. (i) Transcriptional levels of hepatic GLUT2. Values marked with ‘*’ are significant differences (P < 0·05) (n 8). GK, glucokinase; PK, pyruvate kinase; PCK, phosphoenolpyruvate carboxykinase cytosolic; G6Pase, glucose-6-phosphatase catalytic subunit; PDH, pyruvate dehydrogenase E1 subunit (α or β); GLUT2, glucose transporter type 2; SGLT1, sodium/glucose cotransporter 1.
The mRNA levels of intestinal Na+/glucose cotransporter type 1 (SGLT1) were significantly up-regulated in the HS group at 2w, but no significant differences were observed in glucose transporter type 2 (GLUT2) mRNA levels and amylase activity. At 8w, HS diet significantly increased intestinal GLUT2 and SGTL1 mRNA levels, and amylase activity (Fig. 1(g) and (h)). At 2w, the hepatic GLUT2 mRNA levels were significantly up-regulated in the HS group, but no significant differences were observed at 8w (Fig. 1(i)).
Hepatic lipid and total bile acid metabolism
As shown in Fig. 2, hepatic lipid metabolism disorder with the symptoms of TAG and TC accumulation appeared in largemouth bass fed HS diet at 2w. Compared with the LS group, lower plasma TAG and higher hepatic NEFA and TAG levels were observed in the HS group at 2w (Fig. 2(a) and (d)). Meanwhile, the mRNA levels of hepatic fatty acid (ACC1 (acetyl-CoA carboxylase 1) and fatty acid synthase (FASN)) and TAG synthesis (GPAT4 (glycerol-3-phosphate acyltransferase 4) and LPIN1 (phosphatidate phosphatase1))-related genes were significantly up-regulated, while TAG hydrolysis (ATGL (adipose triglyceride lipase) and LPL (lipoprotein lipase)) and β-oxidative (HADH (hydroxyacyl-CoA dehydrogenase), ACADM (acyl-CoA dehydrogenase medium chain) and CPTα (carnitine palmitoyltransferase 1α)) genes were significantly down-regulated (Fig. 2(a)). The accumulation of hepatic TC induced by HS was accompanied with higher plasma HDL-cholesterol levels (Fig. 2(d)) and up-regulated HMGCR (3-hydroxy-3-methylglutaryl-CoA reductase) mRNA levels (Fig. 3(a)). Moreover, up-regulated mRNA levels of bile acid synthesis gene CYP7A1 (cytochrome P450 family 7 subfamily a member 1) and increased contents of TBA in the liver were also induced by HS diet at 2w (Fig. 3).
Effects of HS diet on hepatic lipid metabolism of largemouth bass at postprandial 24 h at 2w and 8w. Plasma NEFA and TAG levels, hepatic NEFA and TAG contents, and transcriptional levels of hepatic FA synthesis (ACC1 and FASN), TAG synthesis (GPAT4, DGAT1 and LPIN1), TAG hydrolysis (ATGL, HSL and LPL) and β-oxidation (HADH, ACADM, CPT1α and PPARα)-related genes at 2w (a) and 8w (b). (c) Liver lipid contents. Plasma LDL-cholesterol, HDL-cholesterol and TC levels, and hepatic LDL-cholesterol and TC contents at 2w (d) and 8w (e). Values marked with ‘*’ are significant differences (P < 0·05) (n 8). HS, high starch; ACC1, acetyl-CoA carboxylase 1; FASN, fatty acid synthase; GPAT4, glycerol-3-phosphate acyltransferase 4; DGAT1, diacylglycerol O-acyltransferase 1; LPIN1, phosphatidate phosphatase1; ATGL, adipose triglyceride lipase; HSL, hormone-sensitive lipase; LPL, lipoprotein lipase.
Effects of HS diet on hepatic TBA metabolism of largemouth bass at postprandial 24 h at 2w and 8w. (a) Transcriptional levels of hepatic cholesterol synthesis (HMGCR) and bile acid synthesis (CYP7A1, FXR and CYP8B1)-related gene. (b) The hepatic TBA contents. Values marked with ‘*’ are significant difference (P < 0·05) (n 8). LS, low starch; HS, high starch; TBA, total bile acid; HMGCR, 3-hydroxy-3-methylglutaryl-CoA reductase; CYP7A1, cytochrome P450 family 7 subfamily a member 1; FXR, farnesoid X-activated receptor; CYP8B1, Cytochrome P450 Family 8 Subfamily B Member 1.
Interestingly, the symptoms of TAG, TC and TBA accumulation symptoms induced by HS diet could be alleviated or disappear in largemouth bass at 8w (Figs. 2 and 3).
Hepatic inflammatory and apoptosis responses along with histological analysis
As shown in Fig. 4(a), three typical hepatic phenotypes were observed with symptoms from light to severe by haematoxylin–eosin staining, PAS staining, sirius red staining and apoptosis signals of cleaved caspase 3 in which (I) no obvious abnormality phenotype with shaped hepatocytes and clearly cell nuclei, and with the negative response to cleaved caspase 3; (II) nuclear dense phenotype, with unclear liver cord, which is usually a precursor to liver fibrosis in the clinic with increased micro-vascular collagen and highlighted cleaved caspase 3 signal; and (III) hepatic fibrosis symptoms, along with increased collagen fibres in red by sirius staining and intensive apoptosis (cleaved caspase 3) around the fibrosis tissues. At 2w, two samples were observed nuclear dense, and two samples with serious fibrosis symptoms in the HS group, while eleven samples were no obvious abnormality in the LS group. At 8w, all the sections were no obvious abnormality in the two groups. The PAS staining showed that hepatic glycogen content was increased as the grade of liver disease aggravated. Moreover, HS-induced hepatic glycogen contents significantly increased at 2w, while no significant difference was observed at 8w (Fig. 4b and c). Ultrastructural observation showed that no obvious abnormal hepatocytes exhibited normal euchromatic nucleus, large stacks of the endoplasmic reticulum (ER), abundant mitochondria with intact cristae and uniformly distributed glycogen granules in the cytoplasm; however, hepatocytes of liver fibrosis revealed extensive cellular damage, including irregular nuclear envelope and dense clumped chromatin, a lot of fat droplets, disruption and swelling of ER membranes, mitochondria with broken cristae (Fig. 4(d)).
Effects of HS diet on hepatic histopathological, inflammatory and apoptosis responses of largemouth bass at 2w and 8w. (a) Three phenotypes of hepatic histopathological examination with symptoms from light to heavy by HE staining. PAS staining for glycogen examination (the pink spot represented the glycogen particles marked by black arrows); Sirius red staining for hepatic fibrosis (the red showed collagen fibres marked by green arrow); apoptosis signals of cleaved caspase 3 in red colour, and DAPI for nucleus (bar = 100 μm), in which (I) no obvious abnormality; (II) nuclear dense; and (III) hepatic fibrosis symptoms. (b) Statistical results of liver phenotypes and glycogen accumulation (n 12). Since the samples were damaged during the embedding process, the number of slices was less than 12. (c) Hepatic glycogen content (n 4). (d) Liver ultrastructure. Note the part of mitochondria (M), nucleus (N), endoplasmic reticulum (ER), glycogen granules (Gly) (red arrows showed) and lipid drop (LD). Fibrosis liver showing severe hepatocyte damage. Observe nucleus with increased heterochromatin patches, irregular nuclear envelope, dense clumped chromatin, disruption and swelling of ER membranes, mitochondria with broken cristae, large lipid droplets. (e) Effects of HS diet on the transcriptional levels of hepatic pro- and anti-inflammation-related genes. (f) Effects of HS diet on the transcriptional levels of hepatic apoptosis-related genes. Values marked with ‘*’ are significant difference (P < 0·05) (n 8). HE, haematoxylin and eosin; PAS, periodic acid Schiff; DAPI, 4',6-diamidino-2-phenylindole; LS, low starch; HS, high starch.
The mRNA levels of both pro- (TNFα and IL1β) and anti-inflammatory cytokines (IL10 and TGFβ1 (transforming growth factor β1)) were significantly up-regulated in the HS group at 2w, but no significant differences were observed in IL1β, IL10 and TGFβ1 mRNA levels at 8w (Fig. 4(e)). Compared with 2w, the mRNA levels of HS-induced TNFα were alleviated at 8w, while still significantly higher than the LS group at 8w (Fig. 4(e)).
The mRNA levels of apoptosis-related genes (Caspase 8 and Caspase 9) were significantly up-regulated in the HS group at 2w, while no significant differences were observed between LS and HS groups at 8w. No significant differences were detected on Caspase 3 mRNA levels in the two groups at 2w or 8w (Fig. 4(f)). These results showed that HS might induce short-term hepatic inflammatory and apoptosis responses.
Energy metabolism
Accelerated energy metabolism rate in largemouth bass fed HS diet was observed at 2w (Fig. 5). The significantly increased plasma and hepatic cyclic-AMP contents (Fig. 5(a) and (b)) and up-regulated mRNA levels of the CREB (cAMP-responsive element binding protein) (Fig. 5(c)) promoted the energy metabolism by accelerating the glycolysis and gluconeogenesis processes in the HS group at 2w (Fig. 1(a)). The mRNA levels of tricarboxylic acid cycle-related genes (IDH3a (isocitrate dehydrogenase 3 (NAD+) alpha), SDHa and SDHb (Succinate Dehydrogenase Complex Iron Sulfur Subunit)) were significantly up-regulated in the HS group at 2w. After 8w feeding, the mRNA levels of HS-induced IDH3a were mitigated compared with 2w, while still significantly higher than the LS group, but the mRNA levels of SDHa and SDHb had no significant differences between the two groups (Fig. 5(d)).
Effects of HS diet on hepatic energy metabolism of largemouth bass at postprandial 24 h at 2w and 8w. (a) Plasma cAMP contents. (b) Hepatic cAMP contents. (c) Transcriptional levels of hepatic CREB. (D) Transcriptional levels of hepatic TCA cycle-related genes (CS, IDH3a, SDHa and SDHb). Values marked with ‘*’ are significant difference (P < 0·05) (n 8). cAMP, cyclic-AMP; CREB, cAMP-responsive element binding protein; TCA, tricarboxylic acid; CS, citrate synthase; IDH3a, isocitrate dehydrogenase 3 (NAD+) alpha; SDH, Succinate Dehydrogenase Complex Iron Sulfur Subunit.
Discussion
Generally, carnivorous fish are known as glucose-intolerant with persistent hyperglycaemia after intake of HS diet(Reference Hemre, Mommsen and Krogdahl2,Reference Enes, Panserat and Kaushik29) . Insulin and glucagon are two important endocrine hormones to regulate the glucose homoeostasis of fish by controlling the plasma glucose levels(Reference Polakof, Panserat and Soengas3). Increased plasma glucose levels can stimulate insulin or reduce glucagon secretion to relieve hyperglycaemia stress for maintaining glucose homoeostasis(Reference Hemre, Mommsen and Krogdahl2). Lin et al.(Reference Lin, Shi and Mu5) showed that 20 % wheat starch diets induced a significant rise in plasma glucose and insulin levels at postprandial 6 h in largemouth bass. In contrast, the present study investigated HS significantly increased plasma glucose levels at postprandial 3 h, while it did no affect plasma insulin levels. Similar results also were found in eel(Reference Suárez, Sanz and Bazoco30), silver sea bream (Sparus sarba)(Reference Leung and Woo31) and common dentex (Dentex dentex)(Reference Pérez-Jiménez, Abellán and Arizcun32). Our previous studies showed plasma insulin levels remained unchanged except at 1 h in largemouth bass after a glucose load(Reference Chen, Wu and Gu33). However, after bovine insulin injection, the plasma glucose levels of largemouth bass significantly decreased, then restored to the basic values in 6 h, which further confirmed that insufficient secretion of insulin was the main reason for glucose intolerance(Reference Chen, Wu and Gu33).
The absorption of glucose by the blood from the enteric cavity is facilitated by two key glucose transporters, SGLT1 and GLUT2 (Reference Ghezzi, Loo and Wright34). In addition, GLUT2 can affect the capacity of the glucose transfer between liver and blood(Reference Polakof, Mommsen and Soengas8). Several studies had indicated that dietary starch significantly up-regulated SGLT1 expression in fish, such as rainbow trout(Reference Polakof, Míguez and Soengas35) and yellow catfish (Pelteobagrus fulvidraco)(Reference Zhao, Yang and Chen36). The GLUT2 expression in common carp (Cyprinus carpio) foregut was significantly up-regulated after 3 h with a glucose load(Reference Deng, Yan and Zhao37). HS diet (20 %) significantly up-regulated the expression of hepatic GLUT2 after postprandial 24 h in rainbow trout(Reference Panserat, Plagnes-Juan and Kaushik38). In tilapia (Oreochromis mossambicus), glucose injection did not affect hepatic GLUT2 expression(Reference Liu, Wang and Wan39). Until now, the regulatory relationship between starch levels and hepatic GLUT2 mRNA levels was not clear. Röder et al.(Reference Röder, Geillinger and Zietek40) and Ghezzi et al.(Reference Ghezzi, Loo and Wright34) showed that up-regulated mRNA levels of SGTL1 and GLUT2 might be critical to the development of glucose tolerance. Increased amounts of glucose transporters implied enhanced capacity of the intestine to transport and absorb glucose(Reference Lehmann and Hornby41). In this study, the intestinal SGTL1 and GLUT2 and hepatic GLUT2 mRNA levels were significantly up-regulated in the HS group at 8w, which strongly indicated that largemouth bass had the adaptability to HS diet. Moreover, the intestinal amylase activity was significantly increased at 8w in response to HS diet. A positive correlation between amylase activity and starch levels was used to evaluate dietary starch adaptation in fish by the genome analysis(Reference Heras, Chakraborty and Emerson42,Reference Karasov43) , which indicated that increased amylase activity was beneficial to improve the utilisation of starch diet. These results indicated that 13·56 % of starch diet could be effectively absorbed and utilised by the largemouth bass, which further illustrated that largemouth bass had the adaptability to HS diet.
In the present study, HS intake in short-term (2w) induced up-regulated mRNA levels and enzyme activities of glycolysis (GK and PK) and gluconeogenic (PCK and G6Pase)-related genes. Previous studies showed that the expression of hepatic glycolysis-related genes increased significantly with increasing starch levels, whereas the opposite was obtained for hepatic gluconeogenic enzymes in largemouth bass(Reference Lin, Shi and Mu5,Reference Ma, Mou and Pu24,Reference Zhang, Xie and Wei25) , rainbow trout(Reference Panserat, Plagnes-Juan and Kaushik38) and golden pompano (Trachinotus ovatus)(Reference Zhou, Ge and Niu21). However, Li et al(Reference Li, Liu and Yu44) and Song et al.(Reference Song, Shi and Lin45) reported that hepatic gluconeogenic enzyme mRNA levels or activities were closely related to different dietary starch sources. Therefore, the different regulation of hepatic gluconeogenesis by HS diet should be further investigated. In the present study, an abnormally up-regulated gluconeogenic pathway indicated HS diet-induced short-term glucose metabolic disorder. The poor inhibition of gluconeogenesis was considered as one of the reasons for the glucose intolerance of carnivorous fish(Reference Panserat, Plagnes-Juan and Kaushik38,Reference Panserat, Médale and Breque46,Reference Zhou, Wang and Xie47) . The persistently high level of gluconeogenesis (endogenous glucose production) independent of HS diet may lead to a putative competition between exogenous glucose (feed) and endogenous glucose as the source of energy, which may explain the poor starch utilisation in carnivorous fish(Reference Enes, Panserat and Kaushik7). At 8w, no significant differences were observed in the glycolysis and gluconeogenesis enzymes between LS and HS groups, which implied the largemouth bass improved starch utilisation capacity by regulating enzymatic activity. Adaptation of hepatic glucose metabolic enzymes to HS diet had been also reported in gilthead sea bream (Sparus aurata)(Reference Ekmann, Dalsgaard and Holm19), blunt snout bream (Megalobrama amblycephala)(Reference Li, Wang and Liu20) and golden pompano(Reference Zhou, Ge and Niu21).
In the present study, the glucose metabolism disorder with up-regulated glycolysis and pyruvate aerobic oxidation, which could overproduce acetyl-CoA for energy metabolism, TAG and TC biosynthesis(Reference Shi and Tu48). Significant higher hepatic NEFA and TAG contents but lower plasma TAG were induced by HS diet at 2w, indicating that overproduced acetyl-CoA made a high contribution to TAG synthesis. Meanwhile, the expression of hepatic fatty acid and TAG synthesis-related genes were significantly up-regulated, while TAG hydrolysis and β-oxidative-related genes were significantly down-regulated, which further confirmed HS induced short-term TAG accumulation in the liver. In the present study, we observed that four of twelve samples in the HS group at 2W showed nuclear dense or fibrosis rather than fatty liver pathological symptom, and this could be related to a fast development of hepatic cell damage. We suppose that a continually sampling, like day by day, could help to get an exact time point for the development from fatty liver (including steatohepatitis) to fibrosis. In general, dietary glucose enters the liver, be stored as glycogen or be converted into lipids if in excess(Reference Hemre, Mommsen and Krogdahl2,Reference Polakof, Panserat and Soengas3) . PAS staining is useful for identifying glycogen deposition(Reference Murli Krishna49). In this study, HS diet induced hepatic glycogen accumulation by PAS staining, together with increased content of the hepatic glycogen and lipid at 2w, strongly indicated that HS induced hepatic glycogen and lipid accumulation, then led significantly increased hepatosomatic index. Previous studies investigated that the glycogen and lipid accumulation was positively in response to HS intake, which was an effective way to reduce glucose stress(Reference Polakof, Panserat and Soengas3,Reference Enes, Panserat and Kaushik12) . However, consistent with previous studies on largemouth bass(Reference Lin, Shi and Mu5,Reference Ma, Mou and Pu24,Reference Zhang, Xie and Wei25) , the excess accumulation of glycogen and lipid in the liver may induce liver damage and lead to MLD in this study. Both plasma alanine aminotransferase and aspartate aminotransferase are the main indicators for evaluating liver function. Significant enhancements in the activity of these indicators usually along with liver damage of largemouth bass(Reference Yu, Zhang and Chen26,Reference Guo, Kuang and Zhong50) . In this study, no significant differences were observed in plasma alanine aminotransferase and aspartate aminotransferase activities between the LS and HS group at 2w or 8w, but HS induced liver damage (4/12) by histopathological examination at 2w. Therefore, we inferred that HS-induced liver damage of largemouth bass was within the repairable range. These results indicated that the excess accumulation of hepatic glycogen and lipid-induced acute liver damage in largemouth bass, but the recovery of liver health in the long-term remained to be clarified. The significantly higher hepatic TC content, plasma HDL-cholesterol and HDL-cholesterol/TC ratio were also observed in the HS group at 2w, indicating that overproduced acetyl-CoA also induced hepatic TC accumulation. Meanwhile, hepatic TC deposition also resulted in bile acids accumulation with higher plasma alkaline phosphatase levels and hepatic TBA content. Conversion of cholesterol to bile acids is critical for maintaining cholesterol homoeostasis and preventing the accumulation of TC, TAG and toxic metabolites, and injury in the liver(Reference Russell51). Bile acids are used as metabolic regulators to maintain glucose, lipid and energy metabolism homoeostasis and signalling molecules to protect against inflammation in the liver(Reference Chiang52). In largemouth bass, bile acids supplementation alleviated MLD induced by HS diet via decreasing the hepatic lipid content(Reference Yu, Zhang and Chen26,Reference Guo, Kuang and Zhong50) . Therefore, higher hepatic bile acids content induced by HS diet in short term was a positive response to protect against abnormal TAG and TC accumulation in the liver.
Under normal circumstances, the production and elimination of ROS maintain a dynamic balance in the liver. Once the environmental stress occurs (HS stress), ROS levels can increase dramatically and surpass the removal capacity of the antioxidant system, which may result in oxidative damage(Reference Nieves-Cordones, López-Delacalle and Ródenas53). MDA production increased is considered as a sign of oxidative stress, and SOD and CAT were able to protect cells or tissues from peroxidation(Reference Hasanuzzaman, Bhuyan and Zulfiqar54). In this study, HS induced higher hepatic MDA content, whereas SOD and CAT showed the opposite trend at 2w, which was consistent with the previous studies in largemouth bass(Reference Lin, Shi and Mu5,Reference Ma, Mou and Pu24,Reference Guo, Kuang and Zhong50) . These results suggested that excessive starch intake lowered the hepatic antioxidant abilities of largemouth bass. Growing evidences testified that glycolipid metabolism disorder and abnormal hepatic TAG and TC accumulation could cause strong oxidative stress(Reference Nieves-Cordones, López-Delacalle and Ródenas53,Reference Zhang, Xu and Yu55,Reference Awad, Haleem and Elbakly56) . Moreover, metabolic functions in fish also were impaired under strong oxidative stress(Reference Lin, Shi and Mu5). Therefore, this acute oxidative stress in largemouth bass was owed to the metabolic disorder induced by HS diet, which led to a reduction in activities of CAT and SOD at 2w. Albumin and IgM are the most widely studied immunoglobulin in fish, which could be a good biomarker for evaluating the immune status of fish(Reference Uribe, Folch and Enríquez57). In this study, no significant differences were observed in plasma albumin and IgM levels at 2w or 8w. So, we may infer that although largemouth bass fed HS diet had a risk of oxidative stress, the ROS levels were still under controllable status.
ROS can cause the production of inflammatory responses, ultimately triggering liver apoptosis even necrosis(Reference Nieves-Cordones, López-Delacalle and Ródenas53,Reference Reuter, Gupta and Chaturvedi58) . Two stages of inflammation exist, acute and chronic inflammation. Acute inflammation persists only for a short time and is protective response to eliminate the initial cause of cell injury and initiate tissue repair(Reference Karin and Clevers59). If the inflammation lasts for a longer time, the chronic inflammation sets in and may predispose the host to various tissues damage, including liver, kidney and intestine(Reference Zhang, Chen and Liang27,Reference Lin and Karin60,Reference Wei, Chen and Liang61) . During the wound healing stage, the significant up-regulated pro-inflammatory cytokine (IL17) and anti-inflammatory cytokine (IL10) expression were observed in the liver of grass carp (Ctenopharyngodon idella), whereas the opposite was shown in the recovery stage(Reference Wu, Wang and Cui62). In this study, at 2w, hepatic pro- and anti-inflammatory cytokines mRNA levels were significantly up-regulated in the HS group, while no significant differences were observed at 8w, which indicated HS diet could induce largemouth bass acute inflammation and ‘self-repair’ response. Similar results also were reported in largemouth bass(Reference Yu, Zhang and Chen26) and Japanese sea bass(Reference Zhang, Chen and Liang27). If the inflammation is not overcome or treated effectively, prolonged inflammatory responses lead to hepatic fibrosis and apoptosis(Reference Lin and Karin60,Reference Lee and Friedman63) . Cleaved caspase 3 is the most important key executioner in both exogenous and endogenous apoptosis processes activated by promoter-type Caspase 8 and Caspase 9, respectively(Reference Elmore64). Activated caspases trigger compensatory proliferation, referred to as apoptosis-induced proliferation which maintains tissue homoeostasis following massive stress-induced cell death, regenerating lost tissue(Reference Yu, Zhang and Chen26,Reference Fogarty and Bergmann65) . The significant up-regulated mRNA levels of Caspase 8 and Caspase 9 and more cleaved caspase 3 signals (four of twelve samples with apoptosis signals) in the HS group at 2w, but eliminating apoptosis responses at 8w, implied that apoptosis induced by HS might contribute to the liver damage repair.
The cyclic-AMP, which is produced by glycolysis and tricarboxylic acid, has been firmly established as second messenger molecule to regulate energy homoeostasis(Reference Boulekbache66). This effect appeared to be mediated by activating CREB at the mRNA level(Reference Wahlang, McClain and Barve67). In this study, significantly increased plasma and hepatic cyclic-AMP contents and up-regulated mRNA levels of the CREB and tricarboxylic acid cycle-related genes (IDH3a, SDHa and SDHb) were observed at 2w, indicating largemouth bass may up-regulate energy metabolism in response to HS diet. This was similar with previous results, which showed that HS diet could accelerate tricarboxylic acid cycle in largemouth bass(Reference Zhang, Xie and Wei25). Inflammation could contribute to an increase in energy expenditure in patients with kidney diseases, obesity, anaemia, or haemodialysis(Reference Akohoue, Shankar and Milne68–Reference Rogers, Perfield and Strissel71). The accelerated energy metabolism impeded growth performance and inflammation occurrence in sea bass fed plant protein feed(Reference Zhang, Chen and Liang27). In addition, southern catfish could oxidise unwanted assimilated starch by accelerating energy expenditure when overfed dietary starch(Reference Fu and Xie22), which is a ubiquitous adaptation mechanism in response to HS diet.
According to the above results, accelerated energy metabolism rate and bile acid synthesis were the main ways for largemouth bass to self-repair in response to acute inflammation induced by HS diet. Besides, the lower specific growth rate of largemouth bass in the HS group at 8w could be ascribed to accelerated energy consumption. It was a pity that a longer growth time had not implemented to clarify whether the compensatory growth existed.
Conclusions
In summary, due to insufficient insulin secretion of largemouth bass, intake of HS diet-induced short-term glucose and lipid metabolism disorder along with TAG and TC accumulation symptoms further caused oxidative stress, acute inflammation, apoptosis response and MLD. However, the up-regulated mRNA levels of bile acid synthesis, inflammatory cytokines and energy metabolism-related genes in the acute metabolic disorder stage indicated that the largemouth bass was still in a state of ‘self-repair’. After long-term feeding, largemouth bass owned certain adaptability to starch depending on whether metabolic parameters and liver health were returned to homoeostasis. Otherwise, HS-activated apoptosis pathway ultimately led to liver fibrosis and necrosis. Therefore, the largemouth bass could well utilise the HS feed (13·56 %) regardless of poor growth performance. Further studies should pay attention to the molecular mechanism on regulating metabolism disorder under HS diet stress, which was more beneficial to clarify adaptive mechanisms of HS in carnivorous fish.
Acknowledgements
This study was supported by the National Natural Science Foundation of China (32172981 and 31902382), National Key R&D Program of China (2019YFD0900200 and 2018YFD0900400), The Agricultural Science and Technology Innovation Program of CAAS, China (CAAS-ASTIP-2017-FRI-08) and China Postdoctoral Science Foundation (2021M703544).
P. C.: Conceptualisation, Software, Formal analysis and Writing – Original Draft. X. G.: Data Curation. X. W.: Resources. Y. Z.: Supervision and Project administration. M. X.: Writing – Review, Supervision and Funding acquisition. X. L.: Writing – Review. All authors have read and approved the final manuscript. The authors declare that there is no conflict of interest.
We declare that we have no financial and personal relationships with other people or organisations that can inappropriately influence our work, and there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the content of this paper.
