All experimental procedures were approved by the Ethics of Animal Experiments Committee of the Tongji University School of Medicine (Shanghai, China; approval no. TJMED-012-006). The present study was conducted according to internationally recognized guidelines on animal welfare, as well as the regulations regarding animal welfare in Shanghai, China, and was conducted in accordance with the guidelines of the Chinese Council on Animal Care.

A total of 28 six-week-old male Sprague-Dawley rats (Center of Experimental Animals, Tongji University School of Medicine, Shanghai, China), weighing 140–160 g, were housed in wire-bottomed cages in 22±1°C with a 12-h light/dark cycle. The rats were raised on a commercial pellet diet (Center of Experimental Animals) consisting of 22% protein, 66% carbohydrates and 12% fat, and were provided with access to tap water ad libitum. The rats were randomly divided into three groups: A normal diet control group (NC; n=10), a catch-up growth APS-treated group (APS-G; n=9) and a catch-up growth model group (CUGFR; n=9). The rats in the NC group were raised on an ad libitum pellet diet for 8 weeks, and the rats in the CUGFR group were subjected to a dietary restriction for 4 weeks (60% of the diet intake of the NC group) following which they were fed with a high fat diet (42% calories from fat), which was provided ad libitum. The rats in the APS-G group were orally treated with 700 mg/kg APS (content of effective components, 69%), purchased from the Beijing Centre Biology Co. Ltd., (Beijing, China), for 8 weeks. The other groups were administrated with saline irrigation at the same dosage. The body weight of each rat was measured every 7 days.

The rats were subjected to fasting 15 h prior to intraperitoneal injection with 20% urethane anesthetic (5 ml/kg; Shanghai Hengyuan Biological Technology Co. Ltd, Shanghai, China). The abdominal cavities of the rats were surgically opened following local disinfection. Blood samples (8–10 ml) were obtained from the abdominal aorta, maintained at room temperature for 10 min prior to centrifugation in an Eppendorf (EP) tube containing EDTA (1 mg/ml) and then centrifuged for 20 min at 2000 ×g. The supernatant was placed into a sterile EP tube and preserved at −20°C. The blood samples were used to determine the serum FGF21 levels using an ELISA kit (BioVendor, Asheville, NC, USA). Liver function, blood lipid levels and glycated hemoglobin (HBA1C) levels were also analyzed. Plasma fatty acid, total cholesterol (TC) and triglyceride levels were determined using a NEFAC kit (Wako Pure Chemical Industries, Ltd., Osaka, Japan), which measures oxidase and peroxidase activity (Trinder reaction). Low-density lipoprotein (LDL) and high-density lipoprotein (HDL) were measured using an immunoassay kit (Daiichi Pure Chemicals Co., Ltd., Tokyo, Japan), alanine aminotransferase (ALT) was determined using an IFCC kit (Roche Diagnostics, Basel, Switzerland). HBA1C levels were determined by immunoturbidimetry (cat. no. D-68305; Roche Diagnostics). The rats were fasted for 15 h prior to intraperitoneal injection of 20% urethane anesthetic (5 ml/kg). The abdominal cavities of the rats were surgically opened by pattern clamps and ophthalmology scissors (Shanghai Kang Chau Medical Instrument Co., Ltd, Shanghai, China) followed by local ethanol disinfection (Sinopharm Co., Ltd., Shanghai, China). Subsequently, blood was drawn from the abdominal aorta and anti-coagulated with EDTA (1 mg/ml). Plasma was obtained by centrifuging the blood at 2000 ×g at 4°C for 20 min and frozen at 20°C for further analyses. The liver, epididymal adipose and perirenal fat tissues were separated and weighed under sterile conditions after the rats were sacrificed by decapitation. The liver tissue samples were immediately dissected, fixed in 2% glutaraldehyde (Sinopharm Co., Ltd) and 10% formalin or frozen in liquid nitrogen until further analysis. The liver index (the ratio of liver weight to total body weight of the rats) and internal body fat ratio (epididymal fat weight plus perirenal fat weight) were calculated.

An OGTT test was conducted as follows: A total of 2 g glucose/100 g body weight was orally administered following overnight fasting for 15 h. Blood samples were collected from the rat tail venous plexus 0, 15, 30, 60 and 120 min after glucose treatment, in order to measure the blood glucose and plasma insulin concentration levels. Blood glucose was measured by a glucometer (Roche Accu-Chek Performa; Roche Diagnostics). The blood was then centrifuged and placed into an EP tube containing 1 mg/ml EDTA (Shanghai Zurui Biological Technology Co., Ltd, Shanghai, China), in order to measure the blood insulin levels using an ELISA kit (EMD Millipore, Billerica, MA, USA). The plasma was then either immediately used for experimentation or frozen at −20°C for further analyses.

The frozen rat liver tissue samples were homogenized three times, for 15 sec, in ice-cold radioimmunoprecipitation assay homogenization buffer (Merck Chemical Technology Co. Ltd, Shanghai, China) containing Protease Inhibitor Cocktail Set III (cat. no. 539134; Invitrogen Life Technologies, Inc., Carlsbad, CA, USA) and Phosphatase Inhibitor Cocktail Set V (cat. no. 524629; Invitrogen Life Technologies, Inc.) using a PRO 200 homogenizer (PRO Scientific, Inc., Oxford, CT, USA). The homogenate was centrifuged at 4°C for 30 min at 12,000 × g and the supernatant was used as the total protein concentration. The total protein concentration then underwent protein quantification according to the Bradford method (Protein Assay kit II; cat no. 5000002; Bio-Rad Laboratories, Inc., Hercules, CA, USA) according to the manufacturer's instructions. Protein assay dye reagent concentrate (Bio-Rad) and BSA (Thermo Fisher Scientific, Waltham, MA, USA) were used as the protein standard. Protein (40 µg) was separated by 4–12% polyacrylamide gels [depending on the protein of interest; NuPAGE 4–12% Bis-Tris Gel 1.0 mmx12 well; MOPS SDS running buffer (20X); both from Invitrogen Life Technologies, Inc.] and transferred onto nitrocellulose membranes (iBlot transfer stacks nitrocellulose, regular; Invitrogen Life Technologies) using a semidry transfer protocol. Precision Plus Protein Dual Color standards (Bio-Rad) and Novex^® Sharp Pre-stained Protein standard (Invitrogen) were used as the protein ladder. Following membrane blocking with Odyssey Blocking Buffer (LI-COR Biotechnology, Lincoln, NE, USA), the membranes were incubated with the appropriate primary antibody: Rabbit anti-NF-κB p65 polyclonal immunoglobulin (Ig)G (1:200; cat. no. sc-372; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), mouse anti-SIRT1 monoclonal antibody (1:2,000; cat. no. ab110304; Abcam, Cambridge, UK), and rabbit anti-human FGF21 polyclonal antibody (0.2–1 µg/ml; cat. no. LS-B5864; 1:500 dilution; LifeSpan BioSciences, Inc., Seattle, WA, USA) antibodies, diluted with fluorescently-labeled secondary antibody blocking buffer [IRDye^® 800CW goat anti-mouse IgG (H+L); 1:10,000; cat. no. 925-32210; or IRDye^® 800CW goat anti-rabbit IgG (H+L); 1:10,000; cat. no. 925-32211; both from LI-COR Biotechnology). The membranes were then washed twice with phosphate-buffered saline (PBS; Decent Biotech, Göttingen, Germany) with 0.1% Tween 20 (Sigma, Santa Clara, CA, USA) to remove all residual antibodies, and scanned using an Odyssey Infrared Imaging system with IR fluorescence scan Odyssey 2.1 software (LI-COR Biotechnology). Quantitative values were obtained from the densitometric measurements of the western blotting bands using Image Quant software (Tool, 1D gel analysis; GE Healthcare Bio-Sciences, Pittsburgh, PA, USA).

The liver mRNA expression levels of PGC-1α and PPARα were assessed by RT-qPCR. The total RNA was extracted using TRIzol reagent (Invitrogen Life Technologies, Inc.; cat. no. 47122). cDNA was reverse-transcribed from 2 µg DNase-treated total RNA using an iScript cDNA Synthesis kit (Bio-Rad Laboratories, Inc.; cat no. 170-8891) according to the manufacturer's instructions. The following primer sequences, obtained from Invitrogen Life Technologies, were used: Forward, 5′-CGCACAACTCAGCAAGTCCTC-3′ and reverse, 5′-CCTTGCTGGCCTCCAAAGTCTC-3′ for rat PGC-1α; and forward, 5′-GTGGCTGCTATAATTTGCTGTG-3′ and reverse, 5′-GAAGGTGTCATCTGGATGGGT-3′ for PPARα. The cDNA (diluted to 1:20) served as a template for the RT-qPCR; 100 µM primer solution was diluted to 5 pmol/µl, and qPCR was performed using SYBR Premix ExTaq (Takara Biotechnology Co., Ltd., Dalian, China). The PCR mixture contained 10.5 µl SYBR Premix ExTaq, 0.5 µl of each primer, 1 µl cDNA and 7.5 µl deionized water. PCR was performed with thermocycling conditions of 95°C for 30 sec, 40 cycles at 95°C for 5 sec, and 60°C for 30 sec in a DNA cycler apparatus (ABI Prism 7900HT; Applied Biosystems, Thermo Fisher Scientific). Genes were quantified using the ΔΔCt method with GAPHD as a housekeeping gene. For each set of reactions, the samples were analyzed in triplicate.

Following sacrifice, the livers of the rats were excised, and the liver wet weights were measured. A section of hepatic tissue (~3×3×10 mm) was removed via a horizontal incision 1 cm away from the edge of the right lobe. The hepatic tissue samples were fixed in 10% formalin, dehydrated with a graded series of ethanol, ethanol uranyl acetate and acetone (all from Sinopharm Co., Ltd.), and stained with hematoxylin and eosin (H&E; n=3 rats/group). Observation and image capture were conducted using a CKX31 light microscope (Olympus Corporation, Tokyo, Japan).

Three 1-mm³ liver sections were fixed with 2.5% glutaraldehyde and stored at 4°C. The samples were subsequently analyzed at the Electron Microscopy Laboratory at the Shanghai University of Traditional Chinese Medicine (Shanghai, China). Following fixing, samples were rinsed with 0.1 M phosphate buffer and treated with 1% OsO₄ solution (Alfa Aesar Chemical Co., Ltd., Shanghai, China) for 3 h, dehydrated with a graded series of ethanol (30, 50, 80 and 90%) and propylene oxide, and embedded in epoxy resin (Shanghai Resin Factory Co., Ltd, Shanghai, China). Semi-thin (thickness, 500 nm) hepatic sections were stained with 1% toluidine blue (Sinopharm Co., Ltd) for histopathology. Suitable areas of ultrathin (thickness, 80–90 nm) hepatic sections were selected for ultrastructural analysis. These were sectioned using a diamond knife (diamond knife(cat no. 706602; Leica Microsystems, Wetzlar, Germany), mounted on a copper grid (Shanghai Chemical Experiment Audio Supplies Company Ltd, Shanghai, China), and stained with uranyl acetate and lead citrate. The sections were analyzed using a Tecnai-12 Bio T min TEM (Philips, Amsterdam, Netherlands), and observation and image capture were performed at an acceleration voltage of 80 kV.

Statistical differences between the groups were evaluated using Student' unpaired t-test. The homeostasis model assessment of insulin resistance (HOMA-IR) was used, where: HOMA-IR = [fasting plasma glucose (mmol/l) × fasting plasma insulin (mU/l)]/22.5. The data were presented as the mean ± standard deviation. Data analyses were performed using SPSS 18.0 (SPSS, Inc., Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference.

The growth curve obtained from weighing the rats weekly demonstrated that their growth was arrested during the period of 60% food restriction, and rapidly increased following re-feeding on a high fat diet for 1 week; however, the rat body weights of the CUGFR group increased less compared with the NC group. The growth curve obtained from the rat weights demonstrated that the APS-G group exhibited a significant trend towards weight loss, as compared with the CUGFR group. Conversely, the rats in the CUGFR group significantly gained weight compared to the rats in the NC group (Fig. 1).

Prior to treatment with APS, glucose tolerance curves demonstrated marginally lower (although not statistically significant) serum glucose levels in the CUGFR group rats at 0, 15 and 30 min compared with the NC group. At 60 min, following glucose gavage, blood glucose levels were significantly higher than the NC group; these glucose levels failed to return to pre-meal state (Fig. 2A). The insulin curve demonstrated significantly lower insulin levels at 15 min in the CUGFR group compared with the NC group (P<0.05) and that insulin increase was delayed (Fig. 3A). The insulin levels of the APS-G group significantly decreased at 15 min, and the area under the curve of glucose increased, although no statistically significant difference was observed (Fig. 2B, and Fig. 3A and B).

Following treatment with APS, no statistically significant difference was observed between the blood glucose levels of all three groups, and the blood glucose levels of the APS-G rats did not return to their pre-meal state (Fig. 2C and D). The insulin levels of the CUGFR group significantly increased, when compared with the NC group (P<0.01), and treatment with APS moderately reduced this increase, although the results were not statistically significant (Fig. 3D). After 15 min the levels of insulin were significantly lower in the APS-G group, when compared with the CUGFR group, and no significant decline in these insulin levels occurred throughout the experiment (Fig. 3C). The HOMA-IR and fasting insulin index were significantly higher in the CUGFR group, as compared with the NC group, and following treatment with APS, the two decreased significantly (Table I).

During early catch-up growth, the levels of thyroglobulin (TG), TC, LDL and HBA1C were significantly higher in the CUGFR group as compared with those in the NC group, while levels of HDL were significantly lower than those in the NC group. The levels of free fatty acid (FFA) and ALT in the GUGFR group were not significantly different from those in the NC group. At the late stage of catch-up growth, the levels of LDL, FFA and ALT in the CUGFR group were markedly increased compared with those in the NC group (Table II). No statistically significant increase was observed in the levels of serum lipid and HBA1C following treatment with APS (Table I).

The liver index and visceral fat content ratio of the rats were significantly higher (P<0.01 and P<0.05, respectively) in the CUGFR group, as compared with the NC group. Following 8 weeks of APS treatment, liver weight in the APS-G group reduced significantly compared with the CUGFR group, however the decrease in the liver index was not significant. Although the visceral fat content ratio of the CUGFR group was significantly higher than that of the NC group, treatment with APS did not significantly affect the visceral fat content (Table III).

Liver H&E staining demonstrated large quantities of lipid droplet deposition in the liver tissue samples of the CUGFR group, as compared with the NC and APS-G groups. Following treatment with APS, the lipid droplet deposition of the APS-G group decreased, as compared with the CUGFR group, and the tissue samples exhibited lipid droplet fusion (Fig. 4).

TEM examination of the liver samples (Fig. 5) demonstrated that damaged mitochondria and lipid droplets were present in the hepatocytes of the CUGFR (Fig. 5C and D) and APS-G groups (Fig. 5E and F). The groups also exhibited clusters of damaged mitochondria, multi lamellar bodies, and large lipid droplets joined to form large lipid droplets and surrounded by autophagosomal membranes. In the CUGFR group, numerous lipid droplets were present, including fused lipid droplets, and the number of mitochondria and endoplasmic reticula was decreased in the cells. In addition, the cellular structure was severely impaired. Furthermore, the nuclear membranes were not continuous, exhibiting visible holes.

Following treatment with APS, the number of lipid droplets decreased in the APS-G group (Fig. 5E and F), however numerous impaired mitochondria and endoplasmic reticula remained visible, and the liver cells exhibited expanded endoplasmic reticula and mitochondrial coagulation-associated degeneration. The cell membranes were thick, but continuous (Fig. 5).

Following treatment with APS, the serum FGF21 levels markedly decreased in the CUGFR group, whereas the serum levels of FGF21 in the APS-G group significantly increased compared with the CUGFR group (Fig. 6; P<0.05).

The expression levels of PGC-1α markedly increased in the CUGFR group when compared with the NC group, and were marginally increased by treatment with APS (Fig. 7A). The expression levels of PPARα in the CUGFR group were significantly lower as compared with the NC group, and these expression levels increased following treatment with APS (Fig. 7B; P<0.05). In the CUGFR group, the protein expression levels of FGF21 and SIRT1 were significantly lower than the NC group (Fig. 7C and D P<0.05), whereas the expression levels of NF-κB were significantly increased. In the APS-G group, the previously low expression levels of FGF21 and SIRT1 were increased (with the difference in SIRT1 being significant compared with the CUGFR group; P<0.05). APS treatment appeared to exert no significant effect on the activity levels of NF-κB, where the expression levels of NF-κB had previously been significantly increased in the CUGFR group, as compared with the NC group (Fig. 7E and F).