Forty-six 8-week-old male Sprague-Dawley (SD) rats obtained from the Laboratory Animal Center of Xi'an Jiaotong University (Xi'an, China) were used in the present study. All the rats were housed under specific pathogen-free conditions with free access to water. All the experiments were carried out in adherence to a protocol approved by the Institutional Animal Care and Use Committee of Xi'an Jiaotong University. Of the 46 rats, some were fed a standard chow diet (Laboratory Animal Center of Xi'an Jiaotong University, Xi'an, China) and were assigned to the normal control group (n=10), while the others were used to establish a model of NAFLD. For the induction of NAFLD, the rats were fed a high-fat diet (HFD; 10% lard, 2% cholesterol, 5% saccharose, 0.5% hog bile extract and 82.5% standard chow diet) for 12 weeks. The 36 rats fed the HFD were assigned to the HFD control group (n=12), the GW501516 treatment group (n=12), or the pioglitazone treatment group (n=12). Therapeutic intervention with GW501516 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) or pioglitazone (Santa Cruz Biotechnology) was initiated after 8 weeks on the HFD. GW501516 (10 mg/kg body weight/day, in water) and pioglitazone (10 mg/kg body weight/day, in water) were administered orally once a day to the rats in the respective groups from the 8th week until the 12th week. The rats in the HFD control group received the same amount of normal saline. Blood samples were collected from the tail vein and stored at −70°C until biochemical analysis. At the end of each experiment, the rats were sacrificed by an intravenous administration of pentobarbital (overdose). The body weight, liver wet weight and body length of the rats were then measured. The hepatosomatic index was calculated as follows: hepatosomatic index = liver wet weight (g)/body weight (g). The body mass index was calculated as follows: body mass index = body weight (kg)/body length (m)². The livers were cut into a number of sections and stored at −80°C until tissue examination.

Serum fasting blood glucose (FBG), triglyceride (TG), total cholesterol (TC), low-density lipoprotein (LDL), high-density lipoprotein (HDL), alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transpeptidase (GGT) and alkaline phosphatase (ALP) levels were measured using an automatic biochemical analyzer. The serum insulin levels were measured using an enzyme-linked immunosorbent assay (ELISA) kit (EIA2048; DRG International, Inc., Springfield, NJ, USA) according to the manufacturer's instructions. Insulin resistance was calculated using the homeostasis model assessment of insulin ressistance (HOMA-IR) as follows: fasting insulin (mIU/l) x fasting blood glucose (mmol/l)/22.5, as previously described (26). The serum insulin-like growth factor-1 (IGF-1) levels were measured using an ELISA kit (IB39431) purchased from IBL (Minneapolis, MN, USA), according to the manufacturer's instructions.

Histological analyses were performed according to a the methods described in a previous study (27). Briefly, the liver tissues were fixed in 10% buffered formaldehyde and then embedded in paraffin. The paraffin-embedded tissues were cut into 5-µm-thick sections and were then stained with hematoxylin and eosin (H&E). To visualize the accumulation of fat droplets, sections of frozen liver tissue were stained with Oil Red O, and hematoxylin was introduced to labelmark the cell nuclei.

The 5-µm-thick paraffin-embedded sections of liver tissue were deparaffinized by changing the xylene twice, and this procedure was followed by rehydration in an ethanol gradient. Following antigen retrieval and non-specific binding blocking, the slides were incubated with primary antibodies diluted at 1:800 for 60 min at 37°C. After washing 3 times, the slides were then incubated with HRP-conjugated rabbit anti-rat IgG (Cat. no. ab6734; Abcam, Cambridge, UK) diluted at 1:500 for 60 min at 37°C. Color development was performed using a DAB Horseradish Peroxidase Color Development kit (Beyotime, Nantong, China). Hematoxylin was used for nuclear compartment visualization. Rabbit anti-rat sterol regulatory element binding protein-1c (SREBP-1c) and glucose transporter type 2 (GLUT-2) antibodies were purchased from Santa Cruz Biotechnology. The results were quantified using a color density assay using iVision software (version 4.0.9; BioVision Technologies, Exton, PA, USA).

Total RNA from the rat livers was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. cDNA was synthesized from 5 µg of the total RNA using a PrimeScript^® RT reagent kit (Takara, Dalian, China) with oligo(dT) primers as per the manufacturer's instructions. The quantitative PCR (qPCR) reactions were performed on a Bio-Rad iQ5 real-time thermal cycler using a SYBR^® Premix Ex Taq™ II kit (Takara) in a 25-µl reaction system. All reactions were performed in triplicate, and the results are represented as relative mRNA expression data, which were calculated using the 2^−ΔΔCT method, as previously described (28). The sequences of the primers used were as follows: SREBP-1c forward, 5′-CGC TAC CGT TCC TCT ATC AAT GAC-3′ and reverse, 5′-AGT TTC TGG TTG CTG TGC TGT AAG-3′; GLUT-2 forward, 5′-TCC GCT TGC TCC TCC TCC TAC-3′ and reverse, 5′-ACA CCGA TGT CAT ATC CGA ACT GG-3′; diacylglycerol acyltransferase (DGAT) forward, 5′-TCC GCC TCT GGG CAT TC-3′ and reverse, 5′-GAA TCG GCC CAC AAT CCA-3′; CPT-1 forward, 5′-TAT GTG AGG ATG CTG CTT CC-3′ and reverse, 5′-CTC GGA GAG CTA AGC TTG TC-3′; ACOX1 forward, 5′-GTT GAT CAC GCA CAT CTT GGA-3′ and reverse, 5′-TCG TTC AGA ATC AAG TTC TCA ATT TC-3′; ACOX3 forward, 5′-TGG AGA AGG AAC GAG AAC TGA AC-3′ and reverse, 5′-ACA TGT GGA GGA CAC ATT TGT TG-3′; and β-actin forward, 5′-CAA CTT GAT GTA TGA AGG CTT TGG T-3′ and reverse, 5′-ACT TTT ATT GGT CTC AAG TCA GTG TAC AG-3′. All the primers were synthesized by BGI Tech (Shenzhen, China).

Data are expressed as the means ± SEM from a minimum of 3 experiments. Differences between groups were analyzed using the Student's t-test. A value of P<0.05 was considered to indicate a statistically significant difference.

Pioglitazone, a prescription drug of the class TZD, which is used in the treatment of diabetes, was introduced in the present study as a positive control. After 12 weeks of being fed the HFD, the hepatosomatic index and body mass index of the rats were both significantly elevated (P<0.05) compared to the rats in the normal control group (Fig. 1). However, 2 weeks of treatment with GW501516 or pioglitazone significantly reduced the hepatosomatic index (P<0.05) compared to the HFD group, although the hepatosomatic index was still significantly higher than that of the normal control group (P<0.05; Fig. 1A). After 4 weeks of treatment with GW501516 or pioglitazone, the hepatosomatic index was almost reduced to the level of the normal control group (Fig. 1A). Treatment with pioglitazone did not reduce the body mass index (Fig. 1B). However, a significant decrease in body mass index was observed in the rats that were treated with GW501516 for 4 weeks (Fig. 1B).

After 4 weeks of treatment with normal saline, GW501516 or pioglitazone, the livers of the rats in all the groups were collected and analyzed using Oil Red O (Fig. 2A) and H&E (Fig. 2B) staining. The histological examination of the livers of the rats from the normal control group exhibited a normal structure and architecture. No obvious lipid droplets or lobular inflammatory cell infiltration were observed (Fig. 2). However, typical vesicular steatosis, hepatocellular ballooning, and lipid droplets were observed in the HFD group (Fig. 2). After 4 weeks of treatment with GW501516 or pioglitazone, these pathological changes were significantly alleviated. In addition, it was clear that GW501516 had a more potent effect than pioglitazone, as lipid droplets, hepatocellular ballooning and inflammatory cell infiltration were scarcely observed in the group treated with GW501516 (Fig. 2).

In order to investigate whether treatment with GW501516 affects glucose metabolism, serum was collected from the rats at the end of each treatment for the determination of the levels of fasting blood glucose, fasting insulin and IGF-1. The HOMA-IR index was calculated based on the levels of fasting blood glucose and fasting insulin. As expected, a significant increase in the serum levels of fasting blood glucose and fasting insulin, as well as in the HOMA-IR index was observed in the rats in the HFD group compared to the rats of the normal control group. In addition, a significant decrease in the levels of IGF-1 was observed in the HFD group compared to the normal control group (Fig. 3). However, 4 weeks of treatment with pioglitazone significantly decreased the levels of fasting blood glucose and fasting insulin, and the HOMA-IR index, and increased the levels of IGF-1 compared to the HFD group (Fig. 3). Nonetheless, these levels and indexes still differed significantly from those of the normal control group (Fig. 3). However, 4 weeks of treatment with GW501516 almost restored these indexes to normal levels (Fig. 3). This suggested that glucose metabolism is regulated by GW501516.

To determine the effect of GW501516 on lipid metabolism, the levels of TG, TC, HDL and LDL in the serum were measured. In contrast to the normal control group, the levels of TG, TC and LDL were all significantly increased, and the levels of HDL were significantly decreased in the HFD group (Fig. 4). However, 4 weeks of treatment with pioglitazone significantly decreased the levels of TG, TC and LDL, and increased the levels of HDL in comparison to the HFD group (Fig. 4). Treatment with GW501516 almost restored the TG, TC, LDL and HDL levels to normal levels (Fig. 4). This suggested that lipid metabolism may be regulated by GW501516.

To evaluate liver function, the serum levels of ALT, AST, GGT and ALP were determined. As shown in Fig. 5, the levels of ALT, AST, GGT and ALP were all significantly elevated in the HFD group compared to those of the normal control group, and these were all markedly reduced after 4 weeks of treatment with GW501516 compared to the HFD group (Fig. 5). These data indicate that GW501516 effectively alleviates NAFLD and that it is more effective than pioglitazone, since treatment with pioglitazone did not reduce the ALT, AST, GGT or ALP levels to the same extent.

The above-mentioned findings indicate that treatment with GW501516 regulates glucose and lipid metabolism. To elucidate the underlying mechanisms responsible for the effects of GW501516 on glucose and lipid metabolism, the expression levels of SREBP-1c, a key lipogenic transcription factor involved in lipogenesis (29,30), and GLUT-2 were determined. As expected, the mRNA levels of SREBP-1c were significantly upregulated and those of GLUT-2 were significantly downregulated in the livers of the rats in the HFD group compared to the normal control group (Fig. 6A and B). However, these levels were both restored to almost normal levels after 4weeks of treatment with GW501516 (Fig. 6A and B). A similar trend was observed in relation to the protein expression of SREBP-1c and GLUT-2, as was determined by immunohistochemical staining (Fig. 6C and D). These data suggest that GW501516 modulates glucose and lipid metabolism by regulating SREBP-1c and GLUT-2 expression.

Previous studies have demonstrated that the expression of genes related to lipid metabolism is modulated by PPARs (31–35), and it has been shown that the expression of SREBP-1c, a key lipogenic transcription factor, is regulated by GW501516 (36). As DGAT, CPT-1, ACOX1 and ACOX3 are important rate-limiting enzymes which are related to lipid metabolism, we determined the transcriptional levels of these genes in order to evaluate the effects of GW501516 on lipid metabolism. As shown in Fig. 7, the transcriptional levels of DGAT1, CPT-l and ACOX1 were all significantly upregulated; however, those of ACOX3 were only slightly altered by the HFD. Four weeks of treatment with GW501516 did not greatly alter the transcription levels of DGAT1, in comparison to those of the HFD group. This suggests that triglyceride synthesis is not significantly affected by GW501516, as DGAT1 is the rate-limiting enzyme for the synthesis of TGs from diacylg-lycerols (37). However, the transcriptional levels of CPT-l, ACOX1 and ACOX3 were significantly upregulated following tretment with GW501516 compared to the levels of the HFD group (Fig. 7). These data suggest that the β-oxidation of fatty acids in the mitochondria and peroxisomes is improved by treatment with GW501516, as CPT-l is the rate-limiting enzyme for β-oxidation in the mitochondria (38), and ACOX1/ACOX3 are the rate-limiting enzymes for β-oxidation in peroxisomes (39).