The present study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of Affiliated Hospital of Nanjing University Medical School (Nanjing, China). All experimental protocols and animals were performed in accordance with National Institutes of Health and approved by the Committee on the Ethics of Animal Experiments Defence Research of Affiliated Hospital of Nanjing University Medical School (Nanjing, China) (23). All surgery and sacrifice were performed in a manner to minimize animal suffering.

A total of 120 female Wistar rats (6–8 weeks old; 190±20 g) were purchased from Slack Experimental Animals Co., Ltd. (Shanghai, China). All animals were housed under pathogen-free conditions (license no. SCXK-2008-014). All rats were raised on a 12-h light/dark cycle. The animals had free access to water and the temperature was 20±2°C with a humidity of 50±5%. All rats were provided with a high-fat diet and intraperitoneal injections of vitamin D3 (160,000 U/kg) once a month for a total of five months, following a 7-day acclimatization period. The rats only received the high-fat diet in the next 2 months. The rats were randomly divided into three groups, in which animals received intravenous treatment with PBS, FGF-21 (6 mg/kg/d) or atorvastatin (4.8 mg/kg/d), respectively, once every 3 days. The rats were starved and sacrificed on day 30.

On day 30, the rats were sacrificed by intraperitoneal injection of 5% urethane (1,000 mg/kg). The abdominal aorta, full-length aorta and peripheral blood were obtained in the experimental rats. The tissues and serum samples were collected and stored to further analysis. Serum samples were obtained by centrifugation at 1,500 × g and 4°C for 15 min. The serum levels of low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride and total cholesterol were measured according to the methods described in a previous study (24). The tissues were homogenized using electrically-driven tissue homogenizer (ABI, Grand Island, NY, USA) and prepared for reverse transcription-quantitative polymerase chain reaction (RT-qPCR), immunofluorescence and western blot analyses.

Total RNA was obtained from on endothelial cells of the experimental rats using an RNAeasy Mini kit (Qiagen, Inc., Gaithersburg, MD, USA). The expression levels of FGF-21, chemokine (C-C motif) ligand (Ccl)2, Ccl5, intercellular adhesion molecule 1 (Icam1) and tumor necrosis factor (TNF)α in the cells were measured using β-actin as an endogenous control (25) (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The forward and reverse primers were synthesized by Invitrogen; Thermo Fisher Scientific, Inc. PCR amplification was preliminary denaturation at 94°C for 2 min, followed by 45 cycles of 95°C for 30 sec, annealing temperature reduced to 58°C for 30 sec, and 72°C for 10 min using a volume of 20 µl containing 50 ng of genomic DNA, 200 µM dNTP, 2.5 units of Taq DNA polymerase and 200 µM primers (Table I). The relative changes in mRNA expression were calculated using the 2^−ΔΔCq method (25). The results are expressed as the n-fold change compared with the control.

The endothelial cells in the experimental rats were homogenized in lysate buffer containing protease-inhibitor and were centrifuged at 6,000 × g at 4°C for 10 min. The supernatant of the mixture was used for analysis of the proteins of interest. For the detection of purpose protein, transmembrane proteins were extracted using a Transmembrane Protein Extraction kit (Qiagen, Inc.) according to the manufacturer's protocol. SDS assays were performed as previously described (26). For western blot analysis, primary antibodies: p65 (ab16502; 1:1,000; Abcam, Cambridge, UK), NF-κB (ab97726; 1:1,000; Abcam) were added following blocking in 5% skimmed milk for 1 h at 37°C. This was followed by incubation with horseradish peroxidase-conjugated anti-rabbit IgG (1721019; 1:5,000; Bio-Rad Laboratories, Inc., Hercules, CA, USA) for 24 h at 4°C. The results were visualized by using a chemiluminescence detection system (1705060; Bio-Rad Laboratories, Inc.).

Tissues were fixed for 48 h, dehydrated through a graded series of ethanol, embedded in paraffin wax, and cut into 4 µm tissue sections. The sections were stained with hematoxylin and eosin (H&E) and observed under a light microscope (Olympus BX51; Olympus Corporation, Tokyo, Japan).

The endothelial cells in the experimental rats were incubated with primary antibodies, followed by appropriate secondary antibodies: pro-inflammatory M1 (ab110273; 1:2,000; Abcam) and anti-inflammatory M2 (ab110274; 1:2,000; Abcam) were added as targets for primary antibodies for the immobilized cells for 12 h at 4°C. The horseradish peroxidase-conjugated anti-rabbit IgG (1721019; 1:5,000; Bio-Rad Laboratories, Inc.) was incubated with the endothelial cells following washing with PBS for 2 h at 37°C. In addition, the localization and levels of NF-κB were analyzed using fluorescein isothiocyanate secondary Ab and nuclear counterstaining (DAPI) using a fluorescence microscope (Leica DMi8; Leica Microsystems GmbH, Wetzlar, Germany) (27).

Intracellular superoxide in arterial tissues was analyzed using fluorescence microscopy combined with expressed as number of DHE-positive cells (DAPI staining) and superoxide-sensitive fluorescent dye dihydroethidium (2 mmol/l, DHE; Invitrogen; Thermo Fisher Scientific, Inc.). The activity of NADPH-dependent oxidase in the endothelial cell homogenates was measured using chemiluminescence by NADPH (100 mmol/l) and lucigenin (5 mmol/l).

Data are presented as the mean ± standard deviation. Statistical tests for data analysis included Fisher's exact test, log-rank test, χ² test, and Student's two-tailed t-test. Multivariate statistical analysis was performed using a Cox regression model. Statistical analyses were performed using SPSS version 19.0 (IBM SPSS, Armonk, NY, USA). P<0.05 was considered to indicate a statistically significant difference.

In order to examine the role of FGF-21 on atherosclerosis, the present study first investigated the expression level of FGF-21 in the peripheral blood of atherosclerosis rats, with healthy rats as a control. The results, as presented in Fig. 1A, showed that the expression levels of FGF-21 were downregulated in the rats with atherosclerosis, compared with the healthy rats. However, FGF-21 therapy improved recovery to normal levels of FGF-21. The present study also examined the blood glucose and blood lipids levels to determine the therapeutic effects of FGF-21 on atherosclerosis. As presented in Fig. 1B and C, blood glucose and blood lipids were increased in rats with atherosclerosis, compared with the rats administered with PBS or atorvastatin. The therapeutic effects of FGF-21 on body weight on the rats with atherosclerosis were also examined. The data indicated that FGF-21 inhibited the increase in body weight, compared with PBS or atorvastatin (Fig. 1D). Collectively, these data suggested that FGF-21 was downregulated in atherosclerosis, and that the restoration of FGF-21 had beneficial effects on blood glucose, blood lipids and body weight in the rats with atherosclerosis.

To analyze the effect of FGF-21 on the endothelial cells in the experimental rat, H&E staining was used to observe alteration in the endothelial cells. It was found that FGF-21 treatment markedly altered the ultrastructure of the endothelial cells in the aorta in the rats with atherosclerosis, compared with the rats administered with PBS or atorvastatin (Fig. 2A). In addition, the formation of neointima was analyzed in the carotid arteries of rats with atherosclerosis. The results revealed that FGF-21 significantly improved neointimal formation, and that the neointimal formation was higher, compared with that in either the PBS or atorvastatin-treated groups. (Fig. 2B and C). The present study also examined the effects of FGF-21 on endothelial-dependent relaxation in the experimental rats to confirm the improvement of vessel function. As shown in Fig. 2D, the results revealed that endothelial-dependent relaxation was increased in the injured carotid arteries of the rats with atherosclerosis. Taken together, the therapeutic effects of FGF-21 provided an advantage in the treatment of rats with atherosclerosis.

In order to identify the beneficial effects of FGF-21 on blood vessels, histopathologal changes were detected using H&E and immunohistochemical staining in the rat model of atherosclerosis. The data showed that the unit area of atherosclerotic plaques was significantly lower, compared with the model and atorvastatin groups (Fig. 3A). The blood velocity increased following FGF-21 treatment (Fig. 3B) and, as shown in Fig. 3C, adipocytes were decreased following FGF-21 treatment. There were marked pathological changes to the normal functioning in the aortas of the rats with atherosclerosis following FGF-21 treatment, compared with PBS and atorvastatin (Fig. 3D). Taken together, these findings suggested that FGF-21 treatment led to a visible decrease in pathological changes, including the vessel walls and atherosclerotic plaques, in rats with atherosclerosis.

NF-κB is a crucial regulator in inflammatory responses, which controls the expression of several genes involved in atherogenesis (28). In the present study, the associations between FGF-21 and the NF-κB signaling pathway were analyzed in rats with atherosclerosis. It was observed that FGF-21 inhibited the nuclear import of activated NF-κB (p65) in the vascular endothelial cells of the experimental rats (Fig. 4A). The data indicated that FGF-21 inhibited the activation of NF-κB in vascular cells isolated from rats with atherosclerosis (Fig. 4B). In addition, it was found that FGF-21 treatment decreased the expression levels of anti-inflammatory genes involved in the NF-κB signaling pathway, including Ccl2, Ccl5, Icam1 and TNFα, determined using RT-qPCR analysis (Fig. 4C). The potential benefits of FGF-21 were also examined, which revealed marked NF-κB inactivation in the FGF-21-treated rats (Fig. 4D). Taken together, it was concluded that FGF-21 improved atherosclerosis via the NF-κB signaling pathway.

Further confirming the effects of FGF-21 treatment on rats with atherosclerosis, the data obtained showed that lesion size was reduced by FGF-21 treatment, in addition to a significant decrease in the number of macrophages within the atherosclerotic plaques (Fig. 5A). Regression analysis revealed a correlation of macrophage content with lesion area (r=0.724; P=0.0042) and NF-κB staining (r=0.746; P=0.0031) in the experimental model, indicating a higher inflammatory component in larger lesions. It was also observed that pro-inflammatory M1 and anti-inflammatory M2 were increased in the plaques of the FGF-21-treated rats (Fig. 5B). The results of immunofluorescence demonstrated a higher relative collagen and vascular smooth muscle cell (VSMC) content in the aortic lesions of the FGF-21-treated rats, compared with the PBS and atorvastatin groups (Fig. 5C). Furthermore, superoxide generation in the aortic root region of the experimental animals was analyzed using DHE staining. As demonstrated in Fig. 5D, FGF-21 treatment reduced the number of DHE-labeled nuclei in the plaques, suggesting that FGF-21 treatment inhibited oxidative stress through the NF-κB signaling pathway. Taken together, these data suggested that FGF-21 mediated inflammation and oxidative stress through the NF-κB signaling pathway in rats with atherosclerosis, which may be a potential drug candidate for the treatment of atherosclerosis.