A total of 24 male New Zealand pure breed white rabbits (age, 10 weeks; weight 1.8±0.2 kg) were provided by Qingdao Shandong Kangda group. Animals were housed at 22°C, 50% relative humidity, 0.03% CO₂ and 12 h light/dark cycle. The rabbits were fed standard diet to acclimate for 1 week prior to being randomly divided into three groups (n=8): Group A was the control group, which was administered an ordinary diet; group B was the high fat group; and group C was the ATRA (provided by the Institute of Pharmacology, Anhui Medical University, Hefei, China) intervention group. Groups B and C were administered high fat feed (94% ordinary feed, 1% cholesterol and 5% lard), and all feed was purchased from the Laboratory Animal Center of Anhui Medical University (Hefei, China). The ATRA group received gavage (10 mg/kg/day) from week 2 (18). At the end of week 12, the animals were sacrificed to obtain blood and arterial specimens for fixation and analysis. All procedures were approved by the Internal Animal Care and Use Committee of Anhui Medical University.

For anesthesia, 3% pentobarbital sodium was administered through the auricular vein and the carotid arterial blood collected. After standing for 2 h, the blood was centrifuged for 10 min at 1,006 × g, and the upper serum was removed and stored at −80°C. The animals were dissected through the abdomen, the blood was drained and the aorta was separated as soon as possible, sections of which were embedded with optimal cutting temperature compound (OCT) following the above procedures, other sections were frozen with liquid nitrogen and stored at −80°C until use, prepared for oil red O staining or were fixed with 4% paraformaldehyde at room temperature for 24 h followed by embedding in paraffin and sectioning at 4 µm.

Triglyceride (TG; cat. no. 0949-2008) and total cholesterol (TC; cat. no. 1568-2003) assay kits were provided by Zhejiang Dong'ou Diagnostic Products Co., Ltd. (Wenzhou, China). The plasma radioimmunoassay ET-1 kit (cat. no. D11PJA) was purchased from the Beijing North Institute of Biological Technology (Beijing, China). Low-density lipoprotein (LDL; cat. no. A113-2), high-density lipoprotein (HDL; cat. no. A112-2) assay kits, endothelial nitric oxide synthase assay kit (eNOS; cat. no. H195) and a nitrate reductase NO kit (cat. no. A012) were obtained from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). Sulfo-NHS-LC-biotin was purchased from Pierce (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Anti-β-actin (cat. no. BM0627) antibody was obtained from Wuhan Boster Biological Technology, Ltd. (Wuhan, China), and ET-1 (cat. no. sc-517436), eNOS (cat. no. sc-136977) and horseradish peroxidase-conjugated anti-mouse IgG (cat. no. sc-2005) and anti-rabbit IgG (cat. no. sc-2004) antibodies were provided by Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). Anti-p-eNOS (Ser-1177; cat. no. 9571s) antibody was obtained from Cell Signaling Technology, Inc. (Danvers, MA, USA). Stripping buffer was purchased from Beyotime Institute of Biotechnology (Haimen, China).

An enzyme coupling colorimetric method was utilized to detect the serum TC and TG, while a precipitation method was used to detect the HDL and LDL in model rabbits. All were performed in strict accordance with the manufacturer's instructions for the assay and ELISA kits.

The paraffin sections from each group were dewaxed and hydrated in a graded series and stained at room temperature with hematoxylin for 10 min and eosin for 5 min, which was followed by mounting in neutral resin. Sections were observed under a light microscope and images were captured. For oil red O staining, the harvested whole specimen of arterial wall was placed in the stock solution (oil red powder 0.5 g dissolved in 100 ml IPA), rinsed with distilled water at a ratio of 6:4 respectively, placed in 40% formaldehyde solution and observed under a light microscope to capture images.

A total of 1 ml PBS was added to the frozen aorta, which was cut it into pieces on ice. Subsequently, protein extraction buffer (0.15 mol/l NaCl, 1.5 mmol/l MgCl₂, 10 mmol/l KCl, 10 µg/ml aprotinin, 0.5 µg/ml leupeptin, 3 mmol/l phenylmethylsulfonyl fluoride, 3 mmol/l dithiothreitol, 1 mmol/l Na₃VO₄, 10 mmol/l hydroxymethyl aminomethane, 5 mmol/l ethylenediaminetetraacetic acid, pH 7.5) was added to prepare the homogenate, followed by freeze-thawing three times at −80°C and room temperature, and centrifugation for 15 min at 1,006 × g and 4°C. The supernatant was collected and the Lowry et al (19) was applied to quantitatively detect the protein concentration. The protein samples (15 µl/lane) were subjected to 10% SDS/PAGE and electrophoretically transferred onto a polyvinylidene fluoride membrane. Non-specific binding was blocked by 5% skimmed milk at room temperature for 2 h. Western blotting was performed using β-actin (1:2,000 dilution), ET-1 (1:1,000 dilution), eNOS (1:2,000 dilution) and p-eNOS (Ser-1177; 1:400 dilution) primary antibodies. All primary antibodies were incubated at 4°C overnight and the secondary antibodies (1:8,000 dilution) were incubated with membranes at room temperature for 1 h. Enhanced Chemiluminescence reagent (Beyotime Institute of Biotechnology) was used for visualization. All the values were normalized to the levels of β-actin. ImageJ 2 software (National Institutes of Health, Bethesda, MD, USA) was utilized in the quantitative analysis of western blot images.

The whole blood (5 ml, 3.8% sodium citrate anticoagulation) was centrifuged for 10 min at 1,006 × g, and the upper plasma was removed and stored at −80°C and assayed within 2 weeks. ET-1 was measured by using an ET-1 radioimmunoassay kit. Assay kits were used to detect the contents of NO and the activity of eNOS. All the testing steps were performed in accordance with the manufacturers' instructions.

The permeability assay using the surface biotinylation technique was performed for the aorta intima as described by Zhu et al (20) with certain modifications. The arterial tissue was filled with a freshly made Sulfo-NHS-LC-biotin solution in Hanks Balance Salt Solution (HBS; 137 mM NaCl, 0.4 mM MgSO₄ x 7H₂O, 5.3 mM KCl, 0.5 mM MgCl₂ × 6H₂O, 1 mM CaCl₂, 0.3 mM Na₂HPO₄ x 7H₂O, 0.45 mM KH₂PO₄, 5.5 mM glucose) containing 1 mM CaCl₂ and 1 mM MgCl₂) at 1 mg/ml for 30 min at room temperature. The aortas were rinsed with PBS three times, 5 min each time embedded in OCT and cryosectioned. Frozen sections (6 µm) were incubated with 5% skimmed milk powder (pH 7.4) at 4°C overnight. Subsequently, the milk was removed and 100 µl Rhodamine600 Avidin D (XRITC-avidin, 1:200, cat. no. A-2005; Vector Laboratories, Inc., Burlingame, CA, USA) was added, followed by incubation at 4°C for 2 h. Rhodamine was then removed, and the sections were rinsed with PBS for 10 min, PBS containing 0.05% Tween-20 for 10 min and with pure PBS twice. The sections were dried at room temperature, mounted and sealed. DAPI staining for 10 min at room temperature was used as a histological control. A fluorescence microscope (Nikon E800; Nikon Corporation, Tokyo, Japan) was used for observation and CCD-SPOT digital camera series (Diagnostic Instruments, Inc., Sterling Heights, MI, USA) was used to capture images. ImageJ 2 software (National Institutes of Health) was used to quantitatively analyze the fluorescence intensity of images.

Data are presented as the mean ± standard deviation, and SPSS statistical software version 16.0 (SPSS, Inc., Chicago, IL, USA) was used for statistical analysis. Comparisons between each group were performed using one-way analysis of variance and the Student-Newman-Keuls method. P<0.05 was considered to indicate a statistically significant difference.

Following hematoxylin and eosin and oil red O staining, the artery intima of the control group was smooth and continuous without intimal thickening, the smooth muscle cells arranged in neat rows and the elastic fibers only slightly disordered. Compared with the control group, the intima of the artery wall in the high fat group was visibly thickened with numerous elastic plate fractures and a large number of plaques; inflammatory and foam cells were identified in the plaques. In the ATRA group, a small amount of plaque formation was observed in the arterial wall, and compared with the high fat group, the foam cells and inflammatory cells were reduced (Fig. 1).

Due to the close association between lipid levels and AS, the serum levels of TC, TG, HDL and LDL in each group were examined (Fig. 2). Compared with the control group, the levels of TC, TG, LDL and HDL were significantly higher in the high fat group (P<0.05). By contrast, the levels of TC, TG and LDL in the ATRA group were lower compared with the high fat group, while HDL levels were higher in the ATRA group (P<0.05).

To determine the effect of ATRA treatment on endothelial permeability, the transport of NHS-LC-biotin across the aortic intima into the media was assessed. The NHS-LC-Biotin concentration profiles were obtained as a function of the radial distance through the aortic wall media layer. Only the endothelial surface of the aorta intima was biotinylated in rabbits fed a normal diet, indicating no paracellular leakage of the NHS-LC-biotin (Fig. 3A), and NHS-LC-biotin leakage into the aortic intima layers was increased in the high fat group (Fig. 3B). However, aortic endothelial permeability was clearly attenuated in the ATRA group compared with the high fat group (Fig. 3C). The permeability in each group was quantified (Fig. 3D).

Western blotting results demonstrated that the expression of ET-1 in rabbit arterial tissue in the high fat group was increased compared with the control group, however, its expression was decreased in the ATRA group. ATRA significantly increased the eNOS phosphorylation at Ser-1177 (p-eNOS); the expression of p-eNOS in the high fat group was lower compared with the control group and increased in the rabbits fed with ATRA. ATRA did not affect the expression of eNOS in each group (Fig. 4).

The results demonstrated that the content of NO and eNOS activity in the high fat group was significantly lower compared with the control group (P<0.05), while the level of NO and eNOS activity were increased following ATRA treatment (P<0.05; Fig. 5A and B). The detection results of ET-1 were in contrast to the levels of NO, and the expression of ET-1 in the ATRA group was significantly lower compared with the high fat group (P<0.05; Fig. 5C).