Fetal bovine serum (FBS) was purchased from PAA Laboratories (Colbe, Germany). DMSO was obtained from Sigma (St. Louis, MO, USA), anti-mouse polyclonal antibody HIF-1α was purchased from Novus Biologicals (NB100-105; Littleton, CO, USA) and 4% paraformaldehyde was purchased from Wuhan Boster Biological Technology, Ltd. (Wuhan, China). The bicinchoninic acid (BCA) and VEGF ELISA kit were obtained from Beijing Dingguo Changsheng Biotechnology Co. Ltd. (Beijing, China). The anti-mouse immunohistochemistry kit was purchased from Jingmei Inc. (Shanghai, China), anti-β-actin secondary antibody was purchased from Beijing Zhongshan Jinqiao Biotechnology Co. Ltd. (Beijing, China) and HRP-goat anti-mouse IgG was obtained from Beyotime Institute of Biotechnology (Shanghai, China).

Rhesus monkey retinal fovea vascular endothelial cells (RF/6A) were obtained from our laboratory (Department of Ophthalmology, The Second Xiangya Hospital, Xiangya School of Medicine, Central South University, Changsha). The RF/6A cells were maintained in DME containing 10% FBS. The cells were incubated with 5% CO₂ at 37°C. The study was approved by the Institutional Review Board of The Second Xiangya Hospital, Central South University (Changsha, China).

Bovine serum albumin (BSA) was glycated by incubation with glucose-6-phosphate in phosphate-buffered saline (PBS) for 6 weeks at 37°C, as described previously (7). Dialyzed glycated protein was characterized based on fluorescence at 450 nm upon excitation at 390 nm using a fluorescence spectrometer (model LS-3B; Perkin-Elmer Corp., Norwalk, CT, USA). The endotoxin content in each sample was measured by the Limulus amebocyte lysate assay (E-Toxate; Sigma) and observed to be below detectable levels (0.2 ng/ml). For the control, non-glycated albumin consisted of the same initial preparations of albumin incubated at 37°C in the absence of sugar. The glycation process was performed twice, and the two AGE preparations yielded similar results.

The cells were divided into two groups: the AGE experimental group and the non-glycated control group. The RF/6A cells were incubated with different concentrations of AGEs (0, 50, 100, 200, 400 and 800 mg/l) for 24 h.

The conditioned media VEGF levels were determined using a sandwich ELISA assay according to the manufacturer’s instructions. Briefly, cells were cultured for 24 h, the supernatant was removed and transferred to wells which were coated with a monoclonal antibody to VEGF. VEGF proteins levels were normalized to cell counts.

Briefly, after the cells were cultured for 24 h, the culture solution was removed and cells were washed in PBS three times. The cells were fixed using 4% paraformaldehyde for 10 min. The cells were then incubated with polyclonal antibody to HIF-1α (1:100) and stained with DAB.

After the cells were cultured with various concentrations of AGEs for 24 h, the culture solution was discarded and cells were washed with PBS three times. The cells were scraped and diverted into a 1.5 ml microtube which contained RIPA buffer with protease inhibitor. After being put on ice for 3 min, the mixture was agitated, dissolved and put on ice for a further 30 min. The mixture was centrifuged at 4°C for 25 min [5,220 × g; Microcentrifuge 5415R (5415D); Eppendorf, Stevenage, UK], the top clear liquid layer was removed and western blotting was performed in triplicate.

Values were expressed as the mean ± standard deviation (SD). Statistical analyses were performed using SPSS software V11.5 (SPSS, Inc., Chicago, IL, USA). Continuous data were analyzed by the Levene method. The significance of a difference between groups was evaluated using one-way ANOVA with a post hoc Student-Newman-Keuls multiple comparisons test. The experimental and control groups were compared using a matched Student’s t-test. The correlation between VEGF and HIF-1α expression was analyzed with double variable regression and correlation analysis. P<0.05 was considered to indicate a statistically significant result.

RF/6A cells are a type of monolayer-adherent cell. Under a microscope, the cells represented two different shapes. One type of cell was short and ‘shuttle-like’ with a diminished cell volume. In addition, the cytoplasm was particularly reflective, presented a typical ‘cobblestone’ appearance and was small in number. The other type of cell had a larger volume and was polygonal. This type of cell grew well, was greater in number, had extensive cytoplasm and one or more nucleoli (Fig. 1).

As shown in Fig. 2, the non-glycated control groups secreted low levels of VEGF and there was no statistical significance between the cells treated with different concentrations of non-glycated albumin (P>0.05). By contrast, VEGF secretion was significantly higher in all AGE experimental groups compared with the control groups (P<0.05). In addition, VEGF secretion increased as the AGE concentration increased and the VEGF level reached its highest point at 200 mg/l AGE. At higher concentrations of AGEs, the VEGF secretion gradually decreased, but there remained a significant difference compared with the control group (P<0.05). In summary, VEGF expression was AGE concentration-dependent.

As shown in Fig. 3, HIF-1α protein expression was observed in the AGE experimental groups, where the majority of HIF-1α was located in the nucleus and a small quantity was located in the cytoplasm, which was significantly different from the control group (P<0.05). However, no significant differences (P>0.05) were observed among the experimental groups. In addition, the difference in HIF-1α expression between the AGE group and blank control group was not statistically significant (P>0.05). The non-glycated control groups had low HIF-1α protein expression levels, and there were no significant differences (P>0.05) among the non-glycated control groups. However, the HIF-1α expression levels of the AGE experimental groups were significantly higher compared with those of the non-glycated control groups.

In the present study, we also detected the HIF-1α protein expression by western blot analysis. As shown in Fig. 4, the level of HIF-1α expression was lower in all non-glycated control groups than in the AGE experimental groups. However, different levels of HIF-1α protein expression were detected, and the HIF-1α expression level increased along with the AGE concentration. The maximum level of HIF-1α expression occurred when the cells were treated with 200 mg/l AGEs. The HIF-1α expression level then decreased with increasing AGE concentration. These results were consistent with VEGF expression in the AGE experimental groups.

As shown in Fig. 5, the levels of HIF-1α protein expression in the non-glycated control groups were low, and no significant difference was observed among the control groups (P>0.05). However, the level of HIF-1α expression was significantly higher in all AGE experimental groups than in the control groups (P<0.05). Accordingly, as the AGE concentration increased, HIF-1α protein expression was promoted; the maximum HIF-1α expression occurred when the AGE concentration was 200 mg/l. The expression level then decreased, but with no significant difference (P>0.05). In summary, these results suggest that the HIF-1α expression level was dependent on AGE concentration.

The association between VEGF and HIF-1α was analyzed by a bivariate vector autoregressive model and correlation analysis. The results showed there was a positive correlation between VEGF and HIF-1α, and the regression equation was y = 1198.2× − 241.78, R=0.9881, P<0.01.