METHODS: Thirty healthy adult male SD rats were randomly assigned to the normal control group(CON group, n=10)and the diabetes group(DM group, n=20). After fasting for 12h, the DM group was injected with 1% streptozotocin(STZ)solution, according to 60mg/kg disposable left lower abdominal injection. After 72h, blood was taken from the rat tail vein to detect blood glucose, diabetic model animals were defined as ≥16.7mmol/L. Model rats were randomly divided into diabetes control group(group D)and NAC treatment group(group N). After the model was established, N group of rats were injected with 4μL 1.6μg/μL NAC through the vitreous cavity every week. Rats in CON group and D group were injected with 4μL 0.01mmol/L phosphate buffer saline. All the rats no diet water, group feeding. Body mass and blood glucose were recorded weekly. After the diabetes was modeled, 2mo killed the experimental animals. The thickness of the inner layer of the retina of rats in each group was determined by HE staining. The number of retinal ganglion cells and the level of pigment epithelial derived factor in the retina were measured by immunofluorescence.

RESULTS: The thickness of retinal kernel layer increased in group N compared with group D(P<0.01), and there was no difference between group CON and group(P>0.05). Compared with CON group, the number of retinal ganglion cells decreased in group D(P<0.01), and decreased slightly in group N(P>0.05). Retinal ganglion cells decreased in group D compared with group N(P<0.01). Compared with CON group, PEDF expression decreased in group D(P<0.01), and decreased slightly in group N(P>0.05). The expression of PEDF in group D decreased compared with group N(P<0.01).

CONCLUSION: The protective effect of antioxidant NAC on retinal tissue in early diabetic rats may be due to the up-regulation of PEDF levels in the retina.