Objective    To evaluate the murine model of dry eye induced by hyperosmolar saline.     Methods    Sixty female BALB/c mice at the age of 6-8 weeks were randomly divided into blank group

control group and experimental group

20 in each group. Mice in control and experimental groups were treated with 308 mOsmol/L and 500 mOsmol/L sodium chloride solution

respectively

5 times a day. Mice in blank group were not treated with sodium chloride solution. Schirmer test

fluorescein staining

corneal scoring

rose bengal staining

tear ferns experiment

corneal epithelial HE staining and thickness measurement

conjunctival epithelial PAS staining and Goblet cell counting were conducted on days 0

7

14

28

and 42

respectively. Corneal surface was observed by scanning electron microscopy on day 42.     Results    No significant difference was found in the above parameters on day 0 between the two groups. On day 7

the volume of tears was significantly smaller in experimental group (2.3±0.4 mm) than in blank group (3.0±0.5mm) and control group (3.1±0.5 mm) (P0.05). Spotty and patchy staining was observed on cornea. The score of corneal fluorescein staining and rose bengal staining was higher in experimental group（1.4±1.6）than in blank and control groups (0.9±1.2 and 0.5±0.7) (P0.01). On day 14

the number of Goblet cells was 10.6±2.0

11.8±1.3 and 5.4±1.8

respectively

in 3 groups (P0.01). The basal corneal epithelial cells were abnormally arranged with vacuoles. The thickness of corneal epithelium 21.69±3.41μm

21.23±2.64μm

and 15.82±1.38μm

respectively

in 3 groups with loss of microvilli on cornea (P0.05).     Conclusion    Hyperosmolar saline can cause pathological lesions similar to dry eye on ocular surface of mice

thus playing an important role in pathogenesis of dry eye.