Comparison of Dextran Perfusion and GSI-B4 Isolectin Staining in a Mouse Model of Oxygen-induced Retinopathy.

PURPOSE Oxygen-induced retinopathy (OIR) is a robust and widely used animal model for the study of retinal neovascularization (NV). Dextran perfusion and Griffonia simplicifolia isolectin B4 (GSI-B4) staining are two common methods for examining the occurrence and extent of OIR. This study provides a quantitative comparison of the two for OIR detection. METHODS At postnatal day 7 (PN7), fifteen C57BL/6J mice were exposed to a 75% hyperoxic condition for 5 days and then returned to room air conditions. At PN17, the mice received intravitreal injection of GSI-B4 Alexa Fluor 568 conjugate. After 10 hours, they were infused with FITC-dextran conjugate via the left ventricle. Retinal flat mounts were photographed by confocal microscopy. Areas with fluorescent signals and the total retinal areas were quantified by Image J software. RESULTS Both GSI-B4 and dextran detected the peripheral neovascular area. The mean hyper fluorescence area was 0.33 ± 0.14% of whole retinal area determined by GSI-B4 staining and 0.25 ± 0.28% determined by dextran perfusion. The difference between the two measures was 0.08% (95% CI:-0.59%, 0.43%). The Pearson correlation coefficient between the two methods was 0.386,P =0.035. The mean coincidence rates were 14.3 ± 13.4% and 24.9 ± 18.5% for GSI-B4 and dextran staining, respectively. CONCLUSION Both methods can complement each other in demonstrating and quantitatively evaluating retinal NV. A poor agreement was found between the two methods; GSI-B4 isolectin was more effective than FITC-dextran perfusion in evaluating the extent of retinal NV in a mouse model of OIR.