Purpose: To image the retina of rats with induced diabetes using multiphoton microscopy to better understand how photoreceptors (PRs) are affected by diabetic retinopathy development. Cell density in diabetic animals was compared with that of a control reference group.
Methods: A custom-developed multiphoton microscope (Bueno et al., JBO 2010) was used to record two-photon excitation fluorescence (TPEF) images of unstained rat retinas. Series of TPEF images of the photoreceptor layer at different retinal locations were recorded. Diabetes was induced in five male Wistar rats by injecting a unique dose of streptozotocin (75 mg/kg) what led to a blood glucose level significantly higher than in the control’s animals as determined three weeks later. The animals were then sacrificed, the ocular globes excised with retinas detached from the fundus and fixated with a paraformaldehyde solution overnight. The retinal tissue was flat mounted on a microscope slide and none of the samples was stained. Unstained retina samples of both the control (N=5) group and the group with induced diabetes (N=5) were imaged at 32 retinal locations. The images of the PR layer were analyzed to determine the cell density. Comparisons between control and diabetic specimens were performed.
Results: Single cells provided enough TPEF signal for direct counting and precise analysis. In all samples, the photoreceptor density changed with eccentricity. Significant differences in PR density between the control eyes and those suffering from diabetes were found. In the control group the density ranged between 38200 and 40900 cells/mm2. These values declined to a range between 32400 and 37200 cells/mm2 in the diabetic group. This reduction occurred at each analyzed retinal location (12% on average). Moreover, Voronoi diagrams and Delaunay triangulation allowed us to explore changes in the PR mosaic distribution.
Conclusions: Multiphoton microscopy provided a clear visualization of single PR in the rat retina. Significant differences were found when comparing control retinas and those from animals with induced diabetes. For every retinal location, a reduction in PR density was observed in diabetic specimens. These results may help to a better characterization of retinal alterations during diabetes development in the rat model.