Fluorescein distribution in retinas of normal and diabetic rats.

Quantitative fluorescence microscopy was used to study fluorescein distribution across the blood-retinal barrier in control and streptozotocin diabetic rats at 2 min, 1 and 2 hr after dye injection (12.5 and 125 mg kg-1 i.v.). Fluorescence intensities of choriocapillaris and retina were compared with plasma fluorescein levels. Diabetic rats had only one-half the total plasma fluorescein concentration of controls by 1 hr after injection, but the proportion of free fluorescein was greater in diabetic animals than in controls, providing the total plasma dye levels did not exceed the binding capacity of plasma proteins. Diabetic rats also had more unbound fluorescein glucuronide in plasma. With fluorescence microscopy no focal fluorescein leakage from retinal capillaries or pigment epithelium was seen in any diabetic or control eye. However, the dye was detected in retinas of diabetic and control animals at all intervals except 2 hr after injection of the lower dose when its fluorescence was too low to measure. Initial fluorescein entry occurred from the choroid by diffusion through pigment epithelial cells creating a steep intensity gradient decreasing from outer to inner retinal layers. With time this gradient flattened and then completely reversed, suggesting removal of dye from outer retinal layers and concomitant equilibration of inner layers with a pool of fluorescein in vitreous humor. Although the pattern of transretinal fluorescence distribution was similar in all rats, in specific instances, retinal fluorescence intensity differed significantly between control and diabetic animals. Fluorescence intensity was higher in diabetic than in control rats at 2 min after the 12.5 mg kg-1 dose and lower at 1 hr after the 125 mg kg-1 dose (P less than 0.05). These differences were directly related to transient differences in plasma free fluorescein concentrations, and in both cases, retinal fluorescence in diabetic rats returned to control values in conjunction with return to control levels of their plasma free fluorescein concentrations. The amount of dye detected in diabetic retinas was not in excess of normal levels at any interval after injection when related to concurrent plasma free fluorescein concentrations. These data do not indicate blood-retinal barrier dysfunction in the diabetic rat, but interpretation of the results is limited by the experimental conditions. With the high dose there was probably initial saturation of active transport mechanisms for dye removal, and with the low dose, dye distribution could not be followed to 2 hr when evidence of abnormal accumulation in retina may be more apparent.