Retinal ion regulation in a mouse model of diabetic retinopathy: natural history and the effect of Cu/Zn superoxide dismutase overexpression

Abstract

Purpose: To test the hypotheses that manganese-enhanced MRI (MEMRI) is useful in evaluating intraretinal ion dysregulation in wild-type (WT) and Cu/Zn superoxide dismutase (SOD1) overexpressor mice.

Methods: Central intraretinal ion activity and retinal thickness were measured from high-resolution data of light- and dark-adapted WT C57BL/6 mice (to gauge MEMRI sensitivity to normal visual processing in mice) and dark-adapted diabetic and nondiabetic WT and Cu/Zn superoxide dismutase overexpressor (SOD1OE) mice. Glycated hemoglobin and retinal vascular histopathology were also determined.

Results: In WT mice, light adaptation reduced outer retinal manganese uptake compared with that in dark adaptation; no effect on inner retinal uptake was found. In diabetic WT mice, intraretinal manganese uptake became subnormal between 1.5 and 4 months of diabetes onset and then relatively increased. Central retinal thickness, as determined with MEMRI, decreased as a function of age in diabetic mice but remained constant in control mice. Nondiabetic SOD1OE mice had normal retinal manganese uptake but subnormal retinal thickness and supernormal acellular capillary density. At 4.2 months of diabetes, SOD1OE mice had normal manganese uptake and no further thinning; acellular capillaries frequency did not increase by 9 to 10 months of diabetes.

Conclusions: In emerging diabetic retinopathy, MEMRI provided an analytic measure of an ionic dysregulatory pattern that was sensitive to SOD1 overexpression. The potential benefit of SOD1 overexpression to inhibit retinal abnormality in this model is limited by the retinal and vascular degeneration that develops independently of diabetes.

Figure 1

MEMRI changes in murine inner and outer retinal signal intensity with light and dark adaptation. (A) Representative MEMRI of a WT mouse eye. White arrows: region illustrated in (B). (B) Pseudocolor-linearized images of average retinal signal intensity in central retina of WT mice (left, light adapted, n = 4; right, dark adapted, n = 4). The same pseudocolor scale was used for linearized images in which blue to green to yellow to red represent lowest to highest signal intensity. Dotted black line: boundary between inner and outer retina, as demonstrated in previous studies. Intraretinal locations from which inner retinal (IR) and outer retinal (OR) data were extracted and used in this study are indicated by the black arrows on the right of the dark-adapted linearized image. (C) Summary of IR and OR signal intensities in light- and dark-adapted mice. Numbers of animals used to generate these data are listed above each bar. Error bars represent the SEM. Comparisons were performed between inner retinal signal intensities during light or dark adaptation and between outer retina values during the different adaptation conditions. *P = 0.0025. The y-axis scale starts at 50 because this is the pre–manganese baseline level determined from uninjected mice (data not shown).

Figure 2

Summary of natural history of signal intensities for (A) inner retina (IR) and (B) outer retina (OR). Gray symbols: average WT (combined Jackson and modified Hilltop mice) IR and OR intensities. *P = 0.0001 for 1.5, 2.5, and 4 months of diabetes, 2-tailed, compared with average WT IR (91.0 ± 0.6 AU; n = 21) or OR (89.1 ± 0.6 AU) intensities. ^%P = 0.0001 (5.5 and 7.4 months of diabetes); ^#P = 0.02 (IR) or 0.05 (OR) (0.9 months of diabetes), 2-tailed, compared with average IR (75.8 ± 0.6 AU; n = 21) or OR (76.6 ± 0.6 AU) of mice with 1.5 to 4 months of diabetes. White symbols: modified 4-month-old diabetic Hilltop mice signal intensities are shown for clarity. Error bars are SEM. The y-axis scale starts at 50 because this is the pre–manganese baseline level determined from uninjected mice (data not shown).

Figure 3

Summary of changes in mean MEMRI intraretinal signal intensity in control and diabetic WT and SOD1OE mice. The y-axis scale starts at 50 because this is the pre–manganese baseline level determined from uninjected mice (data not shown). Asterisk: signifi-cant comparisons with control inner and outer retina intensities are shown (P = 0.0001). Error bars are SEM, and numbers over bars are numbers of animals studied.

Figure 4

Plot of number of mean degenerate retinal capillaries (acellular capillaries) in diabetic WT mice (9 months of study; WT+D) compared with that in nondiabetic controls (WT; left) and diabetic SOD1OE mice (SOD1+D) compared with that in nondiabetic SOD1OE controls (SOD1; right). Diabetes-induced capillary degeneration was increased in WT (WT+D) but not SOD1 overexpressor (SOD1+D) mice. Numbers of animals used to generate these data are listed above each bar. Error bars represent the SEM. *P < 0.05 was considered significant.