Diabetic retinal neurodegeneration is associated with mitochondrial oxidative stress and is improved by an angiotensin receptor blocker in a model combining hypertension and diabetes

Abstract

Objective: Diabetic retinopathy displays the features of a neurodegenerative disease. Oxidative stress is involved in the pathogenesis of diabetic retinopathy. This investigation sought to determine whether hypertension exacerbates the oxidative stress, neurodegeneration, and mitochondrial dysfunction that exists in diabetic retinopathy and whether these changes could be minimized by the angiotensin II type 1 (AT(1)) receptor blocker (ARB) losartan.

Research design and methods: Diabetes was induced in spontaneously hypertensive rats (SHRs) and normotensive Wistar-Kyoto (WKY) rats. The diabetic SHRs were assigned to receive or not receive losartan.

Results: The level of apoptosis in the retina was higher in diabetic WKY rats than in the control group, and higher levels were found in diabetic SHRs. The apoptotic cells expressed neural and glial markers. The retinal glial reaction was more evident in diabetic WKY rats and was markedly accentuated in diabetic SHRs. Superoxide production in retinal tissue increased in diabetic WKY rats, and a greater increase occurred in diabetic SHRs. Glutathione levels decreased only in diabetic SHRs. As a consequence, the levels of nitrotyrosine and 8-hydroxy 2'-deoxyguanosine, markers of oxidative stress, were elevated in diabetic groups, mainly in diabetic SHRs. Mitochondrial integrity was dramatically affected in the diabetic groups. The ARB treatment reestablished all of the above-mentioned parameters.

Conclusions: These findings suggest that concomitance of hypertension and diabetes exacerbates oxidative stress, neurodegeneration, and mitochondrial dysfunction in the retinal cells. These data provide the first evidence of AT(1)blockage as a neuroprotective treatment of diabetic retinopathy by reestablishing oxidative redox and the mitochondrial function.

FIG. 1.

A: A representative immunohistochemical identification of TUNEL-positive cells in the retinal sections of control and diabetic WKY rats and SHRs and losartan-treated diabetic SHRs. The positive controls were retinal slides treated with DNase. The majority of positive cells were localized in the outer nuclear layer of the retina, indicating, therefore, that the photoreceptors are the most affected cells. B: Summary of the number of TUNEL-positive cells per retinal section (0.5 ± 0.1 vs. 2.3 ± 0.9 positive cells per retinal section for control WKY vs. diabetic WKY rats, P = 0.03; 2.3 ± 0.9 vs. 4.4 ± 2.0 positive cells per retinal section for diabetic WKY rats vs. diabetic spontaneously hypertensive rats, P = 0.01; 0.6 ± 0.3 vs. 4.4 ± 2.0 positive cells per retinal section, control spontaneously hypertensive vs. diabetic spontaneously hypertensive, P = 0.0003; 4.4 ± 2.0 vs. 2.0 ± 0.4 positive cells per retinal section for diabetic spontaneously hypertensive vs. diabetic spontaneously hypertensive–losartan, P = 0.01). *P = 0.03; † P = 0.01. CT, control; DM, diabetic; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; Los, losartan; ONL, outer nuclear layer; OPL, outer plexiform layer. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 2.

A: Retinal section with a caspase-3–positive cell colabeled with GFAP antigen localized on outer nuclear layer, as indicated by the arrow. This is indicative of the glial nature of the apoptotic cell. B: Caspase-3–positive cell expressing the intermediate filament nestin in the outer nuclear layer of the retina, demonstrating the neural origin of the indicated cell. Scale bars = 50 μm. Both double-labeling immunofluorescence assays were performed in retinal tissue obtained from diabetic hypertensive rats because of its higher number of TUNEL-positive cells observed. DAPI, 4′,6′-diamino-2-phenylindole; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 3.

Evaluation of glial cell reactivity by GFAP immunofluorescence in retinas of control and diabetic WKY and spontaneously hypertensive rats. The presence of diabetes or hypertension alone induced a clear increase in GFAP immunoreactivity throughout the retina. The concomitance of both provoked a further increase in glial reactivity, and the losartan treatment abolished this effect. Bars = means ± SD of percentage of fluorescence per millimeter squared of retina. Scale bars = 50 μm. The graph shows 1.3 ± 0.3 vs. 5.3 ± 0.5% of fluorescence/mm² of retina for control WKY vs. diabetic WKY rats, *P < 0.0001; 5.3 ± 0.5 vs. 4.9 ± 0.08% of fluorescence/mm² of retina for diabetic WKY rats vs. control SHRs, P = 0.4; 4.9 ± 0.08 vs. 10.8 ± 0.5% of fluorescence/mm² of retina for control SHRs vs. diabetic SHRs, †P < 0.0001; 10.8 ± 0.5 vs. 4.7 ± 0.5% of fluorescence/mm² of retina for diabetic SHRs vs. diabetic losartan (Los)-treated SHRs, †P < 0.0001. CT, control; DM, diabetic; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 4.

Total superoxide generation in retinal tissue. Superoxide anion generation of retinal tissue was determined by the lucigenin-enhanced chemiluminescence method, and photoemission was measured every 10 s for 3 min. The peak level of superoxide generation was observed ∼3 min after lucigenin was added to the reaction buffer containing retina from different groups. Diabetes increased superoxide production (RLU · min⁻¹ · mg protein⁻¹) in the retina, and the concomitance of diabetes and hypertension further exacerbated this parameter. The presence of diphenyliodonium (20 μmol/l), an inhibitor of flavin-containing oxidases, did not affect superoxide production, whereas preincubation with rotenone (100 μmol/l) resulted in a marked reduction in superoxide production from retinal tissue. This indicates that mitochondria are an important source of the superoxide in retina tissue. Treatment with the ABR losartan reduced superoxide production to levels observed in the control groups. Bars are the means ± SD. *P = 0.03 vs. control WKY; †P = 0.0002 vs. other groups. CT, control; DM, diabetic; DPI, diphenyliodonium.

FIG. 5.

Concentration of reduced GSH from retinas of control and diabetic WKY and SHR rats and treated diabetic SHRs (μmol/l GSH per μg of retinal protein). There was a significant decrease in GSH levels in diabetic SHRs compared with control SHRs; the ARB treatment in diabetic hypertensive rats prevented this reduction. Bars = means ± SD (79 ± 17 vs. 43 ± 7 μmol/l reduced GSH per μg retina protein for control WKY vs. diabetic WKY rats, P = 0.1; 140 ± 54 vs. 56 ± 10 μmol/l reduced GSH per μg retina protein for control SHRs vs. diabetic SHRs, P = 0.0005). *P = 0.006. CT, control; DM, diabetic; Los, losartan.

FIG. 6.

A: Representative photomicrograph of immunolocalization of nitrotyrosine in control and diabetic WKY and SHRs and diabetic SHRs treated with losartan. The presence of nitrotyrosine is indicated by the brown color, and the positivity is diffuse. Bars represent the means ± SD of score of positivity of nitrotyrosine per retinal section, as defined in research design and methods. Diabetes increased tyrosine nitration in retinal tissue, and concomitance of diabetes and hypertension exacerbates this phenomenon. A significant reduction was observed in treated diabetic SHRs. The graph shows 2.0 ± 0.1 vs. 3.2 ± 0.2 score of positivity for control WKY vs. diabetic WKY rats, *P < 0.0001; 2.0 ± 0.1 vs. 2.7 ± 0.2 score of positivity for control WKY vs. control SHRs, P = 0.0003; 3.2 ± 0.2 vs. 2.7 ± 0.2 score of positivity for diabetic WKY rats vs. control SHRs, P = 0.004). †P = 0.001 vs. other groups. B: Representative photomicrograph of immunohistochemistry for 8-OHdG from retinas of control and diabetic WKY and SHRs and diabetic SHRs treated with losartan. Bars represent the means ± SD of percentage of positive 8-OHdG retinal cells per retinal field, as defined in research design and methods. The presence of 8-OHdG is indicated by the brown color, and the positivity is in the nucleus of the cells. The presence of diabetes or hypertension separately increased the oxidative DNA damage in retinal cell layers, and concomitance of both induced a markedly pronounced effect; the antihypertensive treatment with ARB promoted significant neuroprotection of retinal tissue, abolishing this damage. All the retinal layers were included in the quantitation for both proteins. *P = 0.003 vs. control WKY; †P < 0.0001 vs. other groups. CT, control; DM, diabetic; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; Los, losartan; ONL, outer nuclear layer; OPL, outer plexiform layer. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 7.

A: Western blot analysis of Bcl-2 in total retinal lysated of the studied groups; the bars represent the means ± SD of band densities expressed in arbitrary densitometric units from at least three independent experiments. *P < 0.05. B: Western blot analysis of UCP-2 in immunoprecipitated retinal protein of the studied groups. The bars represent the means ± SD of band densities expressed in arbitrary densitometric units from at least three independent experiments. *P = 0.03. CT, control; DM, diabetic; Los, losartan.