Lowered cortistatin expression is an early event in the human diabetic retina and is associated with apoptosis and glial activation

Abstract

Purpose: Cortistatin (CST), a neuropeptide with strong structural and functional similarities to somatostatin, is abundant in the vitreous fluid, and it is decreased in patients with proliferative diabetic retinopathy. The aims of the present study were to explore whether the retina produces CST, and to compare its expression between diabetic and nondiabetic donors. Retinal neurodegeneration was assessed by measuring glial fibrilar acidic protein (GFAP) by confocal laser microscopy and counting the apoptotic TUNEL positive cells in which nuclear fragmentation as well as condensation were present.

Methods: Human postmortem eyes (10) from five diabetic donors were compared with 10 eyes from five nondiabetic donors, matched by age. CST mRNA (RT-PCR) and CST (confocal laser microscopy) were measured separately in both the neuroretina and retinal pigment epithelium (RPE). Retinal neurodegeneration was assessed by measuring glial fibrillar acidic protein (GFAP) by confocal laser microscopy and counting the apoptotic cells by TUNEL.

Results: CST was found to be produced by the human retina, and higher levels of CST mRNA were found in RPE than in the neuroretina. CST mRNA levels in diabetic donors were significantly lower in both the RPE (p=0.001) and the neuroretina (p=0.03) in comparison with nondiabetic donors. CST immunofluorescence was in agreement with mRNA expression, but the differences were only significant when comparing neuroretinas (p=0.03). Increased GFAP and a higher degree of apoptosis were observed in diabetic retinas in comparison with nondiabetic retinas. These changes were inversely related with CST levels.

Conclusions: CST is expressed in the human retina. There is more CST in the RPE than in the neuroretina. A lower expression of CST exists in diabetic retinas and it is associated with retinal neurodegeneration.

Figure 1

Expression of CST in the retinal pigment epithelia (RPE) and in the neuroretina (N) of diabetic and nondiabetic donors. A: Semiquantitative RT-PCR shows a higher CST expression in the retina (RPE and N) from nondiabetic donors. Human β-actin was used as internal control and, as can be seen, its signal intensity was similar in the retina of diabetic and nondiabetic donors. B: This shows real-time quantitative RT-PCR analysis of CST mRNA in human retinas. The bars represent the mean±SD of the values obtained in the five diabetic and the five nondiabetic donors studied. CST mRNA gene expression was calculated after normalizing with β-actin. The following abbreviations are used in the figure: size marker (M); positive control, human brain (C); retinal pigment epithelium (RPE); and neuroretina (N).

Figure 2

A comparison of protein content of CST in the retina from diabetic and nondiabetic donors. The upper panel shows a comparison of CST-29 immunofluorescence (red) in the human retina between representative samples from a non diabetic donor (A) and a diabetic donor (B). The bar represents 20 μm. Lower panel: quantification of CST-29 immunofluorescence in non-diabetic and diabetic retinas. The following abbreviations are used in the figure: retinal pigment epithelium (RPE); outer nuclear layer (ONL); inner nuclear layer (INL); and ganglion cell layer (GLC).

Figure 3

Representative images of apoptosis in the retina. Upper panel: nondiabetic donor (A: propidium iodide, B: TUNEL immunofluorescence). Lower panel: diabetic donor (C: propidium iodide, D: TUNEL immunofluorescence). The following abbreviations are used in the figure: retinal pigment epithelium (RPE); outer nuclear layer (ONL); inner nuclear layer (INL); and ganglion cell layer (GCL). The bar represents 20 μm.

Figure 4

A comparison of glial fibrillar acidic protein ( GFAP) immunofluorescence between diabetic and nondiabetic donors. The upper panel shows images of representative samples of retina from a non-diabetic donor (A) and a diabetic donor (B). In the diabetic retina, the endfeet of the Müller cells showed abundant GFAP immunofluorescence (green), and the radial processed stained intensely throughout both the inner and the outer retina. The lower panel shows the quantification of GFAP immunofluorescence in non-diabetic and diabetic retinas. The following abbreviations are used in the figure: outer nuclear layer (ONL); inner nuclear layer (INL); and ganglion cell layer (GCL). The bar represents 20 μm.