Inhibition of connective tissue growth factor by small interfering ribonucleic acid prevents increase in extracellular matrix molecules in a rodent model of diabetic retinopathy

Abstract

Purpose: Connective tissue growth factor (CTGF) is a profibrotic factor that induces extracellular matrix (ECM) production and angiogenesis, two processes involved in diabetic retinopathy (DR). In this study, we examined whether insulin therapy or a CTGF-specific small interfering RNA (siRNA) administered to diabetic rats decreased the levels of CTGF and of selected putative downstream genes in the retina.

Methods: Rats with streptozotocin-induced diabetes were used. Animals received either no treatment for 12 weeks or were administered constant insulin therapy. MRNA and protein levels of CTGF and select ECM genes were determined using real-time PCR and western blotting of the retina. Localization of CTGF in the retina was visualized using immunohistochemistry. A group of diabetic rats received intravitreal injection of CTGF siRNA, and the retinas were examined three days later.

Results: CTGF mRNA and protein significantly increased in the retinas of diabetic rats. Immunohistochemistry indicated that retinal Müller cells of diabetic rats expressed CTGF. Hyperglycemia upregulated mRNA levels of fibronectin, laminin β1, collagen IVα3, and vascular endothelial growth factor (VEGF), and this increase was prevented by insulin therapy. Treatment of diabetic rats with CTGF siRNA decreased laminin β1, collagen IVα3 mRNA, and CTGF mRNA and protein but did not affect fibronectin or vascular endothelial growth factor mRNA levels.

Conclusions: These results indicate that CTGF and ECM genes can be regulated using insulin. Importantly, these results also suggest that CTGF regulates changes in ECM molecules in DR.

Figure 1

Connective tissue growth factor (CTGF) expression increased in retina of 8 and 12 week diabetic rats. CTGF mRNA expression in the retinas of non-diabetic and diabetic rats after 8 and 12 weeks of hyperglycemia were analyzed using real-time PCR and normalized to the housekeeping gene acidic ribosomal phosphoprotein P0 (ARPP₀). A: CTGF mRNA levels increased six- and sevenfold at both time points. B: A representative western blot illustrating CTGF and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) protein expression in the retinas of diabetic rats after 8 and 12 weeks of hyperglycemia. Note the increased CTGF protein levels compared to the non-diabetic controls. C: Densitometric analysis of three separate immunoblots indicates a 2.5 and 5.7 fold increase in the CTGF protein after 8 and 12 weeks of hyperglycemia, respectively. (*p<0.05, **p<0.001 diabetic versus non-diabetic) n=non-diabetic; D=diabetic. Proteins ran on a 4%–15% gradient gel. n=3/age.

Figure 2

Insulin therapy in diabetic rats abolished hyperglycemia-induced increases in levels of selected genes and proteins. mRNA expression in the retinas of non-diabetic, diabetic, and diabetic rats treated with insulin were analyzed using real-time PCR and normalized to the housekeeping gene acidic ribosomal phosphoprotein P0 (ARPP₀). A: Real-time PCR analysis of mRNA extracted from the retina. Note the increase in the levels of connective tissue growth factor (CTGF; sevenfold), fibronectin (threefold), laminin β1 (2.5-fold), collagen IVα3 (2.3 fold), and vascular endothelial growth factor (VEGF; twofold) mRNAs in the rat retinas after 12 weeks of hyperglycemia and that this increase was prevented with insulin therapy. B: A representative immunoblot showing CTGF expression in the retinas of normoglycemic control, diabetic, and insulin-treated rats. Proteins were ran on a 15% SDS–PAGE gel, allowing for better protein separation, hence the double band for CTGF. C: Densitometric analysis of three different western blots indicated that the level of CTGF increased following 12 weeks of hyperglycemia, when compared to the non-diabetic controls. In contrast, CTGF expression in the retinas of diabetic rats treated with insulin for 12 weeks remained near the control levels. D: A representative western blot for VEGF and laminin in retinal extracts shows an increase in retinal VEGF and laminin in the hyperglycemic state, and this increase was inhibited in the insulin-treated hyperglycemic rats. E: Densitometric analysis of three different western blots revealed a 2.1-fold increase in VEGF and a 5.2-fold increase laminin. VEGF and laminin levels were similar in the control and insulin-treated diabetic animals. GAPDH and Ponceau S staining were used as a loading control. n=non-diabetic; D=diabetic; D+IN=insulin treated diabetic (*p<0.05, **p<0.001). n=3/group.

Figure 3

Connective tissue growth factor (CTGF) was detected in vimentin-positive Müller cells in the retina of diabetic rats. A: Müller cells labeled with filament protein vimentin throughout the diabetic retina. B: CTGF staining in the 12-week diabetic retina is also seen throughout the retina. C: Merge photomicrograph for vimentin and CTGF shows that CTGF colocalizes with vimentin in the diabetic retina. Bar=60 um. D: High magnification of the area indicated with a dotted rectangle in C. Bar=5 um. V=Vitreous, ONL=Outer Nuclear Layer, S=Sclera.

Figure 4

Hyperglycemia induces connective tissue growth factor (CTGF) expression in Müller cells. A representative photomicrograph illustrates retinas processed for visualization of CTGF. The CTGF level is low in the retinas of the control rats (A) and increased in the retinas of the diabetic rats after 12 weeks of hyperglycemia (B). Insulin therapy for 12 weeks decreased the staining intensity of CTGF (C). All images were captured at the same magnification. V=vitreous side of the retina. Bar=30 um. n=3/group.

Figure 5

siRNA decreased transforming growth factor-β1 (TGF-β)-induced connective tissue growth factor (CTGF) in Rat-2 fibroblasts. CTGF mRNA expression in Rat-2 cells treated with TGF-β and siRNA were analyzed using real-time PCR and normalized to the housekeeping gene acidic ribosomal phosphoprotein P0 (ARPP₀). A: The effect of two separate CTGF siRNAs (I and III) and a control scrambled siRNA on TGF β-induced CTGF expression in Rat-2 fibroblasts was tested. Real-time PCR showed that siRNA I and siRNA III decreased CTGF mRNA by 75% and 86%, respectively. B: Arepresentative western blot documenting that TGF-β induced CTGF protein decreased by siRNA I and III. Proteins were run on the same gel, but not in adjacent lanes. C: Densitometric analysis summarizing the results of three separate western blots shows that siRNA I and siRNA III decreased CTGF protein by 49% and 46%, respectively. (*p<0.05). n=3/group.

Figure 6

Intravitreal injection of labeled siRNA penetrates the retina. A: Confocal microscopy revealed that the scrambled siRNA covalently bound to Cy3 (Red) was distributed throughout the layers of the retina when injected into the eyes of control rats. The label was distributed throughout the retina from the ganglion cell layer (GCL) to the photoreceptor layer (PRL). Nuclei were stained with To-Pro Blue. B: A photomicrograph illustrates panel A with To-Pro Blue removed. Note the distribution of Cy3 labeled siRNA throughout the layers of the retina. C: In contrast, the retinas of rats injected with unlabeled scrambled siRNA did not contain labeled particles, indicating that the label was not due to autofluorescence or artifacts from the injection. Bar=30 um. D: Illustrates higher magnification of inset depicted with dotted rectangle in A. Bar=15 um. IPL=inner plexiform layer, INL=inner nuclear layer. n=3/group.

Figure 7

siRNA treatment significantly decreased hyperglycemia-induced increase in connective tissue growth factor (CTGF) mRNA and protein, glial fibrillary acidic protein (GFAP), collagen IVα3 and laminin-β1 gene expression. mRNA expression for CTGF and selected genes in the retinas of the diabetic rats after 12 weeks of hyperglycemia were analyzed using real-time PCR and normalized to the TATA-binding protein (TBP). A: Real-time PCR revealed that 3 days post intravitreal injection, CTGF siRNA induced a decrease in CTGF (33%), GFAP (44%), collagen IVα3 (71%), and laminin β1 (63%) mRNAs. In contrast, CTGF siRNA did not affect the level of fibronectin or vascular endothelial growth factor (VEGF). B: Immunoblot analysis of CTGF levels in the retinas following a single intravitreal injection of CTGF siRNA or scrambled siRNA into left and right eye, respectively. The concentration of the CTGF protein is lower in eyes injected with CTGF siRNA. C: Densitometric analysis of three independent experiments revealed a 54% decrease in CTGF protein in retinas injected with CTGF siRNA compared to retinas injected with a scrambled (non-specific) siRNA. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as loading control. D: Real-time PCR showed that CTGF siRNA had no effect on CTGF expression 10 days after injection. (*p<0.05). n=5/group.