Effect of Pigment Epithelium–Derived Factor on Glutamate Uptake in Retinal Müller Cells under High-Glucose Conditions

Abstract

Purpose.: A predominant function of Müller cells is to regulate glutamate levels, but it is compromised in diabetic retinopathy (DR). The present study was performed to determine the role of pigment epithelium–derived factor (PEDF) in glutamate uptake in retinal Müller cells in high-glucose conditions.

Methods.: The levels of Kir4.1, PEDF, and GLAST in retinal Müller cells in high-glucose conditions were analyzed by Western blot analysis and real-time RT-PCR, and a glutamate uptake assay was undertaken to investigate the activity of GLAST. Intracellular reactive oxygen species (ROS) generation was measured, and a glutathione peroxidase (GPx) activity assay was performed to determine oxidative stress in the presence of high glucose. After being treated with PEDF in high-glucose conditions, these proteins (Kir4.1 and GLAST) and their mRNAs, glutamate uptake activity, and oxidative stress in Müller cells were investigated. A PEDF siRNA method was used to confirm the effect of PEDF on glutamate uptake activity in Müller cells.

Results.: In high-glucose conditions, glutamate uptake and Kir4.1, PEDF, and GLAST expression in Müller cells decreased significantly. Simultaneously, ROS generation increased, and GPx activity decreased. PEDF treatment inhibited the high-glucose–induced decreases in glutamate uptake and in Kir4.1 and GLAST expression and ameliorated oxidative stress. Moreover, downregulation of PEDF expression by siRNA in Müller cells resulted in decreases in glutamate uptake, Kir4.1 and GLAST expression, and induced oxidative stress.

Conclusions.: These findings suggest that PEDF acts as an antioxidative agent against the decrease in Kir4.1 and GLAST expression and compromise of glutamate uptake activity in retinal Müller cells in diabetic conditions.

Diabetic retinopathy (DR) is a major cause of blindness in the working-age population in developed countries, but the specific mechanism of DR is not completely understood. ^1,2 However, it is known that the pathogenesis of DR includes not only vascular changes but also neuronal damage. ^{3 –9} Furthermore, diabetes-induced changes in retinal neurons and glia may precede the onset of clinically evident vascular injury. ^{10 –12} It is understood that diabetes-induced neuronal injuries are closely related to excessive glutamate levels which are regulated predominantly by Müller cells in the retina. In diabetes, the ability of Müller cells to sustain the normal glutamate levels in the retina are compromised, ^{5 –8,13 –17} but the mechanism is not completely elucidated. An essential step in the regulation of extracellular glutamate is the transport of this amino acid into Müller cells through the high-affinity l-glutamate/l-aspartate transporter (GLAST). Li and Puro ⁷ demonstrated that soon after the onset of diabetes, the function of GLAST in retinal Müller cells is decreased by a possible mechanism involving oxidation. How oxidative stress causes the function of GLAST of Müller cells to decrease in DR is still unclear. 

GLAST is an electrogenic glutamate transporter that is strongly voltage dependent. Depolarization of Müller cells decreases the efficiency of glutamate uptake by Müller cells. ^18,19 Kir4.1 channels are the major inwardly rectifying channels in Müller cells and are widely thought to support K+ and glutamate uptake by Müller cells. Membrane depolarization of Müller cells after functional inactivation of Kir4.1 decreases the efficiency of GLAST in glutamate uptake. ^20,21 According to some reports, oxidative stress conditions, such as ischemia–reperfusion, ocular inflammation, and DR, cause the depolarization of Müller cells as a consequence of functional inactivation or downregulate the inwardly rectifying potassium channel Kir4.1, possibly resulting in the dysfunction of GLAST. ^{20 –22} The depolarization of Müller cells can also be induced by inflammatory lipid mediators such as arachidonic acid and prostaglandins, which are produced in conditions of oxidative stress. ^{23 –25}  

Although the possible mechanisms of dysfunction of GLAST that are induced by oxidative stress in Müller cells are not known entirely, we hypothesized that if the influence of oxidative stress induced by DR on Kir4.1 could be inhibited, at least partially, the function of GLAST would be protected. Pigment epithelium–derived factor (PEDF) is a multifunctional protein that has been demonstrated as a major antiangiogenic factor, a neuroprotective agent, an anti-inflammatory factor, and an antioxidative agent in the eye. ^{26 –30} Banumathi et al. ³¹ showed that PEDF has an antioxidant effect in bovine retinal endothelial cells at a high glucose level that could induce oxidative stress in several retinal cells including Müller cells. ^26,27,31 Previous reports have suggested that PEDF mitigates inflammation and oxidative stress in retinal pericytes exposed to oxidized low-density lipoprotein involved in DR ³² and PEDF inhibits oxidative stress–induced apoptosis and dysfunction of cultured retinal pericytes. ³³ A recent report demonstrated that PEDF could be a therapeutic option for ameliorating the effects of advanced glycation end products and oxidative stress in DR. ³⁰ Whether PEDF acts as an antioxidant agent to inhibit oxidative stress–induced dysfunction of GLAST in Müller cells remains unconfirmed. 

In this study, we hypothesized that counteraction of GLAST malfunction in Müller cells in high-glucose conditions by PEDF is mediated by its antioxidative activity to ameliorate downregulation of Kir4.1. To test this hypothesis, the levels of Kir4.1, PEDF, and GLAST in retinal Müller cells in high-glucose conditions were analyzed by Western blot analysis and real-time RT-PCR, and a glutamate uptake assay was undertaken to investigate the activity of GLAST. Intracellular ROS generation and GPx activity were measured to determine the level of oxidative stress in high glucose. After being treated with PEDF in high-glucose conditions, these proteins (Kir4.1 and GLAST) and their mRNAs, glutamate uptake activity, and oxidative stress in Müller cells were investigated. To confirm the effect of PEDF on glutamate uptake activity in Müller cells, the PEDF siRNA method was used. 

Materials and Methods

Müller Cell Culture and PEDF siRNA Transfection

Müller cells were obtained by a previously described method. ³⁴ Briefly, enucleated eyes from SD rats at postnatal day (PN)5 to PN7 were obtained in sterile conditions, and the eyes were soaked as intact eyeballs in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10 mg/mL gentamicin overnight at room temperature in the dark. They were then incubated in DMEM containing 0.1% trypsin and 70 U/mL collagenase for 60 minutes at 37°C. Culture procedures were undertaken in a sterile laminar flow hood (Klenzaids Bioclean Devices, Mumbai, India), according to previous reports. ^29,35 The retinas were digested with papain/DNase (Papain Dissociation System; Worthington Biochemicals, Lakewood, NJ) for 40 minutes at 37°C and mechanically dissociated, and the cells were released by a series of triturations. Müller cells were easily recognized on the basis of their characteristic bipolar shape. ¹⁵ The enriched populations of morphologically recognizable Müller cells were centrifuged, and the pellets were resuspended in Dulbecco's modified Eagle's medium (DMEM, containing 5 mM glucose, 10% FBS, and 1% penicillin/streptomycin, as described elsewhere ³⁶ ). The cells were allowed to adhere to the culture flasks for approximately 2 hours at 37°C, after which the nonadherent population was removed, and the medium was replaced with fresh DMEM. The cells were maintained at 37°C in 5% CO₂, 95% air with DMEM being renewed approximately every 3 days until confluence (∼10 days), then they were released for subculture with 0.05% trypsin and 0.53 mM EDTA. Each time the culture medium was renewed, the cultures were washed extensively with medium until only a strongly adherent, flat cell population remained. Retinal neurons that initially adhered to the surface of the Müller cells were rinsed off with medium, which was renewed every 10 days in vitro, leaving a monolayer of glial cells. ³⁶ The Müller cells were identified by GS and glia fibrillary acidic protein (GFAP) staining with indirect immunofluorescence and they were used from the second to fourth passages in this study. 

After the cells grew to reach 85% confluence in different plates, to investigate the effect of high glucose on the expression of PEDF, Kir4.1, and GLAST and the uptake of glutamate in retinal Müller cells, the culture medium was changed to DMEM containing 5 mM (normal) or 25 mM (high) glucose ³⁶ without serum. Control incubations containing 20 mM mannitol were run simultaneously to rule out the effect of increased osmolarity. To study the effect of PEDF on the expression of Kir4.1 and GLAST and the uptake of glutamate in retinal Müller cells in high-glucose conditions, 50 or 100 nM recombinant PEDF (BioProducts Inc, Middletown, MD) was added to the Müller cell culture medium. All plates were supplemented with 1% penicillin/streptomycin and incubated at 37°C in a humidified chamber at 95% O₂ and 5% CO₂, for 24 hours. 

To confirm the influence of PEDF on Kir4.1 and GLAST in this study, short interfering RNA (siRNA) against PEDF was used. PEDF siRNA was generated (Silencer siRNA Construction Kit; Ambion, Austin, TX) with the following sequences: anti-PEDF sense, 5′-GGA UUU CUA CUU GGA UGA ATT-3′, and anti-PEDF antisense, 5′-UUC AUC CAA GUA GAA AUC-3′. Transfection was performed with a lipid transfection reagent (siPORT; Ambion) according to the manufacturer's instructions. The siRNA yield was determined by spectrophotometry for each siRNA preparation and amounted to 400 to 650 ng/mL. The scrambled counterpart, with the sequences: anti-PEDF siRNAscr sense, 5′-GAU UGA UGA UUU AGC UGA CTT-3′, and anti-PEDF siRNAscr antisense, 5′-GUC AGC UAA AUC AUC AAU C-3′, served as a negative control. ³⁷  

When the differently treated cells grew to 85% confluence in different plates under normal or high-glucose conditions, expression of PEDF, Kir4.1, and GLAST in retinal Müller cells was tested by Western blot analysis and real-time RT-PCR. A glutamate uptake assay was undertaken to detect the function of GLAST in different treated cells. Intracellular ROS generation and GPx level were measured to determine oxidative stress. Each experiment was repeated in at least four separate cell preparations. 

Western Blot Analysis

Cultured Müller cells with different treatments as described above were homogenized in ice-cold lysis buffer (50 mM Tris-HCl [pH 7.4]; 1% Nonidet P-40; 0.25% Na-deoxycholate; 150 mM NaCl; 1 mM EDTA; 1 mM PMSF; 1 μg/mL each of aprotinin, leupeptin, and pepstatin; 1 mM Na₃VO₄; and 1 mM NaF), and the homogenate was centrifuged at 12,000g for 15 minutes at 4°C, to remove cell debris. Protein concentrations were analyzed by Bradford's technique. ^38,39 Proteins (40 μg) were resolved by polyacrylamide gel and transferred to a polyvinylidene difluoride membrane (Amersham Life Science, West Chester, PA) for electrophoresis, using the techniques described earlier. ⁴⁰ The membranes were blocked by overnight incubation in the wash buffer (50 mM Tris-HCl, 150 mM NaCl, and 0.1% Tween 20 [pH 7.5]) containing 5% fat-free milk. They were then incubated with different primary antibodies (1:1000, polyclonal rabbit anti-GLAST; ADI, San Antonio, TX; 1:200, polyclonal rabbit anti-Kir4.1′ Alomone Laboratories, Jerusalem, Israel; 1:1000, polyclonal rabbit anti-PEDF; Chemicon, Inc., Temecula, CA) for 2 to 4 hours at room temperature. Next, the membranes were washed and incubated with the appropriate horseradish peroxidase-labeled secondary antibody (1:15,000; Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 hour at room temperature, followed by detection with Western blot reagent (Luminol; Santa Cruz Biotechnology, Santa Cruz, CA). To ensure that equal quantities were loaded in each lane, the membranes were blotted with anti-β-actin antibody (Sigma-Aldrich, St. Louis, MO). Immunoreactive bands were quantified by scanning densitometry (Fluor-s Multimager and Quantity One software; Bio-Rad Laboratories, Hercules, CA), and the density of each band was normalized to that of its own β-actin. To avoid biologic variability, extracts were from four separate cell preparations, and to avoid technical variability, each Western blot analysis was performed in triplicate. Data are the averaged values of these replicates with the standard deviation. 

Realtime RT-PCR

Total RNA was extracted from Müller cells with different treatments (TRIzol reagent; Invitrogen Life Technologies, Gaithersburg, MD) and stored at −80°C. A SYBR Green qPCR kit (DyNAmo Flash; Finnzymes Oy, Espoo, Finland) was used according to the manufacturer's instructions. The following primer pairs were used: GLAST sense, 5′-ACTTTGCCTGTCACCTTCCG-3′, and antisense, 5′-ACTGCGTCTTGGTCATTTCG-3′; Kir4.1 sense, 5′-GCA AGA TCT CCC CCT CCG CAG-3′, and antisense, 5′-CAG ACG TTA CTA ATG CGC ACA CT-3′; PEDF sense, 5′-CAGAAGAACCTCAAGAGTGCC-3′, antisense, 5′-CTTCATCCAAGTAGAAATCC-3′; β-actin sense, 5′-CCTCTATGCCAACAC AGTGC-3′ and antisense 5′-GTACTCCTGCTTGCTGATCC-3′. Primers for the genes were used for amplification as described previously. ³⁷ Real-time PCR was performed with a PCR detection system (MyiQTM Single-Color; Bio-Rad Laboratories) by a published method. ^37,41 The cycling conditions were set to an initial denaturation at 95°C for 15 minutes, followed by 40 cycles with denaturation at 94°C for 20 seconds, annealing at 72°C for 30 seconds, and elongation at 72°C for 5 minutes. To verify the specificity of the PCR products, melting curve analyses were performed. The reactions were performed in triplicate for each of four separate cell preparations. The comparative Ct (^ΔΔCt) method was used to obtain quantitative data of relative gene expression, according to the manufacturer's instructions. Values for each gene were normalized to expression levels of β-actin. 

Glutamate Uptake Assay

Cultured Müller cells treated differently, as described above, were prepared. The cells then were washed in PBS and preincubated in Kreb's solution (containing [mM] 119 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 25 NaHCO₃, 2.5 CaCl₂, and 1 MgCl₂) for 30 minutes. Then, the Müller cells with different treatments were exposed to 0.5 μCi/mL of l-[2,3-³H] glutamate (New England Nuclear, Boston, MA) and 10 mM unlabeled glutamate for 60 minutes. The reaction was stopped by washing the cells three times with ice-cold PBS. Müller cells were subsequently lysed in PBS, and small aliquots were removed from each well for the determination of protein content. l-[2,3-³H] glutamate content of the lysates was determined by scintillation counting. Experiments were performed in triplicate. 

Detection of Intracellular Reactive Oxygen Species (ROS) Generation

Cultured Müller cells prepared with different treatments as described above were seeded in 96-well plates and grown to 85% confluence. The generation of intracellular ROS was detected by the dichlorodihydrofluorescein (DCF) method using 5-(and-6)- carboxy-2′,7′-dichlorodihydrofluorescein diacetate (carboxy-H2DCFDA), a cell-permeable indicator for ROS. The cells were gently washed with PBS and incubated with 2 μM carboxy-H₂DCFDA in phenol red–free medium at 37°C for 20 minutes. The medium was discarded, and the cells were washed with PBS. Fluorescence was measured using a spectrofluorometer (LS-5B; Perkin-Elmer, Waltham. MA) with an excitation wavelength of 485 nm and an emission wavelength of 538 nm. 

Assay of GPx Activity

Cultured Müller cells treated differently as described above were prepared. Then cells were washed twice with PBS. One milliliter of 50 mM potassium phosphate buffer with 1 mM EDTA (pH 7.0) was added, and the cells were scraped from the dishes. Cell suspensions were sonicated three times for 5 seconds on ice each time, and the cell sonicates were centrifuged at 10,000g for 20 minutes at 4°C. Cell supernatants were used to determine the activities of GPx. GPx activity was measured with an assay kit (Cayman, Ann Arbor, MI), according to the manufacturer's instructions. The measurement of GPx activity is based on the principle of a coupled reaction with glutathione reductase (GR). The oxidized glutathione (GSSG) formed after reduction of hydroperoxide by GPx is recycled to its reduced state by GR in the presence of NADPH. The oxidation of NADPH is accompanied by a decrease in absorbance at 340 nm. One unit of GPx was defined as the amount of enzyme that catalyzes the oxidation of 1 nanomole of NADPH per minute at 25°C. 

Statistical Analysis

Data are reported as the mean ± SD (SPSS statistical package, ver. 12; SPSS Inc., Chicago, IL). Statistical comparisons were made by unpaired t-test or one-way analysis of variance with post hoc Student-Newman-Keuls multiple comparisons test, for data sets containing more than two groups. Test results were significant at P = 0.05. 

Results

Immunocytochemical Characterization of Cultured Retinal Müller Cells

Under an inverse phase microscope, the cultured Müller cells were observed to have large cell bodies and abundant cytoplasm (Fig. 1A). The body became wider and microfilaments and cytodendrites appeared after three to four passages. We characterized cultured rats Müller cells by their expression of GFAP and GS, as judged by immunocytochemical staining. Cells in this culture system showed positive labeling for GFAP and GS (Figs. 1B, 1C). This immunocytochemical labeling indicates that the cultured cells were Müller cells. 

Figure 1.

 

Inverse phase microscopy shows the cultured Müller cells to have large cell bodies and abundant cytoplasm (A). The body became wider and microfilament and cytodendrite appeared after three to four passages. Retinal Müller cells were identified by their expression of GFAP (B) and glutamine synthetase (GS) (C), as judged by immunocytochemical staining. Bar, 50 μm.

Effect of High Glucose on the Expression of PEDF, Kir4.1, and GLAST and on Glutamate Uptake Activity and Oxidative Stress in Retinal Müller Cells

Protein expression of retinal Müller cells incubated in DMEM containing normal (5 mM glucose) or high (25 mM glucose) concentrations of glucose for 24 hours were tested by Western blot analysis. Decreases in protein expression of approximately 35.65% ± 6.01% (P < 0.01), 39.12% ± 11.51% (P < 0.01), and 47.25% ± 12.03% (P < 0.01) were observed for PEDF, Kir4.1, and GLAST, respectively, in the cells incubated in 25 mM glucose, compared with those incubated in 5 mM glucose (Fig. 2A). The above results were confirmed by real-time RT-PCR (Fig. 2B). In retinal Müller cells cultured with high glucose, mRNAs of PEDF, Kir4.1, and GLAST were all significantly decreased, compared with those incubated in 5 mM glucose. 

Figure 2.

 

Effect of high glucose on expression of PEDF, Kir4.1, and GLAST; glutamate uptake activity; and oxidative stress in retinal Müller cells. (A) PEDF, Kir4.1, and GLAST expression in cells incubated in 5 and 25 mM glucose-containing media for 24 hours was measured by Western blot analysis. Each measurement was performed in triplicate for each of four separate cell preparations. Western blot analysis of PEDF, Kir4.1, and GLAST was normalized by β-actin (mean ± SD, n = 4). (B) Relative mRNA expression of PEDF, Kir4.1, and GLAST was quantified by real-time RT-PCR (mean ± SD, n = 4; error bars, SD). The values obtained in 5-mM glucose treatments were assumed to represent 100% production. (C) A glutamate uptake assay was performed by using the scintillation counting method to determine ³H-glutamate content of the lysates (mean ± SD, n = 4; error bars, SD). (D) The generation of intracellular ROS was detected by the DCF method using carboxy-H₂DCFDA, and GPx activity was measured with an assay kit. The values obtained in 5-mM glucose treatments were assumed to represent 100% production. All the results indicated that high glucose could significantly decrease expression of PEDF, Kir4.1, and GLAST; induce oxidative stress; and impair glutamate uptake in Müller cells. To rule out the effect of increased osmolarity, control incubations containing 20 mM mannitol were run simultaneously. Because of the data, the possible influence of high osmolarity was excluded (5 mM G, 5 mM glucose; 20 mM M, 5 mM glucose+20 mM mannitol; 25 mM G, 25 mM glucose).

We investigated the effect of high glucose on glutamate uptake activity in retinal Müller cells. Data from a glutamate uptake assay (Fig. 2C) showed that high glucose decreased l-[2,3-³H] glutamate uptake activity from 606 ± 69.59 cpm/min/mg protein in control cultures to 415 ± 46.23 cpm/min/mg protein in high-glucose–treated cultures (P < 0.01). 

To determine oxidative stress in high-glucose conditions, ROS generation and GPx activity were measured. As shown in Figure 2D, the generation of intracellular ROS in Müller cells increased significantly, and GPx activity in Müller cells decreased obviously after treatment with 25 mM high glucose compared with 5 mM normal glucose (P < 0.01). 

To exclude the possible influence of high osmolarity on expression of PEDF, Kir4.1, and GLAST and glutamate uptake activity and oxidative stress in retinal Müller cells, control incubations containing 20 mM mannitol were run simultaneously. Because of the data, the possible effects of increased osmolarity were ruled out (Fig. 2). 

Effect of PEDF on Expression of Kir4.1 and GLAST and on Glutamate Uptake Activity and Oxidative Stress in Retinal Müller Cells in High-Glucose Concentrations

To determine whether PEDF can ameliorate the expression of Kir4.1 and GLAST, different concentrations of PEDF (50 or 100 nM) were added to Müller cell cultures in high-glucose conditions and tested by Western blot analysis after 24 hours. The results indicated that PEDF increased expression of Kir4.1 and GLAST significantly, especially at higher concentration of PEDF (Figs. 3A, 3B). Realtime-RT-PCR was also used to confirm these results (Fig. 3C). 

Figure 3.

 

Effect of PEDF on the expression of Kir4.1 and GLAST and on glutamate uptake activity and oxidative stress in retinal Müller cells in high-glucose concentrations. (A) Retinal Müller cells were incubated for 24 hours in 25 mM glucose medium, with different concentrations of PEDF. (B) The data show the absorbance of Kir4.1 and GLAST, adjusted to the expression of the intrinsic protein in each lane (mean ± SD, n = 4; error bars, SD). The values obtained at 0 nM PEDF treatments were considered 100%. (C) Relative mRNA expression of Kir4.1 and GLAST in cells treated as in (A) was quantified by real-time RT-PCR (mean ± SD, n = 4; error bars, SD). The data for treatment in 25 mM glucose without PEDF were assumed to represent 100% production. (D) A glutamate uptake assay was performed, and (E) the generation of intracellular ROS was detected as described in Figure 2. The values obtained at 0 nM PEDF treatments were considered 100%. All the results indicate that PEDF significantly increased the expression of Kir4.1 and GLAST in a dose-dependent manner, starting at 50 nM PEDF concentration. The results also show that PEDF inhibited generation of ROS and ameliorated GPx activity and glutamate uptake activity in Müller cells in high glucose concentrations.

The glutamate uptake assay showed that PEDF treatment could ameliorate glutamate uptake activity in retinal Müller cells in high-glucose conditions, especially at higher concentrations of PEDF (Fig. 3D). The data in Figure 3E show that that PEDF can decrease levels of NOS and ameliorate GPx activity in Müller cells in high-glucose conditions. 

Under normal glucose conditions, 100 nM PEDF did not cause significant changes in expression of Kir4.1 and GLAST in Müller cells. Obvious changes in glutamate uptake activity and oxidative stress in Müller cells in normal glucose were not found after PEDF treatments (Fig. 4). 

Figure 4.

 

Effect of PEDF on expression of Kir4.1 and GLAST and on glutamate uptake activity and oxidative stress in retinal Müller cells in normal glucose concentrations. (A) Retinal Müller cells were incubated for 24 hours in 5 mM glucose medium with 100 nM PEDF. Expression of Kir4.1 and GLAST was measured by Western blot analysis. (B) Relative mRNA expression of Kir4.1 and GLAST was quantified by real-time RT-PCR. The data for treatment in 5 mM glucose without PEDF were assumed to represent 100% production. (C) A glutamate uptake assay was undertaken and (D) the generation of intracellular ROS was detected as described in Figure 2. The values obtained at 0 nM PEDF treatments were considered 100%. The results showed that, in normal glucose conditions, 100 nM PEDF did not cause significant changes in the expression of Kir4.1 and GLAST in Müller cells. Obvious changes in glutamate uptake activity and oxidative stress in Müller cells were not found.

PEDF siRNA Inhibits Expression of PEDF, Kir4.1, and GLAST; Inhibits Glutamate Uptake Activity; and Induces Oxidative Stress in Retinal Müller Cells

To confirm the effect of PEDF on glutamate uptake activity in retinal Müller cells, we used PEDF siRNA to suppress the expression of PEDF mRNA and protein in Müller cells in normal-glucose concentrations. Western blot analysis and real-time RT-PCR were undertaken to measure PEDF, Kir4.1, and GLAST expression. Western blot analysis showed that there were decreases in expression of PEDF, Kir4.1, and GLAST—approximately 48.25% ± 12.11% (P < 0.01), 58.25% ± 14.12% (P < 0.01) or 55.75% ± 7 .13% (P < 0.01), respectively, in Müller cells with PEDF siRNA treatment compared with the protein expression in the control group (Figs. 5A, 5B). Moreover, it was found that when the PEDF gene was silenced, the expression of PEDF mRNA, Kir4.1 mRNA, or GLAST mRNA were decreased (Fig. 5C). From glutamate uptake assay data, we found that when the PEDF gene was silenced in Müller cells in normal glucose concentrations, the glutamate uptake activity of Müller cells was disrupted (Fig. 5D). As shown in Figure 5E, the production of intracellular ROS in Müller cells increased significantly after the PEDF gene was silenced, whereas GPx activity in Müller cells clearly decreased. 

Figure 5.

 

PEDF siRNA inhibits expression of PEDF, Kir4.1, and GLAST; inhibits glutamate uptake activity; and induces oxidative stress in retinal Müller cells. (A) Differently treated retinal Müller cells (siRNA treatment or scrambled siRNA treatment [control]) were incubated in 5 or 25 mM glucose medium for 24 hours. Expression of PEDF, Kir4.1, or GLAST was measured by Western blot analysis. (B) Relative mRNA expression of PEDF, Kir4.1, or GLAST was quantified by real time-RT-PCR. (C) A glutamate uptake assay was undertaken and (D) the generation of intracellular ROS was detected as described in Figure 2. The values obtained for treatment by scrambled siRNA (control) in 5 mM glucose were assumed to represent 100% production. From the data, it was found that PEDF siRNA decreased expression of PEDF, Kir4.1, and GLAST significantly. Simultaneously, downregulation of PEDF inhibited activity of glutamate uptake and induced oxidative stress. It was also found that high-glucose treatment exacerbated the changes observed in PEDF-deficient cells.

To investigate whether high-glucose treatment would exacerbate the changes observed in PEDF-deficient Müller cells, siRNA PEDF-treated Müller cells were incubated in 25 mM glucose for 24 hours simultaneously. The data showed that, in high-glucose conditions, the expression of PEDF, Kir4.1, and GLAST decreased further (Figs. 5A–C). At the same time, glutamate uptake activity and oxidative stress became more destructive (Figs. 5D, 5E). 

Discussion

The results indicate that Müller cells treated with 25 mM high glucose decreased l-[2,3-³H] glutamate uptake activity and PEDF, Kir4.1, and GLAST expression levels and generated higher levels of ROS and lower GPx activity compared with those treated with normal glucose cultures. However, treatment with 50 or 100 nM PEDF together with 25 mM high glucose exposure resulted in a significant increase in l-[2,3-³H] glutamate uptake, Kir4.1, GLAST expression levels, and ameliorated oxidative stress by decreasing generation of ROS and ameliorating GPx activities in Müller cells. Moreover, downregulation of PEDF expression by siRNA in Müller cells in 5 mM normal glucose resulted in significant decreases in glutamate uptake and Kir4.1 and GLAST expression and induced oxidative stress. Another interesting point presented in this study is that high-glucose treatment exacerbated the changes observed in PEDF-deficient Müller cells. 

Several reports indicated that the degenerative processes begin in the neural retina shortly after the onset of diabetes and precede the characteristic vascular changes that prompt the diagnosis of retinopathy. ^{3 –12} Bresnick and Palta ¹⁰ initially proposed that DR is a primary neurosensory disorder. Accumulating evidence has shown that diabetes-induced neuronal injuries are closely related to excessive glutamate levels in the retina. ^{5 –8,13 –15,42 –44} Glutamate is the main excitatory neurotransmitter in the mammalian brain and retina, but it is neurotoxic when present in excessive amounts. Thus, tight control of glutamate levels in the extracellular space is crucial for the prevention of neurotoxicity. It is known that GLAST in Müller cells is mainly responsible for maintaining low synaptic glutamate levels in the normal retina. Several authors have suggested that the activity of the Müller cell glutamate transporter is decreased in experimental diabetes. ^7,14,22,24 The putative mechanism with regard to the malfunction of GLAST in Müller cells in DR involves oxidative stress induced by DR. ⁷ Although the detailed mechanism remains unknown, it may be of critical importance in neuroprotection to remove excess glutamate from the extracellular space in the retina by ameliorating the function of GLAST. 

Some investigators have shown that oxidative stress causes dysfunction of GLAST in Müller cells due to abnormal membrane depolarization caused by functional inactivation or downregulation of the inwardly rectifying potassium channel Kir4.1. ^{20 –22} Data of this study showed that expression levels of Kir4.1 decreased in high-glucose conditions. Simultaneously, GLAST expression levels decreased and activity of glutamate uptake was compromised. These results were consistent with those in previous reports. ^{20 –22} It is critically important to find an effective agent to protect the function of GLAST under oxidative conditions. PEDF is a secreted 50-kDa glycoprotein that was first identified in conditioned media of cultured fetal human retinal pigment epithelial (RPE) cells. ⁴⁵ In recent years, accumulating evidence has suggested that PEDF protects retinal neurons from light-induced damage, oxidative stress, and glutamate excitotoxicity. ^{26,30,31,46,47} Results in recent reports also suggest that PEDF can act as an antioxidative agent to ameliorate some pathologic changes in DR. ^{31 –33} Previous studies have shown that PEDF is involved in the pathogenesis of DR by regulating cytokine secretion in Müller cells, including VEGF, tumor necrosis factor (TNF)-α, and PEDF itself. ^36–28 It is proposed that PEDF plays an important role in regulating the function of GLAST in Müller cells, by acting as an antioxidative agent. From the results of this study, we found that high glucose could decrease PEDF expression levels, which is consistent with our previous results. ³⁷ Treatment with PEDF, together with high glucose exposure, results in a significant increase in glutamate uptake and in Kir4.1 and GLAST expression levels in Müller cells. These results imply that lower PEDF levels are related to downregulation of Kir4.1 and compromise of GLAST in Müller cells. To test this hypothesis, the siRNA PEDF technique was used, and the data showed that downregulation of PEDF expression by siRNA in Müller cells in 5 mM normal glucose resulted in significant decreases in glutamate uptake and Kir4.1 and GLAST expression. It was also found that high-glucose treatment exacerbated the changes observed in PEDF-deficient Müller cells. 

Oxidative stress is known to be involved in the dysfunction of GLAST in Müller cells. Consequently, ROS production was selected as a measure of oxidative stress. ^{31 –33} High glucose increased the levels of ROS in Müller cells, but the effects were inhibited by PEDF. GPx activity decreased when induced by high glucose; however, this effect was counteracted by PEDF treatment, at least partially. GPx is a primary antioxidant enzyme responsible for peroxide detoxification in mammalian cells. Overexpression of GPx protects various types of cells against ROS-induced apoptosis, whereas knockout of GPx enhances susceptibility to ROS-induced cytotoxicity. ³³ Thus, PEDF may restore GPx activity, which results in inhibition of the effect of ROS. To further confirm the role of PEDF, PEDF siRNA was used. The results imply that downregulation of PEDF induced higher levels of ROS and compromised GPx activity. Although ROS is an important product of oxidative stress, its role in suppression of glutamate uptake in Müller cells remains unclear and should be further investigated. 

Our study demonstrates for the first time that the expression of PEDF, Kir4.1, and GLAST in retinal Müller cells is unbalanced in high glucose concentrations, with abnormal expression closely related to oxidative stress. Decreased expression of PEDF induces downregulation of Kir4.1 levels in Müller cells, causing membrane depolarization of Müller cells and dysfunction of GLAST. Our results suggest that PEDF may act as an antioxidative agent against the decrease in Kir4.1 expression and compromise of GLAST activity in retinal Müller cells in DR, by ameliorating GPx activity. Thus, we propose that retinal PEDF is responsible for the task of balancing an antioxidative state against the need to maintain the protection of retinal neuronal and glial cells. 

Footnotes

 Supported by National Nature Science Foundation of China Grants 81170860 and 81100682; Shanghai Municipal Education Committee Project 10YZ38; Shanghai Nature Science Foundation Grant 11ZR1422000; and Shanghai Leading academic Discipline Project Grant S30205.

Footnotes

 Disclosure: B. Xie, None; Q. Jiao, None; Y. Cheng, None; Y. Zhong, None; X. Shen, None