Pigment epithelium-derived factor (PEDF) is a potent angiogenic inhibitor. Previous studies have shown that decreased ocular levels of PEDF are associated with diabetic retinopathy. However, the implication of PEDF expression in diabetic nephropathy has not been revealed. In the present study, we demonstrated for the first time that the expression of PEDF was decreased at both the mRNA and protein levels in the kidney of diabetic rats, whereas transforming growth factor-β (TGF-β) and fibronectin levels were increased in the same diabetic kidneys. As shown by immunohistochemistry, the decrease of PEDF expression occurs primarily in the glomeruli. In vitro studies showed that high concentrations of glucose significantly decreased PEDF secretion in primary human glomerular mesangial cells (HMCs), suggesting that hyperglycemia is a direct cause of the PEDF decrease in the kidney. Toward the function of PEDF, we showed that PEDF blocked the high-glucose-induced overexpression of TGF-β, a major pathogenic factor in diabetic nephropathy, and fibronectin in primary HMCs, suggesting that PEDF may function as an endogenous inhibitor of TGF-β expression and fibronectin production in glomeruli. Therefore, decreased expression of PEDF in diabetic kidneys may contribute to extracellular matrix overproduction and the development of diabetic nephropathy.

Diabetic nephropathy is one of the most important microvascular complications of diabetes and occurs in 30–40% of diabetic patients (1,2). The early changes in diabetic nephropathy are characterized by the thickening of glomerular basement membrane and expanded extracellular matrix (ECM), leading to glomerular hyperfiltration and microalbuminuria (1,3). Although intensified control of hyperglycemia, blood pressure, and hyperlipidemia reduces the risks of diabetic nephropathy, it does not sufficiently prevent the diabetic patients with microalbuminuria from progression to devastating overt diabetic nephropathy, a leading cause of end-stage renal diseases (2,4,5). The exact pathogenesis of diabetic nephropathy remains largely unknown.

Several growth factors have been suggested to be involved in the pathogenesis of diabetic nephropathy—most importantly, transforming growth factor-β (TGF-β) and vascular endothelial growth factor (VEGF) (6,7). TGF-β has been recognized as a modulator of ECM formation. Overexpression of TGF-β in diabetic glomeruli is believed to contribute to the matrix accumulation by increasing synthesis and decreasing degradation of extracellular proteins such as fibronectin (811). VEGF is a major angiogenic factor and is also referred to as the vascular permeability factor, based on its ability to induce vascular hyperpermeability (12,13). Accumulating evidence indicates that VEGF and TGF-β are key pathogenic factors in early stages of diabetic nephropathy (1419). Serum and urinary TGF-β levels have been found to correlate with the severity of microalbuminuria (20,21). Therefore, the increase of the systemic TGF-β level has been suggested as a marker for diabetic nephropathy (22).

Pigment epithelium-derived factor (PEDF) is a multifunctional serine proteinase inhibitor (23). Although PEDF was first identified in the eye, it is widely distributed in a variety of organs, such as brain, spinal cord, liver, heart, placenta, bone, pancreas, and prostate (2426). As a potent anti-angiogenic factor, PEDF plays an important role in maintaining the homeostasis of the ocular vascular system (2527). A recent study has shown that PEDF also reduces VEGF-induced vascular hyperpermeability (28). In diabetic retinopathy, decreased endogenous PEDF and increased VEGF levels break the delicate balance between angiogenic stimulators and inhibitors, resulting in retinal vascular hyperpermeability and retinal neovascularization (2931). Systemic injection of PEDF or virus-mediated delivery of PEDF significantly inhibits retinal angiogenesis (32,33). Recent studies demonstrated that downregulation of PEDF may also contribute to the development of renal and prostate cancer (34,35).

In contrast to the extensive studies of PEDF in ocular tissues, the function of PEDF in the kidney has not been studied. In the present study, we have determined the PEDF expression in the normal and diabetic rat kidneys and explored the possible association between PEDF levels and pathogenesis of diabetic nephropathy.

Brown Norway (BN) rats were purchased from Charles River Laboratories (Wilmington, MA). Care, use, and treatment of all animals in this study were in strict agreement with the guidelines set forth in the Care and Use of Laboratory Animals by the University of Oklahoma.

Induction of experimental diabetes.

BN rats (8 weeks of age) were used in the experiment. Diabetes was induced by an intraperitoneal injection of streptozotocin (STZ) (50 mg/kg in 10 mmol/l of citrate buffer, pH 4.5) (Sigma, St. Louis, MO) into BN rats after an overnight fast. Control rats received an injection of citrate buffer alone. Blood glucose levels were measured at 24 h after the injection and monitored every 3 days thereafter. Only the animals with blood glucose concentrations >350 mg/dl were considered diabetic.

Cell culture.

Primary human glomerular mesangial cells (HMCs) were purchased from Cambrex Bio Science Walkersville (Walkersville, MD). The cells were cultured in mesangial cell basal medium (Cambrex) with 0.1% GA-1000 (Cambrex) and 5% fetal bovine serum (FBS) at 37°C in a humidified 5% CO2 atmosphere. Cells of passages from 6 to 10 were used in the experiments. After reaching 80% confluence, cells were exposed to medium with 5% FBS for 12 h before the treatments with high glucose or PEDF.

Isolation of glomeruli.

Rats were deeply anesthetized, and the kidneys were immediately removed. The cortex was excised, cut into fine fragments, and homogenized. After being passed through consecutive stainless steel screens of 150- and 75-μm pore size, the glomeruli were suspended in 1 × PBS and collected by centrifuge at 2,000g for 3 min.

Quantitative real-time RT-PCR.

The total RNA was isolated from tissues using the RNeasy Mini-Isolation Kit following the manufacturer’s protocol (Qiagen, Santa Clarita, CA). Primers (PEDF-F, 5′-aagtcatatgggaccaggccc-3′; PEDF-R, 5-′ttacccactgccccttgaagt-3′) were designed from mRNA sequences spanning >1-kb introns using the Primer3 software (http://frodo.wi.mit.edu/cgi-bin/primer3_www.cgi). RT reaction used 1.0 μg total RNA, oligo-dT primer, and MuLV reverse transcriptase in a final volume of 20 μl and conducted at 42°C for 60 min, followed by a denaturation at 95°C for 5 min. The real-time PCR used 1 μl of the RT product and 3 pmol of primers and was performed using a GeneAmp RNA PCR kit and SYBR Green PCR Master Mix (Applied Biosystems). The PCR mix was denatured at 95°C for 10 min, followed by 40 cycles of melting at 95°C for 15 s and elongation at 60°C for 60 s. Fluorescence changes were monitored after each cycle. Amplicon size and reaction specificity were confirmed by 2.5% agarose gel electrophoresis. All reactions were performed in triplicate. The average CT (threshold cycle) of fluorescence units was used to analyze the mRNA levels. The PEDF mRNA levels were normalized by 18s ribosomal RNA levels. Quantification was calculated as follows: mRNA levels (percent of control) = 2Δ(ΔCT), with ΔCT = CT, PEDFCT, 18s and Δ(ΔCT) = ΔCT, normal sample − ΔCT, STZ-induced diabetic sample.

Western blot of PEDF.

Western blot analysis was performed using a monoclonal anti-PEDF antibody (Chemicon, Temecula, CA) as described previously (31).

Immunohistochemistry study.

Immunohistochemistry was performed on frozen tissue sections. Briefly, the sections were blocked with solution containing 3% BSA (Sigma) and 5% rabbit serum (Jackson Immunoresearch). After incubation with 1:800 dilution of an anti-PEDF antibody (Chemicon) or an anti-synaptopodin antibody (Biodesign, ME) for 1 h, the sections were thoroughly washed and incubated with 1:200 fluorescein isothiocyanate–conjugated rabbit anti-mouse antibody (Jackson Immunoresearch) for 1 h. After extensive washing, the sections were visualized and photographed under a fluorescent microscope (Olympus, Humburg, Germany) and confocal microscope (Leica, Mannheim, Germany).

PEDF enzyme-linked immunosorbent assay.

The protein concentration was measured with the BioRad DC protein assay (BioRad Laboratories, Hercules, CA). The amounts of PEDF in the tissue extracts, serum, and conditioned medium were determined using a commercial enzyme-linked immunosorbent assay (ELISA) kit specific for PEDF (Chemicon) according to the manufacturer’s instructions. Briefly, the tissue samples were homogenized in 1 × PBS and centrifuged (TL; Beckman) at 50,000 rpm (rotor type: TLA 100.3) for 20 min at 4°C. The serum or tissue extract was incubated with 8 mol/l urea (Sigma) for 1 h and diluted 1:200 before being applied to the plate. After incubation at 37°C for 1 h and the extensive washing, the plate was incubated with 100 μl of a biotinylated mouse anti-PEDF antibody for 1 h, followed by incubation with 100 μl streptavidin peroxidase conjugate for 1 h. After the addition of TMB/E (3,3′,5,5′-tetramethylbenzidine) for 5–10 min, the plate was read immediately at 450 nm by a Wallac-Victor3 1420 microplate reader (Perkin-Elmer Wallac). For standardization, the PEDF concentration was normalized by the total protein concentration in the samples.

ELISA specific for TGF-β.

TGF-β protein was quantified using the commercial Quantikine TGF-β1 ELISA Kit (R&D Systems, Minneapolis, MN). Briefly, the tissue extracts were activated with 1 N HCl for 10 min, followed by neutralization with 1.2 N NaOH. The activated samples were applied to the plate precoated with soluble type II receptor and incubated at room temperature for 3 h. After extensive washing, horseradish peroxidase–conjugated anti–TGF-β antibody was added and incubated for another 1.5 h. Then the chromogen was added, and the plate was read at 450 nm. The results were expressed as picograms per milligram of total protein.

ELISA specific for fibronectin.

Fibronectin protein was quantified by competitive sandwich ELISA (Assaypro, Winfield, MO). Briefly, samples were diluted and applied to the plate coated with anti-fibronectin antibody, and the same amount of biotin-labeled fibronectin was immediately added to the wells. After incubation for 1.5 h and extensive washing, the horseradish peroxidase–conjugated streptavidin was added to the wells and incubated for 30 min. Then the chromogen was added and the plate was read at 450 nm. The results were expressed as micrograms per milligram of total protein.

Cell proliferation assay.

The tetrazolium dye–reduction assay (MTT; 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide; Sigma) was used to determine cell survival and proliferation rate according to the manufacturer’s protocol. Briefly, the primary HMCs were seeded in 12-well plates at a density of 5 × 104 cells/well 24 h before the treatments. Then the growth medium was replaced by a medium containing 1% FBS with PEDF (concentrations of 2.5–160 nmol/l). After incubation for 72 h, cells were washed with PBS, and 100 μl/well MTT solution was added and incubated at 37°C for 4 h. The formazan crystals that formed were dissolved by incubation with DMSO (1 ml/well) overnight. Absorption was measured at 550 nm, and the number of viable cells was calculated according to the standard curve. Experiments were performed in triplicate.

Evaluation of rat microalbuminuria.

The 24-h urine collected from each diabetic rat and age-matched control was centrifuged at 2,000g for 10 min. The concentration of albumin in the supernatant was measured by ELISA according to the manufacturer’s protocol (Bethyl Laboratories, Montgomery, TX). The total amount of albumin in 24-h urine was calculated accordingly.

Statistical analysis.

Statistical analysis used the Student’s t test and ANOVA with post hoc test. Statistical difference was considered significant at a P value of <0.05.

High-level expression of PEDF in the kidney of normal rats.

PEDF was recently found to express in the kidney as well as in other organs, but its expression levels and cellular localization in the kidney have not been determined previously (35). Because the liver is regarded as the major source of systemic PEDF (36,37) and the retina is a well-known site of PEDF expression and function, we first compared PEDF levels in the kidney with those in the liver and retina. Western blot analysis showed that PEDF in the kidney was at a level comparable to that in the liver and was much higher than that in the retina (Fig. 1A). Quantitative analysis using ELISA confirmed the results from Western blot analysis (Fig. 1B). A significant difference was observed between the PEDF levels in the liver, kidney, and retina (F = 194.41, P < 0.0001). There is no difference between the PEDF levels in the kidney and those in the liver (249.21 ± 34.45 vs. 221.19 ± 40.38 ng/mg total protein, P > 0.05, n = 6). The PEDF level in the retina (51.21 ± 13.30 ng/mg total protein) was significantly lower than that in the kidney and liver (P < 0.01, n = 6). These results suggested that the kidney expresses high levels of PEDF.

High levels of PEDF in glomeruli.

To determine the cellular localization of PEDF in the kidney, we compared the PEDF expression in the cortex and medulla in normal rat kidney. Western blot analysis showed that the PEDF levels in renal cortex were substantially higher than those in the medulla (Fig. 2A). PEDF ELISA confirmed that the PEDF level in the cortex was 324.98 ± 51.29 ng/mg total protein, significantly higher than that in the medulla (173.58 ± 5.79 ng/mg total protein, P < 0.01, n = 6) (Fig. 2B). Immunohistochemistry using a monoclonal antibody specific for PEDF showed that PEDF was predominantly expressed in the glomeruli (Fig. 2C-a), along the parietal glomerular capsule and basement membrane (Fig. 2C-b). In the medulla, the PEDF signal was also detected at lower levels in the tubular basement membrane and interstitial tissue (Fig. 2C-c). To further identify the cellular origin of PEDF, we immunolabeled PEDF and synaptopodin, which is accepted as a podocyte marker, in the consecutive sections of the same kidney. The results showed that the distribution pattern of synaptopodin and PEDF are similar but not identical. PEDF (Fig. 2D-a), but not synaptopodin (Fig. 2D-b), was found on the parietal glomerular capsule.

Decreased expression of PEDF in the kidney of STZ-induced diabetic rats.

BN rats with STZ-induced diabetes developed polyuria and microalbuminuria at 1–6 weeks after the onset of diabetes, reflecting the impaired function of the glomeruli (Table 1). To determine if the PEDF expression change is implicated in diabetic nephropathy, we compared the PEDF levels in the kidney and serum in STZ-induced diabetic rats with that in age-matched control animals. Both Western blot analysis and specific ELISA demonstrated that the PEDF protein levels were significantly decreased in diabetic cortex and medulla, when compared with that in the nondiabetic control animals (Fig. 3A and B). The decreased PEDF expression was also detected at the mRNA level in the diabetic cortex and medulla (Fig. 3C). In the same tissue samples, however, TGF-β and fibronectin levels were significantly increased in the diabetic kidneys (Fig. 3D and E).

When compared with the age-matched nondiabetic control animals, the PEDF levels in the serum were significantly decreased at the late stage of diabetes (3.54 ± 0.83 ng/mg total protein in 12-week diabetic rats vs. 8.35 ± 1.13 ng/mg in age-matched normal rats, n = 4, P < 0.01) but not at the early stage in diabetic rats (7.00 ± 2.12 ng/mg total protein in 3-week diabetic rats vs. 7.08 ± 1.17 ng/mg in normal rats, n = 4, P > 0.1), suggesting that the PEDF decrease in diabetic kidney occurs before the decrease in serum PEDF levels.

An immunohistochemistry study was performed on kidney sections from rats diabetic for 2 weeks, 4 weeks, and 6 months and nondiabetic controls. The average body weight of diabetic rats was 20–30% lower than that of age-matched nondiabetic control animals. The results showed the decrease of PEDF in the diabetic kidneys at both the early stage (2 and 4 weeks after diabetes onset; Fig. 4A and D) and late stage (6 months after diabetes onset; Fig. 4B) of diabetes. Moreover, the decrease of PEDF expression in the diabetic kidney was primarily observed in the glomeruli (Fig. 4A, B, and D), whereas the PEDF expression was not affected in the medulla of the same diabetic kidney (Fig. 4B). Western blot of PEDF in the isolated glomeruli confirmed that PEDF levels decreased in diabetic glomeruli compared with that in nondiabetic controls. At 2 weeks after diabetes onset, PEDF was decreased dramatically, whereas no detectable podocyte loss was observed (Fig. 4D), indicating that the decrease of PEDF levels in the early stage of diabetic nephropathy is not a result of podocyte loss.

High glucose decreased PEDF secretion in primary HMCs.

To reveal the cause of decreased expression of PEDF in diabetic kidney, we determined the effect of high glucose on the expression of PEDF in primary HMCs. HMCs were cultured with different concentrations of d-glucose (Sigma) for 96 h with the same concentrations of d-mannitol as osmotic controls. The PEDF secreted into the medium was quantified by ELISA. The results showed that high glucose significantly decreased PEDF secretion in HMCs in a dose-dependent manner (F = 14.6, P = 0.013), whereas the mannitol had no significant effects on PEDF secretion (F = 0.96, P = 0.45) (Fig. 5).

PEDF blocked high-glucose–induced TGF-β overexpression in primary HMCs.

Because TGF-β is recognized as one of the major mediators of the proliferation of mesangial cells and the overproduction of ECM in diabetic nephropathy, we also examined the effects of PEDF on TGF-β secretion by HMCs. Under high-glucose conditions (30 mmol/l), the TGF-β secretion was significantly increased, compared with the normal glucose and mannitol controls (Fig. 6A). PEDF at concentrations of 40–160 nmol/l significantly downregulated TGF-β expression in a dose-dependent manner (F = 40.5, P < 0.0001). The inhibitory effects of PEDF on TGF-β expression occurred at 24 h and lasted for at least 48 h (Fig. 6B).

PEDF blocked high-glucose-induced fibronectin secretion in primary HMCs.

In the early stage of diabetic nephropathy, the overproduction of ECM proteins, such as fibronectin and collagen, is a major causative factor responsible for glomerular hyperfiltration (38). In primary HMCs, exposure to high glucose (30 mmol/l) for 48 h led to significant increases of fibronectin secretion, compared with mannitol control (Fig. 7A). At low doses (5–40 nmol/l), PEDF decreased fibronectin secretion in a dose-dependent manner in HMCs cultured in the high-glucose medium (F = 39.98, P < 0.0001) (Fig. 7A). At 24 and 48 h after the addition of 40 nmol/l PEDF, the fibronectin secretion was decreased to 23 and 18% of control, respectively (Fig. 7B).

PEDF does not affect the growth of HMCs.

The viability of glomerular mesangial cells and enlargement of the kidney are known as the major pathological changes in the early stage of diabetic nephropathy. To explore whether the decrease of PEDF expression in the glomeruli contributes to the pathogenesis of diabetic nephropathy, we studied the effects of PEDF and high glucose on mesangial cell proliferation using primary HMCs. The results (data not shown) showed that neither high glucose nor PEDF of the doses from 2.5 to 160 nmol/l altered viable cell numbers in HMCs, suggesting that PEDF’s effects on downregulation of fibronectin and TGF-β are not through the inhibition of mesangial cell proliferation.

PEDF is a 50-kDa glycoprotein initially identified in human retinal pigment epithelial (RPE) cells. Its function as a neurotrophic factor and angiogenic inhibitor has been well studied in ocular tissues (2327,2931). Its implication in diabetic retinopathy has been established (25,26,2931). This study revealed for the first time that PEDF may function as an endogenous inhibitor of TGF-β in the kidney, and decreased PEDF expression in diabetic kidney may contribute to the development of diabetic nephropathy.

Previous studies have shown that PEDF is present in a variety of ocular tissues, such as the interphotoreceptor matrix and ganglion cell layer of the retina, epithelium of the cornea, and ciliary epithelium (23,25,39). Expression of PEDF was also found in human brain and spinal cord of the neural system and several nonneural tissues, including the liver, placenta, heart, and skeletal muscle, suggesting that PEDF’s function may not be limited to ocular tissues (25). Recently, Abramson et al. (35) detected the expression of PEDF in the murine kidney. The present study confirmed the expression of PEDF in rat kidney at both the mRNA and protein levels. Moreover, we quantitatively compared the PEDF levels in the rat kidney, liver, and retina. Surprisingly, our results showed that PEDF levels in the kidney are as high as those in the liver, which is considered the major source of systemic PEDF. PEDF levels in the kidney and liver are much higher than those in the retina. The high level of PEDF in the kidney underscores its significance for renal functions.

The cortex and medulla have different structure and function in the kidney. The immunohistochemical analysis demonstrated that PEDF is predominantly present at the glomerular capsule and basement membrane in the cortex, whereas the PEDF signal was relatively weaker at the tubular basement membrane and interstitial tissue in the medulla. This finding suggests a possible role of PEDF in the regulation of glomerular functions. This cellular localization of PEDF is different from that in Abramson’s report, which mentioned that PEDF was mainly expressed in tubular epithelial cells but not in the glomeruli of mice kidney. It is not clear what causes the disparity between our results and those of Abramson et al. in respect to the cellular localization of PEDF in the kidney, because the figure showing the PEDF signal in the glomeruli was not presented in their article (35). Different species used in our studies and theirs may be a possible reason for the disparity.

Functional studies have demonstrated that PEDF is a multifaceted factor with potent antiangiogenesis activity and neuroprotective function (25). PEDF inhibits endothelial cell migration induced by VEGF and fibroblast growth factor (26,27). A recent study showed that in PEDF gene knockout mice, the microvascular density in the kidney is significantly increased compared with the wild-type mice, suggesting that PEDF may play a role in the regulation of renal vasculature development and maintenance of renal homeostasis (35).

Previous studies from both diabetic patients and animal models have demonstrated that decreased PEDF levels are involved in diabetic retinopathy (29,30,40,41). Because of the close relationship between diabetic nephropathy and diabetic retinopathy, we proceeded to determine in the present study whether PEDF levels in the kidney are also decreased in a diabetic animal model. The results showed that the PEDF expression is significantly decreased in the kidney at both the mRNA and protein levels in STZ-induced diabetic rats, which have exhibited early diabetic nephropathy changes including albuminuria and polyuria. Moreover, the PEDF decrease was mainly observed in the glomeruli, whereas the PEDF levels in the tubular region are less affected. The location of the PEDF changes in diabetic kidney seems to suggest a possible association of the decreased PEDF expression with hyperfiltration of glomeruli and microalbuminuria in diabetic rats.

Diabetes is a complicated metabolic disorder and involves multiple changes in vivo. To identify the causative factor responsible for the PEDF decrease in diabetic glomeruli, we first determined the effect of high glucose, a major change in diabetes, on PEDF expression in primary HMCs. The results showed that high glucose significantly decreases PEDF expression while increasing the expression of TGF-β and fibronectin in HMCs, indicating that the decreased expression of PEDF in diabetic kidney may be ascribed to the direct effects of hyperglycemia. It remains to be determined how high glucose downregulates PEDF expression in HMCs.

TGF-β is a well-studied pathogenic factor of diabetic nephropathy (8,9,14). TGF-β is known to be upregulated in diabetic kidney, which contributes to the proliferation of mesangial cell and ECM production, the major pathological changes in early diabetic nephropathy (8,11,16,42). Several potential therapeutic agents exert their effects through blockade of renal TGF-β overexpression/function in diabetic kidney (9,42,43). To further elucidate the function of PEDF in diabetic kidney, we investigated the effects of PEDF on TGF-β secretion in HMCs. The results showed that PEDF significantly blocked the high-glucose–induced TGF-β overexpression, suggesting that PEDF may act as an endogenous inhibitor of TGF-β expression via a paracrine or autocrine regulation in normal kidney. This finding also suggests that the decreased PEDF levels in diabetic kidney may be responsible, at least in part, for the TGF-β overexpression.

One of the pathological functions of TGF-β in diabetic nephropathy is to promote the overproduction of ECM by mesangial cells, which is also closely correlated with microalbuminuria (1,2). Because mesangial cells are the major producer of ECM, we used primary HMCs as a model to determine if PEDF also blocks the function of TGF-β in the induction of ECM production secretion. The results showed that high glucose (30 mmol/l) significantly increased fibronectin secretion. PEDF blocks the high-glucose–induced fibronectin overproduction in a dose-dependent manner. Further, we tested whether the effect of PEDF on fibronectin production was through the inhibition of mesangial cell proliferation. The results showed that PEDF had no effect on mesangial cell growth rate. These results suggest that PEDF blocks ECM secretion from mesangial cells without interfering with the cell proliferation.

In summary, the present study demonstrated for the first time that PEDF expression is implicated in diabetic nephropathy. PEDF may play an important role in preventing mesangial ECM overproduction and pathological growth factor upregulation in the kidney. The decreased expression or the dysfunction of PEDF may be involved in the pathogenesis of diabetic nephropathy. Therefore, PEDF may have therapeutic potential in the treatment of diabetic nephropathy.

FIG. 1.

Quantitative comparison of PEDF levels in the kidney, liver, and retina of normal rats. A: Western blot analysis of PEDF: The tissue samples were obtained from 9-week-old BN rats. Equal amounts (50 μg) of total protein from each sample were blotted with a specific anti-PEDF antibody. The same membrane was stripped and reblotted with an antibody specific to β-actin. B: Quantitative analysis of PEDF by ELISA: PEDF was quantified by ELISA and normalized by the total protein concentration. Average PEDF levels were expressed as nanograms per milligram of total protein in the tissue (means ± SD, n = 6).

FIG. 1.

Quantitative comparison of PEDF levels in the kidney, liver, and retina of normal rats. A: Western blot analysis of PEDF: The tissue samples were obtained from 9-week-old BN rats. Equal amounts (50 μg) of total protein from each sample were blotted with a specific anti-PEDF antibody. The same membrane was stripped and reblotted with an antibody specific to β-actin. B: Quantitative analysis of PEDF by ELISA: PEDF was quantified by ELISA and normalized by the total protein concentration. Average PEDF levels were expressed as nanograms per milligram of total protein in the tissue (means ± SD, n = 6).

FIG. 2.

Localization of PEDF in the renal tissues. A: Western blot analysis of PEDF. The cortex and medulla were carefully dissected from 9-week-old BN rats. Total protein (50 μg) from each sample was blotted with a specific anti-PEDF antibody. The same membrane was stripped and re-blotted with an antibody specific to β-actin. B: Quantitative analysis of PEDF by ELISA. The protein levels of PEDF were quantified by ELISA and normalized by the total protein concentration. Average PEDF levels were expressed as nanograms per milligram of protein in the tissue (means ± SD, n = 6). C: Immunohistochemistry of PEDF in normal rat kidney. In the cortex (a, 200×, and b, 600×), PEDF signal was mainly detected in the glomeruli along the parietal glomerular capsule and basement membrane. In the medulla (c, 200×), PEDF signal was observed at the tubular basement membrane and interstitial tissue, but was much weaker than that in the glomeruli. D: Immunohistochemistry in normal rat glomeruli showing the similar but not identical patterns of PEDF (a, 400×) and synaptopodin (b, 400×) signals. The arrowhead indicates the PEDF signals at the parietal capsule of glomeruli.

FIG. 2.

Localization of PEDF in the renal tissues. A: Western blot analysis of PEDF. The cortex and medulla were carefully dissected from 9-week-old BN rats. Total protein (50 μg) from each sample was blotted with a specific anti-PEDF antibody. The same membrane was stripped and re-blotted with an antibody specific to β-actin. B: Quantitative analysis of PEDF by ELISA. The protein levels of PEDF were quantified by ELISA and normalized by the total protein concentration. Average PEDF levels were expressed as nanograms per milligram of protein in the tissue (means ± SD, n = 6). C: Immunohistochemistry of PEDF in normal rat kidney. In the cortex (a, 200×, and b, 600×), PEDF signal was mainly detected in the glomeruli along the parietal glomerular capsule and basement membrane. In the medulla (c, 200×), PEDF signal was observed at the tubular basement membrane and interstitial tissue, but was much weaker than that in the glomeruli. D: Immunohistochemistry in normal rat glomeruli showing the similar but not identical patterns of PEDF (a, 400×) and synaptopodin (b, 400×) signals. The arrowhead indicates the PEDF signals at the parietal capsule of glomeruli.

FIG. 3.

Decreased expression of PEDF in the kidney of diabetic rats. A: Western blot analysis of PEDF in the kidney. Equal amounts (50 μg) of kidney proteins from diabetic rats and normal controls were used for PEDF Western blot analysis. B: Quantitative analysis of PEDF in the cortex and medulla by ELISA. The cortex and medulla were dissected from the kidneys of rats with diabetes for 6 weeks and age-matched normal controls. PEDF protein in the tissue extract was quantified by ELISA and normalized by the total protein concentration. The PEDF level was expressed as nanograms per milligram of protein in the tissue (means ± SD, n = 4). C: Quantification of the PEDF mRNA in the cortex and medulla by real-time RT-PCR. The total RNA was isolated from the tissue. The mRNA level of PEDF was determined by real-time RT-PCR and normalized by 18s RNA levels. The average mRNA level was expressed as a percentage of control (means ± SD, n = 4). Protein levels of TGF-β (D) and fibronectin (E) in the kidney extracts from rats with diabetes for 6 weeks and age-matched controls were quantified by specific ELISA and normalized by the total protein (means ± SD, n = 4).

FIG. 3.

Decreased expression of PEDF in the kidney of diabetic rats. A: Western blot analysis of PEDF in the kidney. Equal amounts (50 μg) of kidney proteins from diabetic rats and normal controls were used for PEDF Western blot analysis. B: Quantitative analysis of PEDF in the cortex and medulla by ELISA. The cortex and medulla were dissected from the kidneys of rats with diabetes for 6 weeks and age-matched normal controls. PEDF protein in the tissue extract was quantified by ELISA and normalized by the total protein concentration. The PEDF level was expressed as nanograms per milligram of protein in the tissue (means ± SD, n = 4). C: Quantification of the PEDF mRNA in the cortex and medulla by real-time RT-PCR. The total RNA was isolated from the tissue. The mRNA level of PEDF was determined by real-time RT-PCR and normalized by 18s RNA levels. The average mRNA level was expressed as a percentage of control (means ± SD, n = 4). Protein levels of TGF-β (D) and fibronectin (E) in the kidney extracts from rats with diabetes for 6 weeks and age-matched controls were quantified by specific ELISA and normalized by the total protein (means ± SD, n = 4).

FIG. 4.

Immunohistochemistry analysis of PEDF in the kidney of diabetic rats. A: Immunostaining for PEDF in 9-week-old normal rat kidneys (b) and diabetic rat kidneys after diabetes onset for 2 weeks (d) and 4 weeks (f). The panels (a, c, and e) represent phase contrast for the same field of b, d, and f, respectively. Magnification: 100×. Arrows indicate the locations of glomeruli. B: Immunostaining for PEDF in the kidney of 8-month-old normal rat (a and b) and age-matched diabetic rats 6 months after the onset of diabetes (c and d). a and c: The glomeruli (600×); b and d: the medulla (400×). The results showed that in the normal rat cortex, PEDF was highly expressed in the glomeruli (A-b). In diabetic rat kidney, dramatic decreases of PEDF expression were observed in the glomeruli (A-d, A-f, and B-c) but not in the medulla (B-d). C: Western blot analysis of PEDF in isolated glomeruli from rats with diabetes for 8 weeks and age-matched nondiabetic controls. Total protein (50 μg) from each sample was blotted with a specific anti-PEDF antibody. The same membrane was stripped and reblotted with an antibody specific to β-actin. D: Immunohistochemistry of PEDF and synaptopodin in the kidneys of rats with diabetes for 2 weeks and nondiabetic controls. a and c: PEDF staining; b and d: synaptopodin staining. Compared with the staining in normal kidneys (a and b), the PEDF expression was dramatically decreased (c), whereas no detectable podocyte loss (d) was observed in the same diabetic kidney (200×).

FIG. 4.

Immunohistochemistry analysis of PEDF in the kidney of diabetic rats. A: Immunostaining for PEDF in 9-week-old normal rat kidneys (b) and diabetic rat kidneys after diabetes onset for 2 weeks (d) and 4 weeks (f). The panels (a, c, and e) represent phase contrast for the same field of b, d, and f, respectively. Magnification: 100×. Arrows indicate the locations of glomeruli. B: Immunostaining for PEDF in the kidney of 8-month-old normal rat (a and b) and age-matched diabetic rats 6 months after the onset of diabetes (c and d). a and c: The glomeruli (600×); b and d: the medulla (400×). The results showed that in the normal rat cortex, PEDF was highly expressed in the glomeruli (A-b). In diabetic rat kidney, dramatic decreases of PEDF expression were observed in the glomeruli (A-d, A-f, and B-c) but not in the medulla (B-d). C: Western blot analysis of PEDF in isolated glomeruli from rats with diabetes for 8 weeks and age-matched nondiabetic controls. Total protein (50 μg) from each sample was blotted with a specific anti-PEDF antibody. The same membrane was stripped and reblotted with an antibody specific to β-actin. D: Immunohistochemistry of PEDF and synaptopodin in the kidneys of rats with diabetes for 2 weeks and nondiabetic controls. a and c: PEDF staining; b and d: synaptopodin staining. Compared with the staining in normal kidneys (a and b), the PEDF expression was dramatically decreased (c), whereas no detectable podocyte loss (d) was observed in the same diabetic kidney (200×).

FIG. 5.

High-glucose–induced decrease of PEDF secretion in cultured HMCs. HMCs were treated with different concentrations of d-glucose and d-mannitol (osmotic control) for 96 h. The medium was harvested by centrifugation, and the PEDF in the supernatant was measured by PEDF ELISA. All the assays were run in triplicate, and the PEDF concentration was normalized by the total protein in the medium. The results were expressed as nanograms PEDF per milligram total protein in the medium (means ± SD). Values statistically different from the normal glucose controls are indicated by *P < 0.05 and **P < 0.01.

FIG. 5.

High-glucose–induced decrease of PEDF secretion in cultured HMCs. HMCs were treated with different concentrations of d-glucose and d-mannitol (osmotic control) for 96 h. The medium was harvested by centrifugation, and the PEDF in the supernatant was measured by PEDF ELISA. All the assays were run in triplicate, and the PEDF concentration was normalized by the total protein in the medium. The results were expressed as nanograms PEDF per milligram total protein in the medium (means ± SD). Values statistically different from the normal glucose controls are indicated by *P < 0.05 and **P < 0.01.

FIG. 6.

PEDF-induced downregulation of TGF-β expression in HMC. A: HMCs were incubated with high glucose (30 mmol/l) for 48 h; then PEDF was added to different concentrations (2.5–160 nmol/l) and incubated for another 48 h. TGF-β in the medium was measured by ELISA. B: After incubation with high glucose (30 mmol/l) for 48 h, 40 nmol/l PEDF was added to the medium of HMCs. The medium was harvested at 24 and 48 h after the addition of PEDF. The TGF-β concentrations in the medium were determined by ELISA. The results were expressed as picograms per milligram total protein in the medium (means ± SD).

FIG. 6.

PEDF-induced downregulation of TGF-β expression in HMC. A: HMCs were incubated with high glucose (30 mmol/l) for 48 h; then PEDF was added to different concentrations (2.5–160 nmol/l) and incubated for another 48 h. TGF-β in the medium was measured by ELISA. B: After incubation with high glucose (30 mmol/l) for 48 h, 40 nmol/l PEDF was added to the medium of HMCs. The medium was harvested at 24 and 48 h after the addition of PEDF. The TGF-β concentrations in the medium were determined by ELISA. The results were expressed as picograms per milligram total protein in the medium (means ± SD).

FIG. 7.

PEDF-induced decrease of fibronectin secretion from HMCs. A: After incubation of HMCs with high glucose (30 mmol/l) for 48 h, PEDF was added to HMC culture to different concentrations (5–40 nmol/l) and incubated for another 48 h. The medium was collected for fibronectin ELISA. B: After incubation of HMCs with 30 mol/l glucose for 48 h, 40 nmol/l PEDF was added to the culture medium. The medium was harvested at 24 and 48 h after the addition of PEDF. The fibronectin levels in the medium were determined by ELISA. The results were expressed as micrograms per milligram of total protein in the medium (means ± SD). Values statistically different from the cells cultured in 30 mmol/l glucose without PEDF are indicated by *P < 0.05 and **P < 0.01.

FIG. 7.

PEDF-induced decrease of fibronectin secretion from HMCs. A: After incubation of HMCs with high glucose (30 mmol/l) for 48 h, PEDF was added to HMC culture to different concentrations (5–40 nmol/l) and incubated for another 48 h. The medium was collected for fibronectin ELISA. B: After incubation of HMCs with 30 mol/l glucose for 48 h, 40 nmol/l PEDF was added to the culture medium. The medium was harvested at 24 and 48 h after the addition of PEDF. The fibronectin levels in the medium were determined by ELISA. The results were expressed as micrograms per milligram of total protein in the medium (means ± SD). Values statistically different from the cells cultured in 30 mmol/l glucose without PEDF are indicated by *P < 0.05 and **P < 0.01.

TABLE 1

Microalbuminuria and general conditions in diabetic rats

Blood glucose (mg/dl)Body weight (g)24-h urine volume (ml)Albumin in 24-h urine (μg)
Duration of diabetes     
    Before diabetes 106.80 ± 13.61 139.40 ± 6.54 4.13 ± 1.48 8.74 ± 4.20 
    1 week 498.50 ± 90.91 134.83 ± 10.16 31.25 ± 13.90 32.36 ± 3.95 
    2 weeks 504.17 ± 89.16 137.17 ± 17.94 34.25 ± 22.56 38.59 ± 16.36 
    4 weeks 462.00 ± 91.66 144.60 ± 16.04 35.25 ± 28.64 76.95 ± 24.50 
    6 weeks 510.20 ± 54.20 144.20 ± 21.76 47.00 ± 44.78 90.75 ± 37.45 
Blood glucose (mg/dl)Body weight (g)24-h urine volume (ml)Albumin in 24-h urine (μg)
Duration of diabetes     
    Before diabetes 106.80 ± 13.61 139.40 ± 6.54 4.13 ± 1.48 8.74 ± 4.20 
    1 week 498.50 ± 90.91 134.83 ± 10.16 31.25 ± 13.90 32.36 ± 3.95 
    2 weeks 504.17 ± 89.16 137.17 ± 17.94 34.25 ± 22.56 38.59 ± 16.36 
    4 weeks 462.00 ± 91.66 144.60 ± 16.04 35.25 ± 28.64 76.95 ± 24.50 
    6 weeks 510.20 ± 54.20 144.20 ± 21.76 47.00 ± 44.78 90.75 ± 37.45 

Data are means ± SD (n = 4–6).

J.J.W. and S.X.Z. contributed equally to this work.

This study was supported by National Institutes of Health Grants EY015650 and EY12231, Juvenile Diabetes Research Foundation Grant 1-2003-627, and a research award from the American Diabetes Association.

1.
Raptis AE, Viberti G: Pathogenesis of diabetic nephropathy.
Exp Clin Endocrinol Diabetes
109
:
S424
–S437,
2001
2.
American Diabetes Association: Diabetic nephropathy.
Diabetes Care
23
:
S69
–S72,
2000
3.
Sakharova OV, Taal MW, Brenner BM: Pathogenesis of diabetic nephropathy: focus on transforming growth factor-beta and connective tissue growth factor.
Curr Opin Nephrol Hypertens
10
:
727
–738,
2001
4.
The Diabetes Control and Complications (DCCT) Research Group: Effect of intensive therapy on the development and progression of diabetic nephropathy in the Diabetes Control and Complications Trial.
Kidney Int
47
:
1703
–1720,
1995
5.
The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Research Group: Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of intensive therapy.
N Engl J Med
342
:
381
–389,
2000
(erratum in N Engl J Med 342:1376, 2000)
6.
Chiarelli F, Santilli F, Mohn A: Role of growth factors in the development of diabetic complications.
Horm Res
53
:
53
–67,
2000
7.
Cooper ME: Interaction of metabolic and haemodynamic factors in mediating experimental diabetic nephropathy.
Diabetologia
44
:
1957
–1972,
2001
8.
Goldfarb S, Ziyadeh FN: TGF-beta: a crucial component of the pathogenesis of diabetic nephropathy.
Trans Am Clin Climatol Assoc
112
:
27
–32,
2001
9.
Greener M: Targeting TGF could counter diabetic nephropathy.
Mol Med Today
6
:
376
,
2000
10.
Ng DP, Warram JH, Krolewski AS: TGF-beta 1 as a genetic susceptibility locus for advanced diabetic nephropathy in type 1 diabetes mellitus: an investigation of multiple known DNA sequence variants.
Am J Kidney Dis
41
:
22
–28,
2003
11.
Tamaki K, Okuda S: Role of TGF-beta in the progression of renal fibrosis.
Contrib Nephrol
139
:
44
–65,
2003
12.
Dvorak HF, Brown LF, Detmar M, Dvorak AM: Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis.
Am J Pathol
146
:
1029
–1039,
1995
13.
Aiello LP: Vascular endothelial growth factor and the eye: biochemical mechanisms of action and implications for novel therapies.
Ophthalmic Res
29
:
354
–362,
1997
14.
Iglesias-de la Cruz MC, Ziyadeh FN, Isono M, Kouahou M, Han DC, Kalluri R, Mundel P, Chen S: Effects of high glucose and TGF-beta1 on the expression of collagen IV and vascular endothelial growth factor in mouse podocytes.
Kidney Int
62
:
901
–913,
2002
15.
Gambaro G, Ceol M, Del Prete D, D’Angelo A: GLUT-1 and TGF-beta: the link between hyperglycaemia and diabetic nephropathy.
Nephrol Dial Transplant
15
:
1476
–1477,
2000
16.
Lane PH, Snelling DM, Langer WJ: Streptozotocin diabetes elevates all isoforms of TGF-beta in the rat kidney.
Int J Exp Diabetes Res
2
:
55
–62,
2001
17.
Kim BS, Goligorsky MS: Role of VEGF in kidney development, microvascular maintenance and pathophysiology of renal disease.
Korean J Intern Med
18
:
65
–75,
2003
18.
Senthil D, Choudhury GG, McLaurin C, Kasinath BS: Vascular endothelial growth factor induces protein synthesis in renal epithelial cells: a potential role in diabetic nephropathy.
Kidney Int
64
:
468
–479,
2003
19.
Bortoloso E, Del Prete D, Gambaro G, Dalla Vestra M, Sailer A, Baggio B, Antonucci F, Fioretto P, Anglani F: Vascular endothelial growth factor (VEGF) and VEGF receptors in diabetic nephropathy: expression studies in biopsies of type 2 diabetic patients.
Ren Fail
23
:
483
–493,
2001
20.
Pfeiffer A, Middelberg-Bisping K, Drewes C, Schatz H: Elevated plasma levels of transforming growth factor-beta 1 in NIDDM.
Diabetes Care
19
:
1113
–1117,
1996
21.
Ellis D, Forrest KY, Erbey J, Orchard TJ: Urinary measurement of transforming growth factor-beta and type IV collagen as new markers of renal injury: application in diabetic nephropathy.
Clin Chem
44
:
950
–956,
1998
22.
Mogyorosi A, Kapoor A, Isono M, Kapoor S, Sharma K, Ziyadeh FN: Utility of serum and urinary transforming growth factor-beta levels as markers of diabetic nephropathy.
Nephron
86
:
234
–235,
2000
23.
Tombran-Tink J, Shivaram SM, Chader GJ, Johnson LV, Bok D: Expression, secretion, and age-related downregulation of pigment epithelium-derived factor, a serpin with neurotrophic activity.
J Neurosci
15
:
4992
–5003,
1995
24.
Karakousis PC, John SK, Behling KC, Surace EM, Smith JE, Hendrickson A, Tang WX, Bennett J, Milam AH: Localization of pigment epithelium derived factor (PEDF) in developing and adult human ocular tissues.
Mol Vis
7
:
154
–163,
2001
25.
Tombran-Tink J, Barnstable CJ: PEDF: a multifaceted neurotrophic factor.
Nat Rev Neurosci
4
:
628
–636,
2003
26.
Tombran-Tink J, Barnstable CJ: Therapeutic prospects for PEDF: more than a promising angiogenesis inhibitor.
Trends Mol Med
9
:
244
–250,
2003
27.
Dawson DW, Volpert OV, Gillis P, Crawford SE, Xu H, Benedict W, Bouck NP: Pigment epithelium-derived factor: a potent inhibitor of angiogenesis.
Science
285
:
245
–258,
1999
28.
Liu H, Ren JG, Cooper WL, Hawkins CE, Cowan MR, Tong PY: Identification of the antivasopermeability effect of pigment epithelium-derived factor and its active site.
Proc Natl Acad Sci U S A
101
:
6605
–6610,
2004
29.
Boehm BO, Lang G, Volpert O, Jehle PM, Kurkhaus A, Rosinger S, Lang GK, Bouck N: Low content of the natural ocular anti-angiogenic agent pigment epithelium-derived factor (PEDF) in aqueous humor predicts progression of diabetic retinopathy.
Diabetologia
46
:
394
–400,
2003
30.
Gao G, Li Y, Zhang D, Gee S, Crosson C, Ma J: Unbalanced expression of VEGF and PEDF in ischemia-induced retinal neovascularization.
FEBS Lett
489
:
270
–276,
2001
31.
Gao G, Li Y, Gee S, Dudley A, Fant J, Crosson C, Ma J-X: Down-regulation of VEGF and up-regulation of PEDF: a possible mechanism for the anti-angiogenic activity of plasminogen kringle 5.
J Biol Chem
277
:
9492
–9497,
2002
32.
Mori K, Duh E, Gehlbach P, Ando A, Takahashi K, Pearlman J, Yang HS, Zack DJ, Ettyreddy D, Brough DE, Wei LL, Campochiaro PA: Pigment epithelium-derived factor inhibits retinal and choroidal neovascularization.
J Cell Physiol
188
:
253
–263,
2001
33.
Stellmach V, Crawford SE, Zhou W, Bouck N: Prevention of ischemia-induced retinopathy by the natural ocular antiangiogenic agent pigment epithelium-derived factor.
Proc Natl Acad Sci U S A
98
:
2593
–2597,
2001
34.
Doll JA, Stellmach VM, Bouck NP, Bergh AR, Lee C, Abramson LP, Cornwell ML, Pins MR, Borensztajn J, Crawford SE: Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas.
Nat Med
9
:
774
–780,
2003
35.
Abramson LP, Stellmach V, Doll JA, Cornwell M, Arensman RM, Crawford SE: Wilms’ tumor growth is suppressed by antiangiogenic pigment epithelium-derived factor in a xenograft model.
J Pediatr Surg
38
:
336
–342,
2003
36.
Uehara H, Miyamoto M, Kato K, Ebihara Y, Kaneko H, Hashimoto H, Murakami Y, Hase R, Takahashi R, Mega S, Shichinohe T, Kawarada Y, Itoh T, Okushiba S, Kondo S, Katoh H: Expression of pigment epithelium-derived factor decreases liver metastasis and correlates with favorable prognosis for patients with ductal pancreatic adenocarcinoma.
Cancer Res
64
:
3533
–3537,
2004
37.
Tombran-Tink J, Mazuruk K, Rodriguez IR, Chung D, Linker T, Englander E, Chader GJ: Organization, evolutionary conservation, expression and unusual Alu density of the human gene for pigment epithelium-derived factor, a unique neurotrophic serpin.
Mol Vis
2
:
11
,
1996
38.
Weston BS, Wahab NA, Mason RM: CTGF mediates TGF-beta-induced fibronectin matrix deposition by upregulating active alpha5beta1 integrin in human mesangial cells.
J Am Soc Nephrol
14
:
601
–610,
2003
39.
Tombran-Tink J, Chader GG, Johnson LV: PEDF: a pigment epithelium-derived factor with potent neuronal differentiative activity.
Exp Eye Res
53
:
411
–414,
1991
40.
Duh EJ, Yang HS, Haller JA, De Juan E, Humayun MS, Gehlbach P, Melia M, Pieramici D, Harlan JB, Campochiaro PA, Zack DJ: Vitreous levels of pigment epithelium-derived factor and vascular endothelial growth factor: implications for ocular angiogenesis.
Am J Ophthalmol
137
:
668
–674,
2004
41.
Ogata N, Nishikawa M, Nishimura T, Mitsuma Y, Matsumura M: Unbalanced vitreous levels of pigment epithelium-derived factor and vascular endothelial growth factor in diabetic retinopathy.
Am J Ophthalmol
134
:
348
–353,
2002
42.
Lopez-Casillas F: [Transforming growth factor-beta (TGF-beta), keystone in diabetic nephropathy].
Rev Invest Clin
52
:
487
–490,
2000
43.
Chen S, Iglesias-de la Cruz MC, Jim B, Hong SW, Isono M, Ziyadeh FN: Reversibility of established diabetic glomerulopathy by anti-TGF-beta antibodies in db/db mice.
Biochem Biophys Res Commun
300
:
16
–22,
2003