Abstract

Thiazolidinedione (TZD), a ligand for peroxisome proliferator-activated receptor-γ (PPAR-γ), exerts anti-inflammatory effects independently of the insulin-sensitizing effect. In the present study, we tested the hypothesis that TZD prevents the progression of diabetic nephropathy by modulating the inflammatory process. Five-week-old Sprague-Dawley rats were divided into three groups: 1) nondiabetic control rats (non-DM), 2) diabetic rats (DM), and 3) diabetic rats treated with pioglitazone (DM+pio). Diabetes was induced by injection with streptozotocin (STZ). The DM+pio group received 0.0002% pioglitazone mixed in chow for 8 wk after induction of diabetes. Blood glucose and HbA1c were elevated in diabetic rats but did not change by treatment with pioglitazone. Pioglitazone reduced urinary albumin excretion and glomerular hypertrophy, suppressed the expression of transforming growth factor (TGF)-β, type IV collagen, and ICAM-1, and infiltration of macrophages in the kidneys of diabetic rats. Furthermore, renal NF-κB activity was increased in diabetic rats and reduced by pioglitazone. PPAR-γ was expressed in glomerular endothelial cells in the diabetic kidney and in cultured glomerular endothelial cells. High-glucose conditions increased the expression of ICAM-1 and the activation of NF-κB in cultured glomerular endothelial cells. These changes were reduced by pioglitazone, ciglitazone, and pyrrolidine dithiocarbamate, an inhibitor of NF-κB. However, pioglitazone did not show the changes in the presence of PPAR-γ antagonist GW9662. Our results suggest that the preventive effects of pioglitazone may be mediated by its anti-inflammatory actions, including inhibition of NF-κB activation, ICAM-1 expression, and macrophage infiltration in the diabetic kidney.

  • macrophages
  • ICAM-1

diabetic nephropathy is a major cause of end-stage renal failure in many developed countries. Recent studies have suggested the emerging role of inflammatory processes in the pathogenesis of diabetic nephropathy in addition to other well-known mechanisms such as advanced glycation end-products (AGE) (42), activation of protein kinase C (21), acceleration of the polyol pathway (6), and transforming growth factor (TGF)-β (33). Macrophage infiltration and increased expression of leukocyte adhesion molecules are seen in the kidneys of patients with diabetic nephropathy in addition to mesangial matrix expansion and interstitial fibrosis (14).

ICAM-1 is one of the major adhesion molecules that promote leukocyte attachment to the endothelium and their transmigration by its binding to β2-integrins on leukocyte cell surfaces. We have recently described the importance of ICAM-1-dependent infiltration of macrophages into the kidney in the pathogenesis of diabetic nephropathy in a series of studies (14, 28, 35, 36, 38, 44). Our studies also showed that the expression of ICAM-1 is rapidly induced and maintained for a long time in renal tissues after induction of diabetes in experimental type 1 diabetic rats (36). Furthermore, macrophage infiltration was blocked by anti-ICAM-1 antibody, confirming that ICAM-1 mediates macrophage infiltration into the diabetic kidney (36). Glomerular infiltration of macrophages was noted in both type 1 and type 2 models of diabetes (5). Other studies showed that depletion of leukocytes by irradiation improved glomerular dysfunction in the early phase of diabetic nephropathy in animal models of type 1 diabetes (32). Furthermore, we have demonstrated that ICAM-1-deficient mice are protected from renal injury after induction of diabetes, suggesting that the inflammatory process is a critical factor for the development of diabetic nephropathy (28).

Thiazolidinediones (TZDs), ligands for the nuclear receptor peroxisome proliferator-activated receptor (PPAR)-γ, are widely used for diabetes therapy as insulin-sensitizing agents. Some recent studies have shown the preventive effects of TZDs on diabetic nephropathy in human and animal models (26, 41). TZDs have been reported to reduce albuminuria in patients with type 2 diabetes and in diabetic animals (3, 8, 17). PPAR-γ is found in a variety of nonadipose tissues and cells, including skeletal muscle, heart, kidney, vascular smooth muscle cells, neutrophils, lymphocytes, and macrophages (9, 20, 22, 39, 43). Low-grade inflammation is known to contribute to the progression of atherosclerosis (31). Moreover, TZDs are known to reduce the progression of atherosclerosis through its anti-inflammatory effects (10). It has recently been shown that TZDs are effective in slowing the progression of glomerular fibrosis in a non-insulin-resistant ⅚ nephrectomized rat model (23).

The aim of the present study was to test the hypothesis that TZDs prevent the development of diabetic nephropathy through inhibition of the inflammatory processes including activation of NF-κB, ICAM-1 expression, and macrophage infiltration.

MATERIALS AND METHODS

Experimental protocol.

Male Sprague-Dawley rats aged 5 wk were divided into three groups: 1) nondiabetic control rats (non-DM; n = 12); 2) streptozotocin (STZ)-induced diabetic rats (DM; n = 6); and 3) diabetic rats treated with pioglitazone (0.0002% mixed in standard diet; DM+pio; n = 6). Rats of both diabetic groups were injected intravenously with STZ (65 mg/kg body wt) in citrate buffer (pH 4.5). Blood glucose levels were determined at 3 and 7 days after STZ injection, and only rats with blood glucose concentrations >16 mmol/l were used in the study. The DM+pio group received pioglitazone mixed in chow starting from the day before the STZ injection. Rats of the non-DM and DM groups received normal chow. All rats had free access to standard chow and tap water. All procedures were performed according to the Guidelines for Animal Experiments at Okayama University Medical School, Japanese Government Animal Protection and Management Law (no. 105), and Japanese Government Notification on Feeding and Safekeeping of Animals (no. 6). All rats were killed at 8 wk after the induction of diabetes, and the kidneys were harvested, weighed, and fixed in 10% formalin for periodic acid-methenamine silver (PAM) and azan staining. The remaining tissues were embedded in optimal cutting temperature compound (Sakura Finetechnical, Tokyo, Japan) and immediately frozen in acetone cooled on dry ice. For the EMSA, tissues were snap frozen in liquid nitrogen and stored at −80°C.

Metabolic data.

We measured body weight, blood pressure, blood glucose levels, hemoglobin A1c (HbA1c), 24-h urinary albumin excretion (UAE), and creatinine clearance at 1, 2, 3, 4, and 8 wk. Blood pressure was measured using the tail-cuff method. HbA1c was measured using the high-pressure liquid chromatography method, and serum creatinine was measured using the enzymatic method (7). Urine samples were collected over a 24-h period in individual metabolic cages. UAE in a 24-h urine collection was measured by nephelometry using anti-rat albumin antibody (ICN Pharmaceuticals, Aurora, OH). Creatinine clearance was calculated in individual rats.

Light microscopy.

Renal tissues were fixed in 10% formalin and embedded in paraffin. Paraffin sections (4 μm) were then stained with PAM. To evaluate glomerular size, 100 glomeruli/group were examined. Glomerular surface area was measured by manually tracing Bowman's capsule and using Photoshop Software Version 6 (Adobe Systems, San Jose, CA) and Scion image-analysis software (Scion, Frederick, MD). The results are expressed as means ± SE.

Immunoperoxidase staining of macrophages.

Infiltration of macrophages was evaluated by immunoperoxidase staining using an ABC kit (Vector Laboratories, Burlingame, CA) as described previously (36). Briefly, fresh-frozen sections were cut at 4-μm thickness. To reduce background interference, nonspecific binding was blocked by incubation with 10% normal rabbit serum in Tris-buffered saline for 20 min. Nonspecific staining was blocked by 15-min incubation with avidin and then biotin using an avidin-biotin blocking kit (Vector Laboratories). Endogenous peroxidase activity was inhibited by 20-min incubation with methanol containing 0.3% H2O2. Sections were first incubated with a monoclonal antibody (mAb) against rat monocytes/macrophages (ED1) for 12 h at 4°C. The sections were then incubated with biotin-labeled goat anti-mouse IgG for 30 min. Biotinylated horseradish peroxidase was applied for 30 min. Peroxidase activity was developed in 3, 3-diaminobendine. Mayer's hematoxylin was added as a counterstain.

Intraglomerular ED1-positive cells were counted in 200 glomeruli/group under a high-power field by two independent observers with no prior knowledge of the experimental design. The average number per glomerulus was used for the estimation.

Immunofluorescence staining for ICAM-1, type IV collagen, PPAR-γ, and endothelial cells in rat kidney.

The expression of ICAM-1, type IV collagen, PPAR-γ, and endothelial cells was detected by indirect immunofluorescence as described previously (28). Briefly, fresh-frozen sections (4-μm thick) were stained with a mAb against rat ICAM-1 (1A29), anti-mouse type IV collagen mAb, a polyclonal antibody against PPAR-γ, or a mAb against RECA-1 for 12 h at 4°C. They were then stained with FITC-conjugated rabbit-anti-mouse IgG, FITC-conjugated goat anti-rabbit IgG, FITC-conjugated donkey anti-goat IgG, or rhodamine-conjugated rabbit anti-mouse IgG for 30 min at room temperature. The stained sections were observed under a confocal laser fluorescence microscope (LSM-510, Carl Zeiss, Jena, Germany). ICAM-1 and type IV collagen immunofluorescence intensities were quantified by the modified method described previously, and color images were saved as PICT files using the LSM-510 microscope. The brightness of each image file was uniformly enhanced by Photoshop, followed by analysis using Scion image software. TIFF image files were inverted and opened in grey scale mode. The ICAM-1 and type IV collagen indexes were calculated using the formula {[X (density) × positive area (μm2)]/glomerular total area (μm2)}, where the staining density is indicated by a number from 100 to 256 in gray scale.

RNA preparation from renal cortex of rats.

Total RNA was prepared from renal cortex of rats using an RNeasy Midi kit using the instructions provided by the manufacturer (Qiagen, Valencia, CA).

Quantitative real-time RT-PCR analysis for renal cortex of rats.

Quantitative real-time RT-PCR was used to quantify the amounts of TGF-β and β-actin mRNAs in renal tissues. Real-time PCR reactions were performed as described previously (27). The amounts of PCR products were normalized with a housekeeping gene (β-actin) to calculate the relative expression ratios for TGF-β chain mRNA. The following oligonucleotide primers specific for rat TGF-β (GenBank accession no. X52498) and β-actin (accession no. NM031144) were used: TGF-β, 5′-GCAACAACGCAATCTATGAC-3′ (sense) and 5′-CCTGTATTCCGTCTCCTT-3′ (antisense) and β-actin, 5′-CCTGTATGCCTCTGGTCGTA-3′ (sense) and 5′-CCATCTCTTGCTCGAAGTCT-3′ (antisense). Each experiment was performed twice.

EMSA for renal cortex.

Tissue extracts and EMSA for the analysis of transcription factor NF-κB were performed as described earlier (24). Briefly, frozen kidney cortical sections were minced and suspended in 1 ml of Tris-buffered saline (TBS) buffer [25 mM Tris·HCl (pH 7.4), 130 mM NaCl, and 5 mM KCl] and homogenized with tube homogenizer 303 (Ikeda Scientific, Tokyo, Japan). The homogenates were centrifuged, and the pellets resuspended in 1 ml of buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM PMSF, and 1 mM DTT) and chilled on ice for 20 min. Next, 100 μl of 10% Nonidet P-40 were added and vigorously vortexed. The nuclear fraction was collected by centrifugation and resuspended in 100 μl of buffer B [20 mM HEPES (pH 7.9), 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF]. For EMSA, 50 μg of nuclear extract were incubated with 1 ng of [γ-32P]ATP-labeled oligonucleotide, containing an NF-κB binding site (5′Œ-AGTTGAGGGGACTTCCCAGGC-3′Œ) in 20 μg binding buffer [10 mM Tris·HCl (pH 7.5), 50 mM NaCl, 0.5 mM EDTA, 0.5 mM DTT, 1 mM MgCl2, 4% glycerol, and 50 mg/ml dI-dC], for 30 min at room temperature. Individual samples were then electrophoresed on a 4% polyacrylamide gel. The gel was dried and exposed to X-ray film (Hyperfilm-MP, Amersham Pharmacia Biotech, Piscataway, NJ). The relative intensity of the autoradiogram was determined by a scanning densitometer. The results of the densitometric scanning are presented as means ± SE (n = 6 for each group). To assess the specificity of the reaction, we performed a competition assay with a 50-fold excess of unlabeled consensus sequences of NF-κB. The unlabeled probes were added to the binding reaction 10 min before the labeled probe. Each experiment was performed three times.

Cell culture.

Human glomerular microvascular endothelial cells (GE cells) were obtained from the Applied Cell Biology Research Institute (Kirkland, WA) and cultured in CS-C complete medium (Cell Systems, Kirkland, WA) supplemented with 19.4 mM d-glucose,10% FCS, and growth factor in a 5% CO2 incubator at 37°C.

For the preparation of RNA, cells were incubated in fresh media with 100 ng/μl PMA (Sigma, St. Louis, MO) for 24 h. Cells were detached with trypsin/EDTA and trypsin inhibitor solution (Cell Systems) and washed three times with PBS. Total RNA was prepared from cells using an RNeasy Midi kit (Qiagen) as described above.

For other experiments, cells were trypsinized and passaged to gelatin-precoated 100-mm dishes. We changed the medium from CS-C complete medium to EGM-MV2 medium (Cambrex, East Rutheford, NJ) containing 5.5 mM d-glucose and 10% FCS with growth factor 24 h before the next protocols.

RT-PCR analysis: the expression of PPAR-γ in human GE cells.

We used 1 μg RNA from these cells for the RT reaction using an RT-PCR kit (PerkinElmer, Foster City, CA). To confirm the expression of PPAR-γ in GE cells, we used specific primers for PPAR-γ (LightCycler-Primer Set, Roche Molecular Biochemicals, Mannheim, Germany), as described previously (37). The PCR protocol was as follows: initial denaturation at 95°C for 10 min, followed by 35 cycles at 95°C for 10 s, annealing at 68°C for 10 s, and extension at 72°C for 16 s.

Effect of pioglitazone, ciglitazone, pyrrolidine dithiocarbamate, and GW9662 for high glucose-induced ICAM-1 expression on GE cells.

To clarify the relationship between ICAM-1 expression and NF-κB activation, we used pyrrolidine dithiocarbamate (PDTC; Sigma), an inhibitor of NF-κB. We also used ciglitazone (Biomol, Plymouth Meeting, PA) as another TZD and a specific antagonist of PPAR-γ, GW9662 (Sigma). To evaluate the hyperglycemic effects, we used the agents of d-glucose (Sigma).

After 24-h preincubation in medium with 5.5 mM d-glucose, GE cells were exposed to the following experimental conditions for 24 h; 1) 5.5 mM d-glucose (normal glucose; NG), 2) 30 mM d-glucose (high glucose; HG), 3) 30 mM d-glucose with 10 μM pioglitazone, 4) 30 mM d-glucose with 10 μM ciglitazone, 5) 30 mM d-glucose with 100 μM PDTC, and 6) 30 mM d-glucose with 10 μM pioglitazone and 10 μM GW9662. After incubation, cells were collected using a cell scraper and were lysed in TNE Buffer [50 mM Tris·HCl (pH 7.4), 1% Nonidet P-40, 20 mM EDTA] for 30 min on ice. The immunoblot assay was performed as previously described (44). The blots were then visualized and analyzed as previously described (40).

Effect of pioglitazone, ciglitazone, PDTC, and GW9662 for high glucose-induced NF-κB activation on GE cells.

After 24-h preincubation in medium with 5.5 mM d-glucose, GE cells were exposed to the following experimental conditions for 24 h; 1) 5.5 mM d-glucose, 2) 30 mM d-glucose, 3) 30 mM d-glucose with 10 μM pioglitazone, 4) 30 mM d-glucose with 10 μM ciglitazone, 5) 30 mM d-glucose with 100 μM PDTC, and 6) 30 mM d-glucose with 10 μM pioglitazone and 10 μM GW9662. After incubation for 24 h, nuclear extracts were prepared as previously described (13, 38). The blots were then visualized and analyzed as described above.

Antibodies.

The primary antibodies were anti-rat ICAM-1 mAb (1A29; Seikagaku, Tokyo, Japan); anti-mouse type IV collagen mAb (Seikagaku); anti-rat monocyte/macrophage mAb (ED1) (Serotec, Oxford, UK); anti-human PPAR-γ Ab (Santa Cruz Biotechnology, Santa Cruz, CA); anti-rat RECA-1 mAb (Serotec); and anti-human ICAM-1 Ab. The secondary antibodies were FITC-conjugated rabbit-anti-mouse IgG; FITC-conjugated goat anti-rabbit IgG (Jackson Immunoresearch Laboratories, West Grove, PA); fluorophore-labeled donkey anti-goat IgG; fluorophore-labeled rabbit anti-mouse IgG (Molecular Probes, Eugene, OR); and horseradish peroxidase-linked anti-mouse or anti-rabbit IgG (Amersham Biosciences).

Statistical analysis.

All values are expressed as means ± SE. Differences between groups were examined for statistical significance using one-way ANOVA followed by Scheffé's test. A P value <0.05 denoted the presence of a statistically significant difference.

RESULTS

Metabolic data.

At 8 wk after induction of diabetes, there were no significant differences in the systolic blood pressure of the three groups. HbA1c of both DM and DM+pio groups were significantly higher than that of the non-DM group. There was no significant difference in HbA1c and creatinine clearance between the DM and DM+pio group (Table 1). Body weights of both diabetic groups were lower than those for the non-DM group. There was no significant difference in the kidney weights of the three groups.

View this table:
Table 1.

Metabolic data at 8 wk after induction of diabetes

UAE.

UAE increased progressively in diabetic rats during the study (Fig. 1). However, pioglitazone treatment significantly reduced the mean UAE (272 ± 52 μg/day) compared with the DM group at 4 wk after the induction of diabetes (693 ± 106 μg/day, *P < 0.05).

Fig. 1.

Time course of changes in urinary albumin excretion (UAE). UAE increased progressively in the untreated diabetic (DM) group during the 8-wk observation period following induction of diabetes. Pioglitazone (pio) treatment (DM+pio) significantly reduced UAE at 4 wk compared with the DM group. Non-DM, controls. Values are means ± SE. *P < 0.05 vs. untreated DM rats.

Kidney morphology.

At 8 wk after the induction of diabetes, the glomerular surface area was increased in the DM group (8,990 ± 113 μm2) compared with the non-DM group (8,204 ± 126 μm2, *P < 0.0001). Pioglitazone treatment significantly reduced the glomerular surface area (8,214 ± 121 μm2) compared with the DM group (*P < 0.0001) (Fig. 2, AD).

Fig. 2.

Periodic acid-methenamine silver (PAM) staining of kidney sections for glomerular size (tuft area). Original magnification ×400. Shown are representative glomeruli from non-DM controls (A), DM rats (B), and DM+pio rats (C). D: pioglitazone suppresses the increase in glomerular size compared with DM rats. *P < 0.0001.

Glomerular macrophage infiltration.

The number of macrophages (ED1-positive cells) in glomeruli was higher in the DM group than in the non-DM group. Pioglitazone treatment significantly reduced the number of glomerular infiltrating macrophages (Fig. 3, AC and G).

Fig. 3.

Immunoperoxidase staining for macrophages (ED1-positive cells) in glomeruli (AC). DF: immunofluorescence staining for ICAM-1 in glomeruli. Kidney tissues were harvested from non-DM controls (A and D), DM (B and E), and DM+pio rats (C and F). Also shown are number of macrophages (ED1-positive cells; G) and immunofluorescence intensity for ICAM-1 (ICAM-1 index; H) in glomeruli of non-DM, DM, and DM+pio rats. Values are means ± SE. *P < 0.0001.

Expression of ICAM-1, type IV collagen, and TGF-β in kidneys.

In diabetic rats, ICAM-1 expression was increased in GE cells. However, it was markedly reduced in the DM+pio group compared with the DM group (*P < 0.0001) (Fig. 3, DF and H). A similar trend was noted for type IV collagen (*P < 0.0001) (Fig. 4, AD). TGF-β gene expression in renal tissues was significantly enhanced in diabetic rats compared with nondiabetic rats, and pioglitazone suppressed the increase in mRNA at 8 wk after the induction of diabetes (*P < 0.001) (Fig. 5).

Fig. 4.

Histological expression of type IV collagen in glomeruli (AC). A: non-DM rats. B: DM rats. C: DM+pio rats. D: type IV collagen expression estimated by fluorescence intensities was markedly reduced in the DM+pio group compared with the DM group. Values are means ± SE. *P < 0.0001.

Fig. 5.

Quantitative analysis of TGF-β mRNA in kidney sections. Values are means ± SE of 6 rats/group. *P < 0.001.

NF-κB activation in the renal cortex.

Using EMSA, we analyzed the renal binding activity of NF-κB DNA, which is one of the regulators of ICAM-1. The results of densitometric scanning for NF-κB activation are shown as relative intensity [to lane 4 (background level), Fig. 6, A and B]. NF-κB activity was increased in the kidneys of the DM group (relative intensity: 3.074 ± 0.266) compared with the non-DM group (1.000 ± 0.240, *P < 0.05). Pioglitazone treatment reduced renal DNA binding activity of NF-κB compared with the DM group (relative intensity to non-DM group: 1.592 ± 0.339, *P < 0.05, Fig. 6, A and B).

Fig. 6.

A: EMSA for NF-κB in rat kidneys. Lanes 13: NF-κB activation. Lane 1, non-DM; lane 2, DM; lane 3, DM+pio. Lanes 4 and 5: specificity of NF-κB DNA binding. Lane 4, pretreated with excess cold NF-κB; lane 5, pretreated with excess cold mutant NF-κB. B: NF-κB activity was increased in the DM group and markedly reduced by pioglitazone. Values are means ± SE. Individual samples were electrophoresed; n = 6/group. Only representative data are shown for each group. *P < 0.05.

Expression and localization of PPAR-γ in renal tissues.

Double immunofluorescence staining was performed for PPAR-γ and RECA-1 in renal tissues of rats. PPAR-γ was stained in glomeruli (Fig. 7, A and C). The merged image of PPAR-γ and RECA-1 staining showed that RECA-1-positive cells (Fig. 7D) included PPAR-γ-positive cells (Fig. 7E), suggesting the expression of PPAR-γ in endothelial cells. Immunohistochemistry did not show a significant difference in PPAR-γ staining between DM and non-DM groups.

Fig. 7.

A and C: peroxisome proliferator-activated receptor-γ (PPAR-γ) protein expression in a glomerulus of diabetic rats. PPAR-γ staining is visible in glomerular cells. B: negative control for PPAR-γ staining using FITC-conjugated anti-rabbit IgG without primary antibody. D: RECA-1 staining. E: images of PPAR-γ and RECA-1 staining were merged. F: expression of PPAR-γ mRNA in GE cells. RT-PCR products were resolved on a 1% agarose gel. The DNA ladder was included as a marker to indicate the size of the PCR products of PPAR-γ. Specific cDNAs were synthesized from human hematopoietic cell lines (positive controls) and glomerular endothelial (GE) cells.

Expression of PPAR-γ mRNA in GE cells.

RT-PCR analysis showed that PPAR-γ mRNA was expressed in GE cells (Fig. 7F).

Effect of pioglitazone, ciglitazone, PDTC, and GW9662 for high glucose-induced ICAM-1 expression on GE cells.

ICAM-1 expression was significantly increased in HG groups compared with NG groups. High glucose-induced ICAM-1 expression was significantly inhibited by pioglitazone, ciglitazone, and PDTC. In addition, pioglitazone did not suppress ICAM-1 expression in the presence of GW9662 (Fig. 8, A and B). The relative amount of ICAM-1 was as follows: NG, 1.00 ± 0.20; HG, 9.33 ± 1.44; HG+pioglitazone, 2.86 ± 0.61; HG+ciglitazone, 3.17 ± 0.78; HG+PDTC, 2.07 ± 0.24; and HG+pioglitazone+GW9662, 9.38 ± 1.17.

Fig. 8.

Pioglitazone, ciglitazone, and pyrrolidine dithiocarbamate (PDTC) attenuated high glucose-induced ICAM-1 expression (A and B) and NF-κB activity (C and D) on GE cells. GW9662 inhibited these suppressive effects of TZDs and PDTC on GE cells (AD). *P < 0.0001, **P < 0.005, ***P < 0.05.

Effect of pioglitazone, ciglitazone, PDTC, and GW9662 for high glucose-induced NF-κB activation on GE cells.

The activity of NF-κB was significantly increased in HG groups compared with NG groups. High glucose-induced NF-κB activation was significantly inhibited by pioglitazone, ciglitazone, and PDTC. In addition, pioglitazone did not suppress the NF-κB activity in the presence of GW9662 (Fig. 8, C and D). The relative amount of NF-κB was as follows: NG, 1.00 ± 0.002; HG, 1.64 ± 0.15; HG+pioglitazone, 1.13 ± 0.004; HG+ciglitazone, 1.07 ± 0.04; HG+PDTC, 1.01 ± 0.03; and HG+pioglitazone+GW9662, 1.63.1 ± 0.14.

DISCUSSION

In the present study, we showed that treatment with pioglitazone ameliorated albuminuria and glomerular hypertrophy without changing blood glucose levels in STZ-induced diabetic rats. Pioglitazone treatment markedly decreased ICAM-1 expression and macrophage infiltration in diabetic glomeruli. Pioglitazone also reduced the augmented expression of TGF-β and type IV collagen in diabetic kidneys. Furthermore, we found that pioglitazone suppressed the activity of NF-κB, a major transcription factor of many proinflammatory genes. The localization of PPAR-γ in glomeruli was determined by immunohistochemistry in rat kidney. Expression of PPAR-γ in GE cells was confirmed by RT-PCR in cultured GE cells. High glucose-induced ICAM-1 expression and NF-κB activity were suppressed by pioglitazone, ciglitazone, and NF-κB inhibitor on GE cells. Furthermore, PPAR-γ antagonist inhibited these suppressive effects of TZDs and NF-κB inhibitor on GE cells. Our findings suggest that pioglitazone shows an anti-inflammatory effect in diabetic kidneys and prevents the development of nephropathy, independently of blood glucose levels.

TZDs are known to effectively block inflammatory reactions in vitro and in vivo, suggesting that PPAR-γ has an anti-inflammatory effect (25). PPAR-γ is expressed not only in adipocytes but also in monocytes/macrophages and endothelial cells. In this study, we found PPAR-γ expression in GE cells. Native or synthetic PPAR-γ ligands inhibit the production of several inflammatory mediators, nitric oxide synthase, gelatinase B1, scavenger receptor A, an antagonist to AP-1, signal transducer and activator of transcription, and NF-κB transcription activity (16, 20). PPAR-γ negatively regulates the binding of monocytes to proinflammatory adhesion molecules expressed on the surface of endothelial cells and their subsequent infiltration into the subendothelial space (29). The cytokine-induced expression of VCAM-1 and ICAM-1 on endothelial cells and monocyte chemoattractant protein (MCP)-1-directed transendothelial migration of monocytes are both potentially inhibited by PPAR-γ ligands (29).

In our study, TZD treatment reduced activation of the transcription factor NF-κB, which is one of the major factors involved in ICAM-1 transcription. In the diabetic state, many factors such as AGEs, shear stress, and oxidative stress contribute to NF-κB activation (11). Hofmann et al. (15) indicated that insufficient glycemic control increases NF-κB activation in peripheral blood mononuclear cells isolated from patients with type 1 diabetes. In type 2 diabetic patients with obesity, TZD caused a reduction in intranuclear NF-κB binding activity and induced an increase in the expression of inhibitor-κB, which binds to NF-κB and reduces the ability of this transcription factor to move from the cytosol into the nucleus (1, 2).

Elevated TGF-β levels in glomeruli contribute significantly to the pathogenesis of diabetic glomerular lesions (19, 34). In our study, pioglitazone inhibited ICAM-1 expression on endothelial cells by suppressing NF-κB, leading to reductions in the number of recruited macrophages into diabetic glomeruli and in expression of TGF-β and type IV collagen. The TGF-β reduction may be due to a decreased number of macrophages because macrophages per se are known to secrete TGF-β and to stimulate mesangial cells to produce TGF-β (30). On the other hand, TZDs prevent PKC activity and reduce the expression of TGF-β and extracellular matrix proteins in diabetic glomeruli (18). Furthermore, TGF-β-induced fibronectin expression was inhibited by pioglitazone (12). Zheng et al. (45) found that PPAR-γ activation suppressed the increased expression of type I collagen mRNA and protein mediated by TGF-β in mesangial cells. Other studies have suggested that TZDs act directly on mesangial cells and proximal tubular cells and thus prevent tissue damage in the diabetic kidney (4, 12). In addition, some studies have suggested that TZDs prevent glomerular hyperfiltration and reduce albuminuria in diabetic animal models (18). In our experiments, the effect of TZD on renal hemodynamics could also play a role in renoprotection. To clarify the relevance of renal hemodynamics in renoprotective effects of pioglitazone, we measured the weekly creatinine clearance for 4 wk after induction of diabetes. Although there was no significant difference in the creatinine clearance of the three groups each week, we found a tendency that the creatinine clearance of DM groups and DM+pio groups was increased at 1 wk after induction of diabetes and pioglitazone suppressed these changes (data not shown). Suppression of glomerular hyperfiltration might be involved in the renoprotective effects of pioglitazone in early stage of diabetic nephropathy.

In conclusion, TZDs may exert renoprotective effects at least partly due to their anti-inflammatory actions through inhibition of NF-κB activation, ICAM-1 expression, and macrophage infiltration in the diabetic kidney.

GRANTS

This study was supported in part by a Grant-in-Aid for Scientific Research (C15590850 and C17590824 to K. Shikata) from the Ministry of Education, Science, Culture, Sports and Technology of Japan and Health Sciences Research Grants conducted by the Ministry of Health Labor and Welfare.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract