Thiazolidinediones are ligands for peroxisome proliferator–activated receptor (PPAR)-γ, widely used as insulin sensitizer in type 2 diabetic patients and implicated in apoptosis, cell proliferation, and cell cycle regulation. Here, the effect of thiazolidinediones on G1-phase cell cycle arrest, the hallmark in diabetic nephropathy, was investigated. Eight-week-old male Otsuka Long-Evans Tokushima fatty rats were treated with pioglitazone (1 mg · kg body wt−1 · day−1) until 50 weeks of age and compared with insulin treatment. Although similar HbA1c levels were observed in both groups, pioglitazone significantly inhibited glomerular hypertrophy and mesangial matrix expansion and reduced urinary albumin excretion compared with the insulin-treated group. In addition, pioglitazone significantly reduced the number of glomerular p27Kip1-positive cells. Because prominent expression of PPAR-γ was observed in podocytes in glomeruli and cultured cells, conditionally immortalized mouse podocyte cells were cultured under 5.5 and 25 mmol/l d-glucose supplemented with pioglitazone. Pioglitazone inhibited cell hypertrophy revealed by [3H]thymidine and [3H]proline incorporation, and pioglitazone reversed high glucose–induced G1-phase cell cycle arrest, i.e., an increase in G0/G1 phase and decrease in S and G2 phases. Pioglitazone suppressed high glucose–induced phosphorylation of p44/42 mitogen-activated protein kinase and reduced Bcl-2 and p27Kip1 protein levels. Besides glucose-lowering action, pioglitazone ameliorates diabetic nephropathy via cell cycle–dependent mechanisms.
The peroxisome proliferator–activated receptors (PPARs) PPAR-α, PPAR-δ, and PPAR-γ are a family of nuclear receptors that bind to fatty acid–derived ligands and activate the transcription of genes that govern various biological processes such as lipid metabolism, glucose metabolism, adipocyte differentiation, cell cycle, and apoptosis (1,2). PPAR-γ is mainly expressed in adipose tissues where it plays a role in lipid metabolism and adipogenesis and is also expressed in liver, skeletal muscle, intestine, colon, kidney, and cell types throughout the body, including monocytes and macrophages (3). The unsaturated fatty acids bind all three PPARs, whereas saturated fatty acids are poor PPAR ligands in general (4). The ligands for PPAR-γ include several prostanoids such as 15-deoxy-Δ (12,14)-prostaglandin J2 and 15-hydroxy-eicosatetrenoic acid, metabolites of arachidonic acid (5,6). The pharmacological ligands for PPAR-γ are the thiozolidinediones (TZDs), which are widely used as an insulin sensitizer in type 2 diabetic patients and have been shown to possess potent anti-inflammatory and antineoplastic actions (1).
PPAR-γ expressed in adipocytes plays a pivotal role in adipocyte proliferation and differentiation, i.e., cell cycle regulation. After hormonal induction, growth-arrested preadipocytes re-enter into the cell cycle, and after several rounds of clonal expansion, the adipocytes again arrest proliferation and undergo terminal differentiation. In the first hours of adipocyte differentiation, an increase in E2F transcription factor activity has been observed, and the expression of the genes, such as cyclin D1, c-Myc, or cyclin E is increased (7). Retinoblastoma protein (pRb) is involved in two phases of adipocyte differentiation; inactivation of pRb enables clonal expansion, whereas the growth arrest after the expansion requires active pRb. This function of pRb can be compensated by overexpression of C/EBP-α and PPAR-γ, which would mediate the cell cycle arrest after clonal expansion (1,8).
Studies based on various tumor cell lines, smooth muscle cells, and adipocytes have implicated PPAR-γ in cell cycle withdrawal. PPAR-γ activation decreased binding of E2F/DP heterodimers to its target genes, and PPAR-γ ligands are shown to inhibit phosphorylation of pRb and abrogate G1/S-phase transition in vascular smooth muscle cells (9,10). In addition, PPAR-γ upregulates the cyclin-dependent kinase (CDK) inhibitors p18 and p21 during adipogenesis and mediates G1 cell cycle arrest during adipogenesis (11). TZD treatment resulted in G1 cell cycle arrest, inhibition of proliferation, and apoptosis of various malignant cell lines such as human pancreatic cancer cell line, MCF-7 breast carcinoma cells, T24 bladder cancer cells, myelomonocytic U937 cells, and human gastric cancer cell line MKN45 (1,8).
PPAR-γ ligands have been shown to limit infarct size after experimental coronary occlusion (12) and to suppress the development of cardiac myocyte hypertrophy in both in vitro and in vivo settings (13,14). PPAR-γ agonists favorably affect cardiac gene expression in type 2 diabetic rats via activation and upregulation of PPAR-γ expression (15). Similar cellular hypertrophy is prominent in diabetic nephropathy, and its cell cycle–dependent mechanism has been well investigated. The cell cycle is dysregulated in the diabetic state, and G1-phase cell cycle arrest is believed to be responsible for the glucose-induced mesangial cell hypertrophy and increase in the de novo protein synthesis and consequential accumulation of extracellular matrix proteins (16). There is a growing body of evidence that specific CDK inhibitors p27Kip1 and p21Cip1 are critically involved in the G1-phase cell cycle arrest in cultured cells (17,18) exposed to high-glucose ambience, in experimental type 1 and type 2 diabetes (19,20), and also in p27Kip1-null (−/−) mice (21,22).
PPAR-γ agonist has been shown to inhibit diabetic glomerular hyperfiltration and albuminuria in streptozotocin-induced diabetic rats via diacylglycerol–protein kinase C–extracellular signal–regulated kinase (ERK) pathway (23) and to suppress transforming growth factor-β1–induced fibronectin expression in cultured mesangial cells (24). In the line of evidence, it is conceivable that TZDs ameliorate the diabetic nephropathy via cell cycle–dependent mechanism. In the current investigation, we examined the effect of long-term pioglitazone administration on diabetic nephropathy in Otsuka Long-Evans Tokushima fatty (OLETF) rats compared with OLETF rats treated with insulin or voluntary exercise. Furthermore, we evaluated cell cycle regulation in mouse podocyte cell (MPC) culture treated with high glucose and pioglitazone. We concluded that the administration of TZDs has an additional benefit via cell cycle–dependent mechanism, and podocytes are one of the major targets of TZDs in diabetic kidney disease.
RESEARCH DESIGN AND METHODS
Eight-week-old male OLETF rats, a model of type 2 diabetes, were purchased (Otsuka, Tokushima, Japan). OLETF rats were divided into four groups: OLETF rats (OLETF group, n = 10), OLETF rats subjected to voluntary exercise, i.e., free access to running wheel (voluntary exercise group, n = 10), OLETF rats treated with pioglitazone (pioglitazone group, n = 10), and OLETF rats treated with insulin (insulin-treated group, n = 10). Nondiabetic Long-Evans Tokushima Otsuka (LETO) rats fed with rodent diet (MF; Oriental Yeast, Tokyo, Japan) were used as controls. Pioglitazone was mixed in the diet, and the pioglitazone group orally received ∼1 mg pioglitazone · kg body wt−1 · day−1 during the whole period of the experiments. In the insulin-treated group, neutral protamine Hagedorn insulin (10–40 units) was administered to maintain fasting glucose levels of 80–150 mg/dl. We killed 25 rats at 30 weeks of age and 25 rats at 50 weeks of age and subjected them to following studies.
Analysis of metabolic data.
Blood samples were collected from tail vein after a 12-h fast, and blood glucose, HbA1c (A1C), serum creatinine, serum total cholesterol, serum triglyceride, and plasma insulin were measured at 30 and 50 weeks of age. Individual mice were placed in metabolic cages to obtain 24-h urine collections, and daily urinary albumin excretion levels were measured with nephelometric analysis (Organon Teknika-Cappel, Durham, NC). Plasma insulin levels were measured with mouse insulin radioimmunoassay (Linco Research, St. Charles, MO).
Renal tissues were fixed in 10% formaldehyde and embedded in paraffin, and 4-μm-thick sections were prepared. The sections were stained with periodic acid-Schiff (PAS). Glomerular tuft area and mesangial matrix area were measured with image analysis software (Optimas version 6.5; Media Cybernetics, Silver Spring, MD). The cross section yielding the maximum diameter of the glomerulus was photographed and converted into a digital image, and a total of 50 glomeruli were examined from each animal. In addition, ultrastructural examination of glomeruli was carried out using kidney tissues processed for electron microscopy as previously described (25).
Four-micrometer-thick sections of formalin-fixed, paraffin-embedded tissues were deparaffinized and rehydrated, and endogenous peroxidase was blocked by incubation in 3% hydrogen peroxide and methanol. Sections were pretreated by microwave for 20 min in citrate buffer for antigen retrieval. Nonspecific binding was blocked by incubation for 30 min in 10% rabbit serum. The tissues were then incubated with transforming growth factor (TGF)-β1, anti-p27Kip1, and p21Cip1 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). After PBS wash, sections were incubated with a biotinylated secondary antibody, ABC-Elite Reagent (Vector Laboratories, Burlingame, CA). The percentage of p27Kip1- and p21Cip1-positive cells in all glomerular cells was evaluated, and at least 50 glomeruli in each animal were analyzed.
Four-micrometer frozen sections were prepared and fixed with cold acetone for 3 min. For evaluation of mesangial matrix accumlation, 1:100 diluted rabbit anti-rat type IV collagen polyclonal antibody (Sigma, St Louis, MO) was used for primary antibody and followed by incubation with fluorescein isothiocyanate–labeled goat anti-rabbit IgG (Chemicon, Temecula, CA). Digital images were obtained using confocal laser fluorescence microscope (LSM-510; Carl Zeiss, Jena, Germany), and type IV collagen expression was quantified with the formula (density indicated by 0–255 in gray scale × positive area μm2) as previously described (26).
To proliferate MPC lines (27) provided by Dr. Peter Mundel, the cells were cultured on a type I collagen–coated flask (BD Falcon) with RPMI-1640 (Invitrogen, Carlsbad, CA) containing 10% FCS (Invitrogen) and 50 units/ml recombinant mouse interferon-γ (BD Biosciences, Palo Alto, CA) at 33°C. After confluence, the cells were induced to differentiate into podocyte lineage by shifting them to 37°C and culturing in Dulbecco’s modified Eagle’s medium (DMEM) (Sigma) containing 10% FCS without interferon-γ, i.e., nonpermissive conditions. After 7 days under nonpermissive conditions, we successively cultured MPC under nonpermissive conditions in DMEM containing 10% FCS. After 7 days of culture of MPC under nonpermissive conditions, we further cultured differentiated MPC in DMEM containing 0.5% FCS (Invitrogen) for 24 h, and quiescent mature podocytes were incubated with 5.5 mmol/l normal glucose, 25 mmol/l high glucose, high glucose with 100 nmol/l pioglitazone, high glucose with 10 μmol/l pioglitazone, and 5.5 mmol/l glucose plus 19.5 mmol/l mannitol for 1 day. They were then subjected to the following studies, i.e., cell cycle analysis, [3H]thymidine and [3H]proline incorporation, and Western blot analysis.
SV40 MES 13 (glomerular mesangial cells from an SV40 transgenic mouse, CRL-1927; American Type Culture Collection, Rockville, MD) was grown in a mixture of DMEM, Ham’s F12 medium, and 5% fetal bovine serum as previously described (28).
Cell cycle analysis by laser scanning cytometer.
MPC seeded on type I collagen–coated chamber slides (BD Falcon) were fixed with 100% ethanol for 15 min, and then DNA was stained with 50 μg/ml propidium iodide (Sigma) and 200 μg/ml RNase A (Sigma) for 15 min at 37°C avoiding the light. RNase A was boiled at 95°C for 10 min before use. Stained cells were analyzed on a laser scanning cytometer (Olympus Optical, Tokyo, Japan) by measuring total and peak intensity of propidium iodide fluorescence in each nucleus (29). We analyzed ∼4,000 unperturbed MPC populations and generated full and detailed resolution of the cell cycle, i.e., percentage of the cells in G0/G1, S, G2, and M.
[3H]thymidine and [3H]proline incorporation.
Cells were pulse-radiolabeled with [3H]thymidine or [3H]proline (1 μCi/ml; Amersham Pharmacia, Piscataway, NJ) at 6 and 18 h in before the end of culture, respectively. Cells were then washed with PBS, incubated with 10% ice-cold trichloroacetic acid for 30 min at 4°C, and solubilized in 0.5 mol/l NaOH, and the incorporation was counted by liquid scintillation counter (TRI-CARB 2300TR; Packard, Meriden, CT).
At the end of MPC culture, cells were washed with PBS and homogenized with lysis buffer (20 mmol/l Tris-HCl, pH 7.4, 100 mmol/l NaCl, 10 mmol/l benzamidine-HCL, 10 mmol/l ε-amino-n-caproic acid, 2 mmol/l phenylmethylsulfonyl fluoride, and 1% Triton X-100). Forty micrograms protein was subjected to SDS-PAGE under reducing conditions and electroblotted onto Hybond P polyvinylidine fluoride membranes (Amersham Biosciences, Piscataway, NJ). The membrane blots were immersed in a blocking solution containing 5% nonfat dry milk and Tris-buffered saline with Tween (0.05% Tween 20, 20 mmol/l Tris-HCl, and 150 mmol/l NaCl, pH 7.6). Then, membranes were incubated with rabbit polyclonal anti-p27Kip1 in a 1:100 dilution (Santa Cruz Biotechnology), anti–p44/42 mitogen-activated protein kinase (MAPK) in a 1:1,000 dilution (Cell Signaling Technology, Beverly, MA), anti–β-actin antibody in a 1:500 dilution (Sigma), and mouse monoclonal anti–phospho p44/42 MAPK in a 1:1,000 dilution. They were then incubated with anti-rabbit IgG conjugated with horseradish peroxidase in a 1:20,000 dilution (Amersham) for polyclonal antibodies and anti-mouse IgG conjugated with horseradish peroxidase in a 1:20,000 dilution (Amersham) for monoclonal antibodies. The blots were washed three times with Tris-buffered saline with Tween, immersed in ECL Plus Western Blotting Detection Reagents (Amersham), and then exposed to Hyperfilm ECL (Amersham).
Quantitative real-time PCR.
Total RNA was purified from glomeruli isolated by sieving method and cultured cells using QIAzol Reagent (Qiagen, Hilden, Germany). Murine mesangial cell line (MES 13) (American Type Culture Collection), MPC, and human glomerular endothelial cells (Cell System, St. Katarinen, Germany) were cultured by the instruction of manufacturers. cDNAs were synthesized from 1 μg of total RNA and analyzed using LightCycler-FastStart DNA master SYBR Green I system (Roche Diagnostics, Basel, Switzerland) and specific primers. The relative abundance of mRNAs was standardized with β-actin mRNA as the invariant control. The primer sets for mouse PPAR-α, PPAR-γ, and TGF-β1 were purchased from Nihon Gene Research Lab (Sendai, Japan). The primers for PPAR-δ were purchased from Takara BIO (Otsu, Japan).
Data are expressed as the means ± SE and analyzed by the unpaired Student’s t test or a one-way ANOVA by Fisher’s t test when multiple comparisons against the control were required. P < 0.05 was regarded as statistically significant. The data were analyzed with Dr. SPSS II for Windows, release 11.0.1J.
Pioglitazone reduces urinary albumin excretion in OLETF rats.
The body weight of OLETF rats peaked at 30 weeks of age, the time notable hyperinsulinemia was observed (Fig. 1A and B). Thereafter, OLETF rats progressively lost weight for the next 20 weeks along with the decline of plasma insulin levels and rise in A1C, fasting blood glucose, cholesterol, and triglyceride levels (Fig. 1B–F). In the pioglitazone- and insulin-treated groups, OLETF rats progressively gained body weight, and endogenous plasma insulin levels gradually increased during the entire experimental period in the pioglitazone group. At 50 weeks of age, plasma insulin levels in the pioglitazone group were higher than in the other groups. The increment in body weight over the experimental period was prominent in the pioglitazone-treated group compared with insulin-treated OLETF rats. Plasma glucose and A1C levels in the pioglitazone- and insulin-treated groups were comparable with the LETO and voluntary exercise groups (Fig. 1C and D). In contrast, the administration of insulin did not alter the lipid profiles in OLETF rats, whereas pioglitazone significantly reduced both cholesterol and triglyceride levels similar to those in LETO and voluntary exercise rats (Fig. 1E and F). There were no significant differences in blood pressure in the voluntary exercise (116 ± 7.12 mmHg), pioglitazone-treated (124 ± 5.63 mmHg), and insulin-treated (126.8 ± 12.9 mmHg) groups throughout the study period.
Creatinine clearance decreased in OLETF rats compared with LETO rats (Fig. 1G), whereas daily urinary albumin excretion significantly increased in OLETF rats compared with LETO rats at 50 weeks of age (Fig. 1H). Pioglitazone treatment significantly reduced urinary albumin excretion in OLETF rats to similar levels of voluntary exercise rats at 30 and 50 weeks of age. Insulin administration also reduced albumin excretion; however, it was not statistically significant.
Pioglitazone ameliorates glomerular hypertrophy in OLETF rats.
PAS-positive mesangial matrix area and glomerular size of OLETF rats significantly increased compared with LETO rats at 30 weeks of age, and prominent nodular-like lesion appeared in OLETF rats at 50 weeks of age (Fig. 2A and C). Morphometric analysis clearly indicated that pioglitazone significantly reduced both glomerular cross-sectional area and the mesangial matrix index, i.e., the ratio of mesangial matrix area divided by tuft area (Fig. 2B and D). The therapeutic effect of pioglitazone was almost comparable with the voluntary exercise group; however, in the insulin-treated group, glomerular hypertrophy was not significantly inhibited, and apparent sclerotic glomeruli were observed at 50 weeks. The amelioration of the expansion of mesangial matrix by the treatment of pioglitazone was also confirmed by electron microscopy (Fig. 3A–E). At 30 weeks of age, the mesangial matrix areas were reduced both in pioglitazone and voluntary exercise groups compared with OLETF rats, and the therapeutic effect of insulin on mesangial matrix expansion did not surpass the effect of pioglitazone and voluntary exercise. At 50 weeks of age, interstitial lesions such as fibrosis and tubular atrophy were prominent in OLETF rats, and such interstitial changes were ameliorated in voluntary exercise and pioglitazone groups. In the insulin-treated group, interstitial changes were not ameliorated as voluntary exercise and pioglitazone groups (Fig. 2E).
Pioglitazone ameliorates glomerular type IV collagen and TGF-β1 in OLETF rats.
At the 30 weeks of age, type IV collagen–positive area in glomeruli significantly increased in OLETF rats compared with LETO rats (Fig. 4A). Morphometric analysis clearly indicated that pioglitazone significantly reduced the type IV collagen index, the area summation of pixel density. Although insulin treatment reduced the type IV collagen index, it did not enter the statistically significant levels (Fig. 4B). TGF-β1–positive area of OLETF rats significantly increased compared with LETO rats (Fig. 4C). Pioglitazone significantly reduced TGF-β1–positive staining area like voluntary exercise and insulin-treated groups. Similarly, glomerular mRNA expression of TGF-β1 was significantly upregulated in OLETF rats compared with LETO rats (Fig. 4D), and it was significantly reduced in voluntary exercise, pioglitazone, and insulin-treated groups.
Pioglitazone reduced glomerular and interstitial ED-1–positive cells in OLETF rats.
The intraglomerular and interstitial infiltration of macrophages was evaluated with ED-1 staining. The number of infiltrated ED-1–positive cells significantly increased in OLETF rats compared with LETO rats both in glomeruli (Fig. 5A and B) and renal interstitium (Fig. 5C and D). The treatment with pioglitazone and voluntary exercise in OLETF rats significantly reduced ED-1–positive glomerular and interstitial cells.
Pioglitazone decreased glomerular p27Kip1-positive cells in OLETF rats.
The expression of CDK inhibitors (p27Kip1 and p21Cip1) was evaluated. The glomerular expression of CDK inhibitors, p27Kip1- and p21Cip1-positive cells, significantly increased in OLETF rats compared with LETO rats (Fig. 6A–D). p27Kip1- and p21Cip1-positive cells consisted of both podocytes and mesangial cells (Fig. 6A and C). Although tubular cells revealed positive staining for p27Kip1 and p21Cip1, there was no significant difference between OLETF and LETO rats (data not shown). The treatment with pioglitazone and voluntary exercise in OLETF rats significantly reduced both p27Kip1- and p21Cip1-positive glomerular cells; however, insulin treatment did not significantly decrease p27Kip1- and p21Cip1-positive cells (Fig. 6B and D).
PPAR-α/δ/γ mRNA expression in podocyte and mesangial cell lines.
To quantify mRNA expression of PPARs, total RNAs were isolated from mesangial cell line (MES 13), MPC line, and human glomerular endothelial cells, and mRNA expression was evaluated by real-time RT-PCR.
In MES 13, the treatment with high glucose significantly upregulated mRNA expression of PPAR-α/δ/γ. Pioglitazone treatment upregulated mRNA expression of PPAR-α/δ/γ in MES 13 with normal glucose condition; however, it reversed high glucose–induced upregulation of PPAR-α/δ/γ mRNA (Fig. 7A–C).
Prominent mRNA expression of PPAR-γ was observed in MPC (7.45 ± 0.2, relative mRNA ratio) compared with MES 13 (1.0 ± 0.2) and human glomerular endothelial cells (0.5 ± 0.12). In contrast, mRNA expression of PPAR-α and PPAR-δ was prominent in MES 13, and relatively less mRNA expression was observed in MPC (Fig. 7A–C). Although pioglitazone treatment upregulated mRNA expression of PPAR-γin MPC, it did not alter the high glucose–induced upregulation of PPAR-γ mRNA (Fig. 7C).
The localization of PPAR-γ expression in glomeruli was evaluated in LETO, OLETF, and pioglitazone groups at 30 weeks of age. PPAR-γ–positive glomerular cells were observed in all three cell types, i.e., mesangial cells, podocytes, and glomerualr endothelial cells in LETO rats. PPAR-γ–positive cells were significantly increased especially in podocytes in OLETF and pioglitazone groups (Fig. 7D).
Pioglitazone assists cell cycle progression in MPCs with high glucose–induced G1 cell cycle arrest.
To investigate whether pioglitazone affects high glucose–induced cell cycle arrest in vitro, we cultured differentiated MPC in DMEM containing 0.5% FCS (Invitrogen) for 24 h, and quiescent mature podocytes were incubated with 5.5 mmol/l normal glucose, 25 mmol/l high glucose, high glucose with 10 μmol/l pioglitazone, high glucose with 100 nmol/l pioglitazone, and 5.5 mmol/l normal glucose plus 19.5 mmol/l mannitol for 1 day. The cell cycle distribution of MPC was then analyzed by laser scanning cytometer after staining with propidium iodide. After serum deprivation, 94.8% of MPC in high glucose was in G0/G1 phase, 1.1% in the S phase, and 2.7% in G2 phase; whereas 89.4 and 90.8% of MPC was in the G0/G1 phase under normal glucose and mannitol, respectively (Fig. 8). Incubation with pioglitazone altered cell cycle distribution, i.e., reduction in number of the cells in G0/G1 phase and increase in S and G2 phases, in a dose-dependent manner. Under high-glucose condition, pioglitazone reversed G1 cell cycle arrest independent of a glucose-lowering effect (Fig. 8).
Pioglitazone reduces high glucose–induced protein synthesis of MPC.
To assess the cellular hypertrophy of MPCs, DNA and protein synthesis was assessed by [3H]thymidine and [3H]proline incorporation. [3H]thymidine incorporation was inhibited under the condition of high glucose, and treatment of pioglitazone partially recovered [3H]thymidine incorporation to MPCs without statistical significance (Fig. 9A). In contrast, high glucose increased de novo protein synthesis, i.e., [3H]proline (dpm)/cellular protein (μg), and pioglitazone significantly reduced protein synthesis in a dose-dependent manner (Fig. 9B). Hyperosmotic condition with mannitol did not alter the [3H]thymidine and [3H]proline incorporation.
Pioglitazone reduces p27Kip1 via p44/42 MAPK and bcl-2–dependent pathway.
In MPC culture, high glucose upregulated p27Kip1 protein levels, whereas hyperosmotic stimulus with mannitol did not change the expression of p27Kip1. High-glucose environments also activate several protein kinase pathways, including MAPK and ERK pathways and it has been reported that high glucose stimulates p44/42 MAPK (Erk 1,2) that phosphorylates p27Kip1 protein. Thus, we further assessed phosporylation status of p44/42 MAPK in cultured MPCs. High glucose induced phosphorylation of p42 MAPK using phospho-specific anti–p44/42 MAPK antibody (Fig. 9C). Osmotic stimuli by mannitol did not alter the phosphorylation of both p44/42 MAPK. It has been reported that high-glucose environments also activate Bcl-2, the overexpression of which leads to upregulation of p27Kip1, and thus, we further assessed Bcl-2 expression in cultured MPCs. High glucose increased the expression of Bcl-2 (Fig. 9A), and osmotic stimuli by mannitol did not alter the expression of Bcl-2.
The treatment with pioglitazone inhibited expression of p27Kip1, phosphorylation of p42 MAPK, and expression Bcl-2 to the basal levels. Taken together, pioglitazone inhibited the phosphorylation of p44/42 MAPK, protein levels of Bcl-2, and thus protein levels of p27Kip1, and it may be one of the molecular mechanisms of therapeutic potential of pioglitazone on high glucose–induced hypertrophy of podocytes.
Morphometric analyses on renal specimens from type 1 and type 2 diabetic patients revealed characteristic cell hypertrophy, particularly of mesangial cells in early diabetic nephropathy (30,31). Hypertrophy is biochemically defined as an increase in protein synthesis without active DNA replication (30,31). High glucose stimulates the expression of TGF-β (22,32) and connective tissue growth factor (17,33), and both factors activate MAPKs (18,34,35) in the process of mesangial cell hypertrophy (30). The activation of MAPKs leads to increased transcriptional activity of cyclin D and induction of CDK inhibitors through transcriptional (p21Cip1) and posttranslational (p27Kip1) mechanisms. Mesangial cells reenter the cell cycle but fail to progress through G1/S-phase because of induction of CDK inhibitors. Hypertrophy of mesangial cells is a well-recognized hallmark in diabetic milieu; however, recent studies demonstrated that podocyte hypertrophy is observed in differentiated podocyte cell line and Zucker diabetic fatty rats (36) and db/db mice revealed by electron microscopic observations. In addition, increase in p21Cip1 and p27Kip1 was predominantly observed in podocytes in Zucker diabetic fatty rats and also in db/db mice by immunohistochemistry (36). In type 1 and type 2 diabetes, a decrease in podocyte number is well correlated with both microalbuminuria and disease progression of diabetic nephropathy (37). Recently, it has been reported that angiotensin II receptor blocker inhibits p27Kip1 expression in glucose-stimulated podocytes and diabetic glomeruli (38). Thus, high glucose–mediated podocyte injury associated with G1-phase cell cycle arrest takes center stage in the pathogenesis of diabetic glomerulopathy.
In OLETF rats, chronic administration of pioglitazone reduced albuminuria and inhibited glomerular hypertrophy, accumulation of extracellular matrix, and infiltration of macrophages into renal tissues. The administration of pioglitazone inhibited protein expression of type IV collagen and mRNA levels of TGF-β. In OLETF rats, immunohistochemistry indicated that intraglomerular p27Kip1- and p21Cip1-positive cells predominantly increased compared with nondiabetic LETO rats. The administration of pioglitazone reduced such p27Kip1-positive cells in glomeruli in OLETF rats. Because pioglitazone reduced plasma glucose levels, the reduction of glomerular p27Kip1-positive cells may be due to the improvement of plasma glucose levels. Similar control of A1C levels were observed in pioglitazone, voluntary exercise, and insulin-treated groups; however, the improvement of glomerular hypertrophy and accumulation of mesangial matrix compared with OLETF rats was prominent in voluntary exercise and pioglitazone groups. In addition, p27Kip1-positive cells in glomeruli were significantly reduced in voluntary exercise and pioglitazone groups. Thus, the administration of pioglitazone has an additional benefit in the treatment of diabetic nephropathy in addition to glycemic control.
Because PPAR-γ predominantly expressed in podocytes in OLETF rats and cultured cells compared with mesangial cells and glomerular endothelial cells, MPC line was used to further investigate whether therapeutic effects are cell cycle dependent (27). Most of the standard cell culture methods induce cellular proliferation but provoke the loss of differentiation, and they are not suitable for the investigation of high glucose–induced G1-phase cell cycle arrest. However, MPCs, conditionally immortalized cells, divided under permissive condition and then fully differentiated into mature podocyte in nonpermissive conditions (27). Furthermore, a laser scanning cytometer enabled us to scan MPCs on the glass slides and measure total and peak intensity of propidium iodide fluorescence in nuclei, by which a full and detailed resolution of the cell cycle of unperturbed MPC population was obtained (29). Pioglitazone treatment normalized high glucose–induced G1-phase cell cycle arrest in a dose-dependent manner, and pioglitazone treatment caused successful replication of MPCs. By laser scanning cytometer, DNA amount and condensation can be observed, and we cannot find apoptotic death of MPCs by pioglitazone treatment. By cell cycle–dependent mechanism, pioglitazone ameliorated MPC hypertrophy, because high glucose–induced de novo protein synthesis was inhibited and DNA replication was unaltered.
As shown in OLETF rats, high-glucose condition induced p27Kip1 protein levels in MPC, and pioglitazone reduced its protein levels. High glucose–induced G1-phase cell cycle arrest is mediated by MAPKs that upregulate CDK inhibitors (18,35,39), and we further checked activation of p44/42 MAPK (18). p44/42 MAPK was activated by high-glucose condition, and pioglitazone reduced phosphorylation of p44/42 MAPK. Thus, pioglitazone reduced p27Kip1 protein via the p44/42 MAPK-dependent pathway in high-glucose conditions. Another important molecule in cell cycle regulation is Bcl-2, and its overexpression resulted in the accelerated G1 cell cycle arrest and delayed G1/S transition (40). In the current study, high-glucose conditions upregulated the Bcl-2 expression in MPC, and the addition of pioglitazone in the culture media decreased expression of Bcl-2. Because Bcl-2 acts directly to increase levels of p27 to inhibit cell cycle progression in quiescent cells (40), high glucose–induced Bcl-2 overexpression may contribute to upregulation of p27 and the G1 cell cycle arrest in MPC. The gene encoding Bcl-2 has peroxisome proliferator-responsive element–and PPAR-γ–increased bcl-2 protein and mRNA; however, administration of TZDs decreased levels of bcl-2 and induced apoptosis independently of PPAR-γ in various cells such as granulosa cells and prostate cancer cells (41,42). In MPC culture with high-glucose condition, the amelioration of overexpression of bcl-2 may contribute to the reversal of high glucose–induced p27 and p44/42 MAPK and successful progression of G1/S cell cycle.
Here, we have shown that pioglitazone promotes and assists cell cycle progression and successful replication in diabetic state, where cell cycle progression is halted despite of cell cycle entry. The successful cell cycle progression may be dependent on amelioration of plasma glucose levels as well as the glucose-independent mechanism revealed by comparing the renal histology in OLETF rats treated with insulin and pioglitazone. The MPC culture experiments revealed that pioglitazone inhibited cell cycle–dependent hypertrophy of the podocytes by reducing p27Kip1 protein levels by downregulation of p44/42 MAPK and Bcl-2. Although antiproliferative and apoptotic properties by inducing G1 cell cycle arrest are associated with PPAR-γ activation by TZDs in various cell types, it seems that the effect of TZDs on cell cycle, proliferation, differentiation, and apoptosis depends on the cell type or conditions such as hyperglycemia. In this investigation, we newly suggest an application of TZDs to control cell cycle–dependent cellular hypertrophy observed in metabolic syndrome such as renal cell hypertrophy in diabetes, cardiac hypertrophy in hypertension, and smooth muscle cell hypertrophy in atherosclerosis.
J.W. has received Grant-in-Aid for Scientific Research (C) from the Ministry of Education, Science and Culture, Japan (14571025, 17590829). H.M. has received Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Science and Culture, Japan (18390249). This work has received support from the Uehara Memorial Foundation, The Naito Foundation, the ONO Medical Foundation, and the Japan Heart Foundation/Pfizer Grant for Research on Hypertension, Hyperlipidemia, and Vascular Metabolism.