Diabetic nephropathy (DN) eventually develops in ∼30% of patients with both type 1 and type 2 diabetes, and once present will progress in many of these individuals to end stage renal disease. Because of the rapidly increasing prevalence of type 2 diabetes, this is leading to a massive increase in the incidence of end stage renal disease in the United States and elsewhere (Ritz, 1999). Blockade of the angiotensin system is currently the standard treatment for slowing progression of renal disease in DN (Hollenberg, 2002), although this has no inherent beneficial effect on the diabetes. Where angiotensin-converting enzyme inhibitors (ACEI) were first used, these are now being replaced, or supplemented with angiotensin type 1 receptor blockers. Nevertheless, the incidence of DN in type 2 diabetes continues to increase (Ritz, 1999).

In recent years, it has also become evident that prolonged tight glycemic control slows down progression of DN in type 1 diabetics (The Diabetes Control and Complications Research Group, 1995), and some (but not all) studies suggest that good metabolic control is protective to DN in type 2 diabetes (Di Landro et al., 1998). The recent development of the peroxisome proliferator-activated receptor γ (PPARγ) agonists has provided a novel means of improved glycemic control. These drugs (the thiazolidinediones) enhance sensitivity of the peripheral insulin receptor to glucose in type 2 diabetes, leading to reduced plasma glucose and insulin levels and reductions in glycosylated Hb. These drugs also exert a beneficial effect on the lipid profile, antihypertensive, and inhibit collagen type 1 production by the glomerular mesangial cells (Guan and Breyer, 2001; Rosak, 2002).

The present study was conducted to test the relative benefit of long-term (6-month) ACEI versus PPARγ agonist therapy as well as in combination, in a rat model of severe type 2 diabetes, the inbred obese Zucker (ZDF/Gmi, fa/fa) rat. When untreated these rats develop marked kidney damage compared with their lean littermates (Vora et al., 1996; Baylis, 2001).

Materials and Methods

Studies were conducted on 30 inbred obese male Zucker (ZDF/Gmi, fa/fa) rats purchased from Genetic Models, Inc. (Indianapolis, IN) at age 5 weeks. Baseline measurements of blood glucose and urinary protein excretion (see below) were made at 6 weeks of age, and at 8 weeks of age rats were randomized to one of the following four groups and followed for 6 months: 1) controls (n = 9); 2) rats treated with PPARγ agonist rosiglitazone (Rosi, oral; 3 mg/kg/day) (n = 6); 3) rats treated with ACE inhibitor cilazapril (oral 10 mg/kg/day) (n = 9); and 4) rats treated with both drugs (n = 6). For the two different drugs, submaximally effective doses were selected based on their specific pharmacological effects (reduction of insulin resistance for PPARγ agonist; antihypertensive effect for ACEI). One Rosi-treated and two control rats died during the 6-month observation period and were not included in any analysis. Body weight and water intakes were measured every 48 h and the drug concentration adjusted according to water consumption, to give a constant dose of drug.

All rats were allowed water and were fed Purina 5008 diet ad libitum, to maintain a hyperglycemic state in the obese rats. Eight-hour urine collections were made at monthly intervals and urine was frozen for later analysis. Tail vein blood samples were taken at two monthly intervals during the study for analysis of blood glucose and blood urea nitrogen. After 6 months of observation, rats were prepared for terminal acute micropuncture studies for measurement of glomerular blood pressure and renal hemodynamics. Anesthesia was induced with i.p. Inactin; for the grossly obese rats treated with Rosi, an induction dose of 180 mg/kg b.wt. was given followed by frequent i.v. supplementation (40 mg/kg), as needed. Controls and cilazapriltreated rats received the regular i.p. induction dose of 120 mg/kg with i.v. supplementation at 20 mg/kg, as required. Both jugular veins were catheterized, one for i.v. Inactin and one for an infusion of 0.9% NaCl (∼0.5% b.wt.) + [³H]inulin (2 μC/ml) and PAH (0.5 g/100 ml) was infused at 50 μl/kg b.wt. via T piece. Before starting the NaCl/inulin/PAH infusion, an arterial cannula was placed in the right carotid artery, and a 0.2-ml blood sample was removed for later estimation of plasma creatinine. Then the trachea was cannulated, the abdomen opened by a ventral midline incision, and the urinary bladder was catheterized. The perirenal fat was gently cleared from the left kidney, and the kidney was stabilized for micropuncture as described previously (Deng and Baylis, 1995). Paraffin wax blocks were inserted into 5 to 12 midproximal tubule segments on the kidney surface, so that stop-flow pressure could be measured proximal to the block.

After an initial bolus and 40-min infusion (∼2 ml/h) of a solution of 0.9% NaCl containing [³H]inulin and PAH (PAH concentration in the infusate was varied according to the group so that plasma PAH was always below 4 mg/dl), two timed 15- to 20-min urine collections were made with midpoint arterial blood samples (0.2 ml). During whole kidney clearance periods, measurements were made of stop-flow pressure and free flow proximal tubule pressure. Then, the rat was sacrificed, and the kidneys, heart, liver, and pancreas were removed and fixed in 10% buffered formalin for at least 24 h for later histological and immunohistological analysis (kidneys only). Measurement of urine creatinine was by kit 555 (Sigma-Aldrich, St. Louis, MO). Blood glucose was measured by glucose Trinder reagent (Sigma-Aldrich) (Trinder 1969). PAH was analyzed colorimetrically, [³H]inulin activity was determined by scintillation counting, and plasma protein concentration by refractometer, as described previously (Smith et al., 1945; Deng and Baylis, 1995).

For histology and immunohistochemistry, kidneys were trimmed, processed, and embedded in Paraplast. Kidney sections (2–3 μm in thickness) were cut and stained with H&E or periodic acid Schiff reagent. Immunostaining for expression of CD 68 using the monoclonal antibody Ki-M6 specific for monocytes/macrophages (BMA T-1005, 1:100 dilution; Biomedicals AG, Switzerland) (Parwaresch et al., 1986) was performed on kidney sections of four animals each per treatment group. An avidin biotin peroxidase complex technique was applied according to the manufacturer's specifications (ready-to-use detection kits and an automated slide stainer; Ventana Medical System, Strasbourg, France). Renal changes were scored on a scale of 1+ to 4+ (1+, minimal; 2+, slight; 3+, moderate; 4+, marked to massive) as adapted from Zbinden (1976). Tubulointerstitial changes were defined as tubular degeneration, regeneration, and/or atrophy, tubular dilation, hyaline casts, interstitial inflammatory infiltrates, or fibrosis. In addition, for quantification of glomerular sclerosis, kidney sections (5 μm thick) were cut and stained with periodic acid Schiff reagent as described by us previously (Deng and Baylis, 1995). Glomerular sclerosis was defined as segmental or global increases in glomerular matrix, accumulation of hyaline material, and loss of local capillary organization. All data are given as mean ± S.E. Statistical significance was assumed when p < 0.05, using one- and two-way analysis of variance and Wilcoxon Rank sum test for quantitative assessment of glomerular injury. All other pathology findings were expressed as “summary incidence of grading” for each finding.

Results

As shown in Fig. 1, blood glucose was normal in obese rats at 6 to 8 weeks of age (time 0) but rose markedly in untreated animals and reached a stable elevated level by age 14 to 16 weeks (month 2). A similar profile was seen in the rats treated with ACEI, whereas blood glucose remained normal in both groups of rats receiving PPARγ agonist. All PPARγ agonist-treated rats were massively obese and body weight increased profoundly over the 6-month treatment period (Fig. 2), in some cases exceeding 1 kg before sacrifice. In contrast, the untreated rats and rats given ACEI alone showed only moderate weight gain with age. We did not evaluate body composition, but the 24-h urinary creatinine excretion gives an index of skeletal muscle mass. The absolute values were similar in all four groups but when factored for body weight creatinine excretion was significantly higher in untreated or ACEI-treated rats (33.1 ± 3.5 and 29.1 ± 1.6 mg/24 h/kg b.wt.) versus those given PPARγ agonist alone or in combination (11.8 ± 2.3 and 13.4 ± 2.4; p versus untreated [lt]0.001 for both), suggesting greatly increased percentage of adipose tissue in the PPARγ agonist-treated animals. Unfortunately, the high, variable levels of plasma lipids prevented meaningful plasma creatinine measurements.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Plasma glucose (milligrams per deciliter) at 6 to 8 weeks of age (before initiation of treatment) and over the 6-month observation period in untreated (open circles), ACEI-treated (closed circles), PPARγ agonist-treated (open triangles), and PPARγ agonist + ACEI-treated (closed triangles) rats.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Body weight (grams) at 6 to 8 weeks of age (before initiation of treatment) and over the 6-month observation period in untreated (open circles), ACEI-treated (closed circles), PPARγ agonist-treated (open triangles), and PPARγ agonist + ACEI-treated (closed triangles) rats.

The total urine protein excretion rose markedly in the untreated rats, reflecting development of progressive kidney damage (Fig. 3) and was significantly higher (p < 0.05) than all other groups at 6 months of observation. The ACEI was effective in reducing urinary protein excretion although significant proteinuria (p < 0.01 versus both PPAR Groups) was still evident after 6-months treatment. In contrast, rats given PPARγ agonist (either ± ACEI), developed no proteinuria over the entire treatment period.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Total urinary protein excretion (milligrams per 24 h) at 6 to 8 weeks of age (before initiation of treatment) and over the 6-month observation period in untreated (open circles), ACEI-treated (closed circles), PPARγ agonist-treated (open triangles), and PPARγ agonist + ACEI-treated (closed triangles) rats.

Histological evaluation of the kidneys revealed pronounced glomerular sclerosis in untreated rats with some level of injury in 26 ± 1% of glomeruli. Rats receiving ACEI alone had less damage (17 ± 1%; p < 0.05 versus control), but this was still greater than in either of the PPARγ agonist-treated groups (PPARγ agonist alone, 9 ± 1% damaged glomeruli; p < 0.01 versus ACEI alone, and PPARγ agonist + ACEI, 10 ± 1%; p < 0.01 versus ACEI alone). Figure 4 shows that in addition to a higher percentage of the total glomeruli being affected by damage, the grade of injury is also greater in the untreated rats, with significantly more 3+ and 4+ damage versus ACEI alone (p < 0.05), who in turn have more 2+,3+, and 4+ injury (p < 0.05) than either PPARγ agonist-treated group.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Percentage of damaged glomeruli in each of the four groups at each level of severity of damage. 1+ denotes less that 25% of the glomerulus damaged; 2+ denotes 26 to 50% damage, 3+ denotes 51 to 75% damage, and 4+ denotes 76 to 100% damage.

In untreated rats, in addition to pronounced glomerular sclerosis, moderate tubulointerstitial changes were observed. These consisted of tubular atrophy, degeneration and regeneration, tubular dilation, hyaline casts, and interstitial mononuclear cell infiltrates and fibrosis. Slight epithelial cell vacuolation was observed in distal tubules. These vacuolated cells looked swollen and clear with prominent cell boundaries and condensed nuclei (Fig. 5A). Treatment with either PPARγ agonist or ACEI alone or with a combination of both drugs clearly reduced the development of renal lesions (Fig. 5, B–D). In addition to the drug-induced reduction in the incidence and severity of glomerular sclerosis, associated tubulointerstitial changes were improved in parallel. The tubulointerstitial protection was greater in rats receiving a combination of PPARγ agonist and ACEI than in animals given PPARγ alone and was the least pronounced in animals receiving ACEI alone. Treatment with PPARγ agonist alone or in combination with ACEI completely inhibited development of tubular epithelial cell vacuolation. In contrast, treatment with ACEI alone had no impact on the incidence and severity of tubular epithelial cell vacuolation. The incidence of cells expressing CD 68 within the glomeruli using the monocyte/macrophage marker Ki-M6 (Parwaresch et al., 1986) was slight in untreated rats, minimal in rats given PPARγ agonist alone or ACEI alone, and absent in rats receiving the combination of ACEI and PPARγ agonist. Hyperplasia of the juxtaglomerular apparatus was minimal in rats given ACEI as reported by others (Owen and Molon-Noblot, 1998) and was marked in rats receiving the combination of ACEI and PPARγ agonist, with the change extending along afferent arterioles to affect interlobular arteries. The reason for this greater severity of hyperplasia of the juxtaglomerular apparatus in the combination group is unknown. Pelvic dilation was present in untreated rats and in all drug-treated groups to a similar extent.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Renal cortex of obese ZDF fa/fa rats, H&E staining. Scale bar, 200 μm. A, untreated rat showing tubular atrophy, tubular dilation, hyaline casts, interstitial fibrosis and inflammatory infiltrates, and glomerular sclerosis. B, PPARγ agonist-treated rat, normal morphology. C, ACEI-treated rat, focal tubular degeneration and regeneration, tubular dilation with hyaline casts, glomerular sclerosis, and hyperplasia of juxtaglomerular apparatus. D, combination treatment of PPARγ agonist and ACEI: normal morphology except for hyperplasia of juxtaglomerular apparatus.

The data given in Table 1 was derived from the acute study conducted at the end of the 6-month treatment period. Arterial Hct was lower in both groups of rats receiving PPARγ agonist than in untreated or ACEI-treated rats and because the PPARγ agonists are reported to produce fluid retention (Lebovitz, 2002), this may account for the lower Hct, although there were no differences in plasma protein concentration or the colloid osmotic pressure of the plasma. Untreated rats had elevated systemic BP but normal glomerular BP (P_GC), whereas ACEI alone or with PPARγ agonist lowered systemic BP without altering P_GC, which remained normal. The group receiving PPARγ agonist alone had an elevated BP similar to untreated rats and P_GC was also similar to other groups. The inulin clearance was lowest in untreated rats and GFR significantly improved in rats given ACEI alone or with PPARγ agonist. There was a borderline improvement with PPARγ agonist alone (Fig. 6).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6.

Glomerular filtration rate, factored for kidney weight, measured from the inulin clearance (C_IN, milliliters per minute per gram kidney weight) after the 6-month observation period in each group. *, p < 0.05 versus untreated Zucker rats.

Discussion

There was a clear benefit in treatment with either the PPARγ agonist or ACEI in reducing the structural and functional changes in the kidney in this rat model of obesity and type 2 diabetes. The main finding of the present study was that PPARγ agonist gave better protection versus ACEI on both proteinuria (p < 0.005) and glomerular sclerosis (p < 0.02) in association with improved metabolic control. There was an additive protection against tubulointerstitial damage and macrophage/monocyte infiltration, a borderline improvement of GFR, and no additional benefit on proteinuria and glomerular sclerosis when the PPARγ agonist was combined with ACEI versus PPARγ agonist alone. The inbred obese Zucker (ZDF/Gmi fa/fa) rat fed the Purina 5008 diet develops severe diabetes and extensive kidney damage. However, the renal damage does not exhibit all the characteristics of diabetic nephropathy. In remaining intact glomeruli, there is no evidence of mesangial expansion or thickening of the glomerular basement membrane, and the lesions seen were more indicative of focal segmental glomerular sclerosis (Baylis, 2001). In addition, this inbred strain of Zucker rats has a genetic predisposition to develop hydronephrosis, seen to an equal extent in both obese and lean rats (Vora et al., 1996; Baylis, 2001). However, most likely the majority of the renal injury in the obese inbred Zucker rats is due to the metabolic disturbances, because the present study demonstrates that chronic treatment with PPARγ agonist prevents the proteinuria and structural damage.

Given the wide range of actions of the PPARγ agonists, there are a number of possible mechanisms for the superior renal protection seen here. The glycemic control is undoubtedly a factor because hyperglycemia exerts multiple damaging actions on the kidney (Brownlee et al., 1988; Larkins and Dunlop, 1992; Phillips et al., 1995; West 2000), and good glycemic control is renoprotective in clinical studies in diabetics (The Diabetes Control and Complications Research Group, 1995; Di Landro et al., 1998). The vacuoles seen in the renal distal tubular epithelial cells of untreated rats were probably the result of abnormal accumulation of glycogen (resembling Armanni-Ebstein lesion), widely seen in renal tubules of diabetics. PPARγ agonist treatment completely eradicated these reflecting the normalization of blood glucose (Wehner 1998). Furthermore, experimentally induced hyperglycemia (with streptozotocin) in normoglycemic, hyperinsulinemic outbred Zucker rats, causes early glomerular hypertrophy (Park and Meyer 1995), a key event in progression of chronic renal disease (Daniels and Hostetter, 1990). Indeed, glomerular hypertrophy is reported as an early pathological event in obese Zucker fa/fa and ZDF rats (Coimbra et al., 2000; Hoshi et al., 2002).

In addition to the hyperglycemia, the untreated obese Zucker rats exhibit severe hyperlipidemia and dyslipidemia, with low-density lipoprotein cholesterol and triglycerides at 20× that seen in leans (Baylis 2001). Hyperlipidemia is one likely stimulus underlying progressive podocyte damage and subsequent glomerular sclerosis in fa/fa rats (Coimbra et al., 2000). Although not measured in the present study, the lipid lowering actions of the PPARγ agonist (Guan and Breyer, 2001; Rosak, 2002) likely contributed to the improved protection compared with rats treated with ACEI alone. In the outbred Zucker rat that is not frankly hyperglycemic, chronic treatment with lipid-lowering agents affords considerable protection (Kasiske et al., 1992).

Systemic hypertension is another risk factor for development of kidney disease, and BP was elevated in the untreated inbred obese Zucker rats in the present study. Of note, however, the P_GC was not elevated and although rats receiving ACEI alone exhibited a significant fall in BP, P_GC was not lowered. Furthermore, the rats receiving PPARγ agonist alone showed no change in either BP or P_GC, despite the fact that these animals enjoyed superior renoprotection versus those given ACEI alone. These findings suggest that glomerular hypertension was not an important contributory factor in this model of renal disease.

Several groups have reported that the thiazolidinediones protect the kidney in rat models of type 1 and type 2 diabetes (Fujii et al., 1997; Buckingham et al., 1998; McCarthy et al., 2000). In addition to the improved metabolic control, it is possible that PPARγ agonists have more direct beneficial actions on the kidney because constitutively expressed PPARγ are found in glomeruli, particularly mesangial cells (Asano et al., 2000; Guan et al., 2001; Nicholas et al., 2001) and PPARγ agonists reduce type I collagen synthesis in cultured glomerular mesangial cells (Routh et al., 2002; Zheng et al., 2002). In addition to directly inhibiting glomerular collagen synthesis, PPARγ agonist may also act indirectly, via its anti-inflammatory action. Glomerular sclerosis positively correlates with the occurrence of interstitial and glomerular monocytes/macrophages and activated mesangial cells (Nikolic-Paterson et al., 1994; Young et al., 1995; Lavaud et al., 1996). In the present study, the incidence of cells expressing CD 68 within the glomeruli (determined with the monocyte/macrophage marker Ki-M6), is blunted with either the ACEI or PPARγ agonist alone, and was absent in rats given PPARγ agonist in combination with ACEI. Thus both PPARγ agonists and ACEI might act via inhibition of recruitment or activation of these cells. In fact, PPARγ agonists are protective in the nondiabetic, 5/6th renal ablation model of glomerulosclerosis (Ma et al., 2001) and exert direct anti-inflammatory and antiproliferative actions on the circulation (Ishibashi et al., 2002), suggesting that the renal protective actions extend beyond improvement of the metabolic profile.

Overall, long-term administration of the PPARγ agonist was found to be powerfully protective in the complex model of chronic renal disease seen in the inbred obese Zucker rat. One disadvantage of the PPARγ agonists is that they potently stimulate adipocyte formation (Guan and Breyer, 2001; Rosak, 2002), leading to massive obesity. Of note, however, this obesity is not, in itself, damaging to the kidneys, because rats in the present study that received PPARγ agonist were massively obese without kidney damage. In fact the PPARγ agonist treated rats seemed “healthy” despite massive obesity.