Glomerular 20-HETE, EETs, and TGF-β1 in diabetic nephropathy

Abstract

The early stage of diabetic nephropathy (DN) is linked to proteinuria. Transforming growth factor (TGF)-β1 increases glomerular permeability to albumin (P_alb), whereas 20-HETE and EETs reduce P_alb. To investigate the impact of hyperglycemia and hyperlipidemia on 20-HETE, EETs, and TGF-β1 in the glomeruli, rats were divided into four groups: ND rats were fed a normal diet, HF rats were fed a high-fat diet, STZ rats were treated with 35 mg/kg of streptozotocin, and HF/STZ rats were fed a HF diet and treated with STZ. After 10 wk on these regimens, blood glucose, urinary albumin, serum cholesterol, serum triglyceride levels, and the kidney-to-body weight ratio were significantly elevated in STZ and HF/STZ rats compared with HF and ND rats. STZ and HF/STZ rats had histopathologic changes and abnormal renal hemodynamics. Expression of glomerular CYP4A, enzymes for 20-HETE production, was significantly decreased in STZ rats, whereas expression of glomerular CYP2C and CYP2J, enzymes for EETs production, was significantly decreased in both STZ and HF/STZ rats. Moreover, glomerular TGF-β1 levels were significantly greater in STZ and HF/STZ rats than in HF and ND rats. Five-week treatment of STZ rats with clofibrate induced glomerular CYP4A expression and 20-HETE production, but reduced glomerular TGF-β1 and urinary protein excretion. These results demonstrate that hyperglycemia increases TGF-β1 but decreases 20-HETE and EETs production in the glomeruli, changes that may be important in causing glomerular damage in the early stage of DN.

diabetic nephropathy (DN) is a progressive kidney disease caused by angiopathy of capillaries in the glomeruli. The development and progression of DN are determined by many mechanisms. Persistent hyperglycemia is a major initiator. Hyperglycemia-induced metabolic imbalance, abnormal hemodynamics, and other factors act independently or synergistically with hyperglycemia, thus contributing to the alteration of glomerular function (10, 37). A glomerulus and its surrounding Bowman's capsule constitute a renal corpuscle, the basic filtration unit of the kidney. The glomerular filtration barrier permits filtration of water, electrolytes, and small molecules while restricting filtration of albumin and larger molecules (32). Any damage to the glomerular filtration barrier may result in hyperfiltration and histological changes, consequently causing proteinuria, an important indicator of the early stage of DN.

It is known that 20-hydroxyeicosatetraenoic acid (20-HETE) and epoxyeicosatrienoic acids (EETs), metabolites of arachidonic acid (AA) by the cytochrome P-450 (CYP), have important physiological functions, including inhibition of sodium transport in the nephron and vasoconstriction or vasodilation of blood vessels (35). Research results indicate that renal 20-HETE production in the rat is catalyzed primarily by the CYP4A isoforms (23), whereas CYP2C and CYP2J isoforms are responsible for renal EETs formation (35). EETs can be further hydrolyzed by soluble epoxide hydrolase (sEH) to the corresponding dihydroxyeicosatrienoic acids (DHETs). Recent reports demonstrate that 20-HETE and EETs have a protective effect on glomerular mesangial and epithelial cells (22). Endogenously or exogenously formed 20-HETE and EETs also have an essential function in decreasing glomerular permeability to albumin (P_alb) (19, 36). However, transforming growth factor (TGF)-β has a destructive effect on glomerular filtration, which leads to the induction of podocytes (28) and cell apoptosis (18) and directly increases P_alb (30). In addition, renal hypertrophy and fibrosis are mediated by increased TGF-β activity (29). Moreover, a previous study by Dahly-Vernon et al. (6) demonstreated that TGF-β1 significantly inhibits 20-HETE production in the isolated glomeruli and the treatment of 20-HETE agonist opposes the effect of TGF-β1 to increase P_alb.

Because hyperglycemia can also induce hyperlipidemia in diabetic patients and animal models (31), it becomes difficult to evaluate the relative contribution of hyperglycemia and/or hyperlipidemia to the signs of DN. Thus, we used high-fat (HF), streptozotocin (STZ), and HF/STZ treatments to investigate the impact of hyperglycemia and/or hyperlipidemia on glomerular 20-HETE, EETs, and TGF-β1 production in these animals. Fibrate drugs, which have been extensively used to lower triglyceride levels in hyperlipidemia patients, also have been shown to increase 20-HETE production (5). In the present study, we used clofibrate to induce glomerular CYP4A expression and 20-HETE production and studied its effects on glomerular TGF-β1 levels and urinary protein excretion in diabetic rats.

METHODS

Animals.

Three-week-old male Sprague-Dawley rats (Harlan, Indianapolis, IN) were divided into four groups. The ND group consisted of 3-wk-old rats fed a normal diet for 10 wk (41), the STZ group consisted of ND rats given one intraperitoneal dose of 35 mg/kg of STZ in citrate buffer (pH 4.5) when they were 8 wk old, the HF group consisted of 3-wk-old rats fed an HF diet (Bioserv, Frenchtown, NJ) for 10 wk (34), and the HF/STZ group consisted of HF rats treated with STZ when they were 8 wk old. Biochemical and physiological measurements were performed at 13 wk old. All animal protocols were approved by the Institutional Animal Care and Use Committee and were in accordance with the requirements stated in the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Blood parameters, urinary parameters, Western blot analysis, and renal hemodynamic measurements.

Blood glucose concentrations were determined by glucometer (Prestige Smart System, Fort Lauderdale, FL) after 4 days of STZ treatment. Rats with blood glucose levels higher than 300 mg/dl were considered diabetic and included in the experiments. Rats given different treatments were placed in individual metabolic cages for urine collection. Urine and blood samples were used to determine urinary protein by a Bio-Rad Proteinuria Assay kit (Bio-Rad, Hercules, CA), albumin by an ELISA kit (Exocell, Philadelphia, PA), and creatinine levels by an enzymatic colorimetric method described previously (11). Cholesterol and triglyceride levels were measured by chemical assays (Wako, VA). Western blot analysis of different proteins was done as described previously (13). We measured TGF-β1 levels by using an ELISA kit (R&D Systems, Minneapolis, MN) (27). Renal hemodynamics in rats were monitored under anesthesia (2% isoflurane). Briefly, we placed one polyethylene (PE-240) catheter into the bladder to collect urine, one in the femoral artery (PE-50) to measure and record mean arterial pressure (MAP) with a pressure transducer, and one (PE-50) in the femoral vein for the infusion of agents. We administered a priming dose of 0.5 ml of fluorescein isothiocyanate (FITC) inulin (8 mg/ml in PBS) over 2 min. We then infused FITC inulin (12 mg/h iv) to maintain a constant concentration of FITC inulin. We performed a left laparotomy and placed a transonic flow probe (Transonic System, Ithaca, NY) over the left renal artery to measure renal blood flow (RBF). We obtained MAP, RBF, and renal vascular resistance (RVR) from a computerized data collection system (EMKA Technologies, Falls Church, VA) and used the concentration of FITC inulin in the plasma and urine to calculate the glomerular filtration rate (GFR) (13).

Histological analysis.

We isolated and weighed the kidneys from treated and control rats. Excised kidneys were fixed in neutral buffered formalin for routine paraffin embedding and subsequent staining with hematoxylin Jones periodic acid-Schiff (PAS) and Masson's trichrome. To quantify mesangial expansion, the PAS-positive area in the mesangium was examined under a microscope using Metamorph software (Universal Imagine, Downingtown, PA). The color threshold was set in PAS-positive staining areas of individual glomerular tufts. We analyzed 30 glomerular tufts in each animal. The PAS-positive area was expressed as percentage of glomerular tuft area (mesangial area) (1, 39).

Isolation of glomeruli.

Glomeruli were isolated by a rapid-sieving technique described previously (17). Specifically, rats were anesthetized and the abdominal aorta was cannulated with a PE-50 catheter below the left renal artery. Blood flow to the kidneys was interrupted and the kidneys were perfused with 10 ml of ice-cold dissection solution containing (in mM) 135 NaCl, 3 KCl, 1.5 CaCl₂, 10 MgCl₂, 2 KH₂PO₄, 5.5 glucose, and 10 HEPES (pH 7.4). Kidneys were rapidly removed and hemisected, and the medulla was excised. Pieces of renal cortex were pressed through a 180-μm stainless-steel sieve with the barrel of a 30-ml syringe. The cortical tissue passing through the sieve was flushed through a 150-μm sieve and glomeruli were collected on a 100-μm nylon sieve. The preparation of glomeruli was >99% free of any tubular tissues.

AA metabolism in isolated glomeruli.

AA metabolism in glomeruli was assessed by HPLC as described previously (17). Isolated glomeruli were incubated with 0.1% Tween 80 in a 100 mM potassium phosphate buffer (pH 7.4) containing 10 mM MgCl₂ and 1 mM EDTA for 15 min at 4°C. The glomeruli were washed twice with buffer, spun down by centrifuge, and incubated with [1-¹⁴C] AA (0.1 μCi/ml, 10 μM) in 1 ml of potassium phosphate buffer containing 1 mM NADPH in a shaking bath (60 min; 37°C). Extraction and HPLC analysis were done as described previously (34).

Quantitative real-time PCR.

The mRNA levels of CYP isoforms and sEH were determined by quantitative real-time PCR. Total RNA was isolated from glomeruli using a Nucleospin RNA II kit (Clontech, Mountain View, CA). Total RNA (200 ng) was reverse-transcribed using a High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA). Predeveloped TaqMan gene expression assays of the following proteins were purchased from Applied Biosystems (assay IDs): CYP4A1, Rn00598510_m1; CYP4A2, Rn01417066_m1; CYP4A3, Rn00598412_m1; CYP4A8, Rn00581081_m1; CYP2C11, Rn00569868_m1; CYP2C23, Rn00582954_m1; CYP2J3, Rn00598500_m1; CYP2J4, Rn00576482_m1; and sEH, Rn00674574_m1. The analyses were performed on an ABI 7500 Fast Real-Time PCR System (Applied Biosystems) following the manufacturer's recommended 2-step thermal cycling program. During each extension step, oligonucleotide probes with reporter and quencher dyes were degraded by 5′-3′ exonuclease activity of Taq polymerase; the level of released reporter dye was measured at the end of the cycle to determine the amount of specific amplicon. The PCR amplification cycle at which the reporter dye fluorescence passed the selected threshold (C_T) was calculated by the provided program. The expression of mRNA of each protein was normalized to eukaryotic 18S rRNA, and, relative to expression in ND rats, was calculated based on the comparative C_T method, expressed as (−ΔΔC_T).

Clofibrate treatment.

To investigate the potential association of 20-HETE and TGF-β1, STZ rats were treated with either clofibrate (240 mg·kg⁻¹·day⁻¹) or vehicle (corn oil) for 5 wk. Clofibrate was dissolved in corn oil and given intragastrically. After 5 wk of treatment, we performed functional and biochemical analyses.

Statistical analysis.

All values are expressed as means ± SE. All data were analyzed by SPSS computer software (SPSS, Chicago, IL). We used one-way ANOVA and Student-Newman-Keuls tests for multiple comparisons or independent Student's t-test for unpaired groups. Statistical significance was set at P < 0.05.

RESULTS

Physiological parameters.

In STZ and HF/STZ rats, kidney size, blood glucose, serum triglyceride, and urinary protein levels were significantly greater than in ND rats. In contrast, HF had no effect on kidney size, blood glucose levels, or serum triglyceride compared with ND rats (Table 1). HF/STZ rats had heavier kidneys and higher serum triglyceride levels than STZ rats (Table 1). At the end of the study period, body weights of STZ and HF/STZ rats are significantly decreased compared with ND and HF rats (Table 1). Urinary albumin excretion in different groups is shown in Fig. 1A. STZ and HF/STZ rats had significantly higher urinary albumin excretion compared with ND rats, whereas HF treatment had slightly drop on this parameter. Also, there was a significant increase in the ratio of kidney-to-body weight (renal hypertrophy) in both STZ and HF/STZ rats compared with ND rats (Fig. 1B).

Table 1.

Physiological parameters in ND, STZ, HF, and HF/STZ rats

Variable	ND	STZ	HF	HF/STZ
Blood glucose, mg/dl	110±15	385±38*	118±19	334±6*‡
Body wt, g	312±6	243±9*	350±9*	275±9*
Kidney wt, g	1.4±0.05	1.7±0.06*	1.3±0.06	2.0±0.17*†‡
Serum cholesterol, mg/dl	93±7	125±10*‡	101±6	159±14*‡
Serum triglyceride, mg/dl	20±4.4	84±19*	32±3*	210±31*†‡
Urinary protein, mg/24 h	15.5±1	57.4±20*	4.7±0.7*	30.9±0.9*†‡
Mesangial area, %	15±1.7	20±1.2*	19±1.6*	21±1.1*

Values are means ± SE. ND, 3-wk-old rats fed normal diet for 10 wk; STZ, 8-wk-old rats (ND) treated with streptozotocin (35 mg/kg ip) for 5 wk; HF rats, 3-wk-old rats fed high-fat diet for 10 wk; HF/STZ, 8-wk-old HF rats treated with STZ (35 mg/kg ip) for 5 wk.

^*P < 0.05 vs. ND.

^†P < 0.05 vs. STZ.

^‡P < 0.05 vs. HF.

Fig. 1.

Physiological parameters in diabetic and control rats. Urinary albumin excretion (A) and kidney weight-to-body weight ratio (B) in 13-wk-old rats given different treatments are shown. Values are means ± SE (n = 6). *P < 0.05 vs. normal-diet (ND) rats. #P < 0.05 vs. high-fat (HF) rats. STZ, streptozotocin.

Renal histological analyses.

To examine morphological changes in the kidneys, we used PAS or Masson's hematoxylin counterstaining of sections of kidneys isolated from rats given different treatments. The mesangial area (Table 1 and Fig. 2A) was significantly increased in the glomeruli of HF, STZ, and HF/STZ rats compared with ND rats, and HF/STZ rats had the largest mesangial area. STZ and HF/STZ rats had strong Masson's trichrome staining in their glomeruli, whereas HF rats had slightly glomerular fibrosis (Fig. 2B). In addition, STZ, HF, and HF/STZ rats had higher expression levels of glomerular fibronectin than ND rats (Fig. 2C).

Fig. 2.

Histological examination of diabetic and control rats. Representative micrographs show PAS staining (A) and Masson staining (positive staining, blue; B) of renal sections isolated from 13-wk-old rats given different treatments. C: expression of glomerular fibronectin in different groups.

Renal hemodynamics.

To examine the effects of HF, STZ, and HF/STZ treatments on renal function, we used 13-wk-old ND, STZ, HF, and HF/STZ rats in a renal hemodynamic study. MAP was significantly higher in HF rats than in ND rats, while HF/STZ caused the return of MAP to control levels (Fig. 3). HF treatment of rats significantly increased RBF, whereas HF/STZ treatment significantly decreased RBF. Similarly, STZ treatment caused a decrease of RBF from 9.0 ± 1 to 7.4 ± 0.9 ml/min. In STZ and HF/STZ rats, RVR values were significantly higher than in ND rats. GFR values in HF rats were significantly higher than in ND rats (1.3 ± 0.04 vs. 1.1 ± 0.04 ml/min), while those in HF/STZ rats were further elevated from 1.3 ± 0.04 ml/min in HF rats to 1.5 ± 0.05 ml/min. Similarly, STZ treatment caused an increase in GFR from 1.1 ± 0.04 in ND rats to 1.8 ± 0.14 ml/min.

Fig. 3.

Renal hemodynamic measurement in diabetic and control rats. Mean arterial pressure (MAP; A), renal blood flow (RBF; B), renal vascular resistance (RVR; C), and glomerular filtration rate (GFR; D) in 13-wk-old rats given different treatments are shown. Values are means ± SE (n = 4). *P < 0.05 vs. ND rats. #P < 0.05 vs. HF rats.

Real-time PCR analysis of glomerular CYP isoforms and sEH mRNA expression.

We used real-time PCR to determine the effects of different treatments on the expression of the glomerular CYP isoforms and sEH mRNA. HF treatment slightly increased the expression of CYP4A1, CYP4A2, CYP4A3, and CYP4A8, whereas STZ treatment slightly decreased expression of CYP4A2, CYP4A3, and CYP4A8. The decrease of CYP4A2 in STZ rats was the only one of these changes that attained statistical significance (Fig. 4A). Similarly, HF increased the expression of CYP2C11 and CYP2C23, whereas treatment with STZ or HF/STZ slightly decreased the expression of CYP2C11, CYP2J3, and CYP2J4 (Fig. 4B). There was no significant change in the expression of sEH in any group. In summary, HF rats had higher expression levels of glomerular CYP4A and CYP2C mRNA, whereas STZ and HF/STZ rats had lower expression levels of these isoforms in their glomeruli.

Fig. 4.

Quantitive expression of glomerular cytochrome P-450 (CYP) isoforms and soluble epoxide hydrolase (sEH) mRNAs in diabetic and control rats. Real-time RT-PCR analysis of glomerular CYP isoforms and sEH for 20-HETE (A) and EETs (B) production in 13-wk-old rats given different treatments is shown. Values are means ± SE (n = 4). *P < 0.05 vs. ND rats.

Protein expression of glomerular CYP isoforms, sEH, and TGF-β1.

We used Western blot analysis to determine the effects of different treatments on ptotein expression of the glomerular CYP isoforms and sEH mRNA. Expression of glomerular CYP4A, CYP2C11, CYP2C23, and CYP2J was significantly decreased in STZ rats (Fig. 5A). However, sEH protein levels in these rats were similar to those in ND rats (Fig. 5A). In HF rats, there was no significant change in the protein expression of CYP4A, CYP2C11, or sEH; there was decreased expression of CYP2J, but slightly increased expression of CYP2C23. In HF/STZ rats, there was significant downregulation of CYP2C11 and CYP2J, but no change in the expression of CYP4A, CYP2C23, or sEH. To determine the relationship between TGF-β1 and glomerular damage, we examined TGF-β1 levels, finding that they were significantly elevated in both STZ and HF/STZ rats. In contrast, HF treatment did not affect glomerular TGF-β1 levels (Fig. 5B).

Fig. 5.

Glomerular expression of CYP isoforms and sEH and transforming growth factor (TGF)-β1 production in diabetic and control rats. A: Western blot analysis of glomerular CYP isoforms and sEH for 20-HETE and EETs production in 13-wk-old rats given different treatments. The numbers at the top are relative changes of expression compared with those in ND rats. B: glomerular TGF-β1 levels in 13-wk-old rats given different treatments. Values are means ± SE (n = 4). *P < 0.05 vs. ND rats. #P < 0.05 vs. HF rats.

AA metabolism in the glomeruli.

To characterize the production of CYP-derived eicosanoids in glomeruli, we examined AA metabolism by HPLC in glomeruli of these rats. Incubation of isolated glomeruli from ND rats with [¹⁴C]-AA and NADPH produced DHETs, 20-HETE, and EETs (Fig. 6A). The largest peaks, with a retention time of 10.5 min coeluted with 20-HETE standard, indicated that 20-HETE is the major metabolite in rat glomeruli. We next examined 20-HETE production in glomeruli. In STZ rats, glomerular 20-HETE production was significantly reduced by 45% as compared with that in ND rats (Fig. 6B). In addition, 20-HETE production was slightly decreased in the glomeruli of HF/STZ rats compared with HF rats, but this did not reach statistical significance. We did not determine DHETs and EETs production in the glomeruli of different groups because we could not do so accurately. The peaks of DHETs and EETs are very low and close to the background (Fig. 6A).

Fig. 6.

Glomerular 20-HETE production in diabetic and control rats. A: representative reverse-phase HPLC elution profiles of metabolites formed during incubation of arachidonic acid (AA) with glomeruli isolated from ND rats. B: bar graph shows ω-hydroxylase activity in the glomeruli of ND, STZ, HF, and HF/STZ rats. Values are means ± SE (n = 3). *P < 0.05 vs. ND rats.

Effects of clofibrate treatment.

Since there is decreased production of 20-HETE in the glomeruli of STZ rats but not in HF/STZ rats (Figs. 5 and 6) and 20-HETE opposes the action of TGF-β1 on the increase of glomerular permeability to albumin (6), we determined whether clofibrate treatment affects glomerular CYP4A expression levels, 20-HETE production, TGF-β1 levels, and urinary protein excretion in STZ rats. After treating STZ rats with either clofibrate (240 mg·kg⁻¹·day⁻¹ ig) or vehicle for 5 wk, we found that clofibrate caused the reduction of serum cholesterol levels compared with levels in the vehicle-treated group (85 ± 9 vs. 129 ± 7 mg/dl, n = 4, P < 0.05). Clofibrate treatment also alleviated the kidney enlargement in STZ rats (1.5 ± 0.2 in clofibrate-treated group vs. 1.8 ± 0.1 g in vehicle-treated group, n = 4, P < 0.05). In addition, clofibrate treatment significantly induced glomerular CYP4A expression and increased ω-hydroxylase activity (Fig. 7). Clofibrate treatment decreased TGF-β1 levels in STZ rats (Fig. 7C). Clofibrate also reduced urinary protein excretion in STZ rats (60 ± 10 vs. 40 ± 4 mg/day, n = 4, P < 0.05). However, clofibrate treatment did not affect blood glucose levels (425 ± 37 vs. 395 ± 25 mg/dl, n = 4), MAP (96 ± 6 vs. 91 ± 2 mmHg, n = 3), or GFR (2.1 ± 0.3 vs. 1.9 ± 0.2 ml/min, n = 3) of STZ rats.

Fig. 7.

Effect of clofibrate on glomerular CYP4A expression, glomerular ω-hydroxylase activity, and urinary protein levels in STZ rats. A: representative immunoblots of glomerular CYP4A in STZ rats treated with vehicle or clofibrate. The number at the top is relative change of expression compared with that in vehicle-treated rats. B: effect of clofibrate treatment on glomerular ω-hydroxylase activity in STZ rats. C: effect of clofibrate treatment on glomerular TGF-β1 levels. Values are means ± SE (n = 4). *P < 0.05 vs. vehicle control.

DISCUSSION

The HF diet model mimics HF food consumption by people, producing characteristics similar to those of human obesity, including impaired glucose tolerance and insulin resistance (9, 12). In addition, we recently demonstrated that the HF diet treatment leads to hypertension (13, 41, 42), which together with obesity is a risk factor for the development of type 2 diabetes (12). Type 1 diabetes is characterized by an absolute deficiency of insulin. Both types of diabetes can result in DN. Either type 1 or type 2 diabetes is accompanied by hyperglycemia and hyperlipidemia, and it becomes difficult to estimate the relative contributions of hyperglycemia and hyperlipidemia to the development of DN. In this study, we used STZ and HF/STZ rats as our experimental models because STZ treatment provides a well-established chemically induced diabetic model (4, 7). Our results demonstrate that STZ and HF/STZ rats have significant physiologic and histological changes including hyperglycemia, hyperlipidemia, proteinuria, renal hypertrophy, increased GFR, and increased mesangial area (Table 1, Figs. 1, 2, 3). Based on the results in the present study and characteristics of DN described in the literature (7, 20), it is clear that STZ and HF/STZ rats are the models of the early stage of DN. In contrast, 10 wk of HF treatment did not result in hyperglycemia, renal hypertrophy, and proteinuria (Table 1 and Fig. 1).

It has been suggested that in diabetes abnormal renal hemodynamics contribute to the development of DN (20). We found that STZ and HF/STZ rats had abnormal renal hemodynamics, including increased GFR but decreased RBF (Fig. 3). Increased GFR or glomerular hyperfiltration is an important feature of early stage of DN (20). It has been postulated that constriction of the efferent arteriole by angiotensin II is an important mechanism for causing glomerular hyperfiltration (20). In addition, Veelken et al. (33) showed that there is increased expression of endothelial nitric oxide (NO) synthase in STZ rats and that treatment with NO synthase inhibitor causes a reduction of GFR in STZ rats. These results demonstrate that increased NO production in the kidneys contributes to glomerular hyperfiltration and the development of DN. Taken together, it is possible that angiotensin II and NO are important factors in causing glomerular hyperfiltration in diabetic rats. Although our demonstration that STZ treatment caused a decrease in RBF is in accord with the finding of Chen et al. (3), other investigators showed that RBF is significantly increased in STZ rats (20, 33). The reasons for the inconsistency between these results are not known. It could be, however, that the age and species of animals, the duration of hyperglycemia, and the dose of STZ administration contribute to this inconsistency.

Although the exact mechanisms whereby diabetes causes nephropathy are not fully understood, it is recognized that diabetes induces TGF-β1 levels in the kidneys (38) and TGF-β1 contributes to the development and progression of DN (43). In the kidneys, TGF-β1 promotes glomerular hypertrophy and regulates glomerular production of collagen and fibronectin, important contributors to the causation of glomerular damage in DN (43). Recently, Dahly-Vernon et al. (6) showed that TGF-β1 contributes to the development of glomerular injury in Dahl salt-sensitive rats by increasing glomerular P_alb through the inhibition of glomerular 20-HETE production. We therefore examined the expression levels of CYP isforms and sEH and determined TGF-β1 levels in the glomeruli of all groups of rats. In STZ rats, expression of CYP4A, CYP2C11, CYP2C23, and CYP2J was significantly decreased, whereas no change in sEH expression was detected (Figs. 4 and 5). In HF/STZ rats, there was decreased expression of CYP2C11 and CYP2J, but no change in expression of CYP4A and sEH. We do not know why CYP4A was decreased in STZ rats but not in HF/STZ rats. However, these results in HF/STZ rats are in accord with Dey et al.'s (8) findings that obese Zucker rats have decreased expression of glomerular CYP2C11 but no change in CYP4A or sEH expression compared with levels in Zucker lean rats. It has been established that, in rats, 20-HETE synthesis is catalyzed primarily by CYP4A isoforms (35); EETs synthesis is catalyzed by CYP2C and CYP2J isoforms; sEH is the major exzyme for the conversion of EETs to DHETs (14). Thus, decreased glomerular expression of CYP4A, CYP2C, and CYP2J isoforms is responsible for the decreased production of glomerular 20-HETE and EETs in STZ and HF/STZ rats. We found that glomerular TGF-β1 levels were significantly increased in STZ and HF/STZ rats, but not in HF rats. The increase in TGF-β1 levels in STZ and HF/STZ rats was associated with hyperglycemia, hyperlipidemia, proteinuria, and renal hypertrophy. These results demonstrate that diabetes induced by STZ or HF/STZ treatments increases glomerular TGF-β1 levels, contributing to the early stage of DN, whereas HF treatment does not affect glomerular TGF-β1 levels or cause DN.

It has been established that 20-HETE and EETs affect vascular tone (vasoconstriction or vasodilation), tubular sodium transport in vitro, and renal function and blood pressure in vivo (15, 35, 40). However, the function of these eicosanoids in the glomeruli has only recently been recognized. Ito and Roman (17) demonstrated significant production of 20-HETE and EETs in isolated glomeruli. They also found that glomerular 20-HETE production was greater in rats fed a low-salt diet than in rats given a high-salt diet. Similarly, we found that isolated glomeruli produced higher amounts of 20-HETE than EETs (Fig. 6A). In addition, glomerular 20-HETE production was significantly decreased in STZ rats compared with levels in ND and HF rats (Fig. 6B), which is in accord with the CYP4A expression levels shown in Fig. 5. Although we were able to determine glomerular ω-hydroxylase activity in different groups (Fig. 6B), we could not do so in epoxygenase activity. The main reason is that the DHETs and EETs peaks found in HPLC are very low and very close to the background (Fig. 6A). In addition, we found that DHETs and EETs formation in the glomeruli of the same group of rats has high variability. For example, the activity of DHETs + EETs in glomeruli isolated from ND rats is ∼3 ± 1.2 pmol·min⁻¹·mg⁻¹. Under the circumstance, we cannot determine EETs production accurately from different groups. McCarthy et al. (19) showed that puromycin aminonucleoside (PAN), an agent that is injurious to renal function, caused an increase in P_alb, which was blocked by 20-HETE. More recently, Williams et al. (36), using liquid chromatography mass spectrometry, detected significant amounts of endogenous 20-HETE and EETs in glomerui and demonstrated that 20-HETE and EETs are important in maintaining P_alb. In the present study, we showed that decreased production of CYP-eicosanoids in the glomeruli of diabetic rats is associated with proteinuria and histological damage (Table 1, Fig. 2). These results demonstrate that decreased synthesis of 20-HETE and EETs in the glomeruli could be a critical factor in the causation of proteinuria.

Studies demonstrated that clofibrate not only induces CYP4A expression and 20-HETE production through peroxisome proliferator-activated receptor α (PPARα), but also has anti-hypertensive effects in animal models. For example, peroxisome proliferator response elements are located in the promoter region of CYP4A genes (2, 21); clofibrate induced renal tubular 20-HETE production and attenuated HF-induced hypertension in rats (41); clofibrate increased renal 20-HETE production and improved pressure natriuresis in Dahl salt-sensitive rats (24, 26). Since we found glomerular CYP4A expression was decreased in STZ rats but not in HF/STZ rats, we treated STZ rats with clofibrate to examine whether increased glomerular 20-HETE can affect glomerular TGF-β1 levels. We found that clofibrate treatment significantly increased expression of glomerular CYP4A and increased glomerular 20-HETE production in STZ rats (Fig. 7). These results are consistent with studies showing that PPARα is expressed in glomeruli (16) and clofibrate induced glomerular CYP4A expression and protected glomeruli from PAN-induced injury (19). We also found that clofibrate alleviated the increase in levels of TGF-β1 (Fig. 7C), which is associated with both the reduction of urinary protein and kidney enlargement. Interestingly, fenofibrate, another fibrate drug, has been shown to improve the signs of DN in db/db mice (25). Taken together, these findings indicate that the effects of clofibrate on glomerular TGF-β1 and urinary protein in STZ rats are associated with the induction of glomerular 20-HETE production.

In summary, our results demonstrate that STZ and HF/STZ treatments cause early signs of DN, including proteinuria, renal hypertrophy, abnormal hemodynamics, and histological damage. The development of DN as a consequence of STZ and HF/STZ treatment is associated with increased glomerular production of TGF-β1 and decreased production of 20-HETE and EETs. Since TGF-β1 increases P_alb, whereas 20-HETE and EETs decrease P_alb, this study provides the possibility that decreased synthesis of these eicosanoids and increased production of TGF-β1 in diabetic glomeruli affect the permeability of the glomerular barrier and contribute to glomerular damage in the early stage of DN. Thus, manipulation of the synthesis of 20-HETE and EETs by pharmacological inducers, inhibitors, or the gene transfer method will be needed to elucidate the interactions among these eicosanoids and TGF-β1 and their contributions to glomerular function in DN.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grant R01-HL-082733 to M.-H. Wang.