Farnesoid X Receptor Modulates Renal Lipid Metabolism, Fibrosis, and Diabetic Nephropathy

Male C57BL/6J mice were obtained from The Jackson Laboratories (Bar Harbor, ME). They were maintained on a 12-h light/dark cycle and fed a control low-fat 10 kcal% fat diet (LFD; D12450B) or a high-fat 60 kcal% saturated (lard) fat diet (HFD; D12492) obtained from Research Diets (New Brunswick, NJ) for 1 week with the treatment of 1) vehicle only; 2) the synthetic FXR ligand GW4064 (21), 30 mg/kg body wt i.p. injection, twice daily; or 3) the natural FXR ligand cholic acid 0.5% wt/wt in diet.

Male Lepr^db/db mice with type 2 diabetes and their nondiabetic control Lepr^db/m mice on the C57BL/6J or C57BLKS/J genetic background were obtained from The Jackson Laboratories. C57BLKS/J-db/db and the same background nondiabetic mice were treated with GW4064 or vehicle only (40% 2-hydroxypropyl-β-cyclodextrin; Sigma) via intraperitoneal injection (same dose as above) for 1 week. The C57BL/6J-db/db and control mice were fed with control diet only (D12450B; Research Diets) or control diet supplemented with cholic acid 0.5% wt/wt for 12 weeks.

The animal studies and the protocols were approved by the institutional review boards at the Denver VA Medical Center and the University of Colorado Health Sciences Center.

Blood glucose levels were measured by means of a Glucometer Elite XL (Bayer, Tarrytown, NY). Plasma triglyceride and total cholesterol were measured with kits from Wako Chemical (Richmond, VA). Urine albumin and creatinine concentrations were determined using kits from Exocell (Philadelphia, PA). Results were expressed as the urine albumin-to-creatinine ratio (milligrams per milligram).

The renal cortex was separated from the medulla and cut into small fragments under sterile conditions. Glomeruli were isolated by differential sieving based on the report of Harper et al. (22). The purity of glomeruli was checked by light microscopy to ensure that they were >99% free of any tubular tissue. Proximal tubules were isolated according to the method of Brezniceanu et al. (23) by use of the Percoll gradient. Proximal tubular cells were characterized by their histological appearance as described previously (24). This procedure yielded a highly purified preparation of proximal tubules (>97% by microscopy).

Total RNA was isolated from the kidneys using SV total RNA isolation system from Promega (Madison, WI), and cDNA was synthesized using reverse transcript reagents from Bio-Rad Laboratories (Hercules, CA). The mRNA level was quantified using Bio-Rad iCyCler real-time PCR machine. Cyclophilin was used as internal control, and the amount of RNA was calculated by the comparative threshold cycle (C_T) method as recommended by the manufacturer. All the data were calculated from triplicate reactions. Primer sequences used have been described previously (3) or are available from the authors upon request.

Kidneys were homogenized at 4°C in homogenization buffer (20 mmol/l Tris-Cl, pH 7.4, 75 mmol/l NaCl, 2 mmol/l EGTA, 2 mmol/l EDTA, 1 mmol/l Na₃VO₄, and 1 mmol/l dithiothreitol) supplemented with a protease inhibitor cocktail consisting of 104 mmol/l AEBSF [4-(2-aminocthyl)-benzenesulfonyl fluoride], 0.08 mmol/l aprotinin, 2 mmol/l leupeptin, 4 mmol/l bestatin, 1.5 mmol/l pepstatin A, and 1.4 mmol/l E-64 (Sigma-Aldrich, St. Louis, MO). Nuclear extracts were prepared as we have previously described (4–7).

Equal amount of protein samples were subjected to SDS-PAGE (10% wt/vol), and they were then transferred to nitrocellulose membranes. After blocking with 5% fat-free milk powder with 1% Triton X-100 in Tris-buffered saline (20 mmol/l Tris-Cl and 150 mmol/l NaCl, pH 7.4), the blots were incubated with antibody against SREBP-1 (SC-8984; 1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA). Corresponding secondary antibody was visualized using enhanced chemiluminescence (Pierce, Bradford, IL). The signals were quantified with a Phosphor Imager with chemiluminescence detector and the accompanying densitometry software (Bio-Rad).

Lipids from the kidneys were extracted by the method of Bligh and Dyer, as we have previously described (4–7). Triglyceride and cholesterol content was measured using kits from Wako Chemicals.

Mice were anesthetized and perfused as previously described (4–7). Paraffin sections were stained for periodic acid Schiff (PAS) and imaged with an Olympus microscope. Frozen sections were used for immunostaining of fibronectin and imaged with a laser scanning confocal microscope (Zeiss LSM 510).

Mouse mesangial cell line was obtained from the American Type Culture Collection (Manassas, VA). The cells were maintained in Dulbecco's modified Eagle's medium/F12 containing 5% fetal bovine serum (FBS), 100 μg/ml streptomycin, 100 units/ml penicillin, and 4 mmol/l glutamine, with a normal d-glucose concentration of 100 mg/dl. The cells were incubated in a humidified atmosphere of 5% CO₂ at 37°C and passed every 3 days by trypsinization. For FXR overexpression experiments, mouse mesangial cells were transfected with pCMX-mFXR or empty pCMX vector (provided by Dr. David Mangelsdorf) in the medium containing 5% FBS with or without GW4064 and harvested 24 h after transfection for mRNA quantitative analysis and nuclei extraction as described above or previously (4–7). For FXR agonist treatment, cells at 80% confluence were made quiescent by incubation overnight with medium in 0.2% FBS and then treated for 48 h with different conditions as described in the figure legends.

Results are presented as the means ± SE for at least three independent experiments. Data were analyzed by ANOVA and Student-Newman-Keuls tests for multiple comparisons or by Student's t test for unpaired data between two groups. Statistical significance was accepted at the P < 0.05 level.

To examine the localization of FXR in the kidney, we isolated glomeruli and proximal tubules from mouse kidney. Real-time PCR detected the expression of FXR, SREBP-1c, and SREBP-2 transcripts in both glomeruli and proximal tubules to a significant extent. The expression levels in glomeruli were 30% of those in proximal tubule cells for both FXR and SREBP-1c and 45% for SREBP-2 (Table 1). The relative mRNA level of glomeruli marker podocin and proximal tubule marker NaPi IIa confirmed the purity of isolated fragments. In addition, we observed mRNA expression of FXR and SREBP-1c by real-time PCR in both cultured mouse mesangial cell line and podocyte cell line.

To determine the role of FXR activation on renal SREBP-1 regulation, C57BL/6J mice were fed the LFD or HFD with the FXR agonists for 1 week. In LFD-fed mice, treatment with the FXR agonist GW4064 or natural ligand cholic acid induced upregulation of the SHP, a well-characterized FXR target gene (9), and caused downregulation of SREBP-1c mRNA, a master regulator of lipogenic genes in the kidney (Fig. 1). Furthermore, in agreement with our published results (4), HFD induced increased expression of renal SREBP-1, which was prevented by the treatment with GW4064 or cholic acid (Fig. 1).

The above data demonstrating that FXR agonists modulate renal SREBP-1 expression in HFD-fed C57BL/6 mice prompted us to examine the function of FXR in a diabetic mouse model. The db/db mice have been extensively used as model of obesity and type 2 diabetes. To investigate how FXR activation impacts the renal gene expression in an acute way, we performed 1-week treatment with FXR agonist GW4064 in 7-week-old C57BLKS/J-db/db mice when severe hyperglycemia had developed. Administering GW4064 had no significant effect on ad libitum (Table 2) or 16-h fasting (data not shown) blood glucose concentrations in control and db/db mice after a 1-week treatment. Furthermore, GW4064 treatment did not change the circulating triglyceride level. However, plasma total cholesterol concentration was significantly decreased in the treated mice.

Dysregulation of renal lipid metabolism has been shown to play an important role in the pathogenesis of diabetic nephropathy. In agreement with an earlier study on FVB-db/db mice (5), we found increased renal expression of SREBP-1c mRNA and its target enzyme stearoyl-CoA desaturase-1 (SCD-1) in C57BLKS/J-db/db mice. Treatment with the FXR agonist significantly attenuated the increased expression of SREBP-1c and SCD-1 mRNA in the kidneys of db/db mice (Table 2).

In diabetic kidney diseases, glomerulosclerosis, tubulointerstitial fibrosis, and proteinuria were associated with increased renal expression of 1) mesangial matrix proteins (e.g., fibronectin), 2) fibrosis markers fibroblast-specific protein-1 (FSP-1) and α-smooth muscle actin (α-SMA), 3) the profibrotic growth factors (e.g., TGF-β and PAI-1), 4) the proinflammatory cytokines tumor necrosis factor-α (TNF-α) and monocyte chemoattractant protein-1 (MCP-1), and 5) the NADPH oxidase system (e.g., Nox2). The significant increase in the renal mRNA expression of genes involved in those pathways has been noted in 8-week-old untreated C57BLKS/J-db/db mice (Table 2). Treatment with GW4064 prevented or significantly attenuated the increases in mesangial matrix protein, profibrotic growth factors, and pro-inflammatory and pro-oxidative factors in db/db mice, indicating a renal-protective role for FXR activation.

Because cholic acid diet lowered blood glucose level significantly in our own observation on C57BLKS/J-db/db mice after 1 week of treatment (530.7 ± 29 before vs. 211.5 ± 14.1 mg/dl after, P < 0.01) and based on another report (14), we used C57BL/6J-db/db mice for our long-term treatment. This strain develops a moderate hyperglycemia and exhibits a remission from hyperglycemia after 12 weeks of age (25). In this regard, this would allow us to determine the effects of cholic acid independent of major changes in serum glucose.

We treated db/db mice with cholic acid at 16 weeks of age when they have already developed proteinuria and mesangial expansion to determine whether treatment with cholic acid is able to prevent the progression of diabetic kidney disease. After 12 weeks of db/db mice receiving control diet (that is diet without cholic acid), there was clear evidence of renal injury (Fig. 2) despite the normalization of blood glucose levels (Table 3). However, cholic acid caused significant decreases in proteinuria (Table 3) and glomerulosclerosis as determined by PAS staining (Fig. 2A–D) and immunofluorescence microscopy for fibronectin, a major mesangial matrix protein (Fig. 2E–H). Consistent with the immunofluorescence imaging, cholic acid also prevented the increase in renal fibronectin mRNA level in db/db mice (Table 3). In addition, cholic acid significantly attenuated the increase in renal mRNA expression of FSP-1 and α-SMA in db/db mice, both upregulated during development of tubulointerstitial fibrosis (26,27) (Table 3).

Treatment with cholic acid also downregulated the expression of SREBP-1c and SCD-1 mRNA in the kidneys of db/db mice (Table 3). This was accompanied by a significant decrease in renal triglyceride and cholesterol content (Table 3).

Consistent with the observations following short-term treatment with GW4064, cholic acid also decreased the renal expression of several key genes in profibrotic growth factors, proinflammatory cytokines, and oxidative stress enzymes, whose higher expressions are relevant to the pathogenesis of diabetic kidney disease (Table 3).

To determine whether FXR can directly modulate renal gene expression, we overexpressed FXR in mouse mesangial cells by transfection with the expression plasmid construct pCMX-mFXR. We found that adding GW4064 alone extremely upregulated the expression of FXR target gene SHP and decreased the SREBP-1c mRNA level (Fig. 3), indicating that there was enough active endogenous FXR in mesangial cells. Overexpression of FXR without adding GW4064 also increases SHP expression, although in a much lower level, but causes a greater decrease in SREBP-1c when compared with GW4064 alone. The activation of FXR without exogenous GW4064 is believed to be mediated by the presence of bile acid in the serum added to the medium. Furthermore, adding GW4064 in the medium of transfected cells does not cause a further decrease in SREBP-1c inhibition but induces marked stimulation of SHP expression compared with FXR overexpression or GW4064 treatment of untransfected cells.

To further determine whether FXR plays a direct role in regulating renal gene expression in the presence of a diabetic milieu, we examined the effects of the FXR agonist GW4064 in mouse mesangial cells grown in the presence of high glucose (250 or 450 mg/dl d-glucose). We found that a high glucose medium induced dramatic increase of FXR expression and that GW4064 further increased FXR expression (Fig. 4). High glucose also increased FXR target gene SHP expression, but the major upregulation was only seen in the presence of GW4064. In addition, whereas there were glucose-dependent increases in the expression of lipogenic genes, acetyl-CoA carboxylase (ACC) and liver-type pyruvate kinase (L-PK), and in expression of profibrotic growth factor TGF-β and proinflammatory cytokine interleukin-6 (IL-6), GW4064 prevented the gene induction by high-glucose condition, similar to renal protective role of GW4064 or cholic acid in db/db mice. Equimolar concentration of l-glucose had no effects on gene expression modulation, indicating that the effects of glucose are not mediated by increases in osmolality per se. Cell viability assay (G1780; Promega) showed that the effects of GW4064 treatment were not a result of cytotoxicity (data not shown).

There are well-established changes in renal function and structure in diabetes, including mesangial expansion and podocyte injury and loss, resulting in glomerulosclerosis, tubulointerstitial fibrosis, proteinuria, and a decline in glomerular filtration rate (28–33). Several hormonal and metabolic factors, including angiotensin II, TGF-β, VEGF, proinflammatory cytokines, oxidative stress, and advanced glycation end products, have been shown to modulate diabetes-related renal disease (34,35). In addition, renal accumulation of lipids has also been proposed to play a role in the pathogenesis of diabetic nephropathy (1–7). Moreover, we were able to demonstrate that increased renal expression of SREBP-1 is the key factor linking increased fatty acid synthesis and accumulation of lipids to development of nephropathy (2).

FXR is a member of the nuclear hormone receptor superfamily. High expression of FXR is restricted to liver, intestine, adrenal gland, and kidney. The primary bile acids, chenodeoxycholic acid and cholic acid, are the highest affinity endogenous activating ligands for FXR. In the enterohepatic system, FXR activation plays an important role in maintaining glucose, lipid, and bile acid homeostasis (36,37). In addition, FXR agonists prevent liver fibrosis (17,38).

The kidney has been implicated in both reabsorption and secretion of bile acids (39). However, the role of FXR activation in regulation of renal metabolism and prevention of diabetic nephropathy has not been determined.

The novel finding of our study is that we have determined that treatment with FXR agonists prevents the HFD-induced increases in renal expression of SREBP-1, the major factor leading to lipid accumulation during development of nephropathy. In the db/db model of mice with type 2 diabetes, we have further determined that in addition to modulating renal lipid metabolism, FXR agonists also regulate the renal expression of profibrotic growth factors, proinflammatory cytokines, and oxidative stress enzymes, while decreasing glomerulosclerosis and proteinuria.

Previous studies in our laboratory have associated decreased expression of FXR and its target SHP in the kidneys with a significant increase in renal SREBP-1 mRNA abundance in mouse models of type 1 diabetes (3). Those data suggest a potential negative regulation of SREBP-1–mediated renal lipid metabolism by highly regulated FXR in the kidney. In the current study, we provide further evidence to show that FXR activation by administration of FXR agonists can prevent the increase in renal SREBP-1 expression and activity in HFD-fed mice, a model of diet-induced obesity and insulin resistance (40). Treatment of db/db mice with GW4064 or cholic acid also causes decreases in renal SREBP-1 and SCD-1 expression, and long-term cholic acid treatment prevents the diabetes-associated increases in renal triglyceride and cholesterol content. These effects of FXR are similar to its action in the liver, where it has been shown to modulate fatty acid and triglyceride synthesis in part by decreasing the expression of SREBP-1 (10,11). The molecular basis for FXR to regulate SREBP-1 remains elusive. The FXR-SHP-SREBP-1c regulatory cascade has been proposed to link FXR activity to triglyceride levels in liver and serum (10). Our data in the kidney also support this pathway but do not rule out the possibility that renal lipid accumulation in diabetic nephropathy results from multiple changes in both synthetic and catabolic pathways (41). However, in mesangial cell culture studies, we found that increased expression of SHP was not necessarily leading to reduced SREBP-1c expression. Increased molecular expression of FXR caused a larger decrease in SREBP-1c expression than activation of constitutive level of FXR by GW4064, whereas FXR activation by GW4064 induced a higher increase in SHP mRNA expression. In addition, in high glucose–treated cells, mRNA expression of FXR, SHP, and SREBP-1c was increased (Fig. 4) (2). It is worth noting that the acute effects observed in cultured cells are not always consistent with the in vivo models (12).

Administration of FXR agonists may have systemic effects, including modulating serum glucose and lipids, and metabolic pathways in a number of relevant organs, including the liver, pancreas, muscle, and adipose tissue. In our experiments, GW4064 failed to show hypoglycemic effect on db/db mice, in contrast with the previous study (13). We think the use of animals of different ages and administration of compound in both studies makes a difference. In this regard, it has been shown that GW4064 had no effect on plasma glucose level in ob/ob mice (42). On the other hand, the renal effects of cholic acid treatment were obtained in the mice without hyperglycemia. Therefore, our data suggested that FXR exerts its protective role independently of blood glucose changes. Both GW4064 and cholic acid were found to lower the plasma cholesterol level in db/db mice. How the serum lipid affects the renal pathology is not fully understood. Considering the ability of renal de novo lipid synthesis and its role in diabetic nephropathy, the hypolipidemic effect of FXR agonists may conceivably share the common mechanism with the effect of cholic acid on lowering renal lipid content in db/db mice. The direct renal effect for FXR was also manifested by the cell culture studies. In high glucose–treated renal mesangial cells, FXR activation by GW4064 prevented glucose-induced lipogenic gene expression. Overexpression of FXR or treatment with GW4064 in normal glucose condition repressed the SREBP-1c expression. These data indicate that FXR exerts direct action on the modulation of renal lipid metabolism.

Treatment of db/db mice with cholic acid for 12 weeks significantly improved glomerulosclerosis, tubulointerstitial fibrosis, and proteinuria. The relevant renal gene regulation was also observed in acute treatment of db/db mice with GW4064. Both GW4064 and cholic acid have been shown to have FXR-specific effects from studies in FXR-null mice (13,14,43). However, cholic acid can activate genes by FXR-independent pathways (44,45). Further studies will clarify FXR-dependent and -independent mechanisms by which cholic acid protects against renal injury.

The potential mechanisms for the renal protective effects of FXR include 1) decrease in renal lipid accumulation, 2) decrease in the expression of profibrotic growth factors, 3) decrease in the expression of proinflammatory cytokines, and 4) decrease in oxidative stress. Although FXR agonists have also been shown to modulate hepatic fibrosis, their potential anti-inflammatory and anti-oxidative effects have not been previously reported.

Activation of hepatic stellate cells is thought to mediate the process of liver fibrosis through secretion of extracellular matrix proteins, and activation of FXR reduces the secretion of extracellular matrix proteins by these cells (38). In rat models of extrahepatic and intrahepatic cholestasis, the FXR agonist GW4064 protects against cholestatic liver damage in part by decreasing the expression of TGF-β (16). Our results, therefore, consistently indicate that in the db/db model of diabetic nephropathy, treatment with the FXR agonist causes decreased expression of 1) profibrotic growth factors TGF-β, CTGF, and PAI-1; 2) FSP-1; 3) α-SMA; and 4) fibronectin. FXR agonist treatment also decreased TGF-β expression in renal mesangial cell culture. These effects of FXR are associated with prevention of renal fibrosis.

Although diabetic nephropathy has not been considered to be an immune-mediated renal disorder, an inflammatory mechanism has been suggested to play a role in the development of diabetic nephropathy (35,46). Anti-inflammatory agents, such as mycophenolate mofetil, prevent the development of glomerular injury in streptozotocin-induced diabetic rats (47). Urinary MCP-1 excretion increases in patients with early diabetic nephropathy, and acute-phase markers of inflammation are associated with diabetic kidney disease in patients with type 2 diabetes (48,49). In this study, we found that FXR agonist treatment downregulated many of the genes that are involved in the inflammatory process, such as TNF-α, IL-6, and MCP-1, in db/db mice. The same effect was also observed for IL-6 after FXR agonist treatment of mesangial cells grown in the presence of high glucose to mimic the diabetic environment. The mechanisms by which FXR activation regulate the proinflammatory cytokines are not as well understood. However, it is interesting that proinflammatory cytokines can also induce downregulation of nuclear receptors such as FXR during infection and inflammation, accompanied by abnormalities in lipid metabolism that are similar to those that occur in many common disorders, such as diabetes, obesity, and metabolic syndrome (50,51).

In summary, our findings in this study with FXR agonists identify the effect of FXR activation on the prevention of diabetic nephropathy and potential metabolic pathways that may mediate the diabetes-related alterations in renal lipid metabolism, fibrosis, inflammation, and oxidative stress (Fig. 5). These results therefore indicate that 1) long-term treatment of animals with FXR agonists is possible and 2) treatment of diabetic mice with FXR agonists prevents or significantly attenuates the development of mesangial expansion, accumulation of extracellular matrix proteins, and proteinuria at least in part by inhibiting fatty acid synthesis and the expression of profibrotic growth factors, proinflammatory cytokines, and NADPH oxidase in the kidney. This new mechanistic perspective may lead to important new therapeutic considerations and new therapeutic goals for the treatment of diabetic nephropathy.

This study was supported by grants from the Juvenile Diabetes Research Foundation, the American Heart Association, the National Institutes of Health (R01-DK-062209 and R01-AG-026529), and VA Merit Review (to M.L.).

We thank Prof. David Mangelsdorf (University of Texas Southwestern Medical Center, Dallas) for providing us with pCMX-mFXR or empty pCMX vector and Timothy M. Willson, Patrick R. Maloney, and Jim L. Collins (Discovery Research GlaxoSmithKline) for providing us with GW4064.