Bilirubin and biliverdin protect rodents against diabetic nephropathy by downregulating NAD(P)H oxidase

Transcription

1 & 21 International Society of Nephrology Bilirubin and biliverdin protect rodents against diabetic nephropathy by downregulating NAD(P)H oxidase Masakazu Fujii 1, Toyoshi Inoguchi 1,2, Shuji Sasaki 1, Yasutaka Maeda 1, Jing Zheng 1, Kunihisa Kobayashi 1 and Ryoichi Takayanagi 1 1 Department of Internal Medicine and Bioregulatory Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan and 2 Innovation Center for Medical Redox Navigation, Kyushu University, Fukuoka, Japan We recently found a markedly lower prevalence of vascular complications, including kidney disease, in diabetic patients with Gilbert syndrome, a congenital form of hyperbilirubinemia, suggesting a beneficial effect of bilirubin (BIL) on diabetic nephropathy. To directly examine this, we determined whether hereditary hyperbilirubinemic Gunn j/j rats and biliverdin (BVD)-treated diabetic mice were resistant to the development of renal disease. Both rodent models had less albuminuria and complete protection against the progression of mesangial expansion accompanied by normalization of transforming growth factor-b1 and fibronectin expression. Simultaneously, there was normalization of urinary and renal oxidative stress markers, and the expression of nicotinamide adenine dinucleotide phosphate (NAD(P)H) oxidase subunits in the kidney. In cultured vascular endothelial and mesangial cells, BIL and BVD significantly inhibited NADPH-dependent superoxide production, and both high glucose- and angiotensin II-induced production of reactive oxygen species. Collectively, our findings suggest that BIL and BVD may protect against diabetic nephropathy and may lead to novel antioxidant therapies for diabetic nephropathy. Kidney International (21) 78, ; doi:1.138/ki ; published online 4 August 21 KEYWORDS: albuminuria; diabetic nephropathy; NADPH oxidase; oxidative stress Correspondence: Toyoshi Inoguchi, Department of Internal Medicine and Bioregulatory Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka , Japan. toyoshi@intmed3.med.kyushu-u.ac.jp Received 24 November 29; revised 13 May 21; accepted 25 May 21; published online 4 August 21 Worldwide, diabetic nephropathy is a leading cause of endstage renal failure. Recent epidemiological studies have revealed that diabetic nephropathy is also associated with increased risk of cardiovascular events. A therapeutic approach targeting its causative mechanisms urgently needs to be established. In recent years, oxidative stress has been considered to be an important pathogenic factor in the development of diabetic vascular complications, including nephropathy. 1 4 Accumulating evidence has shown that many protein, lipid, and DNA markers of oxidative stress are increased in kidney and vascular tissues from animals and patients with diabetes. 3 5 However, trials using various known antioxidants failed to protect against diabetic nephropathy in humans. Bilirubin (BIL) has been recognized as an endogenous antioxidant for many years. 6 Its formation is mediated by the ubiquitously expressed heme oxygenase, the rate-limiting enzyme involved in heme catabolism. Heme oxygenase participates in heme breakdown to generate biliverdin (BVD), free ferrous iron, and carbon monooxide. Subsequently, BVD is rapidly converted to BIL by BVD reductase. Recently, increasing evidence has suggested that heme oxygenase and its reaction product BIL may serve as important endogenous agents with cytoprotective activity against oxidative stress injury. 7 9 The anti-oxidative action of BIL has been attributed to its highly efficient radicalscavenging effect, 6 whereas recent reports have shown that its hydrophobic tetrapyrrole structure may inhibit the activity of nicotinamide adenine dinucleotide phosphate (NAD(P)H) oxidase. 1,11 Notably, we and other investigators have shown that the activation of NAD(P)H oxidase may be a major cause of increased oxidative stress in diabetic renal and vascular tissues Therefore, we speculated that BIL might have a beneficial effect on diabetic nephropathy. In this context, we recently reported a markedly lower prevalence of nephropathy and other vascular complications in diabetic patients with Gilbert syndrome, a congenital hyperbilirubinemia, along with reduced markers of oxidative stress and inflammation. 16 However, this association study does not necessarily implicate BIL as causative factor in the observed Kidney International (21) 78,

2 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy Table 1 Body weights, blood glucose, and serum total/direct bilirubin levels in Gunn j/ and j/j rats at baseline, 8 weeks, and 24 weeks after STZ-injection j/ j/ j/j j/j Control (n=1) DM (n=1) Control (n=1) DM (n=1) Weight (g) weeks 264.± ± ± ±9.8 8 weeks 439.± ±11.1 a 4.± ±16.6 b 24 weeks 548.5± ±9.7 a 444.4± ±19.4 b Blood glucose level (mg/dl) weeks 123.6± ± ± ±9.6 8 weeks 17.9± ±28.4 a 113.± ±15.8 b 24 weeks 188.9± ±11.1 a 178.9± ±2.8 b Ur Alb/Cr ratio mg/g Cre Serum total bilirubin level (mg/dl) 8 weeks.18±.4.15± ±.4 7.1±.43 b Serum direct bilirubin level (mg/dl) 8 weeks.6±.1.7±.3.64±.3.51±.3 Abbreviations: DM, streptozotocin-induced diabetic; STZ, streptozotocin. Data are means±s.e. a o.1 vs j/ control. b o.1 vs j/j control. lower prevalence of diabetic nephropathy. Therefore, to test this hypothesis and explore the underlying molecular mechanisms, we investigated (1) whether hereditary hyperbilirubinemic Gunn rats are resistant to the progression of nephropathy after the onset of diabetes, (2) whether administration of BVD, the precursor of BIL, protects against diabetic nephropathy in mice, a rodent model of type 2 diabetes. The underlying molecular mechanisms were also explored in these experimental animal models and in vitro culture studies. RESULTS Hyperbilirubinemic Gunn rats The characteristics of experimental rats at, 8, and 24 weeks after onset of diabetes are shown in Table 1. The total BIL levels remained constant after an age of 4 weeks in Gunn j/j rats and Gunn j/ þ rats. At 8 weeks after the onset of diabetes, the total serum BIL levels were.15±.2 and.18±.4 mg/dl in diabetic and non-diabetic Gunn j/ þ rats, respectively, and 7.1±.43 and 9.47±.4 mg/dl in diabetic and non-diabetic Gunn j/j rats, respectively (Table 1). Such marked increments in serum BIL levels in Gunn j/j rats reflect increases in the levels of unconjugated BIL. Diabetic Gunn j/ þ rats exhibited marked increases in the amounts of urinary albumin excretion compared with nondiabetic Gunn j/ þ rats at 8 weeks (Po.1) and 24 weeks (Po.1) after the onset of diabetes, whereas diabetic Gunn j/j rats exhibited significantly less urinary albumin excretion than diabetic Gunn j/ þ rats at 8 weeks (Po.5) and 24 weeks (Po.1) (Figure 1). To explore the mechanism underlying these beneficial effects of hyperbilirubinemia, we measured systemic oxidative stress markers, urinary 8-hydroxy-2 -deoxyguanosine (8-OHdG) excretion and 8-epi-prostaglandin F 2a levels. j / j / j j / j / j j / j / j j / j / j Contr DM Contr DM 8 weeks 24 weeks Figure 1 Ur Alb/Cr ratio in homozygous j/j (n=1) and heterozygous j/ Gunn rats (n=1) at 8 and 24 weeks after the onset of diabetes. A 24-h urine sample was collected using metabolic cages. The well-mixed urine was centrifuged at 75 g for 5 min and then stored at 81C until analysis. Ur Alb/Cr ratio (mg/g creatinine) was measured as described in the Materials and Methods. Results are expressed as means±s.e. Po.5; Po.1; Po.1. Contr, non-diabetic; DM, streptozotocininduced diabetic; Ur Alb/Cr, urinary albumin/creatinine. Urinary excretion levels of both 8-OHdG and 8-epi-prostaglandin F 2a were significantly higher in diabetic Gunn j/ þ rats than in non-diabetic Gunn j/ þ rats at 8 weeks. (Figure 2a and b). The diabetes-induced increases in both oxidative markers at 8 weeks were completely prevented in diabetic hyperbilirubinemic Gunn j/j rats, which showed levels comparable to those in non-diabetic Gunn j/ þ rats. Immunostaining analysis of 8-OHdG in renal tissues revealed that the staining intensities in diabetic Gunn j/ þ rats were significantly higher than those in control Gunn j/ þ rats, in both glomeruli and tubules at 24 weeks (Figure 2c f). These increases in 8-OHdG staining intensities in glomeruli and tubules were completely prevented in diabetic Gunn j/j rats, which showed levels comparable to those in control Gunn j/ þ rats (both non-significant vs control). Furthermore, we examined the expression of renal NAD(P)H oxidase components. To examine the localization and expression levels of NOX4 protein, we carried out immunostaining analysis. The staining intensities for NOX4 protein were stronger in the renal glomeruli and tubules of diabetic Gunn j/ þ rats than in those of control Gunn j/ þ rats (Figure 3a and b). Western blotting analysis confirmed that the protein levels for NOX4 were significantly increased in the kidneys of diabetic Gunn j/ þ rats compared with control Gunn j/ þ rats (Po.1; Figure 3c). All of these changes were completely prevented in diabetic Gunn j/j rats, which showed levels comparable to those in control Gunn j/ þ rats. 96 Kidney International (21) 78,

3 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy μg/g Cre 25 μg/g Cre 25 Urinary 8-OHdG excretion j / j / j j / j / j j / j / j j / j / j j / j / j j / j / j Urinary 8-epi-PG F 2α excretion j / j / j j / j / j Contr DM Contr DM Contr DM Contr DM 8 weeks 24 weeks 8 weeks 24 weeks j / j / j j / j / j j / DM j / j DM j / DM j / j DM Relative glomerular 8-OHdG intensity Relative renal tuble 8-OHdG intensity 2 1 j / j / j j / j / j j / j / j j / j / j Contr DM Contr DM Figure 2 Urinary oxidative stress markers in homozygous j/j (n=1) and heterozygous j/ (n=1) Gunn rats at 8 or 24 weeks after the onset of diabetes. Urinary 8-hydroxy-2 -deoxyguanosine (8-OhdG) excretion (mg/g creatinine) (a), and urinary 8-epi-prostaglandin (8-epi-PG) F 2a excretion (mg/g creatinine) (b) were measured as described in the Materials and Methods. Results are expressed as means±s.e. Po.5; Po.1. Immunostaining analysis of renal 8-OHdG. Renal 8-OHdG contents were evaluated by immunostaining analysis. The pictures are representative of the samples of glomeruli (c) and tubules (d) from non-diabetic heterozygous Gunn j/ þ rats (n ¼ 1), diabetic heterozygous Gunn j/ þ rats (n ¼ 1), non-diabetic homozygous Gunn j/j rats (n ¼ 1), and diabetic homozygous Gunn j/j rats (n ¼ 1). The immunostaining intensities of 8-OHdG in the renal glomeruli (e) and tubules (f) were semi-quantitatively evaluated using Scion imaging software (Scion, Frederick, MD, USA). Results are expressed as the mean percentages of the levels in non-diabetic heterozygous Gunn j/ þ rats±s.e. Po.1. Contr, non-diabetic; DM, streptozotocin-induced diabetic. Real-time PCR analysis showed that the mrna levels for NOX4 and other components, p22phox and p47phox, were also significantly increased in the kidneys of diabetic Gunn j/ þ rats compared with control Gunn j/ þ rats (Figure 3d f). All of these changes were completely prevented in diabetic Gunn j/j rats, which showed levels comparable to those in control Gunn j/ þ rats. Kidney International (21) 78,

4 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy j / j / j j / j / j j / DM j / j DM j / DM j / j DM 3 NOX4 2 1 NOX4 β-actin 67 kda j / j / j j / j / j Contr DM j / j / j j / DM j / j DM NOX4 mrna p22phox mrna j / j / j j / j / j j / j / j j / j / j j / j / j j / j / j Contr DM Contr DM Contr DM Figure 3 Immunostaining analysis of NOX4 protein levels in renal tissues. The pictures are representative of the samples of glomeruli (a) and tubules (b). Similar results were obtained in all samples from the kidneys of each group (n ¼ 1). NOX4 protein levels were evaluated by western blot analysis. Pictures are representative of western blots for NOX4 in renal tissues (c). NOX4 levels were normalized to the level of b-actin and the results are expressed as the mean percentages of the levels in control heterozygous Gunn j/ þ rats±s.e. (For all groups, n ¼ 1.) Po.1. The levels of mrna for NOX4, p22phox, and p47phox in renal tissues. Total RNA was extracted from the kidneys of each group (n ¼ 1) of rats at 24 weeks after the onset of diabetes. NOX4, p22phox, and p47phox mrna levels were measured by real-time reverse transcriptase PCR (d f). The levels of mrna were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in control heterozygous Gunn j/ þ rats±s.e. Po.5; Po.1. Contr, non-diabetic; DM, streptozotocin-induced diabetic. p47phox mrna We investigated the effect of hyperbilirubinemia on mesangial expansion at 24 weeks after the onset of diabetes. The glomerular structure in diabetic Gunn j/ þ rats showed accelerated mesangial expansion compared with that observed in control Gunn j/ þ rats (Figure 4a). The periodic acid-schiff-positive and nuclei-free mesangial area was markedly increased in the glomeruli of diabetic Gunn j/ þ rats (Po.1; Figure 4b). Diabetic Gunn j/j rats showed complete prevention of mesangial expansion. We also examined the effect of hyperbilirubinemia on the levels of transforming growth factor-b1 (TGF-b1) and fibronectin. The levels of TGF-b1/fibronectin mrna and protein were significantly increased in the kidneys of diabetic Gunn j/ þ rats compared with control Gunn j/ þ rats (Figure 4c f). 98 Kidney International (21) 78,

5 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy j / j / j Relative mesangial area j / j / j j / j / j Contr DM j / DM j / j DM TGF-β1 mrna Fibronectin mrna j / j / j j / j / j j / j / j j / j / j Contr DM Contr DM TGF-β1 3 Fibronectin 2 1 j / j / j j / j / j j / j / j j / j / j Contr DM Contr DM TGF-β1 Fibronectin β-actin j / j / j j / DM j / j DM 39 kda 22 kda Figure 4 Mesangial expansion in homozygous j/j and heterozygous j/ Gunn rats (a, b). At 24 weeks after the onset of diabetes, renal sections were stained with periodic acid-schiff. The pictures are representative of samples from non-diabetic heterozygous Gunn j/ þ rats (n ¼ 1), non-diabetic homozygous Gunn j/j rats (n ¼ 1), diabetic heterozygous Gunn j/ þ rats (n ¼ 1), and diabetic homozygous Gunn j/j rats (n ¼ 1). Semi-quantitative analysis of the mesangial area. Results are expressed as the mean percentages of the areas in nondiabetic heterozygous Gunn j/ þ rats±s.e. Po.1. Transforming growth factor-b1 (TGF-b1) and fibronectin mrna and protein levels (c f). Total RNA and protein were extracted from the kidneys of each group of rats 24 weeks after the onset of diabetes (for all groups, n ¼ 1). TGF-b1 and fibronectin mrna levels were measured by real-time reverse transcriptase PCR (c, d). The levels of mrna were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in control heterozygous Gunn j/ þ rats±s.e. TGF-b1 and fibronectin protein levels were evaluated by western blot analysis (e, f). Pictures are representative of western blots of TGF-b1 and fibronectin in the renal tissues. The levels of protein were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in control heterozygous Gunn j/ þ rats±s.e. Po.5; Po.1; Po.1. Contr, non-diabetic; DM, streptozotocin-induced diabetic. In diabetic Gunn j/j rats, the increases in TGF-b1 and fibronectin levels were significantly prevented. BVD administration Next, we investigated the effect of BVD administration on albuminuria, oxidative stress, and renal mesangial expansion in mice. BVD administration (5 mg/kg) orally for 2 weeks and 12 weeks did not significantly affect body weights or blood glucose levels (Table 2). Oral administration of this dose of BVD did not induce a significant elevation in serum BIL levels (Table 2), although an intraperitoneal injection of the same dose induced a slight elevation in serum BIL levels at.5 h, with rapid return to the basal levels at 6 h after injection (data not shown). Urinary albumin excretion significantly increased in nontreated mice as compared with control db/ þ mice at 2 weeks and 12 weeks after the start of BVD administration (Figure 5). BVD administration significantly attenuated such increases in urinary albumin excretion in mice. Each oxidative stress marker (urinary 8-OHdG, 8-epiprostaglandin F 2a excretion, and immunostaining analysis of Kidney International (21) 78,

6 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy Table 2 Body weights, blood glucose, and serum total/direct bilirubin levels in and mice at baseline, 2 weeks, and 12 weeks after treatment. (n=8) (n=8) BVD (n=8) BVD (n=8) mg/g Cre 12 1 Weight (g) weeks 28.4±.4 5.2±.5 a 29.9± ±.6 b 2 weeks 29.7± ±.5 a 29.8± ±.4 b 12 weeks 32.1± ±1.1 a 33.± ±.6 b Blood glucose level (mg/dl) weeks 124.6± ±2.5 a 131.3± ±11.2 b 2 weeks 145.8± ±16.6 a 121.3± ±22.7 b 12 weeks 154.2± ±13.2 a 145.8± ±43.5 b Ur Alb/Cr ratio Serum total bilirubin level (mg/dl) weeks.74±.8.7±.5.74±.7.76±.6 2 weeks.98±.3.99±.4.95±.3.98±.3 Serum direct bilirubin level (mg/dl) weeks.26±.5.24±.3.25±.2.26±.2 2 weeks.22±.2.2±.2.21±.3.21±.2 Abbreviation: BVD, biliverdin. Data are means±s.e. a o.1 vs non-treated. b o.1 vs treated. 8-OHdG in renal tissues) was significantly higher in mice than in control db/ þ mice. BVD administration completely normalized these markers in mice to control levels (Figure 6a f). We also confirmed the effect of BVD to intracellular production of superoxide in the renal tissues by dihydroethidium (DHE) stain. The oxidized DHE signals were significantly higher in mice than those in control mice at 12 weeks (Figure 6g j). BVD administration completely normalized the oxidized DHE signals to the control levels. Then, we evaluated the NADPH oxidase activities, using the lucigenin-enhanced chemiluminescence method as previously described. 17 BVD administration significantly suppressed diabetes-induced activation of NADPH oxidase in crude cortex homogenates up to control levels (Figure 6k). Furthermore, to examine the levels of NOX4 protein, we carried out immunostaining analysis and western blot BVD 2 weeks 12 weeks analysis. The protein levels of NOX4 were also significantly increased in renal tissues of mice. BVD administration completely normalized all of these changes in mice to the control levels (Figure 7a c). The levels of mrna for NOX4 and other components, p22phox and p47phox, were also significantly increased in the kidney of mice as compared with control mice (Po.5; Po.5; Po.5; Figure 7d f). Likewise, the levels of mrna for NAD(P)H oxidase components in glomeruli were significantly increased in mice as compared with control mice(po.1; Po.1; Po.1; Figure 7g i). These abnormalities were completely normalized by BVD administration. We evaluated the levels of mrna for heme BVD Figure 5 Effect of biliverdin (BVD) treatment on Ur Alb/Cr ratio in mice (n ¼ 8) and mice (n ¼ 8). BVD (5 mg/kg) was administered to and db/ þ mice starting at 12 weeks of age for 2 or 12 weeks. A 24-h urine sample was collected using metabolic cages. The well-mixed urine was centrifuged at 75 g for 5 min and then stored at 81C until analysis. Ur Alb/Cr ratio (mg/g creatinine) was measured as described in the Materials and Methods. Po.5, Po.1, Po.1. BVD, BVD treated mice; Ur Alb/Cr, urinary albumin/creatinine. Figure 6 The effect of biliverdin (BVD) treatment on urinary 8-hydroxy-2 -deoxyguanosine (8-OhdG) excretion and urinary 8-epi-prostaglandin (8-epi-PG) F 2a excretion. Urinary 8-OHdG excretion (mg/g creatinine) (a), and urinary 8-epi-PG F 2a excretion (mg/g creatinine) (b) were measured after 2 and 12 weeks treatment, as described in the Materials and Methods. Results are expressed as means±s.e. (For all groups, n ¼ 8.) Po.5, Po.1, Po.1. Immunostaining analysis of renal 8-OHdG. The pictures are representative of the samples of glomeruli (c) and tubules (d) from non-treated db/ þ mice (n ¼ 8), non-treated mice (n ¼ 8), BVD-treated db/ þ mice (n ¼ 8), and BVD-treated mice (n ¼ 8). The immunostaining intensities of 8-OHdG in the renal glomeruli and tubules were semi-quantitatively evaluated using Scion imaging software (Scion) (e, f). Results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. Po.5; Po.1. The effect of BVD treatment on ROS production was detected by dihydroethidium (DHE) stain in the renal tissues. The pictures are representative of the samples of glomeruli (g) and tubules (h) from non-treated db/ þ mice (n ¼ 8), non-treated mice (n ¼ 8), BVD-treated db/ þ mice (n ¼ 8), and BVD-treated mice (n ¼ 8). The emission intensities of DHE oxidation in the renal glomeruli and tubules were semi-quantitatively evaluated using Scion imaging software (Scion) (i, j). Results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. Po.1. Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase activity assay for crude cortex homogenates from non-treated db/ þ mice (n ¼ 8), non-treated mice (n ¼ 8), BVD-treated db/ þ mice (n ¼ 8), and BVD-treated mice (n ¼ 8) (k). After establishing a background rate for 5 min, NADPH was added and the increase in lucigenin-enhanced chemiluminesence was monitored by Flex Station 3, Molecular Devices, Tokyo, Japan. Results are expressed as means±s.e. Po.5, Po.1 vs mice. 91 Kidney International (21) 78,

7 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy oxygenase-1 in all these groups, but there was no significant change among them. Mesangial expansion was found in mice at the age of 24 weeks (Po.1; Figure 8a and b), with increased mrna and protein levels of TGF-b1 (Figure 8c and e) and fibronectin (Figure 8d and f) in the kidney of mice, as well as in glomeruli (Figure 8g and h). These abnormalities were completely prevented by BVD administration. In vitro effect of BIL and BVD on NAD(P)H oxidase activities The effect of BIL and BVD on NAD(P)H oxidase activities in cultured human mesangial cells were evaluated by the lucigenin method. Pre-treatment of the cells with both BIL and BVD for 48 h reduced NAD(P)H oxidase activities in a dose-dependent manner (Figure 9a and b). In addition, intracellular oxidative stress was evaluated by the 2,7 - dichlorofluorescein diacetate staining method as previously Urinary 8-OHdG excretion μg/g Cre μg/g Cre 3 BVD 2 weeks BVD Urinary 8-epi-PG F 2α excretion BVD BVD 12 weeks 2 weeks 12 weeks BVD BVD Relative glomerular 8-OHdG intensity BVD Relative renal tuble 8-OHdG intensity BVD Figure 6 continued on the following page Kidney International (21) 78,

8 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy BVD BVD Figure 6 Continued. Relative glomerular DHE oxidation intensity Biliverdin (BVD) Lucigenin NADPH DPI 14 NADPH-dependent superoxide generation (RLU/min/mg protein) BVD Relative renal tuble DHE oxidation intensity BVD BVD DPI Figure 7 The effect of biliverdin (BVD) treatment on NOX4 protein levels in renal tissues. Immunostaining analysis of NOX4 protein levels in renal tissues. The pictures are representative of the samples of glomeruli and tubules (a, b). Similar results were obtained in all samples from the kidneys of each group. (For all groups, n ¼ 8.) NOX4 protein levels were evaluated by western blot analysis (c). The levels of protein were normalized to the level of b-actin and the results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. (For all groups, n ¼ 8.) Po.1. The expression of mrna levels for the NOX4, p22phox, and p47phox in renal tissues. Total RNA was extracted from the kidneys after 12 weeks of BVD treatment and the levels of mrna were measured by real-time reverse transcriptase PCR (RT-PCR) (d f). The levels of mrna were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. (For all groups, n ¼ 8.) Po.5; Po.1. The expression of mrna levels for NOX4, p22phox, and p47phox in the glomeruli. Total RNA was extracted from the glomeruli after 12 weeks of BVD treatment. The each extracted mrna level from the glomeruli was measured by real-time RT-PCR (g i). The levels of mrna were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. (For all groups, n ¼ 8.) Po Kidney International (21) 78,

9 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy described. 18 2,7 -Dichlorofluorescein diacetate staining showed that exposure of the cells to angiotensin II (; 1 n for 4 h and high glucose levels (45 mg/dl) for 24 h increased intracellular oxidative stress as compared with the control levels (Figure 9c and d). These increases in intracellular oxidative stress were completely normalized to the control levels by treatment with BIL for 1 h (1 m. BVD BVD 3 NOX4 2 1 NOX4 β-actin 67 kda NOX4 mrna BVD p22phox mrna BVD p47phox mrna BVD BVD BVD NOX4 mrna p22phox mrna p47phox mrna BVD BVD BVD Kidney International (21) 78,

10 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy Finally, we investigated the effect of BIL and BVD on the expression of NOX4, TGF-b1, and fibronectin in cultured human mesangial cells. stimulated the expression levels of NOX4 mrna in a time-dependent manner (Figure 1a). Treatment with BIL and BVD completely inhibited the increased expression of NOX4 mrna, whereas BVD Relative mesangial area BVD TGF-β1 mrna Fibronectin mrna BVD BVD g TGF-β1 mrna TGF-β1 Fibronectin BVD BVD BVD TGF-β1 Fibronectin β-actin BVD h Fibronectin mrna BVD 39 kda 22 kda 914 Kidney International (21) 78,

11 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy N-acetylcysteine and a-lipoic acid did not, even at the maximal concentration. The protein levels of NOX4 were also inhibited by BIL and BVD (Figure 1b). We confirmed that the increased expression of mrnas of TGF-b1 and fibronectin induced by was also completely prevented by BIL and BVD in all time courses, whereas N-acetylcysteine and a-lipoic acid did not (Figure 1c and d). DISCUSSION In this study, to reveal the protective effect of BIL on diabetic nephropathy, we first used streptozotocin-induced diabetic homozygous Gunn rats, which exhibit a marked elevation of plasma unconjugated BIL levels owing to a genetic deficiency of uridine diphosphate glucuronosyl transferase-1. The present study showed that diabetic hyperbilirubinemic Gunn j/j rats exhibited significantly less urinary albumin excretion than diabetic non-hyperbilirubinemic Gunn j/ þ rats. In addition, diabetic Gunn j/j rats did not develop renal mesangial expansion, which is one of the most striking morphological features of diabetic nephropathy, 6 months after the onset of diabetes, whereas diabetic Gunn j/ þ rats developed typical mesangial expansion. TGF-b is a key cytokine that mediates the changes, which leads to extracellular matrix accumulation and glomerular expansion in diabetes. 19 We also found that the increased expression of TGF-b and fibronectin, a major matrix protein, was prevented in diabetic Gunn j/j rats. These findings suggested that hyperbilirubinemic Gunn j/j rats were resistant to the progression of functional and morphological features of nephropathy after the onset of diabetes. Second, we used mice, a rodent model of type 2 diabetes to evaluate the effect of administration of BVD, a precursor of BIL. We administered BVD (5 mg/kg) orally instead of BIL, because BVD is relatively more water-soluble than BIL. Oral administration of this dose of BVD did not induce a significant elevation in serum BIL levels. Nevertheless, the present study showed that administration of BVD (5 mg/kg) protected against both albuminuria and renal mesangial expansion in mice. As serum BVD enters cells rapidly and is converted to BIL by BVD reductase, 2 it is likely that the beneficial effect of BVD administration may be because of increased levels of intracellular BIL levels generated from exogenously administered BVD, rather than increased levels of serum BIL or BVD. The present study further suggested that the mechanism underlying these beneficial effects of hyperbilirubinemia and BVD administration may be the inhibition of oxidative stress, evaluated by systemic oxidative stress markers (urinary excretion of 8-OHdG and 8-epi-prostaglandin F 2a ), 8-OHdG staining, and DHE oxidation levels, which are relatively superoxide-sensitive in renal tissues. As BIL has been recognized to possess radical-scavenging activity, 6 the inhibition of oxidative stress may be, at least in part, owing to its radical-scavenging activity. Notably, the serum BIL levels in diabetic Gunn j/j rats were slightly, but significantly, lower than those in non-diabetic Gunn j/j rats. Such a decrease in serum BIL levels induced by diabetes may be a reflection of its consumption owing to its radical-scavenging activity, although this speculation should be confirmed in future studies. However more importantly, we revealed for the first time that BVD administration induced downregulation of NAD(P)H oxidase components (NOX4, p22phox, p47phox) in diabetic kidneys, glomeruli, and human mesangium cells. Although the nature of the sources of reactive oxygen species (ROS) overproduction in diabetes has not been precisely defined, we and other investigators have indicated that nonphagocytic NAD(P)H oxidases may be the major sources of increased ROS production in the vascular tissues of diabetic animals and patients. 12,13,15 The non-phagocytic NAD(P)H oxidase comprises of a membrane-associated cytochrome b558 composed of NOX family proteins (gp91phox, NOX1, NOX4) and p22phox, and several cytosolic regulatory components, p47phox, p67phox, and Rac 1 or Rac The isoform NOX4 was cloned from the kidney, where it was found to be highly expressed. 22,23 It has been suggested that NOX4, as a major source of ROS production in the kidney, could have a role under pathological conditions. We previously reported that increased expression of NOX4 may have an important role in increased ROS production in the kidneys of streptozotocin-induced diabetic rats Gorin et al. 27 confirmed the increased expression of NOX4 in the diabetic kidney, and further revealed that downregulation of NOX4 induced by antisense oligonucleotides completely Figure 8 The effect of biliverdin (BVD) treatment on mesangial expansion (a, b). After 12 weeks of treatment, renal sections were stained with periodic acid-schiff. The pictures are representative of samples from non-treated db/ þ mice (n ¼ 8), non-treated mice (n ¼ 8), BVD-treated db/ þ mice (n ¼ 8), and BVD-treated mice (n ¼ 8). Semi-quantitative analysis of the mesangial area. Results are expressed as the mean percentages of the areas in non-treated db/ þ mice±s.e. Po.1. Transforming growth factor-b1 (TGF-b1) and fibronectin mrna and protein levels (c f). Total RNA and protein were extracted from the kidneys of each group of mice after 12 weeks BVD treatment (for all groups, n ¼ 8). TGF-b1 and fibronectin mrna levels were measured by real-time reverse transcriptase PCR (RT-PCR) (c, d). The levels of mrna were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. TGF-b1 and fibronectin protein levels were evaluated by western blot analysis (e, f). Pictures are representative of western blots of TGF-b1 and fibronectin in the renal tissues. The levels of protein were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. Po.5; Po.1; Po.1. The expression of mrna levels for the TGF-b1 and fibronectin in the glomeruli. The extracted mrna from the glomeruli were measured by realtime RT-PCR (g, h). The levels of mrna were normalized to the levels of b-actin and the results are expressed as the mean percentages of the levels in non-treated db/ þ mice±s.e. Po.5; Po.1. Kidney International (21) 78,

12 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy Lucigenin NADPH DPI Bilirubin 1 μmol/l μmol/l μmol/l nmol/l Lucigenin NADPH DPI Biliverdin 1 μmol/l 1 μmol/l 1 μmol/l 1 nmol/l Control bilirubin (1 n Bilirubin DPI 1 nmol/l 1 nmol/l 1 nmol/l 1 μmol/l 1 μmol/l 2 Glucose Bilirubin 1 mg/dl 45 mg/dl 45 mg/dl 1 μmol/l DPI Figure 9 The effect of bilirubin (BIL) (a) and biliverdin (BVD) (b) on nicotinamide adenine dinucleotide phosphate (NAD(P)H) oxidase activities in cultured human renal mesangial cells. BIL or BVD were incubated each with concentrations as indicated for 48 h. After establishing a background rate for 5 min, NADPH was added and the increase in lucigenin-enhanced chemiluminescence was monitored by Flex Station 3. Results are expressed as means±s.e. from four independent experiments. Po.5, Po.1 vs NADPH. The effect of BIL on increased reactive oxygen species (ROS) production by (angiotensin II) (c) and high glucose (d) stimulation in cultured human endothelial cells as evaluated by 2,7 -dichlorofluorescein diacetate staining. The obtained fluorescence images were converted to gray scale images and quantitatively analyzed using NIH image software (Scion, Frederick, MD, USA). Results are expressed as the mean percentages of the levels in control±s.e. from four independent experiments. Po.5; Po.1. attenuated oxidative stress in the kidneys of streptozotocininduced diabetic rats. Taken together, the present results suggest that downregulation of NAD(P)H oxidase components, especially NOX4, by hyperbilirubinemia and BVD administration may have an important role in the inhibition of oxidative stress in the kidneys of diabetic rodents. We also confirmed the inhibitory effect of hyperbilirubinemia and BVD on NAD(P)H oxidase activities by the lucigenin method. The signaling mechanisms underlying the rapid activation of non-phagocytic NAD(P)H oxidase in vascular cells have been relatively well established. One of the most important physiological regulators is. 28 rapidly activates phospholipase C to increase intracellular calcium and diacylglycerol levels, which cause the activation of protein kinase C. Activated protein kinase C phosphorylates p47phox and induces the release of superoxide from NOX components. We and other investigators have also reported that high glucose levels stimulate superoxide production from vascular endothelial cells and smooth muscle cells via a protein kinase C-dependent activation of NAD(P)H oxidase In the diabetic state, high glucose levels and activation of local renin-angiotensin systems may induce the rapid activation of NAD(P)H oxidase in renal and vascular tissues of diabetic patients and animals. 29,3 A previous report showed that the hydrophobic tetrapyrrole structure of BIL can inhibit NAD(P)H oxidase activity directly in a cellfree system, 1 probably via inhibition of the interactions 916 Kidney International (21) 78,

13 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy NOX4 mrna h 12 h 18 h 24 h NOX NOX4 β-actin 67 kda (1 1 7 (1 1 7 BiIirubin (1 1 6 (1 1 7 BiIiverdin (1 1 6 (1 1 7 NAC (5 1 3 (1 1 7 α-lipoic acid (1 1 4 Angiotensin II Biliverdin Bilirubin NAC α-la 6 3 TGF-β1 mrna (% fo control) Fibronectin mrna (1 1 7 (1 1 7 BiIirubin (1 1 6 (1 1 7 BiIiverdin (1 1 6 (1 1 7 NAC (5 1 3 (1 1 7 α-lipoic acid (1 1 4 (1 1 7 (1 1 7 BiIirubin (1 1 6 (1 1 7 BiIiverdin (1 1 6 (1 1 7 NAC (5 1 3 (1 1 7 α-lipoic acid (1 1 4 Figure 1 The effect of bilirubin (BIL) and biliverdin (BVD) on angiotensin II () (1 nmol/l stimulated expression of NOX4 (a), transforming growth factor-b1 (TGF-b1) (c), and fibronectin (d) mrna levels in cultured human mesangium cells. The extracted mrna levels were measured by real-time reverse transcriptase PCR and levels of each mrna were normalized to the levels of b-actin. Results are expressed as the mean percentages of the levels in control±s.e. Po.5; Po.1; Po.1 vs control. NOX4 protein levels were also evaluated by western blot analysis (b). The levels of protein were normalized to the level of b-actin and the results are expressed as the mean percentages of the levels in control±s.e. Po.5 vs. between regulatory components and NOX family members. In the present study, we showed that BIL and BVD inhibited NAD(P)H activities in cultured renal mesangial cells, and inhibited high glucose-induced and -induced ROS production. The inhibitory effect of BIL and BVD on the expression of NAD(P)H oxidase components were also confirmed. All these findings suggest that BIL and BVD may inhibit oxidative stress in diabetic kidneys by their inhibitory effects on both the activation and the overexpression of NAD(P)H oxidase. Furthermore, the present study revealed that BIL and BVD were more effective in preventing -induced expression of NAD(P)H oxidase, TGFb, and fibronectin in cultured mesangial cells than maximal doses of other antioxidants such as N-acetylcysteine and a-lipoic acid, supporting the effectiveness of BIL and BVD on diabetic nephropathy. The detailed molecular mechanisms for the beneficial effects of BIL and BVD should be clarified in future studies. Accumulating evidence has shown that subjects with higher BIL levels have a lower incidence of cardiovascular disease and strokes Recently, we showed a markedly lower prevalence of nephropathy as well as cardiovascular Kidney International (21) 78,

14 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy disease in diabetic patients with Gilbert syndrome. 16 One report has also shown that serum BIL level is associated with microalbuminuria in patients with type 2 diabetes. 34 These epidemiological findings showing the relationship between serum BIL level and prevalence of diabetic nephropathy are consistent with the present findings implicating that hyperbilirubinemia and BVD administration protect against diabetic nephropathy. In conclusion, we showed for the first time that hyperbilirubinemia and BVD administration protected against diabetic nephropathy via inhibition of oxidative stress, at least in part by downregulating renal NAD(P)H oxidase, in animal models of diabetes. The present findings may lead to novel antioxidant therapies for early stages of diabetic nephropathy. MATERIALS AND METHODS Animals Male homologous Gunn j/j rats and age-matched heterozygous Gunn j/ þ rats were purchased from Clea Japan, Tokyo, Japan. At 8 weeks of age, half of the Gunn j/j rats (n ¼ 1) and half of the Gunn j/ þ rats (n ¼ 1) were injected intraperitoneally with streptozotocin (Sigma, St Louis, MO, USA) at 8 mg/kg body weight. At 1 or 2 days after injection, the development of diabetes was verified by the presence of hyperglycemia (plasma glucose levels mmol). Male C57BL/KsJ mice and their age-matched lean littermates, db/ þ mice, were purchased from Clea Japan. All mice were bred under pathogen-free conditions at Kyushu University Animal Center, Fukuoka, Japan. The animals had free access to tap water and standard chow (Clea Japan). At 12 weeks of age, half of the (n ¼ 8) and half of the db/ þ (n ¼ 8) mice were randomly chosen to receive powder diet (Clea Japan) supplemented with BVD (5 mg/kg; Frontier Scientific, Logan, UT, USA) and the rest of the mice received powder diet without any supplementation for 2 or 12weeks. Powder diet was stored at 41C and kept away from excessive light. All protocols were reviewed and approved by the Ethics Committee of Animal Experiments, Graduate School of Medical Science, Kyushu University. Human mesangium cells and culture Normal human mesangium cells were obtained from Lonza (Walkersville, MD, USA) and cultured in BulletKit containing 5% fetal bovine serum, 5 mg/ml gentamicin, and 5 mg/ml amphotericin-b in a humidified air/5% CO 2 atmosphere at 371C. All cells were studied at passage 4. For more details, see online Supplement. Blood and urine analysis Plasma concentrations of total and direct BIL were measured and a 24-h urine sample was collected using metabolic cages to analyze urinary albumin and oxidative stress makers. The well-mixed urine was centrifuged at 75 g for 5 min, purged of air with a steam of nitrogen to prevent artificial formation of oxidative stress products, and then stored at 81C until analysis. Relating to these analyses, see online Supplement for details. Immunohistochemistry Immunostaining for 8-OHdG and NOX4 in the kidney was performed as previously described. 24 Methods are described in the online appendix section. DHE stain in the renal tissues To detect ROS in the renal glomeruli and tubules, DHE (Invitrogen, Carlsbad, CA, USA) stain was performed as previously described 35 with a slightly modified procedure. The detailed method is summarized in the online appendix section. Isolaion of glomeruli Isolation of mice glomeruli was performed using Dynabeads M-45 tosylactivated and Magnetic Particle Concentrator (Dynal AS, Oslo, Norway) as previously described. 36 We performed a slightly modified procedure. The detailed method is summarized in the online appendix section. RNA extraction and quantitative reverse transcriptase-pcr Methods are described in the online appendix section. Morphological study The mesangial matrix was determined by the presence of periodic acid-schiff-positive and nuclei-free areas. The detailed method is summarized in the online appendix section. Western blot analysis Methods are described in the online appendix section. Measurement of NAD(P)H oxidase activities NAD(P)H oxidase activities in renal cortex tissues and mesangial cell homogenates were evaluated by the lucigenin-enhanced chemiluminescence method. Intracellular oxidative stress was evaluated by the 2,7 -dichlorofluorescein diacetate staining method. Both detailed methods are summarized in the online appendix section. Statistical analysis All data are expressed as means±s.e.m. Statistical analysis was performed with Student s t-test or one-way analysis of variance with Fisher s protected least significant difference test. Po.5 was considered statistically significant. DISCLOSURE All the authors declared no competing interests. ACKNOWLEDGMENTS This work was supported by a Grant-in-Aid for Scientific Research (No ) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan, and by the Special Coordination Funds for Promoting Science and Technology (SCF; funding program Innovation Center for Medical Redox Navigation ). SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at REFERENCES 1. Baynes JW. Role of oxidative stress in development of complications in diabetes. Diabetes 1991; 4: Oberley LW. Free radicals and diabetes. Free Radic Biol Med 1988; 5: Ha H, Kim C, Son Y et al. DNA damage in the kidneys of diabetic rats exhibiting microalbuminuria. Free Radic Biol Med 1994; 16: Kakimoto M, Inoguchi T, Sonta T et al. Accumulation of 8-hydroxy-2 - deoxyguanosine and mitochondrial DNA deletion in kidney of diabetic rats. Diabetes 22; 51: Kidney International (21) 78,

15 M Fujii et al.: Bilirubin/biliverdin ameliorate DM nephropathy 5. Wolff SP, Jiang ZY, Hunt JV. Protein glycation and oxidative stress in diabetes mellitus and ageing. Free Radic Biol Med 1991; 1: Stocker R, Yamamoto Y, McDonagh AF et al. Bilirubin is an antioxidant of possible physiological importance. Science 1987; 235: Stocker R. Induction of haem oxygenase as a defence against oxidative stress. Free Radic Res Commun 199; 9: Abraham NG, Kappas A. Heme oxygenase and the cardiovascular-renal system. Free Radic Biol Med 25; 39: Morita T. Heme oxygenase and atherosclerosis. Arterioscler Thromb Vasc Biol 25; 25: Kwak JY, Takeshige K, Cheung BS et al. Bilirubin inhibits the activation of superoxide-producing NADPH oxidase in a neutrophil cell-free system. Biochim Biophys Acta 1991; 176: Lanone S, Bloc S, Foresti R et al. Bilirubin decreases nos2 expression via inhibition of NAD(P)H oxidase: implications for protection against endotoxic shock in rats. FASEB J 25; 19: Inoguchi T, Li P, Umeda F et al. High glucose level and free fatty acid stimulate reactive oxygen species production through protein kinase C dependent activation of NAD(P)H oxidase in cultured vascular cells. Diabetes 2; 49: Hink U, Li H, Mollnau H et al. Mechanisms underlying endothelial dysfunction in diabetes mellitus. Circ Res 21; 88: E14 E Kim YK, Lee MS, Son SM et al. Vascular NADH oxidase is involved in impaired endothelium-dependent vasodilation in OLETF rats, a model of type 2 diabetes. Diabetes 22; 51: Inoguchi T, Sonta T, Tsubouchi H et al. Protein kinase C-dependent increase in reactive oxygen species (ROS) production in vascular tissues of diabetes: role of vascular NAD(P)H oxidase. J Am Soc Nephrol 23; 14: S227 S Inoguchi T, Sasaki S, Kobayashi K et al. Relationship between Gilbert syndrome and prevalence of vascular complications in patients with diabetes. JAMA 27; 298: Li Y, Zhu H, Kuppusamy P et al. Validation of lucigenin (bis-nmethylacridinium) as a chemilumigenic probe for detecting superoxide anion radical production by enzymatic and cellular systems. J Biol Chem 1998; 273: Tsubouchi H, Inoguchi T, Sonta T et al. Statin attenuates high glucoseinduced and diabetes-induced oxidative stress in vitro and in vivo evaluated by electron spin resonance measurement. Free Radic Biol Med 25; 39: Ziyadeh FN. Role of transforming growth factor beta in diabetic nephropathy. Exp Nephrol 1994; 2: Stocker R. Antioxidant activities of bile pigments. Antioxid Redox Signal 24; 6: Lassegue B, Clempus RE. Vascular NAD(P)H oxidases: specific features, expression, and regulation. Am J Physiol Regul Integr Comp Physiol 23; 285: 277 R Shiose A, Kuroda J, Tsuruya K et al. A novel superoxide-producing NAD(P)H oxidase in kidney. J Biol Chem 21; 276: Geiszt M, Kopp JB, Varnai P et al. Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci USA 2; 97: Etoh T, Inoguchi T, Kakimoto M et al. Increased expression of NAD(P)H oxidase subunits, NOX4 and p22phox, in the kidney of streptozotocininduced diabetic rats and its reversibity by interventive insulin treatment. Diabetologia 23; 46: Sonta T, Inoguchi T, Tsubouchi H et al. Evidence for contribution of vascular NAD(P)H oxidase to increased oxidative stress in animal models of diabetes and obesity. Free Radic Biol Med 24; 37: Fujii M, Inoguchi T, Maeda Y et al. Pitavastatin ameliorates albuminuria and renal mesangial expansion by downregulating NOX4 in mice. Kidney Int 27; 72: Gorin Y, Block K, Hernandez J et al. Nox4 NAD(P)H oxidase mediates hypertrophy and fibronectin expression in the diabetic kidney. J Biol Chem 25; 28: Griendling KK, Minieri CA, Ollerenshaw JD et al. Angiotensin II stimulates NADH and NADPH oxidase activity in cultured vascular smooth muscle cells. Circ Res 1994; 74: Inoguchi T, Nawata H. NAD(P)H oxidase activation: a potential target mechanism for diabetic vascular complications, progressive beta-cell dysfunction and metabolic syndrome. Curr Drug Targets 25; 6: Kitada M, Koya D, Sugimoto T et al. Translocation of glomerular p47phox and p67phox by protein kinase C-beta activation is required for oxidative stress in diabetic nephropathy. Diabetes 23; 52: Breimer LH, Wannamethee G, Ebrahim S et al. Serum bilirubin and risk of ischemic heart disease in middle-aged British men. Clin Chem 1995; 41: Hopkins PN, Wu LL, Hunt SC et al. Higher serum bilirubin is associated with decreased risk for early familial coronary artery disease. Arterioscler Thromb Vasc Biol 1996; 16: Perlstein TS, Pande RL, Creager MA et al. Serum total bilirubin level, prevalent stroke, and stroke outcomes: NHANES Am J Med 28; 121: e Targher G, Bosworth C, Kendrick J et al. Relationship of serum bilirubin concentrations to kidney function and albuminuria in the United States adult population. Findings from the National Health and Nutrition Examination Survey Clin Chem Lab Med 29; 47: Chen HY, Chen TY, Lee MY et al. Melatonin decreases neurovascular oxidative/nitrosative damage and protects against early increases in the blood-brain barrier permeability after transient focal cerebral ischemia in mice. J Pineal Res 26; 41: Takemoto M, Asker N, Gerhardt H et al. A new method for large scale isolation of kidney glomeruli from mice. Am J Pathol 22; 161: Kidney International (21) 78,