Molecular and Cellular Biochemistry

, Volume 382, Issue 1, pp 185–191

NOX1 abet mesangial fibrogenesis via iNOS induction in diabetes

Authors

  • Ling Gao
    • Department of EndocrinologyRenmin Hospital of Wuhan University
  • Weilu Huang
    • Department of EndocrinologyRenmin Hospital of Wuhan University
    • Department of EndocrinologyRenmin Hospital of Wuhan University
Article

DOI: 10.1007/s11010-013-1733-4

Cite this article as:
Gao, L., Huang, W. & Li, J. Mol Cell Biochem (2013) 382: 185. doi:10.1007/s11010-013-1733-4

Abstract

Both NADPH oxidase (NOX) and inducible nitric oxide synthase (iNOS) are the main sources of reactive oxygen species in kidney. However, their interactions in oxidative stress and contributions to kidney fibrosis during diabetic nephropathy have not been studied. Human mesangial cells were treated with normal glucose (5.6 mmol/L), high glucose (30 mmol/L) in the presence or absence of AGE (200 mg/L). Protein expressions of NOX1, NOX2, NOX4, and iNOS were examined by immunoblotting. NOX was genetically silenced with specific RNAi to study the interactions between NOX and iNOS in diabetic milieu. Superoxide (O·−) and peroxynitrite (ONOO·−) productions were assessed by dihydroethidium and hydroxyphenyl fluorescein, respectively. Fibrotic factors were determined by biochemistry assay. Superoxide, peroxynitrite, TGF-β, and fibronectin productions as well as the protein expressions of NOX1, NOX2, NOX4, and iNOS were increased in the diabetic milieu (high glucose 30 mmol/L plus AGE 200 mg/L). However, abolishment of iNOS induction with 1400W or iNOS RNAi would restore peroxynitrite, TGF-β, and fibronectin productions completely to basal level and attenuate superoxide production. Moreover, NOX1 inhibition not only prevented iNOS induction but also abrogated changes consequent to iNOS induction such as mesangial fibrogenesis.

Keywords

Inducible nitric oxide synthase (iNOS)Human glomerular mesangial cells (HMCs)Advanced glycosylation end products (AGEs)Oxidative stress (OS)Transform growth factor-β (TGF-β)Fibronectin

Introduction

Diabetic nephropathy (DN) is one of the most common and serious complications of diabetes, as well as the leading cause of death for diabetic patients [1]. Although the pathogenesis of DN is not yet clear, kidney fibrosis has long been believed a prominent feature of DN [2]. It has been demonstrated that iNOS is the major subtype of NOS dwelling in kidney, whose induction plays a critic role in the early stage of DN [3]. However, the immediate consequence of iNOS induction in diabetes and its upstream signaling remains elusive.

The accumulation of advanced glycation end products (AGEs) has been implicated in the progression of diabetes. AGE formation is increased by accelerated glycation reaction due to hyperglycemia. Recent studies indicate that NAD(P)H oxidase is one of the major sources of ROS in renal cells and serves as the mediator between AGE and iNOS induction in muscle cells [4]. NAD(P)H oxidase is generally composed of one catalytic and several regulatory subunits, in which the catalytic subunits, termed NOX proteins. In this family, NOX1, NOX2, and NOX4 appear to be abundant in kidney [5]. Therefore, the relation of different NOXs activations and iNOS induction, their derived ROS, the subsequent oxidative stress and fibrotic factors induced by AGEs were examined in cultured human mesangial cells (HMCs). Moreover, since both AGE and high glucose are present in diabetes, the combination effects of these two factors were also investigated as a reflection of high glucose environment.

Subjects and methods

Cell culture

Institutional Review Board (IRB)/Ethics Committee approval was obtained and the study was in adherence with the Declaration of Helsinky. HMCs (supplied by Xiangya School of Medicine, Central South University, China) were cultured in dulbecco modified eagle’s medium (DMEM) containing glucose (5.6 mmol/L) with 15 % fetal bovine serum (FBS), 100 μg/mL streptomycin, 100 unit/mL penicillin, 2 mmol/L glutamine at 37 °C, in a humidified 5 % CO2 atmosphere. Cells of passages from 3 to 6 were used in the experiments. After reaching 90 % confluence, cells were tranquilized with a medium containing no FBS for 48 h, and then exposed to different groups: Control (5.6 mmol/L glucose + 200 mg/L BSA), high glucose (30 mmol/L glucose + 200 mg/L BSA), AGE (5.6 mmol/L glucose + 200 mg/L AGEs), high glucose plus AGE (30 mmol/L glucose + 200 mg/L AGEs) for 48 h.

Superoxide measurement

After the treatment, cells were collected and re-suspended into 96-well microtitre plates for 2 h incubation with 10 μM dihydroethidium (DHE). The fluorescence was measured with microplate reader (Perkin Elmer 1420) at excitation 450 nm and emission 535 nm.

Intracellular ONOO·− measurement

Hydroxyphenyl fluorescein (HPF) was used to detect ONOO (peroxynitrite). Cell suspension (200 μL, 107/mL) and 10 μM HPF were added to 96-well plates which were incubated for 30 min at 37 °C in the dark. The fluorescence was measured with microplate reader (Perkin Elmer 1420) at excitation 485 nm and emission 585 nm.

Measurements of TGF-β1 and FN by enzyme-linked immunosorbent assay (ELISA)

The protein levels of TGF-β 1 and FN in cultured cells were quantified spectrophotometrically at a wavelength of 450 nm using the TGF-β 1 ELISA kit (Boster, Wuhan, China), and the FN ELISA kit (USCNLIFE, Wuhan, China) according to the protocols of the manufacturers.

Immunoblotting analysis

In brief, cells were lysed in 50 mM Tris–HCl buffer (pH 8, containing 0.2 % NP-40, 180 mM NaCl, 0.5 mM EDTA, protease inhibitors, 1 M DTT, and 100 mM phenylmethylsulfonyl fluoride). Equal amounts of lysates were separated by 10 % SDS-PAGE and transferred to PVDF membranes. Membranes were blocked with 5 % milk overnight, and blotted with primary antibodies against iNOS (mouse, Abcam, 1:1,000); NOX1 (Goat, Abcam, 1:1,000), NOX2 (Rabbit, Abcam, 1:1,000) and NOX4 (Rabbit, Abcam, 1:1,000) separately following standard procedures.

In vivo RNA interference of NOX1, NOX2, NOX4, and iNOS

The siRNAs against NOX1, NOX2, NOX4, or iNOS were obtained from Invitrogen (Grand Island, NY, USA). HMGs at 95 % confluency were incubated with 50 nmol/L siRNA/oligofectamine (invitrogen, China) mixtures according to manufacture protocol for 48 h.

Statistical analysis

Data were calculated and expressed as mean ± SD. Student’s t test and ANOVA were used to analyze the data and p value <0.05 is considered significant.

Results

The effects of high glucose and AGE on the O·− production

The production of O·− (superoxide) in HMCs was evaluated by DHE fluorescence staining (Fig. 1a). The fluorescence/O·− production was significantly increased by high glucose (p < 0.05). Moreover, it was further increased by AGE (p < 0.01). On the other hand, iNOS inhibitor, 1400W restored superoxide generation induced by elevated glucose and AGE, but only partially.
https://static-content.springer.com/image/art%3A10.1007%2Fs11010-013-1733-4/MediaObjects/11010_2013_1733_Fig1_HTML.gif
Fig. 1

Effects of hyperglycemia and AGE on ROS and fibrotic factors production. Human glomerular mesangial cells (HMCs) were treated with normal glucose (5.6 mM), high glucose (30 mM), AGEs (5.6 mM + 20 mg/dL), high glucose, and AGEs (30 mM + 200 mg/L) medium in the presence or absence of iNOS inhibitor (1400W, 100 μM). a The superoxide production in various conditions detected by DHE; b The peroxynitrite production in various conditions detected by HPF; c The TGF-βproduction in various conditions assessed by biochemical assay; d The fibronectin production in various conditions assessed by biochemical assay. *p < 0.05

The effects of high glucose and AGE on the peroxynitrite production

Peroxynitrite was generated when nitric oxide reacts with superoxide. It was measured by HPF fluorescence staining (Fig. 1b). HPF fluorescence was not altered by high glucose alone, but was significantly increased in the presence of AGE. However, iNOS inhibitor, 1400W completely abolished the peroxynitrite overproduction by AGE.

The effects of high glucose and AGE on fibrotic factors production

In a similar pattern to peroxynitrite production, both TGF-β and fibronectin productions were increased in AGE group and high glucose plus AGE group. However, glucose alone did not have any effect on these fibrotic factors production. On the other hand, 1400W resumed both TGF-β and fibronectin productions back to normal range. (p < 0.05) (Fig. 1c, d).

The effects of hyperglycemia and AGE on NOX1, NOX2, and NOX4 protein expressions

NOX1, NOX2, and NOX4 expressions in HMCs were examined by immunoblotting. In Fig. 2, only NOX2 and NOX4 were upregulated by high glucose. In the presence of AGE, NOX2 and NOX4 expressions were not further increased, but remained significantly higher than control. However, iNOS and NOX1 expressions were not altered by high glucose alone. When AGE was added, both iNOS and NOX1 were upregulated. In addition, 1400W did not change the response of NOX1, NOX2, and NOX4 to AGE.
https://static-content.springer.com/image/art%3A10.1007%2Fs11010-013-1733-4/MediaObjects/11010_2013_1733_Fig2_HTML.gif
Fig. 2

Effects of hyperglycemia and AGE on NOX1, NOX2, NOX4, and iNOS protein expression. a NOX1, NOX2, NOX4, iNOS, and β-actin expression were examined in HMCs treated with normal glucose (5.6 mM), high glucose (30 mM), AGEs (5.6 mM + 200 mg/L), high glucose, and AGEs (30 mM + 200 mg/L) medium. The density analysis of b iNOS/β-actin, c NOX1/β-actin, d NOX2/β-actin, and e NOX4/β-actin from immunoblots (n = 3, data are expressed as the mean ± SEM of three experiments). *p < 0.05

The effect of iNOS inhibition on AGE-induced NOXs upregulations

To clarify the interaction between iNOS and different NOX isoforms, RNAi of iNOS was transfected to HMCs and NOXs expressions were examined with immunoblot (Fig. 3). Clearly, the iNOS induction by AGE was canceled by iNOS RNAi. Similar to what was observed with 1400W, NOX1, NOX2, and NOX4 upregulation by AGE were not altered by cancelation of iNOS induction.
https://static-content.springer.com/image/art%3A10.1007%2Fs11010-013-1733-4/MediaObjects/11010_2013_1733_Fig3_HTML.gif
Fig. 3

Effects of iNOS knockdown on different NOX isoform expressions in the presence of AGE. a After transfection with iNOS RNAi, NOX1, NOX2, and NOX4 expression were examined in HMSc treated with normal glucose (5.6 mM), high glucose (30 mM), AGEs (5.6 mM + 200 mg/L), high glucose, and AGEs (30 mM + 200 mg/L). b The density analysis of iNOS/β-actin from immunoblots (n = 3, data are expressed as the mean ± SEM of three experiments). *p < 0.05

The differential effects of specific NOX knockdown on iNOS induction by AGE

In the presence of high glucose and AGE, the specific NOX1, NOX2, and NOX4 RNAi were used to transfect HMCs and the protein expression of iNOS was examined to further explore the interaction between iNOS and NOXs (Fig. 4). The iNOS induction by high glucose and AGE was abolished by NOX1 knockdown but not altered by NOX2 and NOX4 knockdowns.
https://static-content.springer.com/image/art%3A10.1007%2Fs11010-013-1733-4/MediaObjects/11010_2013_1733_Fig4_HTML.gif
Fig. 4

The effects of specific NOX isoform knockdown on iNOS induction. HMCs were treated with normal glucose (5.6 mM) or high glucose and AGE (30 mM + 200 mg/L) in the presence of different specific NOX isoform (NO1, NOX2, NOX4) RNAi or scramble RNAi. The iNOS expression and NOX protein expression were examined with immunoblot. a The density analysis of iNOS/β-actin and NOX1/β-actin from immunoblots (n = 3, data are expressed as the mean ± SEM of three experiments). b The density analysis of iNOS/β-actin and NOX2/β-actin from immunoblots (n = 3, data are expressed as the mean ± SEM of three experiments). c The density analysis of iNOS/β-actin and NOX4/β-actin from immunoblots (n = 3, data are expressed as the mean ± SEM of three experiments). *p < 0.05

The effect of NOX1 knockdown on ROS and fibrotic factors production by AGE

In a similar pattern to iNOS inhibition, the knockdown of NOX1 resumed the peroxynitrite, TGF-β, and fibronectin productions back to the basal level in the presence of high glucose and AGE (Fig. 5). As expected, NOX1 knockdown only partially restored the superoxide production since other NOXs are still provocative.
https://static-content.springer.com/image/art%3A10.1007%2Fs11010-013-1733-4/MediaObjects/11010_2013_1733_Fig5_HTML.gif
Fig. 5

The effect of NOX1 inhibition on ROS and fibrotic factors production. a ROS production indexed by DHE fluorescence was examined in HMCs treated with normal glucose (5.6 mM) or high glucose and AGEs (30 mM + 200 mg/L) in the presence of either specific NOX1 RNAi or scramble RNAi. b, c & d Peroxynitrite, TGF-β and fibronectin productions were also examined in the above conditions. (n = 3, data are expressed as the mean ± SEM of three experiments). *p < 0.05

Discussion

Kidney fibrosis is the key feature of DN, and ROS derived from NOX or NOS may be implicated in the process [6, 7]. However, the mechanisms of ROS generation in DN and its role in kidney fibrosis are not fully understood. In this study, we demonstrate that NOX1–iNOS interaction played a major role in mesangial fibrogenesis in diabetes.

Numerous studies have demonstrated matrix proteins accumulation under diabetes milieu either in mesangial cells or the kidney of diabetic model [8, 9]. In accordance with that, we have observed prominent amount of fibronectin production in mesangial cells treated with high glucose plus AGE after 48 h as evidence of fibrosis. The major fibrotic factor, TGF-β production was also increased by high glucose plus AGE after 48 h which could contribute to the matrix protein accumulations in mesangial cells.

Some studies have shown that oxidative stress mediated the detrimental effects of diabetes (high glucose, AGE or both) on kidney [10, 11]. Oxidative stress is the imbalance between superoxide production and antioxidant defense. Compared to impairment of antioxidant defense, superoxide overproduction is more of the original cause. Not surprisingly, we observed a slight but robust increase of superoxide production in high glucose treated HMCs detected by DHE staining. In the presence of AGE, superoxide production went up even higher. However, peroxynitrite, the product of superoxide and nitric oxide reaction was also increased significantly by AGE, which suggests possible involvement of nitric oxide synthase (NOS). 1400W, the iNOS inhibitor was then used and it showed that superoxide overproduction is only partially diminished.

Three NOS isoforms have been found in the kidney but only iNOS is reported to be present in the mesangial cells [12, 13]. Early studies from Chang et al. [14] showed that exaggerated peroxynitrite formation, generated from induced iNOS had a pivot role in glomerular lesions in diabetic rats. It also suggested that cytokine release, NF-кB and p38 MAPK-dependent pathways mediated AGE-induced iNOS expression and subsequent nitric oxide production in mesangial cells. Sugimoto et al. [15] reported that both AGE accumulation and iNOS expression are increased in the glomerular mesangial area in STZ-induced diabetic mice and after the use of AGEs inhibitor aminoguanidine, the expression of them both decreased. Our study corroborated the above findings of iNOS in the development of DN or in the AGE-induced fibrogenesis of mesangial cells. However, there is still missing pieces between AGE accumulation and iNOS upregulation which requires further investigation.

Advanced glycosylation end product is the irreversible attachment of reducing sugars onto amino groups of proteins to form AGEs [16, 17]. AGE can be formed either extracellularly or intracellularly. The cellular dysfunction were caused either due to macromolecules modification by AGE directly or activation of signaling pathways through AGE–RAGE axis. NOX is a possible transit linking AGE/RAGE axis and its downstream iNOS and matrix protein production. We found that NOX1, NOX2, and NOX4 all present in the mesangial cells, and all of them were upregulated in response to high glucose plus AGE treatment. It is noteworthy that NOX2 and NOX4 were activated with high glucose alone, but NOX1 and iNOS were activated in the presence of AGE.

Wautier [16] found that AGE/RAGE can induce ROS overproduction via NADPH oxidase activation. The application of NADPH oxidase inhibitor apocynin, and diphenyliodonium (DPI) can effectively inhibit the production of ROS in mesangial cells [18], indicating that the AGE/RAGE-mediated ROS production depended on NADPH oxidase. Tyml and coworkers [19] found that activation of the NADPH oxidase can lead to increased expression of iNOS, while adding the NADPH oxidase inhibitor DPI downregulated iNOS expression suggesting that the expression of iNOS is depended on the activation of NADPH oxidase. Fan et al. [20] found that candesartan, an ARB, reduces iNOS expression and subsequent albuminuria by down-regulating the NADPH oxidase-p47phox component or attenuating RAGE expression in type 2 diabetic KK/Ta mouse kidneys. My previous work has shown that NOX1 activation is the upstream of eNOS uncoupling in aorta of diabetic mice [21]. We wondered similar mechanism may present in mesangial cells during diabetes. First, both 1400W and iNOS RNAi were used to examine whether iNOS inhibition has any effect on NOX activation. However, iNOS inhibition did not abolish the NOX activation by AGE or glucose. Therefore, iNOS induction cannot be the upstream of NOX activation in diabetes. On the other hand, specific NOX1, NOX2, and NOX4 RNAi were used respectively to examine whether they can be the upstream of iNOS induction during high glucose and AGE treatment. It turned out that only NOX1 inhibition completely abolished iNOS induction suggesting NOX1 activation is the upstream of iNOS induction in diabetes. Moreover, NOX1 inhibition would suppress or restore superoxide, peroxynitrite, TGF-β, and fibronectin production in a similar pattern as iNOS inhibition, which further confirmed involvement of NOX1–iNOS pathway in mesangial fibrogenesis during diabetes.

We have for the first time showed that NOX1 abet mesangial fibrogenesis via iNOS induction during diabetes. However, it is possible that other signaling pathways could be involved. Therefore, it would be important to further explore the gap between AGE–RAGE axis and iNOS induction in more detail. Moreover, the coupling status of iNOS is largely ignored and more work need to be done to examine the essential factor for iNOS coupling status: tetrahydrobiopterin level and its synthetic pathways etc.

Acknowledgments

This study is supported by the National Natural Science Foundation of China (Project # 81170767, Dr. Gao).

Copyright information

© Springer Science+Business Media New York 2013