Possible roles of tumor necrosis factor-α and angiotensin II type 1 receptor on high glucose-induced damage in renal proximal tubular cells

Abstract

Recent studies have identified that high glucose-induced renal tubular cell damage. We previously demonstrated that high glucose treatment induced oxidative stress in human renal proximal tubular epithelial cells (RPTECs), and angiotensin II type 1 (AT1) receptor blockers reduce high glucose-induced oxidative stress in RPTEC possibly via blockade of intracellular as well as extracellular AT1 receptor. However, exact roles of tumor necrosis factor (TNF)-α and AT1 receptor on high glucose-induced renal tubular function remain unclear. N-acetyl-beta-glucosaminidase (NAG), concentrations of TNF-α/angiotensin II and p22^phox protein levels after high glucose treatment with or without AT1 receptor blocker or thalidomide, an inhibitor of TNF-α protein synthesis, were measured in immortalized human renal proximal tubular epithelial cells (HK2 cells). AT1 receptor knockdown was performed with AT1 receptor small interfering RNA (siRNA). High glucose treatment (30 mM) significantly increased NAG release, TNF-α/angiotensin II concentrations in cell media and p22^phox protein levels compared with those in regular glucose medium (5.6 mM). Candesartan, an AT1R blocker, showed a significant reduction on high glucose-induced NAG release, TNF-α concentrations and p22^phox protein levels in HK2 cells. In addition, significant decreases of NAG release, TNF-α concentrations and p22^phox protein levels in HK2 cells were observed in high glucose-treated group with thalidomide. AT1R knockdown with siRNA markedly reversed high glucose, angiotensin II or TNF-α-induced p22^phox protein levels in HK2 cells. TNF-α may be involved in high glucose-induced renal tubular damage in HK2 cells possibly via AT1 receptor signaling.

• Angiotensin II type 1 (AT1) receptor
• HK2 cells
• N-acetyl-beta-glucosaminidase
• p22^phox
• tumor necrosis factor alpha

Introduction

Renal proximal tubular cells (RPTCs) play several important physiological roles including glucose reabsorption1^,2. It is now well accepted that there is a local intrarenal renin–angiotensin system (RAS),3 and all components of the RAS are expressed in human, rat and murine RPTCs.4–8 Angiotensin II concentrations, angiotensinogen and renin mRNA expression in the kidney are increased in early diabetes,8–10 indicating that hyperglycemia or augmented intrarenal angiotensin II, or both, may be directly or indirectly responsible for renal proximal tubular hypertrophy and tubulointerstitial fibrosis in diabetes. In addition, our previous study demonstrated that angiotensin II type 1 (AT1) receptor blockers (candesartan, olmesartan or valsartan) showed a significant reduction in high glucose-induced N-acetyl-beta-glucosaminidase (NAG), 8-hydroxy-2′-deoxyguanosine and p22^phox protein levels in renal proximal tubular epithelial cells (RPTECs).11 Therefore, RAS in proximal tubular cells have been suggested to contribute to diabetic nephropathy.12 Cytokines including tumor necrosis factor (TNF)-α involved in the inflammatory process and endothelial damage have emerged as potential markers of progressive diabetic nephropathy.13 TNF-α is expressed, synthesized and released by infiltrating macrophages and by intrinsic kidney cells, namely, endothelial, mesangial, glomerular and tubular epithelial cells.14^,15 Moreover, a recent study demonstrated that aldosterone-enhanced TNF-α expression and secretion in renal proximal tubular cells.16

It is well known that a multicomponent phagocyte-type nicotinamide adenine dinucleotide phosphate [NAD(P)H] oxidase is a major source of reactive oxygen species (ROS) production in many nonspherocytic cells, including renal mesangial cells and tubular cells. TNF-α as well as angiotensin II activates increases various ROS in the mammalian cells.17^,18 A recent study suggested that ΑΤ1 receptor blocker preferentially restored increased free radical production in diabetic kidney.19 We previously demonstrated that significant increases of p22^phox protein [a component of NAD(P)H oxidase] levels were observed in TNF-α-treated group in human RPTCs, and ΑΤ1 receptor blocker significantly decreased TNF-α-induced p22^phox protein levels.20 However, precise mechanism between AT1 receptor and TNF-α on high glucose-induced renal tubular damage is not fully elucidated. In order to investigate the roles of AT1 receptor and TNF-α on high glucose-induced cell damage in immortalized human renal proximal tubular epithelial cells (HK2 cells), we examined the effects of candesartan or thalidomide, an inhibitor of TNF-α protein synthesis on high glucose-induced NAG release, TNF-α levels and the p22^phox protein levels. In addition, we utilized AT1R small interfering RNA (siRNA) to further explore the roles of AT1R on high glucose-induced cell damage.

Materials and methods

Cell preparation

HK2 cell line derived from normal kidney was purchased from ATCC (Manassas, VA). These cells were cultured in growth medium containing DMEM/F-12HAM, 10% FBS, penicillin (100 units/mL), streptomycin (100 µg/mL). HK2 cells between passages 8 and 10 were used in this study. HK2 cells were made quiescent by incubation in serum-free DMEM medium with 5.6 mM glucose for 24–48 h. All the following experiments were performed in quiescent HK2 cells.

High glucose and AT1 receptor blockers/TNF-α inhibitor treatment

HK2 cells were plated on 96-well plate for NAG studies and 6-well plate for Western blotting. On the experimental day, the media in plates or chambers were washed and replaced with fresh DMEM/F-12 media without FBS. Various doses of candesartan (Takeda Pharmaceutical Co., Ltd., Osaka, Japan) and/or TNF-α inhibitor thalidomide (Enzo Life Science, Plymouth Meeting, PA) were added to the media alone or in combination with glucose (30 mM as high glucose treatment) at 24 h prior to the various measurements.

NAG assay

NAG as a marker of lysosomal damage was assessed using the following modified commercial procedure (PPR Diagnostics, London, UK) as previously described.21

TNF-α and angiotensin II enzyme immunoassay

TNF-α and angiotensin II concentrations in cell media were assayed with the commercially available ELISA kits (Human TNF-α ELISA kit, Thermo Fisher Scientific Inc., Yokohama, Japan, and Human/Rat Angiotensin II ELISA, BioVendor, Candler, NC) following the manufacturer’s protocol.

Western blotting

The whole cells were collected and mixed in electrophoresis sample buffer and were incubated at 100 °C for 5 min and electrophoresed by SDS-PAGE. After electrophoresis, separated proteins were transferred to polyvinylidene difluoride membranes followed by blocking with 5% skim milk in TBS solution (0.15 M NaCl/20 mM Tris-HCl, pH 7.6) for 60 min at room temperature. The membrane was incubated with anti-p22^phox (sc-20781; Santa Cruz Biotechnology Inc., Santa Cruz, CA) or anti-α tubulin (sc-5286; Santa Cruz Biotechnology Inc.) at a final concentration of 1:200–1000 for anti- anti-p22^phox or 1:10,000 for anti-α tubulin in 5% skim milk in 1 × TBS solution overnight at 4°C. Subsequently, the membranes were washed for 5 min (× 5) and incubated with biotinylated F (ab')2 fragment of polyclonal swine anti-rabbit immunoglobulin for anti-p22^phox or polyclonal rabbit anti-mouse immunoglobulin for anti-α tubulin (Dako A/S, Glostrup, Denmark) as second antibodies for 30 min at room temperature and washed. The membranes were incubated with streptavidin horseradish peroxidase conjugate (Chemicon, Boronia, VIC, Australia) for 20 min at room temperature and washed in TBS + Tween 20 for 5 min (×3) and TBS for 10 min (× 2). The bands were visualized by Chemi-Lumi One Super (Nakalai Tesque Inc., Kyoto, Japan) and quantitated using a computer-assisted image analyzer (Kodak 1D Image Analysis Software, Scientific Imaging Systems, New Haven, CT).

Gene silencing by siRNA

AT1 receptor siRNA was purchased from Ambion Inc. (Silencer Select pre-designed: ID# s135416). In addition, Silencer select negative control #1 siRNA (Ambion Inc., Carlsbad, CA) was utilized as a negative control siRNA. The transfection was conducted using Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA). Transfection efficiency was assessed by measuring AT1 receptor protein expression by Western blot analysis. The AT1 receptor antibody was purchased from Santa Cruz Biotechnology Inc. (sc-1173).

Data analysis

All results are expressed as mean ± SEM. All analyses were performed with the program SPSS ver.20 statistical package (IBM, Armonk, NY). Data were analyzed by Mann–Whitney U test or ANOVA followed by Fisher’s protected least squares difference test. A level of p < 0.05 was considered to be statistically significant.

Results

Effects of AT1 receptor blocker or thalidomide treatment on high glucose-induced NAG release in HK2 cells

Glucose treatment (30 mM) increased NAG release in HK2 cells compared with glucose treatment (5.6 mM). Candesartan (1 × 10^–7mol/L) or thalidomide (4 × 10^–5mol/L) treatment significantly decreased high glucose-induced NAG release in HK2 cells (Figure 1).

Figure 1. Effects of AT1 receptor blocker or thalidomide treatment on high glucose-induced NAG release in HK2 cells. *Denotes significant alterations at p < 0.05, when compared to control group (5.6 mM glucose) and +denotes significant alterations at p < 0.05, when compared to high glucose treatment (30 mM) group. Notes: Glu, glucose; HG, high glucose treatment (30 mM); Cand, candesartan; and Thal, thalidomide.

Effects of AT1 receptor blocker or thalidomide treatment on high glucose-induced p22^phox protein levels in HK2 cells

Glucose treatment (30 mM) showed significant increases of p22^phox protein levels in HK2 cells compared with glucose treatment (5.6 mM) group. Candesartan (1 × 10^–7mol/L) or thalidomide (4 × 10^–5mol/L) treatment significantly reduced high glucose-induced p22^phox protein levels (Figure 2).

Figure 2. Effects of AT1 receptor blocker or thalidomide treatment on high glucose-induced p22^phox protein levels in HK2 cells. A representative example of Western blotting was shown above. *Denotes significant alterations at p < 0.05 when compared to control group (5.6 mM glucose). +Denotes significant alterations at p < 0.05, when compared to high glucose treatment (30 mM) group. Notes: Glu, glucose; HG, high glucose treatment (30 mM); Cand, candesartan; and Thal, thalidomide.

Effects of high glucose on TNFα concentrations in the cell media

Glucose treatment (30 mM) significantly increased TNF-α concentrations in cell media. Candesartan (1 × 10^–7mol/L) or thalidomide (4 × 10^–5mol/L) treatment significantly decreased high glucose-induced TNF-α concentrations in cell media (Figure 3).

Figure 3. Effects of high glucose on TNF-α concentrations in the cell media in HK2 cells. *Denotes significant alterations at p < 0.05 when compared to control group (5.6 mM glucose) and +Denotes significant alterations at p < 0.05 when compared to high glucose treatment (30 mM) group. Notes: Glu, glucose; HG, high glucose treatment (30 mM); Cand, candesartan; and Thal, thalidomide.

Effects of high glucose on angiotensin II concentrations in the cell media

Glucose treatment (30 mM) induced slight but significant increase of angiotensin II concentrations in cell media (Figure 4).

Figure 4. Effects of high glucose on angiotensin II concentrations in the cell media in HK2 cells. *Denotes significant alterations at p < 0.05 when compared to control group (5.6 mM glucose). Glu, glucose.

Effects of AT1 siRNA treatment on high glucose, angiotensin II- or TNF-α-induced p22^phox protein in HK2 cells

AT1 siRNA significantly reduced AT1 receptor protein compared with negative control siRNA treatment in HK2 cells (Figure 5a). Glucose treatment (30 mM), angiotensin II (1 × 10^–7mol/L) or TNF-α (10 ng/mL) significantly induced p22^phox protein in HK2 cells with negative control siRNA (Figure 5b). In contrast, the increment of p22^phox protein by high glucose, angiotensin II or TNF-α was markedly attenuated by AT1 siRNA (Figure 5c).

Figure 5. Effects of AT1 siRNA treatment on high glucose, angiotensin II- or TNF-α-induced p22^phox protein in HK2 cells. (a) A representative example of knockdown of the AT1R protein using siRNA. Effects of glucose treatment (30 mM), angiotensin II- or TNFα-induced p22^phox protein in HK2 cells with negative control siRNA (b) and with AT1 siRNA (c). *Denotes significant alterations at p < 0.05 when compared to control group (5.6 mM glucose). Glu, glucose and Ang, angiotensin.

Discussion

In this study, high glucose significantly increased NAG release, angiotensin II and TNF-α concentrations. These results are agreement with that high glucose increased angiotensinogen, angiotensin-converting enzyme and AT1R mRNA levels and angiotensin II concentrations in the renal tubular cell culture medium.22 Moreover, a recent study investigated the role of TNF-α/TNF receptors in the development of early and late diabetic nephropathy,23 suggesting that angiotensin II/AT1R and TNF-α may be involved in high glucose-induced renal tubular damage.

High glucose (25 mM) evoked ROS generation and p38 MAPK phosphorylation as well as stimulated immunoreactive rat angiotensin II secretion and angiotensin II mRNA expression in rat immortalized renal proximal tubular cells.24 While, TNF-α markedly induced NAD(P)H oxidase-derived ROS generation in rat embryonic-heart derived H9c2 cells.25 In addition, TNF can induce activation of NAD(P)H oxidase leading to enhanced oxidative stress and decreased bioavailability of NO.26 We also demonstrated that high glucose and TNF-α markedly induced p22^phox protein levels,11^,20 suggesting that angiotensin II/AT1R and TNF-α may be involved in high glucose-induced ROS generation in RPTECs.

The mechanisms on interaction between TNF-α and RAS in high glucose-induced renal tubular damage were investigated. Both in vitro and in vivo studies have demonstrated a cross-talk between angiotensin II and TNF.27^,28 Angiotensin II can act as a potent proinflammatory agent and stimulate the production of chemokines such as MCP-1, and PICs, such as TNF, interleukin (IL)-6 and IL-1β in the brain.29^,30 Olmesartan, an AT1 receptor blocker, has been investigated and found to inhibit the expression of TNF-α in vascular smooth muscle cells.31 In addition, olmesartan resulted in decreased levels of TNF-α in male Wistar albino rats,32 suggesting that the stimulatory role of AT1R on TNF-α production. In this study, the siRNA-mediated AT1 receptor knockdown abolished high glucose, angiotensin II and TNF-α-induced p22^phox protein levels in HK2 cells, supporting a role for AT1R in the development of ROS generation. These results are keeping with the report that the AT1 receptor blocker decreased the TNF-α-dependent increase of 8-isoprostane, phosphorylation of I kappa B alpha and cell death in cultured human umbilical vein endothelial cells.33 Taken together, the results that high glucose treatment induced slight but significant increase of angiotensin II concentrations and thalidomide treatment significantly reduced high glucose-induced p22^phox protein levels in HK2 cells, high glucose may induce angiotensin II, which stimulates TNF-α through AT1 receptor, and the increased TNF-α induces NAD(P)H oxidase, which causes the renal tubular cells damage. Further studies using other kidney injury markers such as KIM-1 or NGAL are required to clarify the mechanisms on interaction between TNF-α and RAS in high glucose-induced renal tubular damage.

In conclusion, high glucose-induced increases in NAG and p22^phox protein were accompanied by increased secretion of angiotensin II and TNF-α in HK2 cells. The increment of p22^phox protein by high glucose, angiotensin II or TNF-α was markedly attenuated by AT1 siRNA. This work shows that AT1 receptor blockade may be useful in preventing TNF-α-induced additional renal tubular damage in diabetic nephropathy.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.