Angiotensin II induces DNA damage via AT1 receptor and NADPH oxidase isoform Nox4

Abstract

Epidemiological studies revealed increased renal cancer incidences and higher cancer mortalities in hypertensive individuals. Activation of the renin–angiotensin–aldosterone system leads to the formation of reactive oxygen species (ROS). In vitro, in renal cells, and ex vivo, in the isolated perfused mouse kidney, we could show DNA-damaging potential of angiotensin II (Ang II). Here, the pathway involved in the genotoxicity of Ang II was investigated. In kidney cell lines with properties of proximal tubulus cells, an activation of NADPH oxidase and the production of ROS, resulting in the formation of DNA strand breaks and micronuclei induction, was observed. This DNA damage was mediated by the Ang II type 1 receptor (AT1R), together with the G protein 
G _α-q/11 . Subsequently, phospholipase C (PLC) was activated and intracellular calcium increased. Both calcium stores of the endoplasmic reticulum and extracellular calcium were involved in the genotoxicity of Ang II. Downstream, a role for protein kinase C (PKC) could be detected, because its inhibition hindered Ang II from damaging the cells. Although PKC was activated, no involvement of its known target, the NADPH oxidase isoform containing the Nox2 subunit, could be found, as tested by small-interfering RNA down-regulation. Responsible for the DNA-damaging activity of Ang II was the NADPH oxidase isoform containing the Nox4 subunit. In summary, in kidney cells the DNA-damaging activity of Ang II depends on an AT1R-mediated activation of NADPH oxidase via PLC, PKC and calcium signalling, with the NADPH subunit Nox4 playing a crucial role.

Introduction

Angiotensin II (Ang II) is one of the oldest known peptide hormones and a major regulator of blood pressure. Ang II and 
aldosterone are the active substances of the renin–angiotensin–aldosterone system (RAAS), which maintain the blood pressure through vasoconstriction and modulation of salt and water homeostasis. In addition to the systemic RAAS, local RAAS have been described in the kidney, the heart and the brain, which are responsible for much higher Ang II concentrations observed in these organs in comparison to the plasma concentration of Ang II (1). Increased activation of the RAAS results in hypertension, atherosclerosis, renal diseases and possibly cancer (2,3). Epidemiological studies have revealed higher cancer mortalities in hypertensive individuals (4). In most hypertensive patients the activity of renin in relation to their sodium balance is either inappropriately normal or even elevated, when it actually should be reduced (5), pointing to a stimulated RAAS.

Several studies have indicated that activation of the RAAS can lead to the formation of reactive oxygen species (ROS) (6,7). The generated ROS could cause DNA damage, which possibly contributes to the formation of malignancies. Most of the studies investigating the source of generated ROS by the RAAS typically focused on the NADPH oxidase of vascular smooth muscle cells. Similar to those of the phagocytes, this is a multi-subunit enzyme, comprised of two catalytic membrane-bound subunits, gp91phox (now known as Nox2) and p22phox, the cytosolic regulatory subunits p47phox, p67phox, p40phox as well as a small G protein, Rac1 or Rac2. Recent investigations have identified six more homologs of Nox2. These members of the Nox family have different tissue expression patterns, cellular localisation and mechanism of activation. In proximal tubules of the kidney, expression of Nox2 and Nox4 is reported (8). Nox2 is activated by phosphorylation of its regulatory subunit p47phox, which leads to conformational changes in the cytosolic subunits. p47phox, p40phox, p67phox and Rac1 then migrate from the cytosol to the membrane and associate with the heterodimer Nox2/p22phox to facilitate electron transfer from NADPH to O₂. Among the Nox family, Nox2 is known to secrete superoxide to the outside of the cell (9). Nox4 is speculated to have a unique mechanism of activation, as it only needs the p22phox subunit for its ROS-producing activity (10), and is regulated at the level of mRNA expression or mRNA translation (11). Nevertheless, recent studies suggested that the p22phox-binding protein, poldip2, can also regulate the activity of Nox4 (12) although the exact mechanism of action is still to be revealed.

Ang II mediates its effects via two receptors, Ang II type 1 receptor (AT1R) and Ang II type 2 receptor (AT2R), both of which are G protein-coupled receptors. AT1R, which mediates the major physiological effects of Ang II, is expressed in a wide variety of cells, whereas AT2R is more predominant in foetal tissues. Our group has shown that treatment of several kidney cell lines (13), as well as perfusion of isolated mouse kidneys (14), with pathophysiological concentrations of Ang II damages DNA in cells. This effect of Ang II could be prevented using an AT1R antagonist. Here, we investigated the signalling pathway that leads to the genotoxic response after Ang II treatment and identified the responsible NADPH oxidase isoform in the human kidney cell line HK-2, which shows properties of proximal tubule cells.

Materials and methods

Chemicals and reagents

Cell culture media and reagents were obtained from PAA Laboratories GmbH (Pasching, Austria) and Invitrogen Life Technologies (Darmstadt, Germany). EpiLife^® calcium-free medium was purchased from Cascade Biologics (Portland, OR, USA). Thapsigargin was obtained from Enzo Life Sciences (Lörrach, Germany). GP antagonist-2A (GP-2A) was purchased from Calbiochem (Darmstadt, Germany). VAS2870 was a gift from Vasopharm GmbH (Würzburg, Germany). If not mentioned otherwise, all other chemicals, including TEMPOL (4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl), were purchased from Sigma-Aldrich (Taufkirchen, Germany).

Oligonucleotides

Primers for polymerase chain reaction (PCR) were designed using the program Primer3 (15) and ordered from MWG Biotech (Ebersberg, Germany). Small-interfering RNA (siRNA) oligonucleotides for Nox2 and Nox4 and control siRNA were purchased from Santa Cruz Biotechnology.

Cell culture

HK-2, a human kidney cell line with properties of proximal tubule cells, was obtained from Dr. G. Garibotto, Nephrology Division, Department of Internal Medicine and Urology Division, University of Genoa, Genoa, Italy and cultured as previously described (16).

Porcine epithelial cells with characteristics of proximal tubule cells (LLC-PK1) were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) and cultured as described before (13).

Human embryonic kidney cells (HEK-293) cells were obtained from ATCC and maintained in Dulbecco’s modified Eagle’s medium (DMEM) with high glucose content supplemented with 10% foetal calf serum, 2mM of l-glutamine, 0.4% antibiotics and 2.5% HEPES.

Immortalised mouse aortic endothelial cell culture

Mouse aortic endothelial cells (MAECs) were isolated from the thoracic aortas of 4-week-old male wild-type C57BL/6 mice (Jackson Laboratories) or p47phox-deficient mice as previously described by Bond et al. (17) and Hwang et al. (18). Briefly, the thoracic aorta was cut into pieces ~1mm × 1mm in area, which were cultured intimal side down on a collagen gel droplet (20 μl of 0.175% type I rat tail collagen) in EGM2-MV Endothelial growth medium 2 for microvascular endothelial cells medium (Lonza) in a six-well plate. The explants were removed after ~3 days, and adherent cells were cultured for another day in MAEC growth medium (DMEM supplemented with 10% foetal bovine serum, 1% penicillin–streptomycin, 50 μg/ml endothelial cell (EC) growth supplement (Sigma-Aldrich), and 1× nonessential Eagle’s amino acid). The adherent cells were then immortalised by infection with polyoma middle T antigen, which specifically immortalizes ECs. Immortalised MAECs (iMAECs) were selected by subculturing for ~2 months in MAEC growth medium supplemented with the neomycin analog G418 (1mg/ml). iMAECs were further purified by cell sorting based on uptake of Dil-labelled acetylated low-density lipoprotein. iMAEC expressed EC markers (VE-cadherin, platelet EC adhesion molecule-1, endothelial nitric oxide synthase and Kruppel-like factor-2), but not the smooth muscle cell marker, smooth muscle cell α-actin (data not shown). iMAECs obtained from C57BL/6 (wild-type) or p47phox were maintained in MAEC growth medium.

All cells were maintained at 37°C in a humidified atmosphere with 5% CO₂. For substance combinations, concentrations that were described as effective in the literature and have been found to be non-cytotoxic in preliminary experiments were applied. Furthermore, all substances used were checked for their intrinsic antioxidant activity before use in actual experiments (data not shown).

Measurement of cellular ROS

To evaluate the release of ROS, the cell-permeable fluorogenic probe dihydroethidium (DHE) was used. The cells were seeded on coverslips. In the following day, the cells were treated with the test compounds and 5 µM DHE (Merck Biosciences GmbH, Schwalbach, Germany) for 30min in the dark. After washing with phosphate-buffered saline (PBS), coverslips were mounted on slides, observed and photographed with an Eclipse 55i microscope (Nikon GmbH, Düsseldorf, Germany) and a Fluoro Pro MP 5000 camera (Intas Science Imaging Instruments GmbH, Göttingen, Germany) at 200-fold magnification. The pictures were analysed using ImageJ software (http://rsbweb.nih.gov/ij, (19)).

Comet assay

After treatment, cells were harvested and the comet assay was performed with all cell lines used in this study as described earlier (20). Briefly, the cells were embedded in 0.5% low-melting-point agarose and loaded on fully frosted microscope slides, coated with a layer of high-melting-point agarose. For lysis, slides were immersed in lysing solution (1% Triton X-100, 10% dimethyl sulfoxide, and 89% lysis buffer containing 10mM Tris, pH 10; 1% Na-sarcosine; 2.5M NaCl; and 100mM Na₂EDTA) and incubated at 4°C in the dark for 1h. DNA unwinding and alkali-labile damage expression was allowed in alkaline electrophoresis buffer (300mM NaOH and 1mM Na₂EDTA, pH 13) for 20min at 4°C in the dark, followed by electrophoresis at 4°C, for 20min in a 1V/cm and 300 mA electrical field. Afterwards, the slides were neutralised for 5min in 0.4M Tris (pH 7.5), dried in methanol for 10min and stained using 20 μg/l propidium iodide (Molecular Probes, Eugene, OR, USA). A fluorescence microscope at 200-fold magnification and a computer-aided image analysis system (Komet 5; Kinetic Imaging, Bromborough, UK) were employed for analysis. The percentage of DNA in the tail was used to quantify the DNA damage.

Micronucleus test

The micronucleus test was conducted with all cell lines as described before (21). After substance treatment, medium was removed and replaced by fresh culture medium containing cytochalasin B (2 µg/ml). After a further 20h, cells were harvested, applied onto glass slides by cytospin centrifugation and fixed in methanol (−20°C) for at least 2h. Before counting, cells were stained with acridine orange (62.5 µl/ml in Sørensen buffer containing 15mM Na₂HPO₄ and 15mM KH₂PO₄, pH 6.8) and mounted for microscopy. From each of two slides, 1000 binucleated cells were evaluated for micronuclei-containing cells and the average was calculated. In addition, the cytokinesis-block proliferation index (CBPI) was determined from 1000 cells of each sample (data not shown) using the equation below:

Only in the case of treatment with the calcium chelator, 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid tetrakis(acetoxymethyl ester) (BAPTA)-AM, the CBPI dropped significantly, and therefore the data from this micronucleus test were excluded. In other cases, no significant difference was observed in CBPI.

siRNA transfection

In six well plates, 1.5 × 105 cells were seeded in antibiotic free medium 1 day before transfection with siRNA. The transfection was conducted as described by the manufacturer using siRNA Transfection Reagent (Santa Cruz Biotechnology). At 24h after transfection, the cells in each well were split into three wells. After additional 24h, total RNA from one well was extracted to confirm down-regulation of the desired target gene. The cells in the other wells were treated with either PBS or Ang II.

RNA isolation and PCR

The expression of mRNA was detected using the reverse-transcription PCR (RT-PCR). Total RNA was isolated from the cells with the RNeasy Mini Kit (Qiagen, Hilden, Germany) as described in the provider’s protocol. One microgram of RNA was used for cDNA synthesis using Verso cDNA Synthesis Kit (Thermo Scientific). The following primers were used for amplification of the respective gene fragments: β-actin forward: 5'-CTC TTC CAG CCT TCC TTC CT-3', β-actin reverse: 5'-AGC ACT GTG TTG GCG TAC AG-3' (610bp, annealing temperature: 56°C), Nox2 forward: 5'-TGC AGC CTG CCT GAA TTT CAA C-3', Nox2 reverse: 5'-GAG GCA CAG CGT GAT GAC AAC-3' (391bp, annealing temperature: 56°C), Nox4 forward: 5'-CTG GTG AAT GCC CTC AAC TT-3', Nox4 reverse: 5'-CTG GCT TAT TGC TCC GGA TA-3' (556bp, annealing temperature: 52°C). The PCR was performed using REDTaq^™ ReadyMix^™ PCR Reaction Mix (Sigma-Aldrich).

Statistics

If not mentioned otherwise, data from at least three independent experiments ± standard deviation are depicted. Statistical calculations were performed using Statistica 8 (StatSoft (Europe) GmbH, Hamburg, Germany). Statistical significance between the individual groups was tested using the Mann–Whitney U-test. For quantification of DHE staining data Student’s t-test was used. 
A P value of ≤0.05 was considered significant.

Results

Ang II induces DNA damage, which is dependent on the 
G protein-coupled receptor AT1R

Induction of DNA damage and micronucleus formation was investigated using different concentrations of Ang II (Table I). 
The results from both comet assay and micronucleus test showed a dose-dependent increase with elevating Ang II concentration in all cell lines tested.

Table I.

Effects of treatment of cells with the indicated concentrations of Ang II on % of DNA in tail in the comet assay or frequency of cells with micronucleus

Ang II (nM)	Number of repeats	0	25	50	100	200	400
Comet assay
LLC-PK1	2	3.73±3.21 (1)	5.50±3.26 (1.47)	9.53±1.32 (2.55)	9.95±0.03 (2.66)	10.00±3.68 (2.68)	11.05±2.88 (2.96)
HEK-293	2	3.90±0.14 (1)	3.38±092 (0.87)	4.19±0.66 (1.08)	6.87±2.57 (1.76)	7.38±1.23 (1.89)	9.01±2.40 (2.31)
HK-2	1	1.22 (1)	–	–	2.92 (2.39)	4.21 (3.45)	–
iMAEC WT	4	1.68±1.35 (1)	–	–	–	3.54±2.12 (2.10)	7.50±3.54 (4.46)
iMAEC KO	4	1.47±0.39 (1)	–	–	–	3.63±0.35 (2.46)	5.75±1.28 (3.91)
Micronucleus test
LLC-PK1	2	2.92±0.02 (1)	4.95±1.06 (1.70)	5.22±1.04 (1.79)	6.27±1.25 (2.15)	9.53±2.89 (3.26)	11.77±1.89 (4.03)
HEK-293	1	10.94 (1)	8.41 (0.77)	10.49 (0.96)	13.47 (1.23)	19.99 (1.83)	31.67 (2.89)
HK-2	1	15.37 (1)	21.49 (1.40)	23.44 (1.53)	40 (2.60)	40.46 (2.63)	44.29 (2.88)

Where applicable, the mean of the indicated number of repeat experiments ± SD is reported. Fold increase over the untreated sample is shown in parantheses.

The expression of AT1R in HK-2 cells was confirmed with RT-PCR (data not shown). The obtained sequence was compared with the Genbank database, yielding 99% identity to the database sequence of Homo sapiens AT1R (accession number S77410.1). To verify that the genotoxic effects of Ang II are mediated by the AT1R, AT1R was blocked with the specific antagonist candesartan. Candesartan completely prevented the Ang II-induced DNA damage in HK-2 cells (Figure 1A and B).

Fig. 1.

DNA damage measured in the (A) comet assay and the (B) micronucleus test after treatment of HK-2 cells with 200nM Ang II with or without co-incubation with the indicated concentrations of the AT1R antagonist candesartan (Cand) for 4h, (C and D) the Gα-q inhibitor GP-2A and (E and F) the phospholipase C inhibitor U-73122. The same concentration of an inactive analog of the phospholipase C inhibitor, U-73343, was also investigated to confirm the specificity of the inhibitor. MN, micronuclei; BNC, binucleated cells. *P ≤ 0.05 vs. Ang II-only treated sample (Mann–Whitney U-test).

Also competitive inhibition of the G protein coupled to AT1R, G_α-q/11, with the specific inhibitor GP-2A resulted in a decrease of Ang II-induced DNA damage to the control level (Figure 1C and D).

Inhibition of PLC protects the cells from Ang II-induced DNA damage

G_α-q/11 is known to ultimately activate phospholipase C (PLC). Involvement of PLC in Ang II-induced DNA damage was verified using the inhibitor U-73122, which protected the cells against DNA breaks and micronuclei formation (Figure 1E and F), while U-73343, the inactive analog of the inhibitor, failed to show a similar effect at the tested concentrations.

Protein kinase C and calcium are involved in Ang II-induced DNA damage

The products of PLC-action on membrane lipids, inositol-1,4,5-triphosphate (IP3) and diacylglycerol (DAG) are activators of protein kinase C (PKC). Induction of PLC therefore often leads to PKC activation. Indeed, inhibition of PKC with the specific inhibitor, sphingosine, decreased the Ang II-induced DNA damage (Figure 2A and B). Because IP3 is also involved in the release of calcium, the involvement of calcium in the genotoxicity of Ang II was investigated. The induced DNA damage by Ang II could be prevented by the calcium chelator BAPTA-AM (Figure 2C).

Fig. 2.

DNA damage measured in the comet assay and the micronucleus test after treatment of HK-2 cells for 4h with 200nM Ang II with or without co-incubation with the indicated concentrations of the (A and B) PKC inhibitor sphingosine or (C and D) different compounds to intervene calcium signalling. 
In the case of thapsigargin a 30-min pre-incubation was performed. BAPTA hindered the cell division and therefore was excluded from the micronucleus test. BAPTA, 1,2-Bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetrakis(acetoxymethyl ester); CFM, calcium-free medium; TPG, thapsigargin; TMB8, 8-(N,N-diethylamino)octyl 3,4,5-trimethoxybenzoate; 2-APB, 2-aminoethoxydi-phenyl borate; MN, micronuclei; BNC, binucleated cells. *P ≤ 0.05 vs. 
Ang II-only treated sample (Mann–Whitney U-test).

To distinguish between the sources of calcium necessary for the observed DNA damage after Ang II treatment, a variety of compounds were used. After depletion of intracellular calcium stores with thapsigargin, Ang II failed to induce DNA damage (Figure 2C and D). The IP3 receptor (IP3R)-associated calcium channels on the endoplasmic reticulum were blocked with 2-aminoethoxydi-phenyl borate (2-APB) and an IP3R antagonist (TMB8). Both of these compounds successfully reduced the damage by Ang II (Figure 2C and D), confirming the role of endoplasmic calcium stores in the genotoxicity of Ang II. To investigate the potential role of extracellular calcium, cells were incubated in calcium-free medium, where again Ang II failed to induce damage (Figure 2C and D). In summary, these results point to a role of extracellular calcium influx as well as of intracellular calcium stores.

Genotoxicity of Ang II depends on ROS formation by NADPH oxidase

PKC and/or calcium can activate some NADPH oxidase isoforms, which can cause oxidative stress. To reveal the production of oxidants after Ang II treatment, LLC-PK1 cells were stained with the ROS-sensitive dye DHE, which after reaction with ROS forms at least two fluorescent derivatives, 2-hydroxyethidium, specific for its reaction with superoxide, and ethidium (22). Ang II-treated cells showed a higher production of ROS in comparison to the control sample (Figure 3A and B). This effect could be partially prevented using the radical scavenger TEMPOL (4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl), which is a mimetic of superoxide dismutase (SOD). 
A reduction of ROS formation was also observed after treatment with Ang II together with the calcium chelator BAPTA-AM; the PKC inhibitor sphingosine; and the NADPH oxidase inhibitor VAS2870. TEMPOL, which can cross the cell membrane, was also able to protect the cells against genotoxic effects of Ang II (Figure 3C and D). Cell-impermeable SOD caused no reduction of DNA damage (Figure 3C and D). This led to the conclusion that the radicals responsible for Ang II-induced DNA damage are formed and act inside the cells.

Fig. 3.

(A) Induction of ROS formation visualised with the help of the dye dihydroethidium in LLC-PK1 cells. Cells were treated with (I) PBS, (II) 400nM Ang II, (III) 400nM Ang II + 100 µM of the calcium chelator BAPTA-AM, (IV) 400nM Ang II + 10 µM of the PKC inhibitor sphingosine, (V) 400nM Ang II + 
10 µM of the NADPH oxidase inhibitor VAS2870 (VAS), (VI) 400nM Ang II + 50 µM of the radical scavenger TEMPOL. (B) Quantification of the induced ROS by measuring the mean grey value using ImageJ normalised to the control. *p ≤ 0.05 vs. Ang II-only treated sample (Student’s t-test). Results from the (C) comet assay and (D) micronucleus formation after incubation of HEK-293 cells with 200nM Ang II and either 300U of the superoxide scavenger enzyme superoxide dismutase (SOD) or 50 µM of the SOD mimetic TEMPOL for 4h. MN, micronuclei; BNC, binucleated cells. *P ≤ 0.05 vs. Ang II-only treated sample (Mann–Whitney U-test).

As a probable source of ROS formation, NADPH oxidase was inhibited using diphenyleneiodonium chloride (DPI), an inhibitor of flavoprotein enzymes such as NADPH oxidase and the specific NADPH oxidase inhibitor, VAS2870. As depicted in Figure 4A and B, DPI and also VAS2870 were able to reduce DNA damage in the cells, leading to the conclusion that this enzyme is responsible for the observed genotoxicity induced by Ang II.

Fig. 4.

DNA damage measured in the (A) comet assay and the (B) micronucleus test after 4h treatment of HK-2 cells with 200nM Ang II with or without co-incubation with the indicated concentrations of the flavoprotein enzyme inhibitor diphenyleneiodonium chloride (DPI), and with the NADPH oxidase inhibitor VAS2870 (VAS). *p ≤ 0.05 vs. Ang II-only treated sample (Mann–Whitney U-test). DNA damage measured by the comet assay and the micronucleus test after 200nM Ang II treatment for 4h in HK-2 cells transfected with (C and D) Nox2 siRNA or (E and F) Nox4 siRNA. MN, micronuclei; BNC, binucleated cells. The success of siRNA transfection was confirmed in RT-PCR. *P ≤ 0.05 vs. Transfection buffer treated sample (Mann–Whitney U-test).

Using siRNAs, the expression of the Nox2 and Nox4 isoforms was inhibited. The down-regulation of Nox2 and Nox4 was confirmed by RT-PCR. The cells were then treated either with Ang II or with PBS as the negative control. The data represented in Figure 4C and D show that the cells treated with Nox2 siRNA did not show any difference in their DNA damage response compared to the control cells. However, Nox4 siRNA down-regulation prevented DNA damage formation induced by Ang II (Figure 4E and F).

Moreover, the wild-type and p47phox-deficient iMAEC cells were treated with Ang II to further investigate whether p47phox-dependent Nox2 plays a role in Ang II-induced formation of DNA damage. As demonstrated in Figure 5, Ang II induced DNA strand breaks in both, wild-type and p47phox-deficient cells.

Fig. 5.

DNA damage measurement after 4 hours treatment of wild-type and p47phox-deficient immortalised mouse aorta epithelial cells (iMAEC) with 200nM Ang II. *P ≤ 0.05 vs. negative control (Mann–Whitney U-test).

Discussion

Accumulating evidence indicates a role of ROS in the pathology of cardiovascular (23) and renal diseases (24). DNA damage is a consequence of increased ROS production. One potential cause of the observed disturbed oxidative balance can be hypertension and activation of RAAS. It was demonstrated that in patients with mild renal insufficiencies, the angiotensin-converting enzyme inhibitor, trandolapril, reduced all-cause mortality (25). On the other hand, it has been reported that lymphocytes of hypertensive patients showed higher genomic damage than those of control individuals (26). A study in stroke-prone spontaneously hypertensive rats revealed that the amount of urinary 8-oxo-2'-deoxyguanosine (8-oxodG), an oxidatively modified DNA base and a common marker of oxidative stress, was not different from normotensive Wistar-Kyoto rats at 6 weeks of age, but became higher than the controls after the development of severe hypertension (27). Another study showed a significant increase in the total chromosomal damage and the percentage of abnormal metaphases between spontaneously hypertensive and normal Wistar-Kyoto rats (28). The RAAS is an important part of the physiological and pathological response of the cardiovascular and renal system. Ang II, which is one of the two biologically active hormones of this system, is not only responsible for vasoconstriction and blood pressure regulation but is also implicated in inflammation, endothelial dysfunction, atherosclerosis, hypertension and congestive heart failure (29). Some of these effects of Ang II are mediated via production of ROS generated by membrane-bound NADPH oxidase localised in the vascular walls (30).

We have shown that DNA damage was induced by Ang II in cell lines derived from kidney (13), as well as in perfused isolated mouse kidneys (14). The damage included single- and double-strand breaks, chromosomal aberrations, abasic sites and 8-oxodG base modifications. These effects of Ang II were mediated by the AT1R.

Here, in cell lines with properties of proximal tubules, we showed that Ang II treatment leads to intracellular ROS production, which is in line with previous reports in cardiomyocytes and pericytes (31,32). By using two inhibitors of NADPH oxidase, we showed that this enzyme is the major source of ROS, which are responsible for DNA damage caused by Ang II treatment. VAS2870, which is used in this study as a specific NADPH oxidase inhibitor, was identified as a small molecular inhibitor of Nox2 by means of high throughput screening (33), and a number of reports verified the efficient inhibition of ROS production in several cell types (34–37). A recent report describes that cysteine modification may have a causal role in the inhibition of ROS formation by VAS2870 (38). Since all Nox isoforms contain cysteines putatively involved in enzyme function, the inhibitor might not exhibit an isoform specificity. Nox2 and Nox4 are expressed in the proximal tubular cells with Nox4 being the most abundant one (39). Nox4 is located in the perinuclear and endoplasmic reticulum membranes as well as in the plasma membrane (40). The localisation of Nox4 in the vicinity of genetic material makes it a probable candidate as the Nox isoform responsible for the DNA damage. There is still a debate about the exact ROS generated by Nox4. We observed a reduced Ang II-induced genotoxicity with TEMPOL, which is usually regarded as superoxide scavenger. In line with this, some other groups have also reported reduction of Nox4-mediated effects by TEMPOL (41,42). Although often hydrogen peroxide is detected after Nox4 stimulation, this might not be the first or not the only ROS formed by Nox4, since the localisation of Nox4 on intracellular membranes results in the release of superoxide into the lumen of the organelles where it rapidly dismutates into hydrogen peroxide (39). Using siRNAs, only inhibition of Nox4, but not of Nox2, abolished the DNA-damaging effect of Ang II. The results from our siRNA transfection and the involvement of Nox4 explain the inability of SOD to prevent Ang II-induced DNA damage. Furthermore, our results from the p47phox-deficient cell line showed that Ang II-induced induction of DNA damage was independent of p47phox, which is believed to be necessary for activation of Nox2.

Although we can conclude from our results that Nox4 is most probably the Nox isoform which is responsible for the observed DNA damage after Ang II treatment, a Nox2 or Nox4 
knock-out cell line or animal model would provide the ultimate proof for the involvement of a specific Nox isoform. There is evidence of a pro-oxidative action of Nox4 in several animal models. Recently, using Nox4 knock-out mice, it has been shown that Nox4-mediated oxidative stress led to neuronal damage and apoptosis. In the same study, a similar effect was not observed in mice with deficiency in the Nox1 or the Nox2 gene (43). In another study, it was demonstrated that the expression of Nox4 in the kidney of diabetic rats was higher than in the control animals and was accompanied by increased formation of ROS and 8-oxodG base modification (44). It is reported that in TGR(mREN2)27 rats, which show higher Ang II concentration in several organs including the kidney, the expression of Nox4 is increased in kidney and aorta (45).

Inhibition of PKC decreased Ang II-induced ROS formation and DNA damage, although Nox4 was involved but not Nox2. Nevertheless, our findings do not provide a proof about the direct effect of PKC on Nox4 upregulation or activation. Possibly PKC affects some not-yet-known molecules that induce the expression/activation of Nox4. This already is speculated for poldip2, a protein that seems to contribute to the activation of Nox4 having a serine/threonine site suitable for phosphorylation (12), which probably allows it to be controlled by PKC.

PKC itself can be activated by products of PLC. Activation of PLC results in the production of IP3 and DAG. IP3 binds to its receptor at the ER, opening a channel that allows calcium efflux into the cytoplasm. Calcium and DAG are necessary for activation of the classical PKC.

We observed a role for both intra- as well as extracellular calcium in Ang II-induced genotoxicity. Both calcium sources were shown to be linked by the signalling molecule STIM1, which acts to amplify and augment calcium signals activated by Ang II. Following the depletion of calcium stores, STIM1 senses calcium decrease through its EF-hand calcium-binding motif and migrates to the vicinity of the plasma membrane to trigger the activation of store-operated calcium channels in the membrane (46).

PLC is activated by AT1R stimulation. AT1R is a GPCR, which has been shown to be capable of coupling to the α-subunit G_q/11 (47). In line with previous findings (13), the AT1R antagonist candesartan reduced the genotoxicity of Ang II also in HK-2 cells, which confirms that the genotoxicity of Ang II is mediated via AT1R signalling. Supporting results came from inhibition of G_α-q/11. Inhibition of G_α-q/11 signalling abrogated the damage caused by Ang II.

In conclusion, we showed here that Ang II-induced genomic damage is associated with the production of ROS mediated by AT1R, G proteins, PLC, PKC and NADPH oxidase and utilizing intra- as well as extracellular calcium signalling. Based on our results, we propose the following model of signalling in Ang II-induced DNA damage (Figure 6): binding of Ang II to the AT1R activates PLC via stimulation of G proteins, resulting in the activation of PKC in a calcium-dependent manner, which in turn possibly activates NADPH oxidase. NADPH oxidase with the involvement of its Nox4 subunit then produces ROS, which cause DNA damage.

Fig. 6.

Proposed model of signalling in Ang II-induced DNA damage. Binding of Ang II to the AT1 receptor activates PLC via stimulation of G proteins, resulting in activation of PKC in a calcium-dependent manner which in turn, activates NADPH oxidase. NADPH oxidase with involvement of its Nox4 subunit then produces reactive oxygen species, which cause DNA damage. Ang II, angiotensin II; AT1R, angiotensin type 1 receptor; Gα-q/11, G protein subunits Gα-q/11; PLC, phospholipase C; IP3, inositol triphosphate; ER, endoplasmic reticulum; PKC, protein kinase C; ROS, reactive oxygen species.

Due to the presence of a functional local RAAS, the concentration of Ang II in the kidney can be up to 1000 times higher than in the plasma (48,49). This concentration can reach 800nM in experimental hypertension (48). This range of Ang II concentration can induce DNA strand breaks, which are independent of the hemodynamic effect of this compound as we have shown before (14). DNA strand breaks and chromosomal damage may lead to cell death or genetic alterations, the latter of which is considered a risk factor for carcinogenesis. The mechanism of Ang II-induced DNA damage, which we have described here, can serve as a starting point to understand the higher risk of renal cancer incidence in hypertensive patients. It still needs to be investigated if Ang II-dependent hypertension increases tumour development in animal models.