N-acetylcysteine as a potential strategy to attenuate the oxidative stressinduced by uremic serum in the vascular system

Silvia D. Rodrigues a, Karime C. França a, Fernando T. Dallin a, Clarice K. Fujihara b, Aguinaldo J. Nascimento c,Roberto Pecoits-Filho d, Lia S. Nakao a,⁎a Departamento de Patologia Básica, Universidade Federal do Paraná, Centro Politécnico, Curitiba 81531-980, Brazilb Laboratório de Fisiopatologia Renal, Faculdade de Medicina, Universidade de São Paulo, São Paulo 01246-903, Brazilc Programa de Pós-Graduação em Ciências Farmacêuticas,Universidade Federal do Paraná, Curitiba 80210-170, Brazild School of Medicine, Pontifícia Universidade Católica do Paraná, Rua Imaculada Conceição 1155, Curitiba 80215-901, Brazil

a b s t r a c ta r t i c l e i n f o

Article history:Received 24 June 2014Accepted 20 November 2014Available online 10 December 2014

Keywords:N-acetylcysteineOxidative stressSuperoxideUremiaVascular cells

Aims: Chronic kidney disease (CKD) progression is accompanied by systemic oxidative stress, which contributesto an increase in the risk of cardiovascular diseases (CVDs). N-acetylcysteine (NAC) is among the most studiedantioxidants, but its therapeutic benefits in CKD-associated CVDs remain controversial. Here, we investigatedwhether NAC could inhibit the oxidative stress induced by uremia in vitro and in vivo.Main methods: Endothelial and smooth muscle cells were challenged with human uremic or non-uremic sera,and the effects of a pre-treatment with 2 mM NAC were evaluated. Reactive oxygen species (ROS) production,protein oxidation and total glutathione/glutathione disulfide (tGSH/GSSG) ratios were measured. Five-sixthsnephrectomized or sham-operated rats were orally treated (in the drinking water) with 60 mg/kg/day NAC ornot treated for 53 days. Plasma cysteine/cystine reduction potential Eh(Cyss/2Cys) was determined as a novelmark-er of the systemic oxidative stress.Key findings: NAC inhibited all the determined oxidative stress parameters, likely by increasing the tGSH/GSSGratio, in both cell lines exposed to uremic serum. Orally administered NAC attenuated the systemic oxidativestress in uremic rats.Significance: The present results indicate that NAC, by preventing GSH depletion in vascular cells exposed to ure-mic serum and by attenuating the systemic oxidative stress during CKDprogression, emerges as a potential strat-egy to prevent the oxidative stress induced by uremic toxicity in the vascular system.

© 2014 Published by Elsevier Inc.

Introduction

Chronic kidney disease (CKD) represents a relevant risk factor forthe development of cardiovascular diseases (CVDs) [1]. CKD progres-sion is associated with systemic chronic inflammation [2] and oxidativestress [3–5], which, in combination, aremajor contributors to the accel-eration of CVDs [2,6]. Indeed, the severity of CKD is associated with in-creased levels of several oxidative stress biomarkers in the blood [6–11].

The participation of uremic toxins, such as advanced glycation endproducts and indoxyl sulfate, in mechanisms associated with vascularinjury has been demonstrated [12]. Smooth muscle cells are sensitiveto the toxic effects of uremic serum or toxins, which increase the prolif-eration rate [13] and induce cellular senescence [14] and the expressionof calcification proteins [15,16], leading to arterial degeneration [17,18].Endothelial cells cultivated in the presence of uremic serum or uremic

toxins exhibit increased proliferation [19] and apoptosis [20] rates, in-creased expression of pro-inflammatory genes [21] and inflammatoryproteins [22] and increased reactive oxygen species (ROS) production[23,24]. The inhibition by antioxidants of several of these processes[16,18,23,25] demonstrates the participation of ROS in such eventsand, most importantly, supports the idea that uremia triggers excessiveROS production.

We have shown that systemic oxidative stress, as determined eitherby plasma total thiol levels and total protein carbonylation levels [26] orby the cysteine/cystine reduction potential, Eh(Cyss/2Cys) [27], increaseswith CKD progression and decreases after kidney transplant [26], indi-cating that oxidative stress results from uremia. Indeed, the reductionof serum concentrations of indole-derived toxins with the oral adsor-bent AST-120 has been shown to retard CKD progression [28], partiallythrough the inhibition of oxidative stress [29]. In this context, antioxi-dant therapy should be considered as a putative strategy to attenuatethe oxidative stress associated with CKD progression.

Several studies and clinical trials have been performed to assess theeffects of antioxidants in CKD patients. A recent systematic review [30]of these trials indicated that orally ingested N-acetylcysteine (NAC) has

Life Sciences 121 (2015) 110–116

⁎ Corresponding author at: Departamento de Patologia Básica, Room 153, Setor deCiências Biológicas, Centro Politécnico, Universidade Federal do Paraná, Curitiba, 81531-980, Brazil. Tel.: +55 41 3361 1760; fax: +55 41 3266 2042.

E-mail address: lia.nakao@ufpr.br (L.S. Nakao).

http://dx.doi.org/10.1016/j.lfs.2014.11.0240024-3205/© 2014 Published by Elsevier Inc.

Contents lists available at ScienceDirect

Life Sciences

j ourna l homepage: www.e lsev ie r .com/ locate / l i fesc ie

promising effects, decreasing the levels of oxidative stress markers andcardiovascular events [31]. In the 5/6 nephrectomy model, NAC alsodemonstrated renoprotective effects by attenuating proteinuria andrenal tissue damage and improving the glomerular filtration rate [32].Orally ingested NAC is rapidly absorbed by the gastrointestinal tract be-cause of its neutral charge at the acidic pH of the stomach [33,34]. Mostof the NAC undergoes first pass metabolism, intracellular deacetylationand promotes an increased hepatic content of cysteine [33], which actsas an intracellular GSH precursor, at least when GSH demand is in-creased [35]. However, it was also described that after a single dose, aminor fraction of free NAC reaches the systemic circulation and reducesplasma cystine in healthy volunteers, thereby increasing the concentra-tions of plasma cysteine in the first hour [36].

In this study, we show that NAC can be considered as a potentialstrategy to attenuate the oxidative stress induced by uremia in the vas-cular system.We demonstrate that NAC pre-treatment increases the re-sistance of smooth muscle and endothelial cells to uremic toxicity andthat NAC administration inhibits the systemic oxidative stress, as deter-mined by the plasma Eh(Cyss/2Cys), which is a novel biomarker of oxida-tive stress, in the 5/6 nephrectomy model of CKD in rats.

Material and methods

Human uremic and non-uremic serum

Uremic sera were collected at the hemodialysis (HD) unit of theCajuru University Hospital. Exclusion criteria included an age of b18years, history of diabetes mellitus, active infection, autoimmune dis-eases and malignancy and were verified by detailed analysis of medicalrecords and patient interview. The pool was prepared from the plasmaof 29 fasted patients (21males and 8 females) with end-stage renal dis-ease onmaintenance HD for amean period of 30months. Themean agewas 49 ± 15 years. Causes of renal failure included the following:chronic glomerulonephritis (44.8%), nephrosclerosis (24.1%), polycystickidney disease (6.9%), obstructive pyelonephritis (6.9%), obstructiveuropathy (6.9%), congenital stenosis (3.4%), transplant rejection (3.4%)and hypertension (3.4%). A total of 24 mL of peripheral blood was col-lected from each patient at three times, before the first HD session ofthe week. Non-uremic (control) sera were collected in a private labora-tory. Exclusion criteria were identical to those used for the patients, inaddition to the history of kidney disease. The pool was prepared fromthe plasma of 30 fasted subjects (22 males and 8 females) with amean age of 44 ± 11 years. Informed consent to blood collection andserum utilization was obtained from all participants. Uremic or non-uremic sera were pooled and heat inactivated (56 °C, 30 min), as rou-tinely made with fetal calf serum. This protocol was approved by theEthics Committee of the UFPR, protocol number 02/08.

Sera pool characterization

The uremic and non-uremic sera pools were analyzed at ScribnerLaboratory (Curitiba, Brazil) to characterize their uremic, inflammatoryand redox state (Table 1). Increased levels of urea, creatinine, uric acidand phosphorus were observed in the uremic pool, reflecting renal fail-ure in the HD patients. The slightly increased level of glucose in the ure-mic pool is due to a nonfasting condition of some patients attending theHD sessions. Levels of calcium, albumin and cholesterol are compatibleto what would be expected from a clinically stable CKD-5 patient treat-ed with HD. The increased C-reactive protein (CRP) levels are in agree-ment with an inflammatory condition in the uremic pool. Decreasedtotal thiols and increased carbonylated proteins, which were de-termined as previously described [27], in the uremic pool indicate anoxidative status associated with uremia. Collectively, these results con-firm the uremic status and indicate an oxidative milieu in the blood ofthe HD patients.

Cell culture

Rabbit aortic smooth muscle (RASM) and endothelial (RAEC) cellswere kindly provided by Dr. Helena B. Nader (UNIFESP, São Paulo).The cellswere routinely cultured in an F12medium (Cultilab, Campinas,Brazil) containing penicillin and streptomycin (Life Technologies) and10% fetal calf serum (Cultilab).

Cell treatment with human sera pools

Both RASM and RAEC were plated (seeding confluence of approxi-mately 3 × 104 cells/cm2) and incubated over the day until complete ad-hesion. Subsequently, the cells were pre-treated or not treated with2 mMNAC (Sigma-Aldrich, São Paulo, Brazil) in a fresh complete medi-um for 24 h. This concentration is the approximate NAC concentrationreported in human serum following an oral administration of 400–600 mg NAC [37]. To perform a smooth transition between the fetalcalf serum and the human serum, the concentration of the fetal calfserum was decreased to 0.5% for 16 h, after which cells were treatedwith an F12 medium containing 10% human uremic or non-uremicsera pools for 3 or 24 h. Under this condition, RASM and RAEC were vi-able for at least 24 h in the presence of human sera, without any visiblemorphological alteration.

Intracellular ROS detection with fluorescent probes

Cells were exposed to human sera pools in 96-well microplates aspreviously described. The 2′, 7′-dichlorofluorescin diacetate (DCFH2-DA) assay was verified by treating both cell lines with 0.5 mM H2O2

(Merck, Darmstadt, Germany) for 4 h in the complete culture medium.In the final 30 min of the incubation time, 20 μMDCFH2-DA (Sigma-Al-drich) was added to the incubation medium. Intracellular DCF fluores-cence was determined using a microplate reader (Tecan InfiniteM200) set at λexc = 485 nm and λem = 530 nm [38]. Thedihydroethidium (DHE) assay was verified by treating cells with30 mM glucose [39] for 4 h in a routine medium. In the final 30 min ofthe incubation time, 10 μMDHE (Sigma-Aldrich)was added to the incu-bation medium. After washing with PBS, intracellular 2-OH-Et+ (whichis the product of oxidation by superoxide) fluorescence was specificallydetected at λexc = 396 nm and λem = 580 nm [40]. In some assays,20 μM apocynin was added to the cell culture for 15 min at the end ofthe treatment. After this period, cells were washed with PBS and incu-bated with DHE. Mitochondrial superoxide was measured usingMitoSOX Red® (Invitrogen, Carlsbad, CA). This assay was verified bytreating cells with 1 μM rotenone (Sigma-Aldrich) [41] for 4 h in a rou-tine medium. In the final 15 min of the incubation time, 1 μMMitoSOXRed® [39] was added to the incubation medium. After washing withPBS, intracellular MitoOH-Et+ fluorescence was measured at λexc =396 nm and λem = 580 nm [40].

Table 1Biochemical characterization of the non-uremic and uremic sera pools.

Parameters Non-uremic pool(n = 30)

Uremic pool(n = 29)

Urea (mg/dL) 30.1 151.4Creatinine (mg/dL) 0.77 8.83Uric acid (mg/dL) 5.80 7.40Glucose (mg/dL) 83 115Calcium (mg/dL) 11 10Phosphorus (mg/dL) 4.2 6.5Albumin (mg/dL) 4.7 5.1Total cholesterol (mg/dL) 215 164CRP (mg/L) 0.356 1.105Total thiols (μM) 662 268Carbonylated proteins (nmol/mg albumin) 0.77 0.97

111S.D. Rodrigues et al. / Life Sciences 121 (2015) 110–116

Protein carbonylation

Cells were exposed to human sera pools as previously described, in60-mm dishes. Protein carbonyl levels were determined using the 2,4-dinitrophenylhydrazine (DNPH) reaction, as described [40].

Determination of GSSG and total GSH (tGSH)

Cells were exposed to human sera pools in 60-mmdishes, and GSSGand tGSH levels were determined according to previously describedprotocols [42,43].

Animals

MaleWistar rats, weighing between 270–300 g, were obtained fromthe facility at the Universidade Federal do Paraná (UFPR). Five-sixthsnephrectomy or sham-operation were performed as previously de-scribed [44]. At the end of the surgery, 0.2 mL Buscopan®was adminis-tered to each rat by gavage. Seven days after the surgery, each groupwas randomized into two subgroups: those receiving NAC (Lidifarma,Curitiba, Brazil) in the drinking water ad libitum (600 mg/L solution[32] prepared every 2 days in light-protected flasks that provided amean NAC ingestion of approximately 60 mg/kg/day) and those re-ceiving only tap water. Each subgroup (sham; sham + NAC; CKD;CKD + NAC) consisted of 6 animals. The treatment lasted until day 60post-operation. On days 7 and 60, blood was collected from the tailvein, and on day 60, the animalswere sacrificedwith a lethal dose of an-esthetics (ketamine and xylazine, 100mg/kg each, ip). This protocol hasbeen approved by the Research Ethics Committee (CEUA) of the Biolog-ical Sciences Building, protocol number 343/2008.

Plasma isolation

Plasma was isolated from the blood that was collected in heparin/diethylenetriaminepentaacetic acid (DTPA) solution [27], aliquotedand immediately frozen at−80 °C until analysis.

Determination of plasma urea

Plasma urea concentrations were determined on the microplatescale using a commercial kit (Katal, São Paulo, Brazil). NAC interferencein the urea determination was excluded by performing these assays inthe presence of 0–5 mM NAC. No altered value was observed.

Determination of the plasma Eh(Cyss/2Cys)

The plasma reduction potential was determined from plasma cyste-ine and cystine concentrations, the major extracellular redox couple,employing the Nernst equation [27].

Statistical analysis

The results are expressed as themean±SD of at least 3 independentin vitro experiments ormean±SE for the in vivo experiments. Statisticalanalysis was performed with factorial ANOVA followed by Tukey test,unless otherwise specified in the figure legend, using Statistica 10.0(Statsoft) software. P b 0.05 was considered statistically significant.

Results

NAC decreases the intracellular ROS concentration

DCFH2-DA oxidation to DCF was verified by treating both RASM andRAEC with 0.5 mM hydrogen peroxide. This oxidant promoted a DCFfluorescence increase of 145 and 45%, respectively, compared with therespective non-treated cells (Fig. 1). Exposure of RASM and RAEC to

uremic serum for 3 h did not significantly alter the amount of intracel-lular ROS (Fig. 1). However, after 24 h, DCF fluorescencewas significant-ly higher for RASM (32%) and RAEC (68%) incubatedwith uremic serumcompared with non-uremic serum. Pre-incubation of cells with 2 mMNAC led to an inhibition of DCF fluorescence intensity for both uremicand non-uremic sera particularly after 24-h exposure, when uremicserum induced-ROS production was higher. Therefore, these results in-dicate that pre-treatment with NAC inhibits uremia-induced ROS pro-duction, particularly after longer exposure times.

NAC decreases the intracellular superoxide concentration

Superoxide has been shown to be produced by NADPH oxidase invascular cells (see [45] for a review) exposed to the uremic milieu[23]. Therefore, we used DHE to detect intracellular superoxide. As apositive control, both RASMandRAECwere treatedwith 30mMglucosefor 4 h, a condition that has been previously described to trigger super-oxide production in endothelial cells [39]. Hyperglycemia led to an in-crease of 30% superoxide formation in both cell types, as determinedby the 2-OH-Et+ fluorescence (Fig. 2A and B). An increased intracellular2-OH-Et+ fluorescence was observed after 3- and 24-h incubation withuremic serum for both RASM and RAEC, compared with treatment withnon-uremic serum for an identical time (Fig. 2A and B). NAC pre-treatment decreased the2-OH-Et+fluorescence intensity in both vascu-lar cells treatedwith both non-uremic and uremic sera after 3- and 24-hexposure, being statistically significant in RASM cells exposed to uremicserum for 3 and 24 h. To confirm the involvement of NADPH oxidase,cells were incubated with 20 μM apocynin, an inhibitor of the NADPHoxidase. Apocynin treatment abolished DHE oxidation in both RASMand RAEC treatedwith both uremic and non-uremic sera at all analyzedtime points, although statistical significancewas obtained only for RAECat 24-h incubation (Fig. 2A and B). These results confirm that NADPHoxidase is the major source of superoxide in vascular cells in a CKDcondition.

To determinewhethermitochondrial dysfunction contributed to su-peroxide production, a similar experiment was performed usingMitoSOX Red®, which is also oxidized by superoxide to 2-OH-Et+. Theuse of this probe was validated by incubating cells with rotenone(1 μM for 4 h), an inducer of mitochondrial superoxide production. Ro-tenone led to significant increases in 2-OH-Et+ fluorescence of 53 and39% in RASM and RAEC, respectively, compared with the negative con-trol (Fig. 2C and D). The results obtained indicated that 3-h exposure touremic serum did not significantly alter mitochondrial superoxide pro-duction, neither in RASM nor in RAEC. After 24 h, both RASM and RAECexhibited increased 2-OH-Et+ fluorescence when treated with uremicserum and compared with non-uremic serum (26 and 11%, respec-tively), but only in RASM was this difference significant (Fig. 2C andD). NAC pre-treatment significantly inhibited this increase after 24-hexposure to the uremic serum in RASM cells. This result indicates thatmitochondrial superoxide does not appear to contribute importantlyto the oxidative stress in vascular cells exposed to uremic serum,particularly after short exposure times. However, the increased su-peroxide level detected in RASM cells after 24-h incubation can beinhibited by NAC pre-treatment. Collectively, these data demonstratethat NAC pre-treatment prevents the accumulation of superoxide andsuperoxide-derived ROS in both cell types after a uremic stimulus.

NAC inhibits oxidative damage in vascular cells induced by uremic serum

Increased ROS production frequently promotes the oxidation of bio-molecules. Therefore, we investigated whether enhanced levels of pro-tein carbonylation resulted from exposure to uremic serum. Incubationof RASM and RAEC with uremic serum for 3 h promoted increased pro-tein carbonylation levels of 18 and 16%, respectively, whereas incuba-tion for 24 h resulted in increases of 40 and 31%, respectively,compared with non-uremic conditions (Fig. 3A and B). NAC pre-

112 S.D. Rodrigues et al. / Life Sciences 121 (2015) 110–116

treatment prevented these increases in both cell types exposed to bothsera pools.

Glutathione is a very relevant intracellular reductant, and the intra-cellular tGSH/GSSG ratio is typically measured to estimate the intracel-lular redox status. Indeed, the incubation of RASM and RAEC with0.5 mM hydrogen peroxide for 4 h induced decreases of 40 and 80%,

respectively, in this parameter. Incubation of RASM and RAEC with ure-mic, but not non-uremic serum, resulted in a progressive decrease in thetGSH/GSSG ratio over the course of 24 h for both cell types (Fig. 3C andD). In addition, pre-incubation of NAC significantly increased tGSH/GSSG ratios in both RASM and RAEC exposed to both sera pools(Fig. 3C and D). Collectively, these results indicate that uremic toxicity

Fig. 1. ROS detection using DCFH2-DA. (A) RASM and (B) RAECwere pre-treated or not treated with 2mMNAC for 24 h in a complete medium (F12 with 10% FCS) followed by overnightincubationwithmedium containing low serum (0.5%). Subsequently, themediumwas replaced by F12 containing 10% non-uremic or 10% uremic sera pool. Cells were exposed to humansera pools for 3 or 24 h. In the last 30min, 20 μMDCFH2-DAwas added. Cells treatedwith 0.5mMhydrogen peroxide in 10% FCS for 4 hwere used as a positive control. The data representthe mean ± SD of 4 independent determinations. *P b 0.05; **P b 0.005.

Fig. 2. Intracellular superoxide detection using dihydroethidine (DHE)-based probes. Intracellular superoxide was determined using DHE (A, B), whereas mitochondrial superoxide (C,D) was determined using MitoSOX Red®. RASM (A, C) and RAEC (B, D) were pre-treated or not treated with 2 ;mM NAC for 24 h in a complete medium (F12 with 10% FCS) followedby overnight incubation with medium containing low serum (0.5%). Subsequently, the medium was replaced by F12 containing 10% non-uremic or 10% uremic sera pool. Cells were ex-posed to human sera pools for 3 or 24 h. Ten µMDHEwas added in the last 30 min (A, B) or 1 μMMitoSOX Red® in the last 15min (C, D). Alternatively, cells were exposed to non-uremicand uremic sera pools for 3 and 24 h. In the last 15 min, cells were incubated with 20 μM apocynin followed by 30 min-incubation with 10 μMDHE. Cells treated with 30 mM glucose orwith 1 μM rotenone in 10% FCS for 4 h were used as positive controls. The data represent the mean ± SD of 4 independent determinations. *P b 0.05; **P b 0.005.

113S.D. Rodrigues et al. / Life Sciences 121 (2015) 110–116

targets the GSH-based intracellular redox state, and that NAC, by mod-ulating this ratio, improves vascular resistance to the oxidative stress in-duced by uremic serum.

Chronically ingested NAC improves the plasma Eh(Cyss/2Cys) in CKD rats

The Eh(Cyss/2Cys) has been considered a novel oxidative stressparameter. Our results indicated that the plasma Eh(Cyss/2Cys) is 22%higher for CKD rats at day 60 compared with the correspondingsham control at an identical time point (Fig. 4A). In addition, theincreased Eh(Cyss/2Cys) observed at days 7 and 60 post-nephrectomyconfirms the clinical observation that oxidative stress increaseswith CKD progression [26,27]. Chronic administration of NACinhibited this increase in CKD rats but did not demonstrate an effectin sham rats (Fig. 4A). Although no statistical significance was ob-served, the plasma cysteine concentration was decreased in CKDrats at day 60, and NAC treatment corrected this concentration tolevels similar to those found in sham control rats (Fig. 4B). No changein cystine levels was observed (Fig. 4C).

In addition, our results showed that CKD rats treated with NACpresented a significantly higher (P b 0.05) body weight (401 ± 13 g)than non-treated CKD rats (322 ± 4 g) at day 60, as already described[32]. The body weight of the CKD group treated with NAC was similarto non-treated (360 ± 13 g) or NAC-treated sham (410 ± 17 g) rats

(P N0.05). Finally, five-sixths nephrectomy significantly increased ureaconcentrations in plasma by 137% at day 60 post-surgery, comparedwith the corresponding sham control. NAC treatment inhibited this ac-cumulation at day 60 and did not demonstrate an effect in sham rats(Fig. 4D).

Discussion

Our in vitro study, in addition to confirming that uremic serum triggersoxidative pathways in both smoothmuscle and endothelial cells, demon-strated that a pre-treatment of both cell typeswith 2mMNACminimizedthe oxidative damage induced by uremic serum. ROS production, primar-ily by the vascular NADPH oxidase, and the resulting oxidative damageswere inhibited by NAC pre-treatment. The results of the tGSH/GSSGratio indicated that GSHmay be the critical target of uremia-derived oxi-dants. The total GSH/GSSG ratio progressively decreased following the ex-posure of cells to a uremic environment, in contrast to results obtainedunder a non-uremic environment. NAC pre-treatment, by increasing theintracellular tGSH/GSSG ratios, increased their intracellular antioxidantpotency. Thus, our results showed that NAC improves the resistance ofvascular cells to the oxidative stress induced by uremia probably by re-storing the intracellular tGSH/GSSG ratio. This concept has been pointedby a recent paper, which highlighted that successful NAC therapy relieson its ability to replenish GSH in GSH-depleted cells [46]. Indeed, this

Fig. 3. NAC prevents oxidative damage induced by uremic serum. RASM and RAEC were pre-treated or not treated with 2 mM NAC for 24 h in a complete medium (F12 with 10% FCS)followed by overnight incubation with medium containing low serum (0.5%). Subsequently, the medium was replaced by F12 containing 10% non-uremic or 10% uremic sera pool.After 3 and 24 h, RASM (A) and RAEC (B) cells were collected and lysed and the levels of protein carbonyls were determined by DNPH reaction. After 3, 6 and 24 h, RASM (C) andRAEC (D) were collected. The amount of total GSH (tGSH) and of GSSG were photometrically determined by Ellman's reaction and used to calculate the tGSH/GSSG ratios. Cells treatedwith 0.5 mM hydrogen peroxide in 10% FCS for 4 h were used as positive controls for both methods. The data represent the mean ± SD of 4 independent determinations. In C and D,the comparison among cells treated with uremic or non-uremic sera for 3, 6 and 24 h was made using a one-way ANOVA.*P b 0.05; **P b 0.005.

114 S.D. Rodrigues et al. / Life Sciences 121 (2015) 110–116

finding is consistentwith the notion that amajor role ofNAC is to be a cys-teine precursor for GSH synthesis, as well as a disulfide bond reductant[34]. Although the effects of NAC were more intense in cells exposed touremic serum, NAC treatment also decreased the oxidative stress bio-markers in cells treated with non-uremic serum, possibly because thepre-treatment with NAC transiently increases intracellular GSH levels inboth conditions.

The exact mechanisms by which NAC exerts its antioxidant effects,however, are complex and remain elusive [34]. As far as we know, notransport system has been characterized for NAC in vascular cells.Therefore, the observed NAC effects should result from extracellularcystine reduction to cysteine followed by cysteine uptake by a transportsystem, such as the ASC system [47], leading to increased intracellularcysteine concentrations, which is known to be the rate-limiting sub-strate for GSH synthesis [48,49]. Several studies have reported anincreased intracellular GSH concentration promoted by increased intra-cellular cysteine levels [48,50,51]. In agreement, our in vivo resultsshowed that NAC restores the extracellular cysteine levels, and conse-quently the plasma Eh(Cyss/2Cys) in CKD rats. Differently of the in vitroresults, where NAC induced an increased intracellular GSH levels alsoin the non-uremic condition, in vivo NAC did not increase the extracel-lular cysteine levels in sham rats. These findings can be explained bythe regulation exerted by cysteine and cystine transport systems,which maintain a cystine/cysteine redox cycle across the plasmamembrane [52,53]. In cases in which extracellular cysteine levelslie within the normal range (as in sham rats), an acute NAC dosewill elevate the extracellular cysteine concentrations [36] by directlyreducing cystine [48,54]. However, under chronic NAC administra-tion, as in this study, an adaptive response may occur in such shamrats inducing either an enhanced uptake of excessive extracellularcysteine by the ASC system [48–50] or an increased xct-mediated up-take of cystine, produced from the oxidation of excessive cysteine. Inboth cases, the intracellular concentration of cysteine increases,

leading to enhanced export of cysteine to the extracellular milieu[55] but at a controlled rate to maintain the Eh(Cyss/2Cys) within aphysiological range. This important finding indicates that this treat-ment does not disturb the cysteine/cystine redox circuit under non-oxidative conditions. Under conditions of oxidative stress, as thattriggered by the uremia, plasma cysteine is consumed, increasingthe plasma Eh(Cyss/2Cys). In this case, chronic administration of NACpromoted the restoration of normal levels of cysteine in plasmaand of the plasma Eh(Cyss/2Cys). The fact that we have analyzed onlythe relevant redox couples in blood and in cells, that is, extracellularcysteine and cystine concentrations in vivo and intracellular gluta-thione and glutathione disulfide in vitro, is a major limitation of ourstudy. A detailed panel of the levels of all those species, both intraand extracellularly, could provide insights about their dynamicsunder basal, uremic and NAC-treated conditions, in acute and chron-ic administrations.

In addition, NAC administration prevented accumulation of urea inplasma, likely as a result of a reduced renal damage, as previouslyreported [32]. Such renoprotection could result from the improvementof the Eh(Cyss/2Cys) values, which ultimately ensure the proper function-ing of kidney cells [56], and/or result from increased intracellular anti-oxidant power provided by NAC.

Conclusion

Collectively, our results support the view that NAC, by maintain-ing the cysteine and GSH pools within a healthy concentration rangeduring CKD progression, is able to maintain the plasma Eh(Cyss/2Cys)within physiological values and to prevent the oxidative effects in-duced by uremic serum in endothelial and smooth muscle cells. Inconclusion, our studies show that NACmay be considered as a poten-tial strategy to mitigate the oxidative effects of uremia in the vas-cular system.

Fig. 4.NAC effects on the plasma Eh(Cyss/2Cys) and plasma urea concentration. Rats were 5/6 nephrectomized or sham-operated. Seven days later, bloodwas collected to isolate plasma, andNAC treatmentwas initialized (600mg/L in thedrinkingwater, ad libitum). Treatment lasted until day 60post-surgerywhen ratswere sacrificed, and bloodwas again collected. The Eh(Cyss/2Cys) (A) was calculated using the Nernst equation and the concentrations of cysteine (B) and cystine (C) were determined in plasma. Urea concentrations were also determined in thesesamples (D). The data represent the mean ± SE of 6 rats. *P b 0.05; **P b 0.005.

115S.D. Rodrigues et al. / Life Sciences 121 (2015) 110–116

Conflict of interest statement

The authors declare that there are no conflicts of interest.

Acknowledgments

This study was supported by the INCT Redoxoma (573530/2008-4),Fundação Araucária (14620 and 22122) and CNPq (473450/2007-0 and475455/2012-6). Fellowships from CAPES (SDR, KCF and FTD) and fromCNPq (RP and LSN) are also acknowledged.

References

[1] M.W. Taal, Chronic kidney disease 10 years on: what have we learned? Curr. Opin.Nephrol. Hypertens. 21 (2012) 607–611.

[2] P. Stenvinkel, New insights on inflammation in chronic kidney disease-genetic andnon-genetic factors, Nephrol. Ther. 2 (2006) 111–119.

[3] F. Locatelli, B. Canaud, K.-U. Eckardt, P. Stenvinkel, C. Wanner, C. Zoccali, Oxidativestress in end-stage renal disease: an emerging threat to patient outcome, Nephrol.Dial. Transplant. 18 (2003) 1272–1280.

[4] B. Descamps-Latscha, T. Drueke, V.Witko-Sarsat, Dialysis-induced oxidative stress: bio-logical aspects, clinical consequences, and therapy, Semin. Dial. 14 (2001) 193–199.

[5] E. Dounousi, E. Papavasiliou, A. Makedou, K. Ioannou, K.P. Katopodis, A. Tselepis,et al., Oxidative stress is progressively enhanced with advancing stages of CKD,Am. J. Kidney Dis. 48 (2006) 752–760.

[6] J. Himmelfarb, P. Stenvinkel, T.A. Ikizler, R.M. Hakim, Perspectives in renal medicine:the elephant in uremia: oxidant stress as a unifying concept of cardiovascular dis-ease in uremia, Kidney Int. 62 (2002) 1524–1538.

[7] J. Mimić-Oka, T. Simić, L. Djukanović, Z. Reljić, Z. Davicević, Alteration in plasma an-tioxidant capacity in various degrees of chronic renal failure, Clin. Nephrol. 51(1999) 233–241.

[8] E. Maggi, R. Bellazzi, F. Falaschi, A. Frattoni, G. Perani, G. Finardi, et al., Enhanced LDLoxidation in uremic patients: an additional mechanism for accelerated atherosclero-sis? Kidney Int. 45 (1994) 876–883.

[9] I. Ceballos-Picot, V. Witko-Sarsat, M. Merad-Boudia, A.T. Nguyen, M. Thévenin, M.C.Jaudon, et al., Glutathione antioxidant system as a marker of oxidative stress inchronic renal failure, Free Radic. Biol. Med. 21 (Jan 1996) 845–853.

[10] P. Rutkowski, S. Malgorzewicz, E. Slominska, M. Renke, W. Lysiak-Szydlowska, J.Swierczynski, et al., Interrelationship between uremic toxicity and oxidative stress,J. Ren. Nutr. 16 (Jul 2006) 190–193.

[11] Y. Matsuyama, H. Terawaki, T. Terada, S. Era, Albumin thiol oxidation and serumprotein carbonyl formation are progressively enhanced with advancing stages ofchronic kidney disease, Clin. Exp. Nephrol. 13 (2009) 308–315.

[12] P. Brunet, B. Gondouin, A. Duval-Sabatier, L. Dou, C. Cerini, F. Dignat-George,et al., Does uremia cause vascular dysfunction? Kidney Blood Press. Res.(2011) 284–290.

[13] G.Muteliefu, A. Enomoto, T. Niwa, Indoxyl sulfate promotes proliferation of human aor-tic smooth muscle cells by inducing oxidative stress, J. Ren. Nutr. 19 (Jan 2009) 29–32.

[14] T. Niwa, H. Shimizu, Indoxyl sulfate induces nephrovascular senescence, J. Ren. Nutr.22 (Jan 2012) 102–106.

[15] N.X. Chen, K.D. O'Neill, D. Duan, S.M. Moe, Phosphorus and uremic serum up-regulate osteopontin expression in vascular smooth muscle cells, Kidney Int. 62(2002) 1724–1731.

[16] T. Sutra, M.Morena, A.-S. Bargnoux, B. Caporiccio, B. Canaud, J.-P. Cristol, Superoxideproduction: a procalcifying cell signalling event in osteoblastic differentiation of vas-cular smooth muscle cells exposed to calcification media, Free Radic. Res. 42 (2008)789–797.

[17] G. Guilgen, M.L. Werneck, L. De Noronha, A.P.C. Martins, A.M. Varela, L.S. Nakao,et al., Increased calcification and protein nitration in arteries of chronic kidney dis-ease patients, Blood Purif. 32 (2011) 296–302.

[18] S. Yamada, M. Taniguchi, M. Tokumoto, J. Toyonaga, K. Fujisaki, T. Suehiro, et al., Theantioxidant tempol ameliorates arterial medial calcification in uremic rats: Impor-tant role of oxidative stress in the pathogenesis of vascular calcification in chronickidney disease, J. Bone Miner. Res. 27 (2012) 474–485.

[19] M. Serradell, M. Díaz-Ricart, A. Cases, J. Petriz, A. Ordinas, G. Escolar, Uraemic medi-um accelerates proliferation but does not induce apoptosis of endothelial cells inculture, Nephrol. Dial. Transplant. 18 (2003) 1079–1085.

[20] H. Cardinal, M.-A. Raymond, M.-J. Hébert, F. Madore, Uraemic plasma decreases theexpression of ABCA1, ABCG1 and cell-cycle genes in human coronary arterial endo-thelial cells, Nephrol. Dial. Transplant. 22 (Feb 1 2007) 409–416.

[21] M. Serradell, M. Díaz-Ricart, A. Cases, M.J. Zurbano, J. López-Pedret, O. Arranz, et al.,Uremic medium causes expression, redistribution and shedding of adhesion mole-cules in cultured endothelial cells, Haematologica 87 (2002) 1053–1061.

[22] A.E.M. Stinghen, S.M. Gonçalves, E.G. Martines, L.S. Nakao, M.C. Riella, C.A. Aita, et al.,Increased plasma and endothelial cell expression of chemokines and adhesion mol-ecules in chronic kidney disease, Nephron Clin. Pract. 111 (2009).

[23] L. Dou, N. Jourde-Chiche, V. Faure, C. Cerini, Y. Berland, F. Dignat-George, et al., Theuremic solute indoxyl sulfate induces oxidative stress in endothelial cells, J. Thromb.Haemost. 5 (2007) 1302–1308.

[24] J. Carracedo, P. Buendía, A. Merino, S. Soriano, E. Esquivias, A. Martín-Malo, et al.,Cellular senescence determines endothelial cell damage induced by uremia, Exp.Gerontol. 48 (2013) 766–773.

[25] F. Galli, L. Ghibelli, U. Buoncristiani, V. Bordoni, V. D'Intini, S. Benedetti, et al., Mono-nuclear leukocyte apoptosis in haemodialysis patients: the role of cell thiols and vi-tamin E, Nephrol. Dial. Transplant. 18 (2003) 1592–1600.

[26] P.R. Aveles, C.R. Criminácio, S. Gonçalves, A.T. Bignelli, L.M. Claro, S.S. Siqueira, et al.,Association between biomarkers of carbonyl stress with increased systemic inflam-matory response in different stages of chronic kidney disease and after renal trans-plantation, Nephron Clin. Pract. (2010) 116.

[27] S.D. Rodrigues, G.B. Batista, M. Ingberman, R. Pecoits-Filho, L.S. Nakao, Plasma cyste-ine/cystine reduction potential correlates with plasma creatinine levels in chronickidney disease, Blood Purif. 34 (Jan 2012) 231–237.

[28] R. Vanholder, E. Schepers, A. Pletinck, N. Neirynck, G. Glorieux, An update on pro-tein-bound uremic retention solutes, J. Ren. Nutr. 22 (2012) 90–94.

[29] N. Nakagawa, N. Hasebe, K. Sumitomo, T. Fujino, J. Fukuzawa, T. Hirayama, et al., Anoral adsorbent, AST-120, suppresses oxidative stress in uremic rats, Am. J. Nephrol.26 (2006) 455–461.

[30] J.S. Coombes, R.G. Fassett, Antioxidant therapy in hemodialysis patients: a systemat-ic review, Kidney Int. 81 (2012) 233–246.

[31] M. Tepel, M. van der Giet, M. Statz, J. Jankowski, W. Zidek, The antioxidantacetylcysteine reduces cardiovascular events in patients with end-stage renal fail-ure: a randomized, controlled trial, Circulation 107 (2003) 992–995.

[32] M.H. Massola Shimizu, T.M. Coimbra, M. De Araujo, L.F. Menezes, A.C. Seguro,N-acetylcysteine attenuates the progression of chronic renal failure, Kidney Interna-tional, 2005, pp. 2208–2217.

[33] K.R. Atkuri, J.J. Mantovani, L.A. Herzenberg, L.A. Herzenberg, N-acetylcysteine—a safeantidote for cysteine/glutathione deficiency, Curr. Opin. Pharmacol. 7 (Aug 2007)355–359.

[34] Y. Samuni, S. Goldstein, O.M. Dean, M. Berk, The chemistry and biological activitiesof N-acetylcysteine, Biochim. Biophys. Acta 1830 (2013) 4117–4129.

[35] B.H. Lauterburg, G.B. Corcoran, J.R. Mitchell, Mechanism of action of N-acetylcysteine in the protection against the hepatotoxicity of acetaminophen inrats in vivo, J. Clin. Invest. 71 (1983) 980–991.

[36] J.M. Burgunder, A. Varriale, B.H. Lauterburg, Effect of N-acetylcysteine on plasmacysteine and glutathione following paracetamol administration, Eur. J. Clin.Pharmacol. 36 (1989) 127–131.

[37] H. Izzedine, V. Guerin, V. Launay-Vacher, M. Bernard, G. Deray, Effect ofN-acetylcysteine on serum creatinine level, Nephrol. Dial. Transplant. 16 (Jul2001) 1514–1515.

[38] C.P. Lebel, H. Ischiropoulos, S.C. Bondys, Evaluation of the probe 2′,7′-dichiorofluorescin as an indicator of reactive oxygen species formation and oxida-tive stress, Chem. Res. Toxicol. 5 (1992) 227–231.

[39] C. Quijano, L. Castro, G. Peluffo, V. Valez, R. Radi, Enhancedmitochondrial superoxide inhyperglycemic endothelial cells: direct measurements and formation of hydrogen per-oxide and peroxynitrite, Am. J. Physiol. Heart Circ. Physiol. 293 (2007) H3404–H3414.

[40] K.M. Robinson, M.S. Janes, M. Pehar, J.S. Monette, M.F. Ross, T.M. Hagen, et al., Selec-tive fluorescent imaging of superoxide in vivo using ethidium-based probes, Proc.Natl. Acad. Sci. U. S. A. 103 (2006) 15038–15043.

[41] Z. Han, S. Varadharaj, R.J. Giedt, J.L. Zweier, H.H. Szeto, B.R. Alevriadou,Mitochondria-derived reactive oxygen species mediate heme oxygenase-1 expres-sion in sheared endothelial cells, J. Pharmacol. Exp. Ther. 329 (2009) 94–101.

[42] O.W. Griffith, Determination of glutathione and glutathione disulfide using glutathi-one reductase and 2-vinylpyridine, Anal. Biochem. 106 (1980) 207–212.

[43] I. Rahman, A. Kode, S.K. Biswas, Assay for quantitative determination of glutathioneand glutathione disulfide levels using enzymatic recycling method, Nat. Protoc. 1(2006) 3159–3165.

[44] C.K. Fujihara, G. De Nucci, R. Zatz, Chronic nitric oxide synthase inhibition aggravatesglomerular injury in rats with subtotal nephrectomy, J. Am. Soc. Nephrol. 5 (1995)1498–1507.

[45] I. Takac, K. Schröder, R.P. Brandes, The nox family of NADPH oxidases: friend or foeof the vascular system? Curr. Hypertens. Rep. 14 (2012) 70–78.

[46] G.F. Rushworth, I.L. Megson, Existing and potential therapeutic uses for N-acetylcysteine: the need for conversion to intracellular glutathione for antioxidantbenefits, Pharmacol. Ther. 141 (2014) 150–159.

[47] M.S. Kilberg, H.N. Christensen, M.E. Handlogten, Cysteine as a system-specific sub-strate for transport system in rat hepatocytes, Biochem. Biophys. Res. Commun. 88(May 1979) 744–751.

[48] D.T. Phelps, S.M. Deneke, D.L. Daley, B.L. Fanburg, Elevation of glutathione levels inbovine pulmonary artery endothelial cells by N-acetylcysteine, Am. J. Respir. CellMol. Biol. 7 (1992) 293–299.

[49] R.G. Knickelbein, T. Seres, G. Lam, R.B. Johnston, J.B. Warshaw, Characterization ofmultiple cysteine and cystine transporters in rat alveolar type II cells, Am. J. Physiol.273 (1997) L1147–L1155.

[50] T. Ishii, Y. Sugita, S. Bannai, Regulation of glutathione levels in mouse spleen lym-phocytes by transport of cysteine, J. Cell. Physiol. 133 (1987) 330–336.

[51] P.K. Mandal, A. Seiler, T. Perisic, P. Kölle, A. Banjac Canak, H. Förster, et al., Systemx(c)- and thioredoxin reductase 1 cooperatively rescue glutathione deficiency, J.Biol. Chem. 285 (2010) 22244–22253.

[52] C.A. Wagner, F. Lang, S. Bröer, Function and structure of heterodimeric amino acidtransporters, Am. J. Physiol. Cell Physiol. 281 (2001) C1077–C1093.

[53] M. Conrad, H. Sato, The oxidative stress-inducible cystine/glutamate antiporter, sys-tem x (c) (-): cystine supplier and beyond, Amino Acids 42 (2012) 231–246.

[54] S.Whillier, J.E. Raftos, B. Chapman, P.W. Kuchel, Role of N-acetylcysteine and cystinein glutathione synthesis in human erythrocytes, Redox Rep. 14 (2009) 115–124.

[55] R. Banerjee, Redox outside the box: linking extracellular redox remodeling with in-tracellular redox metabolism, J. Biol. Chem. 287 (2012) 4397–4402.

[56] Y.-M. Go, D.P. Jones, Cysteine/cystine redox signaling in cardiovascular disease, FreeRadic. Biol. Med. 50 (Feb 15 2011) 495–509.

116 S.D. Rodrigues et al. / Life Sciences 121 (2015) 110–116