Mitochondrial-targeted antioxidants represent a promising approach for prevention of cisplatin-induced nephropathy

Free Radical Biology & Medicine 52 (2012) 497–506

Contents lists available at SciVerse ScienceDirect

Free Radical Biology & Medicine

j ourna l homepage: www.e lsev ie r .com/ locate / f reeradb iomed

Original Contribution

Mitochondrial-targeted antioxidants represent a promising approach for preventionof cisplatin-induced nephropathy

Partha Mukhopadhyay a,1, Béla Horváth a,e,1, Zsuzsanna Zsengellér b,1, Jacek Zielonka c,1, Galin Tanchian a,Eileen Holovac a, Malek Kechrid a, Vivek Patel a, Isaac E. Stillman b, Samir M. Parikh d, Joy Joseph c,Balaraman Kalyanaraman c, Pál Pacher a,⁎a Laboratory of Physiologic Studies, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, MD 20892, USAb Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA 02215, USAc Free Radical Research Center, Biophysics Department, Medical College of Wisconsin, Milwaukee, WI 53226, USAd Division of Nephrology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA 02215, USAe Institute of Human Physiology and Clinical Experimental Research, Semmelweis University, Budapest, Hungary

⁎ Corresponding author. Fax: +1 301 480 0257.E-mail address: [email protected] (P. Pacher).

1 These authors contributed equally to this work.

0891-5849/$ – see front matter © 2011 Elsevier Inc. Alldoi:10.1016/j.freeradbiomed.2011.11.001

a b s t r a c t

Article history:Received 3 August 2011Revised 23 October 2011Accepted 3 November 2011Available online 10 November 2011

Keywords:NephropathyCisplatinOxidative stressMitochondriaMitochondrial antioxidantsFree radicals

Cisplatin is a widely used antineoplastic agent; however, its major limitation is the development of dose-dependent nephrotoxicity whose precise mechanisms are poorly understood. Here we show not only that mito-chondrial dysfunction is a feature of cisplatin nephrotoxicity, but also that targeted delivery of superoxide dismu-tase mimetics to mitochondria largely prevents the renal effects of cisplatin. Cisplatin induced renal oxidativestress, deterioration of mitochondrial structure and function, an intense inflammatory response, histopathologicalinjury, and renal dysfunction. A single systemic dose of mitochondrially targeted antioxidants, MitoQ or Mito-CP,dose-dependently prevented cisplatin-induced renal dysfunction. Mito-CP also prevented mitochondrial injuryand dysfunction, renal inflammation, and tubular injury and apoptosis. Despite being broadly renoprotectiveagainst cisplatin, Mito-CP did not diminish cisplatin's antineoplastic effect in a human bladder cancer cell line.Our results highlight the central role of mitochondrially generated oxidants in the pathogenesis of cisplatinnephrotoxicity. Because similar compounds seem to be safe in humans, mitochondrially targeted antioxidantsmay represent a novel therapeutic approach against cisplatin nephrotoxicity.

© 2011 Elsevier Inc. All rights reserved.

Cisplatin is commonly used to treat malignancies. Cisplatin bindsto DNA, forming inter- and intrastrand cross-links, resulting in defec-tive DNA templates and arrest of DNA synthesis in rapidly dividingcancer cells [1]. The major limitation of cisplatin chemotherapy isthe development of dose-dependent nephrotoxicity in about 30% ofpatients, preventing the administration of high doses to take fulladvantage of its chemotherapeutic efficacy [2,3]. Increased oxidativestress and inflammation have been implicated in cisplatin-inducedrenal tubular cell injury [4–6] as cisplatin accumulates predominantlyin tubular cells and undergoes metabolism; however, the precisemechanisms of cisplatin's renal toxicity are not yet well understood,and efficient approaches to attenuate this dose-limiting side effectare sorely needed [7].

In this study, we used a well-established mouse model of cisplatin-induced nephropathy [5,6,8–12] to investigate the role of mitochondrialdysfunction in cisplatin-induced kidney injury.We characterized cispla-tin's effects on renal oxidative stress, the local inflammatory response,and mitochondrial structure and function. Thereafter, we tested the

rights reserved.

efficacy of two well-characterized membrane-permeative small-molecule compounds that deliver superoxide dismutase (SOD) mi-metics preferentially into mitochondria in vivo, MitoQ and Mito-CP[13–15].

Material and methods

Animals and drug treatment

All animal experiments conformed to National Institutes of Health(NIH) guidelines and were approved by the Institutional Animal Careand Use Committee of the National Institute on Alcohol Abuse andAlcoholism (Bethesda, MD, USA). Six- to 8-week-old male C57Bl/6Jmice were obtained from The Jackson Laboratory (Bar Harbor, ME,USA). All animals were kept in a temperature-controlled environ-ment with a 12-h light–dark cycle and were allowed free accessto food and water at all times and were cared for in accordancewith NIH guidelines. Mice were sacrificed 72 h after a single injectionof cisplatin (cis-diammineplatinum(II) dichloride, 25 mg/kg ip;Sigma). Two mitochondrial antioxidants, MitoQ and Mito-CP, weresynthesized as described [13,16]. MitoQ and Mito-CP were stored inethanol at 50 mg/ml, further diluted in saline, and administered at

498 P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

0.3–10 mg/kg or as described, ip, once and starting 1 h before thecisplatin administration.

Renal function monitoring

Serum levels of creatinine and blood urea nitrogen (BUN) weremeasured using a VetTest 8000 blood chemistry analyzer (Idexx Lab)[5,12].

Electron microscopy

Anesthetized animals were perfusion-fixed with 1.25% glutaralde-hyde in 0.1 M cacodylate buffer. Kidneys were harvested and processedin standard fashion (Epon embedded) for transmission electronmicros-copy using a Jeol 1011 electron microscope.

Histological examination

After fixation of the kidneys with 10% formalin, renal tissues weresectioned and stained with periodic acid–Schiff (PAS) reagents forhistological examination. Tubular damage in PAS-stained sectionswas examined under the microscope and scored based on the per-centage of cortical tubules showing epithelial necrosis: 0, normal; 1,b10%; 2, 10–25%; 3, 26–75%; 4, >75%. Tubular necrosis was definedas the loss of the proximal tubular brush border, blebbing of apicalmembranes, tubular epithelial cell detachment from the basementmembrane, or intraluminal aggregation of cells and proteins as de-scribed [5]. For myeloperoxidase (MPO) staining slides were deparaf-finized and hydrated in descending gradations of ethanol, followed byantigen retrieval procedure. Next, sections were incubated in 0.3%H2O2 in PBS to block endogenous peroxidase activity. The sectionswere then incubated with anti-MPO (Biocare Medical, Concord, CA,USA) or anti-malondialdehyde (Genox, Baltimore, MD, USA) anti-bodies overnight at 4 °C in a moist chamber. Biotinylated secondaryantibodies and ABC reagent were added as per the kit's instructions(Vector Laboratories, Burlingame, CA, USA). Color development wasinduced by incubation with a DAB kit (Vector Laboratories) for3–5 min, and the sections were counterstained with nuclear fast redas described [5]. Finally, the sections were dehydrated in ethanoland cleared in xylene and mounted. The specific staining was visual-ized and images were acquired using an IX-81 microscope (Olympus,Center Valley, PA). The morphometric examination was performed ina blinded manner by two independent investigators.

In situ enzyme chemistry

After removal, kidneys were bi-valved and frozen immediately inisopentane cooled in liquid nitrogen. The tissues were cryosectioned(6 μm thick) and stained for NADH and cytochrome c oxidase (COX)activities, as described previously [17,18].

Renal terminal deoxynucleotidyl transferase-mediated nick-end labeling(TUNEL) immunohistochemistry

Paraffin sections were dewaxed and in situ detection of apoptosisin the renal tissues was performed using the TUNEL assay as per theinstructions provided with the kit (Roche Diagnostics, Indianapolis,IN, USA). The nuclei were labeled with Hoechst 33242 [12].

Isolation of mitochondria from tissues

Mitochondria were isolated from the kidneys of mice treated withvehicle or cisplatin and/or Mito-CP using a tissue mitochondrial isola-tion kit (Pierce Biotechnology, Rockford, IL, USA).

Renal DNA fragmentation ELISA

The quantitative determination of cytoplasmic histone-associatedDNA fragmentation was determined using an ELISA kit (RocheDiagnostics).

Determination of renal caspase 3/7, poly(ADP-ribose) polymerase(PARP), and MPO activities and nitrotyrosine (NT) and4-hydroxynonenal (HNE) content

Renal caspase 3/7, PARP, and MPO activities and NT and HNE con-tent in kidney homogenates were determined as described [5,12,19].

Real-time PCR analyses

RNA was isolated and prepared for cDNA as described [5,19]. Real-time PCR was performed in an ABI HT7900 instrument using Sybergreen technology. Each amplified sample in each well was analyzedfor homogeneity using dissociation curve analysis. The primers usedwere previously described [5,12].

Immunoblot analyses

Kidney tissues were homogenized in mammalian tissue proteinextraction reagent (Pierce) supplemented with protease and phos-phatase inhibitors (Roche Diagnostics). Blots were probed withNOX2 antibody obtained from BioLegend (San Diego, CA, USA).NOX4 monoclonal and polyclonal rabbit antibodies were obtainedfrom Abcam (Cambridge, MA, USA). Blots were incubated with pri-mary antibody at the recommended dilution and incubated overnightat 4 °C. After subsequent washing with PBS–Tween 20, the mem-branes were probed with the appropriate secondary antibodies con-jugated with horseradish peroxidase (Pierce) and incubated 1 h atroom temperature. Then the membranes were developed using aSuperSignal–West Pico substrate chemiluminescence detection kit(Pierce). To confirm uniform loading, membranes were stripped andreprobed with β-actin (Chemicon, Ramona, CA, USA). Immunoblotsobtained using a NOX4 monoclonal antibody showed patterns similarto those obtained with the use of a polyclonal antibody; however, thepolyclonal antibody appeared to be more specific (fewer additionalbands). Quantification of immunoblots was carried out by the Quan-tity One program (Bio-Rad, Hercules, CA, USA) and normalized toactin.

Cell culture of cancer cell line T24; cytotoxicity/cell survival assay

T24, a transitional cell carcinoma fromhuman urinary bladder, waspurchased from the American Type Culture Collection. The T24 cellline was cultured in McCoy's medium with 10% fetal calf serum and1% penicillin–streptomycin as recently described [12]. The effects ofcisplatin and Mito-CP, or their combination, on cell survival wereassessed using the XTT assay (Cell Proliferation Kit II; Roche Diagnos-tics). Cells were seeded into 96- or 24-well culture plates. Twenty-fourhours later, cells were treatedwith various concentrations of cisplatin,Mito-CP, or their combination as described in the figure legends.Untreated cells were cultured in parallel as a negative control. After72 h of incubation, cells were treated with a 50-μl aliquot of the XTTsolution (1 ml XTT labeling/20 μl electron-coupling reagent) in eachwell at the end of the experiment. After 2 h of incubation, the absor-bance was measured at both 492 nm and a reference wavelength(690 nm) as described [12].

Clonogenic assay

Colony-forming (clonogenic) assays of T24 cancer cells treatedwith vehicle, cisplatin, and Mito-CP, or their combination, were

499P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

performed as described previously [20] and the concentrations of cis-platin and Mito-CP are indicated in the figures. Colony-forming unitswere calculated from four experiments in each group.

Sample preparation and EPR measurements

Plasma samples were analyzed without any processing. Frozenkidneys (single, randomly chosen from a pair) were weighed and ho-mogenized in ice-cold 0.1 M phosphate buffer containing 100 μMdiethylenetriaminepentaacetic acid (3 μl per 1 mg of tissue). Afterhomogenization, the samples were quickly frozen by immersion inliquid nitrogen and kept at −80 °C until the analysis. Plasma samplesor tissue homogenates were thawed and immediately transferredinto electron paramagnetic resonance (EPR) capillaries and the EPRspectra scanned. For analysis of the total amount of Mito-CP11

Fig. 1. Mitochondrial antioxidants prevent cisplatin-induced renal dysfunction and mitochodamage as evidenced by the attenuation of the increase in BUN and creatinine levels and/otration to mice. Electron microscopy of the renal cortex (C) demonstrates the preservationtreated control (i and iv), intact proximal tubular epithelium with preserved mitochondriais shown. Note the epithelial swelling and vacuolization accompanied by preservation of thswelling and disruption of their cristae in the cisplatin-treated group. Nuclear structure apmitochondrial structures in the cisplatin-treated mouse kidney. Original magnifications: i, ii,ment membrane; M, mitochondria). Results are means±SEM of 5–9 experiments/group. *

(nitroxide Mito-CP and its reduced form, hydroxylamine Mito-CPH),aliquots of thawed samples were mixed 1:1 (by volume) with di-methyl sulfoxide (DMSO) and vortexed with 1 mM K3Fe(CN)6(added by 1:100 dilution of 0.1 M aqueous solution of K3Fe(CN)6)before EPR measurement. EPR spectra were recorded at room tem-perature using a Bruker EMX spectrometer operating at 9.85 GHzand equipped with a Bruker ER 4119HS-WI high-sensitivity resona-tor. Typical spectrometer parameters were scan range 80 G, timeconstant 1.28 ms, sweep time 42 s, modulation amplitude 1.0 G,modulation frequency 100 kHz, microwave power 20 mW. Micropi-pettes (50 μl for plasma samples and 100 μl for tissue homogenates)were used as sample tubes. Spectra were averaged over five scans.As the EPR signals due to the presence of ascorbyl radical wereoverlapping with the center line of the Mito-CP spectrum, the quanti-fication was done by comparison of the intensities of the low-

ndrial injury. (A) MitoQ and (B) Mito-CP ameliorate cisplatin-induced profound kidneyr (C) tubular damage (electron microscopy) in the kidney 72 h after cisplatin adminis-of mitochondrial structure after mitochondrial antioxidant administration. In sham-

is shown. In cisplatin-treated kidney (ii), a proximal tubule with extensive acute injurye brush border. Higher magnification (v) shows a large number of mitochondria withpears maintained. Mito-CP treatment (iii and vi) preserves the tubular epithelium andand iii—3000×; iv, v, and vi—20,000× (BB, brush border; N, nucleus; TBM, tubular base-Pb0.05 vs vehicle; #Pb0.05 vs cisplatin.

Fig. 2. Mito-CP attenuates cisplatin-induced kidney damage. As shown in the representative images cisplatin induced profound histopathological renal injury 72 h after adminis-tration to mice, evidenced by protein cast, vacuolation, and desquamation of epithelial cells in the renal tubules using PAS staining, which were largely prevented by Mito-CP treat-ment given from 1 h before the cisplatin injection. The calculation of the tubular damage score is described under Material and methods. Results are means±SEM of 6 or 7experiments/group. *Pb0.01 vs vehicle; #Pb0.01 vs cisplatin.

500 P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

field EPR peak of Mito-CP with the Mito-CP solutions of knownconcentrations, measured under the same conditions. All dataare presented as mean values and error bars represent standarddeviations.

Fig. 3. Mito-CP attenuates cisplatin-induced mitochondrial dysfunction: in situ enzyme chesnap-frozen kidneys 72 h after vehicle, cisplatin, and cisplatin + Mito-CP treatment. Enzymvented by Mito-CP treatment (representative images of n=5/condition are shown). G, glostaining) 72 h after vehicle, cisplatin, and cisplatin + Mito-CP treatment (representative of

Statistical analysis

Results are expressed asmeans±SEM. Statistical significance amonggroups was determined by one-way ANOVA followed by Newman–

mistry. (A) Staining (brown) for cytochrome c oxidase enzyme activity on sections ofatic activity is greatly decreased after cisplatin treatment in the cortex, which was pre-merulus; original magnification 40×. (B) NADH dehydrogenase enzyme activity (bluen=5/condition). Original magnification 40×.

501P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

Keuls post hoc analysis using GraphPad Prism 5 software (SanDiego, CA,USA). Probability values of Pb0.05 were considered significant.

Results

Mitochondrially targeted antioxidants prevent cisplatin-induced renaldysfunction, histopathological alterations, and tubular mitochondrialinjury

Intraperitoneal administration of a single dose of mitochondrialantioxidants, MitoQ and Mito-CP, 1 h before cisplatin treatmentdose-dependently (0.3 to 10 mg/kg) prevented the cisplatin-inducedrenal dysfunction (attenuated the increase in BUN and creatininevalues; Figs. 1A and B). Based on this result, a fully effective dose ofMito-CP (3 mg/kg ip) was used in subsequent mechanistic studies.

Fig. 4. Mito-CP attenuates cisplatin-induced oxidative and nitrative stress. Cisplatin significcells, (B) HNE protein adduct and (D) 3-nitrotyrosine levels in the kidney tissues or (C) isoenhanced oxidative/nitrative stress 72 h (or 6 h if indicated) after its administration to mimeans±SEM of 6–15/group. *Pb0.05 vs vehicle; #Pb0.05 vs cisplatin.

Next, we evaluated renal cortical tubules with transmission electronmicroscopy to determine the effects of cisplatin on mitochondrialultrastructure with or without Mito-CP pretreatment. Cisplatin in-duced extensive acute tubular damage (Figs. 1C, ii, and 2). Highermagnification (Fig. 1C, v) revealed swollen mitochondria with disrup-tion of cristae, suggesting mitochondrial injury. Mito-CP treatmentcompletely prevented cisplatin-induced mitochondrial structuraldamage and tubular injury (Figs. 1C, iii and vi, and 2).

To determine the localization and severity of mitochondrial dys-functionwithin the kidney, wemeasured the in situ activity of the elec-tron transport chain enzyme complex component COX (Fig. 3A). In thisenzymatic assay, incubation of tissue sections with substrate that iscombinedwith the cytochrome c oxidase enzyme (oxygen acting as ac-ceptor) produces a series of reactions resulting in brown staining at thesite of enzyme activity. The degree of brown staining therefore reflects

antly increased renal (A) malondialdehyde (red-brown) staining in damaged tubularlated mitochondria, and (E) mRNA and (F) protein expression for NOX2/4, indicatingce. These changes could be largely prevented by treatment with Mito-CP. Results are

502 P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

cytochrome c oxidase activity, which was high in cortical tubules anddramatically reduced 72 h after cisplatin administration.

Just as prior reports on cisplatin nephrotoxicity had identified cor-tical tubules as the most histologically affected region of the kidney[21–23], our study of in situmitochondrial function also demonstratedthat the most prominently affected tubules were in the cortex. Mito-CP pretreatment prevented the decrease in mitochondrial enzymeactivity. To confirm these results, we repeated the in situ study withanother enzyme component of the mitochondrial electron transportchain, NADH dehydrogenase (Fig. 3B). This technique produced re-sults analogous to those of the COX assay, again revealing severemito-chondrial dysfunction in cortical tubules that was nearly completelyprevented by Mito-CP. These results provide strong evidence that cis-platin disrupts the normal structure and function of renal tubular mi-tochondria and that systemic delivery of mitochondrially targetedantioxidants can largely prevent these pathological alterations withinthe kidney.

Mito-CP attenuates cisplatin-induced oxidative and nitrative stress

Whereas several cell culture studies have implicated tubular mito-chondria as a source of oxidant stress after cisplatin [4,24,25], lesswork has been focused on testing and exploiting this hypothesis invivo. We found that cisplatin exposure significantly increased renal

Fig. 5. Mito-CP attenuates cisplatin-induced inflammation. Cisplatin significantly increased m(C and D) myeloperoxidase staining and activity, and (E and F) adhesion molecule ICAM-1 aministration to mice, indicating enhanced inflammatory response. These changes could be*Pb0.05 vs vehicle; #Pb0.05 vs cisplatin.

oxidative/nitrative stress as evidenced by enhanced accumulation ofmalondialdehyde (a marker of lipid peroxidation/oxidative stress;red-brown staining) in damaged tubular cells (Fig. 4A), elevated HNEand nitrotyrosine levels in kidney homogenates (Figs. 4B and D), andelevated HNE in isolated kidney mitochondria (Fig. 4C). Mito-CP inhib-ited cisplatin-induced oxidative stress both in the kidney tissues(Figs. 4A, B, and D) and in isolated kidney mitochondria (Fig. 4C).Mito-CP also inhibited the late/secondary expression of mRNA and pro-tein of reactive oxygen species (ROS)-generating NOX2 and NOX4 in-duced by cisplatin (Figs. 4E and F).

Mito-CP mitigates cisplatin-induced acute and late inflammatoryresponse

Because proinflammatory chemokines/cytokines and expressionof adhesion molecules are critical mediators of renal dysfunctionafter cisplatin exposure [6,9,10,26,27], we investigated the effects ofMito-CP on these processes as well. Cisplatin markedly increased theexpression of the proinflammatory chemokines MCP-1 and MIP1α/2(Figs. 5A and B), the proinflammatory cytokine tumor necrosis factorα (TNF-α; Fig. 5F), the adhesion molecule ICAM-1 (Fig. 5E), andrenal myeloperoxidase staining and activity (an indicator of leukocyteinfiltration; Figs. 5C and D). These changes were all significantly re-duced by pretreatment with Mito-CP.

RNA expression of proinflammatory chemokines (A) MCP-1 and (B) MIP1α and MIP2,nd proinflammatory cytokine TNF-α mRNA expression in the kidneys 72 h after its ad-largely prevented by treatment with Mito-CP. Results are means±SEM of 6–16/group.

503P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

Mito-CP attenuates cisplatin-induced cell death in the kidney

Cisplatin significantly increased apoptotic (caspase 3/7, DNA fragmen-tation, TUNEL) and poly(ADP-ribose) polymerase-dependent cell death(Figs. 6A–D) in the kidneys, both of which were prevented by Mito-CP.

The development of cisplatin-induced intrarenal oxidative stress,inflammation, tubular cell death, and renal dysfunction istime-dependent

Consistent with our recent observations [12], an intraperitonealdose of cisplatin induced a time-dependent increase in HNE, a markerof lipid peroxidation, in the kidney detectable 6 h after its administra-tion, which coincided with an acute proinflammatory chemokine re-sponse and increased expression of the adhesion molecule ICAM-1(not shown, see [12]). NT, a marker of reactive nitrogen species(e.g., peroxynitrite) generation and/or protein nitration [28], also in-creased from 6 to 12 h after cisplatin administration in the kidneys[12]. The expression of mRNAs of ROS-generating NAD(P)H oxidaseisoforms NOX2 and NOX4 occurred only from 24 to 48 h after cisplat-in administration [12]. This again coincided with the increased leuko-cyte infiltration in the kidneys and markedly increased TNF-α levelsand associated second wave of ROS/reactive nitrogen species (RNS)[12]. The peak of delayed inflammatory response also correlatedwith increased apoptotic and necrotic cell death in the kidneys andrenal dysfunction [12]. These results suggest that increased ROS gen-eration and oxidant-induced damage in the kidneys precede the in-flammatory response and dysfunction.

Pharmacokinetics of Mito-CP: EPR analyses

After injection of Mito-CP11 it undergoes a partial reduction in theblood to an EPR-silent hydroxylamine, as evidenced by the increase inthe EPR signal intensity caused by the addition of ferricyanide. Ferri-cyanide oxidizes hydroxylamines back to an EPR-active form. In

Fig. 6. Mito-CP attenuates cisplatin-induced renal cell death. (A–D) Cisplatin significantly iafter its administration to mice, which could be largely prevented by treatment with Mito-staining (the TUNEL-positive nuclei are light blue). Results are means±SEM of 6–8/group.

plasma, Mito-CP is cleared within 12 h of injection (Figs. 7A and B).The lack of the EPR signal at ≥12 h is not due to the reduction toEPR-silent hydroxylamine, as ferricyanide is unable to restore the sig-nal intensity. In the kidney tissue, Mito-CP is present mostly in the re-duced form up to 6 h after injection (Figs. 7C and D). At ≥12 h afterinjection no EPR signal was detected in any kidney tissue sampleeven after the addition of ferricyanide (Figs. 7C and D).

Mito-CP does not inhibit cisplatin-mediated T24 cancer cell death

Several renoprotective strategies against cisplatin may be less ap-plicable clinically because of reduced antineoplastic effect [7]. Be-cause one of the major indications for cisplatin is for treatment ofurinary tract cancers, we determined if Mito-CP interferes with thetherapeutic effect of cisplatin in the human bladder carcinoma cellline T24 (Fig. 8). Cisplatin induced concentration-dependent celldeath in cancer cells measured by XTT assay after 72 h (Fig. 8A),which was actually further enhanced by Mito-CP (Fig. 8A). Further-more, Mito-CP also promoted the cisplatin-induced decrease intumor cell colony formation (Fig. 8B). These results are consistentwith recently published studies in other cancer cells [29,30]. There-fore, Mito-CP did not diminish the chemotherapeutic efficacy of cis-platin. In contrast, Mito-CP actually enhanced this effect of cisplatin.

Discussion

In this study we investigated whether mitochondrially targetedantioxidants could protect against cisplatin-induced kidney injuryusing a preclinical mouse model of cisplatin-induced nephropathy.Previous mechanistic studies using this model have variously impli-cated tubular apoptosis and cell-cycle changes, mitochondrial injury,oxidant stress, and local inflammation as key upstream events that re-sult in renal dysfunction [5,7,12], but the interrelationships amongthese have remained unclear. We hypothesized that cisplatin inducesmitochondrial oxidant stress, which in turn results in mitochondrial

ncreased various markers of apoptotic and PARP-dependent cell death in kidneys 72 hCP. (D) An example of representative TUNEL staining colocalized with nuclear Hoechst*Pb0.05 vs vehicle; #Pb0.05 vs cisplatin.

Fig. 7.Mito-CP pharmacokinetics in plasma and kidney. (A–D) EPR analyses of the pharmacokinetics of Mito-CP11 in mouse plasma and kidney after ip injection of Mito-CP11 at thedose of 10 mg/kg. Representative EPR spectra of plasma and tissue homogenates mixed with DMSO and 1 mMK3Fe(CN)6 are shown in (A) and (C), respectively. The results of quan-titative analysis are shown in (B) (concentration of Mito-CP11 in plasma) and (D) (concentration of Mito-CP11 in kidney homogenates, normalized to the protein concentration).

504 P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

injury and dysfunction, tubular cell death, and local inflammation thatall culminate in renal dysfunction. To test this, we administeredmembrane-permeative small-molecule SOD mimetics that target mi-tochondria 1 h before cisplatin challenge and observed improvementsin each of these pathogenic events including, most importantly, globalrenal function. These findings suggest that mitochondrial ROS may bean indispensable mediator of cisplatin nephrotoxicity.

In agreement with prior reports, we found that cisplatin inducedmarked tubular injury, increased inflammatory cell infiltration and

Fig. 8. Mito-CP enhances cisplatin-mediated T24 cancer cell death. Cisplatin induced (A) conforming units, which were enhanced in the presence of Mito-CP, which by itself was able to*Pb0.05 vs vehicle; #Pb0.05 vs corresponding cisplatin, $Pb0.05 vs vehicle.

oxidative/nitrative stress, and impaired renal function [5,6,8–12]. Asrevealed by transmission electron microscopy, one of the most prom-inent features of cisplatin-induced tubular damage and nephropathywas the swelling of mitochondria with disruption of cristae in practi-cally all tubular cells. Two independent studies of in situ electrontransport chain activity not only confirmedwidespreadmitochondrialdysfunction, but localized the mitochondrial lesion to cortical tubules.Most critically, an intervention designed to target mitochondrial oxi-dant production largely prevented the functional and structural

centration-dependent cell death in T24 cancer cells or (B) a decrease in tumor colony-attenuate slightly the cell viability in cancer cells. Results are means±SEM of 4/group.

505P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

lesions in tubular mitochondria, suggesting that cisplatin's ability toinduce mitochondrial production of oxidants is key to its ability to in-jure mitochondria.

Increased oxidative stress was one of the earliest features we ob-served during the development of cisplatin-induced nephropathy,which was followed by a marked inflammatory cell infiltration and asecondary wave of ROS generation [12]. The secondary ROS mecha-nism probably involves the phagocyte NAD(P)H oxidase isoformgp91phox/NOX2, as well as NOX4/renox (the NAD(P)H oxidase iso-form considered to be a major source of ROS generation in the kidneysunder pathological conditions) because the expression of these ROS-generating enzymes was significantly increased in the kidneys fromthe second day of cisplatin administration [12]. The increased NOX2expression was consistent with enhanced leukocyte infiltration inthe kidneys of cisplatin-treated mice around damaged tubular struc-tures [12]. Cisplatin-induced ROS generation may also promote theexpression of adhesion molecules through the activation of NF-κB.Indeed, we found increased expression of adhesion molecule ICAM-1and enhanced chemokine signaling (MCP-1, MIP1α/2) in the kidneyof cisplatin-treated mice. These chemokines may further facilitate mi-gration and adhesion of inflammatory cells to the activated endotheli-um or damaged tubular cells. These activated immune cells then mayrelease various chemokines, cytokines, and ROS/RNS, further amplify-ing the inflammatory cascade and injury [28,31,32]. We also detectedmarked increases in mRNA expression of the proinflammatory cyto-kine TNF-α, consistent with exacerbated inflammatory processes.

+SOD

PR

AG

Damage

Nucleus

NAD+

Cytoplasm

Fr

DNAbreaks

(inactive)

PARP-1(active)

NAD+

PARP-1

Repair

Excessive damMild damage

ONOO_

NeutrophilLymphocyte

Macrophage

Pro-inflammatory and pgene expression (TNFαα, c

adhesion molecules, iNOS, N

iNOS

NFκB

NO O_2

.NOX2, NOX4

H2O2responseInflammatory

ROS

Cisplati

Fig. 9. Cisplatin-induced mitochondrial reactive oxygen species generation triggers inflammgers oxidative stress in the mitochondria of kidney proximal tubular and endothelial cells,sponse. The secondary ROS most probably also involves the phagocyte NAD(P)H oxidaimportant source of ROS in the kidney under pathological conditions). The cisplatin-inducedclear enzyme poly(ADP) ribose polymerase-1 (PARP-1), with consequent tubular and/or endκB. Inflammation may further amplify oxidative/nitrative stress, and these interrelated proc(both apoptotic and necrotic), secondary hypoxia, kidney dysfunction, and failure.

Inflammation may also enhance oxidative stress, and these processesare interrelated, leading to a concerted activation of various mito-chondrial and other (e.g., PARP-1-dependent) cell death pathways,as demonstrated in the cisplatin-induced nephropathy model (seealso Fig. 9).

A single dose of Mito-CP, which was cleared from the serum andkidney within 12 h of its administration, not only attenuated thecisplatin-induced early and delayed ROS generation andmitochondri-al injury, but also blunted the secondarywave of inflammation and as-sociated cell death. Together, these observations strongly indicate thatearlymitochondrial ROS generation triggers the deleterious cascade ofinflammation and tissue injury and suggest fruitful areas for futureexploration. First, if tubular mitochondrial ROS production is an earlyand necessary pathogenic event after systemic cisplatin, how doesits accumulation in tubular cells trigger mitochondrial oxidant gener-ation? Second, mitochondrial fragmentation after cisplatin also seemsto be an important event for tubular cytotoxicity to occur [8]. Consid-ered in the context of the present results, prevention of mitochondrialROS may attenuate the tendency for mitochondria to fragment aftercisplatin. Conversely, fragmentation may accelerate the breakdownof the electron transport chain, resulting in ROS. Related to this, cis-platin inhibits fatty acid oxidation in tubular mitochondria, and trans-genic expression of the metabolic transcription factor peroxisomeproliferation-activated receptor-α (PPAR-α), in proximal tubularcells, prevents cisplatin-induced lipid peroxidation and nephrotoxici-ty [33]. Induction of antioxidant genes downstream of PPAR-α or

AIF

ΔΨ↓Mitochondrion

ee PAR polymer

Apoptosis

DNA fragmentationNecrosis

cyt c

↓ ATP ↓

Caspaseactivation

age

endothelial cellsKidney tubular and

ro-oxidant hemokines,

OX2, NOX4, etc.)

Stress signallingImpaired vascular functionProtein oxidation, nitration

Inactivation of enzymesLipid peroxidation

Dysfunctionnephropathy

Mitochondrion

ROS/RNS

n

atory response, cell death, and kidney dysfunction/nephropathy. Cisplatin initially trig-which is followed by a secondary wave of ROS/RNS generation and inflammatory re-se isoform gp91phox/NOX2, as well as NOX4/renox (isoform considered to be anROS/RNS generation also induces oxidative DNA injury and rapid activation of the nu-othelial cell death, and activation of proinflammatory transcription factors such as NF-esses eventually culminate in more concerted renal tubular and endothelial cell demise

506 P. Mukhopadhyay et al. / Free Radical Biology & Medicine 52 (2012) 497–506

prevention of the secondary wave of inflammation may contribute toits beneficial effects [34]. Third, previous human trials using antioxi-dants to prevent cisplatin toxicity have yielded mixed results[35,36]. We speculate that these approaches may not have efficientlytargeted the subcellular source of oxidants. Finally, pro-oxidant func-tion of cisplatin may also contribute to its antineoplastic effect; there-fore, our demonstration that Mito-CP actually sensitized a humanbladder cancer cell line to cisplatin lends further support to the devel-opment of similar compounds for clinical use. Further studies will beneeded to understand how this sensitization occurs.

In summary, mitochondrial ROS generation may initiate the delete-rious cascade of events induced by cisplatin, culminating in tubular celldeath and diminished renal function (the simplified mechanisms ofcisplatin-induced nephrotoxicity/nephropathy are summarized inFig. 9). Thus, mitochondrially targeted antioxidants represent a novelapproach to prevent this devastating complication of chemotherapy,thereby also increasing its therapeutic window. This work not onlyestablishes an in vivo mechanism implicating cortical tubular mito-chondrial ROS after cisplatin—a pathway previously tested primarilyin cultured cells—but also provides justification for the developmentof this preventative strategy for clinical use. This is particularly excitingas mitochondrially targeted antioxidants such as MitoQ are alreadybeing evaluated in humans for various therapeutic indications with noundue toxic effects (see ClinicalTrials.gov). Our findings may also beapplicable in other forms of acute kidney injury, such as ischemia–reperfusion, for which several investigators have demonstrated a path-ogenic role for mitochondrial dysfunction [37–39].

Acknowledgments

This studywas supported by the Intramural Research ProgramofNIH/NIAAA (to P.P.) and by National Institutes of Health Grant RO1CA152810(B.K.). Dr. Horvath is the recipient of aHungarian Scientific Research Fundfellowship (OTKA-NKTH-EU MB08 80238). The authors are indebted toDan Brown and Lena Ellezian for help with electron microscopy samplepreparation, Dr. George Kunos (the Scientific Director of the NIAAA) forproviding key resources, and the American Society of Nephrology (CarlGottschalk award to S.M.P.). Dr. Pacher dedicates this study to his belovedmother Iren Bolfert, who died from complications of chemotherapy.

References

[1] Wang, D.; Lippard, S. J. Cellular processing of platinum anticancer drugs. Nat. Rev.Drug Discov. 4:307–320; 2005.

[2] Ries, F.; Klastersky, J. Nephrotoxicity induced by cancer chemotherapy with spe-cial emphasis on cisplatin toxicity. Am. J. Kidney Dis. 8:368–379; 1986.

[3] Schrier, R. W. Cancer therapy and renal injury. J. Clin. Invest. 110:743–745; 2002.[4] Davis, C. A.; Nick, H. S.; Agarwal, A. Manganese superoxide dismutase attenuates

cisplatin-induced renal injury: importance of superoxide. J. Am. Soc. Nephrol. 12:2683–2690; 2001.

[5] Mukhopadhyay, P.; Rajesh, M.; Pan, H.; Patel, V.; Mukhopadhyay, B.; Batkai, S.; Gao, B.;Hasko, G.; Pacher, P. Cannabinoid-2 receptor limits inflammation, oxidative/nitrosativestress, and cell death in nephropathy. Free Radic. Biol. Med. 48:457–467; 2010.

[6] Ramesh, G.; Reeves, W. B. TNF-alpha mediates chemokine and cytokine expressionand renal injury in cisplatin nephrotoxicity. J. Clin. Invest. 110:835–842; 2002.

[7] Pabla, N.; Dong, Z. Cisplatin nephrotoxicity: mechanisms and renoprotectivestrategies. Kidney Int. 73:994–1007; 2008.

[8] Brooks, C.; Wei, Q.; Cho, S. G.; Dong, Z. Regulation of mitochondrial dynamics in acutekidney injury in cell culture and rodent models. J. Clin. Invest. 119:1275–1285; 2009.

[9] Zhang, B.; Ramesh, G.; Norbury, C. C.; Reeves, W. B. Cisplatin-induced nephrotox-icity is mediated by tumor necrosis factor-alpha produced by renal parenchymalcells. Kidney Int. 72:37–44; 2007.

[10] Zhang, B.; Ramesh, G.; Uematsu, S.; Akira, S.; Reeves, W. B. TLR4 signaling medi-ates inflammation and tissue injury in nephrotoxicity. J. Am. Soc. Nephrol. 19:923–932; 2008.

[11] Pabla, N.; Dong, G.; Jiang, M.; Huang, S.; Kumar, M. V.; Messing, R. O.; Dong, Z. In-hibition of PKCdelta reduces cisplatin-induced nephrotoxicity without blockingchemotherapeutic efficacy in mouse models of cancer. J. Clin. Invest. 121; 2011.

[12] Mukhopadhyay, P.; Horvath, B.; Kechrid, M.; Tanchian, G.; Rajesh, M.; Naura, A. S.;Boulares, A. H.; Pacher, P. Poly(ADP-ribose) polymerase-1 is a key mediator ofcisplatin-induced kidney inflammation and injury. Free Radic. Biol. Med. 51:1774–1788; 2011.

[13] Dhanasekaran, A.; Kotamraju, S.; Karunakaran, C.; Kalivendi, S. V.; Thomas, S.;Joseph, J.; Kalyanaraman, B. Mitochondria superoxide dismutase mimetic in-hibits peroxide-induced oxidative damage and apoptosis: role of mitochondrialsuperoxide. Free Radic. Biol. Med. 39:567–583; 2005.

[14] Smith, R. A.; Murphy, M. P. Mitochondria-targeted antioxidants as therapies.Discov. Med. 11:106–114; 2011.

[15] Smith, R. A.; Hartley, R. C.; Murphy, M. P. Mitochondria-targeted small moleculetherapeutics and probes. Antioxid. Redox Signal. 15:3021–3038; 2011.

[16] Cassina, P.; Cassina, A.; Pehar,M.; Castellanos, R.; Gandelman,M.; de Leon, A.; Robinson,K. M.; Mason, R. P.; Beckman, J. S.; Barbeito, L.; Radi, R. Mitochondrial dysfunctionin SOD1G93A-bearing astrocytes promotes motor neuron degeneration: preventionby mitochondrial-targeted antioxidants. J. Neurosci. 28:4115–4122; 2008.

[17] Lebrecht, D.; Setzer, B.; Rohrbach, R.; Walker, U. A. Mitochondrial DNA and its re-spiratory chain products are defective in doxorubicin nephrosis. Nephrol. Dial.Transplant. 19:329–336; 2004.

[18] Sheehan, D. Theory and Practice of Histotechnology. MOSBY, Maryland heights,MO, p. 308; 1980.

[19] Mukhopadhyay, P.; Horvath, B.; Rajesh, M.; Matsumoto, S.; Saito, K.; Batkai, S.;Patel, V.; Tanchian, G.; Gao, R. Y.; Cravatt, B. F.; Hasko, G.; Pacher, P. Fatty acidamide hydrolase is a key regulator of endocannabinoid-induced myocardial tis-sue injury. Free Radic. Biol. Med. 50:179–195; 2011.

[20] Franken, N. A.; Rodermond, H. M.; Stap, J.; Haveman, J.; van Bree, C. Clonogenicassay of cells in vitro. Nat. Protoc. 1:2315–2319; 2006.

[21] Li, S.; Basnakian, A.; Bhatt, R.; Megyesi, J.; Gokden, N.; Shah, S. V.; Portilla, D.PPAR-alpha ligand ameliorates acute renal failure by reducing cisplatin-inducedincreased expression of renal endonuclease G. Am. J. Physiol. Renal Physiol. 287:F990–F998; 2004.

[22] Megyesi, J.; Safirstein, R. L.; Price, P. M. Induction of p21WAF1/CIP1/SDI1 in kid-ney tubule cells affects the course of cisplatin-induced acute renal failure. J. Clin.Invest. 101:777–782; 1998.

[23] Tsuruya, K.; Ninomiya, T.; Tokumoto, M.; Hirakawa, M.; Masutani, K.; Taniguchi,M.; Fukuda, K.; Kanai, H.; Kishihara, K.; Hirakata, H.; Iida, M. Direct involvementof the receptor-mediated apoptotic pathways in cisplatin-induced renal tubularcell death. Kidney Int. 63:72–82; 2003.

[24] Kruidering, M.; Van de Water, B.; de Heer, E.; Mulder, G. J.; Nagelkerke, J. F. Cis-platin-induced nephrotoxicity in porcine proximal tubular cells: mitochondrialdysfunction by inhibition of complexes I to IV of the respiratory chain. J. Pharma-col. Exp. Ther. 280:638–649; 1997.

[25] Baek, S. M.; Kwon, C. H.; Kim, J. H.; Woo, J. S.; Jung, J. S.; Kim, Y. K. Differentialroles of hydrogen peroxide and hydroxyl radical in cisplatin-induced cell deathin renal proximal tubular epithelial cells. J. Lab. Clin. Med. 142:178–186; 2003.

[26] Deng, J.; Kohda, Y.; Chiao, H.; Wang, Y.; Hu, X.; Hewitt, S. M.; Miyaji, T.; McLeroy,P.; Nibhanupudy, B.; Li, S.; Star, R. A. Interleukin-10 inhibits ischemic andcisplatin-induced acute renal injury. Kidney Int. 60:2118–2128; 2001.

[27] Ramesh, G.; Reeves, W. B. TNFR2-mediated apoptosis and necrosis in cisplatin-induced acute renal failure. Am. J. Physiol. Renal Physiol. 285:F610–F618; 2003.

[28] Szabo, C.; Ischiropoulos, H.; Radi, R. Peroxynitrite: biochemistry, pathophysiologyand development of therapeutics. Nat. Rev. Drug Discov. 6:662–680; 2007.

[29] Weinberg, F.; Hamanaka, R.; Wheaton, W. W.; Weinberg, S.; Joseph, J.; Lopez, M.;Kalyanaraman, B.; Mutlu, G. M.; Budinger, G. R.; Chandel, N. S. Mitochondrial me-tabolism and ROS generation are essential for Kras-mediated tumorigenicity.Proc. Natl. Acad. Sci. U. S. A. 107:8788–8793; 2010.

[30] Rao, V. A.; Klein, S. R.; Bonar, S. J.; Zielonka, J.; Mizuno, N.; Dickey, J. S.; Keller, P. W.;Joseph, J.; Kalyanaraman, B.; Shacter, E. The antioxidant transcription factor Nrf2negatively regulates autophagy and growth arrest induced by the anticancer redoxagent mitoquinone. J. Biol. Chem. 285:34447–34459; 2010.

[31] Ischiropoulos, H.; Zhu, L.; Beckman, J. S. Peroxynitrite formation frommacrophage-derived nitric oxide. Arch. Biochem. Biophys. 298:446–451; 1992.

[32] Pacher, P.; Beckman, J. S.; Liaudet, L. Nitric oxide and peroxynitrite in health anddisease. Physiol. Rev. 87:315–424; 2007.

[33] Li, S.; Nagothu, K. K.; Desai, V.; Lee, T.; Branham, W.; Moland, C.; Megyesi, J. K.;Crew, M. D.; Portilla, D. Transgenic expression of proximal tubule peroxisomeproliferator-activated receptor-alpha in mice confers protection during acute kid-ney injury. Kidney Int. 76:1049–1062; 2009.

[34] Li, S.; Gokden, N.; Okusa, M. D.; Bhatt, R.; Portilla, D. Anti-inflammatory effect offibrate protects from cisplatin-induced ARF. Am. J. Physiol. Renal Physiol. 289:F469–F480; 2005.

[35] Weij, N. I.; Elsendoorn, T. J.; Lentjes, E. G.; Hopman, G. D.; Wipkink-Bakker, A.;Zwinderman, A. H.; Cleton, F. J.; Osanto, S. Supplementation with antioxidantmicronutrients and chemotherapy-induced toxicity in cancer patients treatedwith cisplatin-based chemotherapy: a randomised, double-blind, placebo-controlled study. Eur. J. Cancer 40:1713–1723; 2004.

[36] Pace, A.; Savarese, A.; Picardo, M.; Maresca, V.; Pacetti, U.; Del Monte, G.; Biroccio,A.; Leonetti, C.; Jandolo, B.; Cognetti, F.; Bove, L. Neuroprotective effect of vitaminE supplementation in patients treated with cisplatin chemotherapy. J. Clin. Oncol.21:927–931; 2003.

[37] Szeto, H. H.; Liu, S.; Soong, Y.; Wu, D.; Darrah, S. F.; Cheng, F. Y.; Zhao, Z.; Ganger,M.; Tow, C. Y.; Seshan, S. V. Mitochondria-targeted peptide accelerates ATP recov-ery and reduces ischemic kidney injury. J. Am. Soc. Nephrol. 22:1041–1052; 2011.

[38] Weinberg, J. M.; Venkatachalam, M. A.; Roeser, N. F.; Nissim, I. Mitochondrial dys-function during hypoxia/reoxygenation and its correction by anaerobic metabolismof citric acid cycle intermediates. Proc. Natl. Acad. Sci. U. S. A. 97:2826–2831; 2000.

[39] Wei, Q.; Yin, X. M.; Wang, M. H.; Dong, Z. Bid deficiency ameliorates ischemicrenal failure and delays animal death in C57BL/6 mice. Am. J. Physiol. Renal Physiol.290:F35–F42; 2006.