free radical biology and medicine 2013 - rescue of...synovial ﬁbroblasts, and chondrocytes during...

Original Contribution

Rescue of proinflammatory cytokine-inhibited chondrogenesis by theantiarthritic effect of melatonin in synovium mesenchymal stem cellsvia suppression of reactive oxygen species andmatrix metalloproteinases

Xiaozhen Liu a,b,1, Yong Xu a,b,1, Sijin Chen c, Zifang Tan a, Ke Xiong c, Yan Li a,b, Yun Ye a,b,Zong-Ping Luo d,e, Fan He a,d,e,n, Yihong Gong a,b,nnQ1

a Department of Biomedical Engineering, School of Engineering, and Sun Yat-sen University, Guangzhou 510006, People’s Republic of Chinab Guangdong Provincial Key Laboratory of Sensor Technology and Biomedical Instruments, Sun Yat-sen University, Guangzhou 510006,People’s Republic of Chinac Nanfang Hospital, Southern Medical University, Guangzhou 510515, People’s Republic of Chinad Department of Orthopaedics, First Affiliated Hospital of Soochow University, Suzhou 215006, People’s Republic of Chinae Orthopaedic Institute, Soochow University, Suzhou 215007, People’s Republic of China

a r t i c l e i n f o

Article history:Received 6 August 2013Received in revised form4 December 2013Accepted 12 December 2013

Keywords:Mesenchymal stem cellsMelatoninChondrogenesisProinflammatory cytokinesSuperoxide dismutaseMatrix metalloproteinasesFree radicals

a b s t r a c t

Cartilage repair by mesenchymal stem cells (MSCs) often occurs in diseased joints in which the inflamedmicroenvironment impairs chondrogenic maturation and causes neocartilage degradation. In thisenvironment, melatonin exerts an antioxidant effect by scavenging free radicals. This study aimed toinvestigate the anti-inflammatory and chondroprotective effects of melatonin on human MSCs in aproinflammatory cytokine-induced arthritic environment. MSCs were induced toward chondrogenesis inthe presence of interleukin-1β (IL-1β) or tumor necrosis factor α (TNF-α) with or without melatonin.Levels of intracellular reactive oxygen species (ROS), hydrogen peroxide, antioxidant enzymes, and cellviability were then assessed. Deposition of glycosaminoglycans and collagens was also determined byhistological analysis. Gene expression of chondrogenic markers and matrix metalloproteinases (MMPs)was assessed by real-time polymerase chain reaction. In addition, the involvement of the melatoninreceptor and superoxide dismutase (SOD) in chondrogenesis was investigated using pharmacologicinhibitors. The results showed that melatonin significantly reduced ROS accumulation and increased SODexpression. Both IL-1β and TNF-α had an inhibitory effect on the chondrogenesis of MSCs, but melatoninsuccessfully restored the low expression of cartilage matrix and chondrogenic genes. Melatoninprevented cartilage degradation by downregulating MMPs. The addition of luzindole and SOD inhibitorsabrogated the protective effect of melatonin associated with increased levels of ROS and MMPs. Theseresults demonstrated that proinflammatory cytokines impair the chondrogenesis of MSCs, which wasrescued by melatonin treatment. This chondroprotective effect was potentially correlated to decreasedROS, preserved SOD, and suppressed levels of MMPs. Thus, melatonin provides a new strategy forpromoting cell-based cartilage regeneration in diseased or injured joints.

& 2013 Elsevier Inc. All rights reserved.

Articular cartilage is a highly specialized tissue that has verypoor self-repair capacity after injury or disease because of itsavascular structure. Approximately 1% of the world’s populationhas either of two forms of major arthritides, that is, osteoarthritis(OA) and rheumatoid arthritis (RA), which may result from trauma,genetic mutations, or damage to the meniscus and ligaments [1]. OAis characterized by the degradation and loss of articular cartilagefunction that might be caused by trauma when left untreated [2].RA is an autoimmune disease characterized by abnormal synovialproliferation and cartilage destruction by the body’s immunesystem [3]. Although the origins of these two diseases are different,they involve similar destructive proteinases and inflammatorymediators that mediate the pathological changes in the joints [4].

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

6768697071727374757677787980818283

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/freeradbiomed

Free Radical Biology and Medicine

0891-5849/$ - see front matter & 2013 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.freeradbiomed.2013.12.012

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; SMSC, synoviummesenchymal stem cell; ROS, reactive oxygen species; SOD, superoxide dismu-tase; GPx, glutathione peroxidase; MMP, matrix metalloproteinase; ECM, extra-cellular matrix; DDC, diethyldithiocarbamate; 2-MeOH2, 2-methoxyestradiol

n Corresponding author at: Soochow University, Orthopaedic Institute, No.708Renmin Road, Suzhou, Jiangsu 215007, China.Fax: þ86 512 67781165.

nnQ2 Corresponding author at: Guangdong Provincial Key Laboratory of SensorTechnology and Biomedical Instruments, Sun Yat-sen University, Guangzhou510006, People’s Republic of China. Fax:þ86 20 39332146.

E-mail addresses: [email protected], [email protected] (F. He),[email protected] (Y. Gong).

1 These authors contributed equally to this work.

Please cite this article as: Liu, X; et al. Rescue of proinflammatory cytokine-inhibited chondrogenesis by the antiarthritic effect ofmelatonin in synovium mesenchymal.... Free Radic. Biol. Med. (2013), http://dx.doi.org/10.1016/j.freeradbiomed.2013.12.012i

Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎

www.sciencedirect.com/science/journal/08915849

www.elsevier.com/locate/freeradbiomed

http://dx.doi.org/10.1016/j.freeradbiomed.2013.12.012



mailto:[email protected]







Proinflammatory cytokines, such as interleukin-1β (IL-1β) andtumor necrosis factor α (TNF-α), which are produced by macrophages,synovial fibroblasts, and chondrocytes during the pathogenesis of OAand RA, induce local inflammation in articular cartilage [5,6]. There isextensive literature that demonstrates elevated levels of IL-1β andTNF-α during the cartilage destruction cascade in synovial tissue andfluid [7]. TNF-α also induces cell apoptosis in chondrocytes, asevidenced by increased expression of caspase-8 [8]. Moreover, bothIL-1β and TNF-α inhibit cartilage matrix synthesis (e.g., aggrecan andtype II collagen), leading to further cartilage destruction [9].

The arthritic inflammatory response induces excessive generationof reactive oxygen species (ROS), which seem to be an importantmediator in the pathogenesis of joint diseases and aggravate thebreakdown of articular cartilage components. A significant increasein ROS is observed in the early stages of OA progression and isaccompanied by loss of cartilage and subchondral bone surface [10].NADPH oxidase, as a major source of ROS, produces superoxide anionsand free radicals to activate several signaling pathways such as themitogen-activated protein kinase pathway; this results in the over-expression of inflammatory cytokines [11]. Moreover, high levels ofROS that are found in inflamed joints not only induce chondrocyte celldeath and oxidative DNA damage [12] but are also responsible for thedestruction of cartilage extracellular matrix (ECM) through the activa-tion of matrix metalloproteinases (MMPs) [13].

Mesenchymal stem cells (MSCs) provide a feasible cell sourcefor cartilage repair because they can be harvested from varioustissues and can differentiate into multiple lineages, includingchondrocytes. Synovium mesenchymal stem cells (SMSCs) havebeen shown to have robust proliferative capacity and superiorchondrogenic potential, thus attracting increased interest in carti-lage tissue engineering and clinical utilization [14]. More impor-tantly, the phenotype profile and chondrogenic potential of SMSCsare not affected by age or OA etiology [15]. However, cartilageregeneration relying on the in situ chondrogenesis of MSCs, inmany instances, occurs in an inflammatory environment andhence is compromised by proinflammatory cytokines. AlthoughMSCs from adipose tissue express anti-inflammatory and immu-nosuppressive characteristics in collagenase-induced OA [16], highlevels of oxidative stress in an inflammatory environment remainchallenging for stem-cell-based cartilage repair, as evidenced bysuppressed differentiation by IL-1β and TNF-α via activation of theNF-κB pathway [17] and inhibited cartilage formation in thepresence of hydrogen peroxide [18]. Therefore, the arthriticmicroenvironment containing proinflammatory mediators causesfailure of cartilage repair. The development of suitable therapiesthat enhance antioxidant and antiarthritic characteristics of MSCswould promote cartilage regeneration in damaged joints.

Melatonin (N-acetyl-5-methoxytryptamine) is primarily synthe-sized in the pineal gland and is well known for its antioxidantproperties involving direct scavenging of free radicals [19]. Exogen-ous melatonin also exerts protective effects against oxidative stressin motoneurons by preserving superoxide dismutases (SODs) [20].The SOD enzyme system processes superoxide radicals to hydrogenperoxide, which is subsequently neutralized by catalase and glu-tathione peroxidases (GPx’s) [21]. Melatonin has been shown toprotect the nervous system by reducing ROS levels [22] andimproving the resistance of MSCs to hydrogen peroxide-inducedapoptosis [23]. In terms of its action on cell differentiation,melatonin has been proven to enhance ECM formation in primarychondrocytes in vitro through activation of the transforming growthfactor β (TGF-β) signaling pathway [24]. In the presence of IL-1β,melatonin improved osteogenic differentiation of MSCs by alleviat-ing the effects of intracellular ROS [25]. Animal studies have alsoestablished the promotive effect of melatonin on bone formationand its inhibitive effect on bone resorption [26]. More importantly,it has been shown that melatonin inhibits the expression of MMP-9

and MMP-3 in the animal model of acute gastric ulcers [27] andsuppresses the activation of MMPs in RAW264.7 macrophages byinhibiting NF-κB translocation [28]. However, limited information isavailable regarding the potential protective effect of melatonin onMSC chondrogenesis in diseased joint microenvironments.

The purpose of this study was to investigate whether melato-nin protects chondrogenic-specific differentiation of SMSCs fromthe deleterious influences of catabolic factors in an in vitroarthritic model. To create the in vitro arthritic cell model, tworepresentative proinflammatory cytokines, IL-1β and TNF-α, wereadded to cell culture to retard chondrogenesis. We hypothesizedthat melatonin would have an antioxidant effect and reverse theinhibitory effect of cytokines on chondrogenic differentiation. Wefocused on examining the production of intracellular ROS, activ-ities of antioxidant enzymes, chondrocyte-specific markers, andlevels of MMPs in the presence of melatonin, IL-1β, and TNF-α.A melatonin receptor antagonist and SOD inhibitors were used todetermine whether the melatonin-induced chondroprotectiveeffect could be blocked. Our long-term goal is to use melatoninto promote stem-cell-based cartilage regeneration, especially forpreventing and treating joint diseases such as OA and RA.

Materials and methods

Reagents

Synovium mesenchymal stem cells were purchased fromScienCell Research Laboratories (Carlsbad, CA, USA). Dexametha-sone, proline, L-ascorbic acid, melatonin, luzindole, dimethyl sulf-oxide (DMSO), diethyldithiocarbamate (DDC), 2-methoxyestradiol(2-MeOH2), and 20,70-dichlorofluorescein diacetate (DCF-DA) wereobtained from Sigma Chemical Co. (St. Louis, MO, USA). TGF-β1,IL-1β, and TNF-α were obtained from PeproTech Asia (Rehovot,Israel).

Cell culture and chondrogenesis of SMSCs

SMSCs were cultured in 12-well tissue culture plates (Corning,Tewksbury, MA, USA) in α-minimum essential medium supple-mented with 10% fetal bovine serum, penicillin (100 U/ml), strep-tomycin (100 μg/ml), and Fungizone (0.25 μg/ml) at 37 1C with 5%CO2. On reaching 95% confluence, the culture medium wasreplaced with serum-free chondrogenic differentiation medium.The medium contained high-glucose Dulbecco’s modified Eagle’smedium with 40 mg/ml proline, 100 nM dexamethasone, 100 U/mlpenicillin, 100 mg/ml streptomycin, 0.25 μg/ml Fungizone, 0.1 mML-ascorbic acid, ITS Premix (BD Biosciences, San Jose, CA, USA), and20 ng/ml TGF-β1. The medium was changed every 3 days and thecells were induced to differentiate for 14 days.

Proinflammatory cytokine-induced cell assay

SMSCs were induced to chondrogenesis and simultaneouslyexposed to 1 μM melatonin (MT), 10 ng/ml IL-1β, or 10 ng/ml TNF-α. Cells incubated only in chondrogenic differentiation mediumserved as a controls (CTRL). To assess the anti-inflammatory effectof melatonin, cells were treated with melatonin on coexposure toIL-1β (IL-1βþMT) or TNF-α (TNF-αþMT). To block the melatoninreceptor signaling pathway, cells were treated with 1 μM luzin-dole as a melatonin receptor antagonist (Luz) on coexposure tomelatonin (MTþLuz), IL-1β (IL-1βþMTþLuz), or TNF-α (TNF-αþMTþLuz). Cells incubated in 0.1% DMSO were included as avehicle control (Sigma).

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

X. Liu et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎2





Measurement of intracellular ROS accumulation

Intracellular ROS production was quantified by a DCF-DAfluorescence method as previously described [29]. Briefly, DCF-DA was dissolved in methanol and then diluted with phosphate-buffered saline to obtain a final concentration of 10 μM. Cells(n¼4; 2�105) were incubated in DCF-DA for 10 min at 37 1C, andthe fluorescence intensity of DCF, which was activated by intra-cellular ROS, was measured by a BD dual laser FACSCalibur (BDBiosciences). Ten thousand events of each sample were collected.The mean fluorescence intensity of DCF was calculated using theWinMDI 2.9 software (Windows Multiple Document Interface forFlow Cytometry).

Measurement of hydrogen peroxide accumulation

The accumulation of hydrogen peroxide in cells was detectedwith an Amplex Red Hydrogen Peroxide Assay Kit (Invitrogen,Carlsbad, CA, USA) according to the manufacturer0s instructions.Cellular extracts (six-well plates; n¼3) were collected and incu-bated with Amplex red reagent for 30 min. Fluorescence wasthen measured with a fluorescence microplate reader (BioTek,Winooski, VT, USA) using excitation at 530 nm and emissiondetection at 590 nm.

Superoxide dismutase activity assay

SOD activity was determined by using a commercially availableSOD assay kit (Sigma) according to the manufacturer0s instruc-tions. Cells (six-well plates; n¼3) were trypsinized and suspendedin cell lysis solution. Total sample proteins were quantified using aBCA Protein Assay Kit (Invitrogen). The supernatant was mixedwith WST solution and incubated at 37 1C for 20 min. Theabsorbance at 450 nm was measured using a microplate reader(BioTek).

Catalase activity assay

Catalase activity was measured with a commercially availablecatalase assay kit (Sigma). Briefly, total proteins were extractedfrom cells (six-well plate; n¼3) and quantified with the BCAProtein Assay Kit. H2O2 (20 mM) was used as the substratesolution and the amount of H2O2 consumed by each sample wasmeasured with a microplate reader by reading the absorbance at520 nm. The activity of catalase was then calculated according tothe manufacturer0s instructions.

Cell viability assay

Cells in 12-well plates (n¼3) were collected and digested at60 1C for 4 h with 125 μg/ml papain in 100 mM phosphate, 10 mMEDTA, pH 6.5, buffer containing 10 mM L-cysteine. The amount ofDNA content was measured with a Quant-iT PicoGreen dsDNAAssay Kit (Invitrogen) with a SynergyMx multimode reader(BioTek) according to the manufacturer’s instructions. The DNAcontent of each group was normalized to that in the control group.

Antioxidant inhibitor assay

Cells were pretreated with antioxidant inhibitors before incu-bation with melatonin or proinflammatory cytokines as previouslydescribed [30]. Cells were treated with 50 μM DDC (CuZn-SODinhibitor) for 1.5 h or 1 μM 2-MeOH2 (Mn-SOD inhibitor) for 24 h.

Alcian blue staining and Sirius red staining

Cells were fixed in 4% paraformaldehyde overnight. Alcian blue(1%; pH 2.5; Sigma) staining was used to detect sulfated glycosa-minoglycans, and Sirius red (1 mg/ml; Sigma) was used to staincollagens. Images were captured by an Olympus IX71 microscope(Olympus, Tokyo, Japan) and processed by Image-ProPlus software(Media Cybernetics, Rockville, MD, USA).

Total RNA extraction and quantitative real-time reversetranscription–polymerase chain reaction (real-time RT-PCR)

Total RNA was extracted from cells (n¼3) on days 7 and 14 usingTRIzol reagent (Invitrogen). For each sample, 1 μg of total RNA wasreverse transcribed with a PrimeScript RT Reagent Kit (TaKaRa,Mountain View, CA, USA). To quantify the mRNA, cDNA equivalentto 20 ng of total RNA was detected by real-time PCR analysis using aGoTaq qPCR Master Mix kit (Promega, Madison, WI, USA). Chondro-genic markers including aggrecan, type II collagen (COL II), cartilageoligomeric matrix protein (COMP), and Sox-9; antioxidant enzymegenes (CuZn-SOD, Mn-SOD, and catalase); and matrix metallopro-teinase genes (MMP-1, MMP-2, MMP-9, and MMP-13) were detected.Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used asan internal standard. The primer sequences are listed in Table 1. Real-time PCR was performed by an ABI7500 real-time PCR detectionsystem (Applied Biosystems, Foster City, CA, USA) following themanufacturer’s protocol. Relative transcript levels were calculatedas χ¼2�ΔΔCt, inwhichΔΔCt¼ΔE � ΔC,ΔE¼Ct exp � Ct GAPDH, andΔC¼Ct ctr1 � Ct GAPDH.

Statistical analysis

All data are expressed as the mean7standard error (SE).Statistical differences between two groups were determined byone-way analysis of variance followed by Student’s unpaired t testwith SPSS software (SPSS, Inc., Chicago, IL, USA). P values less than0.05 were considered statistically significant.

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Table 1Primers used for real-time RT-PCR.

Gene Primer sequence(50–30) GenBank accession number

GAPDH F: AGAAAAACCTGCCAAATATGATGAC NM_002046R: TGGGTGTCGCTGTTGAAGTC

Antioxidant enzyme-related genesCuZn-SOD F: GGTGGGCCAAAGGATGAAGAG NM_000454.4

R: CCACAAGCCAAACGACTTCCMn-SOD F: GGGGATTGATGTGTGGGAGCACG BC012423.1

R: AGACAGGACGTTATCTTGCTGGGACatalase F: TGGGATCTCGTTGGAAATAACAC NM_001752.3

R: TCAGGACGTAGGCTCCAGAAGChondrogenesis-related genesAggrecan F: TGCATTCCACGAAGCTAACCTT NM_001135.3

R: GACGCCTCGCCTTCTTGAACOL II F: TGGACGATCAGGCGAAACC NM_001844.4

R: GCTGCGGATGCTCTCAATCTCOMP F: AAGAACGACGACCAAAAGGAC NM_000095.2

R: CATCCCCTATACCATCGCCASox-9 F: AGCGAACGCACATCAAGAC NM_000346.3

R: CTGTAGGCGATCTGTTGGGMatrix metalloproteinase-related genesMMP-1 F: GGGGCTTTGATGTACCCTAGC NM_001166229

R: TGTCACACGCTTTTGGGGTTTMMP-2 F: GATACCCCTTTGACGGTAAGGA NM_001127891.1

R: CCTTCTCCCAAGGTCCATAGCMMP-9 F: AGACCTGGGCAGATTCCAAAC NM_004994.2

R: CGGCAAGTCTTCCGAGTAGTMMP-13 F: CCAGACTTCACGATGGCATTG NM_002427.3

R: GGCATCTCCTCCATAATTTGGC

X. Liu et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3





123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

TNF-αTNF-α+MTIL-1βIL-1β+MTMT

CTRL

Fig. 1. Effects of melatonin and proinflammatory cytokines on the accumulation of ROS and hydrogen peroxide. (A) Intracellular ROS production in SMSCs under thetreatments with IL-1β and TNF-αwithin 24 h. (B) ROS generation when cells were treated with melatonin (MT), IL-1β, IL-1β and melatonin (IL-1βþMT), TNF-α, and TNF-α andmelatonin (TNF-αþMT). Serum-free chondrogenic medium served as a control group (CTRL). Cells were labeled with DCF-DA and measured with flow cytometry (top). Graylines represent the treatments with control, IL-1β, and TNF-α. Red lines represent the treatments with melatonin, IL-1β and melatonin, and TNF-α and melatonin. The meanfluorescence intensity was quantified (bottom) to represent ROS production. (C) Cellular extracts from SMSCs were labeled with Amplex red probe and the fluorescence wasmeasured to represent hydrogen peroxide production. ROS results are representative of four independent experiments. Hydrogen peroxide results are representative of threeindependent experiments. Values are the mean7SE. *Po0.05; #Po0.05 versus CTRL.






Results

Intracellular ROS and hydrogen peroxide in response to melatoninand proinflammatory cytokines

We first examined the effects of proinflammatory cytokines on ROSproduction in SMSCs. DCF fluorescence detected by flow cytometryshowed that the levels of ROS significantly increased after treatmentwith catabolic factors (Fig. 1A). Treatment with IL-1β (10 ng/ml)induced the highest level of ROS at 15 min; this effect decreased by30min and recovered to the level in the control group by 24 h.However, SMSCs exposed to TNF-α (10 ng/ml) showed a significantincrease in ROS accumulation in a short time, peaked within 30 min,continued to decrease, and reached a low level within 24 h.

To determine whether exogenous melatonin regulated the pro-duction of ROS in the presence of IL-1β or TNF-α, SMSCs were treatedwith melatonin at a concentration of 1 μM (Fig. 1B) in the presence ofIL-1β or TNF-α. The addition of melatonin alone decreased the levelof ROS by 26.6% compared to that in control cells. Similarly,incubation with melatonin attenuated oxidative stress in SMSCs by14.5 and 30.3% on coexposure to IL-1β and TNF-α, respectively. Theaccumulation of hydrogen superoxide was measured with an Amplexred probe. As shown in Fig. 1C, on supplementation with melatonin,the intracellular hydrogen peroxide was reduced by 13.5% comparedto that in the control group, without significance (P¼0.08). In thepresence of IL-1β, melatonin treatment decreased the hydrogenperoxide level by 13.3% compared to that in the IL-1β group.However, melatonin treatment did not affect hydrogen peroxideaccumulation in SMSCs on exposure to TNF-α (P¼0.48).

Effects of melatonin and proinflammatory cytokines on antioxidantenzymes

We investigated the effects of melatonin on SOD mRNA expressionand activity. As illustrated in Fig. 2A, treatment with melatonin

increased SOD activity by 35.0 and 48.7% in contrast to that in thecontrol and IL-1β groups, respectively. However, it did not affect SODactivity in TNF-α-treated SMSCs. When SMSCs were exposed to IL-1βand TNF-α, the mRNA expression of CuZn-SOD in SMSCs decreased by44.4 and 25.8%, respectively, compared to that in the control group.However, melatonin treatment significantly increased CuZn-SODexpression (by 46.0% vs the CTRL group, by 1.5-fold vs the IL-1βgroup, and by 1.1-fold vs the TNF-α group; Fig. 2B). We found thatproinflammatory cytokine treatments increased the level of Mn-SOD,compared to that in the control cells (by 2.2-fold in the IL-1β groupand 11.0% in the TNF-α group). Cotreatment with melatonin alsoincreased gene expression of Mn-SOD (by 2.0-fold vs the CTRL group,by 47.9% vs the IL-1β group, and by 86.5% vs the TNF-α group; Fig. 2C).

Catalase is one of the major antioxidant enzymes that sca-venges hydrogen peroxide; therefore, we further examined theeffect of melatonin on catalase activity and mRNA expression. Theactivity of catalase slightly increased with melatonin treatment(P¼0.11), significantly decreased by 28.1% when exposed to IL-1β,and decreased by 13.0% on exposure to TNF-α. However, melato-nin supplementation restored catalase activity after coexposure toproinflammatory cytokines (by 38.5% vs the IL-1β group and by10.2% vs the TNF-α group; Fig. 2E). The gene expression of catalasewas downregulated by treatments with proinflammatory cyto-kines (by 15.6% in the IL-1β group and by 17.7% in the TNF-αgroup), but melatonin supplementation successfully increasedcatalase levels by 18.0 and 19.1% compared to that in the IL-1βgroup and the TNF-α group, respectively (Fig. 2F).

Melatonin reverses proinflammatory cytokine inhibitionof chondrogenesis

IL-1β had no effect on cell viability during SMSC differentiation intoa chondrogenic lineage, whereas TNF-α significantly decreased cellviability by 38.7% compared to that in the control group. Melatonintreatment alone decreased DNA content by 22.2%. However, the

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 2. The influence of melatonin and inflammatory cytokines on antioxidant enzymes. (A) SOD activity of SMSCs was measured with or without melatonin treatment oncoexposure to 10 ng/ml IL-1β or 10 ng/ml TNF-α. (B) Gene expression of CuZn-SOD was determined by real-time RT-PCR. (C) Gene expression of Mn-SOD was determined byreal-time RT-PCR. (D) Catalase activity of SMSCs was measured with or without melatonin treatment on coexposure to 10 ng/ml IL-1β or 10 ng/ml TNF-α. (E) Gene expressionof catalase was determined by real-time RT-PCR. Antioxidant protein activities and real-time RT-PCR data are representative of three independent experiments. Values arethe mean7SE. *Po0.05; #Po0.05 versus CTRL.






123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

CTRL IL-1β TNF-α

MT IL-1β+MT TNF-α+MT

CTRL IL-1β

MT IL-1β+MT TNF-α+MT

TNF-α

Fig. 3. Chondrogenesis of SMSCs on exposure to melatonin and proinflammatory cytokines. SMSCs were induced to chondrogenic differentiation with or without melatonintreatment on coexposure to 10 ng/ml IL-1β or 10 ng/ml TNF-α. (A) Cell viability of SMSCs was determined by PicoGreen dsDNA assay and the DNA content was normalized tothe level of control group. (B) Synthesis of glycosaminoglycans in SMSCs was assessed by Alcian blue staining. (C) Synthesis of collagens was assessed by Sirius red staining.DNA results are representative of three independent experiments. Values are the mean7SE. Scale bar, 100 μm in (B) and (C). *Po0.05; #Po0.05 versus CTRL.






addition of melatonin reversed the TNF-α-induced inhibitory effect oncell viability by 47.7% compared to that in the TNF-α group (Fig. 3A).

To evaluate cartilage-specific ECM production, sulfated glycosa-minoglycans (Fig. 3B) and collagens (Fig. 3C) were detected by Alcianblue staining and Sirius red staining, respectively. When differen-tiated SMSCs were stimulated with IL-1β or TNF-α, they showed asignificantly diminished deposition of glycosaminoglycans and col-lagens, but melatonin treatment induced a glycosaminoglycan-richand collagen-rich matrix surrounding the differentiated cells, indi-cating that melatonin rescued the formation of cartilage ECM, whichwas inhibited by proinflammatory cytokines.

We next examined mRNA expression of chondrocyte-specificmarker genes, such as COL II, aggrecan, COMP, and Sox-9, on days7 and 14 by quantitative RT-PCR. After treatment with IL-1β or TNF-α, gene expression of aggrecan significantly decreased by 95.1 and74.2%, respectively, compared to that in the control group (Fig. 4A);gene expression was remarkably restored by melatonin treatment(7.6-fold vs the IL-1β group and 2.7-fold vs the TNF-α group). Thetrends for COL II (Fig. 4B) and COMP (Fig. 4C) mRNA expression weresimilar to those for aggrecan gene expression. Sox-9 is an essentialtranscription factor in the early stage of chondrogenic differentiation.Treatment with IL-1β or TNF-α inhibited the level of Sox-9 (by 21.3and 61.0%, respectively, vs the CTRL group), but melatonin treatmentupregulated Sox-9 gene expression by 87.4% in the presence of TNF-α(Fig. 4D). The data on the expression of aggrecan, type II collagen,COMP, and Sox-9 on day 7 also indicated that melatonin promotedchondrogenesis (Supplementary Figs. 1A–1D).

Melatonin suppresses the expression of MMPs

To assess the effect of melatonin on cartilage ECM degradation,gene expression of MMP-1, MMP-2, MMP-9, and MMP-13 was

examined. MMP-1 is a critical proteinase that targets nativecollagens in articular cartilage. We found that IL-1β treatmentmarkedly increased the expression of MMP-1 by 18.7-fold com-pared to that in the control group. The addition of melatoninsignificantly decreased MMP-1 expression by 20.2% in the pre-sence of IL-1β (Fig. 5A). MMP-2 expression was significantlyupregulated by both IL-1β and TNF-α treatments (by 65.4 and15.6%, respectively, vs the CTRL group), but was downregulated inthe presence of melatonin (by 25.9% vs the IL-1β group and 34.1%vs the TNF-α group; Fig. 5B). However, neither IL-1β treatment norTNF-α treatment affected MMP-9 expression, whereas melatonintreatment reduced MMP-9 by 56.3% compared to TNF-α treatment(Fig. 5C). Melatonin treatment significantly reduced MMP-13expression by 37.4% compared to that in the CTRL group. The levelof MMP-13 was also significantly increased by treatment withIL-1β or TNF-α (by 45.4 and 32.3%, respectively, vs the CTRL group),but was downregulated by melatonin treatment (by 42.8% vs theIL-1β group and 59.7% vs the TNF-α group; Fig. 5D). The trend in theexpression of MMP genes on day 7 was the same as that obtainedon day 14 (Supplementary Figs. 2A–2D).

The melatonin receptor antagonist blocks antioxidantand chondroprotective effects

To confirm that the effect of melatonin on chondrogenesis andMMPs was mediated through a melatonin receptor pathway,luzindole, a specific melatonin receptor antagonist, was added tothe chondrogenic differentiation medium. Collagen synthesis wassuppressed by the addition of luzindole in the presence of IL-1β orTNF-α, as evidenced by Sirius red staining (Fig. 6A). QuantitativeRT-PCR showed that treatment with luzindole significantly blocked

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 4. Melatonin upregulated mRNA expression of chondrogenic-specific marker genes, including (A) aggrecan, (B) type II collagen, (C) COMP, and (D) Sox-9. Real-timeRT-PCR data are representative of three independent experiments. Values are the mean7SE. *Po0.05; #Po0.05 versus CTRL.






the effect of melatonin on gene expression of aggrecan by 88.1 and19.2% in the presence of IL-1β and TNF-α, respectively (Fig. 6B).The levels of COMP (Fig. 6C) and Sox-9 (Fig. 6D) also showed the

same tendency. In addition, the expression of MMP-1, MMP-2,and MMP-13 was significantly upregulated by luzindole treat-ment (Fig. 7A–C). Furthermore, intracellular ROS suppression by

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 5. Melatonin downregulated mRNA expression of MMP genes, including (A) MMP-1, (B) MMP-2, (C) MMP-9, and (D) MMP-13. Real-time RT-PCR data are representativeof three independent experiments. Values are the mean7SE. *Po0.05; #Po0.05 versus CTRL.

DMSO Luz MT MT+Luz

TNF-α+MTIL-1β+MT+LuzIL-1β+MT TNF-α+MT+Luz

Fig. 6. Luzindole as a melatonin receptor antagonist reversed the chondroprotective effect of melatonin in SMSCs in the presence of inflammatory cytokines. SMSCs induced tochondrogenic differentiation were treated with 1 μM melatonin, 10 ng/ml IL-1β, 10 ng/ml TNF-α, or 1 μM luzindole. (A) Synthesis of collagens was assessed by Sirius red staining.(B–D) Gene expression of chondrogenic genes in SMSCs under various treatment conditions was determined by real-time RT-PCR, including (B) aggrecan, (C) COMP, and (D) Sox-9.Scale bar, 100 μm in (A). Real-time RT-PCR data are representative of three independent experiments. Values are the mean7SE. *Po0.05; #Po0.05 versus CTRL.






melatonin treatment was offset by the treatment with luzindolewhen cells were coexposed to IL-1β or TNF-α (Fig. 7D–F).

SOD inhibitors increase ROS levels and suppress chondrogenesis

To verify the roles of CuZn-SOD and Mn-SOD in melatonin-restored chondrogenesis, two SOD inhibitors, DDC and 2-MeOH2,were used to treat cells that were concurrently treated with proin-flammatory cytokines and melatonin. Both inhibitors reversed theantioxidant effect of melatonin by increasing intracellular ROS(Fig. 8A and B). In particular, we found that 2-MeOH2 was morepotent at upregulating oxidative stress. Despite the fact that melato-nin significantly enhanced chondrogenic differentiation, inhibitors ofDDC and 2-MeOH2 downregulated the expression of chondrocyte-specific genes, such as aggrecan (Fig. 8C and D) and Sox-9 (Fig. 8Eand F), indicating that high ROS levels suppressed chondrogeniclineage maturation of SMSCs. Treatment with SOD inhibitors alsoupregulated the mRNA levels of MMP-1 (Fig. 9A and B) and MMP-2(Fig. 9C and D), suggesting that depleting SOD contributed to MMP-dependent cartilage ECM destruction.

Discussion

In many cases, application of stem-cell-based cartilage regen-eration takes place within an inflammatory environment, whichprevents chondrogenic maturation and accelerates cartilage break-down. Release of proinflammatory cytokines induces elevation ofoxidative stress, which is considered to cause chondrocyte senes-cence and contribute to cartilage pathogenesis [31]. Althoughmelatonin has been suggested to protect against free radicalspecies in various cell types and suppress the production ofproinflammatory mediators in lipopolysaccharide-stimulated vas-cular smooth muscle cells [32], the potential anti-inflammatoryeffect on chondrogenesis of MSCs in an arthritic environment was

unknown. To our knowledge, this study is the first to show thatmelatonin rescues impaired human SMSC chondrogenesis inducedby IL-1β and TNF-α. Moreover, the molecular mechanism bywhich this occurs involves the alleviation of intracellular ROS viapreservation of SOD and suppression of MMPs by melatonintreatment. Our findings provide potential applications for melato-nin in the treatment of joint diseases and promotion of cartilagerepair.

An increasing amount of evidence implicates IL-1β and TNF-αas important proinflammatory mediators in joint diseases. Thesecytokines play a vital role not only in modulating subchondralbone resorption but also in shifting cartilage homeostasis towardcatabolism. As shown by our data, both treatments with IL-1β andTNF-α caused significant reduction in cartilage-specific matrixsynthesis and downregulation of chondrogenic marker genes.More interestingly, IL-1β was found to have a more potent effectthan TNF-α on the inhibition of chondrogenesis, in agreementwith a recent report [33]. During chondrogenesis, these cytokinesinitiate the activation and nuclear translocation of NF-κB, whichthen inhibits phosphorylation of Smad3/4, a key signaling med-iator in response to TGF-β [34]. Moreover, our data showed thattreatment with IL-1β and TNF-α suppressed Sox-9, an essentialtranscription factor in early chondrogenesis that is also respon-sible for subsequent inhibition of chondrogenic maturation. There-fore, the inflamed environment present in diseased joints mightresult in failure of cartilage defect repair. This suggests thatblocking the catabolic effect of proinflammatory factors wouldpromote the effectiveness of stem-cell-based cartilage regenera-tion strategies.

The effects of proinflammatory cytokines and melatonin on cellviability were examined. Our study suggested that IL-1β did notaffect cellular viability, whereas TNF-α remarkably inhibited it, asassessed by DNA assays. Consistent with our data, previous studiesdemonstrated that TNF-α induced cell apoptosis via a caspase3- and caspase 8-associated cascade [8]. Melatonin protected MSCs

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 7. Luzindole reversed the antioxidant effect of melatonin in the presence of inflammatory cytokines. (A–C) Expression of MMP genes in SMSCs under various treatmentconditions was determined by real-time RT-PCR, including (A) MMP-1, (B) MMP-2, and (C) MMP-13. (D–F) Intracellular ROS production in SMSCs under various treatmentconditions. Real-time RT-PCR data are representative of three independent experiments. ROS results are representative of four independent experiments. Values are themean7SE. *Po0.05; #Po0.05 versus CTRL.






from cytokines by adjusting the balance between proapoptotic andantiapoptotic factors. Kim et al. [35] showed that melatoninexhibited a protective effect on myoblast cells against NO-induced cell death with increase in Bcl-2 expression and inhibitionof Bax expression. In addition, in this study, we incubated MSCs ina serum-free chondrogenic medium to minimize any serum-induced interference because autologous conditioned serum hasbeen demonstrated to be a potential anti-inflammatory agent fortreating arthritis and muscle injuries [36].

Evidence of increased generation of ROS and subsequentoxidative stress has been reported in pathological cartilage fromOA and RA patients. Proinflammatory cytokines in diseased jointsupregulated oxidative stress reactions and led to chondrocyteapoptosis through the caspase signaling pathway [37]. Eliminationof superoxide radicals is significantly dependent on the expression

of SOD under normal physiological conditions; however, Scottet al. [38] evaluated three SOD genes (CuZn-SOD, Mn-SOD, and EC-SOD) in an OA animal model and reported that OA progressionresulted in significant reduction in the levels of SOD. In the presentstudy, we observed that oxidative stress in SMSCs was elevated byproinflammatory cytokine treatments, but melatonin provided achondroprotective effect by preventing oxidative damage andupregulating the expression of CuZn-SOD and Mn-SOD. The useof luzindole, a melatonin receptor antagonist, significantly coun-teracted the antioxidant effect of melatonin, indicated by theincreased level of ROS in SMSCs. Moreover, hydrogen peroxide,as a product of scavenging of superoxide radicals by the SODenzyme systems, functioned as an important signaling messengerto affect cellular survival and growth. Kim et al. reported [39] thatshort-term treatment with hydrogen peroxide at a concentration

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 8. Treatment with SOD inhibitors reversed the promotive effect of melatonin in SMSC chondrogenesis in the presence of inflammatory cytokines. SMSCs were pretreatedwith 50 μM DDC for 1.5 h or 1 μM 2-methoxyestradiol (2-ME) for 24 h. (A) Intracellular ROS production in SMSCs in the presence of DDC as a CuZn-SOD inhibitor.(B) Intracellular ROS production in SMSCs in the presence of 2-ME as a Mn-SOD inhibitor. (C and D) Gene expression of aggrecan in SMSCs under treatment with DDC (C) or2-ME (D). (E and F) Gene expression of Sox-9 in SMSCs under treatment with DDC (E) or 2-ME (F). Real-time RT-PCR data are representative of three independentexperiments. Values are the mean7SE. *Po0.05; #Po0.05 versus CTRL; ψPo0.05 versus the IL-1β group; ζPo0.05 versus the TNF-α group.






of 100 μM stimulated cell proliferation in fibroblasts throughactivation of Jun-N-terminal kinase and p38 mitogen-activatedprotein kinases. In contrast, long-term incubation with hydrogenperoxide at a concentration even as low as 50 μM resulted in cellgrowth arrest [40]. Melatonin offered protection against cellularoxidative stress injury, particularly that induced by hydrogenperoxide, depending on the inhibition of JAK2/STAT3 phosphor-ylation [41].

Although intracellular ROS generated by NADPH oxidases wasnecessary for chondrogenic differentiation [42], the inhibitoryeffect of external oxidative stress on chondrogenesis of stem cellshas been demonstrated in numerous studies. By creating oxidativestress with hydrogen peroxide, Pei et al. [40] showed thatchondrogenic maturation of human adult stem cells was signifi-cantly suppressed and that the level of type X collagen, a typicalhypertrophy marker, increased. Our data also suggest that highlevels of intracellular ROS cause suppressed chondrogenic differ-entiation, but that melatonin treatment successfully restored theinhibitory effect owing to its antioxidant properties. The decreasedexpression of chondrogenic marker genes in the presence of SODinhibitors confirms that the expression of superoxide radicals is animportant determinant for the poor outcome of MSC chondrogen-esis in OA or RA. The high level of cellular hydrogen peroxide wasresponsible for the senescence and diminished osteogenic poten-tial of adult MSCs because of its effect on c-MAF and PKC-p21activities [43]. The anti-inflammatory effect of melatonin waspossibly related to ERK and p38 signaling pathways because thephosphorylation of these kinases was suppressed in the presenceof melatonin [44]. However, the underlying molecular mechanismthat regulates this process needs to be elucidated in future studies.

To determine whether SOD participates in the anti-inflammatoryeffects of melatonin, we analyzed the levels of CuZn-SOD andMn-SOD and found that they decreased after treatment with IL-1βor TNF-α, suggesting a relationship between suppression ofchondrogenic differentiation and reduction in SOD expression.Gavriilidis et al. [45] demonstrated that Mn-SOD depletion wasdetected in OA cartilage chondrocytes and caused mitochondrialdysfunction and arthritic pathogenesis. In addition, we observedthat expression of SOD was reversed by melatonin treatment,possibly depending on the mitochondrial cytochrome P450 path-way [46]. SMSCs were treated with DDC and 2-MeOH2 to block theantioxidant functions of SOD. The results suggested not only thatchondrogenesis was impaired but also that the levels of MMPsincreased. Therefore, the reversal of proinflammatory cytokine-induced inhibition of chondrogenic differentiation by melatoninwas correlated with low oxidative stress through the SOD anti-oxidant enzyme systems. Moreover, the homeostasis of ROS is alsodependent on the activation of hydrogen peroxide removalenzymes such as catalase and glutathione peroxidases, of whichGPx1 is involved in the insulin signaling pathway and GPx2 playsan important role in carcinogenesis [47]. Jallali et al. [48] demon-strated that, in aging chondrocytes, impaired catalase activityresulted in elevation of hydrogen peroxide levels and substantialcell death. The loss of antioxidative capacity in bone marrow MSCswas ascribed to the low level of selenite in standard cell cultures.Supplementation with selenite restored GPx activity and rescuedcell survival by reducing hydrogen peroxide accumulation [49].We speculate that melatonin reduces the increase in free radicalsinitiated by inflammatory cytokines, subsequently increasing theproduction of hydrogen peroxide, which is then eliminated by

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 9. Treatment with SOD inhibitors upregulated mRNA expression of MMPs in the presence of inflammatory cytokines and melatonin. (A and B) Gene expression of MMP-1 in SMSCs under treatment with DDC (A) and 2-MeOH2 (B). (C and D) Gene expression of MMP-2 in SMSCs under treatment with DDC (C) and 2-MeOH2 (D). Real-time RT-PCR data are representative of three independent experiments. Values are the mean7SE. *Po0.05; #Po0.05 versus CTRL; ψPo0.05 versus the IL-1β group; ζPo0.05 versusthe TNF-α group.






catalase and GPx’s. This maintains the balance of ROS in MSCs;however, the role of GPx’s in melatonin-mediated cytoprotectionand enhanced chondrogenesis will be analyzed in future studies.

Evidence has been provided that degradation of cartilagematrix in joint diseases involves MMPs that include severalsubtypes (e.g., collagenases, gelatinases, stromelysins, metrilysins)depending on their specific substrates. MMP-1 (collagenase-1) andMMP-13 (collagenase-3) are rarely expressed in healthy jointtissues and are important mediators of cartilage destruction inOA and RA. They uniquely target native type II collagen, whichcompromises joint tensile strength and makes collagen chainsmuch more susceptible to further degradation by other MMPs [50].MMP-13 is also considered to be a chondrocyte hypertrophymarker; its expression is accompanied by that of denatured typeII collagen during the development of cartilage pathology [51].With cytokine treatment, the expression of MMP-1 and MMP-13markedly increased but was successfully inhibited by melatonin.In addition, MMP-2 (gelatinase A) and MMP-9 (gelatinase B)recently attributed more interest in OA progression, because theytarget fully or partially denatured type II collagen other than gelatin[52]. Our data confirmed that the mRNA expression of MMP-2 wasupregulated by exposure to proinflammatory factors; however,MMP-9 showed no significant difference in the presence of IL-1β,possibly because MMP-2 was present at a much higher level andwas more deleterious than MMP-9 in OA-diseased joints [53].

The inhibitory effect of melatonin on the activation of MMPshas been demonstrated in RAW 264.7 and BV2 cells. Activation ofNF-κB is required to induce the expression of MMPs, and melato-nin effectively inhibits NF-κB translocation and binding activity tosuppress MMP levels [28]. Our data suggest that melatoninsignificantly suppressed the levels of MMPs, including MMP-1,MMP-2, and MMP-13, accompanied by restoration of MSC chon-drogenesis. Although the relationship between oxidative stressand MMPs is unclear in OA and RA, Martinez-Lemus et al. [54]suggested that, in cremaster arterioles, activation of MMPs wasdependent on the production of ROS, whereas inhibition of ROSprevented the activity of MMPs. In this study, the elevatedproduction of MMPs involved a high level of ROS as an obligatorysecondary messenger [55]. Thus, scavenging of free radicals bymelatonin possibly contributed to the decrease in MMP levels andthe delay of cartilage degradation. Collectively, with the stronginhibition of MMPs, use of melatonin could be a promisingtherapeutic strategy to protect neocartilage undergoing repairfrom proteolytic enzymes.

In summary, we demonstrated that melatonin reversed theinhibitory effects of IL-1β and TNF-α on SMSC chondrogenesis.Interestingly, the chondroprotective effect of melatonin was poten-tially related to the decreased level of oxidative stress, the preservedexpression of SOD, and the suppressed levels of cytokine-inducedMMPs in restraining the degradation of cartilage-specific ECM.Our results suggest that therapeutic strategies that use melatoninmay be effective in treating joint disease and promoting MSC-basedcartilage regeneration.

Authors’ contributions

All authors were involved in drafting the article or revising itcritically for important intellectual content, and all authorsapproved the final version to be published. Dr. He and Dr. Gonghave full access to all of the data in the study and take responsi-bility for the integrity of the data and the accuracy of the dataanalysis. Study conception and design were done by F.H. and Y.G.Acquisition of data was done by X.L., Y.X., S.C., Y.Y., K.X., Z.T., F.H.,and Y.G. Analysis and interpretation of data were done by X.L., Y.X.,

S.C., Y.Y., Y.L., Z.T., F.H., and Y.G. Manuscript writing and revisingwere done by X.L., Y.X., Z.-P.L., F.H., and Y.G.

Acknowledgments

This project was supported by the National Natural ScienceFoundation of China (Nos. 51203194, 51103182, and 11072165);Specialized Research Fund for the Doctoral Program of HigherEducation (No. 20120171120034); Scientific Research Foundationfor the Returned Overseas Chinese Scholars, State Education Ministry,Guangdong Natural Science Foundation (No. S2012040007933);Foundation for Distinguished Young Talents in Higher Education ofGuangdong (No. 2012LYM_0002); and The Open Foundation ofLaboratory Sun Yat-sen University (No. KF201210).

Appendix A. Supplementary Information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.freeradbiomed.2013.12.012.

References

[1] Sangha, O. Epidemiology of rheumatic diseases. Rheumatology (Oxford) 39(Suppl. 2):3–12; 2000.

[2] Dieppe, P. Therapeutic targets in osteoarthritis. J. Rheumatol. Suppl. 43:136–-139; 1995.

[3] Brennan, F. M.; McInnes, I. B. Evidence that cytokines play a role in rheumatoidarthritis. J. Clin. Invest. 118:3537–3545; 2008.

[4] Loeser, R. F. Molecular mechanisms of cartilage destruction: mechanics,inflammatory mediators, and aging collide. Arthritis Rheum. 54:1357–1360;2006.

[5] Tyler, J. A. Articular cartilage cultured with catabolin (pig interleukin 1)synthesizes a decreased number of normal proteoglycan molecules. Biochem.J. 227:869–878; 1985.

[6] Saklatvala, J. Tumour necrosis factor alpha stimulates resorption and inhibitssynthesis of proteoglycan in cartilage. Nature 322:547–549; 1986.

[7] Kubota, E.; Imamura, H.; Kubota, T.; Shibata, T.; Murakami, K. Interleukin1 beta and stromelysin (MMP3) activity of synovial fluid as possible markersof osteoarthritis in the temporomandibular joint. J. Oral Maxillofac. Surg.55:20–27; 1997.

[8] Lopez-Armada, M. J.; Carames, B.; Lires-Dean, M.; Cillero-Pastor, B.; Ruiz-Romero, C.; Galdo, F.; Blanco, F. J. Cytokines, tumor necrosis factor-alpha andinterleukin-1beta, differentially regulate apoptosis in osteoarthritis culturedhuman chondrocytes. Osteoarthritis Cartilage 14:660–669; 2006.

[9] Murakami, S.; Lefebvre, V.; de Crombrugghe, B. Potent inhibition of the masterchondrogenic factor Sox9 gene by interleukin-1 and tumor necrosis factor-alpha. J. Biol. Chem. 275:3687–3692; 2000.

[10] Xie, L.; Lin, A. S.; Kundu, K.; Levenston, M. E.; Murthy, N.; Guldberg, R. E.Quantitative imaging of cartilage and bone morphology, reactive oxygenspecies, and vascularization in a rodent model of osteoarthritis. ArthritisRheum. 64:1899–1908; 2012.

[11] Chenevier-Gobeaux, C.; Lemarechal, H.; Bonnefont-Rousselot, D.; Poiraudeau,S.; Ekindjian, O. G.; Borderie, D. Superoxide production and NADPH oxidaseexpression in human rheumatoid synovial cells: regulation by interleukin-1beta and tumour necrosis factor-alpha. Inflammation Res. 55:483–490; 2006.

[12] Del Carlo Jr M.; Loeser, R. F. Nitric oxide-mediated chondrocyte cell deathrequires the generation of additional reactive oxygen species. Arthritis Rheum.46:394–403; 2002.

[13] Henrotin, Y.; Kurz, B.; Aigner, T. Oxygen and reactive oxygen species incartilage degradation: friends or foes? Osteoarthritis Cartilage 13:643–654;2005.

[14] De Bari, C.; Dell0Accio, F.; Tylzanowski, P.; Luyten, F. P. Multipotent mesench-ymal stem cells from adult human synovial membrane. Arthritis Rheum.44:1928–1942; 2001.

[15] Scharstuhl, A.; Schewe, B.; Benz, K.; Gaissmaier, C.; Buhring, H. J.; Stoop, R.Chondrogenic potential of human adult mesenchymal stem cells is indepen-dent of age or osteoarthritis etiology. Stem Cells 25:3244–3251; 2007.

[16] ter Huurne, M.; Schelbergen, R.; Blattes, R.; Blom, A.; de Munter, W.; Grevers,L. C.; Jeanson, J.; Noel, D.; Casteilla, L.; Jorgensen, C.; van den Berg, W.; vanLent, P. L. Antiinflammatory and chondroprotective effects of intraarticularinjection of adipose-derived stem cells in experimental osteoarthritis. ArthritisRheum. 64:3604–3613; 2012.

[17] Wehling, N.; Palmer, G. D.; Pilapil, C.; Liu, F.; Wells, J. W.; Muller, P. E.; Evans, C. H.;Porter, R. M. Interleukin-1beta and tumor necrosis factor alpha inhibit chondro-genesis by human mesenchymal stem cells through NF-kappaB-dependentpathways. Arthritis Rheum. 60:801–812; 2009.

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132








[18] Zakany, R.; Szijgyarto, Z.; Matta, C.; Juhasz, T.; Csortos, C.; Szucs, K.; Czifra, G.;Biro, T.; Modis, L.; Gergely, P. Hydrogen peroxide inhibits formation ofcartilage in chicken micromass cultures and decreases the activity of calci-neurin: implication of ERK1/2 and Sox9 pathways. Exp. Cell Res. 305:190–199;2005.

[19] Tan, D. X.; Manchester, L. C.; Terron, M. P.; Flores, L. J.; Reiter, R. J. Onemolecule, many derivatives: a never-ending interaction of melatonin withreactive oxygen and nitrogen species? J. Pineal Res. 42:28–42; 2007.

[20] Chang, H. M.; Huang, Y. L.; Lan, C. T.; Wu, U. I.; Hu, M. E.; Youn, S. C. Melatoninpreserves superoxide dismutase activity in hypoglossal motoneurons of adultrats following peripheral nerve injury. J. Pineal Res. 44:172–180; 2008.

[21] Brigelius-Flohe, R.; Maiorino, M. Glutathione peroxidases. Biochim. Biophys.Acta 1830:3289–3303; 2013.

[22] Peyrot, F.; Ducrocq, C. Potential role of tryptophan derivatives in stressresponses characterized by the generation of reactive oxygen and nitrogenspecies. J. Pineal Res 45:235–246; 2008.

[23] Mias, C.; Trouche, E.; Seguelas, M. H.; Calcagno, F.; Dignat-George, F.; Sabatier,F.; Piercecchi-Marti, M. D.; Daniel, L.; Bianchi, P.; Calise, D.; Bourin, P.; Parini,A.; Cussac, D. Ex vivo pretreatment with melatonin improves survival,proangiogenic/mitogenic activity, and efficiency of mesenchymal stem cellsinjected into ischemic kidney. Stem Cells 26:1749–1757; 2008.

[24] Pei, M.; He, F.; Wei, L.; Rawson, A. Melatonin enhances cartilage matrixsynthesis by porcine articular chondrocytes. J. Pineal Res. 46:181–187; 2009.

[25] Liu, X.; Gong, Y.; Xiong, K.; Ye, Y.; Xiong, Y.; Zhuang, Z.; Luo, Y.; Jiang, Q.; He, F.Melatonin mediates protective effects on inflammatory response induced byinterleukin-1 beta in human mesenchymal stem cells. J. Pineal Res. 55:14–25;2013.

[26] Histing, T.; Anton, C.; Scheuer, C.; Garcia, P.; Holstein, J. H.; Klein, M.; Matthys,R.; Pohlemann, T.; Menger, M. D. Melatonin impairs fracture healing bysuppressing RANKL-mediated bone remodeling. J. Surg. Res. 173:83–90; 2012.

[27] Ganguly, K.; Swarnakar, S. Induction of matrix metalloproteinase-9 and -3 innonsteroidal anti-inflammatory drug-induced acute gastric ulcers in mice:regulation by melatonin. J. Pineal Res. 47:43–55; 2009.

[28] Chang, C. C.; Tien, C. H.; Lee, E. J.; Juan, W. S.; Chen, Y. H.; Hung, Y. C.; Chen, T. Y.;Chen, H. Y.; Wu, T. S. Melatonin inhibits matrix metalloproteinase-9 (MMP-9)activation in the lipopolysaccharide (LPS)-stimulated RAW 264.7 and BV2 cellsand a mouse model of meningitis. J. Pineal Res. 53:188–197; 2012.

[29] He, H.; Liu, X.; Peng, L.; Gao, Z.; Ye, Y.; Su, Y.; Zhao, Q.; Wang, K.; Gong, Y.; He, F.Promotion of hepatic differentiation of bone marrow mesenchymal stem cellson decellularized cell-deposited extracellular matrix. Biomed. Res. Int.2013:406871; 2013.

[30] Dumay, A.; Rincheval, V.; Trotot, P.; Mignotte, B.; Vayssiere, J. L. The super-oxide dismutase inhibitor diethyldithiocarbamate has antagonistic effects onapoptosis by triggering both cytochrome c release and caspase inhibition. FreeRadic. Biol. Med. 40:1377–1390; 2006.

[31] Henrotin, Y. E.; Bruckner, P.; Pujol, J. P. The role of reactive oxygen species inhomeostasis and degradation of cartilage. Osteoarthritis Cartilage 11:747–755;2003.

[32] Shi, D.; Xiao, X.; Wang, J.; Liu, L.; Chen, W.; Fu, L.; Xie, F.; Huang, W.; Deng, W.Melatonin suppresses proinflammatory mediators in lipopolysaccharide-stimulated CRL1999 cells via targeting MAPK, NF-kappaB, c/EBPbeta, andp300 signaling. J. Pineal Res. 53:154–165; 2012.

[33] Heldens, G. T.; Blaney Davidson, E. N.; Vitters, E. L.; Schreurs, B. W.; Piek, E.;van den Berg, W. B.; van der Kraan, P. M. Catabolic factors and osteoarthritis-conditioned medium inhibit chondrogenesis of human mesenchymal stemcells. Tissue Eng. Part A 18:45–54; 2012.

[34] Roman-Blas, J. A.; Stokes, D. G.; Jimenez, S. A. Modulation of TGF-betasignaling by proinflammatory cytokines in articular chondrocytes. Osteoar-thritis Cartilage 15:1367–1377; 2007.

[35] Kim, C. H.; Kim, K. H.; Yoo, Y. M. Melatonin protects against apoptoticand autophagic cell death in C2C12 murine myoblast cells. J. Pineal Res.50:241–249; 2011.

[36] Wehling, P.; Moser, C.; Frisbie, D.; McIlwraith, C. W.; Kawcak, C. E.; Krauspe, R.;Reinecke, J. A. Autologous conditioned serum in the treatment of orthopedicdiseases: the orthokine therapy. BioDrugs 21:323–332; 2007.

[37] Yasuhara, R.; Miyamoto, Y.; Akaike, T.; Akuta, T.; Nakamura, M.; Takami, M.;Morimura, N.; Yasu, K.; Kamijo, R. Interleukin-1beta induces death in

chondrocyte-like ATDC5 cells through mitochondrial dysfunction and energydepletion in a reactive nitrogen and oxygen species-dependent manner.Biochem. J. 389:315–323; 2005.

[38] Scott, J. L.; Gabrielides, C.; Davidson, R. K.; Swingler, T. E.; Clark, I. M.; Wallis, G. A.;Boot-Handford, R. P.; Kirkwood, T. B.; Taylor, R. W.; Young, D. A. Superoxidedismutase downregulation in osteoarthritis progression and end-stage disease.Ann. Rheum. Dis. 69:1502–1510; 2010.

[39] Kim, B. Y.; Han, M. J.; Chung, A. S. Effects of reactive oxygen species onproliferation of Chinese hamster lung fibroblast (V79) cells. Free Radic. Biol.Med. 30:686–698; 2001.

[40] Pei, M.; Zhang, Y.; Li, J.; Chen, D. Antioxidation of decellularized stem cellmatrix promotes human synovium-derived stem cell-based chondrogenesis.Stem Cells Dev. 22:889–900; 2013.

[41] Duan, W.; Yang, Y.; Yi, W.; Yan, J.; Liang, Z.; Wang, N.; Li, Y.; Chen, W.; Yu, S.;Jin, Z.; Yi, D. New role of JAK2/STAT3 signaling in endothelial cell oxidativestress injury and protective effect of melatonin. PLoS One 8:e57941; 2013.

[42] Kim, K. S.; Choi, H. W.; Yoon, H. E.; Kim, I. Y. Reactive oxygen species generatedby NADPH oxidase 2 and 4 are required for chondrogenic differentiation.J. Biol. Chem. 285:40294–40302; 2010.

[43] Ho, P. J.; Yen, M. L.; Tang, B. C.; Chen, C. T.; Yen, B. L. H2O2 accumulationmediates differentiation capacity alteration, but not proliferative decline, insenescent human fetal mesenchymal stem cells. Antioxid. Redox Signaling18:1895–1905; 2013.

[44] Lim, H. D.; Kim, Y. S.; Ko, S. H.; Yoon, I. J.; Cho, S. G.; Chun, Y. H.; Choi, B. J.; Kim,E. C. Cytoprotective and anti-inflammatory effects of melatonin in hydrogenperoxide-stimulated CHON-001 human chondrocyte cell line and rabbitmodel of osteoarthritis via the SIRT1 pathway. J. Pineal Res. 53:225–237; 2012.

[45] Gavriilidis, C.; Miwa, S.; von Zglinicki, T.; Taylor, R. W.; Young, D. A.Mitochondrial dysfunction in osteoarthritis is associated with down-regulation of superoxide dismutase 2. Arthritis Rheum. 65:378–387; 2013.

[46] Inarrea, P.; Casanova, A.; Alava, M. A.; Iturralde, M.; Cadenas, E. Melatonin andsteroid hormones activate intermembrane Cu,Zn-superoxide dismutase bymeans of mitochondrial cytochrome P450. Free Radic. Biol. Med. 50:1575–-1581; 2011.

[47] Florian, S.; Krehl, S.; Loewinger, M.; Kipp, A.; Banning, A.; Esworthy, S.; Chu, F. F.;Brigelius-Flohe, R. Loss of GPx2 increases apoptosis, mitosis, and GPx1 expres-sion in the intestine of mice. Free Radic. Biol. Med. 49:1694–1702; 2010.

[48] Jallali, N.; Ridha, H.; Thrasivoulou, C.; Underwood, C.; Butler, P. E.; Cowen, T.Vulnerability to ROS-induced cell death in ageing articular cartilage: the roleof antioxidant enzyme activity. Osteoarthritis Cartilage 13:614–622; 2005.

[49] Ebert, R.; Ulmer, M.; Zeck, S.; Meissner-Weigl, J.; Schneider, D.; Stopper, H.;Schupp, N.; Kassem, M.; Jakob, F. Selenium supplementation restores theantioxidative capacity and prevents cell damage in bone marrow stromal cellsin vitro. Stem Cells 24:1226–1235; 2006.

[50] Swarnakar, S.; Paul, S.; Singh, L. P.; Reiter, R. J. Matrix metalloproteinases inhealth and disease: regulation by melatonin. J. Pineal Res. 50:8–20; 2011.

[51] Tchetina, E. V.; Kobayashi, M.; Yasuda, T.; Meijers, T.; Pidoux, I.; Poole, A. R.Chondrocyte hypertrophy can be induced by a cryptic sequence of type IIcollagen and is accompanied by the induction of MMP-13 and collagenaseactivity: implications for development and arthritis. Matrix Biol. 26:247–258;2007.

[52] Galasso, O.; Familiari, F.; De Gori, M.; Gasparini, G. Recent findings on the roleof gelatinases (matrix metalloproteinase-2 and -9) in osteoarthritis. Adv.Orthop. 2012:834208; 2012.

[53] Soder, S.; Roach, H. I.; Oehler, S.; Bau, B.; Haag, J.; Aigner, T. MMP-9/gelatinaseB is a gene product of human adult articular chondrocytes and increased inosteoarthritic cartilage. Clin. Exp. Rheumatol. 24:302–304; 2006.

[54] Martinez-Lemus, L. A.; Zhao, G.; Galinanes, E. L.; Boone, M. Inward remodelingof resistance arteries requires reactive oxygen species-dependent activation ofmatrix metalloproteinases. Am. J. Physiol. Heart Circ. Physiol. 300:H2005–2015;2011.

[55] Del Carlo, M.; Schwartz, D.; Erickson, E. A.; Loeser, R. F. Endogenous produc-tion of reactive oxygen species is required for stimulation of human articularchondrocyte matrix metalloproteinase production by fibronectin fragments.Free Radic. Biol. Med. 42:1350–1358; 2007.

123456789

1011121314151617181920212223242526272829303132333435363738394041424344454647484950515253

54555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899

100101102103104105






free radical biology and medicine 2013 - rescue of...synovial ﬁbroblasts, and chondrocytes during...

Documents