n-acetyl-l-cysteine inhibits wear particle-induced prosthesis loosening

Journal of Surgical Research 168, e163–e172 (2011)doi:10.1016/j.jss.2010.12.006

N-acetyl-L-cysteine Inhibits Wear Particle-Induced Prosthesis Loosening

Qing Fang, M.D.,2 Huayi Wang, M.D.,2 Shu Zhu, M.D., and Qingsheng Zhu, M.D., Ph.D.1

Institute of Orthopedic Surgery, Xijing Hospital, Fourth Military Medical University, Xi’an, China

Submitted for publication October 13, 2010

Background. To study whether N-acetyl-L-cysteine(NAC) has beneficial effects on wear particle-inducedosteoclastogenesis and osteolysis.Materials andMethods. In vitro, cells were cultured

for 5 d and were then exposed to polymethylmethacry-late (PMMA) particles orwere pretreated withNAC 1 hprior to stimulation with optimal PMMA particles. Af-ter 48 h, the number of osteoclasts was determined bytartrate-resistant acid phosphatase (TRAP) staining.In vivo, a murine calvarial model of PMMA particle-induced osteolysis, was used. PMMAparticles were im-planted on the calvariae ofC57BL/J6mice, andNAC (10and 50 mg/kg) was given intraperitoneally every day.Two weeks later, the calvariae were removed and pro-cessed for micro-CT and histomorphometry analysis.Results. TRAP staining showed that the osteoclasto-

genic response was dose-dependent with PMMA parti-cles. Compared with the PBS group, the PMMA groupshowed a significant decrease in bone mineral density(BMD), bone volume fraction (BVF), cortical meanthickness (CMT), and cortical area/total area (Ct)(P < 0.05). Treatment with NAC (10 and 50 mg/kg)attenuated the PMMA particle-induced decrease inBMD,BVF,CMT, andCt (P< 0.05 versusPMMAgroup).NAC inhibited the osteoclastogenesis and osteolysisthat is caused by wear particles. The TRAP (D) osteo-clast number and the osteolysis area in PBS, PMMA,NAC (10 mg/kg), and NAC (50 mg/kg) were 6.0 ± 0.6,22.5 ± 1.2, 15.8 ± 0.7, 8.7 ± 1.0 and 0.075 ± 0.011, 0.340 ±0.014, 0.231 ± 0.018, 0.142 ± 0.026 mm2, respectively(P < 0.05).

Conclusion. Our in vitroand in vivowork shows thatNAC may effectively inhibit osteoclastogenesis andmay suppress wear particle-induced osteolysis,indicating that NAC may be useful in the prevention

1 To whom correspondence and reprint requests should be ad-dressedat the InstituteofOrthopedicSurgery,XijingHospital,FourthMilitary Medical University, Xi’an 710032, China. E-mail: [email protected].

2 These authors contributed equally to the study.

e16

or treatment of wear particle-induced prosthesisloosening. � 2011 Elsevier Inc. All rights reserved.

Key Words: N-acetyl-L-cysteine; PMMA; osteoclasto-genesis; osteolysis; prosthesis loosening.

INTRODUCTION

Each year over a million total joint arthroplasties areperformed because of the joint destruction caused byvarious pathologic conditions, including rheumatoidarthritis and osteoarthritis. Periprosthetic osteolysiswith resultant bone loss and prosthesis loosening isthe biggest problem limiting the durability of total jointreplacements [1, 2]. Osteoclasts play a crucial role inbone remodeling because these cells are the predomi-nant type of cells involved in osteolysis. Cell functionsare regulated by several cytokines, such as growth fac-tors, vitamins and inflammatory factors, including tu-mor necrosis factor alpha (TNF-a), interleukin-6, andinterleukin-1beta (IL-1b) [3]. Wear particles such aspolymethylmethacrylate (PMMA) [4], a major compo-nent of the implant supporting cement generatedfrom the prosthesis, are phagocytosed by macrophagesand initiate an inflammatory response that induces therelease of proinflammatory cytokines, such as TNF-a,IL-6, and IL-1b. This release subsequently leads tothe recruitment of activated osteoclasts to the bone-implant interface. These cytokines stimulate the differ-entiation of osteoclast precursor cells as well as theactivation of mature osteoclasts [3]. TNF-a is presentin periprosthetic membrane tissue obtained from looseimplants and the macrophage expression of TNF-a invitro is increased by exposure to implant-associatedparticles [5, 6]. Additionally, a knock out study revealedthat mice failing to express the p55 TNF receptor areprotected from the profound osteoclastic osteolysis afterthe subperiosteal implantation of PMMA particles [7].Recent studies have demonstrated that the NF-kB

0022-4804/$36.00� 2011 Elsevier Inc. All rights reserved.

3

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.jss.2010.12.006

http://dx.doi.org/10.1016/j.jss.2010.12.006

JOURNAL OF SURGICAL RESEARCH: VOL. 168, NO. 2, JUNE 15, 2011e164

signaling pathway is activated in macrophages and os-teoclast precursor cells after exposure to implant de-bris, suggesting that NF-kB signaling may play a rolein particle-directed osteoclastogenesis [8, 9]. Upstreamblockade of NF-kB signaling with RANK:Fc [10] andOPG [3] inhibits particle-induced osteoclast formationand particle-induced osteolysis.

N-acetyl-L-cysteine (NAC) is shown to act directly asa reducing agent and indirectly by stimulating thesynthesis of other antioxidant enzymes, such as gluta-thione [11, 12]. At the intracellular level, NAC is theprecursor of glutathione (GSH) synthesis. Previousstudies have shown that NAC acts negatively to regu-late NF-kB activity via impeding the phosphorylationof IkB [13].

Given the effect of NAC on NF-kB activity, we under-took the present study to determine if NAC couldinhibit PMMA particle-induced osteoclastogenesis invitro and PMMA particle-directed osteoclastic osteoly-sis in vivo. The hope is to provide a clue into the preven-tion or treatment of periprosthetic osteolysis andaseptic loosening following total joint replacement.

MATERIALS AND METHODS

N-acetyl-L-cysteine, 1, 25-(OH)2D3, TRAP staining kit, andE-toxate kit were purchased from Sigma (St. Louis, MO, USA).

Polymethylmethacrylate Particles

PMMA particles were provided by Professor Canhui Lu (SichuanUniversity, Chengdu, China). Scanning electron microscopy (HitachiS23400; Osaka, Japan) revealed particle sizes (range, 0.2 mm to 1 mmin diameter) and shapes. The majority of the particles had slightly ir-regular shapes and smooth surfaces. The majority of these submicronparticles are in aggregates and in the biologically active size rangecompared with particles isolated from periprosthetic membrane tis-sue [14]. The particles were washed continuously in 25% nitric acidand absolute ethanol for 24 h respectively in a shaking device andwere then rinsed 4 times with phosphate-buffered saline (PBS) to re-move adherent endotoxins according to Ragab [15]. No contaminationwith lipopolysaccharide (LPS) was detected by a commercial E-toxatekit in this PMMA particle preparation. Stock solutions were preparedwith alpha minimal essential medium (a-MEM) at a concentration of100 mg/mL, and treatment concentrations are indicated for eachexperiment.

Cell Culture

Osteoclastogenic whole bonemarrow cultures were obtained by iso-lating whole bone marrow cells from the tibia and femur of 5–7-d-oldSprague-Dawley rats (Laboratory Animal Centre of the Fourth Mili-tary Medical University, Xi’an, China) [9, 16]. After cervical disloca-tion, the rats were sterilized in 75% ethanol for 10 min. The longbones were taken out, and the soft tissues attached to the boneswere removed. The bones were minced and rinsed in a-MEM. Thecells were dissociated and collected from the bone fragments. The cellswere then centrifuged (1500 3 g for 10 min). The cells were used asunfractionated bone cells and were resuspended in a-MEM contain-ing 10% fetal bovine serum (FBS) plus 1% penicillin and streptomycin(GIBCO; Invitrogen Corp., Carlsbad, CA, USA).

Cell Treatments

Cells were seeded in 24-well cell culture plates in the presence of10 nmol/L1, 25-(OH)2D3 and were incubated at 37�C with 5% CO2.After 5 d, cells were subjected to experimental conditions for 48 h.The experimental conditionswere the following: (1) PMMAwas addedto the culture medium at a final concentration of 0 (control), 0.1, 0.2,0.4, and 0.6mg/mL, respectively, and (2) first, the cellswere incubatedfor 1 h with NAC at a final concentration of 0, 5, 10, 15, or 20 mmol/L,respectively, and then the cells were incubated with or without0.6 mg/mL PMMA particles. Cells cultured in the absence of PMMAand SIN were used as control.

TRAP-Staining

After the experiments were complete, the cell cultures were fixedand TRAP-stained according to the manufacturer’s directions.TRAP positive (pink or purple color) multinuclear (three or more nu-clei) cells were counted as osteoclasts. The number of osteoclast wascounted in every well. All experimental conditions were performedin triplicate or quadruplicate.

Animal Experiments

Amurine calvarial model of PMMA particle-induced osteolysis wasused in this study and was based on the original model of calvarial os-teolysis [17, 18]. The institutional guidelines for the care and use oflaboratory animals at the Fourth Military Medical University werestrictly followed. Twenty-eight healthy C57BL/J6 mice aged 6–8 wkwith a mean weight of 19.8 6 2.1 g (18.4–21.8 g) were randomly di-vided into four groups: (1) the PBS group (negative control group),(2) the PMMA group, where the mice received 30 mg of PMMA parti-cles implantation, (3) the PMMAþNAC10 group, where the micereceived 30 mg of PMMA particles plus 10 mg/kg of NAC intraperito-neal injection every day from d 1 to d 14, and (4) the PMMAþNAC50group, where the mice received 30 mg of PMMA particles plus 50 mg/kg of NAC intraperitoneal injection every day from d 1 to d 14. The an-imals had free access to water and food andwere kept in a 12 h on/12 hoff specific pathogen-free animal room. Each mouse was anesthetizedwith an intraperitoneal injection of pentobarbital. Under sterile con-ditions, a longitudinal incisionwasmade on the scalp between the twoexternal ears, and the external cranial periosteum was exposed. Theperiosteumwas removed until coronal, sagittal, and lambdoid suturesof the calvariae were visible. A total amount of 30 mg of PMMA parti-cles in 300 mL PBS was applied on the surface of the calvariae. Theskin was closed with simple interrupted suture to prevent leakageof the particles. In the PBS group, the animals underwent operationwith 300 mL sterile PBS without particles. After the operation, themice were warmed up, recovered, and sent back to the animal room.Twoweeks after the operation, the animalswere killed in aCO2 cham-ber. No complications occurred. All wounds healed uneventfully. Thecalvaria were removed with care not to split the sagittal suture. Thespecimens were freed of soft tissue, and the underlying brain wasfixed in 10% phosphate-buffered formalin before further analysis.

Micro-Computed Tomography

After sacrificing the mice, the skulls were analyzed by three-dimensional micro-focus computed tomography (micro-CT) (eXploreLocus SP; GE Healthcare, Waukesha, WI, USA). Micro-CT analysismainly focused on the osseous properties in the range of the sagittalsuture in the mice skulls. The X-ray source was set at a voltage of80 kV and a current of 80 mA. To perform accurate quantitative anal-ysis of the skulls and their micro-architecture, Micro-view ABA 2.1.2(GE Healthcare) software was used. In the comparison of the skulls,the following morphometric parameters were considered: bone min-eral density (BMD), bone volume fraction (BVF), cortical mean thick-ness (CMT), and cortical area/total area (Ct). Three-dimensionalimages of the osteolysis area were also reconstructed based on themicro-CT analysis results.

FANG ET AL.: NAC INHIBITS WEAR PARTICLE-INDUCED PROSTHESIS LOOSENING e165

Histologic Analysis

The harvest samples were decalcified in 14% EDTA, dehydrated ingraded alcohols, and embedded in paraffin. Each specimen was sec-tioned to a thickness of 4 mm on the sagittal plane at 2 mm lateralto the midsagittal suture. Each section was sampled twice, 150 mmapart, producing four sections per animal. The sections were stainedwith hematoxylin and eosin (HE) or TRAP. The sections stained withHE or with TRAP were examined under light microscopy (DMLA; Le-ica, Wetzlar, Germany). Histomorphometric analysis was performedusing QWin V3.1 software (Leica, Germany). The coronal suturearea in the HE-stained sections was determined by tracing the areaof soft tissue between the parietal bones, including any resorptionpits on the superior surface of the calvariae visible in the same field,and was quantified by the modified method described by Childs[19]et al. The number of osteoclasts was counted in the TRAP stained sec-tions. The results from the four sections for each animal were aver-aged, and the average coronal suture area and the number ofosteoclast for each group of animals were determined [20].

Statistics

Results were expressed as mean 6 standard deviation (SD). SPSS16.0 (Chicago, IL) statistical software was used to analyze the databy one-way analysis of variance (ANOVA), followed by Dunnett T3-test to determine significance. All results were considered to besignificant at the 5% critical level (P < 0.05).

RESULTS

PMMA Particles Stimulate Osteoclastogenesis in vitro

Implant particulate debris is a potent stimulus for os-teoclast differentiation and mature osteoclast function[21]. In this assay, whole bone marrow cultures wereutilized to test the effect of PMMA particles on osteo-clast formation. SD rat whole bone marrow cultureswere plated and maintained in the presence of10 nmol/L 1, 25-(OH)2D3 for 5 d and were then exposedto PMMA particles. After 48 h of particle exposure, weobserved many TRAP-positive giant multinuclearosteoclast-like cells formed compared with the controlgroup. This osteoclastogenic response was dose-dependent and was optimal at d 5 when whole bonemarrow cultures were treated for 48 h with PMMA par-ticles at a concentration of 0.6 mg/mL (P< 0.05, Fig. 1).

NAC Suppressed Osteoclastogenesis in vitro

Whole bone marrow cultures were pretreated withNAC for 1 h, and then the cultures were stimulatedwith PMMA particles (0.6 mg/mL). NAC, applied ata concentration of 5 mmol/L, could block PMMAparticle-induced osteoclast formation evident by the re-duction of TRAP-positive, multinucleated osteoclasts.When the concentration was increased to 15 mmol/L,NAC significantly inhibited the osteoclastogenesis.When we applied NAC at 20 mmol/L, we found NACcould inhibit survival of the cells (P < 0.05, Fig. 2).

Micro-CT Analysis

The murine calvarial osteolysis model evaluated ifNAC would improve particle-induced osteolysis. Thethree-dimensional reconstruction of bone micro-architecture in the murine skull by micro-CT is shownin Figure 3. In the PMMA group, the particle-inducedosteolysis was apparent. In the NAC groups, however,the particle-induced osteolysis was reduced comparedwith the PMMA group. The PBS group showed no pro-nounced osteolysis. Treatment withNAC prevented theparticle-induced effects on bone metabolism and bonemicro-architecture in the murine skull.

The quantification of bone changes in the murineskull is shown in Figure 4. Compared with the PBSgroup, the PMMA group showed a significant decreasein BMD, BVF, CMT, and Ct by 44.69% 6 2.61%(20.56 6 0.61 versus 11.36 6 0.46 mg/cc), 48.68 67.65% (12.67 6 0.93 versus 6.58 6 0.42), 26.08% 63.12% (0.349 6 0.008 versus 0.257 6 0.009 mm), and40.04% 6 4.22% (10.50 6 0.52 versus 6.29 6 0.39), re-spectively. Treatment with NAC (10 mg/kg) attenuatedthe PMMA particle-induced decrease in BMD, BVF,CMT, and Ct by 40.37% 6 6.01% (15.93 6 0.35 versus11.36 6 0.46 mg/cc), 27.57% 6 13.06% (8.37 6 0.33versus 6.58 6 0.42), 18.31% 6 7.99% (0.304 6 0.013versus 0.257 6 0.009 mm), and 13.66% 6 9.92% (7.12 60.46 versus 6.296 0.39), respectively. In a similar fash-ion, treatment with NAC (50 mg/kg) significantly pre-vented the PMMA particle-induced decrease in BMD,BVF, CMT and Ct by 59.37%6 6.63% (18.096 0.25 ver-sus 11.366 0.46 mg/cc), 59.55%6 12.20% (10.486 0.60versus 6.586 0.42), 22.06%6 3.12% (0.3146 0.010 ver-sus 0.257 6 0.009 mm), and 31.42% 6 9.37% (8.24 60.31 versus 6.29 6 0.39), respectively (P < 0.05). Whileeither treatment increasedBMD,BVF, CMTandCt rel-ative to the PMMA group, NAC(50 mg/kg) group wasfound to exert a greater effect on BMD, BVF, CMT,and Ct by 13.55% 6 2.17% (18.09 6 0.25 versus15.93 6 0.35 mg/cc), 25.47% 6 9.93% (10.48 6 0.60versus 8.37 6 0.33), 3.47% 6 5.45% (0.314 6 0.010versus 0.304 6 0.013 mm), and 15.80% 6 3.26%(8.24 6 0.31 versus 7.12 6 0.46) compared with NAC(10 mg/kg) group.

Bone Histomorphometry

NAC Inhibited Osteolysis Caused by Wear ParticlesIn Vivo

Little inflammatory reaction and little resorptionwere observed in the PBS group (Fig. 5A and E). Therewas a pronounced inflammatory reaction with highlyvascularized granulation containing macrophages andmultinuclear giant cells in the PMMA group (Fig. 5B

FIG. 1. PMMA particles stimulate osteoclastogenesis in vitro Murine whole bone marrow cultures in the presence of 10 nmol/L1,25(OH)2D3 were maintained for 5 d and were then treated with control media (A) or media containing PMMA particles (B). After 48 h of par-ticle exposure, cultures were fixed, stained, and analyzed by light microscopy (TRAP stain, 340). Quantification of giant osteoclasts (mean 6SD) (C) is shown. * P < 0.05 versus control group, n ¼ 4.


and F). Marked osteolysis was located adjacent to theseinflammatory reaction tissues. The bone volume ofmurine calvarium was reduced extensively by osteoly-sis in this group. In the NAC (10 and 50 mg/kg) treatedgroups, bone destruction and resorption were lessmarked than that of the PMMA group but were morepronounced than that of the PBS group (Fig. 5C, G, D,and H). The osteolysis area was quantitatively evalu-ated according to the coronal suture area includingany visible resorption pits in each sample. In the PBSgroup, the osteolysis area was 0.075 6 0.011 mm2.The skull implanted with PMMA showed the greatestosteolysis area, 0.340 6 0.014 mm2, which was signifi-cantly higher than that of the PBS group (P < 0.05).In the NAC (10 and 50mg/kg) treated group, the osteol-ysis area was 0.231 6 0.018 mm2 and 0.142 60.026 mm2, respectively. The NAC (10 and 50 mg/kg)treated group had an obviously smaller osteolysisarea than the PMMA group (P < 0.05, Fig. 6A).

NAC Inhibited Osteoclastogenesis Caused by WearParticles In Vivo

The TRAP (þ) osteoclast number within the experi-mental areas in the PBS group was 6.0 6 0.6. In thePMMA group, the number rose to 22.5 6 1.2. In theNAC (10 mg/kg) group, the number was 15.8 6 0.7,while in the NAC (50 mg/kg) group, it was 8.7 6 1.0.There was a significant difference in the osteoclastnumber between the PBS and the PMMA groups (P <0.05). NAC (10 and 50 mg/kg) treatment significantlysuppressed osteoclastogenesis induced by PMMAparticles (P < 0.05, Fig. 6B).

DISCUSSION

Periprosthetic osteolysis and prosthesis loosening, asa biologic consequence of particulate wear debris limitlong-term durability of total joint replacement,

FIG. 2. NAC blocks PMMA particle-induced osteoclastogenesis in vitro. Osteoclastogenic whole bone marrow cultures were isolated andmaintained for 5 d in media containing 10 nmol/L 1,25(OH)2D3. Cell were pretreated for 1 h with NAC at the indicated concentrations(NAC 5, 10, 15, 20 mmol/L) and then incubated with or without PMMA (0.6 mg/mL) for an additional 48 h. Cultures were fixed on d 8,TRAP stained and analyzed by light microscopy (TRAP stain, 340). (A) Control, (B) PMMA, (C) NAC5þPMMA, (D) NAC10þPMMA, (E)NAC15þPMMA, (F) NAC20þPMMA, (G) Quantification of osteoclasts. Quantification of giant osteoclasts (mean 6 SD, n ¼ 4) are shown.*P < 0.05 versus PMMA group.


although enormous efforts have been made to improvebearing biomaterials, implant designs, and steriliza-tion methods [22–24]. In clinical practice, mechanicalfactors, fluid pressure, and wear particles have beenreported to result in osteolysis and loosening [25]. Par-ticularly, implant-derived particles are thought to bea critical cause of osteolysis and loosening that leadsto the failure of a total joint replacement [26–28].

Osteoclast plays a direct and central role in wearparticle-induced osteolysis and loosening. Redlichet al. have found that severe TNF-dependent arthritisdoes not cause bone destruction in Fos–/– mice that

lack osteoclasts [29]. Clearly, inhibiting osteoclastogen-esis or mature osteoclast survival is an important path-way in improving wear particle-induced osteolysis andloosening.

Numerous studies have revealed that the NF-kB sig-naling pathway plays an important role in osteoclasto-genesis and osteolysis. This central role of NF-kB issupported by the fact that although p50- or p52-deficient mice have no obvious bone disorder, micedoubly deficient in the NF-kB subunits p50 and p52developed severe osteopetrosis due to a defect in osteo-clast differentiation, indicating a critical function of p50

FIG. 3. Micro-CT three-dimensional image of bone resorption in (A) PBS, (B) PMMA, (C) PMMAþNAC 10 (mg/kg), and (D) PMMAþNAC50 (mg/kg) treated mice.


and p52 in osteoclastogenesis [30]. Clohisy et al. havedemonstrated that PMMA particles rapidly induceNF-kB DNA-binding affinity in murine osteoclast pre-cursor cells [8] via a mechanism mediated partly byTNF. These studies strongly suggested that the regula-tion of transcription factor NF-kB and its activity mightbe essential for osteoclastogenesis and osteolysis. NF-kB is a transcription factor that exists in virtually alleukaryotic cell types. In most rest cells, NF-kB residesin the cytoplasm in a high affinity association withthe inhibitory protein IkB [31–33]. IkB associationwith cytosolic NF-kB prevents nuclear translocationand DNA binding of the transcription factor. Alterna-tively, various stimuli can result in IkB phosphoryla-tion causing dissociation from NF-kB and the releaseof active NF-kB for nuclear translocation and the regu-lation of gene expression. Therefore, the blockade of IkBdissociation fromNF-kBmay be used as a strategy to in-hibit the NF-kB signaling pathway. Certain pharma-ceutical agents may function as NF-kB inhibitors byaltering IkB phosphorylation and/or degradation.

NAC is a nontoxic dietary glutathione precursor andantioxidant compound. It may act negatively to regu-late NF-kB activity by impeding the phosphorylationof IkB, leading to the decreased expression of NF-kB-regulated antiapoptotic genes, such as Bcl-2. Bcl-2 islocated on the outer mitochondrial membrane andacts to promote cell survival. Either through interac-tion with Bcl-2 or through direct interaction with the

mitochondria, NAC may greatly reduce mitochondrialmembrane potential in a concentration dependent-fashion and may markedly down-regulate mitochon-drial Bcl-2 expression. This in turn promotesmitochondria-mediated apoptosis and thus exerts pro-found effects on cell survival [31–36]. In the presentstudy, our in vitro and in vivo data showed that NACcould inhibit PMMA particle-induced osteoclastogene-sis and osteolysis.

When mouse bone marrow cells were cultured withPMMA particles in vitro, many TRAP-positive, multi-nuclear osteoclasts formed. While in bone marrow cul-tures pretreated with NAC for 1 h before beingstimulated with PMMA particles (0.6 mg/mL), it wasfound that NAC at a concentration of 5 mmol/l couldinhibit PMMA particle-induced osteoclast formation,evidenced by the reduction in TRAP-positive, multinu-clear osteoclasts. When its concentration increased to15mmol/l, NAC significantly inhibited osteoclastogene-sis. The presence of 20 mmol/l NAC completely blockedPMMA particle-induced osteoclast formation, evi-denced by the complete lack of TRAP-positive, multinu-clear osteoclasts. This provides an important clue fortreating wear particle-induced osteolysis with NAC,given that increased osteoclastogenesis is central toparticle-induced osteolysis [26–29].

In the in vivo experiment, we used the wear particle-induced mouse calvarial osteolysis model to study themechanism of wear particle-induced prosthesis

FIG. 4. Change of bone parameters of mice skull analyzed bymicro-CT. (A) BMD, bonemineral density, (B) BVF, bone volume fraction, (C)CMT, cortical mean thickness, (D) Ct, cortical area/total area data are presented as mean6 SD (n¼ 7). *P< 0.05, comparison between the twogroups connected by the line.


loosening and the effect of drugs treating prosthesisloosening. When 30 mg of PMMA particles were im-planted in the calvariae of C57BL/J6 mice, we observeda pronounced inflammatory reaction with highlyvascularized granulation containing macrophages,TRAP-positive multinuclear osteoclasts, and markedosteolysis located adjacent to these inflammatory reac-tion tissues. In the NAC-treated group, the inflamma-tory reaction, bone destruction, and resorption areawere less marked than that of the PMMA group butwere more pronounced than that of the PBS group.

To evaluate whether NAC would improve particle-induced osteolysis in the murine calvarial osteolysismodel, we used Micro-CT. The three-dimensional re-construction of bone micro-architecture in the murineskull by Micro-CT is shown in Figure 3. In the PMMAgroup, particle-induced osteolysis was apparentcompared with the PBS group, where there was no pro-nounced osteolysis. In the NAC groups, the particle-induced osteolysis was reduced compared with thePMMA group, indicating that treatment with NAC

prevented the particle-induced effects on bone metabo-lism and bone microarchitecture in the murine skull.The quantification of bone changes in the murine skullis shown in Figure 4. Compared with the PBS group,the PMMA group showed significant decrease inBMD, BVF, CMT, and Ct. The decreases in BMD,BVF, CMT, and Ct were associated with decreasedbone formation and increased osteolysis according toHefferan [37] et al. In this study, treatment with NAC(10 mg/kg) attenuated the PMMA particle-induced de-crease in BMD, BVF, CMT and Ct. In a similar fashion,treatment withNAC (50mg/kg) significantly preventedthe PMMA particle-induced decrease in BMD, BVF,CMT, and Ct. Treatment with NAC (50 mg/kg) alsoresulted in a differential improvement in BMD, BVF,CMT, and Ct compared with the NAC (10 mg/kg)treated group. While treatment increased BMD, BVF,CMT, and Ct relative to the PMMA group, the NAC(50 mg/kg) group was found to exert a greater effecton BMD, BVF, CMT, and Ct compared with the NAC(10 mg/kg) group. All of these showed that NAC could

FIG. 5. Histologic appearance of particle induced in vivomouse calvarial osteolysis on d 14. PMMA induced significant bone resorption (HEstain,340) (B) and osteoclastogenesis (TRAP stain,340) (F) than the PBS group (A, E). This inflammatory reactionwas effectively inhibited byNAC (10 mg/kg) (C, G) and NAC (50 mg/kg) (D, H) treatment (n ¼ 7).


FIG. 6. In vivo wear particles induced osteolysis and osteoclasto-genesis. PMMA particles induced osteolysis area was much biggerthan the control (A). PMMA particles induced formation of moreTRAP (þ) cell than the control (B). While NAC could effectively in-hibited wear particles induced osteolysis and osteoclastogenesis(n ¼ 7). *P < 0.05, comparison between the two groups connected bythe line.


inhibit wear particle-induced bone loosening. BecauseNF-kB is a transcription factor that exists in virtuallyall eukaryotic cell types and NAC may act negativelyto regulate the NF-kB activity by impeding the phos-phorylation of IkB leading to decreased expression ofNF-kB-regulated antiapoptotic genes, such as Bcl-2, itmay have a range of side effects on the survival of othercells. Extensive pharmacologic evaluation of NAC is re-quired to fully determine its effects in the treatment ofwear particle-induced prosthesis loosening.

In conclusion, the present study demonstrated thatNAC has inhibitory effects on wear particle-induced os-teolysis through the inhibition of osteoclastogenesis.Our findings suggest thatNACmay represent a promis-ing agent for the prevention and treatment of asepticloosening following total joint arthroplasty.

ACKNOWLEDGMENTS

The authors thank Professor CanHui Lu from Sichuan University(Sichuan, China) for providing us with PMMA particles. The authorsalso thankDr. Liwen Li and Dr. Dawei Zhang for their technical guid-ance and ProfessorWenli Yan for revising the language in this article.

REFERENCES

1. Yamanaka Y, Abu-Amer W, Foglia D, et al. NFAT2 is an essen-tial mediator of orthopedic particle-induced osteoclastogenesis.J Orthop Res 2008;26:1577.

2. WarashinaH, Sakano S, Kitamura S, et al. Biological reaction toalumina, zirconia, titanium and polyethylene particles im-planted onto murine calvaria. Biomaterials 2003;24:3655.

3. Clohisy JC, Frazier E, Hirayama T, et al. RANKL is an essentialcytokine mediator of polymethylmethacrylate particle-inducedosteoclastogenesis. J Orthop Res 2003;21:202.

4. Shardlow DL, Stone MH, Ingham E, et al. Cement particlescontaining radio-opacifiers stimulate pro-osteolytic cytokineproduction from a human monocytic cell line. J Bone JointSurg Br 2003;85:900.

5. Perry MJ, Mortuza FY, Ponsford FM, et al. Properties of tissuefrom around cemented joint implants with erosive and/or linearosteolysis. J Arthroplasty 1997;12:670.

6. Stea S, Visentin M, Granchi D, et al. Cytokines and osteolysisaround total hip prostheses. Cytokine 2000;12:1575.

7. Brooks RA, Wimhurst JA, Rushton N. Endotoxin contaminationof particles produces misleading inflammatory cytokine re-sponses from macrophages in vitro. J Bone Joint Surg Br 2002;84:295.

8. Clohisy JC, Teitelbaum S, Chen S, et al. Tumor necrosis factor-a mediates polymethylmethacrylate particle-induced NF-kB ac-tivation in osteoclast precursor cells. J Orthop Res 2002;20:174.

9. Soysa NS, Alles N. NF-kB functions in osteoclasts. BiochemBiophys Res Commun 2009;378:1.

10. Childs LM, Paschalis EP, Xing L, et al. In vivo RANK signalingblockade using the receptor activator of NF-kB: Fc effectivelyprevents and ameliorates wear debris-induced osteolysis via os-teoclast depletion without inhibiting osteogenesis. J BoneMinerRes 2002;17:192.

11. Nakamura Y, Shimetani A, Fujii H, et al. Glutathione can effi-ciently prevent direct current-induced cytotoxicity. J Endod2008;34:693.

12. Oka S, Kamata H, Kamata K, et al. N-acetylcysteine suppressesTNF-induced NF-kB activation through inhibition of IkBkinases. FEBS Lett 2000;472:196.

13. Barnes PJ, Karin M. Nuclear factor-kB: A pivotal transcriptionfactor in chronic inflammatory diseases. N Engl J Med 1997;336:1066.

14. Abbas S, Clohisy JC, Abu-Amer Y. Mitogen-activated protein(MAP) kinases mediate PMMA-induction of osteoclasts.J Orthop Res 2003;21:1041.

15. Ragab AA, VanDeMotter R, Lavish SA, et al. Measurement andremoval of adherent endotoxin from titanium particles andimplant surfaces. J Orthop Res 1999;17:803.

16. Chen JR, Shankar K, Nagarajan S, et al. Protective effects of es-tradiol on ethanol-induced bone loss involve inhibition of reac-tive oxygen species generation in osteoblasts and downstreamactivation of the extracellular signal-regulated kinase/signaltransducer and activator of transcription 3/receptor activatorof nuclear factor-kB ligand signaling cascade. J Pharmacol ExpTher 2008;324:50.

17. Merkel KD, Erdmann JM, McHugh KP, et al. Tumor necrosisfactor-a mediates orthopedic implant osteolysis. Am J Pathol1999;154:203.

18. Schwarz EM, Benz EB, Lu AP, et al. Quantitative small-animalsurrogate to evaluate drug efficacy in preventing wear debris-induced osteolysis. J Orthop Res 2000;18:849.


19. Childs LM, Goater JJ, O’Keefe RJ, et al. Effect of anti-tumornecrosis factor-a gene therapy on wear debris-induced osteoly-sis. J Bone Joint Surg Am 2001;83-A:1789.

20. Masui T, Sakano S, Hasegawa Y, et al. Expression of inflamma-tory cytokines, RANKL and OPG induced by titanium, cobalt-chromium and polyethylene particles. Biomaterials 2005;26:1695.

21. Clohisy JC, Hirayama T, Frazier E, et al. NF-kB signaling block-ade abolishes implant particle-induced osteoclastogenesis.J Orthop Res 2004;22:13.

22. EndoMM, Barbour PS, Barton DC, et al. Comparative wear andwear debris under three different counterface conditions ofcrosslinked and non-crosslinked ultra high molecular weightpolyethylene. Biomed Mater Eng 2001;11:23.

23. Robertsson O, Wingstrand H, Kesteris U, et al. Intracapsularpressure and loosening of hip prostheses. Preoperativemeasure-ments in 18 hips. Acta Orthop Scand 1997;68:231.

24. Sochart DH, Porter ML. The long-term results of Charnley low-friction arthroplasty in young patients who have congenitaldislocation, degenerative osteoarthrosis, or rheumatoid arthri-tis. J Bone Joint Surg Am 1997;79:1599.

25. MuratogluOK, BragdonCR,O’ConnorDO, et al. A novelmethodof cross-linking ultra-high-molecular-weight polyethylene to im-prove wear, reduce oxidation, and retain mechanical properties.Recipient of the 1999 HAP Paul Award. J Arthroplasty 2001;16:149.

26. Maloney WJ, Smith RL, Schmalzried TP, et al. Isolation andcharacterization of wear particles generated in patients whohave had failure of a hip arthroplasty without cement. J BoneJoint Surg Am 1995;77:1301.

27. Schmalzried TP, Jasty M, Harris WH. Periprosthetic bone lossin total hip arthroplasty. Polyethylene wear debris and the

concept of the effective joint space. J Bone Joint Surg Am1992;74:849.

28. Yamauchi K, Hasegawa Y, Iwasada S, et al. Head penetrationinto Hylamer acetabular liner sterilized by gamma irradiationin air and in a nitrogen atmosphere. J Arthroplasty 2001;16:463.

29. Redlich K, Hayer S, Ricci R, et al. Osteoclasts are essential forTNF-a-mediated joint destruction. J Clin Invest 2002;110:1419.

30. Franzoso G, Carlson L, Xing L, et al. Requirement for NF-kB inosteoclast and B-cell development. Genes Dev 1997;11:3482.

31. Fahy BN, SchliemanMG,MortensonMM, et al. Targeting BCL-2 overexpression in various humanmalignancies throughNF-kBinhibition by the proteasome inhibitor bortezomib. CancerChemother Pharmacol 2005;56:46.

32. Orlowski RZ, Baldwin AS Jr. NF-kB as a therapeutic target incancer. Trends Mol Med 2002;8:385.

33. Richmond A. Nf-k B, chemokine gene transcription and tumourgrowth. Nat Rev Immunol 2002;2:664.

34. Reelfs O, Tyrrell RM, Pourzand C. Ultraviolet a radiation-induced immediate iron release is a key modulator of the activa-tion of NF-kB in human skin fibroblasts. J Invest Dermatol 2004;122:1440.

35. Viatour P, Bentires-AljM, Chariot A, et al. NF-kB2/p100 inducesBcl-2 expression. Leukemia 2003;17:1349.

36. Yang J, Su Y, Richmond A. Antioxidants tiron and N-acetyl-L-cysteine differentially mediate apoptosis in melanoma cells viaa reactive oxygen species-independent NF-kB pathway. FreeRadic Biol Med 2007;42:1369.

37. Hefferan TE, Evans GL, Lotinun S, et al. Effect of gender onbone turnover in adult rats during simulated weightlessness.J Appl Physiol 2003;95:1775.

n-acetyl-l-cysteine inhibits wear particle-induced prosthesis loosening

Documents