Differential Intensity-Dependent EffectsofPulsedElectromagnetic Fieldson

RANKL-inducedOsteoclast Formation,Apoptosis, andBoneResorbingAbility in

RAW264.7Cells

PanWang,1Juan Liu,1YuefanYang,2MingmingZhai,1Xi Shao,1ZedongYan,1

Xuhui Zhang,1YanWu,3 LuCao,1BingdongSui,4 Erping Luo,1andDa Jing1*1Department ofBiomedicalEngineering, FourthMilitaryMedicalUniversity, Xi‘an, China2Department ofNeurosurgery, XijingHospital, FourthMilitaryMedicalUniversity, Xi‘an,

China3Institute of Orthopaedics, XijingHospital, FourthMilitaryMedicalUniversity, Xi‘an,

China4ResearchandDevelopment Center forTissueEngineering, FourthMilitaryMedical

University, Xi‘an, China

Pulsed electromagnetic fields (PEMF) have been proven to be effective for promoting bone massand regulating bone turnover both experimentally and clinically. However, the exact mechanismsfor the regulation of PEMF on osteoclastogenesis as well as optical exposure parameters of PEMFon inhibiting osteoclastic activities and functions remain unclear, representing significantlimitations for extensive scientific application of PEMF in clinics. In this study, RAW264.7 cellsincubated with RANKL were exposed to 15Hz PEMF (2 h/day) at various intensities (0.5, 1, 2, and3mT) for 7 days. We demonstrate that bone resorbing capacity was significantly decreased by0.5mT PEMF mainly by inhibiting osteoclast formation and maturation, but enhanced at 3mT bypromoting osteoclast apoptosis. Moreover, gene expression of RANK, NFATc1, TRAP, CTSK,BAX, and BAX/BCL-2 was significantly decreased by 0.5mT PEMF, but increased by 3mT. Ourfindings reveal a significant intensity window for low-intensity PEMF in regulating bone resorptionwith diverse nature for modulating osteoclastogenesis and apoptosis. This study not only enrichesour basic knowledge for the regulation of PEMF in osteoclastogenesis, but also may lead to moreefficient and scientific clinical application of PEMF in regulating bone turnover and inhibitingosteopenia/osteoporosis. Bioelectromagnetics. © 2017 Wiley Periodicals, Inc.

Keywords: PEMF; osteoclastogenesis; bone resorption; osteoclast differentiation; apoptosis

INTRODUCTION

Osteoporosis increases the risk of fragility frac-tures in about 50% of postmenopausal women and30% of older men, which results from dysfunctionsin the osteoclastic bone breakdown and osteoblasticbone formation process [Reid, 2015; Weitzmann andOfotokun, 2016]. Traditional pharmacological agentseither promoting bone formation (e.g., parathyroidhormone, insulin-like growth factor, and growth hor-mone) or inhibiting bone resorption (e.g., calcitonin,estrogen, and bisphosphonate) may partially help pre-vent and reverse osteoporosis [Canalis, 2013; Charlesand Aliprantis, 2014; Reid, 2015]. However, thesepharmacological drugs have issues of high costs orundesirable side effects, which are non-negligible limi-tations for their clinical application [Burge et al., 2007].

Pulsed electromagnetic fields (PEMF) as a kind ofsafe, inexpensive, and noninvasive physical approach

Grant sponsor: National Natural Science Foundation of China;grant numbers: 31500760, 81471806.

Conflict of interest: None.

Pan Wang and Juan Liu contributed equally to this work.

*Correspondence to: Da Jing, Department of Biomedical Engi-neering, Fourth Military Medical University, No. 169 ChangleWest Road, Xi`an, China. E-mail: jingdaasq@126.com

Received for review 7 December 2016; Accepted 25 June 2017

DOI: 10.1002/bem.22070Published online XX Month Year in Wiley Online Library(wileyonlinelibrary.com).

Bioelectromagnetics

� 2017Wiley Periodicals, Inc.

have shown therapeutic potential in various diseasesof the skeletomuscular system [Jing et al., 2013;Hannemann et al., 2014]. Several clinical investigationshave demonstrated that PEMF stimulation could en-hance bone mineral density, accelerate bone fracturehealing, and reduce the risk of fractures [Bassett et al.,1974; Tabrah et al., 1998; Liu et al., 2013]. Substantialstudies by our group and others have also demonstratedthat PEMF could increase in vivo bone mass withobvious enhancement of bone formation rate in osteopo-rotic animals and also enhance in vitro osteoblastactivity and osteoblastic mineralization potential [Luben,1991; Bodamyali et al., 1998; Jing et al., 2013; Jinget al., 2014; Zhai et al., 2016]. Although it has beendocumented that PEMF stimulation has the capacity ofpromoting osteoblastogenesis, we still relatively lackadequate understanding for the regulatory effects andrelated mechanisms of PEMF on osteoclastic activitiesand functions.

In the past few years, several investigators havealso explored the potential of PEMF in regulatingosteoclastic activities, and inconsistent findings havebeen documented. Chang et al. [2006] reported that7.5Hz, 3.0mV/cm PEMF exposure has the ability ofspeeding up apoptosis of osteoclasts. He et al.[2015] found that the formation of osteoclast-like cellswas potently prevented with PEMF exposure at 3.8mTfor 3 days. However, Barnaba et al. [2012] found noobvious impact of PEMF on osteoclast TRAP activityat 0.4mT, 50Hz. Moreover, our previous in vivoinvestigations have also revealed no significant effectof PEMF at 20mT, 15Hz on osteoclastogenesis andbone resorption in osteoporotic rats [Jing et al., 2013,2014]. According to these inconsistent findings, webelieve that significant parameter-dependent effects forthe regulation of PEMF on osteoclastogenesis mayexist. However, we still lack essential basic knowledgeregarding optimal exposure parameters of PEMF oninhibiting osteoclastic activities and functions. More-over, another important question regarding the exactmechanism about how osteoclasts transduce externalPEMF stimulation remains unanswered thus far. Theseissues represent significant limitations for the extensiveand scientific application of PEMF in clinics.

In the present study, osteoclastic activities underPEMF stimulation with various intensities (0.5, 1, 2,and 3mT) were evaluated by using the murinemonocyte/macrophage RAW264.7 cell line. Weaimed to (i) examine whether the effects of PEMF onosteoclastic activities were intensity-dependent; (ii)investigate whether the osteoclastic activities wereinhibited or promoted under PEMF exposure withdifferent intensities; (iii) elucidate the exact pathsabout how PEMF regulated osteoclastic activities (by

inhibiting osteoclast formation or promoting osteo-clast apoptosis) together with relevant molecularmechanisms. We hypothesize in the present studythat the regulatory effects of PEMF exposure onRANKL-induced osteoclast formation, apoptosis, andbone resorbing ability in RAW264.7 cells are inten-sity-dependent.

MATERIALS AND METHODS

Cell Culture and Osteoclast Differentiation

The RAW264.7 cells were purchased from theCell Bank of the Type Culture Collection of ChineseAcademy of Sciences (Shanghai, China). For theosteoclastogenesis experiment, the cells were seededin six-well culture plate (Corning, Corning, NY) at1� 106 cells/ml and cultured in Dulbecco‘s ModifiedEagle medium (DMEM, Hyclone, Logan, UT) withhigh glucose (1�), supplemented with 10% (v/v) fetalbovine serum (GE Healthcare, Pittsburgh, PA) at 37 8Cin a humidified atmosphere of 95% air and 5% CO2.After overnight incubation, the RAW264.7 cells coulddifferentiate into osteoclast-like cells in the presence of50 ng/ml RANKL (Peprotech, Rocky Hill, NJ) duringthe following 7 days. The medium was refreshed every48 h.

Fig. 1. Schematic representation of PEMF generator togetherwith a Hemholz coil assembly with two-coil array. PEMF wave-form generated and utilized in this experiment comprises apulsed burst (burst width, 5 ms; pulse width, 0.2 ms; pulsewait, 0.02 ms; burst wait, 60 ms; pulse rise, 0.3ms; pulse fall,2.0ms) repeated at 15 Hz. Interval distance between two coils(20 cm diameter) was 10 cm, and turn number of enamel-coated copper wire was 80.

2 Wanget al.

Bioelectromagnetics

PEMF Stimulation

Cells were exposed to PEMF stimulation withdifferent intensities by using a PEMF exposure device(GHY-III, FMMU, Xi`an, China; China Patent no.ZL02224739.4). The device was composed of pulsedsignal generator and a Helmholz coils assembly withtwo-coil array (Fig. 1). The Helmholz coils wereplaced within an incubator during the period of PEMFexposure. The PEMF waveform generated and uti-lized in this experiment was comprised of a pulsedburst (burst width, 5ms; pulse width, 0.2ms; pulsewait, 0.02ms; burst wait, 60ms; pulse rise, 0.3ms;pulse fall, 2.0ms) repeated at 15Hz. This waveformhas proven to be effective on inhibiting in vivo boneloss in osteoporotic animals and enhancing in vitroosteoblastic activities in our previous studies [Jinget al., 2014; Wang et al., 2014; Zhai et al., 2016]. A2-Ω resistor was placed in series with the coils, andthe voltage drop across the resistor was observed withan oscilloscope. Based on the obtained current value,the peak intensity of the magnetic fields could becalculated. In addition, the control cultures wereplaced in other inactivated HelmHolz coils with thesame culture conditions. A Gaussmeter (Model 455DSP, Lake Shore Cryotronics, Westerville, OH) wasused to confirm the accuracy of the magnetic fieldmeasurement. The measured background electromag-netic field was 50� 2mT. In order to determine theinduced electric field within the coils, a custom-designed electrical potential detecting circular coil(5 cm coil diameter, 1mm coil diameter, 20 turns)was placed in the midcenter of the Helmholtz coilswith the coil parallel to the Helmholtz coils. Thecurrent detecting coil was connected with the oscillo-scope, and the induced peak electrical field wasdetermined to be approximately 0.5, 1, 2, and 3mV/cm for 0.5, 1, 2, and 3mT magnetic field intensity,respectively. The peak magnetic field exhibited

osteoclasts and apoptotic osteoclasts in either earlyor late stage were counted and analyzed forquantitative results among 20 randomly selectedosteoclast cells under the confocal microscope in

each group. The experiment was repeated threetimes.

Resorption Capability Determination

The sterilized fresh bovine femoral bone diskwith 5mm diameter was placed into the six-wellculture plate. Then, the RAW264.7 cells were seededinto each well. After overnight incubation, the cellswere incubated in the presence of RANKL (50 ng/ml)and treated by PEMF stimulation for 7 days. Then, thecells were removed from the bone disks by using anultrasonic clearing machine (Kejie, Shenzhen, China).The bone chips of each group were fixed in 2.5%glutaraldehyde in phosphate buffer for 24 h. Thespecimens were then dehydrated in graded series ofaqueous ethanol solutions of 50%, 70%, 90%, and100% ethanol for 1 h. After being air-dried, sampleswere mounted on aluminum SEM stubs with silverpaint and sputter-coated with gold using a coater. Thespecimens were examined using a scanning electronmicroscope (Hitachi JSM-4800, Tokyo, Japan). Theareas of 20 pits formed by osteoclasts were measuredusing Image-Pro Plus software (Media Cybernetics,Silver Spring, MD) in each group. The experimentwas repeated three times.

Fig. 2. Effects of 15 Hz PEMF exposure with various intensities (0.5, 1, 2, and 3mT) onRANKL-induced osteoclast differentiation on RAW264.7 cells. Cells were exposed to PEMFstimulation (2 h/day) with various intensities in the presence of RANKL (50 ng/ml) for 7 days,and then fixed and stained for tartrate-resistant acid phosphatase (TRAP). TRAP-positivemultinucleated cells were counted as osteoclasts in 20 random sights in each group. (A^E)Representative images showing osteoclast formation after cells exposed to PEMF stimulationwith various intensities (original magnification, 100�). (F) Statistical comparison of osteoclastnumber per version between Control group and PEMF groups with various intensities (0.5, 1,2, and 3mT). Values are all expressed as mean�S.D. (n¼ 20). �Significant difference fromControl group with P< 0.05.

TABLE 1. The Sequence of Primers Used in the PresentStudy for in Vitro Real-Time Fluorescence Quantitative PCR

Genes Primers Primer sequence (50–30)

TRAP Forward CGATCACAATCTGCAGTACCReverse ACCCAGTGAGTCTTCAGTCC

NFATc1 Forward CGCAAGTACAGTCTCAATGGReverse CAGGTATCTTCGGTCACACT

RANK Forward CACGGTGGATTCTGAGGGCTReverse GGGGAGGCAACTGTCACCTT

CTSK Forward AGAACGGAGGCATTGACTCTReverse GATGGACACAGAGATGGGTC

MMP-9 Forward GCACTGGGCTTAGATCATTCReverse GCTTAGAGCCACGACCATAC

Caspase-3 Forward GCATTGAGACAGACAGTGGGReverse AAGGGACTGGATGAACCACG

BAX Forward ATGGGCTGGACACTGGACTTCReverse TCTTCCAGATGGTGAGCGAGG

BCL-2 Forward TGTGTGGAGAGCGTCAACAGReverse CATCCCAGCCTCCGTTATCC

b-Actin Forward AGGCCAACCGTGAAAAGATGReverse TGGCGTGAGGGAGAGCATAG

4 Wanget al.

Bioelectromagnetics

Real-Time PCR

Total RNA was extracted from the culturedcells using TRIzol RNA isolation reagents (ThermoFisher Scientific, Waltham, MA) and quantifiedwith spectrophotometry (Bio-Rad, Hercules, CA).The cDNA was synthesized from 2mg of RNAusing FastQuant RT kit (Tiangen Biotech, Beijing,China). Real-time PCR was performed in a real-time PCR detection system (Bio-Rad) by using aMaxima SYBR Green qPCR kit (Thermo FisherScientific). The primers used in this study areshown in Table 1. After an initial denaturation at95 8C for 10min, the amplification reaction wasconducted for 40 cycles with annealing at 95 8C for15 s, annealing at 55 8C for 15 s, and extension at55 8C for 15 s. Gene expression levels were nor-malized to b-Actin using 2�DDCT analysis. Theexperiment was repeated three times.

Statistical Analysis

All data are presented as mean� standard devia-tion (S.D.). One-way analysis of variance (ANOVA)was performed for evaluating differences among thefive groups. Bonferroni’s post hoc analysis was used todetermine the significance between each two groups.

Differences were significant at P< 0.05. All statisticalanalyses were performed by using SPSS 13.0 software(SPSS, Chicago, IL).

RESULTS

TRAP Staining

The effects of 15Hz PEMF exposure withvarious intensities (0, 0.5, 1, 2, and 3mT) onosteoclastogenesis via TRAP staining onRAW264.7 cells are shown in Figure 2A–E. Ourresults demonstrate that osteoclastogenesis wassignificantly regulated by PEMF stimulation. Wefound that PEMF exposure at 0.5 and 1mT sup-pressed osteoclastogenesis, whereas PEMF stimula-tion with 3mT largely promoted the formation ofosteoclasts. Statistical comparisons further showthat the average number of formed osteoclasts perversion was significantly decreased by PEMF expo-sure with the 0.5mT group (P< 0.05, �55.6%) and1mT group (P< 0.05, �48.9%) as compared withthe Control group. We also observe that PEMFstimulation with 3mT significantly increased thenumber of osteoclasts as compared with the Controlgroup (P< 0.05, þ33.3%).

Fig. 3. Effects of 15 Hz PEMF exposure with various intensities (0.5, 1, 2, and 3mT) on osteo-clast cytoskeletal organization. RAW264.7 cells were exposed to different PEMF (2 h/day) atdifferent intensities in the presence of 50 ng/ml RANKL for 7 or 10 days. Cells were then fixedand stained with FITC-conjugated phalloidin (green) and DAPI (blue), and then imaged underconfocal microscope. Scale bars in A and B represent 10mm.

PEMFRegulate Osteoclast Activity 5

Bioelectromagnetics

Osteoclast Cytoskeletal Organization

The effects of 15Hz PEMF burst with variousintensities (0, 0.5, 1, 2, and 3mT) for 7 days and10 days on osteoclast cytoskeletal organization duringosteoclastogenesis via F-actin staining are shown inFigure 3. As shown in Figure 3A, smooth cellularoutlines with no filopodia were observed in theControl, 2mT, and 3mT PEMF groups. However, the0.5mT PEMF group exhibited many dendritic pseu-dopodia at the cell edge. Moreover, we also found

several thin filopodia around the cellular periphery inthe 1mT group. The osteoclasts exhibited reticularstructures of microfilaments when they had smoothedges, whereas osteoclasts with dendritic filopodiaexhibited chaotic cellular cytoskeletal structure. It isnotable that a podosome belt was formed in the 3mTPEMF stimulation group. After PEMF exposure for10 days (Fig. 3B), the morphology of most osteoclastsin the 0.5 and 1mT PEMF groups exhibited typicalmature osteoclast organizations with condensed nucleiand nuclear margin at the chromatin, which were

Fig. 4. Effects of 15 Hz PEMF exposure with various intensities (0.5, 1, 2, and 3mT) on osteo-clast apoptosis. RAW264.7 cells were cultured in confocal petri dishes and exposed to 2 h/dayPEMF for 7 days and 10 days, and stained with annexin V-FITC (green) and propidium iodide(red) for evaluating early apoptosis and late apoptosis, respectively, after PEMF exposurefor (A) 7 days and (B) 10 days. Scale bars represent 10mm. (C) Statistical comparisons ofoverall apoptosis rate, early apoptosis rate, and late apoptosis rate between Control groupand PEMF groups with various intensities (0.5, 1, 2, and 3mT). �Significant difference fromControl group with P< 0.05.

6 Wanget al.

Bioelectromagnetics

regarded as important characteristics of early apopto-sis via DAPI staining. However, osteoclasts in the 2and 3mT PEMF groups could not maintain their cellshape, and the cell nuclei and cell cytoskeleton fellinto cell fragmentation, which was a typical phenome-non of late-stage apoptosis.

Osteoclast Apoptosis

PEMF stimulation exhibited significant regu-latory effects on osteoclast apoptosis as shown inFigure 4. Most cells exposed to PEMF with 0.5and 1mT were non-apoptotic or in early stageapoptosis at day 7, whereas most osteoclasts in the2 and 3mT PEMF groups were in late-stageapoptosis as shown in Figure 4A. According tothe quantitative results, we found that 3mT PEMFgroup induced higher ratio of late apoptosis(P< 0.05) and relatively lower ratio of earlyapoptosis in osteoclasts as compared with theControl group. However, PEMF stimulation at0.5 mT showed relatively lower ratio of lateapoptosis and higher ratio of early apoptosis ascompared with the Control group (P< 0.05).

Bone Resorption Capability

To quantify the resorption ability of thederived osteoclasts, we determined the number andarea of pits by seeding osteoclasts on the boneslices (Fig. 5). In comparison with the Controlgroup, the 2 and 3mT PEMF groups showedrelatively higher pit area, whereas the 0.5 and1mT PEMF groups exhibited lower pit area(Fig. 5A–E). Statistical comparison demonstratesthat PEMF stimulation with 0.5mT induced signif-icantly lower average pit area than the Controlgroup (Fig. 5F, P< 0.05). However, PEMF expo-sure with 2 and 3mT significantly increasedaverage pit areas as compared with the Controlgroup (P< 0.05).

Osteoclastogenesis-Associated GeneExpression

Real-time PCR examination results for totalmRNA expression of RANK, cathepsin K (CTSK),matrix metal-loproteinase-9 (MMP-9), BAX,BCL-2, NFATc1, and TRAP are demonstrated inFigure 6. mRNA gene expression levels of RANK,

Fig. 5. Effects of 15 Hz PEMF exposure with various intensities (0.5, 1, 2, and 3mT) on osteo-clast-associated bone resorptive activity. RAW264.7 cells were cultured on bone slices andexposed to PEMF for 7 days. Samples were then examined under scanning electron micro-scope and photographed in the same magnification. Number and area of pits formed by osteo-clasts were measured using Image-Pro Plus software. (A^E) Representative scanningelectron microscope images after cells exposed to PEMF stimulation with various intensities.(F) Statistical comparison of average pit area between Control group and PEMF groups withvarious intensities (0.5, 1, 2, and 3mT). Values are all expressed as mean�S.D. (n¼ 20).�Significant difference from Control group with P< 0.05.

PEMFRegulate Osteoclast Activity 7

Bioelectromagnetics

TRAP, MMP-9, CTSK, BAX, and NFATc1 in the0.5mT group were significantly lower than theControl group (P< 0.05), and BCL-2 gene expressionlevels in the 0.5mT PEMF group were significantlyhigher than those in the Control group (P< 0.05).Significant increases were observed in the 2mTPEMF group in RANK, TRAP, and NFATc1 mRNAexpression (P< 0.05), but not in BAX/BCL-2 ratioand BCL-2, CTSK, and MMP-9 mRNA expression.Moreover, the 3mT PEMF group exhibited signifi-cantly higher RANK, TRAP, BAX, CTSK, MMP-9,and NFATc1 mRNA expression than the Controlgroup (P< 0.05).

DISCUSSION

Osteoclasts play a critical role in maintainingcalcium balance and regulating bone quantity andbone quality [Matsuo and Irie, 2008]. Abnormality ofosteoclast activities is associated with various congen-ital and metabolic diseases (e.g., osteoporosis, osteo-petrosis, periodontal disease, etc.). Thus, it has greatclinical significance. Low-intensity PEMF exposure,as a safe and non-invasive biophysical method, hasbeen proven to be effective for regulating boneturnover and promoting bone mass [Jing et al., 2013;Hong et al., 2014]. However, the exact mechanisms

Fig. 6. Effects of 15 Hz PEMF exposure with various intensities (0.5, 1, 2, and 3mT) on osteo-clastogenesis-related, bone resorption capability-related, and apoptosis-related gene ex-pression via real-time PCR analysis, including (A) RANK, (B) NFATc1, (C) TRAP, (D) CTSK,(E) MMP-9, (F) BAX, (G) BCL-2, and (H) BAX/BCL-2 ratio. Values are all expressed asmean�S.D. (n¼ 3) and relative expression level of each gene was normalized to b-actin.�Significant difference from Control group with P< 0.05.

8 Wanget al.

Bioelectromagnetics

for the regulation of PEMF on osteoclastogenesis aswell as optical exposure parameters of PEMF oninhibiting osteoclastic activities and functions remainunknown. In this study, we evaluated the effects of15Hz PEMF with various intensities (0.5, 1, 2, and3mT) on osteoclast activities and functions. We alsoquantified gene expression levels which were relatedwith osteoclastic formation, resorption capability, andapoptosis under PEMF exposure with different inten-sities. Our results reveal novel findings that boneresorbing capacity was significantly decreased by0.5mT PEMF mainly by inhibiting osteoclast forma-tion and maturation, but enhanced at 3mT by promot-ing osteoclast apoptosis. Moreover, our results alsodemonstrate that osteoclastogenesis-associated geneexpression was also suppressed by 0.5mT PEMF, butenhanced by 3mT PEMF. Considering the confirmedeffects of PEMF on regulating osteoclastogenesis[Chang et al., 2004; Chang et al., 2005; Zhang et al.,2017], no positive control (chemicals or physicalfactors which are already known to regulate osteoclastactivity and function) was used in our present study,which may be a limitation of our study. Our presentfindings reveal an important intensity window forlow-intensity PEMF in regulating bone resorptionwith diverse nature for modulating osteoclastogenesisand apoptosis, and can enrich our basic knowledge forunderstanding the mechanism of PEMF-mediatedinhibition of osteoclast activities.

The differentiation, maturation, and fusion ofosteoclasts as well as osteoclast-mediated boneresorption function are directly related to the integrityof the actin cytoskeleton [Jurdic et al., 2006]. Twomajor types of actin structures can be formed inosteoclasts depending on the substrate they spread:sealing zone on surfaces containing apatite crystalsfor bone resorption, and podosomes on surfaceswithout apatite crystals [Saltel et al., 2008].Podosomes are organized in clusters at the beginningof differentiation, followed by evolvement into dy-namic rings and final stabilization at the cell edge toform a podosome belt [Song et al., 2014; Ti et al.,2015]. The cell periphery of immature osteoclastsshows dendritic filopodial extensions, while matureosteoclasts exhibit smooth plasma membranes [Wanget al., 2015]. Mature osteoclasts can form a resorptivemicroenvironment consisting of a refuled border andan actin ring on the bone-cell interface. Lysosomalproteases, Hþ and Cl�, are secreted into the resorptionlacuna so that bone can be digested effectively[Teitelbaum, 2011; Boyce et al., 2012]. Our F-actinstaining results indicate that osteoclast formation canbe delayed under relatively low-intensity PEMFstimulation (0.5 and 1mT). It has been shown that

dendritic pseudopods of immature osteoclasts weretransformed into smooth-edged podosome belt inmature osteoclasts during cell fusion process on days4–6 [Song et al., 2014]. We thus observed that mostcells still exhibited dendritic filopodia in relativelylow-intensity (0.5 and 1mT) PEMF groups on day 7,indicating that these osteoclasts were still in theimmature stage. In contrast, F-actin cytoskeleton inrelatively high-intensity (2 and 3mT) PEMF groupsand the Control group showed typically matureosteoclast cellular morphology with smooth cell edgeand reticular intracellular structure [Ti et al., 2015].To investigate the longer-term change of these imma-ture osteoclasts, the experiment was expanded to10 days and osteoclastic cytoskeleton structurechanges were analyzed. Our results show that mostosteoclasts in the 0.5mT PEMF group exhibitedtypical mature osteoclastic cytoskeleton organization.Thus, these results indicate that the immature osteo-clasts at day 7 could grow into mature OC cells, andthis process was delayed but not interrupted.

Since the life spans of osteoclasts were three tofourfold shorter than osteoblasts, osteoclast apoptosiswas rarely observed in histological sections [Jobkeet al., 2014]. Osteoclasts displayed rapid cell death inthe absence of trophic factors (e.g., RANKL andM-CSF), which were caused by cellular apoptosis(programmed cell death) [Akiyama et al., 2008].Consistent with the results of cytoskeletal structureanalyses, our apoptosis examination results alsodemonstrate that PEMF exposure with relatively lowintensity (0.5mT) significantly inhibited osteoclastapoptosis, whereas the extent of apoptosis wasaggravated under PEMF exposure with 3mT. At day7, most osteoclasts exposed to relatively low-intensityPEMF stimulation (0.5mT) underwent early stageapoptosis. In contrast, most osteoclasts fell into cellfragmentation (a typical phenomenon of late-stageapoptosis). In conclusion, our results reveal intensity-dependent regulatory effects of PEMF on osteoclasticapoptosis and osteoclastic lifespan under PEMFstimulation.

Although we observed an obvious regulatoryeffect of PEMF exposure with different intensities onosteoclast differentiation and apoptosis, several previ-ous studies have substantiated that osteoclast numberand activity were not linearly correlated with boneresorption rate because of the involvement of non-resorbing osteoclasts in stimulating bone formation[Karsdal et al., 2007]. Thus, we quantified the boneresorptive function by seeding RANKL-inducedosteoclasts on bone slices and measuring the averagearea of pits they formed on bone slices. Our resultsrevealed that the osteoclast-induced bone resorption

PEMFRegulate Osteoclast Activity 9

Bioelectromagnetics

rates were attenuated in low-intensity (0.5mT) PEMFgroup, whereas 3mT PEMF exposure largely pro-moted osteoclast-related bone resorption abilities. Thebone resorption ability of osteoclasts was closelyrelated to the number, apoptosis, maturation, andmicrostructure of osteoclasts. It has been proven thatthe structure of actin filaments in osteoclasts contrib-uted to the attachment of bone surface, and thusinfluenced the bone resorption function [Furlan et al.,2007]. It can be assumed that the cytoskeletonstructure of some osteoclasts formed on the bone slicemay be less mature at day 7 in 0.5mT PEMF group aswell, so that bone resorption capability was recededwhen compared with other groups.

RANKL and RANK are crucial for regulatingosteoclast differentiation [Rubin et al., 1996], andNFATc1 is a master regulator of osteoclast differenti-ation that is activated by RANKL; it regulatesa number of osteoclast-specific gene expressionssuch as cathepsin K, osteoclast-associated receptor(OSCAR), and TRAP [Kim and Kim, 2014]. Toevaluate the bone-resorption capability of derivedosteoclasts, we detected the gene expression levelsof TRAP, CTSK, and MMP-9, which were bone-degrading enzymes secreted by activated osteoclasts[Teitelbaum, 2011]. We also determined the geneexpression of BCL-2 and BAX, which were apopto-sis-related genes: BCL-2 inhibited apoptosis andBAX promoted cell death [Adams and Cory, 2007].In this study, we found that gene expression levelsof RANK, NFATc1, TRAP, CTSK, BAX, and BAX/BCL-2 ratio were significantly decreased after ex-posed to 0.5mT PEMF; nonetheless these geneswere significantly increased when exposed to higher-intensity PEMF stimulation (3mT). The expressionlevel of BCL-2 was significantly increased whenexposed to 0.5mT PEMF. These findings suggest thatosteoclast differentiation, resorption ability, and apo-ptosis were largely inhibited by 0.5mT PEMF, butpromoted by 3mT PEMF.

In conclusion, the present study clearly demon-strates that 15Hz PEMF exposure with differentintensities (0, 0.5, 1, 2, and 3mT) exhibited signifi-cantly distinct regulatory effects on osteoclastogene-sis, and resorption capability of RAW264.7 cell lineinduced osteoclast-like cells. We demonstrate thatbone resorbing capacity was significantly decreasedby 0.5mT PEMF mainly by inhibiting osteoclastformation and maturation, but enhanced at 3mT bypromoting osteoclast apoptosis. Moreover, gene ex-pression of RANK, NFATc1, TRAP, CTSK, BAX,and BAX/BCL-2 was significantly decreased by0.5mT PEMF, but increased by 3mT. Our findingsreveal a significant intensity window for low-intensity

PEMF in regulating bone resorption with diversenature for modulating osteoclastogenesis and apopto-sis. These findings not only provide important experi-mental evidence for choosing the optimal PEMFexposure parameters, but also enrich our basic knowl-edge for understanding the underlying mechanismconcerning PEMF regulate bone turnover and pro-mote bone mass.

ACKNOWLEDGEMENTS

We thank Ying Liu and Tengrui Shi for theirtechnical assistance.

REFERENCES

Adams JM, Cory S. 2007. Bcl-2-regulated apoptosis: Mechanismand therapeutic potential. Curr Opin Immunol 19:488–496.

Akiyama T, Dass CR, Choong PF. 2008. Novel therapeuticstrategy for osteosarcoma targeting osteoclast differentia-tion, bone-resorbing activity, and apoptosis pathway. MolCancer Ther 7:3461–3469.

Barnaba SA, Ruzzini L, Di Martino A, Lanotte A, Sgambato A,Denaro V. 2012. Clinical significance of different effects ofstatic and pulsed electromagnetic fields on human osteo-clast cultures. Rheumatol Int 32:1025–1031.

Bassett CA, Pawluk RJ, Pilla AA. 1974. Acceleration of fracturerepair by electromagnetic fields. A surgically noninvasivemethod. Ann NY Acad Sci 238:242–262.

Bodamyali T, Bhatt B, Hughes FJ, Winrow VR, Kanczler JM,Simon B, Abbott J, Blake DR, Stevens CR. 1998. Pulsedelectromagnetic fields simultaneously induce osteogenesisand upregulate transcription of bone morphogenetic pro-teins 2 and 4 in rat osteoblasts in vitro. Biochem BiophysRes Commun 250:458–461.

Boyce BF, Rosenberg E, de Papp AE, Duong LT. 2012. Theosteoclast, bone remodelling and treatment of metabolicbone disease. Eur J Clin Invest 42:1332–1341.

Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A,Tosteson A. 2007. Incidence and economic burden ofosteoporosis-related fractures in the United States, 2005–2025. J Bone Miner Res 22:465–475.

Canalis E. 2013. Wnt signalling in osteoporosis: Mechanismsand novel therapeutic approaches. Nat Rev Endocrinol9:575–583.

Chang K, Chang HS, Yu YH, Shih C. 2004. Pulsed electromag-netic field stimulation of bone marrow cells derived fromovariectomized rats affects osteoclast formation and localfactor production. Bioelectromagnetics 25:134–141.

Chang K, Chang WH, Huang S, Huang S, Shih C. 2005. Pulsedelectromagnetic fields stimulation affects osteoclast forma-tion by modulation of osteoprotegerin, RANK ligand andmacrophage colony-stimulating factor. J Orthop Res 23:1308–1314.

Chang K, Chang WH, Tsai MT, Shih C. 2006. Pulsed electromag-netic fields accelerate apoptotic rate in osteoclasts. ConnectTissue Res 47:222–228.

Charles JF, Aliprantis AO. 2014. Osteoclasts: More than ‘boneeaters’. Trends Mol Med 20:449–459.

Furlan F, Galbiati C, Jorgensen NR, Jensen JE, Mrak E, RubinacciA, Talotta F, Verde P, Blasi F. 2007. Urokinase plasminogen

10 Wanget al.

Bioelectromagnetics

activator receptor affects bone homeostasis by regulatingosteoblast and osteoclast function. J Bone Miner Res 22:1387–1396.

Hannemann PF, Mommers EH, Schots JP, Brink PR, Poeze M.2014. The effects of low-intensity pulsed ultrasound andpulsed electromagnetic fields bone growth stimulation inacute fractures: A systematic review and meta-analysis ofrandomized controlled trials. Arch Orthop Trauma Surg134:1093–1106.

He J, Zhang Y, Chen J, Zheng S, Huang H, Dong X. 2015. Effectsof pulsed electromagnetic fields on the expression ofNFATc1 and CAII in mouse osteoclast-like cells. AgingClin Exp Res 27:13–19.

Hong JM, Kang KS, Yi HG, Kim SY, Cho DW. 2014.Electromagnetically controllable osteoclast activity. Bone62:99–107.

Jing D, Cai J, Wu Y, Shen G, Li F, Xu Q, Xie K, Tang C, Liu J,Guo W, Wu X, Jiang M, Luo E. 2014. Pulsed electromag-netic fields partially preserve bone mass, microarchitecture,and strength by promoting bone formation in hindlimb-suspended rats. J Bone Miner Res 29:2250–2261.

Jing D, Li F, Jiang M, Cai J, Wu Y, Xie K, Wu X, Tang C, Liu J,Guo W, Shen G, Luo E. 2013. Pulsed electromagneticfields improve bone microstructure and strength in ovariec-tomized rats through a Wnt/Lrp5/beta-catenin signaling-associated mechanism. PLoS ONE 8:e79377.

Jobke B, Milovanovic P, Amling M, Busse B. 2014. Bisphospho-nate-osteoclasts: Changes in osteoclast morphology andfunction induced by antiresorptive nitrogen-containingbisphosphonate treatment in osteoporosis patients. Bone59:37–43.

Jurdic P, Saltel F, Chabadel A, Destaing O. 2006. Podosome andsealing zone: Specificity of the osteoclast model. Eur J CellBiol 85:195–202.

Karsdal MA, Martin TJ, Bollerslev J, Christiansen C, Henriksen K.2007. Are nonresorbing osteoclasts sources of bone anabolicactivity? J Bone Miner Res 22:487–494.

Kim JH, Kim N. 2014. Regulation of NFATc1 in osteoclastdifferentiation. J Bone Metab 21:233–241.

Liu HF, Yang L, He HC, Zhou J, Liu Y, Wang CY, Wu YC, HeCQ. 2013. Pulsed electromagnetic fields on postmeno-pausal osteoporosis in Southwest China: A randomized,active-controlled clinical trial. Bioelectromagnetics 34:323–332.

Luben RA. 1991. Effects of low-energy electromagnetic fields(pulsed and DC) on membrane signal transduction pro-cesses in biological systems. Health Phys 61:15–28.

Matsuo K, Irie N. 2008. Osteoclast-osteoblast communication.Arch Biochem Biophys 473:201–209.

Reid IR. 2015. Short-term and long-term effects of osteoporosistherapies. Nat Rev Endocrinol 11:418–428.

Rubin J, McLeod KJ, Titus L, Nanes MS, Catherwood BD, RubinCT. 1996. Formation of osteoclast-like cells is suppressedby low frequency, low intensity electric fields. J OrthopRes 14:7–15.

Saltel F, Chabadel A, Bonnelye E, Jurdic P. 2008. Actincytoskeletal organisation in osteoclasts: A model to deci-pher transmigration and matrix degradation. Eur J Cell Biol87:459–468.

Song RL, Liu XZ, Zhu JQ, Zhang JM, Gao Q, Zhao HY, ShengAZ, Yuan Y, Gu JH, Zou H, Wang QC, Liu ZP. 2014. Newroles of filopodia and podosomes in the differentiationand fusion process of osteoclasts. Genet Mol Res 13:4776–4787.

Tabrah FL, Ross P, Hoffmeier M, Gilbert F. 1998. Clinical reporton long-term bone density after short-term EMF applica-tion. Bioelectromagnetics 19:75–78.

Teitelbaum SL. 2011. The osteoclast and its unique cytoskeleton.Ann NY Acad Sci 1240:14–17.

Ti Y, Zhou L, Wang R, Zhao J. 2015. Inhibition of microtubuledynamics affects podosome belt formation during osteo-clast induction. Cell Biochem Biophys 71:741–747.

Wang J, An Y, Li F, Li D, Jing D, Guo T, Luo E, Ma C. 2014.The effects of pulsed electromagnetic field on the functionsof osteoblasts on implant surfaces with different topogra-phies. Acta Biomaterialia 10:975–985.

Wang Y, Brooks PJ, Jang JJ, Silver AS, Arora PD, McCulloch CA,Glogauer M. 2015. Role of actin filaments in fusopodformation and osteoclastogenesis. Biochim Biophys Acta1853:1715–1724.

Weitzmann MN, Ofotokun I. 2016. Physiological and pathophysio-logical bone turnover—role of the immune system. Nat RevEndocrinol 12:518–532.

Zhai MM, Jing D, Tong SC, Wu Y, Wang P, Zeng ZB, Shen GH,Wang X, Xu QL, Luo EP. 2016. Pulsed electromagneticfields promote in vitro osteoblastogenesis through a Wnt/-catenin signaling-associated mechanism. Bioelectromag-netics 37:152–162.

Zhang J, Xu H, Han Z, Chen P, Yu Q, Lei Y, Li Z, Zhao M,Tian J. 2017. Pulsed electromagnetic field inhibitsRANKL-dependent osteoclastic differentiation inRAW264.7 cells through the Ca2 þ �calcineurin-NFATc1 signaling pathway. Biochem Biophys ResCommun 482:289–295.

PEMFRegulate Osteoclast Activity 11

Bioelectromagnetics

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具

http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/vip.htmlhttp://www.xuebalib.com/db.phphttp://www.xuebalib.com/zixun/2014-08-15/44.htmlhttp://www.xuebalib.com/