PulsedElectromagnetic FieldMayAccelerateinvitroEndochondralOssification

JueWang,1NaTang,2QiangXiao,1Li Zhang,1Yu Li,1Juan Li,1JunWang,1

Zhihe Zhao,1and LijunTan1*1StateKeyLaboratoryof OralDiseases,West ChinaHospitalof Stomatology, Sichuan

University, Chengdu, China2StomatologyDepartment of SichuanMedicalScienceAcademy, SichuanProvincial

People’sHospital, Chengdu, China

Recapitulation of embryonic endochondral bone formation is a promising alternative approach tobone tissue engineering. However, the time-consuming process is one of the reasons the approach isunpractical. Here, we aimed at accelerating the in vitro endochondral ossification process of tissueengineering by using a pulsed electromagnetic field (PEMF). The rat bone marrow-derived stemcells were chondrogenic or hypertrophic differentiated in a three-dimensional pellet culture system,and treated with different intensities of PEMF (1, 2, and 5mT with modulation frequency 750Hz,carrier frequency 75Hz and a duty ratio of 0.8, 3 h/day for 4 weeks). The effects of PEMF onhypertrophy and endochondral ossification were assessed by safranin O staining, immunohistochem-istry, and quantitative real-time polymerase chain reaction. The results suggest that PEMF at 1, 2,and 5mT may inhibit the maintenance of the cartilaginous phenotype and increase cartilage-specificextracellular matrix degradation in the late stage of chondrogenic differentiation. In addition, amongthe three different intensities, only PEMF at 1mT directed the differentiation of chondrogenic-induced stem cell pellets to the hypertrophic stage and promoted osteogenic differentiation. Ourfindings provide the feasibility to optimize the process of in vitro endochondral ossification withPEMF stimulation. Bioelectromagnetics 36:35–44, 2015. © 2014 Wiley Periodicals, Inc.

Key words: bone repair; bone marrow-derived stem cells; tissue engineering; cartilage;physical stimulation

INTRODUCTION

Attempts to engineer bone tissue generally focuson intramembranous bone formation. However, bonetissue engineering through intramembranous boneformation for clinical application is not successful,partly because of the lack of sufficient vasculatureduring initial stage [Meijer et al., 2007]. Therefore, theconcept of recapitulation of endochondral bone forma-tion has been investigated as a feasible approach[Sasaki et al., 2010; Scotti et al., 2010, 2013]. It is amore physiological process by which the embryonicintermediate cartilaginous template is graduallyreplaced by bone tissue. Because cartilage by nature isequipped to survive in hypoxic conditions, repair ofbone defects over cartilaginous templates may helpovercome the lack of vascularization [Coyle et al.,2009]. This will be of benefit especially for largertissue engineering constructs. However, this approachfor bone tissue engineering has been largely over-looked, mostly because of the long process length timeinvolved [Tortelli et al., 2010]. Thus, to achievefunctional and timely mechanical and morphologicaldevelopment, it is important to find strategies to

accelerate the hypertrophic and endochondral ossifica-tion process.

Pulsed electromagnetic field (PEMF) is a safe,non-invasive, clinically beneficial physical stimulus,which has been used clinically to promote the healingof fracture non-union and osteoarthritis [De Matteiet al., 2007; Griffin et al., 2011; Ryang et al., 2013].Previous studies have revealed that PEMF can enhancethe activity of osteoblasts and promote osteogenic

Grant sponsor: National Natural Science Foundation of China;grant numbers: 30900287, 1030034, 30900286.

Conflicts of interest: None.

*Correspondence to: Lijun Tan, State Key Laboratory of OralDiseases, West China Hospital of Stomatology, Sichuan Universi-ty, 14#, 3rd section, Renmin South Road, Chengdu 610041, China.E-mail: orthotan@gmail.com

Received for review 18 March 2014; Accepted 13 August 2014

DOI: 10.1002/bem.21882Published online 30 October 2014 in Wiley Online Library(wileyonlinelibrary.com).

Bioelectromagnetics 36:35^44 (2015)

� 2014 Wiley Periodicals, Inc.

differentiation of mesenchymal stem cells (MSCs),and can stimulate chondrocyte proliferation and extra-cellular matrix (ECM) synthesis; therefore, PEMF canbe used to control construct differentiation in tissueengineering [Sun et al., 2010; Mayer-Wagner et al.,2011; Zhou et al., 2011; Ongaro et al., 2012; Finiet al., 2013]. Indeed, PEMF is regarded as a conve-nient and effective way to alter the physical microenvi-ronment of bone tissue through electromagnetic effectrather than mechanical energy [Trock, 2000].

To assess the effects of PEMF on hypertrophyand endochondral ossification process, collagen type I,collagen type II, collagen type X, Sox9, bone sialopro-tein (BSP), transforming growth factor-b3 (TGF-b3),osterix (Osx), and runt-related transcription factor 2(Runx2) were used as the differentiation markers inthis study. During embryonic endochondral boneformation, MSCs condense and differentiate intoimmature chondrocytes. The chondrocytes then syn-thesize a matrix that mainly consists of collagen II andaggrecan. TGF-b3 stimulates production of ECMcomponents and is involved in the formation of newcartilage [Blaney Davidson et al., 2006]. Subsequently,the chondrocytes further differentiate and enter theprehypertrophic stage, producing Runx2 [Komori,2010]. Expression of Runx2 stimulates the upregula-tion of hypertrophic markers collagen X. Once colla-gen X is synthesized, the cells are differentiated intothe hypertrophic stage. During this final stage ofdifferentiation, the expression of chondrogenic genes,such as Sox9 and TGF-b3, is downregulated inhypertrophic chondrocytes as a necessary step toinitiate cartilage-bone transition [Hattori et al., 2010].The hypertrophic cells are now characterized byexpression of the osteogenic genes: collagen I, osterix,and BSP [Pelttari et al., 2006; Gawlitta et al., 2010].

In vitro chondrogenic potential of rat bonemarrow-derived stem cells (BMSC) has been docu-mented in monolayer cultures [Worster et al., 2006],three-dimensional scaffold cultures [Coleman et al.,2007], and pellet cultures [Muraglia et al., 2003]. Inthe present study, we constructed the in vitro chondro-genic and hypertrophic differentiation model with ratBMSC pellets and optimized PEMF intensity toaccelerate the in vitro endochondral ossification pro-cess and propose that the procedure is a potentialstrategy for bone formation and repair.

MATERIALS AND METHODS

Cell Culture

BMSC were harvested from the posterior tibiaand femur bone marrow of 2-week-old male Sprague-

Dawley rats (Experimental Animal Center of SichuanUniversity, Chengdu, China), as reported previously[Qi et al., 2009]. Passage 3 cells were used forsubsequent experiments. MSCs were identified asCD34 (�), CD45 (�), CD29 (þ), and CD44 (þ).

Chondrogenic and Hypertrophic Induction ofRat BMSC Pellets Culture

Passage 3 cells were dissociated, centrifuged,and resuspended to a concentration of 5� 105 cells/mlin chondrogenic medium (RASMX-90042; CyagenBiosciences, Guangzhou, China) containing 100ml/Ldexamethasone (RASMX-90042; Cyagen Biosciences),3ml/L ascorbate (RASMX-90042; Cyagen Biosciences),10ml/L Insulin-Transferrin-Selenium (ITS, RASMX-90042; Cyagen Biosciences), 1ml/L sodium pyruvate(RASMX-90042; Cyagen Biosciences), 1ml/L proline(RASMX-90042; Cyagen Biosciences), and 10ml/LTGF-b3 (RASMX-90042; Cyagen Biosciences). Cellswere then centrifuged at 150g for 5min and maintainedat 37 8C in a humidified atmosphere containing 5% CO2.After 24h, cells were aggregated into pellets, and cellpellets were carefully transferred to 24-well plates, fourpellets per well with 1ml chondrogenic medium(RASMX-90042; Cyagen Biosciences). The mediumwas changed every 2–3 days, and cells were cultured for4 weeks. In order to induce a hypertrophic phenotype,BMSC pellets were cultured for an additional week inhypertrophic medium lacking TGF-b3, but supple-mented with 50nM thyroxine (T1775; Sigma-Aldrich,St. Louis, MO), 7.0� 10�3M b-glycerophosphate(RASMX-90021; Cyagen Biosciences), 10�8M dexa-methasone (RASMX-90042; Cyagen Biosciences), and2.5� 10�4M ascorbic acid (RASMX-90042; CyagenBiosciences) after 3 weeks of chondrogenic induction[Mueller and Tuan, 2008; Scotti et al., 2010].

PEMF Stimulation

The intensity of the PEMF stimulus and theculture conditions in each seven experimental groupsof rat BMSC pellets are shown in Table 1. KDSC4000PEMF (Chengdu Miracle Chemical, Chengdu, China)stimulation was applied for 3 h/day for 4 weeks withmodulation frequency of 750Hz, carrier frequency of75Hz, and three different intensities of 1, 2, and 5mT(Fig. 1). The specific PEMF signal (75Hz/750Hz)was chosen based on previous studies [Fini et al.,2005]. The carrier frequency 75Hz and pulse duration1.3ms were commonly used in previous PEMFstudies. The modulation frequency was calculated andset to 750Hz to meet the requirement. The strength ofthe magnetic field (1, 2, or 5mT) was determined bythe magnetic field computing software (ChengduMiracle Chemical) associated with the PEMF device.

36 Wanget al.

Bioelectromagnetics

By entering the operating current and the sampleposition, the software was able to calculate themaximum field strength of certain position, andthe amplitude represents a peak-to-peak value.

Chondrogenically differentiated rat BMSC pel-lets were divided into three different groups based onthe intensity: group PS1, 1mT; group PS2, 2mT; andgroup PS3, 5mT. For group PS1, PS2, and PS3, thecultures were simultaneously treated with PEMF andchondrogenic medium over the 4-week period. GroupPSh was cultured in hypertrophic medium for 1 week

with PEMF stimulation of 1mT, after 3 weeks ofchondrogenic induction. Group PS1 pellets displayedthe most osteogenic effect by 1mT PEMF stimulationamong groups PS1, PS2, and PS3 in the first part ofour study. Therefore 1mT intensity was adopted inGroup PSh. Group C was the negative control consist-ing of uninduced pellets. Group N was the chondro-genic positive control consisting of pellets that werechondrogenic-induced for 4 weeks with no PEMFstimulation. Group Nh was the hypertrophic positivecontrol consisting of pellets that were cultured inhypertrophic medium for 1 week without PEMFstimulation, after 3 weeks of chondrogenic induction.All seven experimental groups were cultured in oneincubator. PEMF stimulation was applied in anotherincubator (no current to the coils for the controlgroups).

MTT Assay for Cell Proliferation Activity ofRat BMSC

Cell proliferation activity was determined usingthe 3-(4, 5-dimethylthiazol-2)-2, 5-diphenyltetrazoliumbromide (MTT) assay kit according to the manufac-turer’s protocol (Amresco, Solon, OH). BMSC wereseeded at 2� 104 cells/well in 200ml a-minimum essen-tial medium (a-MEM, A1049001; Gibco, Carlsbad,CA)/10% fetal bovine serum (FBS, F2442; Sigma–Aldrich, Grand Island, NY) and exposed to differentintensities of PEMF stimulation (1mT, 2mT, 5mT for3 h/day). A no-stimulation group was used as control.Samples (n¼ 10) were collected at the same time ondays 1, 2, 3, 4, and 5 of cell culture for the MTTcolorimetric test [Wang et al., 2013].

Immunohistochemical and HistochemicalStaining

The cell pellets were fixed in 4% paraformalde-hyde for 48 h, rinsed with distilled water, dehydrated ina graded series of ethanols, embedded in paraffin, andcut into 5mm thick sections that were collected onslides. The sections were subjected to hematoxylin eosinand safranin O-fast green staining. Expression of colla-gen type I, collagen type II, collagen type X, and BSPwere detected by immunohistochemical staining methodusing specific primary antibodies as follows: collagentype I (08A001277; MP Biomedicals, Solon, OH),collagen type II (08A001280; MP Biomedicals), colla-gen type X (ab7046; AbCam, Cambridge, UK), BSP(A4232.1/A4232.2; Immundiagnostik, Bensheim,Germany), as well as biotinylated secondary antibodies(Dako, Carpinteria, CA).The immunobinding wasdetected using the appropriate avidin-biotin complex(ABC PK-4000; Vector Laboratories, Peterborough,UK). Image-Pro Plus 6.0 software (Media Cybernetics,

TABLE 1. Experimental Groups, the Intensity of the PEMFStimulus and the Culture Conditions in Each SevenExperimental Groups of Rat BMSC Pellets

Group Procedure

Group C No inductionþNo PEMF stimulationGroup N CI for 4wkþNo PEMF stimulationGroup Nh CI for 3wkþHI for 1wkþNo PEMF stimulationGroup PS1 CI for 4wkþ 1mT PS for 4wkGroup PS2 CI for 4wkþ 2mT PS for 4wkGroup PS3 CI for 4wkþ 5mT PS for 4wkGroup PSh CI for 3wkþHI for 1wkþ 1mT PS for 4wk

Groups C, N, and Nh were control groups.CI, chondrogenic induction; HI, hypertrophic induction; PS,PEMF stimulation.

Fig. 1. Schematic representation of PFMF pulse protocol andPEMF generating device. A: The PEMF output waveform con-sisted of a pulsed burst repeated at 75 Hz (burst duration,13 ms;pulse duration,1.3 ms).B: The PEMF device was comprised of acontrol boxandamagnetic field unit.Themagnetic field unit wasplaced in an incubator (37 8C,5%CO2).BMSCpellets cultured in24-well plates were placed between the top and bottom coils ofthedevicefor thepulsedelectromagnetic fieldstimulationfor3 h/day.

PEMFStimulation of Endochondral Ossification 37

Bioelectromagnetics

Silver Spring, MD) was used for semi-quantitativeanalysis of immunohistochemical images using thefollowing scores: area, integrated optical density (IOD)and mean optical density (MOD¼ IOD SUM/areaSUM).

Real-Time Polymerase Chain Reaction Analysis

Twelve pellets per condition were homogenizedin 1ml Trizol reagent (Invitrogen, Carlsbad, CA).Equal amounts of RNA samples were reverse-tran-scribed using the SYBR PrimeScrip RT reagent Kitwith gDNA Eraser (TaKaRa, Dalian, China) accordingto the manufacturer’s protocol. Each real-time polymer-ase chain reaction (PCR) was carried out in triplicate inan ABI PRISM 7300 Fast Real-time System (AppliedBiosystems, Foster City, CA) using SYBR Premix ExTaq II (TaKaRa) with primers listed in Table 2. ThePCR amplification reaction was carried out for 40cycles by denaturing at 95 8C for 5 s, and annealing at60 8C for 30 s. b-actin was used for normalization.Value of the negative control group (group C) wasdesignated as 1.

Statistical Analysis

All experiments were performed a minimum ofthree times. All values are expressed as mean� stan-dard deviation of three biological repeats. Statisticalcomparisons were made using one-way analysis ofvariance, followed by multiple comparisons usingStudent–Newman–Keuls test (SNK test). A value ofP< 0.05 was considered statistically significant.

RESULTS

Effect of PEMF on Cell Proliferation

The MTT assay showed that PEMF stimulation(1mT, 2mT, 5mT) did not cause significant changes

in rat BMSC proliferation in two-dimensional culturesystem (Fig. 2). Due to the failure of control pellets toform a sufficiently robust ECM in stromal medium(group C), cell proliferation analysis was conductedwith monolayer cultured stem cells, and the data ofgroup C were thus not included in the remainingresults.

Histological Characteristics of PEMF-StimulatedStem Cell Pellets

In group N (4 weeks chondrogenic differentia-tion), regions with cartilaginous morphology werepresent, accompanied by central necrosis and aneosinophilic matrix, which accumulated mainly in theexternal zone. In groups PS1 (PEMF at 1mT) and PS2(PEMF at 2mT), chondrogenic pellets displayedcartilaginous features, including cells in lacunae em-bedded in an abundant matrix. Interestingly, in groupPS3, 5mT PEMF induced notable differences inmorphology, with prominent central necrosis and large

TABLE 2. Primer Sequences Used for Real-Time RT-PCR Analysis

Target gene Primer sequence Accession number

b-Actin Forward 50-ACGGTCAGGTCATCACTATCG-30 NM_031144Reverse 50-GGCATAGAGGTCTTTACGGATG-30

Sox9 Forward 50-TGCTCGGAACTGTCTGGAAACT-30 XM_003750950Reverse 50-GAGGAGGAGGGAGGGAAAACA-30

Col10a1 Forward 50-GATGCCTCTTGTCAGTGCTAACC-30 XM_001053056Reverse 50-GATCTTGGGTCATAGTGCTGCTG-30

Col1a Forward 50-GGCAAGACAGTCATCGAATACA-30 NM_053304Reverse 50-GATGGAGGGAGTTTACACGAAG-30

Runx2 Forward 50-GAAATGCCTCTGCTGTTATGAA-30 NM_001278483Reverse 50-AAAGTGAAACTCTTGCCTCGTC-30

Osx Forward 50-AAGTTCACCTGTCTGCTCTGCTC-30 NM_001037632Reverse 50-GGCTGATTGGCTTCTTCTTCC-30

TGF-b3 Forward 50-AGGTTTTCCGTTTCAATGTGTC-30 NM_013174Reverse 50-TTGGCTATGTGTTCATCAGGTC-30

Fig.2. GrowthcurvesofratBMSCfromcontrolandexperimentalgroups.1, 2, and 5mTrepresent variousdosesof PEMFinterven-tion.

38 Wanget al.

Bioelectromagnetics

empty areas in the ECM aligned beneath the peripheryof the pellet. In group Nh, the hypertrophic pelletsdisplayed a basophilic matrix and marked necrosiscentrally with a ring of peripheral cells. In group PSh,the hypertrophic pellets stimulated with 1mT PEMFshowed a scattered eosinophilic matrix without obvi-ous peripheral cells. In group C, control pellets grownin stromal medium failed to form a sufficiently robustECM (Fig. 3, left panel).

Expression of Chondrogenic Markers Affectedby PEMF

In group N, strong safranin O staining suggestedhigh glycosaminoglycans (GAG) expression, especial-ly in the region of necrosis. The safranin O stainingarea and intensity decreased with increasing intensityof PEMF stimulation. In group Nh, as chondrogenesisprogressed to hypertrophy, reduced level of GAG wasdetected. Higher safranin O staining was present ingroup PSh, the 1mT hypertrophic pellets (Fig. 3, rightpanel).

Collagen type II was detected in PEMF stimulatedpellets as well as in control group N. It was also detectedin group Nh between the internal necrosis area and theouter rim of elongate cells, whereas no sign of collagentype II was present in group PSh (Fig. 4).

Expression of Hypertrophic and OsteogenicMarkers Affected by PEMF

Type X collagen, a marker for chondrocytehypertrophic differentiation, was present throughoutgroup PS1 (1mT) compared with no expression ingroup N after 4 weeks’ chondrogenic induction. But itwas barely expressed in group PS2 (2mT) and groupPS3 (5mT). The hypertrophic group (Nh) was uni-formly positive for type X collagen, whereas only lowand localized levels of type X collagen were detectedin hypertrophic pellets exposed to 1mT PEMFstimulation (group PSh; Fig. 4).

In group N, group PS1, group Nh, and groupPSh, immunohistochemical analysis showed highlevels of homogeneous staining for type I collagen. Ingroups PS2 (2mT) and PS3 (5mT), type I collagenstaining intensity decreased significantly (Fig. 4).

BSP expression was enhanced in group PS1(1mT) compared with group N. It was confinedpredominantly to the outer rim underneath the superfi-cial layer of the pellets, with notably stronger intensityin the periphery of the lacunae and an overlap withtype X collagen. In group PS2 (2mT), decreased BSPexpression was detected compared to group PS1, whileBSP production was not detected in group PS3 (5mT).Group PSh displayed diffused deposition of BSP,while group Nh was negative for BSP (Fig. 4).

Fig. 3. Safranin red O-fast green staining of rat BMSC pellets.PS1, PS2, PS3: chondrogenic groups exposed to PEMF (PS1,1mT; PS2, 2 mT; PS3, 5 mT; 75 Hz frequency) for 3 h/day for4 weeks. Group N: chondrogenic positive control consisting ofpellets that were chondrogenic-induced for 4 weeks with noPEMF stimulation. PSh: cultured in hypertrophic medium for1weekwith1mTPEMF, after 3 weeksof chondrogenic induction.Group Nh: hypertrophic positive control consisting of pelletsthat were cultured in hypertrophic medium for 1 week withoutPEMF stimulation, after 3 weeks of chondrogenic induction.Theasteriskand thearrowindicate centralnecrosisandeosinophilicmatrixthathadaccumulatedmainlyintheexternalzone, respec-tively;Scalebar¼100 mm.

PEMFStimulation of Endochondral Ossification 39

Bioelectromagnetics

Effect of PEMF Treatment on Chondrogenic-and Osteogenic-Related Gene ExpressionProfile of Rat BMSC

In PS groups, PEMF led to a significant decreasein Sox9 and TGF-b3 mRNA expression compared

with group N. A similar effect was seen in group PShcompared with group Nh (Figs. 5 and 6).

Consistent with the immunohistochemistry obser-vations, PEMF treatment significantly induced moreCol10a1 mRNA expression in PS groups comparedwith group N, but significantly lower expression was

Fig. 4. Effect of PEMF treatments on chondrogenic- and osteogenic-related protein expressionin rat BMSC pellets. A: immunohistochemistry results of collagen II, collagen I, collagen X, andBSPinBMSCpelletswithPEMFstimulation.Scalebar¼10mm.B:MeanopticaldensityofcollagenII, collagen I, collagenX, andBSPinimmunohistochemicalstainingof PEMF-stimulatedchondro-genicgroups. �P< 0.05whencomparedtogroupN.C:Meanopticaldensityofcollagen II, collagenI, collagen X, and BSPin immunohistochemical stainingof PEMF-stimulatedhypertrophicgroup.�P< 0.05whencomparedtogroupNh.

40 Wanget al.

Bioelectromagnetics

detected in group PSh compared with group Nh(Figs. 5 and 6).

Real-time RT-PCR showed that expression ofRunx2, a gene critical for chondrocyte maturation andosteoblast differentiation, was upregulated in PS groupsover controls (group N), whereas it was inhibited by1mT PEMF treatment in the group PSh compared withgroup Ph. Gene expression level of osterix, the genethat controls osteoblast lineage commitment and subse-quent osteoblast differentiation, was significantly in-creased by PEMF treatment in both PS groups andgroup PSh (Figs. 5 and 6).

DISCUSSION

Defining the molecular mechanism of the effectof PEMF on chondrogenic and hypertrophic differenti-ation of BMSC has potential value in bone engineer-ing. The MTT assay was used to investigate whetherselected intensities of PEMF stimulation affectedBMSC proliferation at the early stage of preliminarilyexpansion. No significant change was detected com-pared to the controls.

As to the effect of PEMF on chondrogenesis,lower level of GAG (principally a measure of aggre-can) and collagen type II protein expression wasdetected, as well as a significant decrease in Sox9 andTGF-b3 mRNA expression in PEMF groups (PSgroups). This suggests that PEMF may inhibit themaintenance of the cartilaginous phenotype and pro-mote cartilage specific ECM degradation in chondro-genically differentiated rat BMSC pellets [Hattoriet al., 2010].

A decreased collagen type II protein expressionwas also detected in BMSC pellets imposed withPEMF stimulation and hypertrophic medium simulta-neously (PSh group). However, increased GAG expres-sion was detected in PSh group compared with Nhgroup. This indicates that PEMF may inhibit orinterfere with aggrecan (GAG) degradation in thepresence of hypertrophic medium. As the two majorcomponents of cartilage ECM, many proteolyticenzymes are responsible for the degradative events ofcollagen type II and aggrecan, such as matrix metal-loproteinases and the aggrecanases [Ortega et al., 2004;Mackie et al., 2008]. PEMF may have influenced the

Fig. 5. Effect of PEMF stimulation on chondrogenic and osteogenic gene expression in chondro-genicratBMSCpellets.Quantitativereal-timeRT-PCRanalysisofchondrogenic-andosteogenic-relatedgeneexpressionlevelsusingtotalRNAisolated fromPEMF-stimulatedchondrogenicpel-lets. �P< 0.05whencomparedtogroupN(Y-axis:2�DDCt).

PEMFStimulation of Endochondral Ossification 41

Bioelectromagnetics

function of some proteases assisting with removal ofaggrecan. However, there is no literature on this aspectcurrently and further investigations are needed.

Despite the inhibitory effect on chondrogenicphenotype maintenance, PEMF is crucial in promotingthe hypertrophic phase of endochondral ossification.As the only known hypertrophic chondrocyte-specificmarker, the expression level of collagen X is increasedwhen the immature chondrocytes differentiate intohypertrophic chondrocytes, and decreased when thehypertrophic chondrocytes differentiate into osteo-blasts [Zheng et al., 2003; Gawlitta et al., 2010]. In thepresent study, in group PS1 (1mT PEMF), there werehigher levels of collagen X gene and protein thangroup N, indicating that certain intensity PEMFpromoted the hypertrophic differentiation of BMSCpellets. Collagen X was strongly expressed in groupNh but weakly in group PSh, implying that PEMFmight promote hypertrophic BMSC to differentiateinto osteoblasts. PEMF has played a bidirectionalregulation role during endochondral ossification on

collagen type X and accelerated the hypertrophicstage.

Promotion of the final stage of endochondralossification by 1mT PEMF stimulation was shown bya significant increase in expression of BSP, collagentype I, and osterix in response to both chondrogenic(group PS1) and hypertrophic (group PSh) cues.However, no such effects were detected when cellswere exposed to PEMF of 2mT and 5mT. Therefore,certain PEMF stimulus may promote the in vitrohypertrophic and endochondral ossification process,and the intensity of 1mT was optimal in the presentstudy. If wider amplitude range was studied, theoptimal amplitude may lay outside the amplitudestested in the present study.

In summary, we investigated the effects of PEMFon hypertrophy and endochondral ossification of ratBMSC pellets in the late chondrogenic and hypertro-phic stage. PEMF at 1mT inhibited the maintenanceof the chondrocyte phenotype in chondrogenic differ-entiated BMSC pellets, caused the earlier onset of

Fig. 6. Effect of PEMF stimulationon chondrogenic andosteogenic gene expression in hypertro-phicdifferentiatedratBMSCpellets.Quantitativereal-timeRT-PCRanalysisofchondrogenic-andosteogenic-relatedgeneexpressionlevelsusingtotalRNAisolatedfromPEMF-stimulatedhyper-trophicpellets. �P< 0.05whencomparedtogroupNh(Y-axis:2�DDCt).

42 Wanget al.

Bioelectromagnetics

hypertrophy, and accelerated the endochondral ossifi-cation process in hypertrophic chondrocytes. However,the term “endochondral ossification” is usually re-ferred to an in vivo process of bone formation, whichconsists of certain well-established steps. The presentstudy only mimicked one of these processes. Asthe bone formation process cannot be accomplishedin vitro without the invasion of blood vessels,osteoblasts, and bone marrow cells, further investiga-tion of the application of PEMF is necessary todetermine whether similar effects will be achieved invivo and to determine the underlying mechanisms.

REFERENCES

Blaney Davidson EN, Vitters EL, van der Kraan PM, van den BergWB. 2006. Expression of transforming growth factor-beta(TGFbeta) and the TGFbeta signaling molecule SMAD-2Pin spontaneous and instability-induced osteoarthritis: Rolein cartilage degradation, chondrogenesis and osteophyteformation. Ann Rheum Dis 65:1414–1421.

Coleman RM, Case ND, Guldberg RE. 2007. Hydrogel effects onbone marrow stromal cell response to chondrogenic growthfactors. Biomaterials 28:2077–2086.

Coyle CH, Izzo NJ, Chu CR. 2009. Sustained hypoxia enhanceschondrocyte matrix synthesis. J Orthop Res 27:793–799.

De Mattei M, Fini M, Setti S, Ongaro A, Gemmati D, Stabellini G,Pellati A, Caruso A. 2007. Proteoglycan synthesis in bovinearticular cartilage explants exposed to different low-frequen-cy low-energy pulsed electromagnetic fields. OsteoarthritisCartilage 15:163–168.

Fini M, Giavaresi G, Carpi A, Nicolini A, Setti S, Giardino R.2005. Effects of pulsed electromagnetic fields on articularhyaline cartilage: Review of experimental and clinicalstudies. Biomed Pharmacother 59:388–394.

Fini M, Pagani S, Giavaresi G, De Mattei M, Ongaro A, Varani K,Vincenzi F, Massari L, Cadossi M. 2013. Functional tissueengineering in articular cartilage repair: Is there a role forelectromagnetic biophysical stimulation? Tissue Eng Part BRev 19:353–367.

Gawlitta D, Farrell E, Malda J, Creemers LB, Alblas J, Dhert WJ.2010. Modulating endochondral ossification of multipotentstromal cells for bone regeneration. Tissue Eng Part B Rev16:385–395.

Griffin XL, Costa ML, Parsons N, Smith N. 2011. Electromagneticfield stimulation for treating delayed union or non-union oflong bone fractures in adults. Cochrane Database Syst Rev4:CD008471.

Hattori T, Müller C, Gebhard S, Bauer E, Pausch F, Schlund B,Bösl MR, Hess A, Surmann-Schmitt C, von der Mark H, deCrombrugghe B, von der Mark K. 2010. SOX9 is a majornegative regulator of cartilage vascularization, bone marrowformation and endochondral ossification. Development137:901–911.

Komori T. 2010. Regulation of bone development and extracellularmatrix protein genes by RUNX2. Cell Tissue Res 339:189–195.

Mackie EJ, Ahmed YA, Tatarczuch L, Chen KS, Mirams M. 2008.Endochondral ossification: How cartilage is converted intobone in the developing skeleton. Int J Biochem Cell Biol40:46–62.

Mayer-Wagner S, Passberger A, Sievers B, Aigner J, Summer B,Schiergens TS, Jansson V, Müller PE. 2011. Effects of lowfrequency electromagnetic fields on the chondrogenicdifferentiation of human mesenchymal stem cells. Bioelec-tromagnetics 32:283–290.

Meijer GJ, de Bruijn JD, Koole R, van Blitterswijk CA. 2007.Cell-based bone tissue engineering. PLoS Med 4:e9.

Mueller MB, Tuan RS. 2008. Functional characterization ofhypertrophy in chondrogenesis of human mesenchymalstem cells. Arthritis Rheum 58:1377–1388.

Muraglia A, Corsi A, Riminucci M, Mastrogiacomo M, CanceddaR, Bianco P, Quarto R. 2003. Formation of a chondro-osseous rudiment in micromass cultures of human bone-marrow stromal cells. J Cell Sci 116:2949–2955.

Ongaro A, Pellati A, Setti S, Masieri FF, Aquila G, Fini M, CarusoA, De Mattei M. 2012. Electromagnetic fields counteractIL-1b activity during chondrogenesis of bovine mesenchy-mal stem cells. J Tissue Eng Regen Med [Epub ahead ofpublication].

Ortega N, Behonick DJ, Werb Z. 2004. Matrix remodelingduring endochondral ossification. Trends Cell Biol 14:86–93.

Pelttari K, Winter A, Steck E, Goetzke K, Hennig T, Ochs BG,Aigner T, Richter W. 2006. Premature induction of hypertro-phy during in vitro chondrogenesis of human mesenchymalstem cells correlates with calcification and vascular invasionafter ectopic transplantation in SCID mice. Arthritis Rheum54:3254–3266.

Qi MC, Zou SJ, Han LC, Zhou HX, Hu J. 2009. Expression ofbone-related genes in bone marrow MSCs after cyclicmechanical strain: Implications for distraction osteogenesis.Int J Oral Sci 1:143–150.

Ryang We S, Koog YH, Jeong KI, Wi H. 2013. Effects of pulsedelectromagnetic field on knee osteoarthritis: A systematicreview. Rheumatology 52:815–824.

Sasaki J, Matsumoto T, Egusa H, Matsusaki M, Nishiguchi A,Nakano T, Akashi M, Imazato S, Yatani H. 2010. In vitroreproduction of endochondral ossification using a 3Dmesenchymal stem cell construct. Integr Biol (Camb)4:1207–1214.

Scotti C, Tonnarelli B, Papadimitropoulos A, Scherberich A,Schaeren S, Schauerte A, Lopez-Rios J, Zeller R, BarberoA, Martin I. 2010. Recapitulation of endochondral boneformation using human adult mesenchymal stem cells as aparadigm for developmental engineering. Proc Natl AcadSci 107:7251–7256.

Scotti C, Piccinini E, Takizawa H, Todorov A, Bourgine P,Papadimitropoulos A, Barbero A, Manz MG, Martin I.2013. Engineering of a functional bone organ throughendochondral ossification. Proc Natl Acad Sci 110:3997–4002.

Sun LY, Hsieh D-K, Lin P-C, Chiu H-T, Chiou T-W. 2010. Pulsedelectromagnetic fields accelerate proliferation and osteogen-ic gene expression in human bone marrow mesenchymalstem cells during osteogenic differentiation. Bioelectromag-netics 31:209–219.

Tortelli F, Tasso R, Loiacono F, Cancedda R. 2010. The develop-ment of tissue-engineered bone of different origin throughendochondral and intramembranous ossification followingthe implantation of mesenchymal stem cells and osteoblastsin a murine model. Biomaterials 31:242–249.

Trock DH. 2000. Electromagnetic fields and magnets: Investiga-tion treatment of musculoskeletal disorders. Rheum DisClin North Am 26:51–62.

PEMFStimulation of Endochondral Ossification 43

Bioelectromagnetics

Wang Y, Wang J, Bai D, Song J, Ye R, Zhao Z, Lei L, Hao J, JiangC, Fang S, An S, Cheng Q, Li J. 2013. Cell proliferation ispromoted by compressive stress during early stage ofchondrogenic differentiation of rat BMSCs. J Cell Physiol228:1935–1942.

Worster AA, Brower-Toland BD, Fortier LA, Bent SJ, Williams J,Nixon AJ. 2006. Chondrocytic differentiation of mesenchy-mal stem cells sequentially exposed to transforming growthfactor-b1 in monolayer and insulin-like growth factor-I in athree-dimensional matrix. J Orthop Res 19:738–749.

Zheng Q, Zhou G, Morello R, Chen Y, Garcia-Rojas X,Lee B. 2003. Type X collagen gene regulation byRunx2 contributes directly to its hypertrophic chondrocyte–specific expression in vivo. J Cell Biol 162:833–842.

Zhou J, Ming LG, Ge BF, Wang JQ, Zhu RQ, Wei Z, Ma HP, XianCJ, Chen KM. 2011. Effects of 50Hz sinusoidal electro-magnetic fields of different intensities on proliferation,differentiation and mineralization potentials of rat osteo-blasts. Bone 49:753–761.

44 Wanget al.

Bioelectromagnetics