Bioelectromagnetics32:443^452 (2011)

InvolvementofMidkineExpressionin theInhibitoryEffectsof Low-FrequencyMagnetic

FieldsonCancerCells

TingtingWang,1YunzhongNie,1Shuli Zhao,1YuwangHan,2 Youwei Du,3 andYayi Hou1*1ImmunologyandReproductionBiologyLab,MedicalSchoolandStateKeyLaboratory

of PharmaceuticalBiotechnology, NanjingUniversity, Nanjing, China2College of Science, NanjingUniversity ofTechnology, Nanjing, China

3NationalLaboratoryof SolidMicrostructures, NanjingUniversity, Nanjing, China

Effects of magnetic fields (MFs) on cancer cells may depend on cell type and exposure conditions.Gene expression levels are different among cancer cells. However, the effect of MFs on cancer cellswith different gene expressions is still unclear. In this study, the cancer cell lines BGC-823, MKN-45,MKN-28, A549, SPC-A1, and LOVO were exposed to a low-frequency MF. Specific parameters ofMFs were determined. Furthermore, the potential of the MF to influence cancer cell growth withmidkine (MK) expression was evaluated. Cell proliferation and cell cycle were detected using theCCK-8 assay and flow cytometry. Cell ultrastructure was observed by transmission electron micro-scopy. BGC-823 cells with over-expression of MK (BGC-MK cells) and stanniocalcin-1 weregenerated by plasmid construction and transfection. Results showed that exposure to a 0.4-T,7.5 Hz MF inhibited the proliferation of BGC-823, MKN-28, A549, and LOVO cells, but notMKN-45 and SPC-A1 cells. Moreover, the inhibitory effect of the MF on BGC-MK cells was lower(12.3%) than that of BGC-823 cells (20.3%).Analysis of the cell cycle showed that exposure to theMFled to a significant increase in the S phase in BGC-823 cells, but not in BGC-MK cells. In addition,organelle morphology was modified in BGC-823 cells exposed to the MF. These results suggest thatexposure to a 0.4-T, 7.5 Hz MF could inhibit tumor cell proliferation and disturb the cell cycle. Thealteration of MK expression in cancer cells may be related to the inhibitory effect of the MF on thesecells. Bioelectromagnetics 32:443–452, 2011. � 2011 Wiley-Liss, Inc.

Key words: low-frequency magnetic fields; cancer cells; midkine; cell cycle

INTRODUCTION

Electromagnetic devices are widely used to studythe effects of various magnetic fields (MFs) on cancers.Continuous exposure to a 60 Hz sinusoidal magneticfield could induce apoptosis of prostate cancer cells viathe generation of reactive oxygen [Koh et al., 2008].The inhibitory effects of MFs on tumor cell prolifer-ationwere observed in nudemice bearing subcutaneoushuman colon adenocarcinomas [Tofani et al., 2002] andbreast cancer [Williams et al., 2001]. A marked anti-tumor activitywas observed inmice inoculatedwith theEhrlich ascites carcinoma when treated with static andextremely low-frequency alternating MF (1–16.5 Hz,100–300 nT) [Novikov et al., 2009]. However, chronicexposure to 20 kHz triangular MFs did not appear tocause toxicity in rats [Lee et al., 2010]. Breast cancercell proliferation was only slightly affected by a mod-erate-intensity (6 mT) static MF [Dini and Abbro,2005]. Possible mutagenic and carcinogenic effectsof MFs have also been discussed and tested. Meta-analysis of data showed that approximately half ofthe cited studies showed a tumor-inductive effect of

MFs, while the remaining half showed that MFs inhib-ited tumor proliferation in vitro and in vivo [Juutilainenet al., 2006; Juutilainen, 2008]. MF exposure maystimulate or inhibit cell growth depending on the stateof the cells [Olsson et al., 2001]. These data suggest thatinhibitory or stimulant effects of MFs on cell prolifer-ation may depend on cell type and exposure conditions.

Grant sponsors: Scientific Research Foundation of the GraduateSchool of Nanjing University; State Key Laboratory of Pharma-ceutical Biotechnology (ZZYJ-SN-200805)

*Correspondence to: Yayi Hou, Immunology and ReproductiveBiology Lab, Medical School and State Key Laboratory of Phar-maceutical Biotechnology, Nanjing University, Nanjing 210093,China. E-mail: yayihou@nju.edu.cn

Received for review 24 May 2010; Accepted 15 January 2011

DOI 10.1002/bem.20654Published online 24 February 2011 in Wiley Online Library(wileyonlinelibrary.com).

� 2011Wiley-Liss,Inc.

Some investigators have tried to elucidate themechanism of MFs for their anti-tumor effect. Itwas found that the effects of MFs on cell growth proc-esses may be linked to increased calcium signaling[Walleczek and Liburdy, 1990], dependent on receptorsthat mediate signal transduction cascades [Uckun et al.,1995], or attributed to the increased expression ofgrowth regulatory genes [Lin et al., 1994]. MFs couldalso act as a co-inductor of DNA damage [Ruiz-Gomezand Martınez-Morillo, 2009]. Extremely low-fre-quency MFs induced reversible brain DNA damagebut did not elicit the stress response [Mariucci et al.,2010]. Although studies have demonstrated that geneexpression levels are different between cancer tissuesand normal tissues, the effects of MFs on cancer cellswith different gene expression patterns is still unclear.Moreover, accumulating experimental evidencesuggests that anti-tumor effects of MFs occur onlywithin specific frequency and amplitude windows,and may be dependent on several physical parameters[Verheyen et al., 2003]. Thus, the effect ofMF exposureon different types of cancer cells, cells with differentialexpression of some genes, and MFs with specific fre-quency and amplitude needs to be clarified.

Midkine (MK), a heparin-binding growth factor,is highly expressed in various malignant tumors includ-ing gastric cancer [Huang et al., 2007].MK can activatethe stanniocalcin-1 (STC-1) gene in gastric cancerBGC-823 cells. High expression of STC-1 maypromote carcinogenesis and angiogenesis [Gerritsenet al., 2002]. Based on the important value of MKand STC-1 in tumor development, we hypothesize thatMK or STC-1 expression levels may influence the stateof cancer cells during exposure to MFs. Therefore, inthis study we evaluated the proliferative changes ofdifferent cancer cells during exposure to MFs. Weestablished BGC-MK gastric carcinoma cells (parentalBGC-823 cells with increased expression of MK) andBGC-STC1 cells (BGC-823 cells with increasedexpression of STC-1) using plasmid construction andtransfection. The association between MK or STC-1expression and MFs effect was examined. In addition,MFs with specific frequency and amplitude wereselected.

MATERIALS AND METHODS

Cell Culture and Cell Transfection

BGC-823 (poorly differentiated human gastricadenocarcinoma cell line), MKN-45 (poorly differen-tiated human gastric cancer cell line), MKN-28 (well-differentiated human gastric cancer cell line), A549(human lung cancer cell line), SPC-A1 (human lung

adenocarcinoma cell line), and LOVO (human colonadenocarcinoma cell line) cells were obtained from theShanghai Institute of Cell Biology (Shanghai, China).

The construction of pcDNA3.1/MK andpcDNA3.1/STC1 were described in our previous work[Wang et al., 2008a]. The plasmids were transfectedusing Lipofectamine 2000 reagent (Invitrogen, Carls-bad, CA) according to the manufacturer’s instructions.Briefly, approximately 8 � 104 cells/well were grownovernight in 24-well plates. When cells reached 90–95% confluence, they were transfected with 0.8 mgpcDNA3.1/MK, pcDNA3.1/STC1, or pcDNA3.1 inserum-free medium using Lipofectamine 2000. After4 h incubation at 37 8C, 400 ml RPMI-1640 medium(Gibco, Carlsbad, CA) with 10% FBS was added.Stable transfectants were selected in the presence of400 mg/L G418 (Amersco, Solon, OH) during 2 weeksof culture. All these cells were grown in RPMI-1640mediumwith 10% fetal bovine serum (FBS) and 100 U/ml penicillin–streptomycin at 37 8C in a water-satu-rated atmosphere with 5% CO2.

Experimental Magnetic Field

Two pairs of fan-shaped NdFeB permanent mag-nets (N45, Innuovo, Dongyang, China; Fig. 1B) wereattached to a circular iron plate and arranged to establishMF (Fig. 1A). The bottom two magnets rotated atcertain frequency driven by a step motor, which wascontrolled using a functional signal generator. The toptwo magnets rotated synchronously due to the strongmagnetic interaction. Magnetic flux density wasmeasured at the target site using a gauss meter(HT201, Hengtong, Shanghai, China). As shown inFigure 1C, MF at the target site is alternative pulseswith a maximum flux density of about 0.4 T. Thefrequency of MF was 0–15 Hz. This instrument wasfabricated by the National Laboratory of Solid Micro-structures, Nanjing University (Nanjing, China). Theentire magnetic apparatus was located in a hood withhumidity and temperature controller. Control cells wereplaced in a similar apparatus except that there were tworotating iron plates instead of magnets, thus lacking aMF.

Magnetic Field Exposure and Cell Viability Assay

The anti-proliferative effect of MFs was deter-mined using the Cell Counting Kit-8 (CCK-8) assay.Briefly, exponentially growing cells were seeded onto a96-well plate (Costar, Carlsbad, CA; 8 � 103 cells/welland at least six replicatewells). Cellswere kept growingand exposed to MFs with different magnetic fluxdensities, frequencies, and exposure times. Identical,nonenergized exposure chambers were used for shamexposure of control cells in the same room. Twenty

444 Wangetal.

Bioelectromagnetics

microliters CCK-8 solution (Dojindo, Kumamoto, Ja-pan) was added to each well, and the cells were furtherincubated at 37 8C for another 3.5 h. Absorbance wasdetermined using a multi-well spectrophotometer(BioTek, Winooski, VT) at 450 nm (A450). Inhibitoryrate (IR) ¼ (A45Control � A45MF)/A45Control � 100%.All data are mean values from three independentexperiments.

Transmission Electron Microscopy (TEM)

Exponentially growing BGC-823 cells wereseeded onto 10 cm dishes (Costar). After intermittentexposure to MFs for 4 days (2 h/day), cells were har-vested, washed twice, fixed with 4% glutaraldehyde for2 h, post-fixed with 1% OsO4, dehydrated in gradedconcentrations of ethanol [Cai et al., 2007], and thenembedded in resin SPI-Pon 812 (Shell Chemical,Yuhuan, China). Ultra-thin sections were cut(80 nm), counterstained with lead citrate and uranylacetate, and then observed with a Philips CM 12 elec-tron microscope (Philips, Eindhoven, Holland) at80 kV.

RNA Extraction and Reverse Transcription-Polymerase Chain Reaction (RT-PCR)

Cells were collected by centrifugation (300g,5 min) and total RNA was extracted using Trizol

reagent (Invitrogen) following the manufacturer’sinstructions. One microgram total RNA was used tosynthesize cDNA using random primers and Prime-script reverse transcriptase (Takara, Tokyo, Japan).Two microliters of the incubation mixture was usedas the template for the following PCR using the Taqenzymemix kit (Takara). Primers were synthesized andused as before [Wang et al., 2008a]. PCR was carriedout for 28 or 30 cycles of denaturation (30 s at 94 8C),annealing (40 s at 55 8C), and extension (30 s at72 8C). PCR products were detected on 1% agarosegel containing 0.5 mg/L ethidiumbromide. The gelwasphotographed on aUV-transilluminator, and target genesignal was standardized against theb-actin signal usinga digital imaging and analysis system (SmartSpec Plus,Bio-Rad, Hercules, CA).

Western Blot

Cells (1 � 107) were lysed in a buffer containing50 mmol/LTris–Cl (pH8.0), 150 mmol/LNaCl, 0.02%NaN3, 0.1% SDS, 100 mg/L phenylmethylsufonylfluoride (PMSF), 1 mg/L aprotinin, and 1% Triton.Protein was quantified using the Lowry assay. A70 mg protein sample of each was suspended in aloading buffer (0.5 M Tris–HCl 5 ml, 10% glycerol1 ml, 10% SDS 1 g, 5% b-mercaptoethanol 0.5 ml,0.05% blue bromophenol 0.5 ml, and water 3 ml),

Fig. 1. Magneticfieldexposuresystem(A).Twopairsoffan-shapedNdFeBpermanentmagnetswerearranged to establishmagnetic fields (B).MFat the target site isalternativepulseswithamaximumfluxdensityofabout 0.4 T (C).

GeneExpressionandMFEffectonCancerCell 445

Bioelectromagnetics

resolved on 10% polyacrylamide gels, and electrophor-etically transferred to nitrocellulose membranes. Mem-branes were blocked in 5% skim milk and 1% bovineserum albumin. Immunoblotting was performed usingantibodies specific for MK, STC1, and b-actin (Sigma,St. Louis, MO) as primary reagents. A horseradishperoxidase (HRP)-conjugated goat anti-rabbit antibody(1:2000 BA1054, Boster, Beijing, China) was used asthe secondary reagent. Protein bands were detectedusing enhanced chemiluminescence (ECL).

Cell Cycle Analysis

Analysis of the cell cycle was performed using aCycle Test Plus DNA Reagent Kit (Becton Dickinson,San Jose, CA). After exposure to MFs, cells wereharvested, washed twice with PBS and stainedwith 25 mg/ml propidium iodide solution. Flow cytom-etry analysis was performed as described previously[Wang et al., 2008b].

Statistical Evaluation

Values were expressed as mean � standard devi-ations. Statistical analysis was performed using theStudent’s t-test. Values of P < 0.05 were consideredto be statistically significant. Statistical analysis wasdone using the SPSS 11.5 software program (SPSS,Chicago, IL).

RESULTS

Inhibitory Effect of Magnetic Fields on DifferentCancer Cells

Wefirst examined the inhibitory effect of this low-frequency MF with different parameters on BGC-823cells. The maximum magnetic flux density of our MFexposure system was 0.4 T and the frequency of MFwas 0–15 Hz. We found that the inhibitory effect ofMFs onBGC-823 cell growthwas enhanced by increas-ing the magnetic flux density from 0.05 to 0.4 T(Fig. 2A) and increasing the frequency from 0 to7.5 Hz. However, the inhibitory effect was similar from7.5 to 15 Hz (Fig. 2B). Additionally, the inhibitoryeffect of this MF on BGC-823 cells was also time-dependent (Fig. 2C). After exposure to MFs for 2 h/day and intermittently for 0.5, 1 and 2 days, no inhi-bition was observed in BGC-823 cell growth. Eventhough the inhibitory effect of MFs was observed at4, 5, and 6 days, no significant change was found.Therefore, we selected a 0.4 T, 7.5 Hz MF for furtheranalysis, with cells intermittently exposed for 4 days(2 h/day).

Next, we evaluated the inhibitory effect of the0.4 T, 7.5 Hz MF on BGC-823, MKN-28, A549,LOVO, MKN-45, and SPC-A1 cell lines using the

CCK-8 assay. We found that MFs could significantlyinhibit the proliferation of BGC-823, MKN28, A549,and LOVO cells, while no significant effect wasdetected on MKN-45 and SPC-A1 cells (Fig. 3). Theseresults suggest thatMFsmay have a selective inhibitoryeffect on specific cancer cells.

Ultrastructural Changes of BGC-823 Cells AfterTreatment With Magnetic Fields

TheBGC-823 cellmodel iswell established in ourlab [Wang et al., 2008a], and well suited for studies of

Fig. 2. Inhibitory effect of MF exposure on BGC-823 cells withdifferent stimulation conditions. BGC-823 cells were treated withdifferentmagneticfluxdensities(A),differentmagneticfrequencies(B) and different number of days (C). Data points are mean val-ues � standarddeviationsof threeseparateexperiments.

446 Wangetal.

Bioelectromagnetics

ultrastructural changes induced by extrinsic factors.BGC-823 cells exposed to MFs (0.4 T, 7.5 Hz for2 h/day) were therefore examined using transmissionelectronmicroscopy (TEM). Compared to control cells,the cytoplasm of treated cells was characterized bymoderate edema with mitochondrial disorganization,accompanied by degeneration or hypercondensation ofcristae, expansion of the endoplasmic reticulum (ER),villous changes on cytoplasmic surface, appearance ofmyelin-like corpuscles, and decreased electron densityof nuclear chromatin (Fig. 4). These data indicate thatMFs can directly alter the ultrastructure of BGC-823cells, which suggest that the inhibitory effect ofMFs oncancer cells was associated with their intactultrastructures.

Inhibitory Effect of Magnetic Fields onGene-Modified Cancer Cells

To evaluate MK and STC1 expression level inBGC-823 cells and gene-modified BGC-823 cells, RT-PCR andWestern blotting were performed.When com-pared with the parental cells and pcDNA3.1 transfectedcells (BGC-3.1 cells), transfection of BGC-823 cellswith pcDNA3.1/MK (BGC-MK cells) or pcDNA3.1/STC-1 (BGC-STC1 cells) significantly enhanced MKor STC1 expression in BGC-823 cells. These resultsindicated that transfection of pcDNA3.1/MK andpcDNA3.1/STC1 was successful (Fig. 5).

The inhibitory effect of our MF system on thesegene-modified cells was further evaluated using theCCK-8 assay. As expected, MF inhibited the growthof BGC-823, BGC-3.1, and BGC-STC1 cells, and theinhibitory rates were 20.3%, 19.7%, and 18.8%,respectively. However, the inhibitory rate in BGC-MK cells was just 12.3% (Fig. 6), which represents asignificant decrease compared to either parental, con-trol, or STC1-expressing cells. This suggests that over-

expressed MK may change the inhibitory effect of MFon cell growth in BGC-823 cells.

Effect of Magnetic Fields on Cell Cycle inGene-Modified Cancer Cells

MFsmay delay cell cycle progression, thus affect-ing cell growth [Santini et al., 2003]. We examined theeffect of this 0.4 T, 7.5 Hz MF on the cell cycle. With-out exposure to MFs, the S-phase cell populations ofBGC-823, BGC-3.1, and BGC-MK cells were 40.1%,41.3%, and 43.2%, respectively. Exposure toMFs led toa significant increase in the cell population in the Sphase from 40.1% to 47.5% in BGC-823 cells, and41.3% to 47.0% in BGC-3.1 cells. However, no signifi-cant change in the cell cycle was found in BGC-MKcells (Fig. 7). During the S phase of the cell cycle, DNAreplication occurs and the amount of DNA in the cellhas effectively doubled. These results suggest that theS-phase arrest effect of MFs was eliminated in BGC-MK cells.

DISCUSSION

Considerable interest has been focused on theinfluence of MFs on biological systems. Despite dec-ades of research, explanations remain elusive as to themolecular basis of the effects of MFs on cancers. In thisstudy, we showed that a 7.5 Hz MF at a flux density of0.4 T could selectively inhibit the proliferation ofdifferent types of cancer cells. With a focus on gastriccancer BGC-823 cells, we found that the inhibitoryeffect of MFs could be changed by gene modifications.In addition, exposure to this MF influenced themorphology of the organelles in BGC-823 cells.

The rotating MFs used in our study were selectedfrom various possible combinations of frequency,polarization, and flux density. Our results showed thatthe inhibitory effect of MFs on BGC-823 cell growthwas time-dependent, concordant with the increase inthe magnetic flux density from 0.05 to 0.4 T and thefrequency from 0 to 7.5 Hz. Therefore, a 0.4-T, 7.5 HzMF was selected for further analysis. Moreover, inclinical treatment different cancers may display adifferent response to the same treatment. MFs may alsopossess a specific inhibitory effect on some cancer celllines. Therefore, we studied the effects of the MF onBGC-823,MKN-28,A549, LOVO,MKN-45, and SPC-A1 tumor cell lines.After intermittent exposure to a 0.4-T, 7.5 Hz rotating MF for 4 days (2 h/day), the pro-liferation rates of BGC-823, MKN-28, A549, andLOVO cells, but not MKN-45 or SPC-A1 cells, weresignificantly inhibited. In our present study, MKN-28and MKN-45 are both gastric adenocarcinoma celllines. MKN-28 is highly differentiated while MKN-

Fig. 3. Inhibitory effect of MF exposure on different cancer cells.Datapointsaremeanvalues � standarddeviationsofthreesepar-ateexperiments(��P < 0.01,���P < 0.001,comparedwithcontrolcells).

GeneExpressionandMFEffectonCancerCell 447

Bioelectromagnetics

Fig. 4. UltrastructuralchangesofBGC-823cellsexposedtoMF.Cellultrastructure(A:BGC-823cells,original magnification 500�; B: BGC-823 cells,1000�; C: BGC-823 cells exposed to 0.4 T, 7.5 HzMF,500�; andD:BGC-823 cellsexposed to 0.4 T,7.5 HzMF,1000�) wasobservedby transmissionelectron microscopy. Compared with control cells, the cytoplasm of treated cells is characterizedbymitochondrial disorganization, accompanied by degeneration or hypercondensation of cristae(bluearrows),myelin-like corpusclesappearance (redarrow), expansionofendoplasmic reticulum(greenarrows), anddecreasedelectrondensityofchromatinin thenucleus.

Fig. 5. Verification of transfected BGC-823 cells. RT-PCR (A,B) and Western blotting (C) analysisof the expressionsofMKandSTC-1in BGC-823 cellsafter transfection (A,1:BGC-823; 2:BGC-823/vector; 3: BGC-823/MK; B,1: BGC-823; 2: BGC-823/vector; 3: BGC-823/STC-1; C,1: BGC-23/MK;2:BGC-23/STC-1; 3:BGC-823/vector; 4:BGC-823).

448 Wangetal.

Bioelectromagnetics

45 is a poorly differentiated cell line. With their oppos-ing responses to MFs, we speculate that the inhibitoryeffect of MFs depends on the differentiation degree ofcancer cells. Supporting this concept, similar resultswere reported by Tenuzzo et al. [2006]. Furthermore,BGC-823 is a poorly differentiated human gastricadenocarcinoma cell line, A549 is a human lung cancercell line, and LOVO is a human colon adenocarcinomacell line. All these cell lines showed a sensitive responseto MF exposure, which suggests a potential clinicalapplication of MFs in cancer treatment. Therehave been numerous investigations of possible electro-magnetic field effects on cancer cell proliferation.Tofani et al. [2001] reported that static and ELF MFs(1 mTand 50 Hz) could induce tumorgrowth inhibitionand apoptosis in two transformed cell lines, humancolon adenocarcinoma and human breast adenocarci-noma. Raylman et al. [1996] also found that exposure toa strong static MF could slow the growth of humanmelanoma, ovarian, and lymphomacancer cells invitro.Yamaguchi et al. [2006] showed that pulsed magneticstimulation could inhibit tumor development andregulate immune functions in murine models. Theseresults suggest that the biological effects of MFs maydepend on the cell type and exposure conditions.

Gene modifications could change the inhibitoryeffect of MFs on cells. The inhibitory effect of MFs onthe two types of MKN-45 and MKN-28 gastric cancercells was different, and different expressions of MK inMKN-45 and MKN-28 gastric cancer cells have beenreported [Ono et al., 2005]. Moreover, over-expressedMKmay change the resistance of the BGC-823 cells tothe inhibitory effects of MFs on cell growth. Our

previous studies indicated that over-expressed and trun-catedMKcould promote proliferation amongBGC-823cells in vitro and tumor growth in vivo [Wang et al.,2008a]. MK, a heparin-binding growth factor, wasdiscovered by screening for factors that mediate reti-noic acid-induced cell differentiation. Many pieces ofevidence show that MK expression is increased in lung,liver, breast, and gastric cancers [Kadomatsu andMuramatsu, 2004; Obata et al., 2005; Kaifi et al.,2007; Ruan et al., 2007]. Aberrant MK expression isfrequently associated with tumorigenesis. These resultstaken together suggest that different expressions ofMKmay influence the role of MFs on cancer cells.

Stanniocalcin 1 (STC-1) formerly called hypocal-cin or teleocalcin, is a 50 kDa disulfide-linked homo-dimeric glycoprotein that was originally identified infish and secreted from the corpuscles of Stannius[Filvaroff et al., 2002]. Over-expression of STC-1 intransgenic mouse models leads to high serum phos-phate, dwarfism, increased vascular density, mitochon-drial hypertrophy, and increased rates of respiration[Varghese et al., 2002]. STC-1 is differentiallyexpressed in breast, ovary, colon and hepatocellulartumors, and corresponding normal tissues. Gerritsenet al. [2002] demonstrated that the expression ofSTC-1 was increased in colon cancers primarilythrough tumor vasculature, since STC-1 was highlyexpressed during angiogenesis. Nonetheless, despitea clear role of MK in our studies, STC-1 was notsufficient to protect BGC-823 cells from the inhibitoryeffects of MFs. Thus, MK appears to act via alternativedownstream mediators.

MFsmay change the rate of cell cycle progression.There are reports on the effects of MFs on cell cycles.Cell division occurred when a 5-mT, 50 Hz MF wasallowed to act from 48 to 72 h [Antonopoulos et al.,1995]. Static MFs inhibit the proliferation of culturedmyoblast cells [Kim and Im, 2010]. Rat bone marrowstem cells exposed to acute radiation before encounter-ing a static MF showed a significant increase in thenumber of cells in the G(2)/M phase [Sarvestani et al.,2010]. Our experiments showed that MFs exerted ananti-proliferative action, which was concordant with adelay in the progression of cells from theG1/G0 to the Sphase of the cell cycle. Recent research has proved thatMKmay play fundamental roles in the regulation of celldifferentiation and development. However, in BGC-MK cells, the S-phase arrest effect of MFs was elim-inated. The consideration may be that MK was onlytransfected into BGC-823 cells and did not integrateinto the DNA of BGC-823 cells. Some studies revealedthat MFs could act as a co-inductor of DNA damage[Ruiz-Gomez and Martınez-Morillo, 2009] and in vivoELF-MFs induced reversible brain DNA damage while

Fig. 6. Over-expressionofMKchanged theinhibitoryeffect ofMFoncellgrowth.Datapointsaremeanvalues � standarddeviationsof three separate experiments (��P < 0.01, comparedwith BGC-823 cells).

GeneExpressionandMFEffectonCancerCell 449

Bioelectromagnetics

they did not elicit the stress response [Mariucci et al.,2010].

The BGC-823 cell line exhibited significant ultra-structural changes following exposure to the MF,including cytoplasmic edema and mitochondrial swel-ling. The expansion of ER and the decreased electrondensity of chromatin in the nucleus were also observed.Similar effects had been previously reported [Dini and

Abbro, 2005; Teodori et al., 2006]. Since the mitochon-dria are critical for oxidative respiration and the ERregulates stress responses and protein assembly, ultra-structural changes may indirectly account for changesin cell proliferation and/or apoptosis.

In conclusion, our present study demonstrates thatlow-frequency MF exposure could inhibit tumor cellproliferation and disturb the cell cycle and cell

Fig. 7. S-phasearrest effect ofMFwaseliminatedin BGC-MK cells.A:Cellcycle stageof BGC-823cellsandgene-modifiedBGC-823cellsdeterminedusingflowcytometry.B:Cellpopulationsindiffer-ent phaseswere analyzed.Data points aremean values � standard deviations of three separateexperiments (��P < 0.01and ���P < 0.001, comparedwith controlcells).

450 Wangetal.

Bioelectromagnetics

ultrastructure—at least under the conditions of ourexperimental protocol. The alteration ofMKexpressionmay be related to the inhibitory effect of this MF oncancer cells. The inhibitory effect of MFs on cells maybe due to cell-specific differences in differentiation orgene expression. Further study will be necessary toaddress the detailed effects of MFs on cancer cellsincluding ion channel functions and signal transductionpathways. Nonetheless, based on these data, MFs mayrepresent a potentially useful adjunct therapy for theclinic.

ACKNOWLEDGMENTS

We thank Professor Dwayne G. Stupack of theMoores UCSD Cancer Center and Dr. Kathleen LaS-ance of the Cincinnati Children’s Hospital MedicalCenter for proofreading and editing.

REFERENCES

Antonopoulos A, Yang B, Stamm A, Heller WD, Obe G. 1995.Cytological effects of 50 Hz electromagnetic fields onhuman lymphocytes in vitro. Mutat Res 346:151–157.

Cai Y, Liu Y, Yan W, Hu Q, Tao J, Zhang M, Shi Z, Tang R. 2007.Role of hydroxyapatite nanoparticle size in bone cell pro-liferation. J Mater Chem 17(36):3780–3787.

Dini L, Abbro L. 2005. Bioeffects of moderate-intensity staticmagnetic fields on cell cultures. Micron 36(3):195–217.

Filvaroff EH, Guillet S, Zlot C, Bao M, Ingle G, Steinmetz H,Hoeffel J, Bunting S, Ross J, Carano RA, Powell-Braxton L,Wagner GF, Eckert R, Gerritsen ME, French DM. 2002.Stanniocalcin 1 altersmuscle and bone structure and functionin transgenic mice. Endocrinology 143:3681–3690.

Gerritsen ME, Soriano R, Yang S, Ingle G, Zlot C, Toy K, Winer J,Draksharapu A, Peale F, Wu TD, Williams PM. 2002.In silico data filtering to identify new angiogenesis targetsfroma large invitro gene profiling data set. PhysiolGenomics10:13–20.

Huang Y, Cao G, Wang H, Wang Q, Hou Y. 2007. The expressionand location of midkine in gastric carcinomas of Chinesepatients. Cell Mol Immunol 4:135–140.

Juutilainen J. 2008. Do electromagnetic fields enhance the effectsof environmental carcinogens? Radiat Prot Dosimetry132:228–231.

Juutilainen J, Kumlin T, Naarala J. 2006. Do extremely low fre-quency magnetic fields enhance the effects of environmentalcarcinogens? A meta-analysis of experimental studies. Int JRadiat Biol 82(1):1–12.

Kadomatsu K, Muramatsu T. 2004. Midkine and pleiotrophin inneural development and cancer. Cancer Lett 204:127–143.

Kaifi JT, Fiegel HC, Rafnsdottir SL, Aridome K, Schurr PG, Reich-elt U, Wachowiak R, Kleinhans H, Yekebas EF, Mann O,Ichihara-Tanaka K, Muramatsu T, Kluth D, Strate T, IzbickiJR. 2007.Midkine as a prognosticmarker for gastrointestinalstromal tumors. J Cancer Res Clin Oncol 133:431–435.

Kim S, Im W. 2010. Static magnetic fields inhibit proliferation anddisperse subcellular localization of gamma complex protein3in cultured C2C12 myoblast cells. Cell Biochem Biophys57(1):1–8.

Koh EK, Ryu BK, Jeong DY, Bang IS, Nam MH, Chae KS. 2008.A 60-Hz sinusoidal magnetic field induces apoptosis ofprostate cancer cells through reactive oxygen species. Int JRadiat Biol 84(11):945–955.

Lee HJ, Gimm YM, Choi HD, Kim N, Kim SH, Lee YS. 2010.Chronic exposure of Sprague–Dawley rats to 20 kHztriangular magnetic fields. Int J Radiat Biol 6(5):384–389.

Lin H, Goodman R, Henderson AS. 1994. Specific region of the c-myc promoter is responsive to electric and magnetic fields.J Cell Biochem 54:281–288.

Mariucci G, Villarini M, Moretti M, Taha E, Conte C, Minelli A,Aristei C, Ambrosini MV. 2010. Brain DNA damage and 70-kDa heat shock protein expression in CD1 mice exposed toextremely low frequency magnetic fields. Int J Radiat Biol86(8):701–710.

Novikov VV, Novikov GV, Fesenko EE. 2009. Effect of weakcombined static and extremely low-frequency alternatingmagnetic fields on tumor growth in mice inoculated withthe Ehrlich ascites carcinoma. Bioelectromagnetics 30(5):343–351.

Obata Y, Kikuchi S, Lin Y, Yagyu K,Muramatsu T, Kumai H. 2005.Serummidkine concentrations and gastric cancer. Cancer Sci96:54–56.

Olsson G, Belyaev IY, Helleday T, Harms-Ringdahl M. 2001. ELFmagnetic field affects proliferation of SPD8/V79 Chinesehamster cells but does not interact with intrachromosomalrecombination. Mutat Res 493(1–2):55–66.

Ono HA, Davydova JG, Adachi Y, Takayama K, Barker SD, Rey-nolds PN, Krasnykh VN, Kunisaki C, Shimada H, Curiel DT,Yamamoto M. 2005. Promoter-controlled infectivity-enhanced conditionally replicative adenoviral vectors forthe treatment of gastric cancer. J Gastroenterol 40(1):31–42.

Raylman RR, Clavo AC, Wahl RL. 1996. Exposure to strong staticmagnetic field slows the growth of human cancer cells invitro. Bioelectromagnetics 17(5):358–363.

Ruan M, Ji T, Wu Z, Zhou J, Zhang C. 2007. Evaluation ofexpression of midkine in oral squamous cell carcinomaand its correlation with tumour angiogenesis. Int J OralMaxillofac Surg 36:159–164.

Ruiz-Gomez MJ, Martınez-Morillo M. 2009. Electromagneticfields and the induction of DNA strand breaks. ElectromagnBiol Med 28(2):201–214.

SantiniMT,Rainaldi G, FerranteA, Indovina PL,Vecchia P,DonelliG. 2003. Effects of a 50 Hz sinusoidal magnetic field on celladhesion molecule expression in two human osteosarcomacell lines (MG-63 and Saos-2). Bioelectromagnetics 24(5):327–338.

Sarvestani AS, Abdolmaleki P, Mowla SJ, Ghanati F, Heshmati E,Tavasoli Z, Jahromi AM. 2010. Static magnetic fieldsaggravate the effects of ionizing radiation on cell cycleprogression in bone marrow stem cells. Micron 41(2):101–104.

Tenuzzo B, Chionna A, Panzarini E, Lanubile R, Tarantino P,Di Jeso B, Dwikat M, Dini L. 2006. Biological effects of6 mT static magnetic fields: A comparative study in differentcell types. Bioelectromagnetics 27(7):560–577.

Teodori L, Albertini MC, Uguccioni F, Falcieri E, Rocchi MB,Battistelli M, Coluzza C, Piantanida G, Bergamaschi A,Magrini A, Mucciato R, Accorsi A. 2006. Static magneticfields affect cell size, shape, orientation, and membranesurface of human glioblastoma cells, as demonstrated byelectron, optic, and atomic force microscopy. Cytometry A69(2):75–85.

GeneExpressionandMFEffectonCancerCell 451

Bioelectromagnetics

Tofani S, Barone D, Cintorino M, de Santi MM, Ferrara A,Orlassino R, Ossola P, Peroglio F, Rolfo K, Ronchetto F.2001. Static and ELF magnetic fields induce tumor growthinhibition and apoptosis. Bioelectromagnetics 22(6):419–428.

Tofani S, Cintorino M, Barone D, Berardelli M, De Santi MM,Ferrara A, Orlassino R, Ossola P, Rolfo K, Ronchetto F,Tripodi SA, Tosi P. 2002. Increased mouse survival, tumorgrowth inhibition and decreased immunoreactive p53 afterexposure tomagnetic fields. Bioelectromagnetics 23(3):230–238.

Uckun FM, Kurosaki T, Jin J, Jun X, Morgan A, Takata M, Bolen J,Luben R. 1995. Exposure of B-lineage lymphoid cells to lowenergy electromagnetic fields stimulates Lyn kinase. J BiolChem 270:27666–27670.

Varghese R,Gagliardi AD,Bialek PE,Yee SP,WagnerGF,DiMattiaGE. 2002. Overexpression of human stanniocalcin affectsgrowth and reproduction in transgenic mice. Endocrinology143:868–876.

Verheyen GR, Pauwels G, Verschaeve L, Schoeters G. 2003. Effectof coexposure to 50 Hz magnetic fields and an aneugen on

human lymphocytes, determined by the cytokinesis blockmicronucleus assay. Bioelectromagnetics 24(3):160–164.

Walleczek J, Liburdy RP. 1990. Nonthermal 60 Hz sinusoidal mag-netic-field exposure enhances 45Ca2þ uptake in rat thymo-cytes: Dependence on mitogen activation. FEBS Lett 271:157–160.

Wang QL, Wang H, Zhao SL, Huang YH, Hou YY. 2008a. Over-expressed and truncated midkines promote proliferation ofBGC823 cells in vitro and tumor growth in vivo. World JGastroenterol 14(12):1858–1865.

Wang TT, Wei J, Qian XP, Ding YT, Yu LX, Liu BR. 2008b.Gambogic acid, a potent inhibitor of survivin, reverses doce-taxel resistance in gastric cancer cells. Cancer Lett 262:214–222.

Williams CD, Markov MS, Hardman WE, Cameron IL. 2001.Therapeutic electromagnetic field effects on angiogenesisand tumor growth. Anticancer Res 21(6A):3887–3891.

Yamaguchi S, Ogiue-Ikeda M, Sekino M, Ueno S. 2006. Effects ofpulsed magnetic stimulation on tumor development andimmune functions in mice. Bioelectromagnetics 27(1):64–72.

452 Wangetal.

Bioelectromagnetics