RCYKH

ATatact((wmirdmZamfsdegarlD4

KCp

cSS

ptm0

Ed

Basic Research—Technology

4

ole of Free Radicals and Metal Ions in Directurrent-Induced Cytotoxicity

uko Nakamura, DDS,* Keiso Takahashi, DDS, PhD,*azue Satoh, PhD,‡ Akiko Shimetani, DDS,* Hiroshi Sakagami, PhD,† andirofumi Nishikawa, DDS, PhD*

IhnomDm

pdoDm

baapft

mTctAgA

R

nf5JCT

E

(I

bstracthe purpose of this study was to investigate the mech-nism of direct current (DC)-induced cytotoxicity. Toest the working hypothesis that electrolysis productsre responsible for the DC-induced cytotoxicity, theytotoxic effects between the direct and indirect DCreatment against human polymorphonuclear cellsPMNs) was compared. The indirect DC treatmenttreatment with the culture medium exposed to DC)as comparable in cytotoxicity to the direct DC treat-ent, suggesting that electrolysis products have an

mportant role in DC-induced cytotoxicity. Metal ionseleased from different electrodes into the culture me-ium were quantified by the inductively coupled plas-a-mass spectroscopy. Higher concentrations of Ag,

n, and Ni and chromium were released from Ag, Zn,nd stainless steel (St) electrodes, respectively, whereasuch lower concentrations of Ni and Ti were released

rom Ni-Ti electrode. Further, electron spin resonancepectroscopy with spin-trapping agent showed that theirect current with the following metal electrodes gen-rated alkoxyl radical (St and Ni-Ti electrodes), hydro-en radical (Ag and Au electrodes), and both carbonnd alkoxyl radicals (Zn electrode), respectively. Theseesults suggest that free radicals and metal ions re-eased from electrodes contribute to the cytotoxicity ofC treatment used for iontophoresis. (J Endod 2006;32:42–446)

ey Wordsytotoxicity, direct current, ESR, free radicals, ionto-horesis, metal ions

From the *Department of Endodontics, ‡Dental Pharma-ology, Meikai, University School of Dentistry, Sakado,aitama, Japan; †Department of Anatomy, School of Medicine,howa, University, Tokyo, Japan.

Address requests for reprint to Dr. Keiso Takahashi, De-artment of Endodontics, Meikai University School of Den-istry, Keyaki-dai 1-1, Sakado, Saitama 350-0283, Japan. E-ail address: okeiso@dent.meikai.ac.jp.

099-2399/$0 - see front matterCopyright © 2006 by the American Association of

ndodontists.oi:10.1016/j.joen.2005.11.010

(

42 Nakamura et al.

ontophoresis is a drug delivery method that utilizes electric current to facilitate thelocal delivery of the water-soluble ionized drugs at therapeutic concentrations. Weave previously reported that direct current (DC) with antibacterial agents inducesecrotic cell death against host cells around periapical tissues and the cytotoxic effectsf DC varied, depending upon the type of metal electrodes (1). However, the underlyingechanism of this phenomenon is not understood. It is not clear whether the observedC-induced effects on host cells are a direct consequence of DC or are indirectlyediated by the electrochemical changes in the medium.

DC shows multipotential biological effects on host cells (2– 4). It has been re-orted that polymorphonuclear cells (PMNs) exposed to DC and iron generates hy-roxyl radicals (5). These reports suggest that DC may generate free radicals regardlessf the presence or absence of cells or induce free radical generation by PMNs. TheC-generated free radicals may also show antibacterial activity (6, 7). However, theechanism by which DC generates free radicals in the medium remains unclear.

Metallic elements released from the dental alloys and their cytotoxic activities haveeen investigated (8 –10) and suggest that metallic elements are involved in cytotoxicctivity. It has been reported that silver, liberated by low levels of DC, had the highestntibacterial effect (6) and gold is the most suitable electrode for the use of ionto-horetic urinary catheters (11). Further, chlorine-containing compounds are requiredor the antibacterial effects of iontophoresis (12). These reports suggest that antibac-erial activity of DC depends on the metal liberated from electrodes.

Therefore, we hypothesized that electrolyzed products such as free radicals andetal ions released from metal electrodes play a role in the DC-induced cytotoxicity.

he objective of this study was to delineate the possible mechanism of DC-mediatedytotoxicity in vitro. We compared cytotoxicity against PMNs for direct and indirect DCreatment (using only the culture medium exposed to DC), using five metal electrodes,u, Ag, Zn, St, and Ni-Ti. We also investigated which metal ions and free radicals areenerated by DC treatment, using coupled plasma atomic emission spectroscopy (ICP-ES) and electron spin resonance (ESR) spectroscopy, respectively.

Materials and Methodseagents

The following chemicals and reagents were obtained from the indicated compa-ies: RPMI 1640 medium, fetal bovine serum (FBS) (Gibco, Grand Island, NY); sodium

luoride (NaF), deferoxamine mesylate (DFO) (Sigma Chemical Co., St. Louis, MO); 5,-dimethyl-1-pyrroline-N-oxide (DMPO), cell counting kit (Dojin Ltd., Kumamoto,apan); diamine silver fluoride solution (AgF) (Saforaide Toyo Pharmaceutical Ind.,o., Osaka, Japan); iodine zinc iodide solution (JJZ) (Showa Pharmaceutical Ind., Co.,okyo, Japan).

lectrodesThe following metal electrodes were obtained from the indicated companies: Au

pure gold) (Tokuriki Co., Tokyo. Japan); Ag (silver) and Zn (Showa Pharmaceuticalnd., Co.); St (stainless steel) (Dentsply-Sankin, Tokyo, Japan); Ni-Ti (Nickel Titanium)

Tomy Orthodontic Products, Tokyo).

JOE — Volume 32, Number 5, May 2006

P

rw

TD

sBt(

I

Rcet

A

PTwta(

C

2tc(Ccc

A

wfatR(8stp

S

stDd

C

wPstmtZteddto0

QM

wZa�6

FPmŒ

*

Basic Research—Technology

J

reparation of Human Peripheral Blood PMNsPeripheral blood PMNs was prepared, according to our previous

eport (1). PMNs were incubated in RPMI 1640 medium supplementedith 10% FBS.

reatment of DCirect

The PMNs were suspended at 2 � 106/ml in RPMI 1640 mediumupplemented with 10% FBS, in the 24-well culture plate (Falcon 3047,ecton Dickinson, Franklin Lakes, NJ), and subjected to DC (2 mA) for

he indicated periods with occasional agitation, as described previously1).

ndirectThe electrolysis products were prepared by treatment with

PMI1640 medium containing 10% FBS without PMNs under the sameonditions as described above. The PMNs were then incubated with thelectrolyzed medium for the indicated periods of time before beingested for cell viability as described below.

ssessment of CytotoxicityThe cytotoxic activity of the direct or indirect DC treatment against

MNs was assessed, with cell counting kit as described previously (13).wo hundred microliters of the cell suspensions (5 � 105 cells/well)ere transferred to the 96-microwell culture plate (Falcon 3072, Bec-

on Dickinson), and incubated for 4 hours with cell counting kit. Thebsorbance was then measured at 540 nm by using microplate readerDainippon Pharmaceutical Co., Laboratory Products, Osaka, Japan).

oupled Plasma Atomic Emission SpectroscopyTwo milliliters of RPMI 1640 with 10% FBS was placed into the

4-well culture plate (Falcon 3047) and was subjected to DC (2 mA)reatment for 5 minutes as described above. The samples were pro-essed for inductively coupled plasma atomic emission spectroscopyICP-AES) (Shimadzu Co., Kyoto, Japan), as reported previously (14).alibration of the apparatus was achieved with standard solutions, andomputer-controlled data collection was used to assign peaks to spe-ific elements.

ssay for Radical IntensityThere were 130 �l of RPMI1640 with 10% FBS and 20 �l of DMPO

ere placed into a test tube and were subjected to DC treatment (2 mA)or 1 minute. After DC treatment, the reaction medium was placed inton ESR flat cell. The detection and measurement of free radicals in theest samples were done at 25°C using ESR spectrometer (JEOL JESE1X, X-band, 100 kHz modulation frequency) as described previously9). Instrument settings: center field, 336 � 5 mT; microwave power,mW; modulation amplitude, 0.1 mT; gain, 400; time constant, 0.03

econd; scanning time, 2 minutes. The radical intensity was defined ashe ratio of peak height of these radicals to that of MnO as describedreviously (9).

tatistical AnalysisIn the cytotoxic assay, each absorbance value represents mean �

tandard deviation (SD) from three independent wells. Most of data arehe representative of three individual experiments with similar results.ifferences between the two groups were analyzed, using unpaired Stu-

ent’s t test. 0

OE — Volume 32, Number 5, May 2006

Resultsytotoxic Effects of Electrolyzed Medium on PMNs

The indirect DC treatment (with electrolyzed conditioned mediumith Ag, Zn, St, and Au electrodes) showed similar cytotoxic effects onMNs, compared to the direct DC treatment of PMNs with the corre-ponding electrodes (Fig. 1a–c, e). The significant difference betweenhe direct and indirect DC treatment was observed only at 5 and 10

inutes in St and Au electrodes (Fig. 1c, e). The direct/indirect DCreatment with Ag electrodes showed the highest cytotoxicity, and the St,n, and Au showed the intermediate cytotoxicity, whereas Ni-Ti showed

he lowest cytotoxicity (Fig. 1d). The direct DC treatment using Ni-Tilectrode, showed, lightly, but significantly higher cytotoxicity than in-irect DC treatment at 1 and 5 minutes (Fig. 1d). The cytotoxicity of bothirect and indirect DC treatment was enhanced more than 1 minute

reatment, regardless of the type of electrodes, and the stimulatory effectf electrodes was in the order: Ag � (p � 0.05) Au, Zn, St ��(p �.01) Ni-Ti (Fig. 1).

uantification of Generated Metal Ions by DC Treatment withetal Electrodes

The amounts of metal ions released into the electrolyzed mediumere quantified with ICP. Larger amounts of ions were released from St,n, and Ag electrodes, whereas a little or no ions released from Ni-Tind Au electrodes: 210 �g/ml (1.95 mM) silver from Ag electrode, 490g/ml (7.5 mM) zinc from Zn electrode, 240 �g/ml (4.1 mM) Ni and

6 �g/ml (1.3 mM) Cr from St electrode, 0.7 �g/ml (0.01 mM) Ni and

Abs

orba

nce

at 5

40 n

m

a

0 30 60 300 600

e

c

b

d

0.8

0.6

0.4

0.2

*

(sec)

**

**

****

igure 1. Comparison between direct and indirect DC-induced cytotoxicity.MNs were treated with DC for the indicated period. Time course of DC (2A)-induced cytotoxicity with five different electrodes. �: direct DC treatment;: indirect DC treatment; (a) Ag, (b) Zn, (c) St, (d) Ni-Ti, (e) Au. *p � 0.05;

*p � 0.01.

.2 �g/ml (0.004 mM) Ti from Ni-Ti electrode.

Release of Free Radicals and Metal Ions by Direct Current 443

FS

a

c

FAg vs. Ag � AgF; (b) Zn vs. Zn � JJZ; (c) St vs. St � NaF. **p � 0.01.

Basic Research—Technology

4

e

44 Nakamura et al.

igure 2. ESR analysis for DC-induced free radicals (a–f ). (a) Ag, (b) Zn, (c)t, (d) Ni-Ti, (e) Au, (f ) control. (g) The generation of alkoxyl radical by St

0

0.2

0.4

0.6

0.8

1

1.2 Ag electrode

hydrogen radical

Rad

ical

inte

nsity

+ AgF

Zn electrode

carbon radical**

alkoxyl radical

0

0.2

0.4

0.6

0.8

1

1.2

1.4

b

+JJZ +JJZ

**St electrode

alkoxyl radical

+NaF0

0.2

0.4

0.6

0.8

1

1.2

igure 3. Effects of antibacterial agents on the generation of free radicals. (a)

lectrode in the absence or presence of catalase (300 unit/ml) or DFO (5 mM).

JOE — Volume 32, Number 5, May 2006

GE

SbwtDg(ar(w

edtpc

tgeerN1ott

oeihtet

r(igarar

tfiFtDNgs

rg

baciisetalmt

foSKfLm

1

1

1

1

1

1

1

Basic Research—Technology

J

enerated Free Radicals by DC Treatment with Metallectrodes

ESR spectroscopy showed that alkoxyl radicals were generated byt and Ni-Ti electrodes (Fig. 2c, d), hydrogen radicals were generatedy Ag and Au electrodes (Fig. 2a, e). Both carbon and alkoxyl radicalsere generated by Zn electrode (Fig. 2b). St electrode generated 6.6

imes more alkoxyl radical (Fig. 2c) than Ni-Ti electrodes (Fig. 2d). TheFO (5 mM), an iron chelator, reduced the alkoxyl radical intensityenerated by the St electrode to undetectable levels, whereas catalase300 unit/ml) did not (Fig. 2g), thus confirming the generation oflkoxyl radicals by St or Ni-Ti electrodes. JJZ significantly (p � 0.01)educed the generation of alkoxyl radicals, but not the carbon radicalFig. 3b). NaF significantly (p � 0.01) reduced alkoxyl radical,hereas AgF did not reduce the hydrogen radical (Fig. 3).

DiscussionThis study indicates that free radicals and metal ions released as

lectrolysis products from electrodes could be responsible for DC-in-uced cytotoxicity against host cells during iontophoresis. These elec-rolysis products may play a crucial role in the cytotoxicity of ionto-horesis, which can be suppressed by antioxidants and/or metal ionhelators.

Both the direct and indirect DC treatment with four metal elec-rodes displayed comparable cytotoxic effect (Fig. 1a–c, e), sug-esting that the modulation of cell death is mainly because of thelectrolysis products. In contrast, the direct DC treatment with Ni-Tilectrode showed higher cytotoxicity than the corresponding indi-ect DC treatment at 1 and 5 minutes, although cytotoxic activity ofi-Ti electrode was much weaker than that of other electrodes (Fig.d). This result indicates that the DC-induced cytotoxicity dependsn the metal electrodes (Fig. 1), and suggests that generated elec-rolysis products play a crucial role in the cytotoxicity in addition tohe direct DC effect.

Cytotoxicity of released metal ions has been reported to dependn the metal contents (15). Quantification of metal ions in thelectrolyzed medium revealed that high concentrations of metalons, such as Ag, Zn, Ni, and Cr were eluted. These minerals havearmful effect on host cells. In addition, much higher concentra-

ions of Ni and Ti were released from St electrode than from Ni-Tilectrode and these metal ions may be involved in the higher cyto-oxicity of St than that of Ni-Ti (Fig. 1c, d).

ESR spectroscopy showed the differential generation of alkoxyladical, hydrogen radical, and carbon radicals from different electrodesFig. 2). These free radicals may also modify the DC-induced cytotox-city (Fig. 1). We have shown for the first time that alkoxyl radical wasenerated by DC with St and Ni-Ti (Fig. 2c, d). The St electrode gener-ted higher amounts of alkoxyl radical than the Ni-Ti electrode. Freeadicals are involved in inflammatory reactions in both pulp and peri-pical lesions (16,17) and antioxidants may play important roles in theegulation of inflammatory reactions.

It has been reported that free radicals are generated by the reac-ion of sodium hypochlorite with hydrogen peroxide (18). In addition,ree radicals are required for teeth bleaching (19-21). Thus, we need tonvestigate in depth the role of free radicals in endodontic treatment.urther study is required to build up a novel iontophoretic treatmenthat is safe to host cells. We have recently found that the cytotoxicity ofC with Zn and St electrodes was partially inhibited by adding JJZ andaF, respectively (1). This is in accord with the present finding that theeneration of alkoxyl radical was reduced by adding JJZ and NaF, re-

pectively (Fig. 3). These results support the possibility that antibacte-

OE — Volume 32, Number 5, May 2006

ial agents may modify the DC-induced cytotoxicity through reducing theeneration of free radicals.

It has been reported that iontophoresis with silver metals can killacteria in bone (22) and that electrically generated silver ions (as littles 0.018 mM) show antifungal effect, whereas cytotoxic effect on hostells are minimal (23). The electrically generated silver ions stronglynhibited bacterial growth. In contrast, it has been reported that goldontophoresis showed the highest antibacterial effect (24). These ob-ervations in combination with our results suggest the importance ofstablishing optimal treatment regimens for iontophoresis in endodon-ic therapy, to achieve less cytotoxicity against host cells and betterntibacterial effects on biofilm in infected root canals and periapicalesions. The mechanism of induction of antibacterial effect by DC treat-

ent and the relationship between DC-induced cytotoxicity and antibac-erial effects need to be elucidated further.

AcknowledgmentsThis study was supported in part by the program Grants-in-Aid

or Scientific Research (Takahashi No. 15592029) from the Ministryf Education, Science, Sports, and Culture of Japan, and from Japanociety for Promotion of Science, and by the Research Fund fromampou Science Foundation, and by the Miyata Research Fund (A)rom Meikai University. We thank Prof. Kinane DF (University ofouisville School of Dentistry) for his thoughtful review of theanuscript.

References1. Nakamura Y, Takahashi K, Shimetani A, Sakagami H, Nishikawa H. Cytotoxicity of

direct current with antibacterial agents against host cells in vitro. J Endod2005;31:755– 8.

2. Rajasekaran M, Hellstrom WJ, Sparks RL, Sikka SC. Sperm-damaging effects of elec-tric current: possible role of free radicals. Reprod Toxicol 1994;8:427–32.

3. Holan V, Machala O, Dao T, Minowada J. Effects of direct current on T cell activity:modulation of interleukin-2 production. Immunobiology 1994;190:368 –75.

4. Goldman R, Pollack S. Electric fields and proliferation in a chronic wound model.Bioelectromagnetics 1996;17:450 –7.

5. Fagerheim I, Lingaas E. Generation of hydroxyl radicals by polymorphonuclear neu-trophils during exposure to direct current and iron. APMIS 1996;104:493–9.

6. Berger TJ, Spadaro JA, Chapin SE, Becker RO. Electrically generated silver ions:quantitative effects on bacterial and mammalian cells. Antimicrob Agents Chemother1976;9:357– 8.

7. Berger TJ, Spadaro JA, Bierman R, Chapin SE, Becker RO. Antifungal properties ofelectrically generated metallic ions. Antimicrob Agents Chemother 1976;10:856 – 60.

8. Hanawa T, Kaga M, Itoh Y, Echizenya T, Oguchi H, Ota M. Cytotoxicities of oxides,phosphates and sulphides of metals. Biomaterials 1992;13:20 – 4.

9. Kinoshita N, Yamamura T, Teranuma H, et al. Interaction between dental metals andantioxidants, assessed by cytotoxicity assay and ESR spectroscopy. Anticancer Res2002;22:4017–22.

0. Wataha JC, Hanks CT, Craig RG. The in vitro effects of metal cations on eukaryotic cellmetabolism. J Biomed Mater Res 1991;25:1133– 49.

1. Davis CP, Wagle N, Anderson MD, Warren MM. Bacterial and fungal killing by ion-tophoresis with long-lived electrodes. Antimicrob Agents Chemother 1991;35:2131– 4.

2. Davis CP, Wagle N, Anderson MD, Warren MM. Iontophoresis generates an antimi-crobial effect that remains after iontophoresis ceases. Antimicrob Agents Chemother1992;36:2552–5.

3. Hiroi M, Tajima M, Shimojima T, Kashimata M, Miyata T, Sakagami H. Re-evaluationof the culture condition of polymorphonuclear cells for the study of apoptosis induc-tion. Anticancer Res 1998;18:1813– 8.

4. Sjogren G, Sletten G, Dahl JE. Cytotoxicity of dental alloys, metals, and ceramicsassessed by millipore filter, agar overlay, and MTT tests. J Prosthet Dent2000;84:229 –36.

5. Davis WL, Jacoby BH, Craig KR, Wagner G, Harrison JW. Copper-zinc superoxidedismutase activity in normal and inflamed human dental pulp tissue. J Endod1991;17:316 – 8.

6. Baumgardner KR, Sulfaro MA. The anti-inflammatory effects of human recombinant

copper-zinc superoxide dismutase on pulp inflammation. J Endod 2001;27:190 –5.

Release of Free Radicals and Metal Ions by Direct Current 445

1

1

1

2

2

2

2

2

Basic Research—Technology

4

7. Shimauchi H, Takayama S, Narikawa-Kiji M, Shimabukuro Y, Okada H. Production ofinterleukin-8 and nitric oxide in human periapical lesions. J Endod 2001;27:749 –52.

8. Shiozawa A. Characterization of reactive oxygen species generated from the mixtureof NaClO and H2O2 used as root canal irrigants. J Endod 2000;26:11–5.

9. Kawamoto K, Tsujimoto Y. Effects of the hydroxyl radical and hydrogen peroxide ontooth bleaching. J Endod 2004;30:45–50.

0. Kashima-Tanaka M, Tsujimoto Y, Kawamoto K, Senda N, Ito K, Yamazaki M. Gener-ation of free radicals and/or active oxygen by light or laser irradiation of hydrogen

peroxide or sodium hypochlorite. J Endod 2003;29:141–3.

46 Nakamura et al.

1. Dahlstrom SW, Heithersay GS, Bridges TE. Hydroxyl radical activity in thermo-cata-lytically bleached root-filled teeth. Endod Dent Traumatol 1997;13:119 –25.

2. Becker RO, Spadaro JA. Treatment of orthopaedic infections with electricallygenerated silver ions. A preliminary report. J Bone Joint Surg Am1978;60:871– 81.

3. Berger TJ, Spadaro JA, Bierman R, Chapin SE, Becker RO. Antifungal properties ofelectrically generated metallic ions. Antimicrob Agents Chemother 1976;10:856 – 60.

4. Davis CP, Arnett D, Warren MM. Iontophoretic killing of Escherichia coli in static fluid

and in a model catheter system. J Clin Microbiol 1982;15:891– 4.

JOE — Volume 32, Number 5, May 2006