Benzoyl peroxide cytotoxicity evaluated in vitro with the human keratinocyte cell line, RHEK-1

Toxicology 106 (1996) 187-196

Benzoyl peroxide cytotoxicity keratinocyte H. Babich* a, H.L. Zuckerbrauna, ‘Department

of Biology. Stern College for

evaluated in vitro with the human cell line, RHEK- 1

B.J. Wurzburgerb, L. Blaub

Women,

Yeshiva University.

Y.L. Rubina,

245 Lexinglon

E. Borenfreundc,

Avenue. New York, NY 10016. USA

‘Department of Chemistry. Stern College for Women, Yeshiva University, 245 Lexington Avenue. New York, NY 10016. USA ‘The Roekefeiier University, Laborarory Animal Research Center, 1230 York Avenue, New York, NY 10021, USA

Received 11 April 1995; accepted 10 July 1995

Abstract The human keratinocyte cell line, RHEK-1, was used to evaluate the cytotoxicity of benzoyl peroxide (BZP). As determined with the neutral red (NR) cytotoxicity assay, the 24-h midpoint (NR,,) toxicity values, in mM, were 0.11 for BZP and 29.5 for benzoic acid, the stable metabolite of BZP. Irreversible cytotoxicity occurred after a l-h exposure to 0.15 mM BZP and greater. When exposed to BZP for 7 days, a lag in growth kinetics was first observed at 0.06 mM BZP. Damage to the integrity of the plasma membrane was evident, as leakage of lactic acid dehydrogenase occurred during a 4-h exposure to BZP at 0.05 mM and greater. Intracellular membranes were also affected, as extensive vacuolization, initially perinuclear but then spreading throughout the cytoplasm, was noted in BZP-stressed cells. The generation of reactive free radicals from BZP was suggested by the following: the intracellular content of glutathione was lowered in cells exposed to BZP; cells pretreated with the glutathione-depleting agent, chlorodinitrobenzene, were hypersensitive to a subsequent challenge with BZP; lipid peroxidation by BZP was inducible in the presence of Fe*+; and cells previously maintained in a medium amended with vitamin E, an antioxidant, were more resistant to BZP, showed less lipid peroxidation in the presence of BZP + Fe*+ and did not develop the extensive intracellular vacuolization as compared to non-vitamin E maintained cells. Keywords:

Benzoyl peroxide;

Free radicals;

Keratinocytes;

Abbreviations: ANOVA, analysis of variance; BZP, benzoyl peroxide; CDNB, chlorodinitrobenzene; DMEM, Dulbecco’s modified Eagle’s medium; EDTA, ethylenediaminetetraacetic acid; FBS. fetal bovine serum; GSH, reduced glutathione; GSSG, oxidized glutathione; HSD, honestly significant differences; LDH, lactic acid dehydrogenase; NR, neutral red; PBS, phosphate buffered saline; S.E.M., standard error of the mean; TBARS, thiobarbituric acid reactive substances; TCA, trichloroacetic acid. * Corresponding author. 0300-483X/96/%15.00 0 I996 Elsevier Science Ireland SSDI 0300-483X(95)03 189-M

Lipid peroxidation;

Neutral

red assay; Vitamin E

1. Introduction Benzoyl peroxide (BZP) is used in the plastic and rubber industries as a nolvmerization in. , itiator, in the food industry as a bleaching agent for cheese and flour and in pharmaceuticals as an antibacterial agent to treat acne. Public concern over the widespread use of BZP was prompted

Ltd. All rights reserved

188

H. Babich et al. /Toxicology

when research showed that this free radical generating-molecule was a tumor promoter and progressor in mouse skin (Slaga et al., 1981; O’Connell et al., 1986) and that it produced DNA single strand breaks and DNA-protein crosslinks in mouse and human cells, including primary cultures of leukocytes and epithelial cells and immortalized epidermal cell lines (Gensler et al., 1983; Hartley et al., 1985; Saladino et al., 1985; Birnboim, 1986). The toxic effect of BZP, presumably, is mediated by its free-radical derivatives. The metabolism of BZP in mouse keratinocytes proceeds via the initial cleavage of the peroxide bond, yielding benzoyloxyl radicals which, in turn, can either fragment to form phenyl radicals and carbon dioxide or abstract hydrogen atoms from biomolecules, causing DNA damage, lipid peroxidation and thiol oxidation (Kensler et al., 1988). Subsequent studies suggested that BZP is activated to DNA-damaging intermediates via a coppercatalyzed cleavage of the peroxide bond (Swauger et al., 1991). Benzoic acid is the major stable metabolite of BZP in skin (Yeung et al., 1983; Nacht et al., 1988). Compared to its genotoxic potential, much less information is available on the cytotoxic effects of BZP. RNA and, to a lesser extent, DNA syntheses were depressed (Saladino et al., 1985) and metabolic cooperation was inhibited (Lawrence et al., 1984) in cells exposed to BZP. Furthermore, BZP may interfere with mitochondrial respiration (Kennedy et al., 1989). The intent of this research was to further evaluate the cytotoxic effects of BZP. including the mediating influence of vitamin E, an antioxidant, on BZP toxicity. The target cell in these studies the nontumorigenic human epidermal was keratinocyte RHEK-1 cell line, established by exposure to the adenovirus 12-simian virus 40 hybrid virus (Rhim et al., 1985). 2. Materials and methods 2. I Cell line The RHEK-1 cells, obtained from J.S. Rhim (National Cancer Institute, Bethesda, MD), were maintained in Dulbecco’s modified Eagle medium

106 11996) 187-196

(DMEM) supplemented with lo”/0 fetal bovine serum (FBS), 100 U/ml penicillin G, 100 &ml streptomycin sulfate and 5.0 pg/ml fungizone. Cultures were dissociated with 0.1% trypsin/0.04% EDTA and maintained in a humidified atmosphere with 5.5% CO2 at 37°C. 2.2. Test materials Benzoyl peroxide was solubilized in acetone, cumene hydroperoxide and tert-butyl hydroperoxide in methanol, ( k)-wtocopherol (vitamin E) and chlorodinitrobenzene (CDNB) in ethanol and hydrogen peroxide (HlOz) and benzoic acid in medium. All test agents were obtained from Sigma Chemical (St. Louis, MO). 2.3. Neutral red (NR) cell viability assay Individual wells of a tissue culture 96-well microtiter plate were inoculated with 0.2 ml of medium (DMEM + 10% FBS) containing sufficient cells, usually 2-3 x 104, to provide approximately 70% confluence after 24 or 48 h, respectively, of incubation. Thereafter, that growth medium was replaced with exposure medium, consisting of DMEM, 10% Serum Plus (JRH Biosciences, Lenexa, KS), 2% FBS and antibiotics, unamended and amended with varied concentrations of test agent. Six to eight replicate wells were used per concentration of test agent. After 24 h of exposure, cytotoxicity was assessed with the neutral red (NR) assay, which is based on the uptake and lysosomal accumulation of the supravital dye, NR (Borenfreund et al., 1990). The protocol for the NR assay is as follows. Neutral red, as a 4 mg/ml aqueous stock solution, protected from light with foil, was stored at room temperature. Medium (DMEM + lo”/0 Serum Plus + 2% FBS) prepared to contain 0.04 mg/ml NR was incubated overnight and centrifuged to remove fine dye crystals. After the appropriate exposure time, the BZP-containing medium was removed and 0.2 ml of NR-containing medium was added to each well. Incubation of cells with NR was continued for 3 h at 37°C. Cells were then rapidly washed and fixed with 0.5”/;1 formalin/l”/ CaCl* (v/v) and the NR incorporated into viable cells was released into the supernatant with 0.2 ml 1% glacial acetic acid/50% ethanol. Absorbance

H. Bubich et al. /Toxicology

was recorded at 540 nm with a microtiter plate spectrophotometer. Data, expressed as percent of untreated controls f the standard errors of the mean, were used to construct concentrationresponse cytotoxicity graphs. A midpoint toxicity value was reported as NRsO (Borenfreund et al., 1990). The ability of the RHEK-1 cells to recover from an acute exposure to BZP was studied by exposing the cells to varied concentrations of BZP for 1 h, washing with phosphate buffered saline (PBS) and then refeeding with DMEM + 10% FBS for 48 h. The NR assay was performed both on cells exposed for 1 h to BZP and on cells exposed for 1 h to BZP followed by 48 h in BZP-free medium. To study the cytotoxic effect of BZP after a prior exposure of the cells to the gluthathionedepleting agent, CDNB, the RHEK-1 keratinocytes were seeded into microtiter 96-well plates, incubated for 24 h, washed with PBS, treated with 0.005 mM CDNB in PBS for 1 h and then treated for 24 h with BZP in exposure medium. After 24 h, the NR assay was performed. 2.4. Cell proliferation assay The quantitative determination of DNA with the fluorochrome, bisbenzimidazole (Hoechst 33258), was used to assess the effect of BZP on cell proliferation. RHEK-1 cells were seeded into individual wells of a 96-well tissue culture plate and upon attaining 70% confluence, the medium (DMEM + 10% FBS) was aspirated and the monolayer of cells was treated with exposure medium amended with varied concentrations of BZP. At time zero and at 24-h intervals for a 4-day period and then at day 7, the plates were emptied by overturning, placed on absorbent toweling for 5 min and then stored frozen at -80°C. On the day of the assay, 0.1 ml of distilled water was added to the wells, the plates were incubated at 37°C for 1 h, refrozen at -80°C for 15 min and then thawed at room temperature to lyse the cells. Then, 0.1 ml of the fluorochrome at 0.02 mg/ml in buffer, consisting of 10 mM Tris, 1 mM EDTA and 2 M NaCl, pH 7.4, was added to each well. To ensure interaction of the fluorochrome with the DNA, the plates were briefly agitated on a rotary shaker. Thereafter, fluorescence was read with a 96-well

106 (1996)

187-196

189

microplate fluorometer (Cambridge Instruments), at an excitation wavelength of 360 nm and an emission wavelength of 460 nm (Rago et al., 1990). 2.5. Lactic acid dehydrogenase (LDH) release RHEK-1 cells were seeded into individual wells of a 24-well tissue culture plate and upon attaining confluence the medium (DMEM + 10% FBS) was aspirated, the monolayer of cells washed with PBS and 0.5 ml of PBS, unamended and amended with varied concentrations of test agent, was added to each well. After a 4-h exposure, the PBS solution (supernatant PBS) from each well was collected and stored at 4°C. Cell monolayers were treated with 0.5 ml of 1 mg/ml digitonin in PBS for 30 min at room temperature to lyse the cells. LDH activity was measured in the supernatant PBS and in the cell lysate using the CytoTox 96@ Nonradioactive Cytotoxicity Assay Kit from Promega Company (Madison, WI). Released LDH was determined as the percentage of the total LDH activity: i.e., (LDH activity in PBS supernatant) + (LDH activity in PBS supernatant + LDH activity in cell lysate) (Sinensky et al., 1995). 2.6. Lipid peroxidation Confluent monolayers of RHEK-1 cells seeded into 60 mm culture dishes were washed with PBS, treated for 2 h with 2 ml of PBS (control), 1 mM Fe’+ (as FeCl*) in PBS, 2 mM BZP in PBS or a combination of Fe2+ and BZP in PBS. Thereafter, 1 ml of 24% trichloroacetic acid (TCA) was added, the cells were scraped, the resulting suspensions were centrifuged to remove the precipitated proteins and lipid peroxidation was estimated in the supernatant (which contained TCA at a final concentration of 8%) by measuring the concentration of malondialdehyde (Kappus and Artuc, 1987). Then, l-ml aliquots of the supernatant, to which 0.5 ml of 1% thiobarbituric acid in 0.05 M NaOH was added, were heated at 95-100°C for 25 min (Hellstrom et al., 1994). After cooling, the samples were centrifuged at 1286 x g for 10 min and the absorbance of the supematant was measured at 532 nm (Draper et al., 1993). Absorbance was converted to nmol thiobarbituric acid reactive substances (TBARS) using 1,1,3,3-tetramethoxypropane as the standard (Gavin0 et al., 1981).

190

H. Babich et al. /Toxicology

In each experiment, a confluent monolayer of cells was sacrificed by treatment with 4 ml of 0.1 N NaOH, incubated for 1 h at 37°C and cellular protein was estimated with Coomassie brilliant blue G by the Bradford assay (Bradford, 1979). 2.7. Glutathione assay Measurements of cellular glutathione were performed with minor modifications according to the procedures of Anderson (198 1). Cells, maintained in DMEM + 10% FBS and grown to confluence in 60 mm culture dishes, were washed with PBS and refed with PBS, unamended and amended with varied concentrations of BZP. After 0.5 h of exposure, the cells were washed with PBS, lysed with 0.2% Triton X-100 and made to contain 2.5% sulfosalicylic acid in a total volume of 0.12 ml. After harvesting by scraping and centrifugation, acid-soluble extracts were assayed. Reduced glutathione (GSH) was measured spectrophotometrically at 412 nm in 0.06-ml aliquots of the acid-soluble extract by determining the reduction of 5,5’-dithiobis(2-nitrobenzoic acid), prepared in phosphate buffer/EDTA (pH 7.5), to 5-thio-2nitrobenzoic acid and oxidized glutathione (GSSG). With each assay, a standard curve was generated with known amounts of glutathione. 2.8. Vitamin E studies To study the protective effects of vitamin E on BZP toxicity, the RHEK-1 cells, previously maintained for 3-4 passages in medium supplemented with 0.01 mM vitamin E, were seeded (3 x 104/well) to 96-well microtiter plates in growth medium supplemented with vitamin E. After 24 h of incubation, the growth medium was removed, the cells were washed twice with PBS and the exposure medium without vitamin E, but with varied concentrations of BZP, was added; after a 24 h incubation, the NR assay was performed. For studies on lipid peroxidation, cells maintained in vitamin E were seeded into 24-well plates and the protocol, noted previously for non-vitamin E maintained cells, was followed. 2.9, Brightfield microscopy RHEK- 1 cells (3 x 1OS) were seeded alcohol-cleansed 22 x 22 mm glass coverslips

to in

106 (1996) 187-196

35 mm diameter tissue culture dishes containing DMEM + 10% FBS. After 48-72 h, the monolayers were treated with varied concentrations of BZP in DMEM + 10% Serum Plus + 2% FBS for 24 h. Thereafter, the cells were washed with PBS, fixed in methanol and stained with Giemsa’s solution (EM Science, Gibbstown, NJ). Cells were examined microscopically for cellular aberrations. 2.10. Statistics Experiments were performed at least four times. The individual data points in the graphs and tables are presented as the arithmetic mean percent of control + standard error of the mean (S.E.M.). Experimental data were analyzed with one-way analysis of variance (ANOVA) followed by Tukey’s multiple range test for honestly significant differences (HSD). To be considered significant, the P value of the effect being considered must be 50.05. 3. Results The cytotoxicity of a 24-h exposure to BZP was compared to that of the organic peroxides, cumene hydroperoxide and terl-butyl hydroperoxide, hydrogen peroxide and benzoic acid. Fig. 1 shows the concentration-response cytotoxicity curves of these test agents for the RHEK-1 keratinocytes and Table 1 lists their midpoint cytotoxicity or values (i.e. the concentration of test agent NRYI that reduced the uptake of NR by 50%, as compared to untreated control cells). The organic peroxides were more potent than H,O,, which was more potent than benzoic acid. Cells, seeded to two 96-well microtiter plates and incubated for 1 day, were treated with varied concentrations of BZP for 1 h. Thereafter, one plate was processed for the NR assay to determine the percentage of cells surviving after the 1 h exposure to BZP. The NR,, value for this l-h exposure was 0.13 + 0.008 mM BZP. The BZP-containing medium was removed from the second plate, which was refed with medium (DMEM + 10% FBS) lacking BZP and incubated for an additional 48 h. This plate provided information on the growth dynamics of the surviving cells, stressed by a prior acute exposure to BZP. As noted in Fig. 2,

H. Bubich et al. /Toxicology

IO6 (1996j

187-196

191

a

BA

HP

1 Hr exposure

t

1 Hrexposure \ + 48 hr in recovery medium

7

4 tBH 0 ~-

t 0 *

I

0 01

0 10

1 00

~~~

I

0 05

0 00

100 00

1000

Concentration

(mM)

for cells exposed for 1 h to BZP at 0.1 mM or less, the subsequent 48 h growth response of the surviving populations in recovery medium reflected the initial decrease caused by the 1-h exposure to BZP. However, for cells exposed for 1 h to BZP at 0.15 mM and greater, the subsequent 48 h growth response of the surviving populations in recovery medium showed a more pronounced (P < 0.01) possibly irreversible, damage. The effect of BZP on proliferation of the RHEK-1 cells exposed for 7 days is shown in Fig.

cytotoxicity (NRSo) values of the test agents towards keratinocytes after a 24-h exposure

Test agent

N&O

Cumene hydroperoxide Terr-butyl hydroperoxide Benzoyl peroxide Hydrogen peroxide Benzoic acid

0.07 0.08 0.11 0.91 29.5

as the arithmetic

means

0 15

0 20

Fig. 2. Response of the RHEK-l keratinocytes to a l-h exposure to benzoyl peroxide and to the l-h exposure followed by 48 h in benzoyl peroxide-free recovery medium. Individual data points are presented as the arithmetic mean percent of the controls f S.E.M.

3. A rapid linear increase in growth during the first 2 days was noted for the control as well as for the cells treated with 0.02 and 0.04 mM BZP. Although linearity was observed at 0.06 and 0.08 mM BZP, a lag in growth kinetics was also apparent. Furthermore, based on DNA quantitation, BZP affected the total number of cells present after the 7-day exposure (Table 2). In all treatments, a

Table 2 Effect of benzoyl peroxide RHEK-I keratinocytes Benzoyl peroxide

Data are presented

0 10

Benzoyl peroxlde (mM)

Fig. I. Comparative 24-h cytotoxicities of cumene hydroperoxide (CH), ferr-butyl hydroperoxide (tBH), benzoyl peroxide (BZP), hydrogen peroxide (HP) and benzoic acid (BA) towards RHEK-1 keratinocytes as assessed with the neutral red (NR) assay. The data are presented as the arithmetic mean percent of the controls + S.E.M.

Table 1 Midpoint RHEK-1

~~_~ 3

(mW

* 0.005 f 0.006 + 0.008 f 0.028 zt 1.66

+ S.E.M.

(mM)

on the growth

Time required for the DNA content to double

0

86’ 76* 69* 61’

aBased on the time for the fluorescence units to double. bBased on DNA quantitation after 7 days of exposure. *Statistically

significant

at P I 0.05.

of the

Cell population (% of control)b

(h)a

20 23 24 38’ 48*

0.02 0.04 0.06 0.08

kinetics

H. Babich et al. / Toxicology

192

004mM

008mM

0

---

0

1

2

3

4

’

5

6

7

Days

Fig. 3. Response of the RHEK-I to a continuous exposure to 0, 0.02, 0.04, 0.06 and 0.08 mM benzoyl peroxide. Cellular DNA was estimated with the fluorochrome, Hoechst 33258, using an excitation wavelength of 360 nm and an emission wavelength of 460 nm. Each point mean f S.E.M. from determinations tration of BZP.

represents the arithmetic of eight wells per concen-

106 (1996) 187-196

marked decrease in cell proliferation occurred between day 4 and 7 (Fig. 3). In addition to a pronounced adverse effect on cell proliferation, other cytotoxic effects of BZP were observed. The integrity of the cell membrane was affected as shown by measurements of the leakage of LDH from cells treated for 4 h with BZP. Statistically significant levels of LDH in the supematant were noted with BZP at 0.5 mM and higher (Fig. 4). Benzoyl peroxide also caused lipid peroxidation. Lipid peroxidation, after a 2-h incubation, was detected in RHEK-1 keratinocytes treated with a combination of 2 mM BZP + 1 mM Fe2+. Control keratinocytes and keratinocytes treated either with 2 mM BZP or 1 mM Fe2+ alone did not exhibit peroxidation (Table 3). Lipid peroxidation, however, was not accompanied by an increase in cytotoxicity, as determined by the level of LDH release after a similar 2-h incubation. In untreated controls and in cells treated with 1 mM Fe2+, the level of LDH detected in the supernatant was 7 f 0.4 and 7 + 1.9% of the total, respectively. In cells treated with 2 mM BZP (in which lipid peroxidation was not noted) and with 2 mM BZP + 1 mM Fe’+ (in which lipid peroxidation occurred) the level of LDH detected in the supernatant was 71 + 2.7 and 74 + 2.3% of the total, respectively.

Table 3 Lipid peroxidation

in RHEK-I

keratinocytes

Treatment

nmol TBARSimg cellular protein

Cells not maintained in vitamin E

4 Hr exposure

0

I 00

.--~~-T---

I

05

10 Benzoyl

1

15

-.--

20

I mM Fe2+

0 0 0 2.3 + 0.04’

Cells maintained in vitamin E Control 2 mM Benzoyl peroxide I mM Fe*+ 2 mM Benzoyl peroxide + I mM Fe2+

0 0 0 0.7 * 0.14*

Control 2 mM Benzoyl peroxide I mM Fe’+ 2 mM Benzoyl peroxide + r 25

peroxide (mM)

Fig. 4. Leakage of lactic acid dehydrogenase (LDH) from RHEK-1 keratinocytes treated for 4 h with benzoyl peroxide. The data are presented as the arithmetic means l S.E.M.

1

Data are presented as the arithmetic means f S.E.M. *Statistically significant at P < 0.05.

H. Babich et al. /Toxicology Table 4 Responses of RHEK-I cells, not maintained and maintained in vitamin E amended medium, to a 24-h exposure to benzoyl peroxide Benzoyl peroxide (mM)

0.10 mM 0.12 mM

Percent

of control

RHEK-

I cells

46 * 1.2 34 f 3.1

106 (1996)

100

193

187-196

’

(NR assay) RHEK-1 cells maintained in vitamin E

k

72 f 3.1+ 51 f 4.9**

*Statistically significant, at P < 0.01, from cells not maintained in medium amended with vitamin E.

\

CDNB pretreatment

**Statistically significant, at P < 0.05, from cells not maintained in medium amended with vitamin E.

RHEK-1 keratinocytes maintained in vitamin E-free medium exhibited about three times more lipid peroxidation when exposed to 2 mM BZP + 1 mM Fe2+ than cells maintained in medium amended with 0.01 mM vitamin E and subsequently exposed to BZP + Fe2+ (Table 3). Furthermore, cells previously maintained in medium amended with vitamin E were more resistant to a subsequent challenge with BZP than were cells maintained in control (vitamin E-free) medium (Table 4). The intracellular level of glutathione was lowered in RHEK-1 keratinocytes exposed for 0.5 h to BZP at 0.5 mM and higher (Table 5). A l-h pretreatment of the keratinocytes with the glutathione-depleting agent, CDNB (0.005 mM), lowered their intracellular level of glutathione to 15 f 3.5% as compared to untreated cells and enhanced their subsequent sensitivity to BZP (Fig. 5), with the 24-h NR,, lowered from 0.11 f 0.008 to 0.03 f 0.002 mM.

o 00

0 01

0.02

0.03

0 04

0.05

Benzoyl peroxlde (mM)

Fig, 5. without depleter, benzoyl

Comparative response of RHEK-I keratinocytes, and with prior l-h pretreatment with the glutathionechlorodinitrobenzene (CDNB), to a 24-h exposure to peroxide. The data are presented as the arithmetic

means f S.E.M.

Microscopic observation of the RHEK-1 keratinocytes exposed to BZP resulted in an increase in multinucleation and in extensive vacuolization. Such vacuolization initially appeared in the perinuclear region and then spread throughout the cytoplasm. Cells maintained in medium supplemented with vitamin E and subsequently exposed to BZP showed less vacuolization and less multinucleation than non-vitamin E pretreated cells (Fig. 6). 4. Discussion

Table 5 Decrease of intracellular glutathione in RHEK-I exposed to bcnzoyl peroxide for 0.5 h Benzoyl peroxide 0.1 0.5 1.0 1.5 2.0

(mM)

Percent 100 15 54 31 31

Data are presented as the arithmetic means *Statistically significant at P < 0.05.

keratinocytes

of control

f 2.4 l 4.3; + 3.6* f 1.1* f 1.4* f S.E.M.

Benzoyl peroxide and the two other organic peroxides, cumene hydroperoxide and t-butyl hydroperoxide, were about ten times more cytotoxic to the RHEK-1 keratinocytes than was H202, as determined with the NR assay after a 24-h exposure. Saladino et al. (1985), using normal human bronchial epithelial cells in a clonal growth assay, also showed the greater cytotoxicity of BZP as compared to H202. Based on NRSo values for a 24h exposure, BZP was about 250 times more cyto-

194

H. Bahich ef al. / Toxicology

106 (IYY6)

187-196

Fig. 6. (A) RHEK-I cells grown in control, unamended medium. (B) RHEK-I cells exposed for 1 h to 0.135 mM BZP and then maintained for 48 h in medium without benzoyl peroxide. (C) RHEK-1 cells maintained in medium supplemented with 0.01 mM vitamin E. (D) RHEK-I cells, maintained in medium supplemented with 0.01 mM vitamin E, exposed for 1 h to 0.135 mM benzoyl peroxide in medium lacking vitamin E and then maintained for 48 h in medium without vitamin E or benzoyl peroxide. Giemsa stain, original magnification x 270.

toxic to the RHEK-I cells than was benzoic acid, the major stable metabolite of BZP. Swauger et al. (1991) using a cell-free system, noted DNA damage by BZP, but not by benzoic acid. In addition to exerting an overt cytotoxic effect on the RHEK-1 keratinocytes, other deleterious but more subtle effects were noted. (a) As assessed by DNA quantitation, RHEK-1 cells maintained continuously in medium amended with BZP at 0.06 and 0.08 mM exhibited a prolonged lag in their growth. Saladino et al. (198.5) noted that DNA synthesis was depressed in cells exposed to BZP. (b) Irreversible cellular damage occurred after a lh exposure to BZP at 0.15 mM and higher. Such BZP-treated cells were unable to recover from the

BZP-stress and even when maintained for 48 h in normal medium, cytotoxicity continued and progressed. Similarly, ‘continued’ genotoxicity, as quantitated by DNA strand breakage, was evident in human leukocytes exposed to BZP and subsequently maintained in BZP-free medium (Birnboim, 1986). (c) Significant leakage of LDH from RHEK- 1 cells occurred upon a 4-h exposure to BZP at 0.5 mM and above. (d) Cytopathologies were noted in BZP-treated cells. Although some of the control RHEK-1 keratinocytes exhibit multinucleation (Gantt et al., 1987) this condition was increased after exposure to BZP. More pronounced was the extensive vacuolization noted in RHEK- 1 keratinocytes ex-

H. Babich et al. /Toxicology

posed to BZP. Similarly, such extensive vacuolization was noted by Jain et al. (1992) in mouse skin JB6 cells treated with BZP. The l-h and 24-h NR,, values for BZP towards the RHEK-1 keratinocytes were 0.13 and 0.11 mM, respectively. The closeness of these two values, despite differences in the lengths of exposure, suggests that the cytotoxicity of BZP is due to the generation of a short-lived reactive metabolite. Swauger et al. (1991) proposed that the peroxide bond of BZP undergoes cleavage to yield benzoyloxyl and phenyl radicals, which cause cytotoxicity, lipid peroxidation and DNA damage. Such reactive radicals may be the active cytotoxicants responsible for inducing damage in the RHEK-1 cells. The studies reported herein support, in part, the suggestion that the cytotoxicity of BZP is related to the generation of reactive oxidative free radicals. (a) BZP lowered, in a concentration-response pattern, the intracellular level of glutathione in RHEK-1 keratinocytes. Kennedy et al. (1989) showed that BZP reacts rapidly with reduced glutathione (GSH), leading to the formation of oxidized glutathione (GSSG). (b) RHEK-1 keratinocytes preexposed to the glutathione-depleting agent, CDNB, were hypersensitive to a subsequent challenge with BZP. (c) Lipid peroxidation of the RHEK-1 keratinocytes by BZP was inducible in the presence of Fe*+. (d) The cytotoxicity of BZP, the ability of BZP to induce lipid peroxidation in the presence of Fe*+ and the extensive BZP-induced vacuolization were lessened in RHEK-1 cells pretreated with vitamin E, an antioxidant. Vitamin E is oxidized by BZP (Skinner and Parkhurst, 1966). Kennedy et al. (1989) demonstrated that BZP-induced damage to isolated mitochondria was significantly reduced in the presence of vitamin E. BZP per se, however, did not cause lipid peroxidation in the RHEK-1 keratinocytes; lipid peroxidation by BZP was inducible in the presence of Fe*+ Kappus and Artuc (1987) noted, as in the studies reported herein with the RHEK-1 keratinocytes, that only in the presence of Fe*+ did BZP induce lipid peroxidation in primary cul-

106 (1996) 187-196

195

tures of human epidermal keratinocytes. Furthermore, as also noted with the RHEK-1 keratinocytes, the moderate lipid peroxidation that occurred in the presence of BZP and Fe’+ was not accompanied by toxicity. They postulated that keratinocytes are somehow protected against mild peroxidation. Iannone et al. (1993) found no evidence of free radical generation in human keratinocytes exposed to BZP; however, desferrioxamine, a chelator of iron cations, was incorporated into their test system. Conversely, the formation of free radicals have been identified in murine keratinocytes exposed to BZP (Kensler et al., 1988). The role of Fe*+ in BZP-induced lipid peroxidation has been noted also in cell-free systems. Using a microsomal fraction from rabbit dental pulp as the assay system, Terakado et al. (1984) noted lipid peroxidation by BZP in the presence, but not in the absence, of Fe’+. Kennedy et al. (1989), using isolated rat liver mitoshowed that BZP induced lipid chondria, although the concentration of peroxidation, TBARS was low as compared to an initiating system containing BZP with Fe*+. Apparently, studies are warranted to clarify the involvement of Fe*+ in the induction of lipid peroxidation by BZP. Acknowledgments This research was supported, in part, by grants from Schering-Plough Research Institute and Chevron Corporation. Appreciation is expressed to Dr. J.S. Rhim for the RHEK-1 cells and to Cambridge Instruments for the microtiter plate fluorospectrophotometer. References Anderson, M.E. (1981) Determination of glutathione glutathione disulfide in biological samples. Methods

and En-

zymol. 113, 548-555. Birnboim, H.C. (1986) DNA strand breaks in human leukocytes induced by supcroxide anion, hydrogen peroxide and tumor promoters are repaired slowly compared to breaks radiation. Carcinogenesis 7, induced by ionizing 1511-1517. Borenfreund, E., Babich, H. and Martin-Alguacil, N. (1990) Rapid chemosensitivity assay with human normal and tumor cells in vitro. In Vitro Cell. Dev. Biol. 26, 1030- 1034.

196

H. Babich et al. / Toxicology 106 (1996) 187-196

Bradford, M.M. (1979) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. Draper, M.H., Squires, E.J., Mahmoodi, H., Wu, J., Agarwal, S. and Hadley, M. (1993) A comparative evaluation of thiobarbituric acid methods for determination of malondialdehyde in biological materials. Free Rad. Biol. Med. 15. 353-363. Gantt, R., Sanford, K.K., Parshad, R., Price, F.M., Peterson, W.D. and Rhim, J.S. (1987) Enhanced G2 chromatid radiosensitivity, an early stage in the neoplastic transformation of human epidermal keratinocytes in culture. Cancel Res. 47, 1390-1397. Gavino, V.C., Miller, J.S., Ikharebha, SO., Mile, G.E. and Cornwell, D.G. (198 1) Effect of polyunsaturated fatty acids and antioxidants on lipid peroxidation in tissue cultures. J. Lipid Res. 22, 763-769. Gensler, H.L. and Bowden, G.T. (1983) Evidence suggesting a dissociation of DNA strand scissions and late-stage promotion of tumor cell phenotype. Carcinogenesis 4, 1507- I5 11. Hartley, J.A., Gibson, N.W., Zwelling, L.A. and Yuspa, S.H. (1985) Association of DNA strand breaks with accelerated terminal differentiation in mouse epidermal cells exposed to tumor promoters. Cancer Res. 45, 4864-4870. Hellstrom, W.J.G., Bell, M., Wang, R. and Sikka, SC. (1994) Effect of sodium nitroprusside on sperm motility, viability and lipid peroxidation. Fertil. Steril. 61, 1117-l 122. lannone, A., Marconi, A., Zambruno, G., Giannetti, A., Vannini, V. and Tomasi, A. (1993) Free radical production during metabolism of organic hydroperoxides by normal human keratinocytes. J. Invest. Dermatol. 101. 59-63. Jain, P.Y., Fitzpatrick, M.J., Phelps, P.C., Berezesky, 1.K. and Trump, B.F. (1992) Studies of skin toxicity in vitro: doseresponse studies in JB6 cells. Toxicol. Pathol. 20, 394- 404. Kappus, H. and Artuc, M. (1987) Effects of organic peroxides on human epidermal keratinocytes. Bioelectrochem. Bioenerget. 18, 263-270. Kennedy, C.H., Winston, G.W., Church, D.F. and Pryor. W.A. (1989) Benzoyl peroxide interaction with mitochondria: inhibition of respiration and induction of rapid, largeamplitude swelling. Arch. Biochem. Biophys. 271,456-470. Kens!er, T. W., Egner, P.A., Swauger, J.E., Taffe, B.C. and Zweier, J.L. (1988) Formation of free radicals from benzoyl peroxide in murine keratinocytes. Proc. Am. Assoc. Cancer Res. 29, 150. Lawrence, N.J., Parkinson, E.K. and Emmerson, A. (1984) Benzoyl peroxide interferes with metabolic co-operation be-

tween cultured human epidermal keratinocytes. Carcinogenesis 5, 4 19-42 1. Nacht, S,, Yeung, D., Beasley, J.N., Anjo, M.D. and Maibach, H.I. (1981) Benzoyl peroxide: percutaneous penetration and metabolic disposition. J. Am. Acad. Dermatol. 4, 31-37. O’Connell, J.F., Klein-Szanto. A.P., DiGiovanni. D.M.. Fries, J.W. and Slaga, T.J. (1986) Enhanced malignant progression of mouse skin tumors by the free-radical generator benzoyl peroxide. Cancer Res. 46, 2863-2865. Rago, R., Mitchen, J. and Wilding, G. (1990) DNA fluorometric assay in 96-well tissue culture plates using Hoechst 33258 after cell lysis by freezing in distilled water. Anal. Biochem. 191, 31-34. Rhim, J.S., Jay, G., Arnstein, P., Price, F.M., Sanford, K.K. and Aaronson, S.A. (1985) Neoplastic transformation of human epidermal keratinocytes by Ad12-SV40 and Kirsten sarcoma viruses. Science 227, 1250- 1252. Saladino, A.J., Willey, J.C., Lechner. J.F., Grafstrom, R.C., LaVeck, M. and Harris, C.C. (1985) Effects of formaldehyde, acetaldehyde, benzoyl peroxide and hydrogen peroxide on cultured normal human bronchial epithelial cells. Cancer Res. 45, 2522-2526. Sinensky, M.C., Leiser, A.L. and Babich, H. (1995) Oxidative stress aspects of the cytotoxicity of carbamide peroxide: in vitro studies. Toxicol. Lett. 75, 101-109. Skinner, W.A. and Parkhurst, R.M. (1966) Oxidation products of vitamin E and its model, 6-hydroxy-2,2,5,7,8pentamethylchroman. VIII. Oxidation with benzoyl peroxide. J. Org. Chem. 311, 248-251. Slaga, T.J., Klein-Szanto, A.J.P., Triplett, L.L., Yotti, L.P. and Trosko, J.E. (1981) Skin-tumor-promoting activity of benzoyl peroxide, a widely used free radical-generating compound. Science 213, 1023-1024. Swauger, J.E., Dolan, P.M., Zweier, J.L., Kuppusamy, P. and Kensler, T.W. (1991) Role of the benzoyloxyl radical in DNA damage mediated by benzoyl peroxide. Chem. Res. Toxicol. 4, 223-228. Terakado, M., Yamazaki, M., Tsujimoto, Y., Kawashima, T. Nagashima, K., Ogawa, J., Fujita, Y., Sugiya, H., Sakai, T. and Furuyama, S. (1984) Lipid peroxidation as a possible cause of benzoyl peroxide toxicity in rabbit dental pulp a microsomal lipid peroxidation in vitro. J. Dent. Res. 63. 901-905. Yeung, D., Nacht, S., Bucks, D. and Maibach, H.I. (1983) Benzoyl peroxide: percutaneous penetration and metabolic disposition. II. Effect of concentration. J. Am. Acad. Dermatol. 9. 920-924.