(CANCER RESEARCH 48, 3094-3099, June 1, 1988]

Promotion of DNA Strand Breaks in CoculturedMononuclear Leukocytes byProtein Kinase C-dependentProoxidative Interactions of Benoxaprofen,Human PolymorphonuclearLeukocytes, and Ultraviolet Radiation1

Gwen Schwalb, Albertus D. Beyers, Ronald Anderson,2 and Andre E. Nel

Division of Immunology, Department of Medical Microbiology, Institute of Pathology, University of Pretoria [G. S., R. A.]; and the Departments of Medical Physiology[A. D. B.J and Internal Medicine [A. E. N.J, Faculty of Medicine, University of Stellenbosch, Tygerberg; Republic of South Africa

ABSTRACT

At concentrations of 5 «ig/mland greater the nonsteroidal antiinflam-matory drug benoxaprofen caused dose-related activation of lucigenin-

enhanced chemiluminescence in human polymorphonuclear leukocytes(PMNL). Benoxaprofen-mediated activation of lucigenin-enhanced

chemiluminescence by PMNL was increased by UV radiation and wasparticularly sensitive to inhibition by the selective protein kinase Cinhibitor 11-7. To identify the molecular mechanism of the prooxidative

activity of benoxaprofen, the effects of the nonsteroidal antunflammatorydrug on the activity of purified protein kinase C in a cell-free systemwere investigated. Benoxaprofen caused a dose-related activation of pro

tein kinase C by interaction with the binding site for the physiologicalactivator phosphatidylserine, but could not replace diacylglycerol. Whenautologous mononuclear leukocytes (MNL) were cocultured with PMNLand benoxaprofen in combination, but not individually, the frequency ofDNA strand breaks in MNL was markedly increased. LTV radiationsignificantly potentiated damage to DNA mediated by benoxaprofen andPMNL. Inclusion of Superoxide dismutase, H-7, and, to a much lesserextent, catatase during exposure of MNL to benoxaprofen-activated

PMNL prevented oxidant damage to DNA. These results clearly demonstrate that potentially carcinogenic prooxidative interactions, whichare unlikely to be detected by conventional assays of mutagenicity, mayoccur between phagocytes, UV radiation, and certain pharmacologicalagents.

INTRODUCTION

Highly reactive forms of molecular oxygen may play animportant role in tumorogenesis, apparently by acting at thelevel of tumor promotion (1). These oxygen species, such asSuperoxide, hydroperoxy radical, singlet oxygen, hydroxyl radical and hydrogen peroxide are potent inducers of DNA strandbreaks and chromosomal aberrations (2-7). For this reasonactivated phagocytes are potential carcinogens since they generate unstable and stable reactive oxidants (8) which are muta-genie (9) and promote chromosomal abnormalities and malignant transformation in cocultured eukaryotic cells (10,11). Thesuperoxide-generating enzyme of phagocytes is a membrane-associated NADPH-oxidase which is activated by diverse stimuli such as the tumor promoter phorbol myristate acetate,lectins, calcium ionophore, opsonized particles, and complement-derived and /V tormylpcpi ule leukoattractants (7). Theseactivators utilize different transductional mechanisms to acti-

Received 10/23/87; revised 2/11/88; accepted 3/3/88.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' This investigation was supported by individual investigator research grantsawarded by the National Cancer Association and the Arthritis Foundation ofSouth Africa to R. Anderson and by The South African Medical Research Councilto A. E. Nel.' To whom requests for reprints should be addressed, at Institute of Pathology,

P. O. Box 2034, Pretoria 0001, Republic of South Africa.1The abbreviations used are: PKC, protein kinase C; NSAID, nonsteroidal

antiinflammatory drug; PMNL, polymorphonuclear leukocytes; MNL, mononuclear leukocytes; miss. Hanks' balanced salt solution; LECL, lucigen (bis-/V-methylacridinium nitrate)-enhanced chemiluminescence; H-7, 1-(5-isoquinoline-sulfonyl)-2-methylpiperazine; SOD, Superoxide dismutase; PS, phosphatidylserine; ds, double strand.

vate NADPH oxidase such as PKC3 in the case of phorbolmyristate acetate (8, 12-14).

Theoretically pharmacological agents which directly activatemembrane-associated oxidative responses in human phagocytespresent the potential hazard of oxygen toxicity to host tissues.We have previously reported that the NSAID benoxaprofenactivates Superoxide generation by human PMNL in the absence of physiological stimuli and that the prooxidative interactions of this NSAID are potentiated by UV radiation (15-18). Clinical experience with benoxaprofen was relatively briefsince the drug was withdrawn from the international market in1982 due to drug-related deaths and a particularly high incidence of cutaneous side effects (19) including eruptive skintumors on sun-exposed skin (20). Nevertheless we believe thatcontinued laboratory research on benoxaprofen is warrantedand may contribute to the documentation of novel mechanismsof drug-mediated toxicity and carcinogenesis. In this study wehave investigated the effects of benoxaprofen and humanPMNL individually and in combination on the frequency ofDNA strand breaks in cocultured autologous lymphocytes aswell as the molecular mechanism of the prooxidative activityof this NSAID.

MATERIALS AND METHODS

Chemicalsand Reagents.Unless indicatedchemicalsand reagentswere purchased from Sigma Chemical Co., St. Louis, Mo. Benoxaprofen, 2,4-(chlorophenyl)-a-methyl-benzoxazoleacetic acid, was obtained from the Lilly Research Centre Ltd., Earl Wood Manor, Win-dlesham, Surrey, England, dissolved at a concentration of 6 mg/ml in0.05 N NaOH then diluted to a stock solution of 600 ¿ig/mlwithrestoration to pH 7.4. The concentration range tested was 0.1-60 //«/ml which is well within the range of serum concentrations which wereachieved during chemotherapy with this NSAID (21, 22). [>-32P]ATPwas synthesized by interacting ["Pjorthophosphate (Amersham, England) with Gamma-prepA (Promega Biotec.).

Preparation of PMNL and MNL. PMNL and MNL were preparedfrom venous blood taken from healthy adult human volunteers andtreated with preservative-free heparin (5 units/ml). PMNL and MNL

were separated by centrifugation, at 400 x g for 15 min, of heparinizedblood on cushions of Ficoll (Pharmacia, Uppsala, Sweden) metrizoate.Residual erythrocytes in the PMNL preparations were removed bysequential sedimentation with 3% gelatin and selective lysis with 0.85%ammonium chloride. After washing, the PMNL and MNL were resus-pended to 2 x 107/ml in 4.2 IHM 4-(2-hydroxyethyl)-l-piperazineeth-

anesulfonic acid-buffered, indicator-free HBSS, pH 7.4.

Measurement of Oxidant Generation by PMNL. This was measuredusing LECL (23). PMNL (10') were preincubated for 15 min at 37'C

with 0.2 HIMlucigenin in 900 /<!HBSS. LECL was then measured inan I Kit Wallace (Turku, Finland) luminometer (Model 1251) afteraddition of benoxaprofen (0.1-60 /ig/ml final concentrations). LECLreadings were integrated for 5-s intervals and plotted as mV s ' Appropriate benoxaprofen-free and PMNL-free control systems were included. The effects of UV radiation (as described below) and H-7, apotent inhibitor of protein kinase C (24) on benoxaprofen-inducedoxidant generation by PMNL were investigated. PMNL were preincu-

3094

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

DNA DAMAGE BY BENOXAPROFEN AND PHAGOCYTES

bated for 30 min at 37*C with 12.5-100 //M of H-7, a concentration

range which proved to be nontoxic to PMNL and free of scavengingeffects (25). Benoxaprofen (30 ¿ig/ml)was then added and LECLrecorded as described above.

Measurement of DNA Single-Strand Breaks. The formation of single-strand breaks in DNA was measured by alkaline unwinding and determination of ethidium bromide fluorescence using a Hitachi (Tokyo,Japan) fluorescence spectropho to meter (Model 650-10) with excitationat 520 nm and emission at 590 nm according to the method of Birnboimand Jevcak (26). Under the conditions employed ethidium bromidebinds preferentially to double-stranded DNA. PMNL (4 x IO7) werepreincubated at 37*C for 15 min in a volume of 3.4 ml HBSS in 9-mmplastic l'etri dishes, followed by addition of benoxaprofen and MNL (2x IO7)in a final volume of 6 ml HBSS. PMNL-free and benoxaprofen-

free control systems were included. The dishes were incubated for 15min at 37"C after which the nonadherent MNL were separated fromadherent PMNL, enumerated, and resuspended to 10"/nil in 250 HIM

mesoinositol, 10 mMsodium phosphate, l mM MgCh (pH 7.2). 200 ¿ilof MNL suspension were lysed in alkaline medium containing 9 Murea, 10 mMNaOH, 2.5 HIMcyclohexanediaminetetraacetate, and 0.1%sodium dodecyl sulfate. Ethidium bromide fluorescence was determinedafter a 60-m in incubation period at 15*C. The results were calculated

according to the formula, I) (percentage of double-stranded DNA) =(F - Fmin)/(Fm„- Fmin) x 100, where F is the fluorescence of thesample, /•'„„„the background fluorescence determined in samples that

were sonicated at the beginning of the unwinding period in order toinduce maximal unwinding, and /•'„,„,is the fluorescence of samples

kept at pH 11.0, which is below the pH level needed to induce unwinding of double-stranded DNA. This system was also used for testing theeffect of H-7 (50 and 100 MM),catatase (200 units/ml), and SOD (100units/ml) on the frequency of DNA single-strand breaks in MNL

exposed to benoxaprofen (30 fig/ml) and PMNL individually and incombination, with and without exposure to UV radiation.

UV Irradiation of PMNL and MNL. The UV source was a Philip's

MLV irradiation lamp, 300 W (Philips Electronics Ltd, Johannesburg,South Africa), emitting at 50 cm from the lamp UVA 1.69 mW/cm2,UVB 0.88 mW/cm2, and UVC 0.01 mW/cm2. On the basis of prelim

inary experiments a 1-min exposure time was chosen which corresponded to 0.13 J/c-nr' UVA. Briefly, the Petri dishes containing the

PMNL were irradiated for 1 min immediately after the addition ofMNL and benoxaprofen. Thereafter subsequent incubation and assaysfor DNA strand breaks in cocultured MNL were performed as describedabove. The following control systems were included: (a) irradiatedMNL only, (b) irradiated PMNL and MNL, and (c) irradiated MNLand benoxaprofen. For these investigations benoxaprofen was used atconcentrations of 10 and 30 fig/ml.

Protein Kinase C Purification and Assay. PKC was chromatographi-cally purified from brain tissue by using DEAE ion-exchange-, phenylsepharose-, and protamine agarose columns in sequence as previouslydescribed (27).

The activity of PKC was assayed by measuring the incorporation of[32P]ATP into lysine-rich histone (type 11IS) as previously describedwith minor modifications (28). Briefly, the 100-;¿1reaction mixture

contained 20 mM Tris (pH 7.5), 20 ng histone, 2.5 mM EGTA, 10 HIMMgCl2 in the absence or presence of various amounts and combinationsof PS, Ca2*, diolein, and benoxaprofen as indicated. Reactions wereinitiated by the addition of 1 nmol [32P]ATP containing (0.5-1) x 10'cpm and allowed to incubate for 6 min at 30"C before spotting ontoWhatman 3MM paper squares and washing in trichloroacetic acid. 32P

Incorporation in the protein precipitates were estimated by Cerenkovcounting and PKC activity was expressed as nmol "P transferred/min/

mg enzyme protein by subtracting EGTA backgrounds from stimulatedactivity.

Expression and Statistical Analysis of Results. The results are expressed as the mean values ±the standard error of the mean for eachseries of experiments. The numbers of experiments are indicated in thetables and figures. Statistical analyses were performed by Student's t

test (paired t statistic) by comparison of systems containing benoxaprofen with the corresponding matched benoxaprofen-free control system.

RESULTS

Chemiluminescence Responses of PMNL. The kinetics ofbenoxaprofen induced activation of LECL as compared withcontrol are shown in Fig. 1. The response was linear for 2-3min, peaked at 6 min, and thereafter subsided. A dose-responseeffect was also obtained by recording the LECL peak at 6 minafter the addition of varying concentrations of benoxaprofen(Fig. 2). This activation was statistically significant for benoxaprofen concentrations >7.5 ng/ml and yielded stimulationindices of 2.4,3.7,6.5, and 12 for benoxaprofen concentrationsof 7.5, 15, 30, and 60 ng/ml, respectively. Benoxaprofen-me-diated stimulation of LECL was completely eliminated by theinclusion of 100 units/ml of SOD.

Effects of H-7 on Benoxaprofen-activation of LECL. H-7caused dose-related inhibition of the LECL responses of PMNLactivated with 30 fig/mi benoxaprofen (Fig. 3). The inhibitionobtained with 12.5, 25, 50, and 100 /IM H-7 was 47, 63, 79,and 95% of the maximal response, respectively. This translates

101

EI

Ot—

oLLJ

Z

5-

5 10 20TIME — minutes

Fig. 1. Kinetics of LECL following the addition of 60 Mg/ml of benoxaprolento PMNL (•)and of benoxaprofen-free control PMNL (O). Dala, mean values±SEM in mV-s"' of three separate experiments.

30BENOXAPDOFEN (M m1I

60

Fig. 2. Measurement of LECL in PMNL following the addition of varyingconcentrations of benoxaprofen (3.75-60 Mg/ml)- Data, peak values in mV s ',

recorded 6 min after the addition of benoxaprofen; means ±SEM of fourexperiments. Background LECL values for control unstimulated PMNL havebeen subtracted. Statistically significant activation of LECL was observed withbenoxaprofen at concentrations of 7.5 (p < 0.01), 15, 30, and 60 fig/m[ (P •0.005).

3095

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

DNA DAMAGE BY BENOXAPROFEN AND PHAGOCYTES

Fig. 3. Measurement of the effects of PKC inhibitor H-7 (12.5-100 /¿M)onthe peak LECL responses of PMNL activated with 30 ¿ig/mlbenoxaprofen. Data,mean values ±SEM of three determinations in mV-s"1. Background values for

control unstimulated PMNL have been subtracted.

into an K\(, value of 13 MMas indicated and compares wellwith the IC50 value of 7 ¡IMfor PKC inhibition in vitro (29).

Effects of Benoxaprofen and PMNL on DNA Strand Breaksin MNL. Exposure of MNL to autologous PMNL or varyingconcentrations of benoxaprofen individually did not affect thefrequency of DNA strand breaks in MNL. Percentages of intactds DNA for MNL only, MNL + PMNL, and MNL + 60 ng/ml benoxaprofen were 91 ±4,89 ±3, and 88 ±4%, respectively(mean value ±SEM of three separate experiments). In contrast,coincubation with PMNL and benoxaprofen markedly increased the frequency of mononuclear DNA strand breaks (Fig.4). Moreover, this effect was dependent on the concentrationof benoxaprofen, yielding 91 ±4, 76 ±4, 64 ±4, 44 ±3, 30 ±4, and 29 ±5% intact ds DNA in control systems and systemscontaining benoxaprofen concentrations of 5, 10, IS, 30, and60 Mg/ml, respectively.

Effects of UV Radiation on Benoxaprofen-activated PMNL.These results are shown in Tables 1 and 2. Exposure of benox-aprofen-act ivatc-d PMNL to UV radiation significantly poten

tiated the frequency of DNA single strand breaks in coculturedMNL. In PMNL-free control systems containing MNL only orMNL and benoxaprofen no effects of UV radiation were observed. However in the benoxaprofen-free control system con-

100]

80a-

60

D%

40

20

—¡5— 30 w

BENOXAPHOFEN ()ii| nil)

Fig. 4. Measurement of the percentage of double-stranded DNA (l>) remainingin MNL cocultured with benoxaprofen (5-60 ¿ig/ml)only (O) or with benoxaprofen and autologous PMNL (•).Data, mean values (D) ±SEM of three separateexperiments performed in triplicate. Statistically significant increases in thefrequency of DNA strand breaks were observed at benoxaprofen concentrationsof 5 (p< 0.05), 10 (p < 0.025), 15, 30, and 60 ^g/ml (all p < 0.005), respectively.

Table 1 Measurement of effects of benoxaprofen and PMNL with and without aI-min exposure to UY radiation on the structural integrity ofds DNA of

cocultured autologous MNLResults are expressed as the mean percentage of ds DNA remaining (±SEM)

of triplicate determinations of three separate experiments, p < 0.005 and p <0.05 by comparison of systems e with i and of/withy, respectively.

Percentage of double-strandedTest system DNA remaining in MNL

a. MNL onlyb. MNL + 10 iig/ml benoxaprofenc. MNL + 30 /jg/ml benoxaprofend. MNL + PMNLe. MNL + PMNL + 10 ng/m\ benoxaprofenf. MNL + PMNL + 30 fig/ml benoxaprofeng. MNL + UVRh. MNL + PMNL + UVRi. MNL + PMNL + UVR + 10 jig/ml benox

aprofenj. MNL + PMNL + UVR + 30 Mg/ml benox

aprofen

85 ±286 ±283 ±281 ±564 ±429 ±584 ±568 ±537 ±3

14 ±4

Table 2 Measurement of the effects of a 1-min exposure to UV radiation onspontaneous and benoxaprofen-mediated activation of LECL in PMNL

Results are expressed as the mean values ±SEM in mV •sec ' of four different

experiments, p < 0.05, p < 0.01, and p < 0.025 for comparison of systems a withb, c with d, and e with/, respectively.

Test system LECL (r.l.u.)

a. PMNL onlyb. PMNL + UVRc. PMNL + 10 jig/ml benoxaprofend. PMNL + 10 Mg/ml benoxaprofen + UV ra

diatione. PMNL + 30 »ig/mlbenoxaprofenf. PMNL + 30 jig/ml benoxaprofen + UV ra

diation

49 ±664 ±5

155 ±5207 ±27

275 ±34336 ±33

Table 3 Measurement of the effects of SOD (100 units), catatase (200 units), andH-7 (SOand 100 ¡ÕM)on the structural damage inflicted on the DNA of

cocultured MNL by benoxaprofen (30 pg/mil-activated autologous PMNL

Results are expressed as the mean percentage (±SEM)ds DNA remaining oftriplicate determinations of three different experiments. SOD and both concentrations of H-7 caused statistically significant (p < 0.005) protection of DNA.

Test systemPercentage of ds DNA

remaining in MNL

MNL + PMNL 82 ±3MNL + PMNL + benoxaprofen 20 ±5MNL + PMNL + benoxaprofen + catalase 31 ±2MNL + PMNL + benoxaprofen + SOD 70 ±2MNL + PMNL + benoxaprofen + catalase + SOD 74 ±2MNL + PMNL + benoxaprofen + 50 pM H-7 61 ±2MNL + PMNL + benoxaprofen + 100 >.MH-7 69 ±1

taining PMNL and MNL a slight, but statistically significant(p < 0.01) increase in the frequency of DNA strand breaks wasobserved. UV radiation exposure also potentiated the LECLresponses of benoxaprofen-activated PMNL (Table 2) and thiseffect was eliminated by H-7 (IC50 = 11 MM)and SOD.

Effects of Catalase, Superoxide Disimilase, and H-7 on theFrequency of DNA Strand Breaks in MNL Exposed to PMNLand Benoxaprofen. These results are shown in Table 3. SODcaused statistically significant protection of DNA from damagemediated by PMNL and benoxaprofen (30 Mg/ml), while catalase had a modest protective effect. The combination of catalaseand SOD was not significantly better than SOD alone. H-7 atboth concentrations tested (50 and 100 MM)protected MNLfrom DNA single strand breaks inflicted by PMNL and benoxaprofen. Corresponding results with UV radiation-exposed,benoxaprofen-treated PMNL are shown in Table 4. SOD andH-7, but not catalase, protected cocultured MNL from oxidantdamage to DNA mediated UV radiation-exposed, benoxaprofen-activated PMNL.

Interactions of Benoxaprofen with PKC. Benoxaprofen (30Mg/ml) failed to stimulate PKC activity by itself in the absence

3096

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

DNA DAMAGE BY BENOXAPROFEN AND PHAGOCYTES

Table 4 Measurement of the protective effects of SOD, catatase, and H-7 on the ¿r- 80 -istructural integrity ofds DNA in MNL exposed to irradiated, benoxaprofen-

activated (30 tig/ml) PMNLResults of triplicate determinations of three different experiments are ex

pressed as the mean percentage (±SEM)ds DNA. SOD and both concentrationsof H-7 provided statistically significant protection (p< 0.005).

Test systemPercentage of ds DNA

remaining in MNL

a. MNL +PMNLb.MNL + PMNL +benoxaprofenc.MNL + PMNL + benoxaprofen + UV ra

diationd.MNL + PMNL + benoxaprofen + UV ra

diation +catatasee.MNL + PMNL + benoxaprofen + UV ra

diation +SODf.MNL + PMNL + benoxaprofen + UV radiation + catatase +SODg.MNL + PMNL + benoxaprofen + UV ra

diation + 50 nMH-7h.MNL + PMNL -1-benoxaprofen + UV ra

diation + lOOjiM H-782

±430±810±511

±576

±682

±251

±259

±2

Table 5 Measurement of PKC activity in the presence of various combinations ofallosteric effectors with or without 30 tig/ml benoxaprofen

Results are expressed as the mean values ±SEM of triplicate assays. Percentage of change is calculated as increase in PKC activity with benoxaprofen abovePKC activity without benoxaprofen. Two further experiments gave identicalresults.

Allosteric effectors in as-PKC activity (nmol/min/mg)

saysystema.Ca2*,100„Mb.

PS,25ng/mlc.Diolein, 5fig/mld.Ca2* + PS(sameamounts)e.

Ca2* + diolein(sameamounts)f.

Ca2* + PS +diolein(same

amounts)—benoxaprofen3.4

±0.84.4±1.412.2±0.779.9±3.430.5

±0.383.9

±1+

benoxaprofen6

±0.63.9±0.518±1.779.1±1.340.7

±3.580.5

±3.5change76-1147033-4

of Ca2+ and diolein (not shown). When these two allosteric

effectors were present alone or in combination, however, activity was consistently (n = 3 experiments) enhanced in the presence of benoxaprofen (Table 5, a, c, and e). This apparent effectwas negated in the presence of saturating amounts of PS (Table4, d and /), but could again be seen if nonsaturating amountsof this phospholipid were present (Fig. 5A). These results wouldargue that benoxaprofen operates via the PS binding site andwe subsequently demonstrated that benoxaprofen stimulatedPKC activity dose dependently over a similar dose range as wasrequired for inducing LECL in intact PMNL (Fig. 5B). Again,it proved necessary to have no or nonsaturating amounts of PSpresent in the reaction vial to demonstrate the dose effect.Notice that at the highest concentrations (60 Mg/ml) at whichthe drug was tested, no plateau of activity was achieved. Attempts to use higher concentrations were complicated by precipitation in the reaction vials.

DISCUSSION

The mutagenic and carcinogenic potential of phagocyte-derived reactive oxidants (9-11) and the association of chronicinflammation and cancer (30-33) have been documented.PMNL may also contribute to drug toxicity and carcinogenesisby prooxidative mechanisms. Interactions of the phagocyte-derived oxidants hypochlorous acid, sodium hypochlorite, andtaurine chloramine with chlorpromazine, aminopyrine, andphenylhydrazine lead to the generation of potentially toxic freeradical forms of these agents (34). In the present study we havedescribed an alternative mechanism of drug-mediated carcino-

Q. 70-o>

•

c 60E

BI 50

§O

1,5

1— —l—

6,25 25

PS CONCENTRATION (jig/ml)

100

c 50

'eI 40

30-

25-

B

0,9 3,7 15

BENOXAPROFEN CONCENTRATION (jig/ml)

60

Fig. 5. A, determination of I'M activity in the presence of various amountsof PS without (O) or with (•)benoxaprofen. Free <'a'' was 100 »i\i.Data, mean

values ±SEM of three triplicate measurements at every point. Background activityof control tubes incubated in the absence of (.:>'"' and PS has been subtracted. II.

PKC activity in the presence of various amounts of benoxaprofen. The reactionvessels contained 100 UM( V 10 Mg/ml diolein, but PS was absent. Data, means±SEM of triplicate measurements.

genesis due to prooxidative interactions of benoxaprofen, phagocytes, and UV radiation.

Prior to its withdrawal from the international market in 1982clinical experience with benoxaprofen had been brief, but eventful. The drug was reported to be efficacious in the treatment ofrheumatoid arthritis (22) and various chronic inflammatoryskin disorders (19). However, reports of drug-related fatal hep-atorenal toxicity and a particularly high incidence of cutaneousside effects, especially phototoxicity, led to its withdrawal (19).It could be argued that further research on benoxaprofen wouldgenerate little interest. However it is noteworthy that conventional assays of drug toxicity failed to detect the unusually highincidence of cutaneous side effects (phototoxicity, onycholysis,and eruptive skin tumors on sun-exposed areas) which wereassociated with benoxaprofen chemotherapy (19, 20). We believe that important insight into novel mechanisms of drug-mediated toxicity and carcinogenesis can be learned from thebenoxaprofen experience.

The results of this investigation clearly demonstrate the potential of benoxaprofen to stimulate membrane-associated oxi-dative metabolism in human phagocytes leading to the releaseof reactive oxidants which damage DNA in bystander cells. Wehave previously described the activation of Superoxide generation by benoxaprofen in PMNL using a ferricytochrome creduction assay (17). In the present investigation we have useda lucigenin-enhanced chemiluminescence method (23) which

3097

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

DNA DAMAGE BY BENOXAPROFEN AND PHAGOCYTES

has been reported to detect Superoxide and to be approximately10-fold more sensitive than the cytochrome c method (35). Inadditional control experiments (data not shown) we were unableto demonstrate stimulatory effects on the LECL responses ofPMNL of supernatants from MNL which had been pulsed withbenoxaprofen (60 /ig/ml). This clearly demonstrates that theobserved effects on DNA strand breaks are due to PMNL/benoxaprofen interactions and not to stimulatory factors released by drug-treated MNL. The primary mediator of DNAstrand breaks in MNL cocultured with benoxaprofen-activatedPMNL appears to be Superoxide since inclusion of SOD andH-7, an inhibitor of PKC (24) and Superoxide generation (25),protected the MNL. Catalase, on the other hand, was largelyineffective. PMNL and benoxaprofen individually did not alterthe structure of DNA in cocultured MNL. The prooxidativeinteractions of benoxaprofen and human PMNL were significantly potentiated by UV radiation. We have previously reported that this is due to the UVA component of UV radiationand is due to a sensitizing effect on PMNL and apparently notto photoactivation of the benoxaprofen molecule (17). In support of this others have reported that UV radiation exposureprimes the membrane-associated oxidative and degranulationresponses of PMNL to hyperreact to receptor-mediated activation with the synthetic jV-formylated chemotactic tripeptideyV-formyl-methionyl-leucyl-phenylalanine (36). UV radiationexposure of benoxaprofen-activated PMNL was associated witha significantly greater frequency of DNA strand breaks incocultured MNL relative to the corresponding nonirradiatedsystems. A small, but statistically significant increase in DNAstrand breaks was also observed in benoxaprofen-free controlsystems containing PMNL and MNL and is probably attributable to the slight increase in oxidant generation observed inPMNL irradiated in the absence of benoxaprofen. The association of benoxaprofen chemotherapy with the appearance oferuptive skin tumors on sun-exposed areas could have been adirect consequence of prooxidative interactions of skin phagocytes with the NSAID and UV radiation leading to the generation of tumor-promoting reactive oxidants. Importantly theconcentrations used in the present investigation did not exceed60 ng/ml- During chemotherapy with benoxaprofen (600 mgdaily) the mean serum concentrations recorded in 10 patientswith rheumatoid arthritis after 3 and 6 months of treatmentwere 66 ±16 and 81 ±18 ¿ig/ml,respectively (22).

Benoxaprofen-mediated activation of oxidant generation inPMNL was extremely sensitive to inhibition by H-7 with anIC\„value of 13 UM. We have previously reported that benoxaprofen activates PKC activity in crude extracts of humanplatelets (18). Although we proposed that benoxaprofen mayact at the phosphatidylserine binding site, our data were inconclusive ( 18). In the present study, we have extended our previousinvestigations by measuring the effects of benoxaprofen on theactivity of highly purified PKC in the presence of varyingamounts of allosteric effectors.

Benoxaprofen stimulated PKC activity in a dose-dependentmanner in the presence of diolein and nonsaturating amountsof PS. No further stimulation was observed if the PS concentration exceeded 25 ¿¿g/ml,which suggests that the drug mayenhance a PS effect at its allosteric site. Whether this is due tocompetitive binding or to some other more complex stoichio-metric interaction between enzyme-lipid vesicles, substrate andallosteric effectors is unknown at this stage. It has recently beensuggested that with histone, substrate aggregates may form withlipid vesicles in such a way that access of calcium-dependent,membrane-bound enzyme to its substrate is facilitated (37).

Conceivably, benoxaprofen may exert more than one effect inthis complex stoichiometrical environment; this may dependon whether the drug is primarily intercalated in the vesiclemembrane or not. This is clearly important in the intact celllevel where benoxaprofen is able to stimulate pro-oxidativeactivity despite the presence of saturating amounts of PS. Themechanism of priming of PMNL membrane-associated oxidative metabolism by UV radiation remains to be established.

In conclusion we have shown that UV-potentiated prooxidative interactions which occur between human phagocytes andbenoxaprofen lead to the generation of tumor-promoting reactive oxidants. Although benoxaprofen is no longer availablethese observations are relevant to the early detection of thetoxic and carcinogenic potential of new pharmacological agents.

REFERENCES

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Cerutti, P. A. Pro -oxidant states and tumor promotion. Science (Wash. DC),227:375-381, 1985.Kensler, T. W., Bush, D. M., and Kozumbo, W. J. Inhibition of tumorpromotion by a biomimetic Superoxide dismutase. Science (Wash. DC), 221:75-77, 1983.Borek, C., and Troll, W. Modifiers of free radicals inhibit in vitro theoncogenic actions of X-rays, bleomycin, and the tumor promoter 12-0-tetradecanoylphorbol 13-acetate. Proc. Nati. Acad. Sci. USA, 80: 1304-1307, 1983.Bimboim, H. C., and Kanabus-Kaminska, M. The production of DNA strandbreaks in human leukocytes by Superoxide anión may involve a metabolicprocess. Proc. Nati. Acad. Sci. USA, 82:6820-6824, 1985.Weitberg, A. B., Weitzman, S. A., Clark, E. P., and Stessei, T. P. Effects ofanti-oxidants on oxidant-induced sister chromatid exchange formation. J.Clin. Invest., 75:1835-1841, 1985.Birnboim, H. C. DNA strand breaks in human leukocytes induced by super-oxide anión, hydrogen peroxide and tumor promoters are repaired slowlycompared to breaks induced by ionizing radiation. Carcinogenesis (Lond.),7:1511-1517,1986.Schraufstatter, I. U., Hinshaw, D. B., Hyslop, P. A., Spragg, R. G., andCochrane, C. G. Oxidant injury of cells: DNA strand breaks activate poly-adenosine diphosphate-ribose polymerase and lead to depletion of nicotin-amide adenine dinucleotide. J. Clin. Invest., 77: 1312-1320, 1986.Babior, B. M. Oxidants from phagocytes: agents of defense and destruction.Blood, 64:959-966, 1984.Weitzman, S. A., and Stossel, T. P. Mutation caused by human phagocytes.Science (Wash. DC), 212: 546-547, 1981.Weitberg, A. B., Weitzman, S. A., Destrempes, M., Latt, S. A., and Stossel,T. P. Stimulated human phagocytes produce cytogenetic changes in culturedmammalian cells. N. Engl. J. Med., 308:26-30, 1983.Weitzman, S. A., Weitberg, A. B., Clark, E. P., and Stossel, T. P. Phagocytesas carcinogens: malignant transformation produced by human neutrophils.Science (Wash. DC), 227:1231-1233, 1985.Cox, J. A., Jeng, A. Y., Sharkey, N. A., Blumberg, P. M., and Tauber, A. I.Activation of neutrophil nitotinamide adenine dinucleotide phosphate(NADPH)-oxidase by protein kinase C. J. Clin. Invest., 76: 1932-1938,1985.Tauber, A. I. Protein kinase C and the activation of the human neutrophilNADPH-oxidase. Blood, 69: 711-720, 1987.Clark, R. A., Leidal, K. G., Pearson, D. W., and Nauseef, W. M. NADPHoxidase of human neutrophils: subcellular localisation and characterisationof an arachidonate-activatable superoxide-generating system. J. Biol. ( "hem.,262:4065-4074, 1987.Anderson, R., Lukey, P. T., Naude, S. P. E., and Joone, G. Benoxaprofen: apro-oxidant anti-inflammatory drug. Agents Actions, 14: 238-246, 1984.Anderson, R., and Joone, G. Inhibition of polymorphonuclear leucocytemotility by benoxaprofen related to activation of cellular oxidative metabolism. Int. J. Immunopharmacol., 6: 269-274, 1984.Anderson, R., and Eftychis, H. A. Potentiation of the generation of reactiveoxidants by human phagocytes during exposure to benoxaprofen and ultraviolet radiation in vitro. Brit. J. Dermatol., 115: 285-295, 1986.Lukey, P. T., Anderson, R., and Dippenaar, U. H. Benoxaprofen activatesmembrane-associated oxidative metabolism in human polymorphonuclearleucocytes by apparent modulation of protein kinase C. Br. J. Pharmacol.,93: 289-294, 1988.Allen, B. R. Benoxaprofen and the skin. Brit. J. Dermatol., 709: 361-364,1983.Findlay, G. H., and Hull, P. R. Eruptive tumours on sun-exposed skin afterbenoxaprofen. Lancet, /: 95, 1982.Kama!, A., and Koch, I. M. Pharmacokinetic studies of benoxaprofen ingeriatric patients. Eur. J. Rheumatol. Inflamm., 5: 76-81, 1982.Anderson, I. F., Naude, S. P. E., Eftychis, H. A., Anderson, R., and de Klerk,E. The anti-rheumatic effect of benoxaprofen. S. Afr. Med. J., 63:923-925,1982.

3098

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

DNA DAMAGE BY BENOXAPROFEN AND PHAGOCYTES

23. Dahlgren, C, Aniansson, II.. and Magnusson, K. Pattern of l'ornivi -nu-thio-

nyl-leucyl-phenylalanine-induced luminol and lucigenin-dependent chemilu-minescence in human neutrophils. Infect. Immuni)!., 45: 1-5, 198S.

24. Hidaka, II.. Inagaki, M.. Kawamoto, S.. and Sasaki, Y. Isoquinolinesulfon-amides, novel and potent inhibitors of cyclic nucleotide-dependent proteinkinase and protein kinase C. Biochemistry, 23: 5036-5041, 1984.

25. Fujita, I.. Takeshige, K., and Minakami, S. Inhibition of neutrophil super-oxide formation by l-(5-isoquinolinesulfonyl)-2-methylpiperazine (H-7), aninhibitor of protein kinase C. Biochem. Pharmacol., 35: 4555-4562, 1986.

26. Bimboim, H. C., and Jevcak, J. J. Fluorometric method for rapid detectionof DNA strand breaks in human white blood cells produced by low doses ofradiation. Cancer Res., 41: 1889-1892, 1981.

27. Woolen, M. W., Vanderplas, M., and Nel. A. E. Rapid purificai ¡onof proteinkinase C from rat brain: a novel method employing protamine agarose affinitycolumn chromatography. Eur. J. Biochem., 164:461-467, 1987.

28. Kikkawa, I'., Takai, Y., Minakuchi, K.. Inohara, S., and Nishizuka, Y.

Calcium activated phospholipid dependent protein kinase from rat brain. J.Biol. Chem., 257: 13341-13348, 1982.

29. Nel, A. E., Schabort, I., Rheeder, H., Bouic, P., and Wooten, M. W.Inhibition of CD-3 and PHA-mediated T cellular responses to inhibitors ofC-kinase activity and translocation with the aid of H-7. J. Immunol., in press,1988.

30. Manale, B. L., and Brower, T. D. The significance of bacterial flora incarcinoma in chronic osteomyelitis. Surg. Gynecol. Obstet., 136: 63-64,1973.

31. Kuntz, R. E., Cheever, A. M., and Myers, B. J. Proliferarne epithelial lesionsof the urinary bladder of non-human primates infected with Schistosomahematobium. J. Nati. Cancer Inst., 6S: 223-235, 1972.

32. Gilben, J. B. Tumors of the testis following mumps orchitis. J. I ml.. 51:296-300, 1944.

33. Saeed, W., Kim, S., and Burch, B. H. Development of carcinoma in regionalenteritis. Arch. Surg., 108:376-379, 1974.

34. Kalyanaraman, B., and Sohnle, P. G. Generation of free radical intermediatesfrom foreign compounds by neutrophil-derived oxidants. J. Clin. Invest., 75:1618-1622, 1985.

35. Minkenberg, I., and Ferber, E. Lucigenin-dependent chemiluminescence asa new assay for NADPH-oxidase activity in paniculate fractions of humanpolymorphonuclear leukocytes. J. Immunol. Methods, 71:61-67, 1984.

36. Bredberg, A., and Forsgren, A. Long wavelength IV radiation affects chemiluminescence of human polymorphonuclear leucocytes. Photochem. Photo-biol., 41: 337-341, 1985.

37. Bazzi, M. I>., and Nelsestuen, G. L. Role of substrate in imparting calciumand phospholipid requirements to protein kinase C activation. Biochemistry,26:1974-1982, 1987.

3099

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1988;48:3094-3099. Cancer Res   Gwen Schwalb, Albertus D. Beyers, Ronald Anderson, et al.   Leukocytes, and Ultraviolet RadiationInteractions of Benoxaprofen, Human PolymorphonuclearLeukocytes by Protein Kinase C-dependent Prooxidative Promotion of DNA Strand Breaks in Cocultured Mononuclear

  Updated version

  http://cancerres.aacrjournals.org/content/48/11/3094

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/48/11/3094To request permission to re-use all or part of this article, use this link

on March 2, 2020. © 1988 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from