Mouse keratinocytes derived from initiated skin or papillomas are resistant to DNA strand breakage by benzoyl peroxide: a possible mechanism for tumor promotion mediated by benzoyl peroxide

Alkaline elution was used to examine DNA single-strand breaks in cultured normal and carcinogen-altered mouse keratinocytes exposed to 12-O-tetradecanoyl phorbol-13-acetate and benzoyl peroxide. Seven cell lines derived from carcinogen-induced mouse skin papillomas and three cell lines derived from N-methyl-N'-nitro-N-nitrosoguanidine-treated non-tumor bearing mouse skin were resistant to phorbol ester-mediated DNA strand breaks after 6 or 24 h. Normal keratinocytes sustained strand breaks after 24 h but not after 6 h. Benzoyl peroxide induced extensive strand breaks in normal keratinocytes at both 6 and 24 h, and this was associated with marked cytotoxicity. In contrast, 9 of 10 cell lines showed complete or partial resistance to strand breaks following benzoyl peroxide exposure. It is proposed that differential resistance to DNA strand breaks and cytotoxicity among normal and carcinogen-altered keratinocytes provides the biological basis for the promoting action of benzoyl peroxide. Furthermore, sublethal DNA damage in preneoplastic or neoplastic keratinocytes may account for the potency of benzoyl peroxide in causing malignant conversion.     

Hydrogen peroxide formation and DNA base modification by tumor promoter-activated polymorphonuclear leukocytes

This report shows that generation of hydrogen peroxide (H₂O₂)by human polymorphonuclear leukocytes (PMN_s) activated wtihtumor promoters of varying potency as first and second stagepromoters correlates well with activities of these promotersin vivo. Those tested were 12-O-tetradeca-noylphorbol-13-acetate(TPA), a complete promoter, 12-0-retinoylphorbol-13-acetate(RPA), a synthetic TPA derivative almost devoid of first stageactivity in some strains of mice, and mezerein (Mez), a potentsecond stage and much weaker first stage promoter. Mez-stimulatedPMN_s produced up to four times less H₂O₂ whereas RPA-stimulatedPMN_s pro-duced up to 10 times less H₂O₂ than TPA-activated cellswhen used at concentrations between 0.5 and 15 nM to activate7.5–8.5 x 10 PMNs/ml. Phorbol, a non-promoter, was totallyinactive in this assay. Furthermore, the tumor pro-moter-activatedPMNs caused formation of 5-hydroxymethyl-2'-deoxyuridine (HMdU)and thymidine glycol (dTG) in DNA co-incubated with those cells.The amounts of modified thy-midines formed, particularly ofHMdU, correlated well with first stage tumor promoting efficacyand with the amount of H₂O₂ that was generated by promoter activatedPMNs. In comparison with TPA, Mez-or RPA-stimulated PMNs in-ducedformation of 25 or 70% less H₂O₂ and 30 or 75% less HMdU, respectively,under conditions favoring HMdU for-mation. Thus, formation ofeither H₂O₂ by tumor promoter-stimulated phagocytes or HMdUin DNA exposed to those activated cells may serve as a measureof potency as a first stage tumor promoter. Formation of modifiedbases such as HMdU in DNA might constitute the genetic changeimparted by the first stage tumor promoters.     

Radiation-like modification of bases in DNA exposed to tumor promoter-activated polymorphonuclear leukocytes

Oxygen species generated by human polymorphonuclear leukocytes (PMNs) activated by 12-O-tetradecanoylphorbol-13-acetate (TPA) caused the formation of 5-hydroxymethyl-2'-deoxyuridine (HMdUrd), and (+) and (-) diastereoisomers of cis-thymidine glycol (dTG) in DNA that was exposed to them. There were 9 HMdUrds and 31 dTGs formed per 1 X 10(6) thymidine residues. When Fe(II)/ethylenediaminetetraacetic acid was added to TPA-activated PMNs at 0, 10, 15, and 20 min after TPA, HMdUrd formation increased 5-, 13-, 30-, and 35-fold. Although dTG was initially formed in larger amounts than HMdUrd, it eventually decreased but was still 5-, 6-, 5.5-, and 3-5-fold, respectively, higher than in the absence of iron. From 65 to 1800 times more HMdUrd was formed in DNA when autologous plasma was present during incubation of DNA with TPA-activated PMNs than in its absence. The levels of dTG also varied from about the same as HMdUrd to the nondetectable. Reconstituted human serum transferrin used instead of plasma or Fe(II) also supported the formation of HMdUrd and dTG. When DNA was treated with Fe(II)-reduced H2O2 in the absence of PMNs and TPA, both derivatives were formed. However, the same treatment of marker dTG of dTG-containing polydeoxyadenylic-thymidylic acid caused the decomposition of dTG. Thus, the reduction of hydrogen peroxide by Fe(II) complexed to either ethylenediaminetetraacetic acid or amino acids amy be responsible for the formation of HMdUrd and dTG and for subsequent decomposition of dTG in DNA exposed to the TPA-activated PMNs.     

Enhanced malignant progression of mouse skin tumors by the free-radical generator benzoyl peroxide

Chemical carcinogenesis in mouse skin can be divided into the processes of initiation, promotion, and progression. The free-radical generator benzoyl peroxide is moderately active during the promotion stage. Repetitive treatment of mouse benign skin tumors (papillomas) with benzoyl peroxide (20 mg, twice weekly) increased the number of cumulative carcinomas per group by 325% and the number of keratoacanthomas by 44% compared to tumor-bearing Sencar mice treated with the promoter 12-O-tetradecanoylphorbol-13-acetate. The lack of increase in the number of cumulative papillomas per group due to benzoyl peroxide treatment suggests that benzoyl peroxide enhanced the progression of preexisting papillomas. The ability of benzoyl peroxide to enhance the progression of benign tumors to cancer should be considered when determining the human risk from exposure to this widely used chemical agent; in addition, biological assays specifically testing malignant progression may be essential and beneficial for determining an agent's carcinogenic risk.     

Characterization of the induction of ornithine decarboxylase activity by benzoyl peroxide in SENCAR mouse epidermis

The induction of epiderinal ornithine decarboxylase (ODC) activityby benzoyl peroxide (BPO) was characterized to evaluate theusefulness of this effect as a short-term marker of BPO-inducedmouse skin tumor promotion. The maximal induced levels of ODCspecific activity, after a single topical dose of BPO, were>2-fold higher when a cold scraping method was used to prepareepidermis rather than the commonly used heat treatment method.Therefore, the cold scraping method was used for all the workreported here. Application of a single 20 mg dose of BPO tothe dorsal skin of SENCAR mice caused a relatively small inductionof epidermal ODC activity, to a level <1/40 that inducedby a single 2 µg dose of 12-O-tetradecanoylphorbol-13-acetate(TPA). Furthermore, the time-courses of induction were differentafter single doses of TPA and BPO, with peak activities observedat 6 and 24 h after treatment respectively. In contrast, aftera total of five 20 mg doses of BPO were topically applied (onedose every 2 days), ODC activity was transiently induced toan average level >15 times after a single dose. Additionally,on this dosing regimen, the peak of ODC activity shifted to4 h after the last treatment, so that the time-course of ODCinduction resembled that after multiple applications of TPA.The extent of epidermal ODC induction by BPO was found to bea complex function of the frequency of dosing and the numberof treatments. However, when BPO treatments were administeredfrom 1 to 7 days apart, similar maximal induced levels of ODCactivity were eventually achieved after application of multipledoses. Importantly, the dose-response for the induction of ODCactivity by five doses of BPO (applied one dose every 2 days)was highly correlated with published data on the dose-responsefor tumor promotion by this organic peroxide, indicating thatODC induction is a good short-term marker of BPO-induced tumorpromotion in SENCAR mice.     

Activation of N'-nitrosonornicotine by hydrogen peroxide in vitro

Betel-quid ingredients were found to produce reactive oxygen species, such as superoxide anion and hydrogen peroxide, in vitro. We demonstrated that N'-nitrosonornicotine (NNN) can be converted to its active metabolite, hydrogen peroxide, nonenzymatically in the presence of ferrous ions and ethylenediaminetetracetic acid (EDTA) at pH 7.2. Three ultimate metabolites of NNN--NNN-1-N-oxide, 4-hydroxy-4-(3- pyridyl)butyric acid and 4-oxo-4-(3-pyridyl)butyric acid--and nornicotine were detected by high-performance liquid chromatography. 3H-NNN and 14C-NNN interact with calf thymus DNA in the presence of hydrogen peroxide, ferrous ion and EDTA. The results suggest that formation of reactive oxygen species in the presence of NNN may be a key factor in the initiation of oral tumours in tobacco and betel-quid chewers.     

Generation of DNA base modification following treatment of cultured murine keratinocytes with benzoyI peroxide

BenzoyI peroxide (BzPO) is a free radical generating compoundthat acts as a tumor promoter and progressor in mouse skin.BzPO is cleaved in the presence of copper to produce benzoyloxyIand phenyI radicals. Treatment of mutation reporter plasmidswith BzPO and copper yields predominantly single-strand breaksand GT transversion mutations. To explore the role of base modificationsin the possible mammalian mutagenicity of BzPO the formationof 8-hydroxy-2'-deoxyguanosine (8-OHdG) within the DNA of culturedmurine keratinocytes was investigated. Treatment with 10 µMBzPO produced a maximum 3-fold increase in levels of 8-OHdGversus vehicle controls within1-2 h, with significant levelsof 8-OHdG persisting 6 h after initial exposure to BzPO. Pretreatmentwith the copper chelator bathocuproine disulfonic acid reducedthe levelsof 8-OHdG generated by BzOP ot near background. However,treatment with the iron chelator desferal did not. The stablemetabolic product of BzPO benzoic acid was ineffective in producing8-OHdG. Depletion of cellular glutathione with L-buthionine-(S,R)-sulfoximineincreased the amount of BzPO-generated 8-OHdG, while supplementationwith glutathione monoethyI ester reduced the number of 8-OHdGmolecules formed. Collectively, these results suggest that BzPOat non-cytotoxic concentrations undergoes copper-dependent activationto a reactive product to generate 8-OHdG within cultured murinekeratinocytes.     

Strain differences in mouse skin tumorigenesis by the promotion of 12-o-tetradecanoylphorbol-13-acetate or benzoyl peroxide

Strain difference of mouse skin tumorigenesis was studied with 12-O-tetradecanoylphorbol-13-acetate (TPA) and benzoyl peroxide (BzP) both in 8-week-old hypocatalasemic C3H mice (C-s(b)) and in normal C3H/HeNCrj mice (C3H). Two weeks after initiation by 4 mu mole MNNG, both strains of female mice were promoted by 10 nmole TPA or 10 mg, 20 mg BzP twice weekly for 30 and 52 weeks. The incidence of skin tumors in C-b(s) was significantly increased as compared to that in C3H mice promoted by TPA for 30 weeks but after 52 weeks of TPA promotion there was no difference between the 2 strains. By the promotion with 10 mg or 20 mg BzP, the incidence of skin tumor in C3H was 5% at both levels after 30 weeks of promotion and 19% and 38%, respectively, after 52 weeks of promotion. No skin tumors appeared in C-b(s) mice with BzP treatment after 30 weeks of promotion. After 52 weeks of promotion by 10 mg BzP, skin tumors were induced in only 15% of C-b(s) mice. The results show that genetic factors, especially radical scavenging enzymes, controlled susceptibility to skin tumorigenesis by promotion with either TPA or BzP.     

Covalent modification of DNA by daunorubicin

Daunorubicin, a clinically useful antitumor agent, induces mammary adenocarcinoma in Sprague-Dawley rats. As part of an investigation of the mechanism of tumor induction by daunorubicin, the formation of daunorubicin-DNA adducts has been investigated by 32P-postlabeling assay. Rat-liver DNA indubated with either 0.05 or 0.1 mM daunorubicin, rat-liver microsomes, and 5 mM reduced nicotinamide adenine dinucleotide phosphate (NADPH) for 1 h contained covalent DNA adducts in addition to the endogenous adduct profile present in control DNA. With 1.5 mM cumene hydroperoxide serving as a cofactor, higher levels of these two adducts and two additional adducts were formed, all of which most likely were daunorubicin-DNA adducts. This latter treatment also resulted in an intensification of three endogenous DNA modifications over levels occurring in control DNA. Covalent DNA alterations in vivo were studied in rats treated with 20 mg/kg daunorubicin for 2 days and 200 mg/kg on the 3rd day. Daunorubicin-DNA adducts as observed in vitro could not be detected in DNA of liver or mammary epithelial cells. The levels of endogenous modifications in drug-treated rats were increased by 200% in mammary DNA and by 50% in hepatic DNA as compared with controls. It was concluded from these experiments that daunorubicin may be metabolically activated to a reactive metabolite that binds covalently to DNA. These daunorubicin-DNA adducts may not play a role in tumor induction because they were not detectable in vivo. However, the increase in levels of endogenous DNA modifications induced by daunorubicin both in vitro and in vivo is consistent with a role of this class of DNA modification in the carcinogenic process.Abbreviations DNR daunorubicin or daunomycin - CuOOH cumene hydroperoxide - RAL relative adduct levels as defined by Reddy and Randerath [12] This work was supported by a grant from the National Institutes of Health, NCI (CA43233) Present address: Section of Gastrointestinal Oncology and Digestive Diseases, Department of Medical Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA     

Copper-dependent site-specific mutagenesis by benzoyl peroxide in the supF gene of the mutation reporter plasmid pS189.

Benzoyl peroxide (BzPO) enhances tumor promotion and malignant conversion in mouse epidermis. DNA damage may contribute to these processes. BzPO reacts with Cu(I) to produce the benzoyloxyl radical, which in turn causes strand breaks in plasmid DNA. In this study we investigated whether BzPO with or without Cu(I) caused promutagenic DNA damage in the supF gene of the mutation reporter plasmid pS189 replicating in human Ad293 cells. Exposure of pS189 in vitro to BzPO (0.1-1 mM) inhibited plasmid replication; however, addition of Cu(I) (0.1 mM) did not augment BzPO-induced plasmid toxicity. Exposure to BzPO with or without 0.1 mM Cu(I) was also associated with a concentration-dependent increase in mutation frequency, up to > 100-fold above the spontaneous mutation frequency. Supplemental Cu(I) was not required for mutagenesis; however, it both raised the maximal mutation frequency observed and lowered the threshold concentration of BzPO necessary to discern mutagenesis above the spontaneous background. Neither the hydroxyl radical scavengers mannitol or DMSO, the spin trap N-tert-butyl-alpha-phenylnitrone, nor reduced glutathione altered BzPO/Cu(I)-induced mutagenesis; however, mutagenesis was suppressed by the chelator EDTA. Twenty-four of 32 individual BzPO/Cu(I)-induced mutants characterized by sequencing contained point mutations; 22/25 point mutations occurred at G-C base pairs. There were five large deletions and four small deletions. Three additional BzPO-induced mutants contained four point mutations, all occurring at G-C base pairs. Two BzPO/Cu(I)-induced mutational clusters at d(pGGG)-d(pCCC) sites were observed. These data suggest that BzPO may interact with Cu(I) bound to G-C base pairs in DNA to produce site-specific promutagenic DNA damage.     

Hydrogen ion-mediated enhancement of cytotoxicity of bis-chloroethylating drugs in rat mammary carcinoma cells in vitro

Aerobic glycolysis, a metabolic characteristic of malignant cells, can be exploited to increase the concentration of lactic acid selectivity in tumor tissues in vivo by systemic administration of glucose (E. J?hde and M. F. Rajewsky, Cancer Res., 42: 1505-1512, 1982). To investigate whether a more acidic microenvironment can enhance the effectiveness of cytocidal drugs, we have analyzed the colony-forming capacity of M1R rat mammary carcinoma cells exposed to bis-chloroethylating agents in culture as a function of extracellular pH (pHe). At pHe 6.2 the cytotoxicity of 4-hydroperoxycyclophosphamide, as measured by inhibition of colony formation, was potentiated by a factor of approximately 200 as compared to pHe 7.4. Similar results were obtained with mafosfamide, nitrogen mustard, nornitrogen mustard, melphalan, and chlorambucil; not, however, with ifosfamide. As indicated by experiments using the ionophor nigericin for rapid equilibration of pHe and intracellular pH (pHi; measured with pH-sensitive microelectrodes), modulation of drug action by varying pHe primarily resulted from the concomitant decrease in pHi. The acidic microenvironment enhanced cytotoxicity most effectively during the phase of cellular drug uptake and monofunctional alkylation of DNA. DNA cross-link formation appeared to be less affected by pH, and lowering of pHe during the phase of cross-link removal was only marginally effective.     

Nickel(II)- and cobalt(II)-dependent damage by hydrogen peroxide to the DNA bases in isolated human chromatin

Nickel compounds are known to be carcinogenic to humans and animals. Cobalt compounds produce tumors in animals and are probably carcinogenic to humans. The mechanisms of the carcinogenicity of these metal compounds, however, have remained elusive. In the present work, we have investigated the ability of Ni(II) and Co(II) ions in the presence of H2O2 to cause chemical changes in DNA bases in chromatin extracted from cultured cells of human origin. Eleven modified DNA bases in chromatin were identified and quantitated by the use of gas chromatography-mass spectrometry. 2-Hydroxyadenine (isoguanine), which has not previously been shown to occur DNA or chromatin, was also identified. Products identified were typical hydroxyl radical-induced products of DNA bases, suggesting that the hydroxyl radical was involved in their formation. This idea was supported by partial inhibition of product formation by typical scavengers of hydroxyl radical. Partial inhibition of product formation indicated a possible "site-specific" formation of hydroxyl radical by unchelated Ni(II) and Co(II) ions bound to chromatin. Although treatment of chromatin for 1 h with Co(II)/H2O2 caused formation of significant amounts of products, treatment with Ni(II)/H2O2 required incubation times of more than 5 h and an increase in Ni(II) concentration before increases in product amounts above background levels became detectable. In both cases, ascorbic acid did not increase product yields. Glutathione at a physiologically relevant concentration had little overall effect on DNA base modification. Superoxide dismutase increased the yields of most products. Chelation of Ni(II) and Co(II) ions with EDTA almost completely inhibited product formation. Ni(II) in the presence of H2O2 produced greater base damage to the DNA in chromatin than to isolated DNA, unlike other metal ions tested. DNA damage in chromatin caused by Ni(II) and Co(II) ions in the presence of H2O2 may contribute to the established genotoxicity and carcinogenicity of these metal ions.     

Effects of formaldehyde, acetaldehyde, benzoyl peroxide, and hydrogen peroxide on cultured normal human bronchial epithelial cells

The effects of several aldehydes and peroxides on growth and differentiation of normal human bronchial epithelial cells were studied. Cells were exposed to formaldehyde, acetaldehyde, benzoyl peroxide (BPO), or hydrogen peroxide (HPO). The effect of each agent on the following parameters was measured: (a) clonal growth rate; (b) squamous differentiation; (c) DNA damage; (d) ornithine decarboxylase activity; (e) nucleic acid synthesis; (f) aryl hydrocarbon hydroxylase activity; and (g) arachidonic acid and choline release. None of the agents were mitogenic, and their effects were assessed at concentrations which reduced growth rate (population doublings per day) to 50% of control. The 50% of control concentrations for the 6-h exposure were found to be 0.065 mM BPO, 0.21 mM formaldehyde, 1.2 mM HPO, and 30 mM acetaldehyde. BPO-exposed cells were smaller than controls (median cell planar area, 620 sq microns versus 1150 sq microns), and acetaldehyde-exposed cells were larger than controls (median cell planar area, 3200 sq microns). All agents increased the formation of cross-linked envelopes and depressed RNA synthesis more than DNA synthesis. HPO caused DNA single-strand breaks, while formaldehyde and BPO caused detectable amounts of both single-strand breaks and DNA-protein cross-links. Other effects included increased arachidonic acid and choline release due to HPO. The similarities and differences of the effects of these aldehydes and peroxides to those caused by tumor promoters are discussed.     

In vitro DNA modification by the ultimate carcinogen of 4-nitroquinoline-1-oxide: influence of superhelicity

The effect of DNA tertiary structure on in vitro modificationby 4-acetoxy-aminoquinoline-l-oxide (Ac-4-HAQO) was investigated.The reactivity of pAT153 plasmid DNA depended on the conformationalstate of the molecule: it progressively decreased accordingto the decrease of the superhelical tension, being negativelysupercoiled DNA about two times more susceptible than singly-nickedrelaxed DNA. HPLC of the three main Ac-4-HAQO adducts showedthat 3-(deoxyguanosin-N2-yl)-4-aminoquinoline-1-oxide, N-(deoxyguanosin-C8-yl)-4-aminoquinoline-l-oxideand 3-(deoxyadenosin-N6-yl)-4-aminoquinoline-l-oxide accountedfor 50, 25 and 10% of total quinoline DNA base adducts in allDNA conformations tested, except in the negatively super-coiledtopoisomers where they accounted for 80, 15 and 5% respectively.DNA modification by Ac-4-HAQO resulted also in the formationof apurinic/apyrimidinic sites and in strand scissions. Thequantification of these damages revealed that they representan important fraction of all damaging events and that theiryield is also influenced by DNA superstructure. Thus, theselesions must be considered as important DNA damage induced byAc-4-HAQO.     

The promotion of wound healing following chemosurgery (Mohs' technique) by dressings with a lotion of benzoyl peroxide

A lotion of benzoyl peroxide was applied postoperatively to chemosurgical wounds in a small group of patients. The technique of utilizing dressings of benzoyl peroxide is described and illustrated. Enhanced wound healing was observed in all cases. The effects appear to be due, at least in part, to antimicrobial activity, debriding activity, hyperbaric oxygenation, and stimulation of granulation.     

In vitro reaction of N-n-butyl-N-nitrosourea and n-butyl methanesulphonate with guanine and thymine bases of DNA

The reaction of N-n-butyl-N-nitrosourea (BNU) and n-butyl methanesulphonate(BMS) with DNA has been investigated. In addition to the expectedn-butylpurines formed on reaction with BNU some rearranged sec-butyl-adductswere also observed, indicating that a carbonium ion is involvedin the alkylation process by nitrosoureas. Only n-butyl adductswere observed following reaction with BMS. The reaction of thesealkylating agents with the thymine residues in the DNA was alsodetermined. The relative amounts of the various adducts formedby the two butylating agents are similar to those formed bythe corresponding ethyl and propyl compounds.     

Elevated expression and point mutation of the Ha-ras proto-oncogene in mouse skin tumors promoted by benzoyl peroxide and other promoting agents

The two-stage skin carcinogenesis model of initiation and promotion in SENCAR mice has been used to examine the effects of various tumor-promoting agents on the expression of the Ha-ras oncogene in early stages of tumorigenesis in vivo. Papillomas were induced in 7,12-dimethylbenz[a]anthracene (DMBA)-initiated SENCAR mouse epidermis by (i) complete promotion with benzoyl peroxide; (ii) complete promotion with 12-O-tetradecanoyl phorbol-13-acetate (TPA); and (iii) two-stage promotion with TPA for 2 weeks followed by mezerein for 9 weeks. Results of Northern blot hybridization analyses show that early papillomas contain significantly elevated levels of Ha-ras polyadenylated [poly(A)+] RNA, irrespective of the type of tumor promotion regimen used. This pattern holds for promoters of the phorbol ester class as well as for the free radical generating agent benzoyl peroxide. Furthermore, digestion of tumor DNA with diagnostic restriction endonucleases demonstrated that 9-week-old papillomas induced by DMBA contained a point mutation in the 61st codon of one allele of the Ha-ras gene. The results represent the earliest stage in the development of a papilloma at which a Ha-ras point mutation has been reported.     

Release of sulfur mustard-modified DNA bases by Escherichia coli 3-methyladenine DNA glycosylase II

The toxic effects of sulfur mustard have been attributed toDNA modification with the formation of 7-hydroxyethyl-thioethylguanine, 3-hydroxyethylthioethyl adenine and the cross-link,di-(2-guanin-7-yl-ethyl)sulfide. To investigate the action ofbacterial 3-methyladenine DNA glycosylase II (Gly II) on theseadducts, calf thymus DNA was modified with [¹⁴C]sulfur mustardand used as a substrate for Gly II. Gly II releases both 3-hydroxyethylthioethyladenine and 7-hydroxyethylthioethyl guanine from this substrate.In comparison with the activity of Gly II towards methylatedDNA, 3-hydroxyethylthioethyl adenine is released somewhat moreslowly than 3-methyladenine, while 7-hydroxyethylthioethyl guanineis released much more readily than 7-methylguanine. Glycosylaseaction may play a role in protecting cells from the toxic effectsof sulfur mustard.     

Differential oxidative stress induced by two different types of skin tumor promoters, benzoyl peroxide and 12-O-tetradecanoylphorbol-13-acetate

The oxidative stress induced in vivo by benzoyl peroxide (BzPo) or 12-O-tetradecanoylphorbol-13-acetate (TPA) was evaluated in terms of chemiluminescence (CL) emitted by SENCAR mouse skin, a non-invasive method that allows an estimation of overall oxidative stress. The ability of a biomimetic superoxide dismutase, copper(II)(3,5-diisopropylsalicylate)2 (CuDIPS), to inhibit that response was also evaluated. A single application of BzPo to mouse skin resulted in a dose-dependent increase in CL up to 0.083 mumol. Sequential treatment with BzPo in a dose used for tumor promotion resulted in a fall in CL induced by the second topical application. There were no differences between initiated and non-initiated mice in their responses to BzPo-induced CL. CuDIPS, an inhibitor of tumor promotion, was an effective inhibitor of CL in all the protocols evaluated. Conversely, ZnDIPS and DIPS did not inhibit CL. Phenolic antioxidants induced partial inhibition of CL. Unlike BzPo treatment, a single application of TPA up to 105 nmol did not induce an increase in CL, but the second topical application with TPA in a dose used for tumor promotion resulted in a small but significant increase in CL. However, these values of CL were much smaller than the CL induced by BzPo. Our results show a differential response of the skin in terms of the oxidative stress induced by BzPo or TPA.