Chem. Res. Toxicol. 1993,6, 731-738
731
Role of Quinone Methide in the in Vitro Toxicity of the Skin Tumor Promoter Butylated Hydroxytoluene Hydroperoxide Kathryn Z. Guyton,+John A. Thompson,$ and Thomas W. Kensler*p+ Division of Toxicological Sciences, Department of Environmental Health Sciences, Johns Hopkins School of Hygiene and Public Health, Baltimore, Maryland 21205, and Division of Pharmaceutical Sciences, School of Pharmacy, University of Colorado Health Sciences Center, Denver, Colorado 80262 Received May 10, 1993"
An electrophilic quinone methide (2,6-di-tert-butyl-4-methylene-2,5-cyclohexadienone, BHTQM) functions in the toxicity of butylated hydroxytoluene (BHT) in both rodent liver and lung. BHT-QM has also been demonstrated to mediate tumor promotion in mouse skin by another (BHTOOH). metabolite of BHT, 2,6-di-tert-butyl-4-hydroperoxy-4-methyl-2,5-cyclohexadienone In the present study, the role of BHT-QM in the cytotoxicity of BHTOOH was investigated. The toxicity of BHTOOH was potentiated by glutathione depletion and inhibited by thiol compounds, indicating that BHTOOH is activated to a thiol-reactive, toxic intermediate. This activation process was suggested to be iron-dependent by the ability of an Fe(II1)-specific chelator to inhibit BHTOOH toxicity. Comparative study of analogs of BHTOOH in which the 4-methyl group was substituted with CD3, ethyl, isopropyl or tert-butyl supported the hypothesis that BHT-QM mediates this toxicological response. The decreased rate of reactivity of quinone methides that occurs as the 4-alkyl group is enlarged was accompanied by a corresponding reduction in toxic potency. The structural requirements for quinone methide toxicity were also explored with a series of BHTOOH analogs substituted a t the 2- and 6-positions of the molecule. Reducing the lipophilicity of the 2,6-tert-butyl groups is known to increase quinone methide reactivity with glutathione but does not diminish the rate of quinone methide formation from the hydroperoxide. Interestingly, alteration of only one of the tert-butyl groups did not change the toxic potency, whereas removal or replacement of both tert-butyl groups dramatically reduced the toxicity in control cells but not glutathione-depleted cells. These results suggest that the ultimate target for BHT-QM is not glutathione itself, but perhaps instead a sulfhydryl moiety located in a more lipophilic environment. Indeed, covalent binding of BHTOOH to proteins predominated in the noncytosolic fraction. Electron microscopy studies showed prominent mitochondrial and nuclear changes upon treatment with BHTOOH. Taken together, these results provide a composite picture of BHTOOH-mediated toxicity in keratinocytes that may apply more generally both to mouse skin tumor promotion by BHTOOH as well as to the deleterious actions of BHT in its target tissues.
Introduction The extensive metabolism of butylated hydroxytoluene1 (BHT) in its target tissues to reactive intermediates is thought to mediate the adverse effecta of this food antioxidant in biological systems. An early report that BHT undergoes covalent binding to lung macromolecules in a metabolism-dependent manner suggested the involvement of an electrophilic intermediate in the toxicity of BHT (1). Cellular glutathione appears to play a strong
* To whom correspondence should be addressed. t Johns Hopkins School of 8 University of Colorado.
Hygiene and Public Health.
Abstract published in Advance ACS Abstracts, September 1,1993. Abbreviations: BHT, butylated hydroxytoluene;BHTOOH, 2,6-ditert-butyl-4-hydzoperoxy-4-methyl-2,5-cyclohexadienone; BHT-QM,2,6di-tert-butyl-4-methylene-2,5-cyclohexadienone; 4-CDa-BHTOOH,2,6d i - t e r t - b u t y l - 4 - [ a , a ~ - ~ m e t h y l - 2 , 6 c y4-ethyl-BHTOOH, c; 2,6-di-tert-butyl-4-hydroperoxy-4-ethyl-2,5-cyclohexadienone; 4-isopropyl-BHTOOH, 2,6-di-tert-butyl-4-hydroperoxy-4-isopropyl-2,5-cycloheladienone;4-tert-butyl-BHTOOH,2,4,6-tri-tert-butyl-4-hydroperoxy2,5-cyclohexadienone;BHTOH,2,6-di-tert-butyl-4-hydroxp4-methyl-2,5cyclohexadienone; BSO, L-buthionine-(S,R)-sulfoximine; DMF, N,Ndimethylformamide; MTT, 3-(4,5-dimethylthiazo1-2-y1)-2,5diphenyltetrazolium bromide; PBS, 1X Dulbecco's phosphate buffered saline (pH 7.4). 0
role in the detoxication of this reactive intermediate (2). A series of structure-activity studies at the 4-position of the molecule by Mizutani and co-workers have strongly suggested that a quinone methide mediates the toxicity of BHT in both mouse liver and lung (3-5). Moreover, quinone methide formation may also function in the toxicity of BHT in other species. For example, in vivo production of 2,6-di-tert-butyl-4-methylene-2,5-cyclohexadienone (BHT-QM)from BHT has been documented in rat liver (6) and bile from rats (7)as well as mice (8). Indeed, BHT-QM was directly shown to inhibit vitamin K epoxide reductase in vitro and to mediate hemorrhagic death in rats in vivo (9, 10). Thus BHT-QM appears to mediate toxicity in multiple organ systems among rodent species. The quinone methide metabolites of compounds other than BHT can also invoke toxicological responses. The toxicity of eugenol is thought to require formation of a quinone methide (11, 12); like BHT-QM, the quinone methide formed from eugenol may also be detoxified by GSH conjugation (13). Further, the lung toxicity (14)as well as tumor promoting capacity (15) of a tert-butyl-
0893-228~/93/2706-0731$04.00/0 0 1993 American Chemical Society
732 Chem. Res. Toxicol., Vol. 6, No. 5, 1993
hydroxylated analog of BHT is proposed to occur through quinone methide production. BHT-QM can also function in a tumor-promoting capacity. While direct experimental evidence implicating BHT-QM as a mediator of tumor promotion in the target tissues of BHT is lacking, this intermediate is required for tumor promotion by 2,6-dit e r t - butyl-4-hydroperoxy-4-methyl-2,5-cyclohexadienone (BHTOOH) in mouse skin initiated with 7,12dimethylbenz[alanthracene. In fact, the tumor promoting capacity of hydroperoxide analogs of BHTOOH has been shown to be directly correlated with rate of formation of quinone methide from these compounds (16). The interaction of quinone methide intermediates with biological targets may therefore play an important role in the processes of toxicity and tumor promotion. Depending on the treatment regimen, BHT can induce either tumor promotion or acute toxicity in its target tissues. Moreover, selective toxicity in normal cells may be a contributing mechanism in tumor development by this and other tumor promoters. In response to selective cytotoxicity in the normal population, initiated cells may undergo a compensatory stimulation of growth (I7).The production of BHT-QM in the target cell of BHTOOH in the mouse skin, the keratinocyte, is known to lead to tumor promotion. The toxicity of BHTOOH in these target cells may also be caused by the formation of BHT-QM. The present study was undertaken to investigate the hypothesis that BHT-QM mediates the in vitro toxicity of BHTOOH and to explore the cellular determinants of this response.
Guyton et al.
matography on CH3OH-deactivated silica gel (955 dichloromethane/ethyl acetate) to produce a 14% yield of the hydroperoxide. 2-tert-Butyl-4-hydroperoxy-4-methyl-2,5-cyclohexadienone: lH NMR (2H-CHC13)6 6.79 (dd, J = 9.8, 3.2 Hz, Ha), 6.60 (d, J = 3.2 Hz, H3), 6.19 (d, J = 9.8 Hz, &),1.18 (8, CH3), 1.12 [s, C(CH3)3]; UV (acetonitrile) 232 nm; thermospray MS (filament-on, 70:30 acetonitrile/water) m/z 238 [59%, (MH CHsCN)+], 197 (loo%, MH+). 2,6-Dimethyl-4-hydroperoxy2,5-cyclohexadienone: lH NMR (2H-CHC13)6 6.85 (dd, J = 9.5, 3.0H~,Hs),6.62(d,J=3.0H~,H3),6.27(d,J=9.5H~,He),1.90 (8, 2-CH3), 1.35 (8, 4-CH3); UV (2575 CHsCN/water) 231 nm; thermospray MS (filament-on, 2080 CHSOH/water) m/z 172 [loo%, (M + HzO)+l, 155 (51%, MH+). Hydroperoxides were confirmed to be >95% pure from other BHT derivatives by the HPLC procedure of Wand and Thompson (22). Cell Line Derivation, Culture, and Treatment. Murine epidermal papilloma cell line PA was kindly provided by Dr. James E. Strickland (National Cancer Institute). This cell line was derived from a papilloma induced in female SENCAR mice by a 7,12-dimethylbenz[a]anthracene/l2-0-tetradecanoylphorbol-13-acetateinitiation-promotion carcinogenesisprotocol (25). The cells were maintained in a 37 OC humidified atmosphere containing 7% COz in air. The culture medium consisted of Eagle's minimal essential medium (without CaC12) supplemented with chelex-treated fetal calf serum (8%)and 0.05 mM CaC12. For hydroperoxide treatment of the cells, the cellular growth medium was removed and replaced with reduced-serum (0.5%) treatment medium to which hydroperoxide dissolved in DMF had been added. The solvent (DMF) concentration in this medium was held constant at 0.1 % . All studies were performed in subconfluent cultures. Cell Viability Determination by Colorimetric Assay for Reduced MTT. The toxicity of BHTOOH was determined by assay for the reduction of the tetrazolium-based compound MTT Experimental Procedures (26).Cells were plated at a density of 2500 cells/well into 96-well tissue culture plates. GSH depletion (>go%) was achieved by Chemicals and Syntheses. Caution: the following chem18 h pretreatment with 0.5 pM L-buthionine-(S,R)-sulfoximine icals are hazardous and should be handled carefully: phenol, eel4,and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo- (BSO). Cells were treated with the appropriate experimental concentrations of hydroperoxide for 24 h, after which time 0.1 lium bromide (MTT). All syntheses should be conducted in an mg of MTT (50 p L of 2 mg/mL solution in 1X Dulbecco's efficient fume hood. All chemicals were purchased from Sigma phosphate buffered saline, pH 7.4 (PBS)) was added to each (St. Louis, MO) or Aldrich (Milwaukee, WI) unless otherwise well. Followinga 4-h incubation period,the medium was removed noted. Desferal (deferoxaminemesylate)was from CIBA-GEIGY and 100 pL MezSO was added to each well to dissolve the MTT (Summit, NJ). Solvents used in syntheses and HPLC analyses reduction product. Only live cells have the mitochondrial were of the highest quality commercially available, and NJVdehydrogenaseactivity required for reduction of MTT to ita blue dimethylformamide (DMF) was redistilled before use. The formazan product. This product has an absorbance maximum monoethyl ester of GSH was synthesized by the method of of 505 nm in MeZSO, and thus the number of viable cella could Anderson et al. (18). Phenolic analogs of BHT were synthesized be calculated from the blank-subtracted A m values determined as previously described (16,19,20).[ring-l4C1BHT(65.1 mCi/ with a microtiter plate reader. mmol) was prepared by Chemsyn ScienceLaboratories (Lenexa, Trypan Blue Determination of Cell Viability. Thiols and KS) and was kindly provided by Dr. Michael A. Trush (Johns metal chelators alone caused reduction of MTT and, therefore, Hopkins School of Hygiene and Public Health). Synthesis and could not be used in the MTT assay for cell viability. Instead, chemical characterizationof BHTOOH, [ring-WIBHTOOH, 2,6trypan blue exclusion was employed to assess the toxicity of di-tert-butyl-4-[cr,cr,cr-~H~]methyl-2,5-cyclohexadienone (4-CD3BHTOOH) and 2,4,6-tri-tert-butyl-4-hydroperoxy-2,5-cyclo- BHTOOH in the presence of these compounds. Subconfluent cells were treated in 100cmzdisheswith a cytotoxicconcentration hexadienone (4-tert-butyl-BHTOOH) followed the method of of BHTOOH (30pM). Metal chelatorsor soluble thiol compounds Karasch and Joshi (21)as detailed in Guyton et al. (16). The were added simultaneously with BHTOOH to the treatment syntheses of the following compounds have also been described previously: 2,6-di-tert-butyl-4-hydroperoxy-4-ethyl-2,5-cyclo- medium. N-Ethylmaleimide was added together with N-acehexadienone (4-ethyl-BHTOOH), 2,6-di-tert-butyl-4-hydroper- tylcysteine to the treatment medium 5 min prior to the addition of BHTOOH in DMF. Twenty-four hours after BHTOOH oxy-4-isopropyl-2,5-cyclohexadienone (4-isopropyl-BHTOOH) treatment, the cells were rinsed with ice-cold PBS, and 0.08% (22),6-tert-butyl-2-(hydroxy-tert-butyl)-4-methylphenol (23), 2-tert-butyl-4,6-dimethyl-4-hydroperoxy-2,5-cyclohexadi-trypan blue in PBS was then added directly to the adherent enone,and 4-hydroperoxy-2,4,6-trimethyl-2,5-cyclohexadienone cells. The fraction of cells excludingtrypan blue was determined (24). Preparations of 2-tert-butyl-4-hydroperoxy-4-methyl-2,5- in a representative field under high power (32X). Approximately cyclohexadienone and 2,6-dimethyl-4-hydroperoxy-2,5-cyclo- 100 cells per plate were scored by this method. hexadienone were carried out by the photooxygenation of, Covalent Binding. [ring-14C]BHTOOH (0.5pCi per 100cmz respectively, 2-tert-butyl-4-methylphenol and 2,4-dimethylpheplate, correspondingto 4 p M ) was added to subconfluent cultures nol. The general procedure involved irradiation with a 275-W in serum-free medium for covalent binding studies. GSH sun lamp of a solution containing 12.2 mmol of the phenol and depletion (>go%) was achieved by 18 h of pretreatment with 36 mg of the sensitizer 5,10,15,20-tetraphenyl-21H,23H-porphine BSO. Following a 2-h incubation period, the cells were rinsed with PBS until no further radioactivity was recovered in the in 100 mL of CCl, at 4 "C for 45 h. The solvent was removed wash (typically 10 washes). DNA was isolated by the method of by evaporation, and the crude product purified by flash chro-
+
Chem. Res. Toxicol., Vol. 6, No. 5, 1993 733
Toxicity of the Tumor Promoter BHT Hydroperoxide
I
I
+Ethvl-GSH
60
L
60
0
-
BSO
+
BSO
50 40
u 0.1
1
10
Concentration o f BHTOOH
30
7,1 "
20
(pM)
Figure 1. Toxicity of BHTOOH in keratinocyte cell line PA. Cells were cultured in 96-well plates, and cell viability was determined by MTT assay 24 h after exposure to BHTOOH (0). Depletion of GSH (>go%)was achieved by 18 h pretreatment with BSO (m). To raise intracellular GSH levels, cells were preincubated with ethyl ester of GSH for 1 h, whereupon the cells were rinsed with PBS 3 times to remove any residual extracellular GSH prior to treatment with BHTOOH (v).Values represent mean f SD for 12 wells. Gross-Bellardet al. (27)with multiple washes of DNA-containing fractions throughout the procedure. Isolated DNA was resuspended in water for quantification by diphenylamine assay (28) and determination of covalent binding. For the assessment of protein binding, cells were homogenized by freeze-thawing. A particulate fraction of proteins was generated by centrifugation for 3 min at 12000g. Proteins were isolated either from whole homogenate8 or from the soluble and particulate fraction of cells by ice-cold CHsOH precipitation and subsequent centrifugation at l2000g at 4 OC. Pelleted proteins were washed and reprecipitated with ethanol until no further radioactivity was detected in the wash (typically 8-10 washes). Proteins were resuspended in 1N NaOH for assessment of radioactivity covalently bound, and protein content was determined by modified Lowry assay
10 n
BHTOOH 4-CD3
4-Et
4-iPr
4-tBu
BHT
BHTOH
Figure 2. Cytotoxic ICWdeterminations for 4-alkyl analogs of BHTOOH in PA cells. The 4-methyl group of BHTOOH was sequentially enlarged to CDs, ethyl (Et), isopropyl (iPr), and tert-butyl (tBu). The parent compound BHT was also tested. BHTOH, the major detoxication product of the hydroperoxide, was not toxic at any concentration up to 60 pM.The ICs0 for each compound was determined directly from seven-point doseresponse curves BSO, with N = 12 wells for each point, as described in Figure 1. Values represent the mean of two to four separate experiments in which the ICWvalues for BHTOOH were consistent. Standard deviations were 110% of ICs0 values.
*
toxicity of BHTOOH. This ethyl ester will more readily cross the cellular membrane than does GSH, and ita administration has been shown to increase intracellular GSH levels (18). Taken together, these studies indicate a strong role for intracellular GSH in the direct detoxication of reactive intermediates formed from BHTOOH. Structure-Activity Analyses: 4-Alkyl Series. The toxic ICs0 values for a series of analogs of BHTOOH in (29). which the 4-methyl group has been increased in size Electron Microscopy Studies. Following 15 min or 1 h incrementally are shown in Figure 2. Each compound was treatment with a cytotoxic concentration of BHTOOH (30 pM) tested in both control and BSO treated (GSH-depleted) or solvent (DMF), cells were rinsed and fixed in 2% glutaralcells. The 4-fold potentiation of BHTOOH toxicity by dehyde, 1%paraformaldehyde in 0.1 M phosphate buffer (pH GSH depletion (shown in Figure 1)is seen as a reduction 7.4) starting at room temperature and then chilled to 4 "C. After from 12 to 3 pM of the ICs0 for BHTOOH in BSO treated 1h the cultures were washed in phosphate buffer 3X for 15 min cells. As the substituent at the 4-position is changed from each and then fixed for 1h in 1% OsO4,O.l M phosphate buffer. CH3 to CD3 and to the larger alkyl groups ethyl, isopropyl, Cells were washed in three changes of phosphate buffer at room temperature, dehydrated through graded alcohols (the propylene and tert-butyl, the toxic properties of the hydroperoxides oxide steps were omitted), and then imbedded with Poly/812 change in two ways from the parent compound. The IC50 (Polysciences,Inc.). Sections were made with an MJO Diatome in control cells increases, indicating a loss of toxic potency. diamond knife on an LKB-Vultramicrotome, stained with uranyl Further, the degree of potentiation by GSH depletion acetate (Fahmy's) and lead citrate, and viewed at 100 kV on a decreases. The parent compound BHT, which is not Phillips 410 transmission electron microscope. metabolized by keratinocytes to BHTOOH or other reactive intermediates, exhibited minimal toxicity in these Results cells. BHTOH, the major detoxication product of BHTOOH, was inactive as a toxin in this experimental system. Role of Intracellular Thiol Status in the Toxicity of BHTOOH. As shown in Figure 1,BHTOOH caused Structure-Activity Studies: Substitutions at the dose-dependent toxicity in cultured keratinocytes. A 2- and 6-Positions. The toxic ICs0 values for a series of 4-fold potentiation of toxicity was achieved by depletion analogs of BHTOOH in which one or both of the tertof cellular stores of GSH with BSO, a specific and potent butyl groups at the 2- and 6-positions has been changed inhibitor of GSH synthesis (30). GSH depletion apparor removed is shown in Figure 3. The IC50 was determined ently does not significantly impact the metabolism of in both control and GSH-depleted cells. By comparison BHTOOH in these cells; BSO treatment did not diminish with BHTOOH, the hydroxylated metabolite of BHTOOH the GSH peroxidase-dependent metabolic detoxication exhibited a similar toxicity under control and GSHof BHTOOH to 2,6-di-tert-butyl-4-hydroxy-4-methyl-2,5- depleted conditions despite that the reaction rate of the hydroxylated quinone methide with GSH is 6-fold higher cyclohexadienone (BHTOH) as determined by HPLC analyses.2 Preincubation of the cells for 1 h with the than that of BHT-QM (8). For other analogs in which monoethyl ester of GSH offered protection against the only one tert-butyl group was replaced by a methyl group or a hydrogen, the toxicity also remained unchanged. Alteration of both tert-butyl groups, however, generated 2 Guyton and Kensler, unpublished observations.
734 Chem. Res. Toxicol., Vol. 6, No. 5, 1993
60
i
v
I
0-
BSO
+
BSO
Guyton et al.
60
I
50 40
t. 30
30
20
20
10
10
io_
2
?
0
-
n
Table 11. Iron Chelator Desferal Protects against Toxicity by BHTOOH treatment viability (% control)" solvent (0.15% DMF) 98 f 2 BHTOOH (30pM) 2422 + 0.1 mM Desferal 40 10 + 0.5 mM Desferal 92 f 2 + 1.0 mM Desferal 96* 1 BHTOOH (30 pM) 3*2 + 0.1 mM BCSb lf2 + 0.5 mM BCS 4f3 + 1.0 mM BCS 2f2 a Cell viability was assessed by trypan blue exclusion 24 h after exposure to BHTOOH (30pM) and the indicated concentrations of chelators. Values represent mean (N = 3) SE. BCS, bathocuproinedisulfonic acid.
0 BHTOOH t - B U - O H
BDMP
BMP
TMP
DMP
Figure 3. Cytotoxic ICw determinations for 2,6-substituted analogs of BHTOOH in PA cells. In the first BHTOOH analog, the 2-tert-butylgroup of BHTOOH was hydroxylated (t-BuOH). The hydroperoxides of the following phenols were also tested: 2-tert-butyl-4,6-dimethylphenol (BDMP); 2-tert-butyl-4-methylphenol (BMP); 2,4,6-trimethylphenol (TMP); and 2,4-dimethylphenol (DMP). Cytotoxic ICWdeterminations for 2,6substituted analogs of BHTOOH in PA cells were performed as described in Figure 2. Table I. Thiols Protect against Toxicity by BHTOOH treatment viability (% contro1)a solvent (0.1% DMF) 99 f 1 BHTOOH (30 pM) 5h3 + 0.1 mM MPGb 50 f 16 + 0.5 mM MPG 93 f 2 + 1.0 mM MPG 95 3 BHTOOH (30 pM) 4f2 + 0.1 mM NACC 50 f 10 + 0.5 mM NAC 93f 1 + 1.0 mM NAC 95 f 3 + 0.5 mM NAC + 0.45 mM NEMd 4f2 a Cell viability was assessed by trypan blue exclusion 24 h after exposure to BHTOOH (30 pM) and the indicated concentrations of thiols. Values represent mean (N = 3) f SE. b MPG, 2-mercaptopropionylglycine. NAC, N-acetylcysteine. d NEM, N-ethylmaleimide.
analogs that were not particularly toxic under control conditions but had comparable toxicity to BHTOOH when GSH had been depleted. Protection Against the Toxicity of BHTOOH by Thiols. As shown in Table I, the thiol compounds N-acetylcysteine and 2-mercaptopropionylglycineoffered compete protection against a cytotoxic dose of BHTOOH. This protection is presumed to occur through the conjugation of reactive intermediates generated from BHTOOH at the sulfhydrylmoietiesof these compounds. The ability of N-ethylmaleimide, which specifically oxidizes sulfhydryls, to abolish the protective effects of N-acetylcysteine substantiates the role of free sulfhydryl groups in the protective actions of these thiol compounds. Protection Against the Toxicity of BHTOOH by Desferal. The protective effect of the Fe(II1) chelator Desferalagainst the toxicity of BHTOOH is shown in Table 11. In the absence of BHTOOH, the chelators alone were not toxic. By contrast with Desferal, the copper chelator bathocuproinedisulfonic acid did not offer protection against BHTOOH. These results suggest that iron, rather than copper, is involved in the activation of BHTOOH to toxic reactive intermediates. Covalent Bindingof [Ring-W]-BHTOOH in Intact Cells. As shown in Table 111, BHTOOH undergoes
Table 111. Covalent Binding of BHTOOH Enhanced by GSH Depletion* treatment protein binding (nmol/mg) DNA binding control 96.9 f 6.1b not detectablec not detectable BSO 251.3 f 15.2 2.6 fold enhancement with BSO a GSH depletion (NO% ) was achieved by addition of BSO (10 pM) 18 h prior to the addition of 0.5 pCi (4pM) [WIBHTOOH in serum-free medium for 2 h. Values represent mean (N= 3) f SD. Covalent binding to protein was labile to NaOH (1 N). c Not detectable,
