Mutation Research. 279 (1992) 83-89 0 1992 Elsevier Science Publishers B.V. All rights reserved Ol~5-l2l~/~2/$OS.t~t~
x3
MUTGEN 01753
Enhanced mutagenicity of anisidine isomers in bacterial strains containing elevated N-acetyltransferase activity David C. Thompson ‘,*, P. David Josephy b, Joseph and Thomas E. Eling a
W.K. Chu b
’ LaboratovofMolecular Biophysics, National Institute of Enrirorm~ental Health Sciences, Research
Triangle park. NC 27709 (~.SA.) and ’ Department of Chemistry and Biochemistry, Unirersity of Gue!ph. Guelph. Ont. NIG 2WI (Canada)
(Received 2 May 1991) (Revision received 10 September 1991) (Accepted 18 September 1991)
Keywords: p-Anisidine;
o-Anisidine; N-Acetyltransferase
Summary In previous studies on the mutagenicity of anisidine isomers, the orrho isomer was considered to be mutagenic towards standard Ames tester strains, while the para isomer gave equivocal results. In the present study we show that both para- and o&o-anisidine isomers are mutagenic in a Salmonella typhimurium tester strain containing elevated levels of N-acetyltransferase (YG1029). p-Anisidine gave a positive mutagenic response using either hamster S9 or ram seminal vesicle microsomes (RSVM) as an activating system, white o-anisidine gave a positive response only with the hamster S9 fraction. The mutagenic response from p-anisidine was greater than with o-anisidine in each case. In tests with p-anisidine and RSVM, the addition of arachidonic acid was not necessary to observe a mutagenic response. Catalase produced a dose-dependent decrease in the mutagenic response with p-anisidine and RSVM; this indicates that endogenous hydrogen peroxide from the bacteria acts as a substrate for the peroxidase activity of RSVM prostaglandin H synthase. These results demonstrate that both anisidine isomers are mutagenic and that N-acetyltransferase enzymes play an important role in their metabolism to mutagenic species.
Our laboratories are interested in the activation of aromatic amines to mutagenic and carcinogenic metabolites by prostaglandin H syn-
Correspondence: Dr. Thomas E. Eling, Eicosanoid Biochemistry Section, NIEHS, P.O. Box 12233, Research Triangle Park, NC 27709 (U.S.A.).
* Present address: Department of Medical Pharmacology and Toxicology, Texas A and M University, College Station, TX 77843 (U.S.A.).
thase (PHS). This enzyme exhibits peroxidase activity and is broadly distributed in mammalian tissues, including the urinary bladder, which is the site of carcinogenesis for many aromatic amines [ 1,2]. Two isomeric forms of anisidine, ortho-anisidine (Zmethoxyaniline) and paraanisidine (4-methoxyaniline), have markedly different carcinogenic properties. o-Anisidine elicits urinary bladder tumors in both mice and rats while p-anisidine is inactive [3-51. However, panisidine is not considered an established noncar-
cinogcn because it appeared to be associated with an increased incidence of preputial gland tumors in low dose male rats. In bacterial mutagenicity tests o-anisidine gave a positive response in Scrbnonella typhimutium TA1538 161.Studies evaluating possible p-anisidine mutagenicity have produced equivocal results. p-Anisidine produced mutagenic or questionable responses in TAN0 in several laboratories but consistent positive results were not obtained 16-81. Thus, the anisidine isomers represent an interesting case study in testing the correlation of mutagenicity and carcinogenicity results. Recently, new bacterial tester strains of Salmonella typhimrrrizdm (derived from TA1538) have been constructed which have greatly elevated sensitivities to aromatic amines and nitroaromatic compounds; these strains carry plasmid-borne copies of the genes encoding the bacterial enzymes acetyl CoA-dependent arylamine N-acetyltransferase/ arylhydroxylamine U-acetyltransferase (NAT/OAT) or nitroreductase 19,101. Our laboratory recently reported [l l] that the mutagenicity of benzidine is markedly enhanced in a strain containing elevated NAT/OAT activity (TA1538 1,8-DNP pYG121) compared with the strain from which it was derived (TA1538). The N-acetyltransferase gene has been sub-cloned to construct strains with even higher enzyme activities [12]. These new strains include YG1012 (derived from TA1538) and YG1029 (derived from TAZOOI. With YG1012, the improvement over TA1538 is about one thousand-fold in its ability to detect benzidine mutagenicity. The construction of these sensitive new strains has prompted the re-evaluation of mutagenicity data for aromatic amines, especially those, such as the anisidine isomers, which have previously yielded borderline responses. In the present study we describe the mutagenic response of both anisidine isomers in the Ames test using strains YG1012 and YG1029. We used either hamster S9 liver homogenate or ram seminal vesicle microsomes (RSVM, a rich source of PHS) as activating systems. We discuss the possible role of N-acetyltransferase enzymes in the mutagenicity and carcinogenicity of anisidine isomers.
Materials and methods Chemicals
Anisidine isomers (ortho and para) were purchased from Aldrich Chemical Co. (Milwaukee, WI). Mutagens were prepared the day of the experiment, dissolved in dimethyl sulfoxide (DMSO) and added in a volume not exceeding 50 pi/plate. Bovine liver catalase (type C-10, specific activity 2000 U/mg protein), horseradish peroxidase (type VII and bovine serum albumin were obtained from Sigma Chemical Co. (St. Louis, MO). Arachidonic acid was purchased from Nu Chek Prep, Inc. (Elysian, MN). Glucose 6phosphate and NADP+ were obtained from Boehringer-Mannheim Canada (Dorval, P.Q.). Hydrogen peroxide (30%) was obtained from Fisher Scientific (Fairlawn, NJ) and quantitated by UV absorption using a Hewlett Packard Model 8451 diode array spectrophotometer. 5-Phenyl-4pentenyl hydroperoxide (PPHP) and purified PHS were obtained from Oxford Biomedical Research, Inc. (Oxford, MI). Bacterial cell cultures Salmonella typhimurium
tester strains YG1012 (TA1538/1,8-DNP pYG213) and YG1029 (TAlOO pYG 219) were kindly provided by M. Watanabe, National Institute of Hygienic Sciences, Tokyo, Japan. Overnight cultures of YG1012 and YG1029 were grown at 37°C with shaking at 120 rpm in Oxoid Nutrient Broth No. 2 for about 10 h (YG1012) or 14 h (YG1029), until the OD,, reached 1.0. YG1012 and YG1029 were grown in the presence of ampicillin (25 pg/ml), and tetracycline (6.25 pg/ml) respectively.
Preparations fractions
of RSV microsomes
and hepatic S9
Ram seminal vesicle microsomes (RSVM) were prepared and y-sterilized as described previously [ 131. Hepatic post-mitochondrial supernatants (S9 fractions) from uninduced hamster liver were prepared as described [14]. Protein concentration of S9 fractions was determined by the modified Lowry assay of Peterson [153 and ranged from 25 to 41 mg/ml.
X5
Ames way
For the Ames test, mutagen fdissoived in 50 ~1 DMSO), RSVM or S9 (1 mg protein per plate) in buffer (60 mM KCI-100 mM sodium phosphate, pH 7.4, 0.5 ml), and bacterial culture (0.1 ml)
‘10
were combined, in that order. For the RSVMactivated test only, the mixture was incubated at 37°C for 3 min and arachidonic acid was then added, if so indicated, to give a concentration of 100 @M. The system was incubated for a further
L ,-, . . . .. 100 1000 KKloo
Dose (1 t nrnol/plate)
Dose (I+
nmol/plote
Fig. 2
O...
1
Dose(l Fig. 3
10
100
1000
KIx3x
Dose (1+ pg/Plate)
f nmol/plate) Fig. 4
Fig. 1. SP-activated rn~~ag~nicj~ of p-anisidine in YG1029. p-Anisidine, SP (1 mg protein per plate). and bacteria were incubated
at 37°C for 30 min before addition of tap agar and plating. Squares, soiid line: f SS; circles, dashed line: - S9 (direct-acting).
Fig. 2. ~S~M-activated mutagcnic~~ of of p-anisidine in YG1029. ~-Anisidine (in 50 ~1 DMSOf, RSVM U mg per plate), and bacteria were incubated at 37°C for 3 min: arachidonic acid (100 PM) was added, where indicated, to give a concentration of 100 FM, and the system was incubated a further 30 ruin before plating. Squares, solid line: f arach~donic acid: circles, dashed line: - arachidonic acid. The long dashed line (no symbols) reproduces the - S9 (direct-acting) data of Fig. 1.
Fig. 3. S9-activated mutagenicity af o-anisidine in YG1029. o-Anisidjne, S9 (I mg protein per plate), and bacteria were incubated at 37°C for 30 min before addition of top agar and plating. Squares. solid line: f S9; circles. dashed line: -S9 (direct-actin&.
Fig. 4. Effect of catalase on the RSVM-dependent mutagenicity of p-anisidine in YG1029. Catalase (circles. solid line)or bovine serum albumin (squares, dashed line) was added to the buffer, followed by RSVM (1 mg protein per plate) and p-anisjdin~ (10 *moles per plate). No arachidonic acid was added.
30 nun before addition of 2 ml of top agar and plating on minimal glucose agar. Colonies were counted after 48 h, using a 3M 620 automatic colony counter with a size threshold of 0.29 mm. For all figures, doses are plotted using the transformed variable X = (1 + dose) in logarithmic scale: this method allows the zero dose point to be included without breaking the axis. For each figure, data are pooled from at least two separate experiments, and each point represents the average of at least six plates. Error bars represent standard errors and, where not shown, are smaller than the symbols. Reducing cofactor assay
The ability of anisidine isomers to serve as reducing cosubstrates for horseradish peroxidase and PHS was measured using the hydroperoxide 5-phenyl-4-pentenyl hydroperoxide (PPHP) as previously described [ 16,171. Incubations contained 100 pM PPHP, 200 ,xM cofactor (when present) and either 60 nM horseradish peroxidase or 154 nM purified PHS. The horseradish peroxidase incubations were carried out in 0.1 M potassium citrate buffer (pH 5.5) while the PHS incubations were carried out at 37°C in 0.1 M phosphate buffer (pH 7.8). The amount of PPHP reduced in the presence and absence of cofactors was measured by HPLC. Results
Both anisidine isomers were tested for mutagenicity using the new Salmonella typhimurium tester strains, YGl012 and YG1029. With YGl012 neither isomer yielded a positive response using either RSVM or hamster S9 liver fraction as an activation system (data not shown). When tested with YG1029, however, p-anisidine caused a dose-dependent increase in revertants. p-Anisidine was mutagenic with either hamster S9 (Fig. 1) or RSVM (Fig. 2) as an activating system. The number of p-anisidine-dependent revertants observed with the two activating systems tS9, RSVM) was similar. With RSVM, the presence or absence of arachidonic acid (substrate for PHS) had no effect on the number of revertants (Fig. 2, compare solid and dashed curves). In the
absence of any activation system (Fig. 1, circles), p-anisidine produced very little mutagenicity. No significant toxicity was seen on these plates even at the highest dose (10 pmoles/plate). When activated by the hamster S9 system, o-anisidine elicited a dose-dependent increase in revertants (Fig. 3). No significant increase in revertants was observed in the absence of activation (circles). With RSVM activation, no significant o-anisidine-induced reversion was observed, whether arachidonic acid was present or not (data not shown). p-Anisidine gave a stronger mutagenic response with either activation system than did o-anisidine. We tested the possibility that endogenous (bacterial-derived) hydrogen peroxide was acting as a substrate for the RSVM-dependent mutagenicity of p-anisidine. Such hydrogen peroxide may explain the lack of arachidonic acid requirement for mutagenicity. Catalase (which lowers the effective concentration of hydrogen peroxide in the incubations, but does not catalyze p-anisidine oxidation) inhibited the mutagenicity in a dose-dependent manner (Fig. 4). On the other hand, bovine serum albumin (squares) had no effect on the number of revertants; therefore, the catalase effect was not due to a nonspecific protein effect. We also tested horseradish peroxidase as a possible activating system for p-anisidine in YG1029. Horseradish peroxidase catalyzes the oxidation of xenobiotics such as aromatic amines to free radical and other reactive intermediates, and is often used as a model system to mimic PHS reactions. We used a dose of 10 pmoles p-anisidine/ plate and various doses of hydrogen peroxide (O-O.2 pmoles/ plate). The horseradish peroxidase/ hydrogen peroxide system was unable to activate p-anisidine to a mutagen (data not shown). At the two highest doses of hydrogen peroxide examined (0.1 and 0.2 pmoles/ plate), some toxicity was evident. A comparison of the ability of PHS and horseradish peroxidase to utilize the anisidine isomers as reducing equivalents is shown in Table 1. With horseradish peroxidase, both anisidine isomers were efficient electron donors. In contrast, with PHS, neither isomer was a good electron donor. This suggests that the mode of oxida-
TABLE
I
ABILITY OF ANISIDINE ISOMERS TO SERVE OXIDASE REDUCING COFACTORS Reducing substrate
AS PER-
9%PPHP reduced (mean + SE) Prostaglandin H synthase
Control (no reducing substrate)
Horseradish peroxidase
5,l
2+1
Butylated hydroxyanisole
66+2
55+1
p-Anisidine
11+2
x2+2
o-Anisidine
5+1
100+0
is unclear whether the mutagenic species from both activation systems are the same. The metabolism of anisidine isomers by microsomal enzymes has not been thoroughly investigated. Both isomers undergo 0dealkylation [22]. Experin vitro (rat-liver microsomes) showed that the ortho isomer is dealkylated more readily than the para isomer [23]. We have recently described the horseradish peroxidase-catalyzed formation of reactive intermediates from both anisidine isomers which covaimentS
lently bound to nucleic acids and protein [24].
The reactive intermediates formed included elecIncubations contained 100 PM PPHP, 200 PM cofactor and 60 nM trophilic diimine and quinoneimine metabolites horseradish peroxidase in 0.1 M potassium citrate buffer (pH 5.5). (Scheme 1). These metabolites are also formed in For PHS. incubations were carried out at 37°C in 0.1 M phosphate incubations with PHS (data not shown). However, buffer fpH 7.8) using 154 nM of purified PHS. A higher percentthese metabolites do not appear to be responsible age of PPHP reduced in this assay indicates that the test comfor the mutagenicity of anisidine; in mutagenicity pound is a better reducing cofactor. studies using horseradish peroxidase, under conditions where large quantities of the same tion of p-anisidine differs between the two enmetabohtes are produced, p-anisidine was not zymes. mutagenic. The ultimate mutagenic species has yet to be identified in these reactions. One possiDiscussion bility is that PHS forms additional mutagenic metabolites from anisidine which are not formed Both anisidine isomers are mutagenic towards with horseradish peroxidase. The results presented above demonstrate that, the new tester strain YG1029, which contains with the specific activation systems and incubaelevated levels of N-acetyltransferase. This protion conditions used in this study, the mutagenicvides additional evidence that p-anisidine is a ity of p-anisidine is greater than that of the ortho mutagen and demonstrates the utility of YG1029 isomer. RSVM activated only the para isomer to (which incorporates the hisG46 mutational tar-
get) for detecting the mutagenicity of aromatic amines. Surprisingly, no mutagenic response was observed in tester strain YG1012, which incorporates the hisD3052 mutational target; other hisD3052 strains, such as TA98 and TA1538, are highly responsive to aromatic amines such as benzidine, 2-aminofluorene and 2,4_diaminoanisole [l&19]. Our results, however, confirm the sensitivity of hisG46 strains, such as TAlOO, to panisidine reported previously [6,7]. In addition, a similar strain specificity has been reported for a number of other substituted anilines which produce a positive mutagenic response in TAlOO but not TA98 [20,21]. Presumably, oxidation of anisidine is an essential step in the activation of these compounds. The oxidation either hepatic
of p-anisidine can be catalyzed by microsomal enzymes or by PHS. It
a mutagenic product. In contrast, p-anisidine is the inactive isomer in carcinogen bioassays 13-51. This suggests that the mechanisms which result in tumorigenesis in vivo may differ from the mechanisms which result in mutagenesis in vitro in our system. For example, pharmacological differences between the isomers may play a significant role in
vivo. In summary, we have shown that both PaniSidine and o-anisidine are mutagenic in the Ames test using tester strains with elevated levels of NAT/OAT, and using either RSVM or hamster ~9 microsomes as an activating fraction. Although our results further demonstrate the importance of IV-acetyltransferase activity in aromatic amine mutagenicity, they do not solve the intriguing question of the difference between oanisidine and p-anisidine carcinogenicity. More
ss
daimlne
quinoneimine
OCH, diimine
Scheme 1. Peroxidase-dependent
research on the metabolism and disposition of these two isomers m vivo is needed in order to address this question.
Acknowledgement
Research in Dr. Josephy’s laboratory is supported by a grant from the National Cancer Institute of Canada. References 1 Zenser. T.V., and B.B. Davis (19881 Arachidonic acid metabolism by human urothelial cells: Implication in aromatic amine-induced bladder cancer, Prostaglandins Leukot. Essent. Fatty Acids Rev., 31, 199-207. 2 Flammang,T.J., Y. Yamazoe, R.W. Benson, D.W. Roberts, D.W. Potter, D.Z.J. Chu, N.P. Lang and F.F. Kadlubar t 19891 Arachidonic acid-dependent peroxidative activation of carcinogenic arylamines by extrahepatic human tissue microsomes. Cancer Res., 49, 1977-1982. 3 National Cancer Institute, Technical Report No. 89 (1978) Bioassay of o-anisidine hydrochloride for possible carcinogenicity. 4 DHEW Publication No. 7X-1371 (19781 Bioassay of panisidine hydrochloride for possible carcinogenicity. 5 IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Some Aromatic Amines, Anthraquinones and Nitroso Compounds and Inorganic Fluorides Used in Drinking Water and Dental Preparations, Vol. 27, pp. 63-80, 1982. 6 Dunkel, V.C.. E. Zeiger, D. Brusick, E. McCoy, D. McGregor, K. Mortelmans, H.S. Rosenkranz and V.F. Simmon. (19851 Reproducibility of microbial assays, II. Test-
OCH, quinoneimine
pathways of anisidine oxidation.
ing of carcinogens and noncarcinogens in Salmonella typhimurium and Escherichia co/i, Environ. Mutagen., 7 (SuppI. 51, l-248. 7 Haworth. S., T. Lawlor, K. Mortelmans. W. Speck and E. Zeiger, (19831 Salmonella mutagenicity results for 250 chemicals, Environ. Mutagen., 5 (Suppl. 11, 3-142. 8 M.J. Prival, and V.C. Dunkel, (19891 Reevaluation of the mutagenicity and carcinogenicity of chemicals previously identified as “false positives” in the Sulmonclia @phimurium mutagenicity assay, Environ. Mol. Mutagen., 13, l-24. 9 Watanabe, M.. T. Nohmi and M. Ishidate Jr. (19871 New tester strains of Sulmunella typhimurium highly sensitive to mutagenic nitroarenes, Biochem. Biophys. Res. Commun., 147, 974-979. 10 Einisto, P., M. Watanabe, M. Ishidate Jr., and T. Nohmi (19911 Mutagenicity of 30 chemicals in Salmonella fyphimurium strains possessing different nitroreductase or O-acetyltransferase activities, Mutation Res., 259, 95-102. 11 Josephy, P.D., A.L.H. Chiu and T.E. Eling (1989) Prostaglandin H synthase-dependent mutagenic activation of benzidine in a Salmonella typhimurium Ames tester strain possessing elevated N-acetyltransferase levels, Cancer Res., 49, 853-856. 12 Josephy, P.D. (19891 New developments in the Ames assay, High-sensitivity detection of mutagenic arylamines, Bioessays, 11. 108-l 12. 13 Petty, T.W., T.E. Eling, A.L. Chiu and P.D. Josephy (19881 Ram seminal vesicle microsome-catalyzed activation of benzidine and related compounds: Dissociation of mutagenesis from peroxidase-catalyzed formation of DNA-reactive material, Carcinogenesis 9, 51-57. 14 Josephy, P.D., M.H. Carter and M.T. Goldberg, (19851 Inhibition of benzidine mutagenesis by nucleophiles, A study using the Ames test with hamster hepatic S9 activation, Mutation Res., 143, S-10.
x9
15 Peterson, C.L. (1977) A simplification of the protein assay method of Lowry et al. which is more generally applicable, Anal. Biochem., 83, 346-356. 16 Weller, P.E., C.M. Markey, and L.J. Marnett, (1985) Enzymatic reduction of S-phenyl-4-pentenyl hydroperoxide: detection of peroxidases and identification of peroxidase reducing substrates, Arch. Biochem. Biophys.. 243, 633643. 17 Markey, C.M., A. Alward, P.E. Weller and L.J. Marnett (1987) Quantitative studies of hydroperoxide reduction by prostaglandin H synthase, Reducing substrate specificity and the relationship of peroxidase to cyclooxygenase activities, J. Biol. Chem., 262, 6266-6279. 18 Josephy, P.D. (1986) Oxidative activation and mutagenicity of benzidine, Fed. Proc. (FASEB) 45. 2465-2470. 19 de Giovanni-Donnelly, R. (1981) The comparative response of Salmonella typhimurium strains TA1538, TA98 and TAlOO to various hair-dye components, Mutation Res.. 91, 21-25.
20 Zimmer, D., J. Mazurek, G. Petzold and B.K. Bhuyan (1980) Bacterial mutagenicity and mammalian cell DNA damage by several substituted anilines, Mutation Res., 77, 317-326. 21 Kugler-Steigmeier. M.E., U. Friederich, U. Graf, WK. Lutz, P. Paier and C. Schlatter (1989) Genotoxicity of aniline derivatives in various short-term tests, Mutation Res., 211, 279-289. 22 Smith, J.N.. and R.T. Williams (1949) The metabolism of p-phenetidine (p-ethoxyaniline) with some observations on the anisidines (methoxyanilines), Biochem. J., 44,25&255. 23 Schmidt, H.L., M.R. Moeller and N. Weber (1973) Influence of substrates on the microsomal dealkylation of aromatic N-, O- and S-alkyl compounds, Biochem. Pharmacol., 22, 2989-2996. 24 Thompson, D.C., and T.E. Eling (1991) Reactive intermediates formed during the peroxidative oxidation of anisidine isomers, Chem. Res. Toxicol.. 4. 474-481.