The capacity of some nitro- and amino-heterocyclic sulfur compounds to induce base-pair substitutions

The capacity of 27 heterocyclic sulfur compounds to induce base-pair substitutions was investigated with Klebsiella pneumoniae ur- pro- and Salmonella typhimurium TA100 as test organisms. Among the compounds tested, all sulfur compounds with nitro groups and some thiazoles with an amino group were mutagenic. Among the nitrothiazoles, the most potent mutagen was niridazole, followed by 2-acetamido-5-nitrothiazole, 2-bromo-5-nitrothiazole, N-(5-nitrothiazol-2-yl)benzamide, and 2-amino-5-nitrothiazole. Of the nitrothiophenes, 2-nitrothiophene was more mutagenic than 3-nitrothiophene and 2,4-dinitrothiophene. 4-Nitroisothiazole was also mutagenic. Of the aminothiazoles, 2-amino-5-bromothiazole and 2-amino-5-chlorothiazole were mutagenic to both test organisms. With 2-amino-5-(p-nitrophenylsulfonyl)thiazole, a mutagenic action was only found with Salmonella typhimurium TA100, whereas 2-aminothiazole and 2-amino-4-methylthiazole were only mutagenic with Klebsiella pneumoniae. With the other 13 compounds, no mutagenic activity was observed. Of the coccidiostatics, 2-acetamido-5-nitrothiazole was also mutagenic on Escherichia coli K12 and Saccharomyces cerevisiae D4 but non-mutagenic on Salmonella typhimurium TA1530, TA1535, TA1537 and TA98, while 2-amino-5-nitrothiazole was mutagenic on Escherichia coli K12, Salmonella typhimurium TA1530, TA1535 and TA98, and non-mutagenic on strain TA1537 and on Saccharomyces cerevisiae D4.     

Mutagenicity studies of p-substituted benzyl derivatives in the Ames Salmonella plate-incorporation assay

The mutagenicity of 24 benzyl derivatives, containing a variety of substituents and leaving groups, were assayed in strain TA100 using the Ames plate-incorporation assay. p-Nitrobenzyl chloride (12 000 revertants/mumole), p-nitrobenzyl tosylate (6100 revertants/mumole), and p-acetoxybenzyl chloride (100 revertants/mumole) were mutagenic; none of the remaining 21 compounds were mutagenic. p-Nitrobenzyl chloride was also found to be mutagenic in strain TA98 (700 revertants/mumole), but not in strain TA98NR (a strain deficient in nitro reductase activity). p-Acetoxybenzyl chloride was nonenzymatically hydrolyzed to p-hydroxybenzyl alcohol and p-acetoxybenzyl alcohol. These findings suggest that nitrobenzyl derivatives were mutagenic due to nitro reductive metabolism and that p-acetoxybenzyl chloride was mutagenic due to the intermediate formation of p-hydroxybenzyl chloride during the hydrolysis of p-acetoxybenzyl chloride.     

Structural specificity of aromatic compounds with special reference to mutagenic activity in Salmonella typhimurium--a series of chloro- or fluoro-nitrobenzene derivatives

The mutagenicity of 21 chloro- or fluoronitrobenzene compounds and 9 chloro- or fluorobenzene compounds in Salmonella typhimurium (strains TA98, TA1538, TA1537, TA100 and TA1535) was examined. The tests were carried out under the conditions of absence and presence of liver microsomal activation. 15 nitro-group compounds had mutagenic activity; above all, compounds of fluoronitrobenzene were mutagenic for both types of strain. On the other hand, chloronitrobenzene compounds were mutagenic for base-pair substitution strains only. Mutagenic activity was exhibited by all compounds having a chloro or fluoro substituent at the para and ortho position in the nitrobenzene nucleus. All compounds without a nitro substituent showed no mutagenic activity.     

Comparative mutagenicity of aliphatic epoxides in Salmonella

37 aliphatic epoxides comprising 6 subclasses (unsubstituted aliphatic epoxides, halogenated aliphatic epoxides, glycidyl esters, glycidates, glycidyl ethers and diglycidyl ethers) were tested, under code, for mutagenicity in Salmonella strains TA98, TA100, TA1535 and TA1537 and/or TA97 with and without metabolic activation using a standardized protocol. The 4 halogenated aliphatic epoxides and the 4 diglycidyl ethers were all mutagenic. The 2 glycidates were negative in all strain/activation systems used while all 5 glycidyl esters were mutagenic. 3 of the 8 unsubstituted aliphatic epoxides and 11 of the 12 glycidyl ethers were mutagenic. Glycidol also was mutagenic whereas 9,10-epoxyoctadecanoic acid, 2-ethylhexyl ester was not mutagenic. Of the 28 mutagenic compounds, all but neodecanoic acid, 2,3-epoxypropyl ester and 2-ethylhexyl glycidyl ether were detected in TA100 without activation. The latter two were detected only with activation in TA100 and TA1535. The majority of the other 26 chemicals were also mutagenic in TA1535 without activation. Good intra- and interlaboratory reproducibility was seen in the results of each of the 4 chemicals tested in more than one set of experiments. The current results confirm and extend the observations of other investigators regarding structural effects on the mutagenicity of members of the aliphatic epoxide class of chemicals.     

Relationships between structures and mutagenic potencies of 16 heterocyclic nitrogen mustards (ICR compounds) in Salmonella typhimurium

16 heterocyclic nitrogen mustards (ICR compounds), which were synthesized for use as possible antitumor agents by Creech and coworkers, were tested for mutagenicity in Salmonella typhimurium strains TA1535, TA1536, TA1537, TA1538, TA98 and TA100. The compounds were incorporated into the top agar at 5 doses: 0.5, 1, 2.5, 5 and 10 micrograms/plate. All of the compounds were negative in TA1535 except ICR 449, which was positive in all 6 strains. The other 15 compounds were positive in the remaining strains with the following exceptions: ICR 371 and 355 were negative in TA100; ICR 445 was negative in TA98 and TA100; and ICR 360 was negative in TA1537, TA1538, TA98 and TA100. Good qualitative agreement was observed between the mutagenic and antitumor activities of the 16 compounds, and between the mutagenic and carcinogenic activities of the 5 compounds that have been tested for carcinogenicity by Peck and coworkers. However, no significant correlation was found between mutagenic potency in Salmonella and antitumor potency in mice for the 16 compounds. Also, for the 5 compounds that have been tested for carcinogenicity, no significant correlation was found between their mutagenic potency in Salmonella and their carcinogenic potency in mice. In Salmonella, the secondary (2 degrees) amines generally were more mutagenic than their tertiary (3 degrees) amine homologs, although the opposite result has been reported in certain eukaryotes. Relationships between structures and potencies for the different nuclei of the 16 ICR compounds are discussed, as are similarities and differences in strain sensitivities. We conclude that the Salmonella his reversion test is not a good predictor of the antitumor and carcinogenic potencies of these ICR compounds.     

Mutagenic activities of ten imidazole derivatives in <Emphasis Type="Italic">Salmonella typhimurium</Emphasis>

Ten imidazole derivatives were tested for mutagenicity in Salmonella typhimurium strains TA98 and TA100 both in the absence and presence of metabolic activation by the microsomal fraction S9 mix. In a general manner, derivatives tested exhibited a greater mutagenic activity in the TA100 strain comparing to the responses in TA 98. In the standard plate incorporation assay, 8 of these substances (80%) were found to be mutagenic for at least one of the two strains in the presence or absence of metabolic activation. Two compounds showed positive results in TA98 and 6 compounds were also mutagenic in TA100 without S9. In the presence of S9 mix, all of the 10 substances were non-mutagenic in TA98, whereas 4 compounds were positive in TA100. The results suggested the mutagenic potentials of the imidazole derivatives particularly inducing the reversion of base-pair substitutions. According to the structure-activity relationships phenyl groups in position 2 with different substituents can confer the mutagenic activity of the tested compounds. Methyl groups in different positions of these phenyl substituents can cause different types of mutations. This mutagenic effect is observed more clearly when the phenyl group is inhibited with a nitro group.     

Mutagenicity of polycyclic aromatic compounds (PAC) identified in source emissions and ambient air

Several polycyclic aromatic compounds (PAC) including nitrated and oxygenated derivatives of polycyclic aromatic hydrocarbons (PAH) were tested for mutagenic activity in the Salmonella/microsome assay. Among the compounds tested the isomer mix of nitro-1-hydroxypyrenes showed the highest direct mutagenic response in both the Salmonella strain TA98 and TA100 (1251 revertants/micrograms and 463 revertants/micrograms, respectively). The direct-acting mutagenicity of the nitro-1-hydroxypyrene isomer mix was dependent upon reduction of the nitro function as evidenced by the decrease in activity observed with the nitroreductase-deficient and arylhydroxylamine esterifying-deficient tester strains. The oxygenated derivatives of PAH containing aldehyde or keto groups showed weak or no mutagenic responses. In most cases addition of S9 was essential for any mutagenic activity and the strain TA100 was more sensitive than the strain TA98. Within this group, 7H-dibenzo[c,g]fluoren-7-one showed the highest mutagenic effect; 7 and 22 revertants/micrograms using the strains TA98 and TA100, respectively.     

Evaluation of the mutagenic and cytostatic potential of aristolochic acid (3,4-methylenedioxy-8-methoxy-10-nitrophenanthrene-1-carboxylic acid) and several of its derivatives

Aristolochic acid (1), a constituent of Aristolochia species, has been used for medicinal purposes since the Graeco-Roman period. Following the observation that the compound was mutagenic and carcinogenic, it was removed from pharmaceutical products. Consistent with previous reports, we have found that 1 serves as a direct-acting mutagen in Salmonella typhimurium strains TA100, TA102, TA1537 and TM677, but was not active in the nitroreductase-deficient strains TA98NR and TA100NR. However, aristolic acid (2), a compound that differs in structure only by the absence of the nitro group, was also found to be a direct-acting mutagen in Salmonella strains TA98, TA100, TA102, TA1537, and TM677, as well as strains TA98NR and TA100NR. Both compounds (1 and 2) were active mutagens when evaluated with cultured Chinese hamster ovary cells. Thus, in contrast to previous suggestions, the nitro group at position 10 is not required to induce a mutagenic response. Also, a series of structural relatives (the methyl esters of 1 and 2 (3 and 4, respectively), aristolochic acid-D (5), aristolactam (6), aristolactam A-II (7), and aristolactam-N-beta-D-glucoside (8)) were evaluated for mutagenic potential with Salmonella typhimurium strain TM677 and found to be inactive. Since compounds 3 and 4 were found to be active mutagens with Salmonella typhimurium strains TA98, TA100, TA102 and TA1537 (sufficient quantities of compounds 5-8 were not available for testing), differential sensitivity of the tester strains unrelated to mutagenic potential is suggested. Further, compounds 1, 2, and 6-8 were evaluated for potential to inhibit growth with cultured KB or P388 cells. P388 cells were substantially more sensitive, and compound 1 was the most active of the materials tested (ED5 = 0.58 microM). Compound 6 also demonstrated appreciable activity (ED50 = 4.2 microM), as did compound 8 (ED50 = 6.0 microM). It therefore appears that phenanthrene-ring substituents, in addition to the nitro group at position 10, serve important roles for biological potential. In considering the carcinogenic event induced by aristolochic acid, these functionalities should also be taken into account.     

Structure-activity relationships of nitro and methyl-nitro derivatives of indoline, indole, indazole and benzimidazole in Salmonella typhimurium

The mutagenic activities of eleven nitro derivatives and eleven N-methyl-nitro derivatives of indoline, indole, indazole and benzimidazole were investigated in Salmonella TA98 and TA100. The presence of a nitro group at C4 or C7 resulted in only weakly or nonmutagenic compounds, while a nitro group at C2, C5 or C6 usually resulted in measurable mutagenic activity in the non-N-methylated compounds. Methylation of a ring nitrogen usually reduced the mutagenic activity of these nitroheterocyclics except 2-nitro-benzimidazole, which resulted in a better than 300-fold increase in mutagenic activity. A proposed mechanism for the increased mutagenic activity obtained by methylation of imidazole nitrogens may provide insights into the reasons for the potent mutagenicities observed for several similarly methylated cooked-food mutagens.     

Mutagenicity of aminophenyl and nitrophenyl ethers, sulfides, and disulfides.

The mutagenic activity of several aromatic amines and aromatic nitro compounds related to 4,4'-methylenedianiline towards Salmonella typhymurium tester strains TA100 and TA98 was evaluated. The heteroatomic analogs of 4,4'-methylenedianiline which include aminophenyl and nitrophenyl ethers, sulfides and disulfides were assayed in the presence of rat-liver homogenate. The relative mutagenic response of these analogs indicated the following order of activity, --S-- greater than --O-- greater than --CH2--CH2-- greater than or equal to --S--S--. In both tester strains 4-aminophenylsulfone was inactive with and without microsomal activation. The p-nitrophenyl ethers, sulfides and disulfides were relatively strong mutagens without microsomal activation towards TA100. While 4-nitrophenyldisulfide was found to possess significantly different mutagenic activity than 4-nitrothiophenol in TA98, 4-AMINOPHENYl disulfide has similar mutagenic properties to 4-aminothiophenol in both tester straains TA100 and TA98.     

Relationship between mutagenic potency in Salmonella typhimurium and chemical structure of amino- and nitro-substituted biphenyls

All positional isomers of mononitro- and monoaminobiphenyls and those of dinitro-, diamino- and aminonitrobiphenyls, which have one substituent on each benzene ring, were assayed for mutagenicity in Salmonella typhimurium by the Ames method. The results suggest that the structural requirements favoring mutagenic activity are the presence of substituents at the 4-position and their absence at the 2'-position. The introduction of an amino group to the 3'- or 4'-position of 4-nitrobiphenyl or a nitro group to 3'- or 4'-position of 4-aminobiphenyl enhanced the mutagenicity. Among the mutagenic compounds, 4-nitro analogues were mutagenic in strains TA98 and TA100 in the absence of a microsomal metabolic activation system. Strain TA98NR was not reverted by the direct-acting mutagens, whereas strain TA98/1,8-DNP6 was as revertible as strain TA98; these results suggest that the direct-acting mutagenicity involves the reduction of the nitro group by bacterial nitroreductase but does not involve specific esterification enzymes.     

Determining the mutagenic activity of a tar, its vapors and aerosols

The Ames test was performed on Salmonella typhimurium, strain TA98, TA100, TA1535, TA1537, TA1538, to evaluate the mutagenic potential of a tar as well as its vapors and aerosols emitted at 250, 350 and 550 degrees C. Two chemical procedures were used: extractions of aromatics for DMSO; elimination of acids, alcohols and phenols. Weak mutagenic activity was demonstrated at each temperature. Then, using only Salmonella typhimurium strains TA98 and TA100, a study was made on the effects of the mutagenic compounds, benzo[a]pyrene, 2-aminoanthracene, nitrofluorene, methyl methanesulfonate and on the vapors and aerosols emitted at 350 degrees C by road-coating tar. For promutagenic compounds, an enhancing effect was observed before an inhibition effect. For direct mutagenic compounds, only the inhibition effect appeared. The mutagenic and/or carcinogenic activity was usually tested on a pure isolated chemical compound.     

Desmutagenic effect of alpha-dicarbonyl and alpha-hydroxycarbonyl compounds against mutagenic heterocyclic amines

The desmutagenic effects of alpha-hydroxycarbonyl compounds, such as glyceraldehyde, glycolaldehyde, dihydroxyacetone, furfural, 5-hydroxymethylfurfural, maltol, acetol and acetoin and alpha-dicarbonyl compounds, such as diacetyl, glyoxal, methyl glyoxal and 2,3-pentanedione were investigated against the mutagenic heterocyclic amines, such as Trp-P-1, Trp-P-2, Glu-P-1, Glu-P-2 and IQ. Most of the carbonyl compounds suppressed the mutagenicity of heterocyclic amines for S. typhimurium TA98, alpha-dicarbonyl compounds showing a higher desmutagenic effect than alpha-hydroxycarbonyl compounds. Among the alpha-hydroxycarbonyl compounds, glyceraldehyde, glycolaldehyde and dihydroxyacetone showed more effective desmutagenicity, and diacetyl among the alpha-dicarbonyl compounds had the highest desmutagenic effect. These carbonyl compounds alone also showed mutagenicity to S. typhimurium TA100 without S9 mix. The reaction of carbonyl compounds with mutagenic heterocyclic amines also eliminated the mutagenicity of the former for S. typhimurium TA100.     

Mutagenicity of 43 structurally related heterocyclic compounds and its relationship to their carcinogenicity.

43 heteropolycyclic compounds belonging to a homologous series were investigated for mutagenicity. The results are compared with carcinogenicity data obtained with the same batches of compounds under conditions identical for all of them. Mutagenicity was tested in the Ames test with Salmonella typhimurium strains TA1535, TA1537 and TA100 in the presence and absence of liver 10 000 g supernatant from rats treated with Aroclor 1254. Carcinogenicity was tested by injection of the compounds into subcutaneous tissue of XVIInc/Z mice. 18 test compounds showed carcinogenic activity, some strongly, others only weakly. Of these, 17 were detected as mutagens: one weak carcinogen did not revert the Salmonella strains. No quantitative correlation was observed between the extents of the mutagenic and the carcinogenic effects. Of the 25 substances that did not produce tumours, 13 showed mutagenicity (12 in the presence, 2 in the absence, of the liver homogenate). The mutagenic effects of these compounds were quantitatively similar to those of the compounds that produced tumours. The most sensitive strain of Salmonella typhimurium was TA100. It detected all 30 mutagens. TA98 was mutated by 25 compounds, TA1537 by 16 compounds. No mutagenic effects were seen with TA1535. Possible reasons for the high percentage of apparently "false positives" in the Ames test and the lack of a quantitative correlation between the potency of the mutagenic and carcinogenic effects are discussed. It is suggested that the complexity of the metabolism of these heterocyclic compounds may lead to critical differences in metabolism in mouse subcutaneous tissue in vivo and in liver homogenates from rats treated with Aroclor. Therefore the present study will be extended to life-long oral and intrahepatic carcinogenicity tests leading to a higher proportion of metabolism in the liver.     

Effect of various alkyl and unsaturated substituents on the mutagenicity of some nitrophenyl thioethers.

A variety of nitro-substituted phenyl alkyl/aryl thioethers and nitroso-substituted phenyl alkyl/aryl thioethers have been synthesized and tested for their mutagenicity towards Salmonella typhimurium strain TA100, TA98, TA98NR and TA98/1,8-DNP(6) in the absence of S9 mix. The relative order of mutagenicity in TA98 and TA100 among p-nitrophenyl thioethers having alkyl or aryl substituents is allyl>phenyl>benzyl>butyl>propyl>ethyl>methyl. Compounds having an alkyl chain C(6) to C(12) were found to be non-mutagenic. Among the various positional isomers (ortho, meta and para) of nitro-substituted diphenyl thioethers only the compounds having the -NO(2) function at the para position is mutagenic, whereas compounds having a -NO(2) function at ortho and meta are non-mutagenic. However, the reduced intermediate, ortho-nitroso derivative was found to be mutagenic in all the four strains but the meta-nitroso derivative was found to be non-mutagenic. All mutagens were found to be non-mutagenic when tested in nitroreductase deficient strain TA98NR, whereas their nitroso intermediates are found to be mutagenic. A substantial fall in the mutagenic activity is observed when some mutagens are tested in O-acetyltransferase deficient strain TA98/1,8-DNP(6).     

Dependence of mutagenic activity of heterocyclic analogs of pyrene on their chemical structure]

Comparative mutagenic activity of 6 heterocyclic analogs of pyrene was studied. The highest activity was revealed for 2,7-dinitro-4,9-dioxy-5,10-dioxo-4,5,9,10-tetrahydro-4,9- diazapyrene (2,7-DN-DDTDP) and 2,7-dinitro-5,10-dioxo-4,9-dioxapyrene (2,7-DN-DDP) which induced mutations in the tester strain Salmonella typhimurium TA98 and TA100. High mutagenicity of 2,7-DN-DDTDP and 2,7-DN-DDP is conditioned by reduction of nitro groups in the 2,7 position. The carbonyl groups in the 5,10 position were believed to cause bifunctional activity of 2,7-DN-DDTDP and 2,7-DN-DDP and to promote deletion of two G-C in the D3052 site.     

Mutagenicity of (p-nitrophenyl)adenines in Salmonella typhimurium

Adenine derivatives having a p-nitrophenyl group at position 2, 8, or 9 were directly mutagenic towards Salmonella typhimurium strains TA98 and TA100, whereas N6-(p-nitrophenyl)adenine was not mutagenic. 2,9- And 8,9-bis-(p-nitrophenyl)adenines were also mutagenic, but N6,9-bis-(p-nitrophenyl)adenine was not. The study on 13 (p-nitrophenyl)adenine derivatives for their Salmonella mutagenicity indicates that only those having a p-nitrophenyl ring directly linked to the purine ring are mutagenic, implying the importance of the coplanar character of the nitrophenyl and the purine rings. The nitro group seems essential for the mutagenicity, as shown from the results of assays using nitroarene-sensitive and -insensitive Salmonella strains. The mutagenic potency of this class of compounds is high, comparable to that of 2-nitrofluorene.     

The effects of 4'-alkyl substituents on the mutagenic activity of 4-amino- and 4-nitrostilbenes in Salmonella typhimurium

Six derivatives of trans-4-aminostilbene bearing different alkyl groups in the 4'-position and six of the corresponding nitro compounds were synthesized and tested for their mutagenic potency in Salmonella typhimurium strains TA98 and TA100. Regarding the test series in presence of S9-mix, maximum activity was observed for those trans-4-aminostilbenes and trans-4-nitrostilbenes bearing small alkyl substituents like methyl and ethyl. More bulky substituents reduced the mutagenic potential in the order iso-propylethyl>iso-propyl>sec-butyl>tert-butyl). These trends have been compared with quantitative structure activity relationship (QSAR) model predictions, leading to the conclusion that steric demand is an important factor for mutagenicity of substituted aminostilbenes and nitrostilbenes. The unexpected result for the tert-butyl nitrostilbene tested with metabolic activation may be attributed to a different metabolic pathway.     

Mutagenicity of nitro derivatives induced by exposure of aromatic compounds to nitrogen dioxide

Mutagenic nitro derivatives were readily induced when 6 kinds of chemicals were exposed to 10 ppm of nitrogen dioxide (NO2). Single nitro derivatives were formed from pyrene, phenanthrene, fluorene or chrysene. Carbazole and fluoranthene each produced 2 derivatives substituted with nitro groups at different positions. The formation of nitro derivatives was enhanced by exposure of pyrene to NO2 containing nitric acid (HNO3, less than 100-fold enhancement) or sulphur dioxide (SO2, less than 15-fold enhancement). After 24 h of exposure the yields of the nitro derivative were 0.02% with 1 ppm of NO2 in air and 2.85% with NO2 (1 ppm) containing traces of HNO3. The nitro derivatives from all but phenanthrene and carbazole were chemically identified by means of gas chromatography (GC) and mass spectrometry (MS), and the mutagenicity of the 4 kinds of authentic nitro derivatives was tested by using Salmonella strains TA98 and TA1538 with or without the S9 fraction from rat liver treated with Aroclor 1254. The nitro derivative induced from pyrene was determined to be 1-nitropyrene; that of chrysene was 6-nitrochrysene; that of fluorene was 2-nitrofluorene; and those of fluoranthene were 3-nitrofluoranthene, and 8-nitrofluoranthene. Tested with strain TA98 in the absence of the S9 fraction, the first 4 of these derivatives yielded, respectively, 3050, 269, 433 and 13 400 revertants per nmole. Thus, each nitro derivative formed was potentially a direct-acting frameshift-type mutagen. Each compound exposed to NO2 showed a decreased mutagenic activity when tested in the presence of S9 mix. A possible explanation comes from experiments in which 1-nitropyrene was incubated with the S9 mix at 37 degree C for 10 min, and 1-aminopyrene was formed. The mutagenic activity of 1-aminopyrene was appreciable, but only about one-tenth of that of 1-nitropyrene in the Ames test.     

Mutagenicity of anthraquinones in the Salmonella preincubation test

The mutagenicities of 15 naturally occurring anthraquinones were examined in Salmonella typhimurium strains TA98, TA100 and TA2637 by the preincubation method. 7 of the 15 compounds tested, i.e., chrysazin, emodin, islandicin, alizarin, chrysophanol, 2-hydroxyanthraquinone and emodic acid, were strong mutagens in strain TA2637 with metabolic activation. All of these compounds contain 1-3 hydroxyl groups, and some also have methyl groups. Cynodontin, an anthraquinone with 4 hydroxyl groups and 1 methyl group, was only slightly mutagenic in strain TA2637. 2-Hydroxyanthraquinone, alizarin, emodin, islandicin and chrysazin were also mutagenic in strain TA100 with S9 mix. All the bisanthraquinones tested, i.e., skyrin, (+)rugulosin, (-)luteoskyrin, (-)rubroskyrin and sennoside A, were non-mutagenic in this test system with or without metabolic activation. Unsubstituted anthraquinone and anthrone were also non-mutagenic. These results show that hydroxyl substituents are necessary for the mutagenicity of anthraquinones, the optimal substitutions being 1-3 hydroxyl groups per molecule. The 4th hydroxyl group, in the compound cynodontin reduces the mutagenicity considerably.