The essentiality of at least one N-methyl group for the carcinogenic activity of the azo dyes

for the carcinogenic activity of the azo dyes re lated to 4-dimethylaminoazobenzene (DAB) has been well established (13, 15, 80, 31). Further more, the N-dimethyl and N-monomethyl dyes studied have essentially the same carcinogenic ac tivities, while the primary aminoazo dyes are much less active (5, 16) and apparently owe their carcinogenicity to a small amount of N-methyla tion in t@ivo(16). On the basis of studies with DAB and 4-monomethylaminoazobenzene (MAB) (17), the equal activities of the monoand dimethyl dyes are probably explained by a readily reversible demethylation of the N-dimethyl dyes. Demethyl ation of the N-monomethyl dyes also occurs read ily, but in the cases examined it has been only slightly reversible (16, 17). The oxidative nature of the N-demethylation reaction has been shown by the isolation of equimolar amounts of for maldehyde and 8-methyl-4-aminoazobenzene fol lowing the incubation of 3-methyl-MAB with for tified rat liver homogenates (22). In accord with this observation is the rapid expiration of C'402 following the administration of 4-dimethyl-C'4aininoazobenzene or its S'-methyl derivative to rats (2, 11). Studies on the 3'-methyl derivative further showed that a small but significant amount of the C'4 was incorporated into the fl-position of serine and the methyl groups of choline and thus also pointed to formaldehyde as an intermediate in the oxidation in vivo (11). The introduction of ring substituents into DAB and MAB can either increase or decrease the car cinogenic activity of the dye (18, 15, 18, 20, 30). Since it seems likely that oxidized derivatives of the N-.methyl dyes (such as the N-methylol com