1. Methods were developed for the quantitative estimation of 5 to 30 µgm. of aniline, o-aminophenol, p-aminophenol, p-phenylene diamine, monomethyl-p-phenylene diamine, and dimethyl-p-phenylene diamine either alone, in mixtures, or as the reduction products obtained from aminoazo dyes. In the absence of aniline either o-, m-, or p-toluidine could also be estimated. These determinations depended on the formation of Schiff bases between the amines and sodium β-naphthoquinone-4-sulfonate and the separate estimation of the colored derivatives by their different solubility properties and absorption spectra.

  2. On reduction the polar bound dye isolated from the livers of rats fed 4-dimethylaminoazobenzene yielded an amine with the characteristics of aniline and an unidentified polar amine. Similarly, reduction of the 2′-, 3′-, and 4′-methyl derivatives of 4-dimethylaminoazobenzene yielded amines similar to o-, m-, and p-toluidine, respectively, and 2-methyl-4-dimethylaminoazobenzene and 3-methyl-4-monomethylaminoazobenzene appeared to give rise to aniline. These results suggest that the dyes are bound to the protein either through the -N(CH3)2 group or to the ring bearing this group.

  3. The levels of bound dye in the livers of rats fed 4-dimethylaminoazobenzene, its 2-, 2′-, 3′-, and 4′-methyl derivatives, and 3-methyl-4-monomethylaminoazobenzene were determined after dye-feeding periods of 1 to 21 weeks. When the derivatives with the methyl group on the prime ring were considered, there was a striking inverse correlation between carcinogenic activity and the time required to reach a maximum level of bound dye. With relative carcinogenicities of 10 to 12, 6, 2 to 3, and <1 the maxima were attained in approximately 2, 4, 8, and ≥21 weeks for 3′-methyl-4-dimethylaminoazobenzene, 4-dimethylaminoazobenzene, 2′-methyl-4-dimethylaminoazobenzene, and 4′-methyl-4-dimethylaminoazobenzene, respectively. When these data were calculated to micromolar levels by estimating the distillable amine, the amount of bound dye found until the maximum was reached increased with the carcinogenicity of the dye when the methyl group was located on the prime ring. The bound dye levels of livers from rats fed 2-methyl-4-dimethylaminoazobenzene (activity = 0) and 3-methyl-4-monomethylaminoazobenzene (activity < 1) reached maxima after about 12 weeks, but these dyes formed much higher micromolar levels of bound dye than did 4-dimethylaminoazobenzene.

  4. The bound dye-time curves for rats fed 0.060, 0.045, and 0.030 per cent of 4-dimethylaminoazobenzene reached maxima after 2, 4, and 8 and 3, 3, and 8 weeks, respectively, in two series. Since 0.030 per cent of this dye induces tumors much more slowly than the higher levels, these data conform to the inverse relationship between carcinogenic activity and the time required to reach a maximum level of bound dye as found for the C-monomethyl derivatives.

  5. It is suggested that the ascending portion of the bound dye-time curve represents a period in which the liver cells, on the average, can synthesize the proteins which are bound to the dye faster than they are removed by dye-binding and in amounts sufficient for the normal functioning of the tissue. However, once the level of bound dye begins to fall with continued dye-feeding, the dye-binding may have affected one or more specific synthetic mechanisms to such an extent that they cannot keep pace with the combined demands of normal function and removal of protein by the carcinogen. Viable cells may finally result which have completely lost those systems controlling normal growth and hence represent the initial tumor cells. Such a concept is in agreement with the observed absence of protein-bound dye in liver tumors formed during the continuous feeding of the azo dye.


This investigation was aided by grants from the National Cancer Institute, the Jane Coffin Childs Fund for Medical Research, and the American Cancer Society on recommendation of the Committee on Growth of the National Research Council.

This content is only available via PDF.