An aqueous mixture of
p-phenylenediamine dihydrochloride,
resorcinol and
hydrogen peroxide, incubated for 30 min at 37 degrees C, was not mutagenic in the Ames test and in the mouse
lymphoma assay and did not produce
chromosome aberrations in human lymphocytes, whereas the same mixture without
resorcinol was mutagenic in the Ames test and the
chromosome aberration assay, probably due to the formation of
Bandrowski's base. The formation of mutagenic
Bandrowski's base and other reactive products could be demonstrated by thin-layer chromatography and subsequent Ames testing. Preincubation for 0-7 h of 5 different oxidative mixtures of
p-phenylenediamine dihydrochloride and various couplers at 37 degrees C resulted in mutagenicities in the Ames test ranging from 'very strong effects' at 0 or 0.5 h for
2,4-diaminoanisole dihydrochloride to 'not mutagenic' at up to 7 h preincubation time for a recently developed alternative coupler. A comparable ranking of mutagenic potencies of the same mixtures could be obtained by Ames tests with dyed buffalo hair strands, either tested directly on the
agar plates or by testing the
solvent extracts of the colored strands. The optimal combination without any sign of mutagenic tendency was finally tested in a commercially available
hair dye formulation. This
complex mixture was also not mutagenic in the Ames test. Our main conclusions are as follows. (I) Oxidative
hair dye mixtures of
p-phenylenediamine dihydrochloride and non-mutagenic couplers are not mutagenic if tested under normal conditions of use; the formation of mutagenic reaction products (such as
Bandrowski's base) can be prevented if the oxidative reaction time is limited to 30 min and if about equimolar proportions of para to meta compounds are used. (II) Optimal non-mutagenic combinations of primary intermediates (such as
p-phenylenediamine dihydrochloride) and couplers can be evaluated in the Ames test by either time-dependent preincubation of the mixtures or the described test protocol with hair strands.