An evaluation is presented of the effects of 181 chemicals in the (white/white+) (w/w+) eye mosaic assay, an in vivo short-term test measuring genetic damage in somatic cells of Drosophila after treatment of larvae. The genetic principle of this system is loss of heterozygosity for the wild-type reporter gene w+, an event predominantly resulting from homologous interchromosomal mitotic recombination between the two X chromosomes of female genotypes. The w/w+ eye mosaic test detects a broad spectrum of DNA modifications, since all distinct classes of genotoxins are monitored. Non-DNA-reactive chemicals are in principle not detected by this system. Occasional positive responses obtained for chemicals such as amitrole, ethionine and hexachloeroethane are probably not related to the mechanism responsible for their tumorigenicity. The principle outcome of this analysis is the necessity for classification of responses into three categories. (i) Positive, '+ +'. The 92 chemicals (Tables II and III) falling into this category were clearly recombinagenic in the assay, meaning that dose-response relations were obtained (or could have been established as was evident from the strong responses obtained at one or two exposure doses). Among the 92 chemicals were 49 promutagens including volatile chemicals such as vinyl bromide and vinyl chloride. (ii) Marginally positive, '+w'. The definition of a weakly positive response is the absence of a dose - response relationship due to the fact that a weak but reproducible effect, in most cases no more than a doubling of the spontaneous clone frequency, is inherently related to toxicity. The 40 chemicals (Tables IV and V) belonging to this category mainly represented four distinct types. (a) Procarcinogens, such as 2-acetylaminofluorene, dibenz[a,h]anthracene, p-dimethylaminoazobenzene, 2-naphthylamine and safrole, for which metabolic conversion was the apparent problem in the assay. (b) Electrophilic chemicals of high nucleophilic selectivity: acrolein, acrylamide, acrylonitrile, epichlorohydrin, chloroethylisocyanate, 1,2-epoxybutane, N-methyl-n-vinylacetamide, methyl vinylketone, 2-methyl-2-vinyloxirane and methyl vinylsulfone. These chemicals have a relatively low DNA reactivity. Chloroethylisocyanate was active only in the absence of excision repair, suggesting that efficient DNA repair is the cause for the weak genotoxic effectiveness of these genotoxins. (c) Spindle poisons (Table V) were active at rather low but toxic exposure levels. Irregularities in the structure of ommatidia were seen at dose levels producing no more than 2- to 3-fold increases in clone frequencies. (d) The fourth group consisted of chemicals generally regarded as non-genotoxic carcinogens: amitrole, ethionine, ethylurea, tetrachloroethylene and thiourea. Their weak responses were always accompanied by signs of toxicity. (iii) Inactive, '-'. Among the 49 chemicals (Table VI) intactive in the assay were the carcinogens acetamide, auramine-0, chloramphenicol, chloroform, di(2-ethylhexyl)-phthalate, dimethylformamide, ethylene thiourea and testosterone-chemicals for which DNA reactivity is unlikely, For an analysis of the effectiveness of chemicals active in the assay, the number of mosaic clones induced per millimole exposure dose (or per 100 p.p.m. in the case of volatile chemicals) has been compared with the lowest effective dose (LED), i.e. the dose producing an approximately 2- to 3-fold increase in clone frequency over that in the control. The genotoxic effectiveness of chemicals showing strong positive responses (+ +), expressed as the number of mosaic clones induced per millimole chemical, varied over a 10(6)-fold range in dose. Similarly, the LEDs needed to produce a measurable effect covered a > 10(5)-fold range in dose (5 x 10(-4) to 10(2) mM). Among the most effective recombinagens were three of the four aflatoxins tested, of which aflatoxin B1 represents the key chemical in this bioassay. Another noticeable feature of the w/w+ test became evident from a comparison of the LEDs with the highest doses tested (HDTs) for the 49 chemicals inactive in the assay. With a few exceptions, the HDTs represented the maximum doses applicable because either toxicity of the test chemical or its low solubility restricted the application of higher doses. The HDTs for the test chemicals showing no response are located at the upper dose range but clearly not outside the total dose scale spanning from 5 x 10(-4) to 10(2) mM. Since cytotoxic effects to the developing organism apparently put a limit to the selection of high dose levels, the consequence is that the use of non-physiologically high concentrations is largely avoided in this in vivo system.
