Chemically induced mutations observed as mosaics in Drosophila melanogaster

Abstract

This study shows that 2 commonly held explanations for mosaics induced by treating postmeiotic spermatozoa with chemical mutagens are incosistent with an analysis of the percent of mutant tissue produced in the F 1. The first model tested assumes that mosaics are the results of altering one strand of the DNA double helix so that there will be an equal number of mutant and nonmutant nuclei following semiconservative replication. The second model assumes that alkylation of a base in DNA will cause a fixed probability of base pairing mistakes and may produced a mutation at any replication. In phage the probability of mispairing following alkylation is low so that a mutation likely occurs only once in 8 replications; therefore, most of the mutations are fixed after the first 2 divisions. If Drosophila is similar to phage, the mean frequency of mutant nuclei and the frequency distribution of mosaics by class of percent mutant tissue will be distinctly different for the 2 models. Males were fed ethyl methanesulfonate and mutations were detected in the F 1 by the specific locus method at the yellow ( y) and white w loci. Agreement between this experiment and previous work with mosaics produced by chromosome aberrations was found in that the germ line developed from a small sample of poorly mixed nuclei that was independent of the phenotype of head and thorax. The latter finding allowed a selection of mosaics to be made independent of the germ line, and a method of analysis was developed for flies selected because they were mosaic in the head or thorax. Analyses of results from the eye mosaics in this experiment and from 3 experiments with the dumpy (dp) locus previously reported by others but not analyzed in this was were inconsistent with both original models. The average percent of mutant phenotype among the yellow mosaic observed in this experiment was also found inconsistent with both original models, and a frequency distribution of mosaics classified by percent of mutant tissue was inconsistent with both original models and a combination of the two; therefore, both of our initial models and a combination of the two were rejected. The data are consistent with delayed mutation in which an alkylated base has about a 50% probability of mispairing at each replication and producing a fixed mutation or with a multistranded model in which the postmeiotic eukaryote chromosome contains 2 parallel DNA molecules for each locus.