The relation of germ line mosaicism to somatic mosaicism in Drosophila.

HE frequently used sex-linked recessive lethal test based on observing the Tprogeny of the F, I, yeneration provides an objective and sensitive test for F, females having germ lines complete for a lethal mutation on their paternal X chromosome, and applied to the F, generation provides an objective test for F, females with germ lines mosaic for a lethal on their paternal X chromosome. If a mutagen is used that produces cell lineages with only part of the cells mutant, an increase of sex-linked recessive lethals observed in the F, is normally expected (see CARLSON and SorrTHIN 1963; CARLSON 1964; and ALDERSON 1965 for examples and theoretical reviews) ; however, the germ line analyzed by the sexlinked recessive lethal test represents only a small sample of the cleavage nuclei, for only 2 to 11 cleavage nuclei-usually 3 to 7-penetrate the pole region of the embryo to become poteintial germ cell primordia while the ninth cleavage division is in progress, and some of these cells are lost during movements before being established as germ cells in the gonad (SONNENBLICK 1950). If the few primordial germ cells are drawn together from the same region of poorly mixed cleavage nuclei, they would often be either all mutant or all nonmutant nuclei even in a mosaic embr.yo, and mosaic germ lines that could be detected in the F, would represent only EL small proportion of the mosaic embryos. If, on the other hand, the few nuclei that are differentiated as germ cell primordia represent a random sample which would result from thorough mixing of the cleavage nuclei, then most mosaic embryos would also produce mosaic germ lines. Therefore, an understanding of the type of sample from the cleavage nuclei that the germ line represents becomes essential in interpreting the mutagenic action of mutagens that produce cell lineages with both mutant and nonmutant cells. A direct test of mosaicism among cleavage nuclei is not practical, as we do not have suitable markers; consequently, it is necessary to use a somatic marker that we assume is a good estimate of the frequency of mutant cleavage nuclei. Yellow body color ( y ) was selected because its phenotype can be detected in all regions of the adult surface, and embryological studies have shown that the surface of the insect is derived from large and well distributed samples of the original cleavage nuclei (SONNENBLICK 1950). In support of this assumption, data will be presented to show that flies phenotypically completely mutant for yellow rarely contain in their germ line cells with the normal allele of yellow. In order to produce a population of yellow mosaics, we resorted to inducing chromosome