The developmental basis for germline mosaicism in mouse and Drosophila melanogaster.

Data involving germline mosaics in Drosophila melanogaster and mouse are reconciled with developmental observations. Mutations that become fixed in the early embryo before separation of soma from the germline may, by the sampling process of development, continue as part of germline and/or differentiate into any somatic tissue. The cuticle of adult D. melanogaster, because of segmental development, can be used to estimate the proportion of mutant nuclei in the early embryo, but most somatic tissues and the germlines of both species continue from samples too small to be representative of the early embryo. Because of the small sample of cells/nuclei that remain in the germline after separation of soma in both species, mosaic germlines have percentages of mutant cells that vary widely, with a mean of 50% and an unusual platykurtic, flat-topped distribution. While the sampling process leads to similar statistical results for both species, their patterns of development are very different. In D. melanogaster the first differentiation is the separation of soma from germline with the germline continuing from a sample of only two to four nuclei, whereas the adult cuticle is a representative sample of cleavage nuclei. The presence of mosaicism in D. melanogaster germline is independent of mosaicism in the eye, head, and thorax. This independence was used to determine that mutations can occur at any of the early embryonic cell divisions and still average 50% mutant germ cells when the germline is mosaic; however, the later the mutation occurs, the higher the proportion of completely nonmutant germlines. In contrast to D. melanogaster, the first differentiation in the mouse does not separate soma from germline but produces the inner cell mass that is representative of the cleavage nuclei. Following formation of the primitive streak, the primordial germ cells develop at the base of the allantois and among a clonally related sample of cells, providing the same statistical distribution in the mouse germlines as in D. melanogaster. The proportion of mutations that are fixed during early embryonic development is greatly underestimated. For example, a DNA lesion in a postmeiotic gamete that becomes fixed as a dominant mutation during early embryonic development of the F1 may produce an individual completely mutant in the germ line and relevant somatic tissue or, alternatively, the F1 germline may be completely mutant but with no relevant somatic tissues for detecting the mutation until the F2. In both cases the mutation would be classified as complete in the F1 and F2, respectively, and not recognized as embryonic in origin. Because germ cells differentiate later in mammalian development, there are more opportunities for correlation between germline and soma in the mammal than Drosophila. However, because the germ cells and any somatic tissue, like blood, are derived from small samples, there may be many individuals that test negative in blood but have germlines that are either mosaic or entirely mutant.