The laatest in lineaage A novel technique that may simplify mammalian lineage analysis relies on intragenic recombination to generate a functional lacZgene from a mutated aacZ gene in single cells of transgenic mouse embryos.

Much of modern developmental biology involves asearch for mechanisms that determine how embryoniccells move to appropriate positions and acquire distinctphenotypes. Clearly, such inquiries need to rest on a solidfoundation of descriptive information: it is of little use todiscover factors that can influence migration or differen-tiation unless one knows where the cells in questionactually move or what they will normally become.Recent success in elucidating the contributions oflineage and environment to pattern formation in worms,flies and leeches, for example, has depended not only onthe accessibility of these organisms' embryos, but also onthe ability to trace the genealogy of their cells [1].The larger size and greater complexity of vertebrateembryos have made them harder to tackle, but progresshas been steady. For example, late stages of diversificationwithin individual tissues of chicks and rodents have beentraced by infecting cells with recombinant retroviralvectors. In this method, the retroviral genome ismodified to contain a gene encoding a reporter enzymesuch as -galactosidase, encoded by the Escherichia coligene lacZ. The lacZ-containing retroviral genome inte-grates into a chromosome of the infected cell and ispassed on to that cell's progeny, where the gene productis readily detectable by a simple histochemical stain thatturns blue any cell expressing the gene [2,3]. An alterna-tive method is to use a fine glass microelectrode to injecta tracer into an identified progenitor cell. Both fluores-cent and enzymatically identifiable tracers have beeninjected; they dilute with growth, but offer the advantagethat a particular progenitor cell can be targeted foranalysis. Microinjection provides an alternative to theretroviral technique for studying relatively late lineages[4,5], and has also provided a means of tracing the contri-butions made by individual blastomeres to fish and frogembryos [6,7].A major gap in studies of cell lineage has been due tothe limited ability to mark and trace cells in the earlymammalian embryo. Elegant studies using arduous tech-niques such as whole-embryo culture and blastocystinjection have provided some information on cellmovements and fates during germ layer formation [8],but it is fair to say that our knowledge of early lineagedecisions in mammals is rudimentary compared withwhat we now know about lower vertebrates. Theobstacles are formidable. First, infection and injection,although straightforward in ovo, are extraordinarilydifficult to achieve in utero. Second, the embryo is