[Lineage-switching by pluripotent cells derived from adults]

When proceeding normally, embryonic morphogenesis begins with germ layer formation through the process of gastrulation. Each primordial germ layer gives rise to a particular set of lineages. Until recently, it was considered that fate switches between germ layers were impossible. In the last two or three years however, a fair number of such switches have been described (Table I), the most spectacular of which entails the differentiation of neural stem cells into various derivatives. This unexpected plasticity opens important prospects for cell therapy. Stem cells, which are the cells that display this plasticity, are defined by the two properties of self renewal and pluripotency. They are set apart during ontogeny and are responsible for maintaining the homeostasis of a tissue. This notion, first established in the case of hematopoietic stem cells was later extended to other fast renewing cells, such as those in the intestinal epithelium or epidermis, and more recently to cells reputedly non-renewable, i.e. neurons. A new strategy has been described, which has the interesting feature that it can be applied to the isolation of stem cells from various lineages. It consists in sorting out cells on the basis of the efflux of Hoechst 33342 dye (Goodell et al., 1996). When a cell suspension stained with this dye is examined under two distinct wave lengths, a "side population" (SP), characterized by weak fluorescence, can be identified and sorted out. The dye efflux property of these cells is due to the activity of the mdr (multidrug resistance) gene, which encodes a protein responsible for the building of a canal which serves to extrude toxins from the cells. A means of distinguishing a truly multipotent stem cell from a progenitor committed to a specific lineage has been reported. This consists in the expression of the Pax7 gene. Pax7-/- mouse muscles have no satellite cells, i.e. they miss the cells normally responsible for the regeneration of muscle. In contrast they do have an SP population. These SP cells are incapable of differentiating into muscle, but give rise to 10 times more hematopoietic colonies, when cloned in vitro, than SP cells from wild type muscle do. Thus Pax7 appears to be a commitment gene, in the absence of which stem cells cannot become specified to the muscle lineage. As a conclusion, this review emphasizes various features of the recent findings: 1) the unexpected plasticity uncovered in recent years is restricted to the stem cells of each tissue; 2) the switch in phenotype has to be "forced" on these stem cells by drastic experimental conditions enforced in the host: often sublethal irradiation is superimposed on a genetic deficiency. Progress in this field, concerning both conceptual and applied aspects, will require the identification of the factors characterizing the niches which promote integration and fate switches of stem cells, probably a combination of growth factors and intercellular interactions. Finally a key issue, before any therapeutical applications can be considered, is how to control the proliferation of transplanted stem cells in their new environment.