Mammalian chimaeras / Anne McLaren.

  • Anne McLaren
Date:
[1976]
Catalogue details
    122/168 page 112
    112 Distribution of cell populations immediate past, every tissue containing n cells has been derived from n¡2 cells one generation ago. More remotely, every tissue has been derived from a single cell, the zygote, in an ordinary animal or a mosaic, and from two zygotes in a chimaera. Attempts have been made to specify more precisely the stage in development to which the estimate applies; e.g. the stage 'at which the tissues first become differentiated' or 'at which tissue-specific genes are first switched on' or 'at commitment' or 'at inactivation'. The problem has been discussed at length by McLaren (1972Ò) and by Lewis, Summerbell & Wolpert (1972). It is concluded that the approach is only meaningful if applied to a system in which a period of random cell mingling of indefinite duration is succeeded by a period in which the tissue is formed by coherent clonal growth. Subsequent to this moment of alloca¬ tion, cell mingling may take place within the tissue primordium, but no sig¬ nificant movement of cells occurs between the primordium and its neighbours. The binomially derived estimate of initial cell number relates either to the point at which allocation occurs, or to the point at which heterogeneity arises (mid-cleavage for aggregation chimaeras, after implantation for X-inactiva- tion mosaics), whichever happens later in development. There is no a priori reason why allocation should coincide with biochemical or morphological differentiation, or with the activation of tissue-specific genes. One major assumption required for any valid inference to be based on binomial reasoning is that the two cell types should be distributed randomly at the time of sampling. Thus in the derivation of the embryo proper, the three 'progenitor' cells would presumably be located in the inner cell mass of the blastocyst. Yet Garner & McLaren (1974) found that, far from being arranged randomly, the cells of a chimaeric blastocyst formed two coherent clones, one of each component type. The assumption of random distribution of cells has not been tested for any of the later stages of development at which sampling events are likely to occur. The other assumption underlying the binomial approach is that there is no differential cell death or proliferation of one component at the expense of the other. In X-inactivation mosaics, where the two components are genetic¬ ally identical except for their active X chromosome, such an assumption may hold; in chimaeras, it is often demonstrably false. The estimate of three 'progenitor' cells for the embryo proper is based on the premise that 25 % of aggregated embryo-pairs show no evidence of overt chimaerism. In fact the values of this proportion so far reported range from under 20 % to over 60% (Mullen & Whitten, 1971); it will include aggregants from which one component has been lost through technical hazards, such as developmental retardation or defects induced by manipulation or culture, as well as innate differences in mitotic rate or viability. As for differential proliferation, Mintz (1970e) has emphasized how in any particular strain combination, one com¬ ponent or the other is likely to be at a selective advantage in any particular
    No text description is available for this image
    No text description is available for this image
    No text description is available for this image