Rebuttal to Enver, Heyworth, and Dexter

THIS PREARRANGED controversy has ended up with three English authors versus one Australian; the odds seem about right, regardless of whether it is dice, cricket, or hematopoiesis.

The arguments presented by Enver et al are well-reasoned and the ambivalent conclusion reached is rather similar to my own. There are two matters warranting additional comment.

First, all of us accept the data from the recloning of multipotential cells rather uncritically as indicating random commitment. However, what was actually observed was asymmetry in the manner in which various committed progeny were generated. With a genuinely random process, the probability of progenitor cells of a particular type appearing in a developing blast colony composed of progenitor cells should increase as progenitor cell numbers increase in the colony. Current data in our laboratory indicate no correlation between the total number of progenitor cells in such colonies and the presence or absence of progenitors committed to a particular lineage. This seems a powerful argument against random generation.

Second, the question under discussion has a practical aspect that needs to be kept in mind. The question is not so much the relative importance of the roles played by extrinsic regulators or internal gene programming in commitment, but whether administered agents—either cytokines or modulators of nuclear transcription factors—can predictably influence the pattern of generation of committed progenitors. If this is possible, there are future prospects for intelligent clinical application. If not, we remain with a hematopoietic system in which selectivity of mature cell production is achievable using cytokines but at the cost of an inefficient perturbation of populations of less mature precursors, which might be deleterious in certain clinical situations.

Our present uncertainties regarding the questions under discussion would be reduced by work in two areas: (1) a solid, in principle, documentation using normal cells that commitment can be inducible by cytokines and (2) some serious efforts to establish whether selective activation of key nuclear transcription factors can be achieved reproducibly in multipotential cells by extrinsic agents, whatever their nature.