Rebuttal to Metcalf

THE REVIEW OF lineage commitment and maturation presented by Donald Metcalf complements and agrees with many of the points that were raised in own article. Indeed, it seems in the end that it was easier for us three limeys to accomodate the views of our esteemed antipodean colleague than those of each other! Thus, we all agree that, although hematopoietic cell regulatory molecules can play an active role in expansion and maturation of committed cells, a role in lineage commitment has not yet been really identified. Nonetheless, Metcalf has raised some issues that need addressing.

First, we agree that extrinsic influences and, in particular, gradients of inducing molecules influence embryonic development. And, it is reasonable to propose that generation of hematopoietic stem cells in the embryo arises via similar mechanisms. Indeed, candidate inducing agents have been tabled and are currently under intense investigation (see Neave et al1 and refs therein). However, it is also possible that known hematopoietic regulatory molecules may have a role to play in the decision making processes, because, contrary to the suggestion made by Metcalf, very primitive and nonhematopoietic cells do indeed have receptors for, and can respond to, known regulators of hematopoiesis, eg, primordial germ cells to stem cell factor2 and embryonic stem cells to leukemia inhibitory factor.3 Furthermore, because low levels of receptor expression are very difficult to detect, it could be that such primitive cells express receptors for some of the other, more classical, regulators of hematopoiesis. Still, the question remains as to whether or not these act in an inductive capacity.

As we and Metcalf point out, many of the experiments purporting to demonstrate an inductive role of hematopoietic regulatory molecules, may simply be showing selective survival and expansion of previously committed cells. Of course, absence of evidence is not evidence of absence and so we cannot dismiss Metcalf's assertion that as yet undiscovered molecules play instructive roles in the process of early lineage commitment. That said, and until we have more robust data (such as following the fate of individual cells in a homogenous multipotential stem cell population), we believe that the current balance of evidence presented both by us and by Metcalf favors the hypothesis, put forward in our review and elsewhere, that differentiation of stem cells involves mainly stochastic processes but that instructive events probably play a greater role in more lineage-restricted cells.