The wizard of Hox

Comment on Lawrence et al, page 3988

In this issue of Blood, Lawrence and colleagues provide strong evidence that HoxA9 plays a key physiologic role in the proliferation of early mouse hematopoietic stem progenitor cells.

Members of this group showed previously that overexpression of HoxA9 drove expansion of mouse hematopoietic stem progenitor cells. Indeed, many, perhaps all, Hox genes may be sufficient to “pharmacologically” drive stem cell expansion in gain-of-function studies (especially since forced expression of each of several Hox genes resulting from chromosomal translocations can drive leukemia cell proliferation) but only the loss of HoxB4 (or HoxB3 and B4 together) had previously resulted in impaired stem cell functionality. Lawrence and colleagues used elegant in vitro and in vivo analyses to demonstrate normal numbers of immunophenotype-defined hematopoietic stem cells but significantly decreased hematopoietic stem cell proliferation in HoxA9 knockout mice. They made this interpretation acknowledging a limitation: HoxA9 deficiency down-modulates proliferation of hematopoietic progenitor cells, and functional stem cell assay readouts all depend on progenitor cell function in addition to stem cell function. Of interest, when Lawrence et al evaluated the role of the nearest 5′ neighbor of HoxA9, they found no hematopoietic stem cell defect in HoxA10-deficient mice.

The 39 Hox family genes¹  are clustered at 4 chromosomal loci (HoxA1-7, A9-11, and A13 at chromosome 7p15; HoxB1-9 and B13 at 17p21; HoxC4-6 and C8-13 at 12q13; and HoxD1, D3, D4, and D8-13 at 2q31) in humans. All the mammalian Hox genes are highly homologous to the HOM-C genes of Drosophila, discovered by Bridges and Morgan,²  who in 1915 described the “bithorax” mutation, and later Lewis, who noted a “homeotic” mutant where an additional pair of legs was formed in place of the antennae. In mammals as in flies, Hox genes exhibit striking “collinearity” of expression and effect: during embryonic body development Hox gene expression profiles correlate with body segments and with developmental stage in a direct manner (ie, 3′ genes such as HoxA1 and B1 are expressed more anteriorly and earlier in development than their 5′ paralogs, which are expressed later and in more posterior regions). However, there is no “one Hox, one segment” rule; instead, chromosomally adjacent clusters of Hox genes (eg, HoxA9-11) are expressed in clusters at a given developmental stage of a given body segment. This redundancy is likely part of the reason that so few of the Hox genes are individually necessary for any given functional role in hematopoiesis. Moreover, hematopoiesis seems to break the collinearity rule. Clearly, global transcriptomic and proteomic methodologies should be of considerable utility in dissecting the physiologic functions of the Hox family and pathways in hematopoiesis.

As the normal physiology is being pains-takingly illuminated, the ability of certain Hox genes and TAT-Hox proteins to stimulate in vivo and ex vivo hematopoietic stem cell expansion is already nearing clinical application, especially with HoxB4. In this context we wonder the following. (1) Should we prefer any one Hox for clinical stem cell expansion? Specifically, should we test HoxA9 now that we know it is not only sufficient but also necessary for optimal stem cell function in vivo (in mice)? Or will most or all of these Hox paralogs function similarly? (2) Can we further enhance stem cell expansion by combined application of more than one Hox molecule? (3) As we learn more about stem cell functional wiring diagrams, can we productively improve the “context” for stem cell expansion by up-regulating or down-regulating Hox or other pathways? (4) Finally, determining how the Hox genes are regulated may permit efficient stimulation of the physiologic expression profile to optimally drive hematopoietic stem cell expansions. ▪