Unsweetened Notch leads to myeloproliferation

The article by Zhou and colleagues in this issue of Blood highlights the importance of Notch glycosylation in suppression of myelogenesis.

The extracellular domain of all 4 mammalian Notch receptors contains 29 to 36 epidermal growth factor (EGF)–like repeats, many of which are predicted to be modified with 2 unusual carbohydrates: O-linked fucose and O-linked glucose. Work in a number of laboratories has revealed that O-fucosylation is essential for Notch function in many contexts,¹  and a recent report has shown that O-glucosylation is essential for Notch function in Drosophila.²  In addition, elongation of O-fucose by members of the Fringe family of β1,3-N-acetylglucosaminyltransferases modulates Notch activity. Several years ago, Smith et al developed a mouse lacking a key enzyme in GDP-fucose biosynthesis (FX^−/−).³  These mice lack all forms of fucosylation, including O-fucosylation of Notch, and do not survive long after birth unless fucose is added to their diet. A salvage pathway uses the fucose to rescue GDP-fucose biosynthesis, restoring fucosylation. In the absence of fucose, the FX^−/− mice display a profound neutrophilia. Part of this neutrophilia can be explained by loss of selectin ligands (Sialyl Lewis x contains fucose), but proliferation of myeloid progenitor cells suggests that myelopoiesis is being stimulated. Zhou and coworkers have now examined the proliferation of myeloid lineages in FX^−/− mice and attributed it to loss of fucose-dependent Notch activation in myeloid progenitors. The authors present compelling data that suggests a role for Notch activation in suppression of myeloid differentiation, a somewhat controversial area. Another recent publication highlighted the importance of O-fucose at a specific site on Notch1 in T-cell development.⁴  Ge and Stanley generated a mouse in which endogenous Notch1 was replaced with a mutant lacking the O-fucosylation site in the ligand-binding domain (within EGF repeat 12). Homozygotes developed fairly normally but had a reduced number of T cells, suggesting that O-fucosylation of Notch1 at EGF repeat 12 is important for T-cell development.

These results raise a number of interesting questions. All 4 receptors should be unfucosylated in FX^−/− mice, but it is not known which Notch receptor is responsible for suppression of myeloproliferation. As mentioned, modification of O-fucose by Fringe modulates Notch activity. In the absence of Fringe, O-fucose remains a monosaccharide, but in the presence of Fringe it is elongated to a tetrasaccharide. The relevant structures of the O-fucose glycans that are lost in FX^−/− mice are unknown. Because O-fucosylation of EGF repeat 12 in Notch1 plays such an important role in T-cell development, it would be interesting to know if loss of this specific fucose also suppresses myelogenesis. The future of Notch and hematopoiesis certainly looks sweet.