Studies on the leukemogenic process have revealed the implication of at least 2 pathways in the generation of neutrophils, one involving a membrane receptor, GM-CSFR, and the other one, nuclear receptors of the RAR family. In tissue culture, GM-CSF induces CD34+Lin cells to terminal myeloid differentiation. Kit ligand (or SCF) antagonizes this effect whereas retinoids enhance it. Furthermore, cells lacking RARα/RARγ fail to undergo terminal maturation when exposed to GM-CSF or IL-3 in vitro. It came therefore as a surprise that neither gm-csfnor RARα gene knock-out disrupts granulopoiesis in vivo, suggesting redundancy and/or compensatory mechanisms. The interest in GM-CSF resurfaced with the discovery that leukemic cells in juvenile chronic myeloid leukemia (JCML) are hypersensitive to GM-CSF, followed by the demonstration of a genetic interaction betweengm-csf and Nf1, a gene frequently deleted in JCML. In acute promyelocytic leukemia (APL), the involvement of theRARα locus in chromosomal translocations with 5 distinct fusion partners indicates that loss of RARα function is the primary event in leukemogenesis.

Johnson and colleagues (page 746) have now provided unequivocal evidence that GM-CSF and IL-3 enhance both ligand-dependent and ligand-independent transcriptional activity of RA receptors. Moreover, RA receptor activation is exquisitely dependent on GM-CSFR and IL-3R signaling and does not occur with activated c-Kit. This induction correlates with biologic outcome in neutrophil maturation and, together with the RAR gene knock-outs, suggests a critical role for RA receptor activation in IL-3 and GM-CSF signaling. Since loss of RAR function is associated with APL, a colinearity between RAR and GM-CSF implies that a loss of function of GM-CSF might also be associated with myeloid leukemia. Paradoxically, GM-CSF signaling is essential in the development of a myeloproliferative disease inNf1−/− mice. It is possible that signaling pathways have different outcomes depending on the context of the cell, a hypothesis set forth by the authors to explain the differential sensitivity of myeloid leukemias to ATRA therapy. Regardless of what future experiments will tell us, Johnson and colleagues unravel here unsuspected crosstalks between 2 disparate receptor families, bringing a novel twist to our understanding of signaling pathways.

Sign in via your Institution