FANCD2 Participates in JAK2-Dependent Activation of STAT5a

Because hematopoiesis is consistently perturbed in patients with Fanconi anemia (FA) and because FANCC is known to facilitate STAT1 activation we tested the notion that FA proteins might also facilitate activation of STAT5, a critically important signaling molecule for survival and replication of hematopoietic stem cells and progenitors. The numbers of bone marrow hematopoietic progenitor cells are greatly diminished in FA patients and no informative FA hematopoietic cell lines exist that respond to defined myeloid or erythroid growth factors. Consequently, we screened widely available FA fibroblast cell lines for STAT5 activation defects by quantifying nuclear and cytoplasmic tyrosine phosphorylated STAT5 (STAT5-P) in hydrogen peroxide-exposed and EGF-exposed cells using cell fractionation, immunoblotting and deconvolution microscopic methods. In normal fibroblasts and in isogenic complemented cell lines from three complementation groups (C, G, and D2) H2O2 induced JAK2-dependent STAT5 phosphorylation and nuclear translocation at least in part by activating EGFR and IGFR. In fibroblasts deficient in FANCC, FANCG, and FANCD2, cytoplasmic STAT5 was phosphorylated, but not translocated to the nucleus and apoptotic responses to H2O2 were exaggerated. The apoptotic responses to H2O2 were complemented by a constitutively active mutant form of STAT5a (STAT5A-N642H) which did translocate to the nucleus even in mutant cells. Nuclear STAT5-P was also reduced in EGF stimulated FA fibroblasts and in erythropoietin (EPO)-stimulated MO7E-epoR cells in which either FANCC or FANCD2 was knocked down with siRNA. Using co-immunoprecipitation and size exclusion chromatographic methods we found that in normal cells but not in FA cells, H2O2 induced formation of multimeric cytoplasmic complexes containing STAT5, FANCD2, Hsp90, and FANCC. Inactivation of any one of the three FA proteins interdicted the inducible association of the other two with STAT5 and prevented nuclear translocation of STAT5-P. Because we found no evidence of differential genotoxic responses to H2O2 between mutant and complemented cells, we conclude that activation of a JAK2-dependent STAT5a signaling pathway in response to extracellular H2O2 requires inducible intermolecular interactions involving STAT5, FANCC, and FANCD2, interactions that facilitate nuclear translocation of tyrosine phosphorylated STAT5. This is the first report of the participation of FANCD2 in a canonical cytoplasmic signaling pathway. Because optimal STAT5 activation in response to EGF (in fibroblasts) and erythropoietin (in MO7E-EPO-R cells) also induces the formation of STAT5/FANCC complexes, we suggest that failure of this signaling pathway may contribute to the Fanconi anemia phenotype. Results of studies designed to confirm this notion in primary hematopoietic cells will be presented. That the constitutively active mutant form of STAT5a (STAT5A-N642H) was able to undergo nuclear translocation without requiring an intact FANCC/D2/STAT subcomplex suggests that small molecules designed to alter the tertiary structure or binding affinities of native STAT5a might facilitate nuclear translocation even in mutant (FA) cells. The use of such molecules might effectively treat bone marrow failure in FA-C, FA-D2, and FA-G patients.