SUMMARYEaf factors play a crucial role in tumor suppression and embryogenesis. To investigate the potential mechanism of Eaf activity, we performed loss-and gain-of-function assays in zebrafish using morpholino and mRNA injections, respectively. We found that eaf1 and eaf2 inhibit Wnt/β-catenin signaling, thereby modulating mesodermal and neural patterning in the embryo. Moreover, ectopic expression of eaf1 and eaf2 in embryos and cultured cells blocked β-catenin reporter activity. By immunoprecipitation, we also observed that Eaf1 and Eaf2 bound to the Armadillo repeat region and C-terminus of β-catenin, as well as to other β-catenin transcription complex proteins, such as c-Jun, Tcf and Axin, suggesting the formation of a novel complex. In addition, the N-terminus of Eaf1 and Eaf2 bound to β-catenin and exhibited dominant-negative activity, whereas the C-terminus appeared to either harbor a suppression domain or to recruit a repressor. Both the N-and C-terminus must be intact for Eaf1 and Eaf2 suppressive activity. Lastly, we demonstrate a conservation of biological activities for Eaf family proteins across species. In summary, our evidence points to a novel role for Eaf1 and Eaf2 in inhibiting canonical Wnt/β-catenin signaling, which might form the mechanistic basis for Eaf1 and Eaf2 tumor suppressor activity.

FOXO genes are involved in many aspects of development and vascular homeostasis by regulating cell apoptosis, proliferation, and the control of oxidative stress. In addition, FOXO genes have been showed to inhibit Wnt/β-catenin signaling by competing with T cell factor to bind to β-catenin. However, how important of this inhibition in vivo, particularly in embryogenesis is still unknown. To demonstrate the roles of FOXO genes in embryogenesis will help us to further understand their relevant physiological functions. Zebrafish foxo3b gene, an orthologue of mammalian FOXO3, was expressed maternally and distributed ubiquitously during early embryogenesis and later restricted to brain. After morpholino-mediated knockdown of foxo3b, the zebrafish embryos exhibited defects in axis and neuroectoderm formation, suggesting its critical role in early embryogenesis. The embryo-developmental marker gene staining at different stages, phenotype analysis and rescue assays revealed that foxo3b acted its role through negatively regulating both maternal and zygotic Wnt/β-catenin signaling. Moreover, we found that foxo3b could interact with zebrafish β-catenin1 and β-catenin2 to suppress their transactivation in vitro and in vivo, further confirming its role relevant to the inhibition of Wnt/β-catenin signaling. Taken together, we revealed that foxo3b played a very important role in embryogenesis and negatively regulated maternal and zygotic Wnt/β-catenin signaling by directly interacting with both β-catenin1 and β-catenin2. Our studies provide an in vivo model for illustrating function of FOXO transcription factors in embryogenesis.

Studies implicate a potential role for EAF1 in MLL-ELL induced leukemogenesis; however the biological function of EAF1 in this process remains unknown. In this study, we show that knockdown of zebrafish eaf1 by morpholinos caused serious defects in both primitive and definitive hematopoiesis. Using microarray analysis, we identified foxo3b as a target gene suppressed by eaf1. Ectopic expression of foxo3b in embryos mimicked the phenotypes exhibited in eaf1 morphants, except that foxo3b had no effect on runx1 and c-myb expression while eaf1 morphants did not express these markers in the ventral wall of dorsal aorta. Subsequent experiments showed that a dominant negative form of Foxo3b (dn-foxo3b) partially restored primitive hematopoietic defects in eaf1 morphants, suggesting that foxo3b might serve as a key factor for mediating eaf1 function in primitive hematopoiesis. Furthermore, we observed that foxo3b inhibited the transcriptional activity of gata1 and spi1 through protein-protein interaction. Our findings not only suggest a function of eaf1 on hematopoiesis in vivo, but also reveal a novel regulatory pathway, eaf1-foxo3b-gata1/spi1, that may shed light on the role of EAF1 in MLL-ELL induced leukemogenesis.