A Screen for Regulators of Globin Switching in the Zebrafish Embryo

Abstract 826

The switching of the globin genes involves critical transcriptional regulators such as BCL11A, EKLF and SOX6, and the induction of fetal globin has been shown to ameliorate the symptoms of diseases such as sickle cell anemia. Recently, there has been interest in driving iPS cells to produce mature red cells that express adult globin genes in an attempt to make these cells therapeutically useful. Here, to understand hemoglobin switching and the molecular pathways that allow the establishment of an adult fate in embryonic tissues, we utilized a screening approach in the zebrafish model. The concept of the screen is to find transcription factors that are expressed in a stage-specific manner, and manipulate the expression of these genes to alter the cell fate of embryonic erythroid cells. In order to generate a candidate list of genes, microarray analysis was performed on murine yolk sac, fetal liver and adult derived red blood cells and red blood cell precursors, which express unique sets of globin genes. Pair-wise comparison of these populations yielded 879 unique differentially regulated genes. GO term analysis was used to narrow the list to 49 transcription factors. We focused on the transcription factors that might increase adult globin expression in the embryo based on their differential expression in the microarrays. Morpholinos were used to knock down these 24 genes by individually injecting each into one-cell stage embryos, allowing the embryos to reach 24 hpf and performing in situ hybridization for the adult globin gene α_a1. The number of adult globin positive cells present in each embryo was counted for a clutch control group, which on average has 2–4 positive cells per embryo, and three doses of morpholino. We identified 4 genes, Tcf7l2, Ncoa1, Hif1al and E2F5, the knock down of which results in a significant increase in the number of adult globin positive cells in at least one dose of morpholino (control [n=53, mean=6.34], 6ng [n=56, mean=15.07], p=<0.0001; control [n=35, mean=1.543], 4ng [n=56, mean=2.75], p=<0.01; control [n=19, mean=1.368], 12ng [n=16, mean=6.188], p=<0.0001; control [n=44, mean=1.091], 4ng [n=30, mean=2.7], p=<0.05, respectively). Pair-wise knock down of these genes were also tested, and the combinations of Ncoa1 and E2F5, Tcf7l2 and E2F5 and Tcf7l2 and Ncoa1 were found to synergistically increase the number of adult globin expressing cells (control [n=49, mean=0.5306], knock down [n=38, mean=9.895], p=<0.0001; control [n=49, mean=7.633], knock down [n=54, mean=17.41], p=<0.0001; control [n=20, mean=2.95], knock down [n=28, mean=too numerous to count], p=<0.0001, respectively). The combined knock down of Tcf7l2 and Ncoa1 was both the strongest inducer of adult globin expression and had the lowest toxicity of the pair-wise combinations. Further characterization of this phenotype shows that, while many globin genes are up regulated, both of the adult globin genes, α_a1 and β_a1, are upregulated to a higher degree than other globin genes. In order to determine if the Wnt pathway is responsible for phenotype observed with the Tcf7l2 morpholino, we tested the Wnt pathway inhibitors IWR1 and XAV939. Both drugs phenocopied the Tcf7l2 knockdown response. In addition, XAV939 synergies with the Ncoa1 morpholino to enhance the increase in adult globin observed in a similar manner to Tcf7l2 knockdown. These results indicate that modulation of Wnt signaling, rather than a Wnt-independent function of Tcf7l2, is responsible for the phenotype and regulation of globin gene expression. Chip-Seq analysis of Ncoa1 occupancy in the erythroid cell line K562 was performed to examine potential mechanisms of action. Significant binding was observed at the enhancers of the α- and β-globin loci, indicating that the nuclear hormone receptor pathway may be acting directly on the globin loci to modulate globin expression patterns. These results indicate that Wnt signaling in combination with alterations of other pathways regulated by Ncoa1 are responsible for stage-specific globin expression. Our studies have impact on the understanding of globin switching in vertebrates, and could establish new methods to activate specific globins clinically, and to make iPS cells form adult-type tissues.