Runx1 represses pu.1 expression through acting on pu.1 promoter. (A) Schematic diagram of pu.1 promoter. Two adjoining putative runx1 binding sites (red arrowheads) in the distal portion of pu.1 promoter are predicted by promo, Version 3.0 online software. Red bars represent positions of primers designed to test for Runx1 binding. runx1-I amplifies the region containing 2 putative Runx1 sites, whereas runx1-II amplifies the region devoid of runx1 sites. (B) Representative fluorescent images (left panels) of transient eGFP expression at 17.5 hpf in the RBI of WT embryos injected with −9.0pu.1:eGFP (top panels) and −9.0pu.1ΔR:eGFP (bottom panels) constructs. Right panels: Overlays with bright field images. (C) Quantitative RT-PCR for GFP expression at 17.5 hpf in WT embryos injected with −9.0pu.1:eGFP and −9.0pu.1ΔR:eGFP. Units on y-axis represent the relative fold change of GFP expression in WT embryos injected with −9.0pu.1:eGFP and −9.0pu.1ΔR:eGFP. Expression level was normalized with elf1α expression and the amount of injected DNA. Error bars represent SE. (D) Semiquantitative PCR analysis with chromatin before (input) and after immunoprecipitation with anti-Myc antibody or anti-BrdU antibody (negative control). Sequence of mespa gene promoter serves as a negative control. (E) A model for the regulatory network in controlling embryonic neutrophil and macrophage fate segregation. In this model, a graded Pu.1 level specifies embryonic neutrophil and macrophage fates with high Pu.1 activity required for macrophage fate formation and low Pu.1 supporting neutrophil fate formation. High Pu.1 activity might switch on the expression of its binding partner, Irf8, to establish the embryonic macrophage fate. High Pu.1 activity, on the other hand, turns on the expression of Runx1, which is a direct feedback repressor of pu.1 expression. This Pu.1-Runx1 negative feedback loop thus stabilizes a favorable Pu.1 level that is essential for the formation of neutrophil fate.