Fig. 3. The binding of both Mad and Zen to the Race enhancer is required for proper Race expression. (A) DNAse I footprinting analysis of Mad and Zen GST fusion proteins bound to a 255 bp fragment that includes the proximal region of the Race 533 bp enhancer (349-533) and 70 nucleotides from the Bluescript vector. The fragment is end-labeled at the vector end. Lane 1, chemical degradation of the probe on G+As; lanes 2 and 6, DNAse I digestion of the DNA probe. Increasing amounts of Mad (500 ng, 1500 ng and 4500 ng in lanes 3-5 respectively) and Zen (20 ng, 60 ng and 200 ng in lanes 7-9 respectively) were incubated with the fragment prior to DNAse I digestion. The region protected by Mad is depicted as a blue rectangle, the hatched half denoting weaker protection. The regions protected by Zen are shown as red ovals. The nucleotide sequence of the protected regions are shown below the gel with the overlap between the Zen and Mad footprints shown in purple. Putative core binding sites are underlined for Zen (Han et al., 1989) and boxed for Mad (boxes a,b,c,f,g) (Kim et al., 1996) and Medea (boxes d,e) (Xu et al., 1998; Pyrowolakis et al., 2004). The boxes are also marked on the G+A sequence. (B) Schematic representation of the full-length Race 533 bp enhancer fused to a lacZ reporter gene, and a transgenic embryo carrying this construct in situ hybridized with lacZ probes. lacZ expression is identical to the Race pattern. The ring of staining in the head region is an artifact of the vector. (C) Embryo carrying a deletion of the Mad-binding region (nucleotides 432-497). lacZ expression is severely reduced. (D) Embryo carrying mutations in the ATTA core sites of the Zen-binding sites (underlined, see Materials and methods). lacZ expression is absent.