Abstract
Pioneer transcription factors act at the top of gene-regulatory networks by promoting accessible chromatin at the cis-regulatory regions that drive gene expression. Despite their ability to bind closed chromatin, pioneer factors occupy distinct binding sites in different tissues. The pioneer factor Zelda promotes the undifferentiated fate in both the early Drosophila embryo and in the neural stem cells (neuroblasts) of the larval brain. Tissue-specific binding by Zelda identifies cell-type specific enhancers, which are enriched for different DNA-sequence motifs. We investigated the features that promoted cell-type specific occupancy by testing the role of conserved, structured protein domains in the capacity of Zelda to promote the embryonic and neuroblast cell fates. We unexpectedly identified that the most deeply conserved region in Zelda, the second zinc finger, has opposing functions in the embryo and neuroblasts. We showed that this zinc finger is a previously unrecognized DNA-binding domain that is specifically required for Zelda binding to a G-rich motif in larval neuroblasts. The pioneering function of Zelda depends largely on the C-terminal cluster of zinc fingers that promotes binding in the early embryo, suggesting that pioneer function may depend on how Zelda engages the genome. As opposed to co-factor expression or chromatin environment, our data identify tissue-specific usage of two, widely separated DNA-binding domains as the mechanism controlling tissue-specific binding and activity.