TALE family transcription factors (TFs) are ubiquitously expressed and act as cofactors to several classes of
essential TFs – most prominently Hox proteins – in embryogenesis and cellular homeostasis, but their exact
functions remain enigmatic. For instance, two family members – Prep and Meis – bind identical sequence
motifs in vitro and share the ability to dimerize with Pbx proteins, suggesting that they function interchangeably.
However, loss-of-function analyses in several systems suggest that their roles diverge in vivo. Furthermore, a
third TALE family member (TGIF) shares DNA binding motif preference with Prep and Meis, but it is unclear if
these three TFs compete or compensate for each other in vivo. Because most known TFs belong to larger
families that share DNA and protein:protein interaction properties, similar questions about shared and
divergent functions vex our understanding of most TF families. Further, since TF function in vivo is subject to
constraints not encountered in vitro, it is essential to evaluate their functional properties in native systems.
Indeed, numerous fundamental questions about TF activity remain to be addressed in vivo. Perhaps most
importantly, do TFs with similar in vitro binding selectivity select distinct motifs in vivo? What is the mechanistic
basis of divergent in vivo binding selectivity and how does it impact function? The answers to these questions
will have profound implications for our understanding of embryogenesis and disease, but progress in this area
has been hampered by major barriers. Specifically, access to multiple genome-wide data sets for closely
related TFs has been limited – particularly during embryogenesis, when cell numbers are often small. We have
addressed this by generating ChIP-seq and RNA-seq data for members of the TALE family of TFs at multiple
stages of zebrafish development – thereby establishing one of the most comprehensive sets of data available
for a single TF family. In spite of containing near-identical homeodomains and binding to indistinguishable
motifs in vitro, we find that Prep and Meis TFs display divergent binding preferences in vivo. Also, Prep, but not
Meis, occupies a novel non-Hox related genomic element in vivo. These initial observations underscore the
importance of exploring TF function in their native environment and highlight the strong technical and
conceptual position of our research group to pursue these analyses further. Based on our preliminary findings,
we hypothesize that emergent in vivo constraints restrict TALE TF motif selectivity and that dynamic
exchange among TALE members controls transcriptional outcome. To test this hypothesis, we will first
express wild-type and domain-swapped TF constructs in zebrafish to define the mechanistic basis of TALE TF
binding selectivity in vivo. Second, we will use gain- and loss-of-function approaches to manipulate the balance
of TALE TFs in order to define the functional consequences of different TALE TFs occupying the same
genomic binding sites in vivo. Lastly, we will use loss-of-function animals and CRISPR-mediated deletions to
examine the in vivo role for a novel TALE-occupied motif. Since TALE TFs are implicated in several cancers,
our in vivo delineation of the dynamic interplay between TALE family members will directly impact our
understanding of both embryogenesis and oncogenesis. Conceptually, our findings will also be applicable to
other homeodomain TFs (constituting the 2nd largest class of TFs) and other multi-member TF families.