PAX Proteins and Fables of Their Reconstruction

Abstract

The PAX proteins derive their name from the "paired box," a region of homology first described between the Drosophila paired (Prd) and gooseberry (Gsb) proteins and later found to encode a sequence-specific DNA-binding activity. Both Prd and Gsb also contain a homeodomain, and this combination of DNA-binding domains is conserved in ancestral predecessors, reflecting an early "homeodomain-capturing" event. In addition, the prototypic PAX protein was thought to contain 2 additional features, namely the octapeptide (or eh1) motif and PHT (or OAR) domain-both modulate PAX regulatory activity but are not unique to the PAX family. Together with gene duplications and mutagenesis, a domain loss model accounts for the distinct architecture and sequence of extant PAX proteins. Despite the disparate evolutionary history of these 4 conserved motifs, there is a remarkable level of interplay that is modulated by discrete sequences elsewhere in the protein. Here, the implications with respect to the evolution of PAX protein structure and activity are discussed, it is suggested that the sum of these constituent domains is more than the contribution of individual parts. When combined with alternative splicing and posttranslational modifications, this model confers an extraordinary degree of functional diversity to even highly related PAX proteins.