Identifying the molecular basis for functional divergence of duplicated SOX factors controlling endoderm formation and left-right patterning in zebrafish

Endoderm, one of three primary germ layers of vertebrate embryos, makes major contributions to the respiratory and gastrointestinal tracts and associated organs, including liver and pancreas. Placement and patterning of these organs relies on the left-right organiser – known as Kupffer's Vesicle (KV) in zebrafish. Transcription factors Sox32 and Sox17 are members of the zebrafish SoxF subfamily. Sox32 and sox17 arose from duplication of ancestral Sox17 in the teleost lineage. Sox32 induces sox17 expression in early embryos and is required for specification of endoderm and KV progenitors. Zebrafish Sox17 is implicated in KV morphogenesis. In mammals, Sox17 is vital for endoderm organ formation and can induce endoderm progenitor identity. Phenotypic evidence therefore suggests functional similarities between zebrafish Sox32 and Sox17 with mammalian SOX17. We sought to explore the functional differences and potential similarities between these proteins in the early zebrafish embryo. Our results indicate that, unlike Sox32, human SOX17 cannot induce endoderm specification in zebrafish. Furthermore, using hybrid protein functional analyses, we show that Sox32 specificity for the endoderm gene regulatory network is linked to evolutionary divergence in its HMG domain from its paralogue Sox17. Additionally, changes in the C-terminal regions of Sox32 and Sox17 underpin their differing target specificity and divergence in mediating differential gene regulatory programmes. Finally, we establish that specific conserved peptides in the C-terminal domain are essential for the role of Sox17 in establishing correct organ asymmetry. Overall, our results provide novel insights into vertebrate endoderm development, left-right patterning, and the evolution of SoxF transcription factor function. Overall design: To investigate the functional differences between SOX family transcription factors implicated in endoderm formation and left-right patterning (mammalian SOX17 and zebrafish Sox17 and Sox32) we injected equimolar quantities of mRNA encoding these factors into zebrafish embryos at the one-cell stage and performed RNA-seq analyses at 6 hours post-fertilisation. To determine whether differences in the HMG domains between zebrafish Sox17 and Sox32 were critical to their distinct functions we produced mRNAs for Sox32 lacking the HMG domain, or where the Sox32 HMG domain had been substituted for that of Sox17. We then performed RNA-seq analyses as previous.