Gene-transcription factor regulatory networks implicate primary cilia in the evolution of vertebrate sex determination and expand models of epigenetic regulation.

The molecular architecture underlying diverse vertebrate sex-determining systems remains elusive despite fragmentary evidence of changes in upstream regulators and downstream mediators. Here we modeled species-specific regulatory networks of urogonadal development for turtles with contrasting mechanisms [Apalone spinifera - ZZ/ZW genotypic sex determination (GSD), and Chrysemys picta - temperature-dependent sex determination (TSD)] using matched data from time-course sampling. We uncovered key steps in the evolutionary transition of sex determination by testing for conservation or divergence of network modular components. Specifically, we tested these alternative hypotheses: first, transcription factor (TF) hubs and their target genes are conserved between species (null H0); second, the same TF hub acquired a new set of target genes in a species, retaining or not ancestral functions (H1 and variants); third, a new TF hub took over the regulation of the former gene targets of an ancestral TF (H2); and finally, complete overhaul occured where both ancestral TF hubs and their target genes were replaced in a species (H3). Results implicate primary cilia as integrators of environmental signals underlying TSD, because known thermosensitive TSD components (e.g., calcium-redox, pSTAT3, Wnt/Rspo1/β-catenin, Dhh) overrepresented in our results are linked to primary cilia. TFs that evolved between species also regulate primary cilia and point to key changes in their sensory machinery that accompanied TSD-GSD transitions (e.g., calcium/ion channels or membrane transport components in Chrysemys versus structural elements and ciliogenesis in Apalone). This novel Primary Cilia Integration hypothesis expands current models of epigenetic regulation of turtle sexual development, the evolution of plasticity versus canalization, and warrants functional validation.