Summary Compartmentalization in cells is central to the spatial and temporal control of biochemistry. In addition to membrane-bound organelles, membrane-less compartments form partitions in cells. Increasing evidence suggests that these compartments assemble through liquid-liquid phase separation. However the spatiotemporal control of their assembly, and how they maintain distinct functional and physical identities is poorly understood. We have previously shown an RNA-binding protein with a polyQ-expansion called Whi3 is essential for the spatial patterning of cyclin and formin transcripts in cytosol. Here, we show that specific mRNAs that are known physiological targets of Whi3 drive phase separation. mRNA can alter the viscosity of droplets, their propensity to fuse, and the exchange rates of components with bulk solution. Different mRNAs impart distinct biophysical properties of droplets indicating mRNA can bring individuality to assemblies. Our findings suggest that mRNAs can encode not only genetic information, but also the biophysical properties of phase-separated compartments.

RNA promotes liquid-liquid phase separation (LLPS) to build membrane-less compartments in cells. How distinct molecular compositions are established and maintained in these liquid compartments is unknown. Here we report that secondary structure allows mRNAs to self-associate and determines if an mRNA is recruited to or excluded from liquid compartments. The polyQ-protein Whi3 induces conformational changes in RNA structure and generates distinct molecular fluctuations depending on the RNA sequence. These data support a model in which structure-based, RNA-RNA interactions promote assembly of distinct droplets and protein-driven, conformational dynamics of the RNA maintain this identity. Thus, the shape of RNA can promote the formation and coexistence of the diverse array of RNA-rich liquid compartments found in a single cell.

Melanoma is a tumor of transformed melanocytes, which are derived from the embryonic neural crest. It is unknown to what extent the programs regulating neural crest development interact with mutations in the BRAF oncogene, the gene most commonly mutated in human melanoma1. We have utilized the zebrafish embryo to identify initiating transcriptional events upon BRAFV600E activation in the neural crest lineage. Transgenic mitf-BRAFV600E;p53-/- zebrafish embryos demonstrate a gene signature enriched for markers of multipotent neural crest cells, and exhibit a failure of terminal differentiation of neural crest progenitors. To determine if these early transcriptional events were important for melanoma pathogenesis, we performed a chemical genetic screen to identify small molecule suppressors of the neural crest lineage, which were then tested for effects in melanoma. One class of compounds, inhibitors of dihydroorotate dehydrogenase (DHODH) such as leflunomide, led to an almost complete abrogation of neural crest development in the zebrafish and a reduction in self-renewal of mammalian neural crest stem cells. Leflunomide exerts these effects by inhibiting transcriptional elongation of genes required for neural crest development and melanoma growth. When used alone or in combination with a specific inhibitor of the BRAFV600E oncogene, DHODH inhibition led to a marked decrease in melanoma growth both in vitro and in mouse xenograft studies. Taken together, these studies highlight developmental pathways in neural crest cells that have direct bearing upon subsequent melanoma formation.