A reentrant transition in RNA aggregation

RNA molecules aggregate under certain conditions. The resulting condensates are implicated in human neurological disorders, and can potentially be designed towards specified bulk properties in vitro. However, the mechanism for aggregation---including how aggregation properties change with sequence and environmental conditions---remains poorly understood. I will show that a multimerization-based framework for aggregation replicates known experimental and simulation-based results, and makes concrete predictions for the aggregation of unstudied sequences. Our model reveals that the driving force for aggregation is the increased configurational entropy associated with the multiplicity of ways to form bonds in the aggregate. I will show that the simplest system of RNA with a single self-complementary sticker nonetheless exhibits rich phase behavior, including repeat parity-dependent aggregation and a sequence-dependent reentrant phase transition. I will also describe an extension of this system to the case of multiple orthogonal stickers. Our work unifies and extends published results, both explaining the behavior of CAG-repeat RNA aggregates implicated in Huntington’s disease, and enabling the rational design of programmable RNA condensates.