Biomolecular condensates — phase-separated drops of proteins and nucleic acids — provide vital spatial organization within cells. Model systems made up of proteins and/or nucleic acids are useful to elucidate the underlying principles that govern the phase separation of biomolecules. Here, I investigate a model system composed of branched, limited valence particles called DNA nanostars, which can phase separate into a DNA-rich phase and a DNA-dilute phase. Nanostars are unique in that they are composed entirely of DNA; thus, modulating individual nanostar properties is straightforward through sequence engineering of the constituent strands. Specifically, one can easily tune key molecular parameters associated with phase separation such as valence and the strength of interactions between particles. Additionally, because nanostars interact via base pairing, it is facile to relate nanostar phase behavior to the binding free energy of inter-nanostar interactions.
In this dissertation, I discuss three projects in which I examine various aspects of nanostar liquids, as motivated by key behaviors of biomolecular condensates. In the first project, I show that the physical properties of the nanostar dense phase — such as the viscosity, macromolecular density, and surface tension — are similar to the properties of biomolecular condensates. However, because nanostar interactions are driven by base pairing while other biomolecular condensates form via electrostatic attractions, the two condensates exhibit contrary responses to changing salt concentration. In the second project, I show that nanostar drops exhibit selective partitioning of solutes, analogous to behavior displayed by biomolecular condensates. I quantify the effects of adding one or two nanostar binding sites on solute particles and show that this can counter the lengthdependent entropic confinement that affects long solutes when entering the nanostar dense phase. In the third project, I investigate a system that mimics the heterogeneity of composition and interactions of biological condensates. When nanostars are combined with a positively charged polymer, I observe a rich phase diagram in which electrostatic and base pairing interactions cooperate in some cases and compete in others to form multiple distinct phases that can coexist.
Overall, this work demonstrates that nanostar phase behavior is highly predictable when phase separation is driven by nanostar–nanostar base pairing. Additionally, I use nanostar drops to mimic several key properties of biomolecular condensates and am able to draw conclusions about the effects of interaction thermodynamics on phase behavior. These results clarify why the nanostar system is interesting to investigate from a material standpoint as well — the dense phase has a surprisingly low volume fraction due to the rigidity of the nanostar arms and sequence-specific interactions can be used to predictably functionalize the nanostar dense phase. These properties can be exploited to create nanostar materials that target specific cell types, act as molecular sensors, or sequester relatively large solute particles.
