Liquid-liquid phase separation and biomolecular condensates

Eukaryotic cells contain many membrane-less compartments, such as P granules, stress granules and p62 bodies, which are composed of highly concentrated biomacromolecules. These compartments are involved in various cellular processes, including germ cell specification, stress response, and selective autophagy. They are called “biomolecular condensates” and function in organizing cellular biochemical reactions, to make them relatively independent, highly ordered, and highly efficient. According to recent studies, biomolecular condensates exhibit liquid-like properties, such as a spherical shape, fusion upon contact and high dynamics. They arise from “liquid-liquid phase separation” (LLPS) driven by multivalent interactions among macromolecules. There are many ways to achieve macromolecular multivalence: Linear modular domains with similar function, like PRM repeats in N-WASP; oligomerization domains, like the PB1 domain in the selective autophagy receptor p62; similar post-translational modifications at multiple sites; poly-ubiquitination at one site; degenerate binding of RNA to proteins, etc. Additionally, intrinsically disordered regions (IDRs), sometimes called low complexity domains (LCDs), which contain high frequencies of a few specific amino acid types, are a common way to mediate multivalent interactions and drive LLPS. Many IDR-containing proteins, such as the P granule component PGL-3, the nuclear pore protein Nup98, and the RNA-binding protein FUS, were shown to undergo phase separation in vitro and localize to specific biomolecular condensates in vivo. RNAs with disease-related nucleotide repeat expansions, such as CUG repeats, CAG repeats, and GGGGCC repeats, can also form phase-separated puncta in vitro and in patients’ cells. Mechanisms for regulating phase separation include modulating the relationship between the concentration and saturation concentration of biomacromolecules, and altering biomolecular interactions. Regulation of transcription, translation and degradation can change biomacromolecule concentrations. Temperature, crowding agents, and ATP (which functions as a hydrotrope), can impact protein solubility. Protein modifications and novel interactions with other biomacromolecules can impact inter-molecular or/and intra-molecular interactions, thus disrupting phase separation. A special example is the nuclear transport receptor Kapβ2, which suppresses normal and aberrant phase separation of some RNA-binding proteins with IDRs, including FUS, by interacting with their PY-NLS region. The components that are essential for establishing phase separation are called “scaffolds”, and they enrich their “clients” via interaction through their free binding sites. The composition of clients can be altered by changing the stoichiometries of scaffolds. Droplets formed by different biomacromolecules, such as FIB1 and NPM, can co-exist to form immiscible multi-phase condensates, which are analogous to multi-layered membrane-less organelles in vivo, such as nucleoli and stress granules. Many in vitro phase-separation systems have been established, which can induce receptor clustering, activate actin or tubulin polymerization, facilitate enzymatic reactions, etc. Some studies have suggested a link between phase separation and heterochromatin formation or gene control. Based on reconstituted phase separation systems in vitro, and their links to biomolecular condensates in vivo, we can gain deep understanding of the assembly, dissolution, composition, physical properties, biochemical activities and cellular functions of biomolecular condensates.