In nascent and rapidly developing areas of research it is not uncommon that key papers emerge a little before their time. One such paper was the work by Tim Nott, Julie Forman-Kay and Andy Baldwin published in early 2015 in Molecular Cell. This paper represents one of the first examples of a rigorous and detailed dissection of the biophysical determinants of liquid–liquid phase separation (LLPS) both in vitro and in cells. It provides a conceptual and practical blueprint for how experiment, theory and bioinformatics can be seamlessly combined to draw strong and general conclusions about the physical rules and molecular interactions underlying LLPS (the ‘molecular grammar’).
It was in 2009 that Cliff Brangwynne and Tony Hyman’s now classic paper describing P granules of the Caenorhabditis elegans germ cells as liquid droplets was published in Science. This paper sparked the realization that perhaps many membraneless cellular bodies could be described by physical principles traditionally applied to coexisting liquid equilibria. Subsequently, combined effort from many laboratories delineated the key physical principles and molecular interactions that underlie the formation of phase-separated molecular assemblies in cells — now known as biomolecular condensates — from multivalent proteins (see Banani et al. 2017 for review). However, it was Nott et al. that first rigorously and quantitatively connected the specific physical chemistry of a disordered region with its phase diagram — a line of inquiry that was only picked up again a few years later as the underlying molecular grammar was more fully explored.
DDX4 is a mammalian DEAD-box helicase essential for the formation of germ granules. Given the parallels between human germ granules and P bodies in C. elegans, the authors pursued the hypothesis that human DDX4 may itself be capable of undergoing LLPS alone. From these humble beginnings, Nott et al. uncovered an array of (bio)physical principles driving LLPS.
Among various insights, the authors demonstrated that remarkably simple models from polymer physics are sufficient to quantify the salt and temperature-dependence of LLPS for naturally occurring disordered proteins, that charge–charge, cation–pi, and aromatic-mediated interactions play key roles in driving assembly, and that methylation of arginine diminishes these interactions. Moreover, they found that DDX4 droplets can solubilize nucleic acids and revealed that sequence features of the N-terminal disordered domain of DDX4 responsible for its LLPS are conserved across a wide range of proteins found in other membraneless bodies.
“Nott et al. uncovered an array of (bio)physical principles driving LLPS”
Since this work was published, the principles uncovered in this study have held true across a remarkable array of different proteins and biomolecular condensates, including for biomolecules with seemingly disparate physical properties compared with DDX4. Ultimately, the N-terminal disordered domain from DDX4 has emerged as a ‘classic’ model domain for the study of LLPS.
Nott, T. J. et al. Phase transition of a disordered nuage protein generates environmentally responsive membraneless organelles. Mol. Cell 57, 936–947 (2015)
Brangwynne, C. P. et al. Germline P granules are liquid droplets that localize by controlled dissolution/condensation. Science 324, 1729–1732 (2009)
Banani, S. F. et al. Biomolecular condensates: organizers of cellular biochemistry. Nat. Rev. Mol. Cell Biol. 18, 285–298 (2017)
A.S.H. is a scientific consultant with Dewpoint Therapeutics.
About this article
Cite this article
Holehouse, A.S. Mapping from sequence to droplets. Nat Rev Mol Cell Biol 22, 163 (2021). https://doi.org/10.1038/s41580-021-00333-1