The Role of Protein-Protein Interactions and Phase Separation in Biophysics

Liquid-liquid phase separation is a form of polymer chemistry, recently found to be particularly relevant to cell biology, that involves the de-mixing of enclosed molecules into a dense phase and a depleted phase [1]. This results in the formation of various membrane-less organelles, such as P bodies and stress granules. Summarizing several studies of various membrane-less organelles, three basic principles underlie all compartments [2]: first, membrane-less organelles arise from the phase separation of proteins or proteins and nucleic acids from the surrounding milieu; second, membrane-less organelles remain in a liquid-like state but have properties distinct from those of the surrounding milieu; and third, proteins exchange with membrane-less organelles in the span of seconds instead of minutes/hours or longer, as generally occurs for subunits of stable complexes. The dynamic and liquid-like properties of membrane-less organelles present unique opportunities to control the biochemical environment of the cell that are distinct from those provided by canonical macromolecular complexes [3].

At the shortest length of time, membrane-less organelle formation can substantially increase the local concentration of resident chemical species. This may enhance the reaction rates of certain processes and increase overall reaction kinetics inside the structure. On the other hand, changes in the physical properties of membrane-less organelles (e.g., changes in viscoelasticity) can slow the diffusion of the enclosed molecules and decrease the overall reaction kinetics. Membrane-less organelles may also act to decrease cellular activity by sequestering molecules from their sites of action and/or corresponding substrates and inhibiting their activity outside of the structure. Moreover, membrane-less organelles can enhance the overall reaction activity by excluding a negative regulator of certain reactions. Finally, membrane-less organelles can be used to buffer against biological fluctuations by maintaining the concentration of essential components at the solubility threshold of the structure – even if more of this component is added to the system, the overall concentration remains the same by increasing the volume of the structure. Taken together, these scenarios demonstrate that membrane-less organelles can modulate these potential functions and can be switched on and off through the formation and dissolution of a membrane-less organelle. This suggests that the functional control of membrane-less organelles in short, biochemical length scales is likely to be prevalent in condensate biology.

Despite the ever-growing library of proteins known to be able to phase separate in vitro and in vivo, the underlying molecular mechanisms that result in the formation of these membrane-less organelles have not been well defined and are currently heavily debated. One commonality amongst these theories is that proteins associated with membrane-less organelles often exhibit multivalent features that arise in various ways. Folded proteins, with well-defined interaction surfaces, can often form oligomers that bring multivalence with other associative patches [4]. Folded domains can be strung together by flexible linkers to generate linear multivalent proteins [5]. Moreover, intrinsically disordered regions and/or proteins can serve as scaffolds for multiple, distinctive, short linear motifs [6]. Within intrinsically disordered regions, multivalent interactions can be encoded through various uncharged polar side chains (e.g., glutamine), charged amino acids (e.g., glutamic acid), and aromatic residues (e.g., tyrosine). Critical interactions that drive multivalency and phase separation arise from numerous weak interactions that additively counteract the entropic cost for phase separation. These molecular interactions found to be important in phase separation likely include electrostatic, dipole-dipole, pi-pi (π-π), hydrophobic, and hydrogen-bonding interactions [7].

Unfortunately, the biophysical experimentation used to study liquid-liquid phase separation has often been very limiting, in part due to the highly dynamic nature of the resulting membrane-less organelles. There is some reported success with crosslinking approaches, but deciphering the physical basis of membrane-less organelles will ultimately require advancing new technologies or re-purposing existing ones. Recently, I reported the use of hydrogen-deuterium exchange mass spectrometry (HXMS) to uncover local changes in protein structure upon phase separation [8]. HXMS has become a key and powerful technique to monitor and probe protein conformation dynamics and protein interactions in solution. This approach is based primarily on the main chain amide NH hydrogens and relies on the fact that exposure of a protein to D₂O (i.e., heavy water) induces rapid amide H à D (HX) exchange in regions that lack hydrogen bonding in a time-dependent fashion [9]. In terms of the physical basis of membrane-less organelle formation, the most prominent changes in backbone dynamics can be measured as additional protection from HX exchange, primarily localized to discrete portions of proteins that are involved in multivalent interactions.

In the absence of structural information or useful structural models (i.e., cases of multimeric complexes where current biophysical and structural predictions fall short), HXMS provides localization information at a moderately high resolution. I envision that HXMS will be broadly useful to advance our physical understanding of the protein/protein and protein/nucleic acid interactions that drive phase separation and the formation of membrane-less compartments within the cell. Specifically, I foresee the use of HXMS as an essential step to localize the key contacts in the liquid-liquid demixed state; then, this data can be combined with any available structural information, including crystal contacts when they are available, and various mutational analyses to highlight the multivalency that is theorized to be the key driver of phase separation.

[1]: Spannl, Stephanie, Tereshchenko, Maria, Mastromarco, Giovanni J., Ihn, Sean J., and Lee, Hyun O. 2019. “Biomolecular condensates in neurodegeneration and cancer.” Traffic

[2]: Brangwynne, Clifford P. 2013. “Phase transitions and size scaling of membrane-less organelles.” Journal of Cell Biology

[3]: Lyon, Andrew S., Peeples, Willian B., and Rosen, Michael K. 2021. “A framework for understanding the functions of biomolecular condensates across scales.” Molecular Cell Biology

[4]: Li, Pilong, Banjade, Sudeep, Cheng, Hui-Chun, Kim, Soyeon, Chen, Baoyu, Guo, Liang, Llaguno, Marc, Hollingsworth, Javoris V., King, David S., Banani, Salman F., Russo, Paul S., Jiang, Qiu-Xing, Nixon, B. Tracy, and Rosen, Michael K. 2012. “Phase transitions in the assembly of multivalent signalling proteins.” Nature

[5]: Banani, Salman F., Rice, Allyson M., Peeples, William B., Lin, Yuan, Jain, Saumya, Parker, Roy, and Rosen, Michael K. 2016. “Compositional Control of Phase-Separated Cellular Bodies.” Cell

[6]: Nott, Timothy J., Petsalaki, Evangelia, Farber, Patrick, Jervis, Dylan, Fussner, Eden, Plochowietz, Anne, Craggs, Timothy D., Bazett-Jones, David P., Pawson, Tony, Forman-Kay, Julie D., and Baldwin, Andrew J. 2015. “Phase transition of a disordered nuage protein generates environmental responsive membraneless organelles.” Molecular Cell

[7]: Gomes, Edward and Shorter, James. 2019. “The molecular language of membraneless organelles.” Journal of Biological Chemistry

[8]: Bryan, Nikaela W., Ali, Aamir, Niedzialkowska, Ewa, Mayne, Leland, Stukenberg, P. Todd, and Black, Ben E. 2024. “Structural basis for the phase separation of the chromosome passenger complex.” eLife

[9]: Englander, S. Walter. 2006. “Hydrogen Exchange and Mass Spectrometry: A Historical Perspective.” Journal of American Society of Mass Spectrometry