Sequence determinants of intracellular phase separation by complex coacervation of a disordered protein
Mol. Cell 63, 72 - 85.
Pak, C.W., Kosno, M., Holehouse, A.S., Padrick, S.B., Mittal, A., Ali, R., Yunus, A.A., Liu, D.R., Pappu, R.V., and Rosen, M.K.
This paper is actually one of my favourites from my graduate work so far. The Rosen lab discovered (totally by chance) that when they expressed the disordered C-terminal domain of the Nephrin Intracellular Receptor (NICD) in cells they formed liquid like droplets that fuse, flow and undergo dynamic and rapid internal re-arrangement as measured by FRAP. However, this protein was completely soluble in vitro, raising the question of how and why does it phase separate in cells?
NICD has the sequence characteristics of a negative polyelectrolyte, meaning it has a net negative charge. We reasoned that phase-separation may be being driven by complex coacervation, where two highly but oppositely charged polymers are soluble independently, but once they are mixed together they undero complexation to form a phase separated state (liquid or solid) due to charge neutralization and an entropic effect. To test this hypothesis we ran extensive simulations of pairs of NICD chains with and without multiple poly-lysine peptides. In the absence of poly-lysine the NICD chains were repulsive for one another, but in the presence of poly-lysine the NICD and poly-lysine undergo complex coacervation to form a dynamic assembly of NICD and poly-lysine. This result was verified experimentally, and using supercharged GFP developed by David Liu the charge on the GFP was varied to determine what the threshold opposite charge needed to drive complexation was.
Given the lab's previous work on the impact of charge patterning on the conformational behaviour of IDPs , we wondered if generating sequence permutants of NICD that change the charge patterning but hold composition fixed might change the driving force for phase separation. We designed a series of sequences that shuffled the charged residues around, and found that the extent of 'blocky-ness' in terms of the charged residues had a profound impact of the driving forces for complex coacervation both in vitro and in vivo.
Finally, we performed an extensive set of statistical analysis coupled to a huge set of deletion, shuffle, and insertion experiments to dissect what the molecular determinants of phase separation are in terms of the specific residues and residue motifs. We found that charged and hydrophobic residues appear to be important, while polar residues in NICD seem much less critical. Interestingly, the two most important residues were tyrosine and arginine.
This was a massive body of work from a number of different people done over almost two years, and working with the Rosen lab was an absolute pleasure. It is the first time that complex coacervation was demonstrated to drive phase separation in a naturally occurring protein inside the cell, but a large proteome-wide analysis we performed suggests this may be a general mechanism for driving self-assembly in the cytoplasm and on the membrane.
 Das, R.K., and Pappu, R.V. (2013). Conformations of intrinsically disordered proteins are influenced by linear sequence distributions of oppositely charged residues. Proc. Natl. Acad. Sci. U. S. A. 110, 13392 - 13397.