Molecular Grammar Governing Phase Behavior of Intrinsically Disordered Proteins with Prion-like Domains

Abstract

There is growing interest in uncovering the rules that govern the sequence-determinants of phase separation of intrinsically disordered proteins (IDPs). This has direct relevance for reversible formation/dissociation of membraneless organelles and other biomolecular condensates. Here, we focus on a subset of IDPs with prion-like domains (PLDs) whose phase behavior appears to be governed by a rather simple molecular grammar. To the first order, IDPs with PLDs can be reduced to a flexible polymer chain comprising multiple interacting motifs connected by spacers. The minimal motifs are individual amino acids, specifically tyrosine and arginine. We present a mean-field theory that predicts critical concentrations for phase separation as a function of Tyr/Arg motif numbers. The theory, which was tested in numerical lattice simulations, provides a first-order explanation for experimentally measured critical concentrations of a family of IDPs with PLDs. Interestingly, substitutions of Tyr with Phe or Arg with Lys residues drastically weakens the driving forces for phase separation, suggesting that the grammar cannot be simply reduced to the multivalency of cation-pi interactions. Instead, phase separation requires collective interactions amongst a multiplicity of Tyr/Arg motifs that implies involvement of a hierarchy of length scales and orientational effects. Finally, spacers interspersed between the Tyr/Arg motifs serve as secondary modulators of the driving forces for phase separation and as primary determinants of the material properties, specifically the fluidity of droplets. Here too, the choice of optimal spacers is non-random, with a decided preference for Gly residues, which promotes inter-chain overlaps that appear to promote collective interactions. The combined experimental and theoretical analysis enables proteome-wide predictions of critical concentrations of disordered regions with Tyr/Arg motifs and begins to open the door to designing sequences with bespoke phase behavior.