Triblock Proteins with Weakly Dimerizing Terminal Blocks and an Intrinsically Disordered Region for Rational Design of Condensate Properties

Abstract

Condensates are molecular assemblies that are formed through liquid–liquid phase separation and play important roles in many biological processes. The rational design of condensate formation and their properties is central to applications, such as biosynthetic materials, synthetic biology, and for understanding cell biology. Protein engineering is used to make a triblock structure with varying terminal blocks of folded proteins on both sides of an intrinsically disordered mid‐region. Dissociation constants are determined in the range of micromolar to millimolar for a set of proteins suitable for use as terminal blocks. Varying the weak dimerization of terminal blocks leads to an adjustable tendency for condensate formation while keeping the intrinsically disordered region constant. The dissociation constants of the terminal domains correlate directly with the tendency to undergo liquid–liquid phase separation. Differences in physical properties, such as diffusion rate are not directly correlated with the strength of dimerization but can be understood from the properties and interplay of the constituent blocks. The work demonstrates the importance of weak interactions in condensate formation and shows a principle for protein design that will help in fabricating functional condensates in a predictable and rational way.

Fundamental Aspects of Phase-Separated Biomolecular Condensates

“…To further obtain the fusion speed, one needs to determine the fusion times for a range of droplet sizes. The relation between τ fu and R is generally found to be linear (actually proportional;Figure e) (refs , , , , , , , , , , , , − , − , and − ). Inverse fusion speeds have been measured for many condensates (Table and Figure ).…”

“…The measured values range from ∼1 pN/μm for droplets of nucleophosimin 11 and PGL-3 312,318 to ∼250 pN/μm for droplets formed by mixing the recombinant mussel adhesive protein fp-151 (carrying a net positive charge) with hyaluronic acid (a polysaccharide) 315 and droplets of synthetic proteins with a triblock construct F-L-F, where F = a folded domain such as human γD-Crystallin or human fibronectin III domain10, and L = disordered linker. 316 Compared to the surface tension of ∼7 × 10 4 pN/μm at the water−air interface, the interfacial tensions of biomolecular condensates are two to five orders of magnitude lower. A simple reason for this enormous contrast in γ lies in the disparity in across-interface concentration changes (see eqs 41 and 49b): at the water−air interface, the change in water concentration is ∼55 M; at the interphase interfaces of condensates, the changes in macromolecular concentration are of the order of 10 mM.…”

“…311 A related method is based on using an atomic force microscope instead of OT to pull a microbead that is glued to the tip of a cantilever. 315,316 4.2.5. Experimental Data and Interpretations.…”

“…Interfacial tension has also been estimated based on the rupture force, i.e., the force that is required when a trapped microbead, pulled from inside a protein droplet, ruptures the interface and leaves the droplet . A related method is based on using an atomic force microscope instead of OT to pull a microbead that is glued to the tip of a cantilever. , …”

“…Interfacial tension has been measured for a relatively small number of condensates (Table and Figure ). The measured values range from ∼1 pN/μm for droplets of nucleophosimin and PGL-3 , to ∼250 pN/μm for droplets formed by mixing the recombinant mussel adhesive protein fp-151 (carrying a net positive charge) with hyaluronic acid (a polysaccharide) and droplets of synthetic proteins with a triblock construct F-L-F, where F = a folded domain such as human γD-Crystallin or human fibronectin III domain10, and L = disordered linker …”