| Abstract |
Orderliness, speed, and rhythm in biochemistry are vital for cellular function. In order to achieve this, cells implement compartmentalization via several methods, one of which is the formation of membrane-less compartments. These compartments, often referred to as “biomolecular condensates”, are understood to be formed by separation of proteins and other biomolecules into dense and dilute phases. While the formation of the resulting protein-rich condensates is fundamental for spatiotemporal organization of biochemistry within the cell, a vast majority of proteins found to phase separate in vitro do so at a concentration an order of magnitude above their endogenous expression levels. Recently, a theoretical study has shown that membrane binding of phase separating proteins can result in phase separation spatially occurring at the membrane well below bulk saturation concentrations. However, much remains unknown about the formation mechanism and function of these condensates. To that end, for my doctoral project, I used a synthetic system composed of supported lipid bilayers decorated with lipid-bound NTA(Ni) to allow for coordination and thus membrane binding of the well-characterized protein FUS via a C-terminal His-tag. Through this model system I found that 2D phase separation of FUS could occur an order of magnitude below the experimentally determined bulk saturation concentration. FUS was able to form dense and dilute phases in 2D and the transition point to form these phases could be controlled by modulating buffer conditions. Additionally, membrane-bound FUS condensates were able to further recruit FUS from the bulk to form a multilayer of protein through a prewetting transition. With this characterization of 2D phase separation of FUS, I then explored a physiologically relevant protein in the form of PAR-3, a fundamental protein of the PAR polarity system, which is necessary for the establishment of polarity in the C. elegans zygote. I found that full-length PAR-3 was able to phase separate under physiological salt conditions with a Csat of 100nM. Further, I identified a C-terminal predicted prion-like domain to act as a driver for phase separation. Additionally, I determined PAR-3’s affinity and specificity for PI(4,5)P2 and found that it could form 2D condensates upon binding to the membrane at physiological concentrations. Furthermore, these condensates were able to recruit PAR-6 alone and PAR-6 in complex with PKC-3 to the membrane, ultimately resulting the reconstitution of the anterior PAR complex which is known to exist in a condensed clustered form in vivo. Taken together, this work provides insight into a mechanism where phase separation can be locally triggered by membrane binding under sub-saturation concentration, offering a robust and potentially universal mechanism by which cells can spatially control phase separation and pattern cellular membranes. |