2025 Theses Doctoral

Discerning the Influence of Cationic and Aromatic Amino Acid Identity on Biomolecular Condensates

The crowded cellular milieu contains a dry weight of up to 90 percent macromolecules, with proteins making up 20 to 30 percent of that mass. Through billions of years of evolution, eukaryotic cells have developed internal membrane bound organelles that further organize the crowded cytosol, and separate incompatible reactions and processes. This high level of organization lends itself to efficient nutrient uptake, energy production, cell to cell signaling, protein folding, and macromolecule recycling.

Although membrane-bound organelles have been widely studied since the mid 18th century, recent discoveries have uncovered the role of biomolecular condensates (BMC), also known as membraneless organelles (MBO), in a number of cellular functions. BMCs are defined by a dense region of macromolecules that spontaneously coalesce within cells. One of the chief phenomena that regulates condensate formation is liquid-liquid phase separation (LLPS), in which macromolecules will spontaneously demix from their environment and form highly dynamic, macromolecule-rich liquid droplets that are energetically separated from the surrounding dilute phase. The role of BMCs in RNA metabolism, mechanical and thermal stress responses, as well as heterochromatin packing in the nucleus has been widely studied. In addition, their aberrant role in a number of prion-related diseases has also been critically reported on.

At the same time, our understanding of the molecular interactions underpinning BMC behavior has also gained a large degree of interest, with wide ranging advancements in our ability to understand and predict the biomacromolecules that are able to engage with and form BMCs. Some of the more common structural motifs that are found to promote phase separation are large disordered domains, highly charged blocks separated by uncharged or nonpolar regions, and the ability to complex with oppositely charged polymers; primarily RNA. Nevertheless, our understanding of the role primary amino acid content confers on condensate dynamics remains poorly understood.

In this dissertation, I aim to uncover the specific role of aromatic and cationic side chain composition on BMC characteristics. To achieve these goals, I use in vitro assays composed of purified protein and RNA, as well as a BL21 strain of E. coli (NiCo21) as a model organism to study the impact of sequence-level cationic protein variants in a simple living system. I also created a panel of four engineered charge variant proteins (GFP(0) and GFP(+6)) derived from superfolder green fluorescent protein (sfGFP) with poly-lysine or poly-arginine tags appended to the C-terminus.

I first investigated discrepancies within the phase boundary of the engineered proteins in complex with RNA due to the charge and sequence variations. I next uncovered the impact of non-covalent interactions on these condensates, using NaCl, 1,6-hexanediol, and urea to understand the role of simple electrostatics, hydrophobic interactions, and hydrogen bonding, respectively. These studies revealed that although there are minimal changes in the initial phase boundary between arginine and lysine-based variants, there are significant increases to electrostatic interaction strength that is attributable to the change in amino acid composition. Hydrophobics played no role in the behavior of these engineered condensates and hydrogen bonding played an equivalent role across all constructs.

I also found that the in vitro results were quantifiable and replicated in the E. coli experiments. I found that there was a higher ratio of protein in the condensate compared to the cytoplasm in GFP(+6) and arginine tagged variants compared to GFP(0) and lysine tagged variants. In addition, the percent of cells with condensates was found to increase along the same hierarchy. Using an RNA hairpin-specific ligand, I was able to observe mRNA co-localizing with GFP at the poles of cells that form condensates, while DNA was excluded and relegated to the midcell. Using Fluorescent Recovery After Photobleaching (FRAP) on cells with condensates, it was revealed that arginine-based condensates have a much slower recovery rate, and a smaller fraction of mobile protein, showing that protein mobility within the dense phase is lowered in arginine-based condensates.

When expanding the scope from engineered condensates to endogenously phase separating proteins, I once again found that arginine promotes increased condensate formation relative to lysine. I also provide preliminary evidence that of the three aromatic amino acids (phenylalanine, tryptophan, and tyrosine), tryptophan promotes condensate formation in biological systems at a higher rate than the other two amino acids, and results in an overall reduction in protein mobility in the dense phase.

In this dissertation, I provide evidence for rules to further understand the chemical grammar of BMC formation. These studies lay the foundation for the prediction of protein likely to engage with BMCs, and also provides crucial information for the design of synthetic membraneless organelles.