Get ready to dive into the fascinating world of biomolecular condensates! These unique structures are formed by the self-assembly of disordered proteins with low-complexity sequences, and they play a crucial role in cellular organization. But here's where it gets controversial: the aromatic amino acids phenylalanine and tyrosine, known as 'stickers', drive the cohesion of these dense phases, but their relative strengths have been a subject of debate.
Recent studies suggest that tyrosine is the stronger 'sticker', but this contradicts conventional chemical intuition based on hydrophobicity scales and pairwise contact statistics. So, what's the truth?
In this groundbreaking work, we use molecular dynamics simulations and quantum chemistry calculations to unravel this mystery. By studying simple model peptides and side-chain analogues, we've discovered that the hierarchy of sticker strength arises from the lower free energy of transfer of tyrosine into the condensate. This is influenced by both stronger protein-protein interactions and solvation effects.
But here's the twist: as the dielectric constant of the surrounding media approaches that of an apolar solvent, phenylalanine takes the lead as the stronger sticker. This finding highlights the crucial role of the chemical environment in modulating protein-protein interactions, providing a clear explanation for the crossover in sticker strength between these two amino acids.
For decades, cellular organization was primarily understood through membrane-bound compartmentalization, but membrane-less organelles, like stress granules and nuclear speckles, have been gaining attention. These organelles form through phase separation, primarily driven by disordered regions of proteins and nucleic acids. The resulting condensates exhibit liquid-like properties, such as fusion and dripping.
Proteins undergoing phase separation often share common characteristics, including deviations from typical folding compositions and simple sequences with amino acid repeats. These sequences have been termed 'stickers' and 'spacers', with spacers acting as flexible linkers and stickers determining single-chain properties and phase behavior.
Recent experiments have quantified the influence of different stickers, leading to a crucial question: what determines their relative strengths? This matter has been investigated in the context of cationic amino acids, and now we turn our attention to the aromatic residues tyrosine and phenylalanine.
Given their structural similarity, one might expect them to have equal relevance in phase separation. However, experimental evidence suggests that tyrosine is a stronger driver of condensation. This is observed in mutants of FUS, LAF-1, and hnRNPA1, where tyrosine-to-phenylalanine substitutions reduce phase separation propensity.
Understanding the origin of these differences is crucial, as the fractions of phenylalanine and tyrosine in hnRNPA1 variants co-vary, suggesting evolutionary control of condensate properties.
The molecular properties of these amino acids provide valuable insights. Hydrophobicity scales typically rank phenylalanine as the most hydrophobic, consistent with its greater hydration free energy compared to tyrosine. However, using hydrophobicity as a proxy for interaction energy in simulation models proved insufficient to explain their relative strengths.
Statistical contact matrices also failed to capture the correct order of stickiness. On the other hand, potentials of mean force calculated with atomistic force fields show a deeper free energy well for the tyrosine-tyrosine pair. Additionally, tyrosine's extremely low solubility suggests stronger hydrogen bonds in the protein interior.
In summary, experimental results on condensates containing phenylalanine/tyrosine variants align with atomistic force field calculations and solubilities in water, but not with solvation free energies or hydrophobicity scales. This paradox is addressed using a combination of molecular dynamics simulations and quantum chemical calculations.
We estimate transfer free energies of peptides with these aromatic residues into model peptide condensates, finding that they are more favorable for tyrosine. However, this trend reverses in apolar media, where phenylalanine becomes the stronger sticker. DFT calculations confirm stronger interaction energies in tyrosine-tyrosine pairs, but the transfer free energy contribution dominates at low dielectric constants.
Our findings provide a clear explanation for the crossover in sticker strength between tyrosine and phenylalanine in different media, highlighting the role of the chemical environment.
To investigate this further, we performed classical molecular dynamics simulations and quantum chemical calculations. We studied terminally-capped GGXGG peptides with X=F/Y in various media, including water, organic solvents, and peptide condensates. We estimated transfer free energies using an alchemical transformation, finding that transfer into peptide condensates is more favorable for the tyrosine-containing peptide.
This result is consistent with experimental evidence suggesting a greater propensity for tyrosine to form condensates. We also ran simulations of isolated dense phases for gly/ser/tyr and gly/ser/phe condensates, finding similar interaction patterns dominated by aromatic residue contacts.
However, the saturation density and critical temperature were slightly lower in the gly/ser/tyr condensate, indicating a stronger propensity for phase separation.
To explore the role of protein-solvent interactions, we analyzed the distribution of water molecules around aromatic side chains in the dense phase simulations. We found notable differences, with a prominent peak in the radial distribution function for water oxygen around the tyrosine Cζ due to hydrogen bonding.
To investigate the effect of different environments, we performed calculations using a range of solvents with diverse properties. The results showed a dependence on the dielectric constant, with non-polar solvents favoring phenylalanine and polar solvents favoring tyrosine. Quantum calculations confirmed a crossover in interaction strength, with phenylalanine favored in non-protic aliphatic solvents and tyrosine in alcohols and high dielectric environments.
Our work reconciles the apparent contradictions between phenylalanine's lower solvation-free energy and stronger hydrophobicity, and tyrosine's greater contact energies and sticker strength in biomolecular condensates. The crucial factor in their overall propensity for contact formation across different environments is the transfer free energy.
Our findings align with a recent report emphasizing the role of transfer free energy in biomolecular condensates. While our reductionist approach has limitations, such as considering very short peptides and a minimal alphabet of interaction pairs, our results are consistent with those found by other authors.
The methodology proposed in this work can be extended to other types of interactions, and the overall qualitative conclusions on the crossover in interaction strengths as a function of the environment are likely to be maintained. The context-dependence of interaction strengths is related to the concepts of 'pair' and 'bulk' hydrophobicity, and our work stresses the importance of considering both the pair contribution and the transfer free energy contribution.
This research has been financed by grants and supported by various institutions, and the authors gratefully acknowledge conversations with colleagues that inspired this work. The methodology and findings presented here provide a deeper understanding of the complex interactions within biomolecular condensates and highlight the importance of considering the chemical environment in modulating protein-protein interactions.
