Session Y38: Biomolecular Condensates II - Dynamics, Properties, Design

A physics-based model for predicting temperature-dependent phase behavior of proteins

When certain proteins inside the cell are subject to temperature perturbations, they phase separate to form condensates, such as stress granules and P-bodies. Some proteins exhibit upper critical solution temperature (UCST) behavior, forming condensates at low temperatures, whereas others exhibit lower critical solution (LCST) behavior, condensing at higher temperatures. While experiments are able to capture this behavior at a microscopically visible level, elucidating the full biophysics of temperature-driven condensate formation requires sub-molecular insight. Computational approaches, specifically coarse-grained molecular dynamics simulations have proven useful in understanding molecular underpinnings of condensates. However, while some approaches have recapitulated UCST behavior, capturing LCST behavior quantitatively has been a challenge, as this requires models to carefully account for differences in solvation with temperature. We have developed a quantitatively accurate residue-level coarse-grained model to predict temperature-driven protein condensation. By parametrizing our model with atomistic simulations and validating with experimental data, we are able to capture full phase behavior of disordered proteins. This work has direct implications for uncovering biophysics of temperature-dependent phase separation in disordered proteins. Furthermore, our model expands the set of tools that can be used to engineer thermoresponsive biopolymers applications in smart nanomaterials.   

Investigating the molecular driving forces of liquid-liquid phase separation in intrinsically disordered proteins

It is increasingly recognized that liquid-liquid phase seperation (LLPS) serves as a fundamental mechanism employed by cells to establish transient spatiotemporal organization of the cytoplasm which is achieved through the formation of biomolecular condensates. The molecular forces driving this intricate process, predominantly involving intrinsically disordered proteins (IDPs) in the cell, remain inadequately characterised.  Hence, an enhanced understanding of these determinants could offer insight into cellular organisation, as well as shedding light on the pathogenesis of protein condensation diseases.

We have investigated these molecular drivers using a model for intrinsically disordered proteins. By experimentally mapping the LLPS phase boundary, we have shown that the binodal is sensitive to both changes in protein structure and protein-protein interactions. Hence, we shed light on the cooperative relationship between entropic and enthalpic components of the free energy which drive the phase separation of IDPs.   

Modeling phase separation of proteins with compact native states using LatticePy, a package for MCMC simulations on lattices

Liquid-Liquid Phase Separation (LLPS) of proteins has proven to be a ubiquitous feature inside cells, playing critical roles in many biological pathways. While phase separation is commonly observed for intrinsically disordered proteins, predicting if and when well-folded proteins may also undergo LLPS is an open problem. To begin to address this knowledge gap, here we provide the first open-source scalable, replicable, and easy-to-use package, LatticePy, to simulate oligomerization and phase separation for any given lattice-protein sequence. Implementing standard moves for lattice proteins and simple energy functions, we explore the conditions under which proteins with a well-defined, compact native state instead adopt unfolded conformations as part of multiprotein oligomers or condensates.   

Multiscale approaches to model the functions of Ubiquitin and SUMO modifications in protein phase separation

The post-translational modifiers such as mono and poly ubiquitins and SUMOs are known for their ability to modulate protein–protein interactions by becoming covalently attached to other target proteins. Despite the high similarity in the tertiary structure and sequence, they differentially influence the target protein properties. In this work, we employed a multiscale simulation approach that encompasses atomistic to different level coarse-grained modeling techniques with data-driven machine-learning methods, to explore the structural differences and multidimensional energy landscape of ubiquitin and SUMO and their conjugates. We finally study the influence of distinct features of the targets and modifiers on protein phase separation and aggregation, providing molecular-level insight into the corresponding in vitro measurements and instructing further experiments through adjustment of relevant parameters.   

Thermodynamics and Transport Properties of Pure RNA condensates

Pure RNA systems can form  condensates both in vitro and within cells. Such RNA condensation has been associated with disease states. Additionally, RNA assembly via condensation creates  opportunities for new soft material design. However, the biophysical principles governing RNA condensation  remain largely unexplored, as  it is still challenging to characterize RNA–RNA interactions and the resulting RNA condensates. To this end , building on previous work, we have implemented and optimized a model for probing RNA condensation. Significantly, our implementation achieves a  dramatic computational speed up. Exploiting this  feature, we probe the long-time dynamics of RNA condensates Furthermore, we characterize the effects of primary RNA sequence on condensate thermodynamics. Altogether, our work provides new physicochemical codes that govern the material properties of pure RNA condensates.    

RNA-driven Percolation Transitions in Biomolecular Condensates

Phase separation of protein and RNA complexes in living systems is driven by a complex interplay of homotypic and heterotypic interactions mediated by proteins, structured, and unstructured RNAs. We recently reported that RNAs are intrinsically poised to undergo an entropically-driven phase separation coupled to an enthalpically-driven percolation, where a system-spanning networking transition in the dense phase leads to their dynamical arrest and hysteretic phase behavior. Here, we investigate RNA-driven percolation transitions in the context of multicomponent protein-RNA condensates that closely mimic intracellular ribonucleoprotein (RNP) granules. We show that naturally occurring RNAs that form G-quadruplex structures drive an age-dependent liquid-to-solid transition of RNP condensates. Upon physical aging, the RNP condensate fluid network undergoes dramatic rearrangements and features a dynamically arrested viscoelastic solid core surrounded by a terminally viscous fluid shell. The timescale of the RNA-driven percolation transition of RNP condensates is tuned by the RNA chain length, valence, and mutations that modulate the stability of RNA G-quadruplex. Utilizing GC-rich RNAs that are associated with repeat expansion disorders, we further show that RNA-driven percolation transition universally engenders dynamical arrest of RNP condensates in a non-functional state. We posit that this aberrant age-dependent phase transition of RNA-protein condensates can be counteracted by factors and additives, such as ATP-dependent RNA binding proteins that can modulate RNA base pairing and base stacking interactions in the dense phase. The ability of RNA chaperones to alter an RNA’s phase separation and percolated network formation may provide a new lens to view their roles in modulating the physical aging of intracellular biomolecular condensates that are central to RNA biology.

Active remodeling of RNA interactions facilitates nonequilibrium steady states of protein-RNA condensates

Biomolecular condensates are membraneless organelles which organize biochemical processes within the cell. Many biomolecular condensates contain RNA and are involved in RNA metabolism. Base-pairing interactions of RNA can contribute to the normal material and dynamic state of a particular biomolecular condensate or drive a transition of the condensate into a solid-like or aggregate-like state. RNA base-pairing interactions can be tuned and remodeled by DEAD-box helicases, a class of proteins commonly found associated with biomolecular condensates, in an energy-dependent manner. In this work, we assess the effects of active RNA remodeling by a DEAD-box helicase on condensate structure, composition, and phase behavior. Through modulating enzyme activity and changing RNA substrate identity, we tie helicase activity to RNA partitioning, time-dependent condensate morphology, and time-dependent condensate dynamics. We also show that helicase activity can be leveraged to produce secondary RNA phase transitions, suggesting that DEAD-box helicase-RNA systems may offer a new frontier in active matter research.   

Modulating condensate properties and catalytic activity with peptide sequence in coacervate dispersions

Condensation of biopolymers gives rise to distinct aqueous microenvironments that are proposed to help organise cellular as well as prebiotic (bio)chemistry. In addition to the spatial localization of reactants within these biomolecular condensates, reactions may also be modulated by the local physical properties of the condensate microenvironment. These properties in turn emerge from the sequence-encoded interactions of the molecular building blocks, which are often polypeptides and nucleic acids. To simultaneously explore both the potential influence of droplet physical properties on biochemical reactions in demixed systems as well as the control of those properties through polypeptide sequence, we use polyelectrolyte complex coacervate microdroplets prepared from an RNA enzyme and a small library of prebiotic peptide sequences as a minimal model system. Using quantitative phase microscopy and thermodynamic tie-line analysis, we measure the precise molecular composition of these condensates and find that that the phase behaviour is strongly influenced by peptide net charge and polar residue content. We show that small variations in peptide sequence can tune rates and yield of the ribozyme up to 15 times. Furthermore, we show that ribozyme rate constant is anti-correlated to the ribozyme concentration and correlated to the diffusion coefficient within the coacervates. Taken together, these results are a step towards linking polymer sequence to condensate properties and downstream catalytic function.   

Functional membraneless organelles formed by de novo proteins

Protein phase transitions in biology are associated with disease pathogenesis, for example in cataract disease or sickle cell anaemia, but they also occur as part of normal biological processes, such as liquid-liquid phase separation in cells. While protein phase behaviour has been understood from a physics perspective for globular proteins for some time, how we can apply what we already know to biomolecular condensation, or the phase transitions of other protein types is less clear. In my talk, I will describe the design of genetically encoded de novo polypeptides that form membraneless organelles in Escherichia coli and how the designed polypeptide can co-compartmentalize a functional enzyme pair, to create functional membraneless organelle with product formation close to the theoretical limit.

    

Synchronization of chemical reactions in a population of phase-separated droplets

Collective behaviour refers to the actions and interactions of a group of individuals, which results in emergent patterns and behaviour that cannot be explained by individual actions alone. Examples of this emergent behaviour from complex systems are widespread in physics, ecology and biology and include phase transitions in materials and ant or bee colonies displaying swarm intelligence. How is this possible? A method of communication is universally required for a complex system to exhibit collective behaviour. In this project, we explore whether biomolecular condensates formed via liquid-liquid phase separation could act as a means for collective behaviour to emerge within a cellular environment to enable population-level control of chemical reactions relevant to complex biological processes.

 

If such processes are regulated at the cellular scale by condensates, how is communication possible between spatially distinct condensates? How does this coordinate the behaviour of multiple droplets, resulting in a more predictable and stable outcome of chemical reactions, with specific functions, at the population level? As the mechanisms by which communication and population-level regulation may be possible for in vitro/vivo systems that form droplets have not been explored, this is an exciting opportunity to generate a novel understanding of how complex processes are regulated at the cellular scale. Here we have demonstrated that communication via the dynamic exchange of materials to maintain partition concentrations between the dense phases of individual droplets and the surrounding dilute phase is able to efficiently synchronise chemical clock reactions occurring within the droplets.   

Diffusiophoresis promotes phase separation and transport of biomolecular condensates

The internal microenvironment of a living cell is heterogeneous and comprises a multitude of organelles with distinct biochemistry. Amongst them are biomolecular condensates, which are membrane-less, phase-separated compartments enriched in system-specific proteins and nucleic acids. The heterogeneity of the cell engenders the presence of multiple spatiotemporal gradients in chemistry, charge, concentration, temperature, and pressure. Such thermodynamic gradients can lead to non-equilibrium driving forces for the formation and transport of biomolecular condensates. Here, we report how ion gradients impact the transport processes of biomolecular condensates on the mesoscale and biomolecules on the microscale. Utilizing a microfluidic platform, we demonstrate that the presence of ion concentration gradients can accelerate the transport of biomolecules, including nucleic acids and proteins, via diffusiophoresis. This hydrodynamic transport process allows localized enrichment of biomolecules, thereby promoting the location-specific formation of biomolecular condensates via phase separation. The ion gradients further impart active motility of condensates, allowing them to exhibit enhanced diffusion along the gradient. Coupled with reentrant phase behavior, the gradient-induced active motility leads to a dynamical redistribution of condensates that ultimately extends their lifetime. Together, our results demonstrate diffusiophoresis as a non-equilibrium thermodynamic force that governs the formation and active transport of biomolecular condensates.   

Buffering of biomolecular condensate volume and mass in multi-component mixtures.

Cells exhibit significant variations in the concentration of biomolecules, such as proteins and RNA, despite possessing identical genomes. These variations result from intrinsic fluctuations in the expression rates of proteins and RNA, as well as in the expression rates of molecules involved in transcription and translation, such as polymerases. Recently the formation of biomolecular condensates through liquid-liquid phase separation (LLPS) has been proposed as a powerful buffering mechanism to enable local protein concentration to be robust to noise. Here, using theory, simulations, in vitro and in vivo experiments we show orthogonal noise buffering mechanisms, whereby the total volume or mass of condensates is buffered when the global protein concentration changes. Our work provides new insights into the potential role of biomolecular condensates in functional biology.   

Physical Mechanisms of Protofilament Formation

Protein misfolding and aggregation are associated with the onset of neurodegenerative disorders such as Alzheimer’s and Parkinson’s disease. To date no cure exists for neurodegenerative diseases and therapeutic interventions give limited symptomatic relief, rather than prevention. Prefibrillar aggregates (oligomers) are associated with neurotoxicity. A better understanding of the physicochemical properties that govern the assembly mechanism of the early oligomeric species will aid in understanding their role in toxic propagation. 

Here, we introduce a novel coarse-grained model for amyloidogenic polypeptides [1] and use Brownian dynamic simulations [2] to gain insight into the physical mechanisms of assembly into oligomeric species. We find that spherical and elongated oligomers form via different mechanisms. The first are governed by hydrophobic interactions and grow mainly via monomer addition, while the latter additionally undergo structural modifications, frequent breakage and hierarchical assembly. Our results are complemented by atomic force microscopy measurements and explain experimental observations [3].

References

[1] B. Mayer et al. in preparation

[2] I.M. Ilie et al, J. Chem. Phys., 142, 114103 (2015)

[3] F.S. Ruggeri et al. PNAS, 115, 7231(2018)   

Developing methods to classify protein clusters as liquid-liquid phase separated biomolecular condensates in vivo.

Recently, it has been discovered that liquid-liquid phase separation (LLPS) occurs in the cell, driving the demixing of specific biomolecules to form concentrated droplets called condensates. LLPS is now believed to play a role in many key cellular processes including gene regulation. To test this hypothesis, we are studying the transcription factor Bicoid (Bcd) in fly embryos which has structural properties important for LLPS. Bcd also forms a spatial concentration gradient along the embryo which governs a spatial pattern of gene expression and provides a system to study the effect of concentration on condensate formation in vivo. By creating a dedicated confocal image acquisition and analysis workflow, Bcd clusters have been detected and shown to be quantitatively distinguishable from a non-clustering control protein. The diffusion coefficients of these clusters were also measured using particle tracking algorithms, indicating the existence of two populations, and suggesting that some of these clusters are bound to DNA. These results are supported by fluorescence correlation spectroscopy experiments that allow estimating the average number of Bcd molecules present in these clusters, while super-resolution Airyscan imaging enhances the ability to capture the formation and breaking down of these structures.