Session Y06: Sub-cellular Structures: Droplets and Assemblies

Single molecule force spectroscopy of telomeric H3.3 variant nucleosomes

Understanding how histone protein variants affect the dynamics of nucleosome unwinding is key for developing a mechanistic model of their role in chromatin maintenance and disease. Nucleosomes are the basic units of chromatin, consisting of 146 bp of DNA wrapped around an octamer made up of two of each of the four core histones—H2A, H2B, H3, H4. Nucleosomes containing histone H3.3, a variant of the canonical histone H3.1, are enriched in telomeric regions. Mutations in the pathway responsible for depositing H3.3 nucleosomes on telomeres are associated with the Alternative Lengthening of Telomeres (ALT) pathway. The ALT pathway allows cells to overcome senescence induced by telomere shortening, a key protection against cancer. There is currently no mechanistic understanding of the role of the H3.3 variant in the ALT+ cancer pathway. To address this gap, we measure the kinetic unwinding rates of single H3.3 or H3.1 nucleosomes assembled on either the Widom 601 or human telomeric DNA sequences using single molecule optical tweezers. Our studies quantify the role of the H3.3 variant in telomeric chromatin maintenance.   

Effects of macromolecular crowding on protein-mediated looping in DNA

The looping of DNA is a phenomenon that is ubiquitous throughout all kingdoms of life, and has been studied extensively under laboratory conditions. However, the interior of cells is crowded. Currently, there is no method to directly observe the looping of a single DNA molecule inside a cell, and it is exceedingly difficult to stably change the concentration of crowders in a living cell. By adding macromolecular crowders to a microchamber we are able to get a better picture of how DNA behaves inside the crowded cell. We have investigated the crowding effects of Dextran 70 (Dx70), as well as bovine serum albumin (BSA), at various concentrations, on protein-mediated DNA looping through tethered particle motion experiments, using 831 bp-long DNA and the lac repressor protein (LacI). We found that the probability of LacI-mediated DNA looping is reduced by Dx70 which is consistent with the notion that Dx70 stiffens DNA. We also found that a larger concentration of Dx70 was required to decrease looping probability at 200 mM KCl than were necessary for 100 mM KCl. Instead, BSA increased looping, unless Mg2+ was added which decreased looping probability. These observations show that looping is affected strongly by the choice of crowder and ionic conditions.   

Simplifying Protein Collective Vibrational Assignment Through Symmetry

Anisotropic terahertz microspectroscopy (ATM) isolates protein collective vibrations from a highly congested spectrum based on the direction of the vibrational transition dipole.  ATM changes with inhibitor binding and photo excitation reveal that the vibrational bath evolves with functional state.  Assignment of the spectral bands to specific structural motions is needed to understand the impact of these dynamical changes.  The complexity of the anisotropic spectral structure is dependent on the alignment of the molecules.  This alignment is achieved by crystallization, which results in biomacromolecular arrays along with 30-70% water by volume.  We examine crystal symmetry effects on the anisotropic spectra for hen egg white lysozyme crystals (HEWL).  ATM measurements of triclinic, monoclinic and tetragonal HEWL crystals identify the conserved versus unique spectral features. Peaks near 40 cm⁻¹ and 55 cm⁻¹ were common among three lattice systems while a peak near 20 cm⁻¹ was observed only in triclinic CEWL. Further we find that the spectral structure decreases with increasing symmetry, an expected result from dipole cancellation.  The spectral symmetry shows good agreement with normal mode ensemble analysis calculations providing initial assignment of the bands.   

Regulation of gene expression by transcription factor clustering

In developmental systems, like the fruit fly embryo, highly reproducible spatial patterns of gene expression are formed within minutes. This requires fast and precise interpretation of the nuclear concentrations of input transcription factors (TFs) by the regulatory sequence of the target genes. However, these factors, which are freely diffusing proteins in the nucleus, have previously been shown to be heterogeneously distributed. Here, we show that this apparent heterogeneity arises out of accumulation of the TFs into submicron clusters. Using several target genes of a prominent TF, we show that active gene loci coincide with these TF clusters. The physical properties of the clusters and their stability are highly dependent on the DNA regulatory sequences of these target genes, but not on the nuclear concentration of the TF proteins. These results suggest that such cluster formation might confer robustness against input signal fluctuations.   

Optical detection of structural alteration of brain tissues in progressive Alzheimer’s diseases using partial wave spectroscopy (PWS)

 

Alzheimer disease (AD) is the most common form of dementia, characterized pathologically by amyloid plaques and neurofibrillary tangles. One of the earliest overt signs of AD is a loss of cognitive function.  It has been reported that AD affects the nanoscale structure of the brain cells which begins long before the cognitive symptoms appear. However, these alterations are undetectable in the initial stages with currently used bulk diagnostic techniques such as MRI and OCT. Thus, the elucidation of a neuroimaging method that can uniquely characterize these structural disorders at nanoscale is imperative for clinical diagnosis. Recently developed finer-focused partial wave spectroscopy (PWS) is a sensitive technique for probing nanoscale structural alterations in cells/tissues in terms of the average structural disorder strength. Results of PWS technique measurements of brain tissues from an animal model and human subjects show significant increase in the disorder strength with the progression of AD relative to their controls. The increased disorder strength can be explained by the higher mass density fluctuations caused by the rearrangements of macromolecules due to the deposition of the amyloid beta protein and damage in DNA with progress of AD. This structural alteration may be reflected in other clinical parameters for early detection of AD and possible treatment.

 

 

 

 

    

Multiscale Molecular Dynamics Simulations of Amyloidogenic Protein Binding and Folding on Lipid Rafts

Amyloidogenic proteins, e.g., beta-amyloid, are a class of proteins that do not have a fixed three-dimensional structure. These proteins rapidly aggregate in solution and on membrane surfaces.  At present, very little is known about the binding behaviors and disordered-to-ordered transition of amyloidogenic protein aggregates on phase-separated lipid nanodomains, which mimic the structures of biological membranes. Using multiscale (atomistic and coarse-grained) simulations, we explore the structure, dynamics, and energetics of beta-amyloid aggregates of various sizes on highly dynamic and heterogeneous lipid nanodomains, or lipid rafts, at microsecond timescales with atomistic resolution. Our lipid rafts are composed of phase-separated cholesterol-enriched liquid-ordered (Lo) domains surrounded by liquid-disordered (Ld) regions. We are interested in the preference of beta-amyloid binding to the Lo, Ld, or Lo/Ld interfacial domain, and the subsequent surface-induced templated folding of protein on nanostructured surfaces. This work will allow us to have a better understanding of the physics of protein interactions and folding on nanostructured surfaces, and propose new protein structures for future anti-aggregation drug development to treat amyloid diseases.   

Anti-apoptotic condensates of extremotolerance-related proteins

Phase-separated macromolecular assemblies have long been implicated in biological tolerance to environmental stress, from prototypical stress granules to the gel-like or glassy states thought central to achieving anhydrobiosis. Many extremotolerance-associated (ExTol) proteins exhibit features like intrinsic disorder and/or repetitive sequence, which are also often hallmarks of proteins that undergo liquid-liquid phase separation. We screened a library of proteins exhibiting these features, which included many ExTol proteins but also non-ExTol and even de novo designed proteins, for their ability to protect human cells against chemically-induced apoptosis. In addition to identifying protective proteins, we observed that their anti-apoptotic effect can be enhanced by promoting phase separation through modification with inducible oligomerization domains (e.g., FKBP). One of the condensates examined, driven by a synthetic protein, exhibited Caspase-7 partitioning, thereby suggesting effector sequestration as a possible mechanism for condensate-mediated apoptosis protection.   

Receptor-mediated signaling through a prewetting phase-transition

Membrane-bound receptors sense ligands at the plasma membrane, and must amplify and communicate the signals they receive into the three-dimensional cytoplasm. Proteins involved in these signaling cascades often form "signaling clusters" upon activation: structures rich in many of the 3D components required for downstream signaling, as well as specific lipids and membrane bound proteins. Recent work has suggested that liquid-liquid phase separation underpins many signaling clusters. Here we develop a theory of these phenomena through Monte-Carlo simulations and a minimal Landau theory. Our model suggests that signaling clusters are best described as prewet, displaying surface-only phase coexistence outside of the ordinary membrane and bulk protein coexistence regions[1]. Prewet phases confer the ability to 'integrate' the state of the bulk and membrane concurrently. We explore the ramifications of our model wherein signal initially received at many receptors is integrated and transduced by the dynamics of the prewetting transition.   

Interface resistance of biomolecular condensates

Cells possess a host of biomolecular condensates, many of which exhibit dynamic exchange of condensate components with their surroundings. Such material exchange is key to condensate functions, especially for those acting as microreactors. While standard theory suggests that the exchange rate is limited by either external influx or internal mixing, recent experiments showed, surprisingly, that the rate can be limited by the dynamics of molecules at the droplet interface, termed interface resistance. Here, we combined theory and coarse-grained simulations to uncover the physics that gives rise to interface resistance. We first derived an expression for the exchange rate and identified parameter regimes in which the exchange is governed by external, internal, or interface dynamics. We then validated the results via simulating exchange between dilute and dense phases of "sticker-spacer" molecules. Importantly, we found that interface resistance is caused by incident molecules dwelling at the interface without entering the dense droplet. Our work highlights the underappreciated role of interface resistance on condensate exchange dynamics.   

Non-specific adhesive forces reorganize the cytoskeleton around membrane-less organelles

Phase-separation of biomolecules in cells takes place in a complex environment crossed by multiple filaments of the cytoskeleton or chromatin. Here, we study the interactions of stress granules, a phase-separated protein-RNA droplet in the cytosol, with the heterogeneous networks of the cytoskeleton. In the case of the microtubule network, for example, statistical tools similar to the radial distribution function allow for us to quantify long-ranged enhancement in filament density in the vicinity of stress granules. When microtubules are depolymerized, the molecular subunits partition to the surface of the droplet. We interpret the data using a thermodynamic model, revealing a weak non-specific affinity of the subunits to the surface of about 0.1 kT. As filaments polymerize, the adhesion is amplified leading to significant adhesion of filaments to the granule surface, making microtubule rich regions of the cell energetically favorable for stress granules.   

Pattern formation of liquid condensates at surfaces

In living cells, the cytoplasm or the nucleoplasm is confined by membranes separating their chemically distinct inside from their environment. Specific proteins can not only bind to such membranes but can also undergo phase separation in the cytoplasm or nucleoplasm, respectively. These properties suggest an assembly pathway for phase-separated, protein-rich condensates at membranes that are controlled by membrane binding. Interestingly, in living systems, these binding processes are typically regulated by active chemical reactions which involve irreversible steps such as phosphorylation. Here, we theoretically investigate how active binding processes control the liquid condensates using irreversible thermodynamics and active binding kinetics that breaks the detailed balance of the rates. We found that these active binding processes lead to non-spherical droplets wetting on the surface. Moreover, the ripening of condensates is suppressed causing various stationary patterns composed of multiple equally-sized droplets. Strikingly, for small systems or molecules that can be sufficiently densified on the membrane, we report the emergence of condensates continuously oscillating between two opposite membranes. Our findings indicate that active binding in the bulk-surface coupled system can facilitate temporal and spatial control of liquid condensates, which might provide insights into the tempo-spatial organization of membrane-less organelles in the living cells.   

Nanomechanical crowding at the interface between RNA and soft surfaces

Nanomechanical crowding remains theoretically unexplored at the interface of RNA molecules and soft surfaces. Existing RNA molecular models tend to reach very complex ensembles on themselves to be combined to e.g. soft matter mechanics. Here, we introduce a multiscale approach which couples a tractable RNA coarse-grained model with an elastic energy component. Within this approach, we study the specific role of RNA's secondary structure patterns on the deformation of soft surfaces by characterizing representative RNA motifs. We also vary the length of the investigated molecules and propose scaling trends for longer RNAs. Under controlled molecular crowding conditions we analyze the effects of conformational entropy and the interplay between surface energy per monomer and deformation lengths. Our findings add a novel way to address the mechanisms of response of encapsulated RNA inside crowded macromolecular environments, like the ones faced during RNA delivery.   

Theoretical model of efficient phagocytosis driven by curved membrane proteins and active cytoskeleton forces

Phagocytosis is the process of engulfment or internalization of comparatively larger particles by the cell, that plays a central role in our immune system. We study the process of phagocytosis by considering a simplified coarse grained model of a three-dimensional vesicle, having uniform adhesion interaction with a particle, in the presence of curved membrane proteins and active cytoskeleton forces. We note that the complete engulfment is achieved when the bending energy cost of the vesicle is balanced by the gain in the adhesion energy. The addition of curved proteins reduces the bending energy cost at the highly curved leading edge, and thereby the engulfment is achieved at smaller adhesion strength. The curved proteins form the rim of the phagocytic cup that wraps around the particle and makes the engulfment process more efficient than in a protein-free vesicle. The presence of active force allows the formation of the phagocytic cup at even smaller protein density, and the engulfment is achieved more quickly. We consider spherical as well as non-spherical particles (spheroid, sphero-cylinder etc.) and note that the non-spherical particles are more difficult to engulf in comparison to the spherical particles of the same surface area. For non-spherical particles, the engulfment time crucially depends upon the initial orientation of particles with respect to the vesicle. Our results are in agreement with the experiments performed either using artificial particles or the bacterial engulfment by the immune cell.   

Polymer architecture orchestrates the segregation and spatio-temporal organization of replicating E. coli chromosomes in slow growth.

The mechanism and driving forces of chromosome segregation in the bacterial cell cycle of E. coli is one of the least understood events in its life cycle. Using principles of  entropic repulsion between polymer loops confined in a cylinder, we use Monte carlo simulations to show that the  segregation dynamics is spontaneously enhanced by the  adoption of a certain DNA-polymer architecture  as replication  progresses. Secondly, the chosen polymer-topology ensures  its self-organization along  the cell axis while segregation is in progress, such that various chromosomal loci get spatially localized. The  time evolution of loci positions quantitatively match the corresponding experimentally reported results, including observation of the cohesion time and  the ter-transition.  Additionally,  the contact map generated using our bead-spring   model reproduces the  four macro-domains of the experimental Hi-C  maps. Lastly, the proposed mechanism  reproduces the observed universal dynamics as the sister loci separate during segregation. We propose that cross-links (plausibly induced by SMC proteins) at crucial positions along the contour is  enough to provide sufficient entropic forces for segregation within experimentally observed time scales.