Biology subjects

Choi, J.-M.

Publications and source records attributed to Choi, J.-M..

9 recordsLinked to original sources

bioRxiv2025

Single-molecule tweezers decoding hidden dimerization patterns of membrane proteins within lipid bilayers

Dimerization of transmembrane (TM) proteins is an essential biological process within cellular membranes, playing a key role in diverse pathophysiological pathways and serving as a promising therapeutic target. Although often simplified as a two-state transition from freely diffusing monomers to fully formed dimers, the dimerization process after monomer diffusion--the post-diffusion dimerization--is likely more complex due to intricate inter-residue interactions. Here, we introduce a single-molecule tweezer platform to map detailed profiles of the post-diffusion transitions in TM protein dimerization. This approach captures reversible dimerization events of a single TM dimer, revealing hidden intermediate states that emerge following the quiescent phase of monomer diffusion. Profiling the post-diffusion intermediates, kinetics, and energy landscapes--integrated with molecular dynamics simulations--uncovers the dimerization pathway, the effects of residue interactions and lipid bilayers, and the kinetic and energetic contributions of distinct dimerization domains. Furthermore, this platform characterizes selective and localized modulations via peptide binding, underscoring its potential to elucidate the mechanisms of action of TM dimer-targeting drugs at single-molecule resolution.

biophysics↗

bioRxiv2024

Hidden route of protein damage through confined oxygen gas

Oxidative modifications can severely impair protein structure, fold, and function, closely linked to human aging and diseases. Conventional oxidation pathways typically involve the free diffusion of reactive oxygen species (ROS), followed by chemical attacks on the protein surface. Here, we report a hidden route of protein oxidative damage, which we refer to as O2-confinement oxidation pathway. This pathway starts with the initial trapping of dissolved molecular oxygen (O2) within protein cavity spaces, followed by interaction with photosensitizing tryptophan residues. The trapped O2 is then converted to singlet oxygen (1O2), a powerful ROS, through spin-flip electron transfer mechanism under blue light. The generated 1O2 within the protein ultimately attacks the protein core residues through constrained diffusion, accelerating the oxidative damage. This alternative photooxidation pathway through the initial O2 trapping would bypass the antioxidant defense systems which target freely-diffusing ROS, constituting an additional layer of protein oxidative damage in cells and tissues.

biochemistry↗

bioRxiv2023

Biomolecular condensates form spatially inhomogeneous network fluids

The functions of biomolecular condensates are thought to be influenced by their material properties, and these will be determined by the internal organization of molecules within condensates. However, structural characterizations of condensates are challenging, and rarely reported. Here, we deploy a combination of small angle neutron scattering, fluorescence recovery after photobleaching, and coarse-grained molecular dynamics simulations to provide structural descriptions of model condensates that are formed by macromolecules from nucleolar granular components (GCs). We show that these minimal facsimiles of GCs form condensates that are network fluids featuring spatial inhomogeneities across different length scales that reflect the contributions of distinct protein and peptide domains. The network-like inhomogeneous organization is characterized by a coexistence of liquid- and gas-like macromolecular densities that engenders bimodality of internal molecular dynamics. These insights suggest that condensates formed by multivalent proteins share features with network fluids formed by systems such as patchy or hairy colloids.

biophysics↗

bioRxiv2022

Bystander memory-phenotype conventional CD4+ T cells exacerbating autoimmune neuroinflammation

Memory-phenotype (MP) CD4+ T cells are a substantial population of conventional T cells that exist in steady-state mice, and their immunologic functions in autoimmune disease have not yet been studied. In this work, we unveil a unique phenotype of MP CD4+ T cells by analyzing single-cell transcriptomics and T cell receptor (TCR) repertoires. We found that steady-state MP CD4+ T cells exist regardless of germ and food-antigen which are composed of heterogenous effector subpopulations. Distinct subpopulations of MP CD4+ T cells are specifically activated by IL-1 family cytokines and STAT activators, revealing that the cells have TCR-independent effector functions. Especially, CCR6high MP CD4+ T cells are major responders to IL-1{beta} and IL-23 without MOG35-55 antigen reactivity, which gives them pathogenic-Th17 characteristics and allows them to contribute to autoimmune encephalomyelitis. We identified Bhlhe40 in CCR6high MP CD4+ T cells drives the expression of GM-CSF, contributing to CNS pathology in experimental autoimmune encephalomyelitis. Collectively, our findings reveal heterogeneity of MP CD4+ T cells that can contribute to autoimmune neuroinflammation in bystander manner synergistically with antigen-specific T cells.

immunology↗

bioRxiv2019

Computer-aided Binding Mode Prediction and Affinity Maturation of LRR Protein Binder without Structural Determination

Precise binding mode identification and subsequent affinity improvement without structure determination remain a challenge in the development of therapeutic proteins. However, relevant experimental techniques are generally quite costly, and purely computational methods have been unreliable. Here, we show that integrated computational and experimental epitope localization followed by full-atom energy minimization can yield an accurate complex model structure which ultimately enables effective affinity improvement and redesign of binding specificity. As proof-of-concept, we used a leucine-rich repeat (LRR) protein binder, called a repebody (Rb), that specifically recognizes human IgG1 (hIgG1). We performed computationally-guided identification of the Rb:hIgG1 binding mode and leveraged the resulting model to reengineer the Rb so as to significantly increase its binding affinity for hIgG1 as well as redesign its specificity toward multiple IgGs from other species. Experimental structure determination verified that our Rb:hIgG1 model closely matched the co-crystal structure. Using a benchmark of other LRR protein complexes, we further demonstrated that the present approach may be broadly applicable to proteins undergoing relatively small conformational changes upon target binding. Author SummaryIt is quite challenging for computational methods to determine how proteins interact and to design mutations to alter their binding affinity and specificity. Despite recent advances in computational methods, however, in silico evaluation of binding energies has proven to be extremely difficult. We show that, in the case of protein-protein interactions where only small structural changes occur upon target binding, an integrated computational and experimental approach can identify a binding mode and drive reengineering efforts to improve binding affinity or specificity. Using as a model system a leucine-rich repeat (LRR) protein binder that recognizes human IgG1, our approach yielded a model of the protein complex that was very similar to the subsequently experimentally determined co-crystal structure, and enabled design of variants with significantly improved IgG1 binding affinity and with the ability to recognize IgG1 from other species.

bioinformatics↗

bioRxiv2019

LASSI: A lattice model for simulating phase transitions of multivalent proteins

Biomolecular condensates form via phase transitions that combine phase separation or demixing and networking of key protein and RNA molecules. Proteins that drive condensate formation are either linear or branched multivalent proteins where multivalence refers to the presence of multiple protein-protein or protein-nucleic acid interaction domains or motifs within a protein. Recent work has shown that multivalent protein drivers of phase transitions are in fact biological instantiations of associative polymers. Such systems can be characterized by stickers-and-spacers architectures where stickers contribute to system-specific spatial hierarchies of directional interactions and spacers control the concentration-dependent inhomogeneities in densities of stickers around one another. The collective effects of interactions among stickers and spacers lead to the emergence of dense droplet phases wherein the stickers form percolated networks of polymers. To enable the calculation of system-specific phase diagrams of multivalent proteins, we have developed LASSI (LAttice simulations of Sticker and Spacer Interactions), which is an efficient open source computational engine for lattice-based polymer simulations built on the stickers and spacers framework. In LASSI, a specific multivalent protein architecture is mapped into a set of beads on the 3-dimensional lattice space with proper coarse-graining, and specific sticker-sticker interactions are modeled as pairwise anisotropic interactions. For efficient and broad search of the conformational ensemble, LASSI uses Monte Carlo methods, and we optimized the move set so that LASSI can handle both dilute and dense systems. Also, we developed quantitative measures to extract phase boundaries from LASSI simulations, using known and hidden collective parameters. We demonstrate the application of LASSI to two known archetypes of linear and branched multivalent proteins. The simulations recapitulate observations from experiments and importantly, they generate novel quantitative insights that augment what can be gleaned from experiments alone. We conclude with a discussion of the advantages of lattice-based approaches such as LASSI and highlight the types of systems across which this engine can be deployed, either to make predictions or to enable the design of novel condensates.\n\nAuthor SummarySpatial and temporal organization of molecular matter is a defining hallmark of cellular ultrastructure and recent attention has focused intensely on organization afforded by membraneless organelles, which are referred to as biomolecular condensates. These condensates form via phase transitions that combine phase separation and networking of condensate-specific protein and nucleic acid molecules. Several questions remain unanswered regarding the driving forces for condensate formation encoded in the architectures of multivalent proteins, the molecular determinants of material properties of condensates, and the determinants of compositional specificity of condensates. Building on recently recognized analogies between associative polymers and multivalent proteins, we have developed and deployed LASSI, an open source computational engine that enables the calculation of system-specific phase diagrams for multivalent proteins. LASSI relies on a priori identification of stickers and spacers within a multivalent protein and mapping the stickers onto a 3-dimensional lattice. A Monte Carlo engine that incorporates a suite of novel and established move sets enables simulations that track density inhomogeneities and changes to the extent of networking among stickers as a function of protein concentration and interaction strengths. Calculation of distribution functions and other nonconserved order parameters allow us to compute full phase diagrams for multivalent proteins modeled using a stickers-and-spacers representation on simple cubic lattices. These predictions are shown to be system-specific and allow us to rationalize experimental observations while also enabling the design of systems with bespoke phase behavior. LASSI can be deployed to study the phase behavior of multicomponent systems, which allows us to make direct contact with cellular biomolecular condensates that are in fact multicomponent systems.

biophysics↗

bioRxiv2018

Measuring NDC80 binding reveals the molecular basis of tension-dependent kinetochore-microtubule attachments

Proper kinetochore-microtubule attachments, mediated by the NDC80 complex, are required for error-free chromosome segregation. Erroneous attachments are corrected by the tension dependence of kinetochore-microtubule interactions. Here, we present a method, based on fluorescence lifetime imaging microscopy and Forster resonance energy transfer, to quantitatively measure the fraction of NDC80 complexes bound to microtubules at individual kinetochores in living human cells. We found that NDC80 binding is modulated in a chromosome autonomous fashion over prometaphase and metaphase, and is predominantly regulated by centromere tension. We show that this tension dependency requires phosphorylation of the N-terminal tail of Hec1, a component of the NDC80 complex, and the proper localization of Aurora B kinase, which modulates NDC80 binding. Our results lead to a mathematical model of the molecular basis of tension-dependent NDC80 binding to kinetochore microtubules in vivo.

cell biology↗

bioRxiv2017

ProteomeVis: a web app for exploration of protein properties from structure to sequence evolution across organisms’ proteomes

MotivationProtein evolution spans time scales and its effects span the length of an organism. A web app named ProteomeVis is developed to provide a comprehensive view of protein evolution in the S. cerevisiae and E. coli proteomes. ProteomeVis interactively creates protein chain graphs, where edges between nodes represent structure and sequence similarities within user-defined ranges, to study the long time scale effects of protein structure evolution. The short time scale effects of protein sequence evolution is studied by sequence evolutionary rate (ER) correlation analyses with protein properties that span from the molecular to the organismal level.\n\nResultsWe demonstrate the utility and versatility of ProteomeVis by investigating the distribution of edges per node in organismal protein chain universe graphs (oPCUGs) and putative ER determinants. S. cerevisiae and E. coli oPCUGs are scale-free with scaling constants of 1.78 and 1.50, respectively. Both scaling constants can be explained by a previously reported theoretical model describing protein structure evolution (Dokholyan et al., 2002). Protein abundance most strongly correlates with ER among properties in ProteomeVis, with Spearman correlations of -0.51 (p-value<10-10) and -0.46 (p-value<10-10) for S. cerevisiae and E. coli, respectively. This result is consistent with previous reports that found protein expression to be the most important ER determinant (Zhang and Yang, 2015).\n\nAvailabilityProteomeVis is freely accessible at http://proteomevis.chem.harvard.edu/proteomevis/.\n\nContactshakhnovich@chemistry.harvard.edu\n\nSupplementary informationAvailable to download on bioRxiv.

evolutionary biology↗

bioRxiv2016

Tuning the course of evolution on the biophysical fitness landscape of an RNA virus

Predicting viral evolution remains a major challenge with profound implications for public health. Viral evolutionary pathways are determined by the fitness landscape, which maps viral genotype to fitness. However, a quantitative description of the landscape and the evolutionary forces on it remain elusive. Here, we apply a biophysical fitness model based on capsid folding stability and antibody binding affinity to predict the evolutionary pathway of norovirus escaping a neutralizing antibody. The model is validated by experimental evolution in bulk culture and in a drop-based microfluidics device, the \"Evolution Chip\", which propagates millions of independent viral sub-populations. We demonstrate that along the axis of binding affinity, selection for escape variants and drift due to random mutations have the same direction. However, along folding stability, selection and drift are opposing forces whose balance is tuned by viral population size. Our results demonstrate that predictable epistatic tradeoffs shape viral evolution.

evolutionary biology↗