Zi Hao Liu

The unique solvent milieu found in biomolecular condensates can control cellular enzymatic reactions and shift reaction kinetics by modulating reactant concentrations, structural dynamics, and enzyme activities. Here, we explore the interplay of multiple regulatory factors within a condensate to control poly(A) RNA deadenylation, the first and rate-limiting step in mRNA turnover. The deadenylase CNOT7, a subunit of the CCR4-NOT deadenylation complex, localizes to protein- and RNA-rich granules and shows increased degradation activity in vitro in condensates formed by the C-terminal low-complexity disordered region of CAPRIN1, a component of many ribonucleoprotein (RNP) granules. We use a combination of enzymatic assays, kinetic modeling, microscopy, nuclear magnetic resonance (NMR) spectroscopy, and molecular dynamics simulations to deconvolute and define the components that underlie this enhancement. We found that enzyme and RNA are concentrated in condensates relative to buffer, which increases CNOT7 activity, while the equilibrium between CNOT7’s active and inactive states remains unchanged. The concentration-dependent increase in enzymatic rates is counterbalanced by a decrease in the enzyme’s catalytic efficiency, likely due to slower diffusion of CNOT7 and RNA within the condensates, which lessens the probability of enzyme–substrate complex formation. Molecular dynamics simulations reveal CNOT7-CAPRIN1 interactions that rely on conserved CAPRIN1 sequence features, hinting at an evolutionarily conserved role for CAPRIN1 condensation. With this quantitative kinetic analysis, we describe the multifaceted mechanism behind regulation of CNOT7 deadenylation by a condensate environment.

Although machine learning has transformed protein structure prediction of folded protein ground states with remarkable accuracy, intrinsically disordered proteins and regions (IDPs/IDRs) are defined by diverse and dynamical structural ensembles that are predicted with low confidence by algorithms such as AlphaFold and RoseTTAFold. We present a machine learning method, IDPForge ( I ntrinsically D isordered P rotein, FO lded and disordered R egion GE nerator), that exploits a transformer protein language diffusion model to create all-atom IDP ensembles and IDR ensembles that maintains the folded domains. IDPForge does not require sequence-specific training, back transformations from coarse-grained representations, nor ensemble reweighting. The created IDP/IDR conformational ensembles show good agreement with both NMR and SAXS solution experimental data, and options for biasing with experimental restraints are available if desired. IDPForge with these diverse capabilities facilitates integrative and structural studies for proteins that contain intrinsic disorder, and is available as an open source resource for general use.

The Protein Ensemble Database (PED; https://proteinensemble.org) is the primary deposition resource for structural ensembles of intrinsically disordered proteins (IDPs) and intrinsically disordered regions (IDRs). PED 2027 (released in September 2026) substantially expands its scope by integrating ensembles determined or refined using experimental data with clearly identified collections of computational predictions. The release introduces a multi-method collection targeting 3319 DisProt proteins and 2452 manually curated disordered regions, together with AlphaFlex ensembles for 10 022 human proteins containing IDRs. Alternative predictions for the same sequence are retained as independent records with method-specific provenance, enabling comparison across ensemble-generation approaches. A scalable deposition workflow accommodates individual experimental studies and proteome-scale collections while preserving quality control and curator oversight. The PED controlled vocabulary standardizes experimental data acquisition, computational generation, selection, and refinement. Precomputed IDPEnsembleTools and CIDER analyses provide structural and sequence-derived descriptors directly on ensemble pages. A redesigned interface and extended REST API support discovery, visualization, download and programmatic access. Interoperability with UniProt, DisProt, MobiDB, and experimental-data repositories connects ensembles with sequences, annotations, and supporting measurements. PED 2027 therefore provides an interoperable, analysis-ready platform for studying IDP/IDR conformational behaviour.

Eukaryotic cap‐dependent translation initiation is regulated by binding of the predominantly folded eukaryotic initiation factor 4E (eIF4E) to the intrinsically disordered eIF4E binding proteins (4E‐BPs). Here, we report full‐length atomistic conformational ensembles generated by IDPConformerGenerator and optimized by X‐EISDv2 workflow for both apo 4E‐BP2, the neuronal 4E‐BP, and 4E‐BP2 in complex with eIF4E, using data from single‐molecule fluorescence and nuclear magnetic resonance (NMR), together with select coordinates from a 4E‐BP1:eIF4E crystal structure. Structural sampling within dynamic complexes is often underappreciated, with NMR and crystal structure data for 4E‐BP:eIF4E suggesting different degrees of structural heterogeneity. Our ensemble models validated by solution spectroscopy data enable comparison of free 4E‐BP2 and its complex with eIF4E. This shows a delocalization of contacts around canonical regions, which supports previous findings of unidirectional conditional occupancy of the binding sites. Two new contact regions emerged: one between the disordered N ‐termini of eIF4E and 4E‐BP2, which may play an allosteric role in tuning the binding affinity, and the other between the C‐terminus of 4E‐BP2 and an extended region of eIF4E, which is consistent with the extended, dynamic binding interface that we reported previously. These results support a model of translation regulation in which the dynamic 4E‐BP2:eIF4E complex facilitates accessibility of regulatory sites of 4E‐BP2 when bound.

The unique solvent milieu found in biomolecular condensates can control cellular enzymatic reactions and shift reaction kinetics by modulating reactant concentrations, structural dynamics, and enzyme activities. Here we explore the interplay of multiple regulatory factors within a condensate to control poly(A) RNA deadenylation, the first and rate-limiting step in mRNA turnover. The deadenylase CNOT7, a subunit of the CCR4-NOT deadenylation complex, localizes to cytoplasmic RNA granules and shows increased degradation activity in vitro in condensates formed by the C-terminal low complexity disordered region of CAPRIN1, a component of RNA granules. We use a combination of enzymatic assays, kinetic modeling, microscopy, Nuclear Magnetic Resonance (NMR) spectroscopy, and molecular dynamics simulations to deconvolute and define the components that underlie this enhancement. We found that enzyme and RNA are concentrated in condensates relative to buffer, which increases CNOT7 activity, while the equilibrium between CNOT7's active and inactive states remains unchanged. The concentration-dependent increase in enzymatic rates is counterbalanced by a substantial decrease in the enzyme's catalytic efficiency, likely due to slower diffusion of CNOT7 and RNA within the condensates, which lessens the probability of enzyme-substrate complex formation. Molecular dynamics simulations reveal CNOT7-CAPRIN1 interactions that rely on conserved CAPRIN1 sequence features, hinting at an evolutionarily conserved role for CAPRIN1 condensation. With this quantitative kinetic analysis, we describe the multifaceted mechanism behind regulation of CNOT7 deadenylation by a condensate environment.

Eukaryotic cap-dependent translation initiation is regulated by binding of the predominantly folded eukaryotic initiation factor 4E (eIF4E) to the intrinsically disordered eIF4E binding proteins (4E-BPs). Here, we report full-length atomistic conformational ensembles generated by IDPConformerGenerator and optimized by X-EISDv2 workflow for both apo 4E-BP2, the neuronal 4E-BP, and 4E-BP2 in complex with eIF4E, using data from single-molecule fluorescence and nuclear magnetic resonance (NMR), together with select coordinates from a 4E-BP1:eIF4E crystal structure. Structural sampling within dynamic complexes is often under-appreciated, with NMR and crystal structure data for 4E-BP:eIF4E suggesting different degrees of structural heterogeneity. Our ensemble models validated by solution spectroscopy data enable comparison of free 4E-BP2 and its complex with eIF4E. This shows a delocalization of contacts around canonical regions, which supports previous findings of unidirectional conditional occupancy of the binding sites. Two new contact regions emerged: one between the disordered N-termini of eIF4E and 4E-BP2, which may play an allosteric role in tuning the binding affinity, and the other between the C-terminus of 4E-BP2 and an extended region of eIF4E, which is consistent with the extended, dynamic binding interface that we reported previously. These results support a model of translation regulation in which the dynamic 4E-BP2:eIF4E complex facilitates accessibility of regulatory sites of 4E-BP2 when bound.

NMR spectroscopy is a powerful tool for studies of protein dynamics over timescales extending more than twelve orders of magnitude, with motion queried at many of the backbone and sidechain positions in the molecule of interest. NMR experiments can, in principle, provide atomic resolution descriptions of excited conformational states that are sparsely populated and transiently formed—i.e., invisible, and there are examples of such structures in the literature. However, in other cases, the NMR data are not sufficient for generating structural ensembles of rare and transient higher energy conformers of the protein of interest. One example is provided by the precursor form of the proinflammatory cytokine interleukin-18, pro-IL-18, studied here. Although NMR studies show that pro-IL-18 adopts two sparsely populated (<0.5%) and transiently formed (ms lifetimes) excited-state conformations in exchange with a highly populated ground state conformer and localize regions undergoing exchange to a pair of short β-strands that are preserved in at least one of the excited states, additional structural information is not forthcoming. Here, we develop a protocol whereby the NMR data are used to select alternative conformers of pro-IL-18 from ensembles predicted by the generative machine learning model AlphaFlow that are then evaluated through further NMR experiments. The method identifies distinct conformers that correspond to the pair of excited states reported by NMR, combining experiment with computation to characterize the pro-IL-18 energy landscape in ways that are not possible with either approach alone.

Although machine learning has transformed protein structure prediction of folded protein ground states with remarkable accuracy, intrinsically disordered proteins and regions (IDPs/IDRs) are defined by diverse and dynamical structural ensembles that are predicted with low confidence by algorithms such as AlphaFold and RoseTTAFold. We present a new machine learning method, IDPForge ({I}ntrinsically {D}isordered {P}rotein, {FO}lded and disordered {R}egion {GE}nerator), that exploits a transformer protein language diffusion model to create all-atom IDP ensembles and IDR disordered ensembles that maintains the folded domains. IDPForge does not require sequence-specific training, back transformations from coarse-grained representations, nor ensemble reweighting, as in general the created IDP/IDR conformational ensembles show good agreement with bot NMR and SAXS solution experimental data, and options for biasing with experimental restraints are provided if desired. We envision that IDPForge with these diverse capabilities will facilitate integrative and structural studies for proteins that contain intrinsic disorder, and is available as an open source resource for general use.

Understanding how proteins explore their conformational energy landscapes is essential for linking sequence to function, yet current ensemble methods are limited by sampling inefficiency and poor scalability to large disordered systems. Here we introduce a free energy landscape screen (FELS), a conceptually different approach that replaces sampling-centric ensemble fitting with broad exploration of energy landscapes, screening thousands of landscape shapes—from highly funneled to flat and rugged. By systematically biasing and evaluating large conformer pools according to contact propensities derived from experiment, FELS efficiently identifies sets of conformers that best reproduce experimental data and highlights candidates for structural refinement, without being restricted by chain length or amount of disorder. To demonstrate the power of this approach we applied it to a previously intractable system, human tropoelastin (hTE), a ∼700-residue precursor of elastin. FELS provides the first experimentally defined atomistic view of the hTE conformational ensemble, revealing that this protein is intrinsically disordered yet exhibits distinct local secondary structure and specific, transient medium- and long-range contacts that organize its ensemble. These findings reconcile long-standing conflicting models and demonstrate that FELS provides a general, experimentally driven framework for mapping conformational energy landscapes of large proteins across the continuum between structural order and disorder.

NMR relaxation studies of the precursor form of the pro-inflammatory cytokine interleukin-18, pro-IL-18, show that it adopts two sparsely populated (<0.5%) and transiently formed (ms lifetimes) excited state conformations in exchange with a highly populated ground state conformer. Although NMR data localize regions undergoing exchange to a pair of short β-strands that are preserved in at least one of the excited states, additional structural information is not forthcoming. Here we develop a protocol whereby the NMR data is used to select alternative conformers of pro-IL-18 from ensembles predicted by the generative ML model AlphaFlow that are then evaluated through further NMR experiments. The approach identifies distinct excited state conformers and suggests a general method for combining experiment with computation to characterize protein energy landscapes.