Protein Science最新文献

Human transferrin receptor 1 (TfR) is essential for cellular iron homeostasis by internalizing the iron carrier proteins transferrin and ferritin. It is also an entry point for various pathogens, such as South American hemorrhagic fever caused by arenaviruses and the malaria parasite Plasmodium vivax, which utilize TfR to gain access to cells. The receptor is additionally upregulated in many aggressive cancers and at the blood-brain barrier. Altogether, the TfR is a highly relevant target for many medical applications, and novel protein-interacting partners are sought after. A protein design strategy was explored here to develop a small protein that can be used for drug delivery across cell membranes, to investigate blood-brain barrier crossing, study endocytosis, or to block pathogen access to the apical domain. A computationally docked library of small protein scaffolds to the TfR apical domain, the native binding site of the Machupo arenavirus, was a starting point for the design and optimization. The best variants were expressed in a yeast surface display system and assessed for interaction with TfR by flow cytometry. One protein variant, which initially showed a low binding signal, was further optimized by directed evolution to bind to the target receptor at nanomolar concentration. The evolved construct, tagged with the enhanced green fluorescent protein (eGFP) and bacterially expressed, showed uptake similar to that of FITC-coupled transferrin in a cell-based assay. The designed protein can be utilized as a tool to target cell entry via TfR for drug delivery applications or as a foundation for developing antiviral therapeutics against arenaviruses.

Computational design of functional proteins is of both fundamental and applied interest. This study introduces a generative framework for co-designing protein sequence and structure in a unified process by modeling their joint distribution, with the goal of enabling cross-modality interactions toward coherent and functional designs. Each residue is represented by three distinct modalities (type, position, and orientation) and modeled using dedicated diffusion processes: multinomial for types, Cartesian for positions, and special orthogonal group SO(3) for orientations. To couple these modalities, we propose a unified architecture, ReverseNet, which employs a shared graph attention encoder to integrate multimodal information and separate projectors to predict each modality. We benchmark our models, JointDiff and JointDiff-x, on unconditional monomer design and conditional motif scaffolding tasks. Compared to two-stage design models that generate sequence and structure separately, our models produce monomer structures with comparable or better designability, while currently lagging in sequence quality and motif scaffolding performance based on computational metrics. However, they are 1-2 orders of magnitude faster and support rapid iterative improvements through classifier-guided sampling. To complement computational evaluations, we experimentally validate our approach through a case study on green fluorescent protein (GFP) design. Several novel, evolutionarily distant variants generated by our models exhibit measurable fluorescence, confirming functional activity. These results demonstrate the feasibility of joint sequence-structure generation and establish a foundation to accelerate functional protein design in future applications. Codes, data, and trained models are accessible at https://github.com/Shen-Lab/JointDiff.

The classification of novel protein folds remains a central challenge in structural bioinformatics, particularly as deep learning models like AlphaFold2 dramatically expand the universe of predicted protein structures. In this study, we investigated 664 candidate novel fold (CNF) domains from the TED database that both TED and DPAM methods had classified with low confidence. These CNFs span a structurally diverse and largely non-redundant set of domains, most of which lack clear sequence or structural similarity to known folds. Many CNFs appear as insertions into known transmembrane or enzymatic domains, while others occur in modular architectures, co-occurring with interaction or catalytic folds such as β-barrels, zinc fingers, or Rossmann-like domains. Although some CNFs resemble known folds that have undergone topological rearrangements or circular permutations, others result from errors in domain boundary prediction, often due to truncated sequences or tightly packed domain duplications. Our analyses led to the creation of 190 new Pfam families, many classified as domains of unknown function (DUFs), and revealed intriguing cases of zinc-binding and disulfide-rich architectures that contribute to fold space expansion. A small subset of CNFs helped define new superfamilies by linking previously unclassified but structurally related domains. Taken together, this work underscores the importance of integrating structural, evolutionary, and contextual information to resolve challenging fold assignments and provides a roadmap for extending protein classification frameworks into previously uncharted structural territory.

Lysine-specific demethylase 1 (LSD1) plays a crucial role in chromatin organization and gene regulation by removing methyl groups from histone and non-histone substrates. While its catalytic core is well characterized, the functional contributions of its intrinsically disordered N-terminal region remain less understood. Here, we identify a conserved acidic patch within this unstructured domain as a key regulator of LSD1 activity. Our findings suggest that this region influences enzymatic efficiency and interactions with regulatory cofactors, shedding new light on the mechanistic control of LSD1 function in epigenetic modulation.

Intracellular protein aggregation occurs in a highly crowded environment. The intracellular environment is highly heterogeneous, featuring diverse crowder protein surface chemistries along with varying crowder stability and solubility. It remains unclear how these aspects influence protein aggregation. Therefore, we assessed how a crowder protein and its surface properties impact aggregation. We utilize high concentrations of surface-modified proteins based on bovine serum albumin (BSA) to monitor how they influence the aggregation of mutant huntingtin exon 1, enabled by fluorescent proteins (mHttex1-VC) for förster resonance energy transfer (FRET). This system reveals three mechanisms through which bystander proteins direct mHttex1-VC aggregation: (1) monodisperse inert proteins appear to function as crowders, increasing the amount of fibrils and their length and width; (2) marginally soluble proteins strongly enhance mHttex1-VC aggregation and density through coaggregation; and (3) crowders that bind mHttex1-VC or folding-destabilized crowders reduce aggregation. The buffer conditions modulate the effects of the protein surface. Thus, in addition to macromolecular crowding effects, the crowder stickiness, solubility, and stability determine the aggregation of the test protein. We expect these effects to also play a role in cells.

Aryl polyenes (APEs) are critical secondary metabolites in gram-negative pathogens, synthesized by polyketide synthases that rely on acyl carrier proteins (ACPs) as essential cofactors. Despite their importance, the structural and functional dynamics of tandem ACPs-ApeE and ApeF-have remained largely uncharacterized. In this study, we elucidate the molecular mechanisms underlying these ACPs within the APE biosynthetic gene cluster of carbapenem-resistant Acinetobacter baumannii (CRAB), a major threat to global health, revealing features that markedly differ from those in conventional fatty acid or polyketide synthesis systems. We demonstrate that ApeE functions as the primary starter ACP, engaging selectively with benzoyl-ACP synthetase (ApeH) to initiate APE biosynthesis. Through NMR spectroscopy and molecular dynamics simulations, we reveal that ApeE possesses distinctive features-such as a glycine-rich motif, a substrate-binding surface pocket, and a highly mobile NAE-lid domain-that orchestrate dynamic mechanisms required to accommodate and transfer bulky, rigid intermediates. CPMG and CEST experiments further uncover conformational exchange between "in" and "out" states at the prosthetic group attachment site S41 and α3-helix, regulating substrate entry and release in concert with the flexible NAE-lid domain. Functional assays show that key motifs in ApeE facilitate interactions with biosynthetic enzymes, highlighting its specialized role in handling large APE intermediates. Conversely, ApeF features a shallow hydrophobic cavity optimized for efficient malonyl-group transfer. These insights establish a novel molecular framework for ACP function in complex APE biosynthesis, offering promising avenues for targeted antimicrobial development against multidrug-resistant pathogens like CRAB.

The muscle protein titin spans half a sarcomere, from M-line to Z-disk, and is essential for both active and passive stretch. The N2A region of titin plays a critical role in various regulatory processes through its binding interactions. Located at the C-terminus of the N2A region, adjacent to the PEVK region, are the I82 and I83 domains, which are key to binding calpain/p94. However, this interaction is absent in the mdm-mouse model, which contains an 83-amino acid deletion spanning the C-terminus of the I83 domain and the N-terminus of the PEVK region, leading to muscular dystrophy with myositis. This mdm-deletion disrupts the structure of the I83 domain, preventing normal force enhancement in the presence of calcium and inhibiting eccentric contractions. Our lab has demonstrated that the I83 domain exhibits calcium sensitivity at concentrations similar to those found in active muscle. In this current study, we further demonstrate that the tandem I82-I83 domains exhibit cooperative unfolding, as seen by a single unfolding event, and that calcium enhances the stability of the tandem I82-I83 domains. The NMR structure of this construct exhibits a tighter interface between I82 and I83 than is observed in the crystal structure, suggesting that the two structures might represent the structure in the relaxed state versus the structure under force. The calcium response of these domains is hypothesized to affect the function of the N2A region during muscle activation.

Cannabidiol (CBD), the second most abundant of the active compounds found in the Cannabis sativa plant, is nonaddictive and of increasing interest due to its potential therapeutic utilities. The actions of CBD are mediated by cannabinoid receptors and other targets, both G protein-coupled receptors (GPCRs) and non-GPCR proteins. It has recently been shown that CBD is not an orthosteric ligand of the adenosine A_2A receptor (A_2AR), but rather a negative allosteric modulator. Because of the multitude non-conserved allosteric pockets in GPCRs, the binding mode of CBD to A_2AR remains unknown. To fill this knowledge gap, and due to the therapeutic relevance of CBD and A_2AR, we have used metadynamics simulations and site-directed mutagenesis to identify the binding mode of CBD. These theoretical and experimental results show that the allosteric binding site of CBD is along the ligand binding pathway of adenosine to A_2AR, near the orthosteric binding site.

In this study, we explored the design of linear D-tripeptides tailored to bind specific cavities of Gadd45β, chosen as a model protein target. To identify peptides that selectively interact with predicted binding sites, we combined computational modeling with biophysical experiments. Gadd45β was selected since it has emerged as a promising therapeutic target involved in multiple disease pathways, including cancer and inflammation. Computational analysis was first employed to characterize the structural features and potential binding sites of Gadd45β. Guided by these insights, linear D-tripeptides were designed and optimized for specific interactions with the target surface. The resulting candidates were subsequently assessed through a series of biophysical assays to evaluate their binding affinity, selectivity, and potential therapeutic activity. Complementary computational simulations were employed to gain atomistic insight into the dynamics of peptide-protein recognition. This integrated computational-experimental strategy led to the identification of two D-tripeptides, RYR and VWR, that bind Gadd45β at a biologically relevant site, illustrating a general framework for early-stage peptide ligand discovery.

IM30, the inner membrane-associated protein of 30 kDa (also known as Vipp1) is essential for thylakoid membrane biogenesis and/or maintenance in chloroplasts and cyanobacteria. IM30 and its bacterial homolog PspA belong to the ESCRT-III superfamily, proteins previously thought to be restricted to eukaryotes and archaea. Despite low sequence similarity, IM30 shares key structural and functional features with eukaryotic ESCRT-IIIs, including a conserved α1-α2 helical hairpin core and the ability to form oligomeric barrel or rod assemblies that mediate membrane remodeling. Using IM30 variants, we now show that membrane binding of IM30 is driven by electrostatic interactions between the positively charged α1-α3 helical hairpin and negatively charged lipid surfaces, paralleling the role of charged helical regions in some eukaryotic ESCRT-IIIs. This likely is followed by lateral assembly of IM30 into higher-order barrel or rod structures on the membrane. Once assembled, α0 helices within these oligomers engage and stabilize internalized membrane tubules, mirroring membrane interaction strategies of eukaryotic ESCRT-IIIs, which use both N-terminal sequence motifs and charged residues on α1/α2. Thus, our findings demonstrate a conserved membrane binding and remodeling mechanism across the ESCRT-III superfamily, underscoring an evolutionary link in membrane dynamics between pro- and eukaryotes.