共找到 20 条结果
Fluorescence-activating proteins (FAP) have emerged as a novel class of genetically encoded tools for fluorescence-based protein imaging, complementing the existing toolkit consisting of fluorescent proteins and self-labeling tags. FAP have the ability to bind and activate the fluorescence of small molecules, called fluorogens, that are otherwise non-fluorescent, allowing protein localization with high specificity and little background. In this review, we present the engineering of FAP and FAP-based reporters from various protein scaffolds, focusing on the different strategies implemented to design and engineer their properties for specific biological imaging applications.
Proteins and peptides underpin essential biological functions and technological applications, from targeting disease-relevant interactions to providing broad enzymatic activities. However, engineering molecules with desired properties remains difficult, owing to complex sequence-structure-function relationships and the lack of data on specific systems. Experimental selection strategies, including directed evolution, phage display, and mRNA display, address this challenge by leveraging high diversity libraries and iterative enrichment under defined selection pressures. This allows for the identification of candidates without requiring extensive prior knowledge, and can generate extensive datasets for use in machine learning. While many selection systems exist, comparisons across different selection approaches are hindered by the lack of a unifying analytical framework. Here, we developed a toolset of broadly applicable analyses for assessing selection dynamics in multi-round or multi-condition experiments, ranging from position level analysis of sequence properties to full sequence space mappings through protein language model embeddings. Performing analyses across different systems, we identify desirable traits in selection experiments including enrichment of distinct sequence patterns and correlation between enrichment and final desired functions. Notably, even under weak selection regimes with all sequences <1% frequency, functional sequences (e.g., 70nM IC50 binder to SARS-CoV-2 main protease) are still consistently enriched. We also find repeated selections of the same starting library can help differentiate selection effects of varying conditions (e.g., different delivery of metal ligand) from system noise. These findings, along with the toolset, can be used to guide experimental design, interpretation, and troubleshooting across protein and peptide discovery platforms.
Miniprotein designs are emerging as promising antibody alternatives for therapeutic and diagnostic use. Using RFdiffusion, we designed 538 binder candidates targeting both N and C-terminal domains of the SARS-CoV-2 nucleocapsid protein, selecting 19 for recombinant production in E. coli. All were soluble and purified (8-10 mg/L yield), though 13 unexpectedly formed oligomers. CD spectroscopy confirmed proper folding and thermal refolding, with 9 miniproteins exhibiting Tm > 75°C. Binding assay revealed three miniproteins with affinities of 115 nM, 4.8 μM, and 7.32 μM. The 1.8 Å resolution crystal structure of one binder (Gpx62) matched the predicted design. However, predictive metrics (like ipTM of AlphaFold3 and computational simulations) did not align with experimental data. Together, these results reveal unintended oligomerization as a major, previously underappreciated barrier in miniprotein binder discovery, demonstrating that while RFdiffusion reliably predicts structural integrity and stability of miniprotein, its current metrics do not account for the oligomeric behavior that might critically limit binding competence.
While the field of computational protein design has witnessed amazing progression in recent years, folding properties still constitute a significant barrier towards designing new and larger proteins. In order to assess and improve folding properties of designed proteins, we have developed a genetics-based folding assay and selection system based on the essential enzyme, orotate phosphoribosyl transferase from Escherichia coli. This system allows for both screening of candidate designs with good folding properties and genetic selection of improved designs. Thus, we identified single amino acid substitutions in two failed designs that rescued poorly folding and unstable proteins. Furthermore, when these substitutions were transferred into a well-structured design featuring a complex folding profile, the resulting protein exhibited native-like cooperative folding with significantly improved stability. In protein design, a single amino acid can make the difference between folding and misfolding, and this approach provides a useful new platform to identify and improve candidate designs.
Directed evolution for protein engineering, as currently practiced in the biotechnology and pharmaceutical industries, is both tedious and expensive. Computationally driven protein design has the potential to expedite the engineering process and generate high-quality variants at a lower cost than traditional approaches. We investigated the effectiveness of two different computational methods as triaging tools for prioritizing target positions and identifying specific mutations that are likely to improve protein thermodynamic stability. Our benchmarking study used a comprehensive dataset consisting of 174,945 mutations across 180 distinct proteins and evaluated the ESM (Evolutionary Scale Modeling) protein language model alongside a physics-based method, MM/GBSA (Molecular Mechanics Generalized Born Surface Area). We found prediction biases in each method but also determined that these biases can be mitigated by applying the two methods in a complementary manner. We propose a hybrid mutation prioritization and selection strategy that achieves better accuracy than either method alone. Through re-ranking, the combined prioritization strategy attained a higher overall average ROC (receiver operating characteristic) AUC (area under curve) of 0.743 across the dataset compared to either MM/GBSA alone (0.685) or ESM Log Odds alone (0.597). The integrated framework can be adapted and applied to newer AI and physics-based models as the field advances.
A wide variety of protein and peptidomimetic design tasks require matching functional 3D motifs to potential oligomeric scaffolds. For example, during enzyme design, one aims to graft active-site patterns-typically consisting of 3-15 residues-onto new protein surfaces. Identifying protein scaffolds suitable for such active-site engraftment requires costly searches for protein folds that provide the correct side chain positioning to host the desired active site. Other examples of biodesign tasks that require similar fast exact geometric searches of potential side chain positioning include mimicking binding hotspots, design of metal binding clusters and the design of modular hydrogen binding networks for specificity. In these applications, the speed and scaling of geometric searches limits the scope of downstream design to small patterns. Here, we present an adaptive algorithm capable of searching for side chain take-off angles, which is compatible with an arbitrarily specified functional pattern and which enjoys substantive performance improvements over previous methods. We demonstrate this method in both genetically encoded (protein) and synthetic (peptidomimetic) design scenarios. Examples of using this method with the Rosetta framework for protein design are provided. Our implementation is compatible with multiple protein design frameworks and is freely available as a set of python scripts (https://github.com/JiangTian/adaptive-geometric-search-for-protein-design).
Machine learning is a useful computational tool for large and complex tasks such as those in the field of enzyme engineering, selection and design. In this review, we examine enzyme-related applications of machine learning. We start by comparing tools that can identify the function of an enzyme and the site responsible for that function. Then we detail methods for optimizing important experimental properties, such as the enzyme environment and enzyme reactants. We describe recent advances in enzyme systems design and enzyme design itself. Throughout we compare and contrast the data and algorithms used for these tasks to illustrate how the algorithms and data can be best used by future designers.
Protein-based binders have become increasingly more attractive candidates for drug and imaging agent development. Such binders could be evolved from a number of different scaffolds, including antibodies, natural protein effectors and unrelated small protein domains of different geometries. While both computational and experimental approaches could be utilized for protein binder engineering, in this review we focus on various computational approaches for protein binder design and demonstrate how experimental selection could be applied to subsequently optimize computationally-designed molecules. Recent studies report a number of designed protein binders with pM affinities and high specificities for their targets. These binders usually characterized with high stability, solubility, and low production cost. Such attractive molecules are bound to become more common in various biotechnological and biomedical applications in the near future.
The molecular recognition ability of proteins is essential in biological systems, and therefore a considerable amount of effort has been devoted to constructing desired target-binding proteins using a variety of naturally occurring proteins as scaffolds. However, since generating a binding site in a native protein can often affect its structural properties, highly stable de novo protein scaffolds may be more amenable than the native proteins. We previously reported the generation of de novo proteins comprising three α-helices and three β-strands (α3β3) from a genetic library coding simplified amino acid sets. Two α3β3 de novo proteins, vTAJ13 and vTAJ36, fold into a native-like stable and molten globule-like structures, respectively, even though the proteins have similar amino acid compositions. Here, we attempted to create binding sites for the vTAJ13 and vTAJ36 proteins to prove the utility of de novo designed artificial proteins as a molecular recognition tool. Randomization of six amino acids at two linker sites of vTAJ13 and vTAJ36 followed by biopanning generated binding proteins that recognize the target molecules, fluorescein and green fluorescent protein, with affinities of 10(-7)-10(-8) M. Of note, the selected proteins from the vTAJ13-based library tended to recognize the target molecules with high specificity, probably due to the native-like stable structure of vTAJ13. Our studies provide an example of the potential of de novo protein scaffolds, which are composed of a simplified amino acid set, to recognize a variety of target compounds.
Recent years have seen remarkable clinical success with therapies that harness the patient's immune system, with checkpoint inhibitors in oncology as a prominent example. Non-invasive monitoring of immune activation in vivo has the potential to accelerate both basic immunology research and clinical drug development, offering a valuable means to track responses to emerging immunotherapies. CD69 is a rapidly induced activation marker on lymphocytes and other leukocytes, making it an attractive imaging target. For such applications, affibody molecules offer distinct advantages as radio imaging tracers due to their small size, high affinity, and rapid pharmacokinetics, resulting in excellent imaging contrast. Here, we combined directed evolution with computational design to optimise a CD69-binding affibody molecule. An alanine scan of the parental binder informed the construction of a diversification library, which was displayed on Escherichia coli and subjected to iterative MACS and FACS selections with stringent off-rate competition. The top selection hit was subsequently refined through site-directed mutagenesis, including variants suggested by a general protein language model. The resulting lead, Z1525, bound human CD69 with single-digit nanomolar affinity and showed improved thermal stability while retaining solubility and refolding capacity, consistent with suitability for radiolabelling and in vivo targeting. Most importantly, it displayed selective binding to CD69 on stimulated Jurkat cells, with negligible binding to resting cells. These results establish E. coli display with off-rate-driven selection as an efficient strategy for affibody affinity maturation and demonstrate that protein language models can effectively guide improvements in folding stability. Z1525 represents a promising radio imaging tracer candidate for monitoring immune activation in vivo and warrants further preclinical development.
Protein redesign methods aim to improve a desired property by carefully selecting mutations in relevant regions guided by protein structure. However, often protein structural requirements underlying biological characteristics are not well understood. Here, we introduce a methodology that learns relevant mutations from a set of proteins that have the desired property and demonstrate it by successfully improving production levels of two enzymes by Aspergillus niger, a relevant host organism for industrial enzyme production. We validated our method on two enzymes, an esterase and an inulinase, creating four redesigns with 5-45 mutations. Up to 10-fold increase in production was obtained with preserved enzyme activity for small numbers of mutations, whereas production levels and activities dropped for too aggressive redesigns. Our results demonstrate the feasibility of protein redesign by learning. Such an approach has great potential for improving production levels of many industrial enzymes and could potentially be employed for other design goals.
Attempts to create novel ligand-binding proteins often focus on formation of a binding pocket with shape complementarity against the desired ligand (particularly for compounds that lack distinct polar moieties). Although designed proteins often exhibit binding of the desired ligand, in some cases they display unintended recognition behavior. One such designed protein, that was originally intended to bind tetrahydrocannabinol (THC), was found instead to display binding of 25-hydroxy-cholecalciferol (25-D3) and was subjected to biochemical characterization, further selections for enhanced 25-D3 binding affinity and crystallographic analyses. The deviation in specificity is due in part to unexpected altertion of its conformation, corresponding to a significant change of the orientation of an α-helix and an equally large movement of a loop, both of which flank the designed ligand-binding pocket. Those changes led to engineered protein constructs that exhibit significantly more contacts and complementarity towards the 25-D3 ligand than the initial designed protein had been predicted to form towards its intended THC ligand. Molecular dynamics simulations imply that the initial computationally designed mutations may contribute to the movement of the helix. These analyses collectively indicate that accurate prediction and control of backbone dynamics conformation, through a combination of improved conformational sampling and/or de novo structure design, represents a key area of further development for the design and optimization of engineered ligand-binding proteins.
Stabilizing antigenic proteins as vaccine immunogens or diagnostic reagents is a stringent case of protein engineering and design as the exterior surface must maintain recognition by receptor(s) and antigen-specific antibodies at multiple distinct epitopes. This is a challenge, as stability enhancing mutations must be focused on the protein core, whereas successful computational stabilization algorithms typically select mutations at solvent-facing positions. In this study, we report the stabilization of SARS-CoV-2 Wuhan Hu-1 Spike receptor binding domain using a combination of deep mutational scanning and computational design, including the FuncLib algorithm. Our most successful design encodes I358F, Y365W, T430I, and I513L receptor binding domain mutations, maintains recognition by the receptor ACE2 and a panel of different anti-receptor binding domain monoclonal antibodies, is between 1 and 2°C more thermally stable than the original receptor binding domain using a thermal shift assay, and is less proteolytically sensitive to chymotrypsin and thermolysin than the original receptor binding domain. Our approach could be applied to the computational stabilization of a wide range of proteins without requiring detailed knowledge of active sites or binding epitopes. We envision that this strategy may be particularly powerful for cases when there are multiple or unknown binding sites.
Split reporter proteins capable of self-association and reactivation have applications in biomedical research, but designing these proteins, especially the selection of appropriate split points, has been somewhat arbitrary. We describe a new methodology to facilitate generating split proteins using split GFP as a self-association module. We first inserted the entire GFP module at one of several candidate split points in the protein of interest, and chose clones that retained the GFP signal and high activity relative to the original protein. Once such chimeric clones were identified, a final pair of split proteins was generated by splitting the GFP-inserted chimera within the GFP domain. Applying this strategy to Renilla reniformis luciferase, we identified a new split point that gave 10 times more activity than the previous split point. The process of membrane fusion was monitored with high sensitivity using a new pair of split reporter proteins. We also successfully identified new split points for HaloTag protein and firefly luciferase, generating pairs of self-associating split proteins that recovered the functions of both GFP and the original protein. This simple method of screening will facilitate the designing of split proteins that are capable of self-association through the split GFP domains.
Genes encoding membrane proteins have been estimated to comprise as much as 30% of the human genome. Among these membrane, proteins are a large number of signaling receptors, transporters, ion channels and enzymes that are vital to cellular regulation, metabolism and homeostasis. While many membrane proteins are considered high-priority targets for drug design, there is a dearth of structural and biochemical information on them. This lack of information stems from the inherent insolubility and instability of transmembrane domains, which prevents easy obtainment of high-resolution crystals to specifically study structure-function relationships. In part, this lack of structures has greatly impeded our understanding in the field of membrane proteins. One method that can be used to enhance our understanding is directed evolution, a molecular biology method that mimics natural selection to engineer proteins that have specific phenotypes. It is a powerful technique that has considerable success with globular proteins, notably the engineering of protein therapeutics. With respect to transmembrane protein targets, this tool may be underutilized. Another powerful tool to investigate membrane protein structure-function relationships is computational modeling. This review will discuss these protein engineering methods and their tremendous potential in the study of membrane proteins.
After approximately 60 years of work, the protein folding problem has recently seen rapid advancement thanks to the inventions of AlphaFold and RoseTTAFold, which are machine-learning algorithms capable of reliably predicting protein structures from their sequences. A key component in their success was the inclusion of pairwise interaction information between residues. As research focus shifts towards developing algorithms to design and engineer binding proteins, it is likely that knowledge of interaction features at protein interfaces can improve predictions. Here, 574 protein complexes were analyzed to identify the stability features of their pairwise interactions, revealing that interactions between pre-stabilized residues are a selected feature in protein binding interfaces. In a retrospective analysis of 475 de novo designed binding proteins with an experimental success rate of 19%, inclusion of pairwise interaction pre-stabilization parameters increased the frequency of identifying experimentally successful binders to 40%.
Protein switches have potential applications as biosensors and selective protein therapeutics. Protein switches built by fusion of proteins with the prerequisite input and output functions are currently developed using an ad hoc process. A modular switch platform in which existing switches could be readily adapted to respond to any ligand would be advantageous. We investigated the feasibility of a modular protein switch platform based on fusions of the enzyme TEM-1 β-lactamase (BLA) with two different antibody mimetic proteins: designed ankyrin repeat proteins (DARPins) and monobodies. We created libraries of random insertions of the gene encoding BLA into genes encoding a DARPin or a monobody designed to bind maltose-binding protein (MBP). From these libraries, we used a genetic selection system for β-lactamase activity to identify genes that conferred MBP-dependent ampicillin resistance to Escherichia coli. Some of these selected genes encoded switch proteins whose enzymatic activity increased up to 14-fold in the presence of MBP. We next introduced mutations into the antibody mimetic domain of these switches that were known to cause binding to different ligands. To different degrees, introduction of the mutations resulted in switches with the desired specificity, illustrating the potential modularity of these platforms.
Erythropoietin (EPO) suppresses apoptosis and promotes survival by signaling through EPO-R/EPO-R on hematopoietic progenitors or EPO-R/CD131 on non-hematopoietic cells. However, EPO signaling through EPO-R/CD131 is controversial and there is no solved structure of a complex. Here, we constructed a structural model of EPO-R/CD131 and designed several anti-EPO-R, anti-CD131 bispecific proteins that selectively activate EPO-R/CD131. Treatment with these fusion proteins is sufficient to activate STAT5 phosphorylation downstream of EPO-R/CD131 without engaging EPO-R/EPO-R. We demonstrated that proteins with a tandem scFv or bispecific antibody format activate EPO-R/CD131, in contrast to an equimolar mixture of the individual scFvs. Finally, we explored the effect of modifications to binding domain arrangement and linker length and found results consistent with our structural model of an EPO-R/CD131 complex. These findings highlight the utility of bispecific scaffolds in the development of cytokine receptor agonists and provide a foundation for the study of EPO-R/CD131 biology and future clinical development.
During the past decades, advances in protein engineering have resulted in the development of variousin vitroselection techniques (e.g. phage display) to facilitate discovery of new and improved proteins. The methods are based on linkage between genotype and phenotype and are often performed in successive rounds of selection. Since the resulting output depends on the selection pressures used and the applied strategy, parameters in each round must be carefully considered. In addition, studies have reported biases that can cause enrichment of unwanted clones and/or low correlation between abundance in output and affinity. We have recently developed a selection method based on display of protein libraries onStaphylococcus carnosusand isolation of affinity proteins by fluorescence-activated cell sorting. Here, we compared duplicate selections for affinity maturation using equilibrium binding at different target concentrations and kinetic off-rate selection. The results showed that kinetic selection is efficient for isolation of high-affinity binders and that equilibrium selection at subnanomolar concentrations should be avoided. Furthermore, the reproducibility of the selection was high and a clear correlation was observed between enrichment and affinity. This work reports on the reproducibility of bacterial display in combination with FACS and provides insights into selection design to help guide the development of new affinity proteins.
ANTIC ALIGN: is an interactive software developed to simultaneously visualize, analyze and modify alignments of DNA and/or protein sequences that arise during combinatorial protein engineering, design and selection. ANTIC ALIGN: combines powerful functions known from currently available sequence analysis tools with unique features for protein engineering, in particular the possibility to display and manipulate nucleotide sequences and their translated amino acid sequences at the same time. ANTIC ALIGN: offers both template-based multiple sequence alignment (MSA), using the unmutated protein as reference, and conventional global alignment, to compare sequences that share an evolutionary relationship. The application of similarity-based clustering algorithms facilitates the identification of duplicates or of conserved sequence features among a set of selected clones. Imported nucleotide sequences from DNA sequence analysis are automatically translated into the corresponding amino acid sequences and displayed, offering numerous options for selecting reading frames, highlighting of sequence features and graphical layout of the MSA. The MSA complexity can be reduced by hiding the conserved nucleotide and/or amino acid residues, thus putting emphasis on the relevant mutated positions. ANTIC ALIGN: is also able to handle suppressed stop codons or even to incorporate non-natural amino acids into a coding sequence. We demonstrate crucial functions of ANTIC ALIGN: in an example of Anticalins selected from a lipocalin random library against the fibronectin extradomain B (ED-B), an established marker of tumor vasculature. Apart from engineered protein scaffolds, ANTIC ALIGN: provides a powerful tool in the area of antibody engineering and for directed enzyme evolution.