Computational Protein Engineering Research Articles

Automated optimization of the solubility of a hyper-stable α-amylase.

Most successes in computational protein engineering to date have focused on enhancing one biophysical trait, while multi-trait optimization remains a challenge. Different biophysical properties are often conflicting, as mutations that improve one tend to worsen the others. In this study, we explored the potential of an automated computational design strategy, called CamSol Combination, to optimize solubility and stability of enzymes without affecting their activity. Specifically, we focus on Bacillus licheniformis α-amylase (BLA), a hyper-stable enzyme that finds diverse application in industry and biotechnology. We validate the computational predictions by producing 10 BLA variants, including the wild-type (WT) and three designed models harbouring between 6 and 8 mutations each. Our results show that all three models have substantially improved relative solubility over the WT, unaffected catalytic rate and retained hyper-stability, supporting the algorithm's capacity to optimize enzymes. High stability and solubility embody enzymes with superior resilience to chemical and physical stresses, enhance manufacturability and allow for high-concentration formulations characterized by extended shelf lives. This ability to readily optimize solubility and stability of enzymes will enable the rapid and reliable generation of highly robust and versatile reagents, poised to contribute to advancements in diverse scientific and industrial domains.

Codon language embeddings provide strong signals for use in protein engineering

Protein representations from deep language models have yielded state-of-the-art performance across many tasks in computational protein engineering. In recent years, progress has primarily focused on parameter count, with recent models’ capacities surpassing the size of the very datasets they were trained on. Here we propose an alternative direction. We show that large language models trained on codons, instead of amino acid sequences, provide high-quality representations that outperform comparable state-of-the-art models across a variety of tasks. In some tasks, such as species recognition, prediction of protein and transcript abundance or melting point estimation, we show that a language model trained on codons outperforms every other published protein language model, including some that contain over 50 times more parameters. These results indicate that, in addition to commonly studied scale and model complexity, the information content of biological data provides an orthogonal direction to improve the power of machine learning in biology.

Computational protein engineering for systematic enhancement of crystallization propensity

Computational protein engineering for systematic enhancement of crystallization propensity

Counteraction of stability-activity trade-off of Nattokinase through flexible region shifting

The widespread trade-off between stability and activity severely limits enzyme evolution. Although some progresses have been made to overcome this limitation, the counteraction mechanism for enzyme stability-activity trade-off remains obscure. Here, we clarified the counteraction mechanism of the Nattokinase stability-activity trade-off. A combinatorial mutant M4 was obtained by multi-strategy engineering, exhibiting a 20.7-fold improved half-life; meanwhile, the catalytic efficiency was doubled. Molecular dynamics simulation revealed that an obvious flexible region shifting in the structure of mutant M4 was occurred. The flexible region shifting which contributed to maintain the global structural flexibility, was considered to be the key factor for counteracting the stability-activity trade-off. Further analysis illustrated that the flexible region shifting was driven by region dynamical networks reshaping. This work provided deep insight into the counteraction mechanism of enzyme stability-activity trade-off, suggesting that flexible region shifting would be an effective strategy for enzyme evolution through computational protein engineering.

Reducing the Immunogenicity of Pulchellin A-Chain, Ribosome-Inactivating Protein Type 2, by Computational Protein Engineering for Potential New Immunotoxins

Pulchellin is a plant biotoxin categorized as a type 2 ribosome-inactivating protein (RIPs) which potentially kills cells at very low concentrations. Biotoxins serve as targeting immunotoxins (IT), consisting of antibodies conjugated to toxins. ITs have two independent protein components, a human antibody and a toxin with a bacterial or plant source; therefore, they pose unique setbacks in immunogenicity. To overcome this issue, the engineering of epitopes is one of the beneficial methods to elicit an immunological response. Here, we predicted the tertiary structure of the pulchellin A-chain (PAC) using five common powerful servers and adopted the best model after refining. Then, predicted structure using four distinct computational approaches identified conformational B-cell epitopes. This approach identified some amino acids as a potential for lowering immunogenicity by point mutation. All mutations were then applied to generate a model of pulchellin containing all mutations (so-called PAM). Mutants’ immunogenicity was assessed and compared to the wild type as well as other mutant characteristics, including stability and compactness, were computationally examined in addition to immunogenicity. The findings revealed a reduction in immunogenicity in all mutants and significantly in N146V and R149A. Furthermore, all mutants demonstrated remarkable stability and validity in Molecular Dynamic (MD) simulations. During docking and simulations, the most homologous toxin to pulchellin, Abrin-A was applied as a control. In addition, the toxin candidate containing all mutations (PAM) disclosed a high level of stability, making it a potential model for experimental deployment. In conclusion, by eliminating B-cell epitopes, our computational approach provides a potential less immunogenic IT based on PAC.

Genetically Engineered MRI-Trackable Extracellular Vesicles as SARS-CoV-2 Mimetics for Mapping ACE2 Binding In Vivo.

The elucidation of viral-receptor interactions and an understanding of virus-spreading mechanisms are of great importance, particularly in the era of a pandemic. Indeed, advances in computational chemistry, synthetic biology, and protein engineering have allowed precise prediction and characterization of such interactions. Nevertheless, the hazards of the infectiousness of viruses, their rapid mutagenesis, and the need to study viral-receptor interactions in a complex in vivo setup call for further developments. Here, we show the development of biocompatible genetically engineered extracellular vesicles (EVs) that display the receptor binding domain (RBD) of SARS-CoV-2 on their surface as coronavirus mimetics (EVsRBD). Loading EVsRBD with iron oxide nanoparticles makes them MRI-visible and, thus, allows mapping of the binding of RBD to ACE2 receptors noninvasively in live subjects. Moreover, we show that EVsRBD can be modified to display mutants of the RBD of SARS-CoV-2, allowing rapid screening of currently raised or predicted variants of the virus. The proposed platform thus shows relevance and cruciality in the examination of quickly evolving pathogenic viruses in an adjustable, fast, and safe manner. Relying on MRI for visualization, the presented approach could be considered in the future to map ligand-receptor binding events in deep tissues, which are not accessible to luminescence-based imaging.

Combining chemistry and protein engineering for new-to-nature biocatalysis

Biocatalysis, the application of enzymes to solve synthetic problems of human import, has blossomed into a powerful technology for chemical innovation. In the past decade, a threefold partnership, where nature provides blueprints for enzymatic catalysis, chemists introduce innovative activity modes with abiological substrates, and protein engineers develop new tools and algorithms to tune and improve enzymatic function, has unveiled the frontier of new-to-nature enzyme catalysis. In this perspective, we highlight examples of interdisciplinary studies which have helped to expand the scope of biocatalysis, including concepts of enzymatic versatility explored through the lens of biomimicry, to achieve both activities and selectivities that are not currently possible with chemocatalysis. We indicate how modern tools, such as directed evolution, computational protein design and machine learning-based protein engineering methods, have already impacted and will continue to influence enzyme engineering for new abiological transformations. A sustained collaborative effort across disciplines is anticipated to spur further advances in biocatalysis in the coming years.

Concerns with computational protein engineering programmes IPRO and OptMAVEn and metabolic pathway engineering programme optStoic.

It has become customary in engineering to require a modelling component in research endeavours. In addition, as the code for these models becomes more byzantine in complexity, it is difficult for reviewers and readers to discern their value and understand the underlying code. This opinion piece summarizes the negative experience of the author with the IPRO and OptMAVEn computational protein engineering models as well as problems with the optStoic metabolic pathway model. In our hands, these models often fail to predict reliable ways to engineer proteins and metabolic pathways.

Synthetic Biology: A Very Short Introduction by Jamie Davies

When Mary Shelley wrote Frankenstein in 1818, she probably did not see synthetic biology coming. Shelley’s unorthodox use of galvanism and vivisection may have become computational biology, protein engineering and genome editing, but – many scientists would argue – has lost none of its gothic charm. As a truly interdisciplinary field, synthetic biology has been around – at least in its modern molecular incarnation – since Stanley Cohen and Herbert Boyer genetically engineered the first plasmid in 1973. In Synthetic Biology: A Very Short Introduction, Jamie Davies starts us off on the ‘creation of new living systems by design’ with one of many predictable clashes – between the economic forces who swear by the promise of synthetic biology and the bioethicists whose reservations keep this field at arm’s length. As most books in the Very Short Introduction series, Davies’ pocket-size guide provides readers with a thorough and engaging introduction to the biology and methods of synthetic biology while touring its numerous applications, from saving our oceans to changing the way we produce drugs. Davies does a great job of sketching a field concerned with the deliberate redesigning of life. As with other great introductions, by the end of this book, one cannot help oneself, but wanting to know more.

GRAPE, a greedy accumulated strategy for computational protein engineering.

Nature harbors fascinating enzymatic catalysts with high efficiency, chemo-, regio- and stereoselectivity. However, the insufficient stability of the enzymes often prevents their widespread utilization for industrial processes. Not content with the finite repertoire of naturally occurring enzymes, protein engineering holds promises to extend the applications of the improved enzymes with desired physical and catalytic properties. Herein, we devised a computational strategy (greedy accumulated strategy for protein engineering, GRAPE) to enhance the thermostability of enzymes. Through scanning of all point mutations of the structural and evolutionary consensus analysis, a library containing fewer than 100 mutations was established for characterization. After preliminary experimental verification, effective mutations are clustered in a multidimensional physical property space and then accumulated via the greedy algorithm to produce the final designed enzyme. Using the recently reported IsPETase from Ideonella sakaiensis that decomposes PET under ambient temperatures as a starting point, we adopted the GRAPE strategy to come up with a DuraPETase (TM=77°C, raised by 31°C) which showed drastically enhanced degradation performance (300-fold) on semicrystalline PET films at 40°C.

Increased Tumor Penetration of Single-Domain Antibody-Drug Conjugates Improves In Vivo Efficacy in Prostate Cancer Models.

Targeted delivery of chemotherapeutics aims to increase efficacy and lower toxicity by concentrating drugs at the site-of-action, a method embodied by the seven current FDA-approved antibody-drug conjugates (ADC). However, a variety of pharmacokinetic challenges result in relatively narrow therapeutic windows for these agents, hampering the development of new drugs. Here, we use a series of prostate-specific membrane antigen-binding single-domain (Humabody) ADC constructs to demonstrate that tissue penetration of protein-drug conjugates plays a major role in therapeutic efficacy. Counterintuitively, a construct with lower in vitro potency resulted in higher in vivo efficacy than other protein-drug conjugates. Biodistribution data, tumor histology images, spheroid experiments, in vivo single-cell measurements, and computational results demonstrate that a smaller size and slower internalization rate enabled higher tissue penetration and more cell killing. The results also illustrate the benefits of linking an albumin-binding domain to the single-domain ADCs. A construct lacking an albumin-binding domain was rapidly cleared, leading to lower tumor uptake (%ID/g) and decreased in vivo efficacy. In conclusion, these results provide evidence that reaching the maximum number of cells with a lethal payload dose correlates more strongly with in vivo efficacy than total tumor uptake or in vitro potency alone for these protein-drug conjugates. Computational modeling and protein engineering can be used to custom design an optimal framework for controlling internalization, clearance, and tissue penetration to maximize cell killing. SIGNIFICANCE: A mechanistic study of protein-drug conjugates demonstrates that a lower potency compound is more effective in vivo than other agents with equal tumor uptake due to improved tissue penetration and cellular distribution.

Protein engineering: from directed evolution to computational design

By constructing mutant libraries and utilizing high-throughput screening methods, directed evolution has emerged as the most popular strategy for protein design nowadays. In the past decade, taking advantages of computer performance and algorithms, computer-assisted protein design has rapidly developed and become a powerful method of protein engineering. Based on the simulation of protein structure and calculation of energy function, computational design can alter the substrate specificity and improve the thermostability of enzymes, as well as de novo design of artificial enzymes with expected functions. Recently, machine learning and other artificial intelligence technologies have also been applied to computational protein engineering, resulting in a series of remarkable applications. Along the lines of protein engineering, this paper reviews the progress and applications of computer-assisted protein design, and current trends and outlooks of the development.

CaverDock: A Novel Method for the Fast Analysis of Ligand Transport.

Here we present a novel method for the analysis of transport processes in proteins and its implementation called CaverDock. Our method is based on a modified molecular docking algorithm. It iteratively places the ligand along the access tunnel in such a way that the ligand movement is contiguous and the energy is minimized. The result of CaverDock calculation is a ligand trajectory and an energy profile of transport process. CaverDock uses the modified docking program Autodock Vina for molecular docking and implements a parallel heuristic algorithm for searching the space of possible trajectories. Our method lies in between the geometrical approaches and molecular dynamics simulations. Contrary to the geometrical methods, it provides an evaluation of chemical forces. However, it is far less computationally demanding and easier to set up compared to molecular dynamics simulations. CaverDock will find a broad use in the fields of computational enzymology, drug design, and protein engineering. The software is available free of charge to the academic users at https://loschmidt.chemi.muni.cz/caverdock/.

Exploiting correlated molecular-dynamics networks to counteract enzyme activity–stability trade-off

The directed evolution of enzymes for improved activity or substrate specificity commonly leads to a trade-off in stability. We have identified an activity-stability trade-off and a loss in unfolding cooperativity for a variant (3M) of Escherichia coli transketolase (TK) engineered to accept aromatic substrates. Molecular dynamics simulations of 3M revealed increased flexibility in several interconnected active-site regions that also form part of the dimer interface. Mutating the newly flexible active-site residues to regain stability risked losing the new activity. We hypothesized that stabilizing mutations could be targeted to residues outside of the active site, whose dynamics were correlated with the newly flexible active-site residues. We previously stabilized WT TK by targeting mutations to highly flexible regions. These regions were much less flexible in 3M and would not have been selected a priori as targets using the same strategy based on flexibility alone. However, their dynamics were highly correlated with the newly flexible active-site regions of 3M. Introducing the previous mutations into 3M reestablished the WT level of stability and unfolding cooperativity, giving a 10.8-fold improved half-life at 55 °C, and increased midpoint and aggregation onset temperatures by 3 °C and 4.3 °C, respectively. Even the activity toward aromatic aldehydes increased up to threefold. Molecular dynamics simulations confirmed that the mutations rigidified the active-site via the correlated network. This work provides insights into the impact of rigidifying mutations within highly correlated dynamic networks that could also be useful for developing improved computational protein engineering strategies.

Structure Guided Engineering of Highly Specific Chimeric Antigen Receptors for the Treatment of T Cell Lymphomas

IntroductionMature T cell lymphomas are aggressive, treatment resistant cancers that are associated with poor prognosis. Clinical application of immunotherapeutic approaches has been limited by a lack of target antigens that discriminate malignant from healthy T cells. Unlike B cell depletion, pan-T cell aplasia is prohibitively toxic. Previously we reported a targeting strategy based on the mutually exclusive expression of T cell receptor beta-chain constant domains 1 and 2 (TRBC1 and TRBC2). We identified an antibody with unique TRBC1 specificity and demonstrated that anti TRBC1 chimeric antigen receptor (CAR) T cells can ablate cells expressing TRBC1 TCRs while sparing those expressing TRBC2 TCRs. A phase I clinical study investigating the efficacy of our TRBC1 CAR is ongoing.T cell malignancies are clonal, and the ratio of TRBC2 to TRBC1 expressing lymphoma cases is predicted to be approximately 2:1. To treat all cases of T cell lymphoma, a CAR that targets TRBC2 is needed. TRBC1 and 2 are highly homologous. Structural studies suggest that amino acid inversions at positions 4 and 5 of the constant beta chain provide an accessible discriminating portion between the two proteins.Given the structural similarities between TRBC1 and TRBC2 and our characterized binder against TRBC1, we explored generating antibodies with specificity towards TRBC2 via a structure guided computational biology approach; engineering the previously identified TRBC1 antibody and reversing its specificity such that it recognised TRBC2.ResultsThe crystal structure of the TRBC1 specific monoclonal antibody was solved in complex with a TRBC1-TCR to 2.4Å, Figure 1. Through computational biology and protein engineering we rationally designed a mutant version of TRBC1 binder that was specific for TRBC2 and had a 1000 fold decreased affinity towards TRBC1. Flow cytometry analysis of the TRBC2 specific antibody demonstrated the ability to bind to T-cells expressing TRBC2 TCRs. We further showed that the engineered antibody retained favourable biophysical characteristics with high stability (Fab Tm > 65oC) and low aggregation propensity (>99% monomer).We used the engineered monoclonal antibody to generate a 2nd generation anti-TRBC2 CAR. We demonstrated that our anti-TRBC2 CAR showed specificity, cytokine release and cytotoxicity in 72hr co-cultures against TRBC2+ cell lines but not TRBC1+ cell lines or cell lines that did not express TCR on the surface. Anti-TRBC2 CAR T-cells also demonstrated proliferative capacity in long-term co-culture assays.ConclusionsWe have utilised structural biology and rational protein design to generate CAR-T cells capable of specifically targeting TRBC2. The combination of TRBC1 and 2 targeting CAR-T cell products with a patient stratification companion diagnostic assay offers a therapeutic strategy for the treatment of a wide range of, otherwise untreatable, T-cell lymphomas.Figure 1. A. Structural interface between TCR Beta and Fab fragment of TRBC1 specific antibody. B and C. CDR fold of TRBC1 binder and 90o rotation. [Display omitted] DisclosuresOnuoha:Autolus Ltd: Employment, Equity Ownership, Patents & Royalties. Ferrari:Autolus Ltd: Employment, Equity Ownership. Bulek:Autolus Ltd: Employment, Equity Ownership, Patents & Royalties. Bughda:Autolus Ltd: Employment, Equity Ownership, Patents & Royalties. Manzoor:Autolus Ltd: Employment, Equity Ownership, Patents & Royalties. Srivastava:Autolus Ltd: Employment, Equity Ownership. Ma:Autolus Ltd: Employment. Karattil:Autolus Ltd: Employment, Equity Ownership, Patents & Royalties. Kinna:Autolus Ltd: Employment, Equity Ownership. Thomas:Autolus Ltd: Employment, Equity Ownership, Patents & Royalties. Cordoba:Autolus Ltd: Employment; Autolus Ltd: Patents & Royalties; Autolus Ltd: Equity Ownership. Maciocia:Autolus: Equity Ownership, Patents & Royalties: UCLB. Pule:Autolus Ltd: Employment, Equity Ownership, Other: Salary contribution paid for by Autolus, Research Funding; University College London: Patents & Royalties: Patent with rights to Royalty share through UCL.

Biomimetic engineering of conductive curli protein films

Bioelectronic systems derived from peptides and proteins are of particular interest for fabricating novel flexible, biocompatible and bioactive devices. These synthetic or recombinant systems designed for mediating electron transport often mimic the proteinaceous appendages of naturally occurring electroactive bacteria. Drawing inspiration from such conductive proteins with a high content of aromatic residues, we have engineered a fibrous protein scaffold, curli fibers produced by Escherichia coli bacteria, to enable long-range electron transport. We report the genetic engineering and characterization of curli fibers containing aromatic residues of different nature, with defined spatial positioning, and with varying content on single self-assembling CsgA curli subunits. Our results demonstrate the impressive versatility of the CsgA protein for genetically engineering protein-based materials with new functions. Through a scalable purification process, we show that macroscopic gels and films can be produced, with engineered thin films exhibiting a greater conductivity compared with wild-type curli films. We anticipate that this engineered conductive scaffold, and our approach that combines computational modeling, protein engineering, and biosynthetic manufacture will contribute to the improvement of a range of useful bio-hybrid technologies.

Enantio-, Regio-, and Chemoselective Lipase-Catalyzed Polymer Synthesis.

In contrast to chemical routes, enzymatic polymerization possesses favorable characteristics of mild reaction conditions, few by-products, and high activity toward cyclic lactones which make it a promising technique for constructing polymeric materials. Meanwhile, it can avoid the trace residue of metallic catalysts and potential toxicity, and thus exhibits great potential in the biomedical fields. More importantly, lipase-catalyzed polymer synthesis usually shows favorable enantio-, regio-, and chemoselectivity. Here, the history and recent developments in lipase-catalyzed selective polymerization for constructing polymers with unique structures and properties are highlighted. In particular, the synthesis of polymeric materials which are difficult to prepare in a chemical route and the construction of polymers through the combination of selective enzymatic and chemical methods are focused. In addition, the future direction is proposed especially based on the rapid developments in computational chemistry and protein engineering techniques.

Enzymatic clickable functionalization of peptides via computationally engineered peptide amidase

Directed peptides C-terminal modification enabled by the engineered biomolecular catalyst – peptide amidase 12B has been achieved via computational protein engineering. The engineered enzyme exhibits great promising potential in the C-terminal modification of opioid peptides using prop-2-yn-1-amine (PYA) or prop-2-en-1-amine (PEA) as the nucleophile. A variety of opioid peptides could be readily functionalized at the C-terminal chain in high yield in a mild and selective manner. Notably, modified opioid peptides bearing alkynyl moiety could be further functionalized through well-established click reaction.

Genetic Algorithm Based Design and Experimental Characterization of a Highly Thermostable Metalloprotein

The development of thermostable and solvent-tolerant metalloproteins is a long-sought goal for many applications in synthetic biology and biotechnology. In this work, we were able to engineer a highly thermostable and organic solvent-stable metallo variant of the B1 domain of protein G (GB1) with a tetrahedral zinc binding site reminiscent of the one of thermolysin. Promising candidates were designed computationally by applying a protocol based on classical and first-principles molecular dynamics simulations in combination with genetic algorithm optimization. The most promising of the computationally predicted mutants was expressed and structurally characterized and yielded a highly thermostable protein. The experimental results thus confirm the predictive power of the applied computational protein engineering approach for the de novo design of highly stable metalloproteins.

Construction of ATP-Switched Allosteric Antioxidant Selenoenzyme

Rational redesign of allosteric protein offers an efficient strategy to develop switchable biocatalysts. By combining the computational design and protein engineering, a glutathione peroxidase (GPx)-like active center that contains the catalytic selenocysteine (Sec) residue and substrate-binding Arg residue was precisely incorporated into the allosteric domain of adenylate kinase (AKe). The engineered selenoenzyme shows not only high GPx activity but also adenosine triphosphate (ATP)-responsive catalytic property, which is regulated by its opened to closed conformational change upon ATP binding. Theoretical and mutational analysis reveals that the synergistic effect of electrostatic interactions and van der Waals (vdW) interactions for substrate recognition is a major contribution to the high activity. The mitochondrial oxidative damage experiment further demonstrated its antioxidant ability at the subcellular level, offering a potential application toward controllable catalysis in vivo.