Robust prediction of relative binding energies for protein-protein complex mutations using free energy perturbation calculations.

Computational free energy-based methods have the potential to significantly improve throughput and decrease costs of protein design efforts. Such methods must reach a high level of reliability, accuracy, and automation to be effectively deployed in practical industrial settings in a way that impacts protein design projects. Here, we present a benchmark study for the calculation of relative changes in protein-protein binding affinity for single point mutations across a variety of systems from the literature, using free energy perturbation (FEP+) calculations. We describe a method for robust treatment of alternate protonation states for titratable amino acids, which yields improved correlation with and reduced error compared to experimental binding free energies. Following careful analysis of the largest outlier cases in our dataset, we assess limitations of the default FEP+ protocols and introduce an automated script which identifies probable outlier cases that may require additional scrutiny and calculates an empirical correction for a subset of charge-related outliers. Through a series of three additional case study systems, we discuss how protein FEP+ can be applied to real-world protein design projects, and suggest areas of further study.

Robust prediction of relative binding energies for protein-protein complex mutations using free energy perturbation calculations.

Computational free energy-based methods have the potential to significantly improve throughput and decrease costs of protein design efforts. Such methods must reach a high level of reliability, accuracy, and automation to be effectively deployed in practical industrial settings in a way that impacts protein design projects. Here, we present a benchmark study for the calculation of relative changes in protein-protein binding affinity for single point mutations across a variety of systems from the literature, using free energy perturbation (FEP+) calculations. We describe a method for robust treatment of alternate protonation states for titratable amino acids, which yields improved correlation with and reduced error compared to experimental binding free energies. Following careful analysis of the largest outlier cases in our dataset, we assess limitations of the default FEP+ protocols and introduce an automated script which identifies probable outlier cases that may require additional scrutiny and calculates an empirical correction for a subset of charge-related outliers. Through a series of three additional case study systems, we discuss how protein FEP+ can be applied to real-world protein design projects, and suggest areas of further study.

Discovery and Characterization of a Bicyclic Peptide (Bicycle) Binder to Thymic Stromal Lymphopoietin.

Thymic stromal lymphopoietin (TSLP) is an epithelial-derived pro-inflammatory cytokine involved in the development of asthma and other atopic diseases. We used Bicycle Therapeutics' proprietary phage display platform to identify bicyclic peptides (Bicycles) with high affinity for TSLP, a target that is difficult to drug with conventional small molecules due to the extended protein-protein interactions it forms with both receptors. The hit series was shown to bind to TSLP in a hotspot, that is also used by IL-7Rα. Guided by the first X-ray crystal structure of a small peptide binding to TSLP and the identification of key metabolites, we were able to improve the proteolytic stability of this series in lung S9 fractions without sacrificing binding affinity. This resulted in the potent Bicycle 46 with nanomolar affinity to TSLP (KD = 13 nM), low plasma clearance of 6.4 mL/min/kg, and an effective half-life of 46 min after intravenous dosing to rats.

Sequence-based prediction of the intrinsic solubility of peptides containing non-natural amino acids.

Non-natural amino acids are increasingly used as building blocks in the development of peptide-based drugs as they expand the available chemical space to tailor function, half-life and other key properties. However, while the chemical space of modified amino acids (mAAs) such as residues containing post-translational modifications (PTMs) is potentially vast, experimental methods for measuring the developability properties of mAA-containing peptides are expensive and time consuming. To facilitate developability programs through computational methods, we present CamSol-PTM, a method that enables the fast and reliable sequence-based prediction of the intrinsic solubility of mAA-containing peptides in aqueous solution at room temperature. From a computational screening of 50,000 mAA-containing variants of three peptides, we selected five different small-size mAAs for a total number of 37 peptide variants for experimental validation. We demonstrate the accuracy of the predictions by comparing the calculated and experimental solubility values. Our results indicate that the computational screening of mAA-containing peptides can extend by over four orders of magnitude the ability to explore the solubility chemical space of peptides and confirm that our method can accurately assess the solubility of peptides containing mAAs. This method is available as a web server at https://www-cohsoftware.ch.cam.ac.uk/index.php/camsolptm .

Structure-Based Design of Small Imine Reductase Panels for Target Substrates.

Biocatalysis is important in the discovery, development, and manufacture of pharmaceuticals. However, the identification of enzymes for target transformations of interest requires major screening efforts. Here, we report a structure-based computational workflow to prioritize protein sequences by a score based on predicted activities on substrates, thereby reducing a resource-intensive laboratory-based biocatalyst screening. We selected imine reductases (IREDs) as a class of biocatalysts to illustrate the application of the computational workflow termed IREDFisher. Validation by using published data showed that IREDFisher can retrieve the best enzymes and increase the hit rate by identifying the top 20 ranked sequences. The power of IREDFisher is confirmed by computationally screening 1400 sequences for chosen reductive amination reactions with different levels of complexity. Highly active IREDs were identified by only testing 20 samples in vitro. Our speed test shows that it only takes 90 min to rank 85 sequences from user input and 30 min for the established IREDFisher database containing 591 IRED sequences. IREDFisher is available as a user-friendly web interface (https://enzymeevolver.com/IREDFisher). IREDFisher enables the rapid discovery of IREDs for applications in synthesis and directed evolution studies, with minimal time and resource expenditure. Future use of the workflow with other enzyme families could be implemented following the modification of the workflow scoring function.

Analyzing proteolytic stability and metabolic hotspots of therapeutic peptides in two rodent pulmonary fluids.

Peptides and peptide drug conjugates are emerging modalities to treat pulmonary diseases. Peptides are susceptible to proteolytic cleavage. Expression levels of specific proteases in the lung can be significantly increased in disease state and may lead to exaggerated peptide proteolysis. To support optimization of peptides for inhaled administration, we have recently reported a streamlined high-throughput LC-HRMS protocol to determine enzymatic protease stability of peptides. This method has now been complemented with profiling of peptide metabolic stability in two respiratory fluids, a lung supernatant (lung S9) and a bronchioalveolar lavage fluid (BALF) taken from rats. We have tested a set of 28 peptides with high structural diversity, analyzed the whole data set for formed metabolites, and identified the differences of cleavage pattern in the two test fluids. Comparison of our experimental results and literature-derived cleavage site estimates based on e.g. MEROPS show significant differences for a number of peptides. This indicates the need for an experimental workflow using both protease panels and testing of metabolic stability in lung fluid (BALF) to guide peptide optimization and selection of peptides for inhaled in vivo PK/PD studies in our drug discovery projects.

Mechanism of Action of Flavin-Dependent Halogenases.

To rationally engineer the substrate scope and selectivity of flavin-dependent halogenases (FDHs), it is essential to first understand the reaction mechanism and substrate interactions in the active site. FDHs have long been known to achieve regioselectivity through an electrophilic aromatic substitution at C7 of the natural substrate Trp, but the precise role of a key active-site Lys residue remains ambiguous. Formation of hypochlorous acid (HOCl) at the cofactor-binding site is achieved by the direct reaction of molecular oxygen and a single chloride ion with reduced FAD and flavin hydroxide, respectively. HOCl is then guided 10 Å into the halogenation active site. Lys79, located in this site, has been proposed to direct HOCl toward Trp C7 through hydrogen bonding or a direct reaction with HOCl to form an -NH2Cl+ intermediate. Here, we present the most likely mechanism for halogenation based on molecular dynamics (MD) simulations and active-site density functional theory "cluster" models of FDH PrnA in complex with its native substrate l-tryptophan, hypochlorous acid, and the FAD cofactor. MD simulations with different protonation states for key active-site residues suggest that Lys79 directs HOCl through hydrogen bonding, which is confirmed by calculations of the reaction profiles for both proposed mechanisms.

Virtual Screening Expands the Non-Natural Amino Acid Palette for Peptide Optimization.

Peptides are an important modality in drug discovery. While current peptide optimization focuses predominantly on the small number of natural and commercially available non-natural amino acids, the chemical spaces available for small molecule drug discovery are in the billions of molecules. In the present study, we describe the development of a large virtual library of readily synthesizable non-natural amino acids that can power the virtual screening protocols and aid in peptide optimization. To that end, we enumerated nearly 380 thousand amino acids and demonstrated their vast chemical diversity compared to the 20 natural and commercial residues. Furthermore, we selected a diverse ten thousand amino acid subset to validate our virtual screening workflow on the Keap1-Neh2 complex model system. Through in silico mutations of Neh2 peptide residues to those from the virtual library, our docking-based protocol identified a number of possible solutions with a significantly higher predicted affinity toward the Keap1 protein. This protocol demonstrates that the non-natural amino acid chemical space can be massively extended and virtually screened with a reasonable computational cost.

Machine learning/molecular dynamic protein structure prediction approach to investigate the protein conformational ensemble.

Proteins exist in several different conformations. These structural changes are often associated with fluctuations at the residue level. Recent findings show that co-evolutionary analysis coupled with machine-learning techniques improves the precision by providing quantitative distance predictions between pairs of residues. The predicted statistical distance distribution from Multi Sequence Analysis reveals the presence of different local maxima suggesting the flexibility of key residue pairs. Here we investigate the ability of the residue-residue distance prediction to provide insights into the protein conformational ensemble. We combine deep learning approaches with mechanistic modeling to a set of proteins that experimentally showed conformational changes. The predicted protein models were filtered based on energy scores, RMSD clustering, and the centroids selected as the lowest energy structure per cluster. These models were compared to the experimental-Molecular Dynamics (MD) relaxed structure by analyzing the backbone residue torsional distribution and the sidechain orientations. Our pipeline allows to retrieve the experimental structural dynamics experimentally represented by different X-ray conformations for the same sequence as well the conformational space observed with the MD simulations. We show the potential correlation between the experimental structure dynamics and the predicted model ensemble demonstrating the susceptibility of the current state-of-the-art methods in protein folding and dynamics prediction and pointing out the areas of improvement.

How well does molecular simulation reproduce environment-specific conformations of the intrinsically disordered peptides PLP, TP2 and ONEG?

Understanding the conformational ensembles of intrinsically disordered proteins and peptides (IDPs) in their various biological environments is essential for understanding their mechanisms and functional roles in the proteome, leading to a greater knowledge of, and potential treatments for, a broad range of diseases. To determine whether molecular simulation is able to generate accurate conformational ensembles of IDPs, we explore the structural landscape of the PLP peptide (an intrinsically disordered region of the proteolipid membrane protein) in aqueous and membrane-mimicking solvents, using replica exchange with solute scaling (REST2), and examine the ability of four force fields (ff14SB, ff14IDPSFF, CHARMM36 and CHARMM36m) to reproduce literature circular dichroism (CD) data. Results from variable temperature (VT) 1H and Rotating frame Overhauser Effect SpectroscopY (ROESY) nuclear magnetic resonance (NMR) experiments are also presented and are consistent with the structural observations obtained from the simulations and CD. We also apply the optimum simulation protocol to TP2 and ONEG (a cell-penetrating peptide (CPP) and a negative control peptide, respectively) to gain insight into the structural differences that may account for the observed difference in their membrane-penetrating abilities. Of the tested force fields, we find that CHARMM36 and CHARMM36m are best suited to the study of IDPs, and accurately predict a disordered to helical conformational transition of the PLP peptide accompanying the change from aqueous to membrane-mimicking solvents. We also identify an α-helical structure of TP2 in the membrane-mimicking solvents and provide a discussion of the mechanistic implications of this observation with reference to the previous literature on the peptide. From these results, we recommend the use of CHARMM36m with the REST2 protocol for the study of environment-specific IDP conformations. We believe that the simulation protocol will allow the study of a broad range of IDPs that undergo conformational transitions in different biological environments.

Molecular Engineering of Efficacious Mono-Valent Ultra-Long Acting Two-Chain Insulin-Fc Conjugates.

Here, we describe molecular engineering of monovalent ultra-long acting two-chain insulin-Fc conjugates. Insulin-Fc conjugates were synthesized using trifunctional linkers with one amino reactive group for reaction with a lysine residue of insulin and two thiol reactive groups used for re-bridging of a disulfide bond within the Fc molecule. The ultra-long pharmacokinetic profile of the insulin-Fc conjugates was the result of concertedly slowing insulin receptor-mediated clearance by (1) introduction of amino acid substitutions that lowered the insulin receptor affinity and (2) conjugating insulin to the Fc element. Fc conjugation leads to recycling by the neonatal Fc receptor and increase in the molecular size, both contributing to the ultra-long pharmacokinetic and pharmacodynamic profiles.

Automated high-throughput in vitro assays to identify metabolic hotspots and protease stability of structurally diverse, pharmacologically active peptides for inhalation.

The inhalation of peptides comes with the advantage of directly targeting the lung as tissue of interest. However, peptides are often rapidly metabolized in lung tissue through proteolytic cleavage. We have developed an assay workflow to obtain half-life and metabolite ID data for peptides incubated with four proteases abundant in lungs of asthma and COPD patients. The assay system has been validated using 28 structurally diverse linear and cyclic peptides with a molecular weight between 708 and 5808 Da. Experimental conditions for incubation, sample preparation, chromatography, data acquisition and analysis are compatible with the required throughput in early stage peptide projects. Together with co-crystal structures and Ala scans, we are using the described assay workflow to guide the first chemical modifications of peptide hits in early respiratory drug discovery projects.

Impact of PROTAC Linker Plasticity on the Solution Conformations and Dissociation of the Ternary Complex.

The conformational behavior of a small molecule free in solution is important to understand the free energy of binding to its target. This could be of special interest for proteolysis-targeting chimeras (PROTACs) due to their often flexible and lengthy linkers and the need to induce a ternary complex. Here, we report on the molecular dynamics (MD) simulations of two PROTACs, MZ1 and dBET6, revealing different linker conformational behaviors. The simulation of MZ1 in dimethyl sulfoxide (DMSO) agrees well with the nuclear magnetic resonance study, providing strong support for the relevance of our simulations. To further understand the role of linker plasticity in the formation of a ternary complex, the dissociation of the complex von Hippel-Lindau-MZ1-BRD4 is investigated in detail by steered simulations and is shown to follow a two-step pathway. Interestingly, both MZ1 and dBET6 display in water, a tendency toward an intramolecular lipophilic interaction between the two warheads. The hydrophobic contact of the two warheads would prevent them from binding to their respective proteins and might have an effect on the efficacy of induced cellular protein degradation. However, conformations featuring this hydrophobic contact of the two warheads are calculated to be marginally more favorable.

General Protocol for Constructing Molecular Models of Nanodiscs.

Nanodisc technology is increasingly being applied for structural and biophysical studies of membrane proteins. In this work, we present a general protocol for constructing molecular models of nanodiscs for molecular dynamics simulations. The protocol is written in python and based on geometric equations, making it fast and easy to modify, enabling automation and customization of nanodiscs in silico. The novelty being the ability to construct any membrane scaffold protein (MSP) variant fast and easy given only an input sequence. We validated and tested the protocol by simulating seven different nanodiscs of various sizes and with different membrane scaffold proteins, both circularized and noncircularized. The structural and biophysical properties were analyzed and shown to be in good agreement with previously reported experimental data and simulation studies.

Glucocorticoid receptor Thr524 phosphorylation by MINK1 induces interactions with 14-3-3 protein regulators.

The glucocorticoid receptor (GR) is a ligand-dependent transcription factor that plays a central role in inflammation. The GR activity is also modulated via protein-protein interactions, including binding of 14-3-3 proteins induced by GR phosphorylation. However, the specific phosphorylation sites on the GR that trigger these interactions and their functional consequences are less clear. Hence, we sought to examine this system in more detail. We used phosphorylated GR peptides, biophysical studies, and X-ray crystallography to identify key residues within the ligand-binding domain of the GR, T524 and S617, whose phosphorylation results in binding of the representative 14-3-3 protein 14-3-3ζ. A kinase screen identified misshapen-like kinase 1 (MINK1) as responsible for phosphorylating T524 and Rho-associated protein kinase 1 for phosphorylating S617; cell-based approaches confirmed the importance of both GR phosphosites and MINK1 but not Rho-associated protein kinase 1 alone in inducing GR-14-3-3 binding. Together our results provide molecular-level insight into 14-3-3-mediated regulation of the GR and highlight both MINK1 and the GR-14-3-3 axis as potential targets for future therapeutic intervention.

Membrane Permeability in Cyclic Peptides is Modulated by Core Conformations.

Cyclic peptides have the potential to bind to challenging targets, which are undruggable with small molecules, but their application is limited by low membrane permeability. Here, using a series of cyclic pentapeptides, we showed that established physicochemical criteria of permeable peptides are heavily violated. We revealed that a dominant core conformation, stabilized by amides' shielding pattern, could guide the design of novel compounds. As a result, counter-intuitive strategies, such as incorporation of polar residues, can be beneficial for permeability. We further find that core globularity is a promising descriptor, which can extend the capability of standard predictive models.

The C-Type Lysozyme from the upper Gastrointestinal Tract of Opisthocomus hoatzin, the Stinkbird.

Muramidases/lysozymes are important bio-molecules, which cleave the glycan backbone in the peptidoglycan polymer found in bacterial cell walls. The glycoside hydrolase (GH) family 22 C-type lysozyme, from the folivorous bird Opisthocomus hoazin (stinkbird), was expressed in Aspergillus oryzae, and a set of variants was produced. All variants were enzymatically active, including those designed to probe key differences between the Hoatzin enzyme and Hen Egg White lysozyme. Four variants showed improved thermostability at pH 4.7, compared to the wild type. The X-ray structure of the enzyme was determined in the apo form and in complex with chitin oligomers. Bioinformatic analysis of avian GH22 amino acid sequences showed that they separate out into three distinct subgroups (chicken-like birds, sea birds and other birds). The Hoatzin is found in the "other birds" group and we propose that this represents a new cluster of avian upper-gut enzymes.

Structure of Aspergillus aculeatus ß-1,4-galactanase in complex with galactobiose.

ß-1,4-Galactanases are glycoside hydrolases that are involved in the degradation of pectin and belong to family 53 in the classification of glycoside hydrolases. Previous studies have elucidated the structures of several fungal and two bacterial galactanases, while biochemical studies have indicated differences in the product profiles of different members of the family. Structural studies of ligand complexes have to date been limited to the bacterial members of the family. Here, the first structure of a fungal galactanase in complex with a disaccharide is presented. Galactobiose binds to subsites -1 and -2, thus improving our understanding of ligand binding to galactanases.

Molecular Modeling Investigation of the Interaction between Humicola insolens Cutinase and SDS Surfactant Suggests a Mechanism for Enzyme Inactivation.

One of the largest commercial applications of enzymes and surfactants is as main components in modern detergents. The high concentration of surfactant compounds usually present in detergents can, however, negatively affect the enzymatic activity. To remedy this drawback, it is of great importance to characterize the interaction between the enzyme and the surfactant molecules at an atomistic resolution. The protein enzyme cutinase from the thermophilic and saprophytic fungus called Humicola insolens (HiC) is a promising candidate for use in detergents thanks to its hydrolase activity targeting mostly biopolyesters (e.g., cutin). HiC is, however, inhibited by low concentrations of sodium dodecyl sulfate (SDS), an ubiquitous surfactant. In this work, we investigate the interaction between HiC and SDS using molecular dynamics simulations. Simulations of HiC dissolved in different aqueous concentrations of SDS show the interaction between HiC and SDS monomers, as well as the formation and dynamics of SDS micelles on the surface of the enzyme. These results suggest a mechanism of cutinase inhibition by SDS, which involves the nucleation of aggregates of SDS molecules on hydrophobic patches on the cutinase surface. Notably, a primary binding site for monomeric SDS is identified near the active site of HiC constituting a possible nucleation point for micelles and leading to the blockage of the entrance to the enzymatic site. Detailed analysis of the simulations allow us to suggest a set of residues from the SDS binding site on HiC to probe as engineered mutations aimed at reducing SDS binding to HiC, thereby decreasing SDS inhibition of HiC.

In vitro and in silico assessment of the developability of a designed monoclonal antibody library.

Despite major advances in antibody discovery technologies, the successful development of monoclonal antibodies (mAbs) into effective therapeutic and diagnostic agents can often be impeded by developability liabilities, such as poor expression, low solubility, high viscosity and aggregation. Therefore, strategies to predict at the early phases of antibody development the risk of late-stage failure of antibody candidates are highly valuable. In this work, we employ the in silico solubility predictor CamSol to design a library of 17 variants of a humanized mAb predicted to span a broad range of solubility values, and we examine their developability potential with a battery of commonly used in vitro and in silico assays. Our results demonstrate the ability of CamSol to rationally enhance mAb developability, and provide a quantitative comparison of in vitro developability measurements with each other and with more resource-intensive solubility measurements, as well as with in silico predictors that offer a potentially faster and cheaper alternative. We observed a strong correlation between predicted and experimentally determined solubility values, as well as with measurements obtained using a panel of in vitro developability assays that probe non-specific interactions. These results indicate that computational methods have the potential to reduce or eliminate the need of carrying out laborious in vitro quality controls for large numbers of lead candidates. Overall, our study provides support to the emerging view that the implementation of in silico tools in antibody discovery campaigns can ensure rapid and early selection of antibodies with optimal developability potential.