Membrane Permeating Macrocycles: Design Guidelines from Machine Learning.

The ability to predict cell-permeable candidate molecules has great potential to assist drug discovery projects. Large molecules that lie beyond the Rule of Five (bRo5) are increasingly important as drug candidates and tool molecules for chemical biology. However, such large molecules usually do not cross cell membranes and cannot access intracellular targets or be developed as orally bioavailable drugs. Here, we describe a random forest (RF) machine learning model for the prediction of passive membrane permeation rates developed using a set of over 1000 bRo5 macrocyclic compounds. The model is based on easily calculated chemical features/descriptors as independent variables. Our random forest (RF) model substantially outperforms a multiple linear regression model based on the same features and achieves better performance metrics than previously reported models using the same underlying data. These features include: (1) polar surface area in water, (2) the octanol-water partitioning coefficient, (3) the number of hydrogen-bond donors, (4) the sum of the topological distances between nitrogen atoms, (5) the sum of the topological distances between nitrogen and oxygen atoms, and (6) the multiple molecular path count of order 2. The last three features represent molecular flexibility, the ability of the molecule to adopt different conformations in the aqueous and membrane interior phases, and the molecular "chameleonicity." Guided by the model, we propose design guidelines for membrane-permeating macrocycles. It is anticipated that this model will be useful in guiding the design of large, bioactive molecules for medicinal chemistry and chemical biology applications.

Atomic Scale Structure of Self-Assembled Lipidated Peptide Nanomaterials.

ß-Peptides have great potential as novel biomaterials and therapeutic agents, due to their unique ability to self-assemble into low dimensional nanostructures, and their resistance to enzymatic degradation in vivo. However, the self-assembly mechanisms of ß-peptides, which possess increased flexibility due to the extra backbone methylene groups present within the constituent ß-amino acids, are not well understood due to inherent difficulties of observing their bottom-up growth pathway experimentally. A computational approach is presented for the bottom-up modelling of the self-assembled lipidated ß3-peptides, from monomers, to oligomers, to supramolecular low-dimensional nanostructures, in all-atom detail. The approach is applied to elucidate the self-assembly mechanisms of recently discovered, distinct structural morphologies of low dimensional nanomaterials, assembled from lipidated ß3-peptide monomers. The resultant structures of the nanobelts and the twisted fibrils are stable throughout subsequent unrestrained all-atom molecular dynamics simulations, and these assemblies display good agreement with the structural features obtained from X-ray fiber diffraction and atomic force microscopy data. This is the first reported, fully-atomistic model of a lipidated ß3-peptide-based nanomaterial, and the computational approach developed here, in combination with experimental fiber diffraction analysis and atomic force microscopy, will be useful in elucidating the atomic scale structure of self-assembled peptide-based and other supramolecular nanomaterials.

An Active Site Inhibitor Induces Conformational Penalties for ACE2 Recognition by the Spike Protein of SARS-CoV-2.

The novel RNA virus, severe acute respiratory syndrome coronavirus II (SARS-CoV-2), is currently the leading cause of mortality in 2020, having led to over 1.6 million deaths and infecting over 75 million people worldwide by December 2020. While vaccination has started and several clinical trials for a number of vaccines are currently underway, there is a pressing need for a cure for those already infected with the virus. Of particular interest in the design of anti-SARS-CoV-2 therapeutics is the human protein angiotensin converting enzyme II (ACE2) to which this virus adheres before entry into the host cell. The SARS-CoV-2 virion binds to cell-surface bound ACE2 via interactions of the spike protein (s-protein) on the viral surface with ACE2. In this paper, we use all-atom molecular dynamics simulations and binding enthalpy calculations to determine the effect that a bound ACE2 active site inhibitor (MLN-4760) would have on the binding affinity of SARS-CoV-2 s-protein with ACE2. Our analysis indicates that the binding enthalpy could be reduced for s-protein adherence to the active site inhibitor-bound ACE2 protein by as much as 1.48-fold as an upper limit. This weakening of binding strength was observed to be due to the destabilization of the interactions between ACE2 residues Glu-35, Glu-37, Tyr-83, Lys-353, and Arg-393 and the SARS-CoV-2 s-protein receptor binding domain (RBD). The conformational changes were shown to lead to weakening of ACE2 interactions with SARS-CoV-2 s-protein, therefore reducing s-protein binding strength. Further, we observed increased conformational lability of the N-terminal helix and a conformational shift of a significant portion of the ACE2 motifs involved in s-protein binding, which may affect the kinetics of the s-protein binding when the small molecule inhibitor is bound to the ACE2 active site. These observations suggest potential new ways for interfering with the SARS-CoV-2 adhesion by modulating ACE2 conformation through distal active site inhibitor binding.

Improving Membrane Permeation in the Beyond Rule-of-Five Space by Using Prodrugs to Mask Hydrogen Bond Donors.

A wide range of drug targets can be effectively modulated by peptides and macrocycles. Unfortunately, the size and polarity of these compounds prevents them from crossing the cell membrane to reach target sites in the cell cytosol. As such, these compounds do not conform to standard measures of drug-likeness and exist in beyond the rule-of-five space. In this work, we investigate whether prodrug moieties that mask hydrogen bond donors can be applied in the beyond rule-of-five domain to improve the permeation of macrocyclic compounds. Using a cyclic peptide model, we show that masking hydrogen bond donors in the natural polar amino acid residues (His, Ser, Gln, Asn, Glu, Asp, Lys, and Arg) imparts membrane permeability to the otherwise impermeable parent molecules, even though the addition of the masking group increases the overall compound molecular weight and the number of hydrogen bond acceptors. We demonstrate this strategy in PAMPA and Caco2 membrane permeability assays and show that masking with groups that reduce the number of hydrogen-bond donors at the cost of additional mass and hydrogen bond acceptors, a donor-acceptor swap, is effective.

Structural and functional characterisation of a novel peptide from the Australian sea anemone Actinia tenebrosa.

Sea anemone venoms have long been recognised as a rich source of peptides with interesting pharmacological and structural properties. Our recent transcriptomic studies of the Australian sea anemone Actinia tenebrosa have identified a novel 13-residue peptide, U-AITx-Ate1. U-AITx-Ate1 contains a single disulfide bridge and bears no significant homology to previously reported amino acid sequences of peptides from sea anemones or other species. We have produced U-AITx-Ate1 using solid-phase peptide synthesis, followed by oxidative folding and purification of the folded peptide using reversed-phase high-performance liquid chromatography. The solution structure of U-AITx-Ate1 was determined based on two-dimensional nuclear magnetic resonance spectroscopic data. Diffusion-ordered NMR spectroscopy revealed that U-AITx-Ate1 was monomeric in solution. Perturbations in the 1D 1H NMR spectrum of U-AITx-Ate1 in the presence of dodecylphosphocholine micelles together with molecular dynamics simulations indicated an interaction of U-AITx-Ate1 with lipid membranes, although no binding was detected to 100% POPC and 80% POPC: 20% POPG lipid nanodiscs by isothermal titration calorimetry. Functional assays were performed to explore the biological activity profile of U-AITx-Ate1. U-AITx-Ate1 showed no activity in voltage-clamp electrophysiology assays and no change in behaviour and mortality rates in crustacea. Moderate cytotoxic activity was observed against two breast cancer cell lines.

Free Energy Methods in Drug Design: Prospects of "Alchemical Perturbation" in Medicinal Chemistry.

Underpinning all drug discovery projects is the interaction between a drug and its target, usually a protein. Thus, improved methods for predicting the magnitude of protein-ligand interactions have the potential to improve the efficiency of drug development. In this review, we describe the principles of free energy methods used for the calculation of protein-ligand binding free energies, the challenges associated with these methods, and recent advances developed to address these difficulties. We then present case studies from 2005 to 2017, each demonstrating that alchemical free energy methods can assist rational drug design projects. We conclude that alchemical methods are becoming a feasible reality in medicinal chemistry research due to improved computational resources and algorithms and that alchemical free energy predictions methods are close to becoming a mainstream tool for medicinal chemists.

Cyclic Hexapeptide Mimics of the LEDGF Integrase Recognition Loop in Complex with HIV-1 Integrase.

The p75 splice variant of lens epithelium-derived growth factor (LEDGF) is a 75 kDa protein, which is recruited by the human immunodeficiency virus (HIV) to tether the pre-integration complex to the host chromatin and promote integration of proviral DNA into the host genome. We designed a series of small cyclic peptides that are structural mimics of the LEDGF binding domain, which interact with integrase as potential binding inhibitors. Herein we present the X-ray crystal structures, NMR studies, SPR analysis, and conformational studies of four cyclic peptides bound to the HIV-1 integrase core domain. Although the X-ray studies show that the peptides closely mimic the LEDGF binding loop, the measured affinities of the peptides are in the low millimolar range. Computational analysis using conformational searching and free energy calculations suggest that the low affinity of the peptides is due to mismatch between the low-energy solution and bound conformations.

A Cyclic Peptide Inhibitor of the iNOS-SPSB Protein-Protein Interaction as a Potential Anti-Infective Agent.

SPRY domain- and SOCS box-containing proteins SPSB1, SPSB2, and SPSB4 interact with inducible nitric oxide synthase (iNOS), causing the iNOS to be polyubiquitinated and targeted for degradation. Inhibition of this interaction increases iNOS levels, and consequently cellular nitric oxide (NO) concentrations, and has been proposed as a potential strategy for killing intracellular pathogens. We previously described two DINNN-containing cyclic peptides (CP1 and CP2) as potent inhibitors of the murine SPSB-iNOS interaction. In this study, we report the crystal structures of human SPSB4 bound to CP1 and CP2 and human SPSB2 bound to CP2. We then used these structures to design a new inhibitor in which an intramolecular hydrogen bond was replaced with a hydrocarbon linkage to form a smaller macrocycle while maintaining the bound geometry of CP2 observed in the crystal structures. This resulting pentapeptide SPSB-iNOS inhibitor (CP3) has a reduced macrocycle ring size, fewer nonbinding residues, and includes additional conformational constraints. CP3 has a greater affinity for SBSB2 ( KD = 7 nM as determined by surface plasmon resonance) and strongly inhibits the SPSB2-iNOS interaction in macrophage cell lysates. We have also determined the crystal structure of CP3 in complex with human SPSB2, which reveals the structural basis for the increased potency of CP3 and validates the original design.

Fragment-Based Screening of a Natural Product Library against 62 Potential Malaria Drug Targets Employing Native Mass Spectrometry.

Natural products are well known for their biological relevance, high degree of three-dimensionality, and access to areas of largely unexplored chemical space. To shape our understanding of the interaction between natural products and protein targets in the postgenomic era, we have used native mass spectrometry to investigate 62 potential protein targets for malaria using a natural-product-based fragment library. We reveal here 96 low-molecular-weight natural products identified as binding partners of 32 of the putative malarial targets. Seventy-nine (79) fragments have direct growth inhibition on Plasmodium falciparum at concentrations that are promising for the development of fragment hits against these protein targets. This adds a fragment library to the published HTS active libraries in the public domain.

A Practical Guide to Molecular Docking and Homology Modelling for Medicinal Chemists.

Elucidating details of the relationship between molecular structure and a particular biological end point is essential for successful, rational drug discovery. Molecular docking is a widely accepted tool for lead identification however, navigating the intricacies of the software can be daunting. Our objective was therefore to provide a step-by-step guide for those interested in incorporating contemporary basic molecular docking and homology modelling into their design strategy. Three molecular docking programs, AutoDock4, SwissDock and Surflex-Dock, were compared in the context of a case study where a set of steroidal and non-steroidal ligands were docked into the human androgen receptor (hAR) using both rigid and flexible target atoms. Metrics for comparison included how well each program predicted the X-ray structure orientation via root mean square deviation (rmsd), predicting known actives via ligand ranking and comparison to biological data where available. Benchmarking metrics were discussed in terms of identifying accurate and reliable results. For cases where no three dimensional structure exists, we provided a practical example for creating a homology model using Swiss-Model. Results showed an rmsd between X-ray ligands from wild-type and mutant receptors and docked poses were 4.15Å and 0.83Å (SwissDock), 2.69Å and 8.80Å (AutoDock4) and 0.39Å and 0.71Å (Surflex-Dock) respectively. Surflex-Dock performed consistently well in pose prediction (less than 2Å) while Auto- Dock4 predicted known active non-steroidal antiandrogens most accurately. Introducing flexibility into target atoms produced the largest degree of change in ligand ranking in Surflex-Dock. We produced a viable homology model of the P2X1 purireceptor for subsequent docking analysis.