Bioactive Conformation Research Articles

Optimization of the Central α-Amino Acid in Cystobactamids to the Broad-Spectrum, Resistance-Breaking Antibiotic CN-CC-861.

Cystobactamids have a unique oligoarylamide structure and exhibit broad-spectrum activity against Gram-negative and Gram-positive bacteria. In this study, the central α-amino acid of the cystobactamid scaffold was modified to address the relevance of stereochemistry, hydrogen bonding and polarity by 33 derivatives. As demonstrated by three matched molecular pairs, l-amino acids were preferred over d-amino acids. A rigidification to a six-membered system stabilized the bioactive conformation for the on-target Escherichia coli gyrase, but did not improve antimicrobial activity. Compound CN-CC-861, carrying a propargyl side chain, had more than 16-fold lower minimal inhibitory concentration (MIC) values against Enterococcus faecalis, Staphylococci and Acinetobacter strains, compared to known analogues. Moreover, CN-CC-861 retained activity against multidrug-resistant enterococci, displayed strong bactericidal activity, moderate-low frequencies of resistance and in vivo efficacy in a neutropenic thigh infection model with E. coli. Overall, the findings will guide the design of new promising structures with higher activities and broader spectrum.

DNA-templated spatially controlled proteolysis targeting chimeras for CyclinD1-CDK4/6 complex protein degradation.

Constraining proximity-based drugs, such as proteolysis-targeting chimeras (PROTACs), into its bioactive conformation can significantly impact their selectivity and potency. However, traditional methods for achieving this often involve complex and time-consuming synthetic procedures. Here, we introduced an alternative approach by demonstrating DNA-templated spatially controlled PROTACs (DTACs), which leverage the programmability of nucleic-acid based self-assembly for efficient synthesis, providing precise control over inhibitors' spacing and orientation. The resulting constructs revealed distance- and orientation-dependent selectivity and degradation potency for the CyclinD1-CDK4/6 protein complex in cancer cells. Notably, an optimal construct DTAC-V1 demonstrated the unprecedented synchronous degradation of entire CyclinD1-CDK4/6 complex. This resulted in the effective cell cycle arrest in G1 phase, and further therapeutic studies showed its potent anti-tumor effects compared to inhibitors alone. These findings present a novel framework for PROTACs design, offering critical insights that may inform the development of other proximity-induced therapeutic modalities.

The Missing Link(er): A Roadmap to Macrocyclization in Drug Discovery.

Macrocycles are one of nature's preferred choices to generate large but cell-permeable bioactive molecules. Macrocyclization is increasingly prominent in medicinal chemistry beyond natural products, especially for difficult-to-drug targets. However, strategies to best exploit the potential of macrocycles are only beginning to emerge. Here we survey drug discovery campaigns from the past decade that cumulated in advanced macrocyclic drug-like compounds or drug candidates. Most macrocycles were conceived by ring closing based on U- or C-shaped bioactive conformations observed in co-crystal structures. We focus on the key step from linear precursors to the first macrocycle and the follow-up optimization of the resulting macrocyclic scaffold. Conformational control recurrently emerged as a key factor for macrocycle properties and linkers as an opportunity for optimization. With increasingly challenging drug targets, we expect these trends to become more prominent and relevant.

A Multi-Angle Approach to Predict Peptide-GPCR Complexes: The N/OFQ-NOP System as a Successful AlphaFold Application Case Study.

With nearly 700 structures solved and a growing number of customized structure prediction algorithms being developed at a fast pace, G protein-coupled receptors (GPCRs) are an optimal test case for validating new approaches for the prediction of receptor active state and ligand bioactive conformation complexes. In this study, we leveraged the availability of hundreds of peptide GPCRs in the active state and both classical homology and artificial intelligence (AI) based protein modeling combined with docking and AI-based peptide structure prediction approaches to predict the nociceptin/orphanin FQ-NOP receptor active state complex (N/OFQ-NOPa). The In Silico generated hypotheses were validated via the design, synthesis, and pharmacological characterization of novel linear N/OFQ(1-13)-NH2 analogues, leading to the discovery of a novel antagonist (3B; pKB = 6.63) bearing a single ring-constrained residue in place of the Gly2-Gly3 motif of the N/OFQ message sequence (FGGF). While the experimental validation was ongoing, the availability of the Cryo-EM structure of the predicted complex enabled us to unambiguously validate the generated hypotheses. To the best of our knowledge, this is the first example of a peptide-GPCR complex predicted with atomistic accuracy (full complex Cα RMSD < 1.0 Å) and of the N/OFQ message moiety being successfully modified with a rigid scaffold.

Uncovering a Latent Bioactive Interleukin-6 Glycoform.

A bioinspired semisynthesis of human-interleukin-6 bearing N-glycan at Asn143 (143glycosyl-IL-6) was performed by intentional glycosylation effects and protein folding chemistry for regioselective peptide-backbone activation. 143Glycosyl-IL-6 is a genetically coded cytokine, but isolation was difficult owing to a tiny amount. IL6-polypeptide (1-141-position) with an intentionally inserted cysteine at 142-position was expressed in E. coli. The expressed polypeptide was treated with a chemical folding process to make a specific helices bundle conformation through native two-disulfide bonds (43-49 and 72-82). Utilizing the successfully formed free-142-cysteine, sequential conversions using cyanation of 142-cysteine, hydrazinolysis, and thioesterification created a long polypeptide (1-141)-thioester. However, the resultant polypeptide-thioester caused considerable aggregation owing to a highly hydrophobic peptide sequence. After the reduction of two-disulfide bonds of polypeptide (1-141)-thioester, an unprecedented hydrophilic N-glycan tag was inserted at the resultant cysteine thiols. The N-glycan tags greatly stabilized polypeptide-thioester. The subsequent native chemical ligation and desulfurization successfully gave a whole 143glycosyl-IL-6 polypeptide (183-amino acids). Removal of four N-glycan tags and immediate one-pot in vitro folding protocol efficiently produced the folded 143glycosyl-IL-6. The folded 143glycosyl-IL-6 exhibited potent cell proliferation activity. The combined studies with molecular dynamics simulation, semisynthesis, and bioassays predict the bioactive conformation of latent 143glycosyl-IL-6.

Uncovering a Latent Bioactive Interleukin‐6 Glycoform

A bioinspired semisynthesis of human‐interleukin‐6 bearing N‐glycan at Asn143 (143glycosyl‐IL‐6) was performed by intentional glycosylation effects and protein folding chemistry for regioselective peptide‐backbone activation. 143Glycosyl‐IL‐6 is a genetically coded cytokine, but isolation was difficult owing to a tiny amount. IL6‐polypeptide (1‐141‐position) with an intentionally inserted cysteine at 142‐position was expressed in E. coli. The expressed polypeptide was treated with a chemical folding process to make a specific helices bundle conformation through native two‐disulfide bonds (43–49 and 72–82). Utilizing the successfully formed free‐142‐cysteine, sequential conversions using cyanation of 142‐cysteine, hydrazinolysis, and thioesterification created a long polypeptide (1–141)‐thioester. However, the resultant polypeptide‐thioester caused considerable aggregation owing to a highly hydrophobic peptide sequence. After the reduction of two‐disulfide bonds of polypeptide (1–141)‐thioester, an unprecedented hydrophilic N‐glycan tag was inserted at the resultant cysteine thiols. The N‐glycan tags greatly stabilized polypeptide‐thioester. The subsequent native chemical ligation and desulfurization successfully gave a whole 143glycosyl‐IL‐6 polypeptide (183‐amino acids). Removal of four N‐glycan tags and immediate one‐pot in vitro folding protocol efficiently produced the folded 143glycosyl‐IL‐6. The folded 143glycosyl‐IL‐6 exhibited potent cell proliferation activity. The combined studies with molecular dynamics simulation, semisynthesis, and bioassays predict the bioactive conformation of latent 143glycosyl‐IL‐6.

Strengthening an Intramolecular Non-Classical Hydrogen Bond to Get in Shape for Binding.

In this research article, we report on the strengthening of a non-classical hydrogen bond (C-H⋅⋅⋅O) by introducing electron withdrawing groups at the carbon atom. The approach is demonstrated on the example of derivatives of the physiological E-selectin ligand sialyl Lewisx (1, sLex). Its affinity is mainly due to a beneficial entropy term, which is predominantly caused by the pre-organization of sLex in its binding conformation. We have shown, that among the elements responsible for the pre-organization, the stabilization by a non-classical hydrogen bond between the H-C5 of l-fucose and the ring oxygen O5 of the neighboring d-galactose moiety is essential and yields 7.4 kJ mol-1. This effect could be further strengthened by replacing l-fucose by 6,6,6-trifluoro-l-fucose leading to an improved non-classical H-bond of 14.9 kJ mol-1, i.e., an improved pre-organization in the bioactive conformation. For a series of glycomimetics of sLex (1), this outcome could be confirmed by high field NMR-shifts of the H-C5Fuc, by X-ray diffraction analysis of glycomimetics co-crystallized with E-selectin as well as by isothermal titration calorimetry. Furthermore, the electron-withdrawing character of the CF3-group beneficially influences the pharmacokinetic properties of sLex mimetics. Thus, acid-stability, a prerequisite for gastrointestinal stability, could be substantially improved.

Alternative synthetic route for the pharmacophore of anticancer agent: Triazolopyridazine derivative

ATAD2 has received attention as one of the potential oncogene with tumor-promoting aspects in many malignancies. ATAD2 is a highly conserved bromodomain family protein that exerts its biological functions by mainly AAA ATPase and bromodomain. Several small molecule inhibitors have been described in the literature. AZ13824374 (1) recently reported by Holt and co-workers showed promising in vitro (bio-chemical, cellular) and antiproliferative activity in range of breast cancer models. In this work, we described scalable synthetic route for triazolopyridazine derivative (2), a key intermediate of AZ13824374 (1) without using CO in the process. Triazolopyridazine helps to attain the bioactive conformation for AZ13824374 (1) through its crucial interaction with Tyr 1021 of ATAD2. Additionally, triazolopyridazine is extensively used as an intermediate for anticancer agents. This encouraged us to develop cost-effective and scalable process for it.

Gonococcal Mimitope Vaccine Candidate Forms a Beta-Hairpin Turn and Binds Hydrophobically to a Therapeutic Monoclonal Antibody.

The spread of multidrug-resistant strains of Neisseria gonorrhoeae, the etiologic agent of gonorrhea, represents a global health emergency. Therefore, the development of a safe and effective vaccine against gonorrhea is urgently needed. In previous studies, murine monoclonal antibody (mAb) 2C7 was raised against gonococcal lipooligosaccharide (LOS). mAb 2C7 elicits complement-dependent bactericidal activity against gonococci, and its glycan epitope is expressed by almost every clinical isolate. Furthermore, we identified a peptide, cyclic peptide 2 (CP2) that mimicked the 2C7 LOS epitope, elicited bactericidal antibodies in mice, and actively protected in a mouse vaginal colonization model. In this study, we performed structural analyses of mAb 2C7 and its complex with the CP2 peptide by X-ray crystallography, NMR spectroscopy, and molecular dynamics (MD) simulations. The crystal structure of Fab 2C7 bound to CP2 showed that the peptide adopted a beta-hairpin conformation and bound the Fab primarily through hydrophobic interactions. We employed NMR spectroscopy and MD simulations to map the 2C7 epitope and identify the bioactive conformation of CP2. We also used small-angle X-ray scattering (SAXS) and native mass spectrometry to obtain further information about the shape and assembly state of the complex. Collectively, our new structural information suggests strategies for humanizing mAb 2C7 as a therapeutic against gonococcal infection and for optimizing peptide CP2 as a vaccine antigen.

Discovery of Clinical Candidate PF-06648671: A Potent γ-Secretase Modulator for the Treatment of Alzheimer's Disease.

Herein, we describe the design and synthesis of γ-secretase modulator (GSM) clinical candidate PF-06648671 (22) for the treatment of Alzheimer's disease. A key component of the design involved a 2,5-cis-tetrahydrofuran (THF) linker to impart conformational rigidity and lock the compound into a putative bioactive conformation. This effort was guided using a pharmacophore model since crystallographic information was not available for the membrane-bound γ-secretase protein complex at the time of this work. PF-06648671 achieved excellent alignment of whole cell in vitro potency (Aβ42 IC50 = 9.8 nM) and absorption, distribution, metabolism, and excretion (ADME) parameters. This resulted in favorable in vivo pharmacokinetic (PK) profile in preclinical species, and PF-06648671 achieved a human PK profile suitable for once-a-day dosing. Furthermore, PF-06648671 was found to have favorable brain availability in rodent, which translated into excellent central exposure in human and robust reduction of amyloid β (Aβ) 42 in cerebrospinal fluid (CSF).

Macrocyclic Azapeptide Nitriles: Structure-Based Discovery of Potent SARS-CoV-2 Main Protease Inhibitors as Antiviral Drugs.

Given the crucial role of the main protease (Mpro) in the replication cycle of SARS-CoV-2, this viral cysteine protease constitutes a high-profile drug target. We investigated peptidomimetic azapeptide nitriles as auspicious, irreversibly acting inhibitors of Mpro. Our systematic approach combined an Mpro active-site scanning by combinatorially assembled azanitriles with structure-based design. Encouraged by the bioactive conformation of open-chain inhibitors, we conceptualized the novel chemotype of macrocyclic azanitriles whose binding mode was elucidated by cocrystallization. This strategy provided a favorable entropic contribution to target binding and resulted in the development of the extraordinarily potent Mpro inhibitor 84 with an IC50 value of 3.23 nM and a second-order rate constant of inactivation, kinac/Ki, of 448,000 M-1s-1. The open-chain Mpro inhibitor 58, along with the macrocyclic compounds 83 and 84, a broad-spectrum anticoronaviral agent, demonstrated the highest antiviral activity with EC50 values in the single-digit micromolar range. Our findings are expected to promote the future development of peptidomimetic Mpro inhibitors as anti-SARS-CoV-2 agents.

Structural Features Influencing the Bioactive Conformation of Angiotensin II and Angiotensin A: Relationship between Receptor Desensitization, Addiction, and the Blood-Brain Barrier.

The N-terminal portion of the octapeptide angiotensin II (DRVYIHPF; AngII), a vasopressor peptide that favorably binds to, and activates, AngII type 1 receptor (AT1R), has an important role in maintaining bioactive conformation. It involves all three charged groups, namely (i) the N-terminal amino group cation, (ii) the Asp sidechain anion and (iii) the Arg guanidino cation. Neutralization of any one of these three charged groups results in a substantial reduction (<5%) in bioactivity, implicating a specialized function for this cluster. In contrast, angiotensin A (ARVYIHPF; AngA) has reduced bioactivity at AT1R; however, replacement of Asp in AngII with sarcosine (N-methyl-glycine) not only restores bioactivity but increases the activity of agonist, antagonist, and inverse agonist analogues. A bend produced at the N-terminus by the introduction of the secondary amino acid sarcosine is thought to realign the functional groups that chaperone the C-terminal portion of AngII, allowing transfer of the negative charge originating at the C-terminus to be transferred to the Tyr hydroxyl-forming tyrosinate anion, which is required to activate the receptor and desensitizes the receptor (tachyphylaxis). Peptide (sarilesin) and nonpeptide (sartans) moieties, which are long-acting inverse agonists, appear to desensitize the receptor by a mechanism analogous to tachyphylaxis. Sartans/bisartans were found to bind to alpha adrenergic receptors resulting in structure-dependent desensitization or resensitization. These considerations have provided information on the mechanisms of receptor desensitization/tolerance and insights into possible avenues for treating addiction. In this regard sartans, which appear to cross the blood-brain barrier more readily than bisartans, are the preferred drug candidates.

From Protein Structures to Functional Biomimetics

AbstractThe development of complex molecular scaffolds with defined folding properties represents a central challenge in chemical research. Proteins are natural scaffolds defined by a hierarchy of structural complexity and have evolved to manifest unique functional characteristics; for example, molecular recognition capabilities that facilitate the binding of target molecules with high affinity and selectivity. Utilizing these features, proteins have been used as a starting point for the design of synthetic foldamers and enhanced biocatalysts, as well as bioactive reagents in drug discovery. In this account, we describe the strategies used in our group to stabilize protein folds, ranging from the constraint of bioactive peptide conformations to chemical protein engineering. We discuss the evolution of peptides into peptidomimetics to inhibit protein–protein and protein–nucleic acid interactions, and the selective chemical modification of proteins to enhance their properties for biotechnological applications. The reported peptide- and proteomimetic structures cover a broad range of molecular sizes and they highlight the importance of structure stabilization for the design of functional biomimetics.1 Introduction2 Constraining the Conformation of Peptides3 Peptide-Based Covalent Protein Modifiers4 Chemical Protein Engineering5 Conclusions

Syntheses and Structure-Activity Relationship Studies of Antitumor Bicyclic Hexapeptide RA-VII Analogues

A series of antitumor bicyclic hexapeptide RA-VII analogues modified at residue 2, 3, or 6 were prepared by the chemical transformation of the hydroxy, methoxy, or carboxy groups or the aromatic rings of natural peptides RA-II, III, V, VII, and X. Analogues with modified side chains or peptide backbones, which cannot be prepared by the chemical transformation of their natural peptides, and newly isolated peptides from Rubia cordifolia roots were synthesized by using protected cycloisodityrosines prepared by the degradation of bis(thioamide) obtained from RA-VII or the diphenyl ether formation of boronodipeptide under the modified Chan-Lam coupling reaction conditions. Studies of the conformational features of the analogues and the newly isolated peptides and their relationships with cytotoxic activities against the HCT-116, HL-60, KATO-III, KB, L1210, MCF-7, and P-388 cell lines revealed the following: the methoxy group at residue 3 is essential for the potent cytotoxic activity; the methyl group at Ala-2 and Ala-4 but not at D-Ala-1 is required to establish the bioactive conformation; the N-methyl group at Tyr-5 is necessary for the peptides to adopt the active conformation preferentially; and the orientation of Tyr-5 and/or Tyr-6 phenyl rings has a significant effect on the cytotoxic activity.

Chemical Transformation of B- to A-type Proanthocyanidins and 3D Structural Implications.

In nature, proanthocyanidins (PACs) with A-type linkages are relatively rare, likely due to biosynthetic constraints in the formation of additional ether bonds to be introduced into the more common B-type precursors. However, A-type linkages confer greater structural rigidity on PACs than do B-type linkages. Prior investigations into the structure-activity relationships (SAR) describing how plant-derived PACs with B- and complex AB-type linkages affect their capacity for dentin biomodification indicate that a higher ratio of double linkages leads to a greater interaction with dentin type I collagen. Thus, A-type PACs emerge as particularly intriguing candidates for interventional functional biomaterials. This study employed a free-radical-mediated oxidation using DPPH to transform trimeric and tetrameric B-type PACs, 2 and 4, respectively, into their exclusively A-type linked analogues, 3 and 5, respectively. The structures and absolute configurations of the semisynthetic products, including the new all-A-type tetramer 5, were determined by comprehensive spectroscopic analysis. Additionally, molecular modeling investigated the conformational characteristics of all trimers and tetramers, 1-5. Our findings suggest that the specific interflavan linkages significantly impact the flexibility and low-energy conformations of the connected monomeric units, which conversely can affect the bioactive conformations relevant for dentin biomodification.

DMXP: A De Novo Small-Molecule 3D Structure Predictor with Graph Attention Networks.

Generating the three-dimensional (3D) structure of small molecules is crucial in both structure- and ligand-based drug design. Structure-based drug design needs bioactive conformations of compounds for lead identification and optimization. Ligand-based drug design techniques, such as 3D shape similarity search, 3D pharmacophore model, 3D-QSAR, etc., all require high-quality small-molecule ligand conformations to obtain reliable results. Although predicting a small molecular bioactive conformer requires information from the receptor, a crystal structure of the molecule is a proper approximation to its bioactive conformer in a specific receptor because the binding pose of a small molecule in its receptor's binding pockets should be energetically close to the crystal structures. This study presents a de novo small molecular structure predictor (dMXP) with graph attention networks based on crystal data derived from the Cambridge Structural Database (CSD) combined with molecular electrostatic information calculated by density-functional theory (DFT). Two featuring strategies (topological and atomic partial change features) were employed to explore the relation between these features and the 3D crystal structure of a small molecule. These features were then assembled to construct the holistic 3D crystal structure of a molecule. Molecular graphs were encoded using a graph attention mechanism to deal with the issues of the inconsistencies of local substructures contributing to the entire molecular structure. The root-mean-square deviation (RMSDs) of approximately 80% dMXP predicted structures and the native binding poses within receptors are less than 2.0 Å.

Integration of the Butina algorithm and ensemble learning strategies for the advancement of a pharmacophore ligand-based model: an in silico investigation of apelin agonists.

Introduction: 3D pharmacophore models describe the ligand's chemical interactions in their bioactive conformation. They offer a simple but sophisticated approach to decipher the chemically encoded ligand information, making them a valuable tool in drug design. Methods: Our research summarized the key studies for applying 3D pharmacophore models in virtual screening for 6,944 compounds of APJ receptor agonists. Recent advances in clustering algorithms and ensemble methods have enabled classical pharmacophore modeling to evolve into more flexible and knowledge-driven techniques. Butina clustering categorizes molecules based on their structural similarity (indicated by the Tanimoto coefficient) to create a structurally diverse training dataset. The learning method combines various individual pharmacophore models into a set of pharmacophore models for pharmacophore space optimization in virtual screening. Results: This approach was evaluated on Apelin datasets and afforded good screening performance, as proven by Receiver Operating Characteristic (AUC score of 0.994 ± 0.007), enrichment factor of (EF1% of 50.07 ± 0.211), Güner-Henry score of 0.956 ± 0.015, and F-measure of 0.911 ± 0.031. Discussion: Although one of the high-scoring models achieved statistically superior results in each dataset (AUC of 0.82; an EF1% of 19.466; GH of 0.131 and F1-score of 0.071), the ensemble learning method including voting and stacking method balanced the shortcomings of each model and passed with close performance measures.

Consensus Pharmacophore Strategy For Identifying Novel SARS-Cov-2 Mpro Inhibitors from Large Chemical Libraries.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) main Protease (Mpro) is an enzyme that cleaves viral polyproteins translated from the viral genome and is critical for viral replication. Mpro is a target for anti-SARS-CoV-2 drug development, and multiple Mpro crystals complexed with competitive inhibitors have been reported. In this study, we aimed to develop an Mpro consensus pharmacophore as a tool to expand the search for inhibitors. We generated a consensus model by aligning and summarizing pharmacophoric points from 152 bioactive conformers of SARS-CoV-2 Mpro inhibitors. Validation against a library of conformers from a subset of ligands showed that our model retrieved poses that reproduced the crystal-binding mode in 77% of the cases. Using models derived from a consensus pharmacophore, we screened >340 million compounds. Pharmacophore-matching and chemoinformatics analyses identified new potential Mpro inhibitors. The candidate compounds were chemically dissimilar to the reference set, and among them, demonstrating the relevance of our model. We evaluated the effect of 16 candidates on Mpro enzymatic activity finding that seven have inhibitory activity. Three compounds (1, 4, and 5) had IC50 values in the midmicromolar range. The Mpro consensus pharmacophore reported herein can be used to identify compounds with improved activity and novel chemical scaffolds against Mpro. The method developed for its generation is provided as an open-access code (https://github.com/AngelRuizMoreno/ConcensusPharmacophore) and can be applied to other pharmacological targets.

Binding Free Energy Calculation Based on the Fragment Molecular Orbital Method and Its Application in Designing Novel SHP-2 Allosteric Inhibitors.

The accurate prediction of binding free energy is a major challenge in structure-based drug design. Quantum mechanics (QM)-based approaches show promising potential in predicting ligand-protein binding affinity by accurately describing the behavior and structure of electrons. However, traditional QM calculations face computational limitations, hindering their practical application in drug design. Nevertheless, the fragment molecular orbital (FMO) method has gained widespread application in drug design due to its ability to reduce computational costs and achieve efficient ab initio QM calculations. Although the FMO method has demonstrated its reliability in calculating the gas phase potential energy, the binding of proteins and ligands also involves other contributing energy terms, such as solvent effects, the 'deformation energy' of a ligand's bioactive conformations, and entropy. Particularly in cases involving ionized fragments, the calculation of solvation free energy becomes particularly crucial. We conducted an evaluation of some previously reported implicit solvent methods on the same data set to assess their potential for improving the performance of the FMO method. Herein, we develop a new QM-based binding free energy calculation method called FMOScore, which enhances the performance of the FMO method. The FMOScore method incorporates linear fitting of various terms, including gas-phase potential energy, deformation energy, and solvation free energy. Compared to other widely used traditional prediction methods such as FEP+, MM/PBSA, MM/GBSA, and Autodock vina, FMOScore showed good performance in prediction accuracies. By constructing a retrospective case study, it was observed that incorporating calculations for solvation free energy and deformation energy can further enhance the precision of FMO predictions for binding affinity. Furthermore, using FMOScore-guided lead optimization against Src homology-2-containing protein tyrosine phosphatase 2 (SHP-2), we discovered a novel and potent allosteric SHP-2 inhibitor (compound 8).

Applying atomistic neural networks to bias conformer ensembles towards bioactive-like conformations

Identifying bioactive conformations of small molecules is an essential process for virtual screening applications relying on three-dimensional structure such as molecular docking. For most small molecules, conformer generators retrieve at least one bioactive-like conformation, with an atomic root-mean-square deviation (ARMSD) lower than 1 Å, among the set of low-energy conformers generated. However, there is currently no general method to prioritise these likely target-bound conformations in the ensemble. In this work, we trained atomistic neural networks (AtNNs) on 3D information of generated conformers of a curated subset of PDBbind ligands to predict the ARMSD to their closest bioactive conformation, and evaluated the early enrichment of bioactive-like conformations when ranking conformers by AtNN prediction. AtNN ranking was compared with bioactivity-unaware baselines such as ascending Sage force field energy ranking, and a slower bioactivity-based baseline ranking by ascending Torsion Fingerprint Deviation to the Maximum Common Substructure to the most similar molecule in the training set (TFD2SimRefMCS). On test sets from random ligand splits of PDBbind, ranking conformers using ComENet, the AtNN encoding the most 3D information, leads to early enrichment of bioactive-like conformations with a median BEDROC of 0.29 ± 0.02, outperforming the best bioactivity-unaware Sage energy ranking baseline (median BEDROC of 0.18 ± 0.02), and performing on a par with the bioactivity-based TFD2SimRefMCS baseline (median BEDROC of 0.31 ± 0.02). The improved performance of the AtNN and TFD2SimRefMCS baseline is mostly observed on test set ligands that bind proteins similar to proteins observed in the training set. On a more challenging subset of flexible molecules, the bioactivity-unaware baselines showed median BEDROCs up to 0.02, while AtNNs and TFD2SimRefMCS showed median BEDROCs between 0.09 and 0.13. When performing rigid ligand re-docking of PDBbind ligands with GOLD using the 1% top-ranked conformers, ComENet ranked conformers showed a higher successful docking rate than bioactivity-unaware baselines, with a rate of 0.48 ± 0.02 compared to CSD probability baseline with a rate of 0.39 ± 0.02. Similarly, on a pharmacophore searching experiment, selecting the 20% top-ranked conformers ranked by ComENet showed higher hit rate compared to baselines. Hence, the approach presented here uses AtNNs successfully to focus conformer ensembles towards bioactive-like conformations, representing an opportunity to reduce computational expense in virtual screening applications on known targets that require input conformations.