Shape-Based Virtual Screening of a Billion-Compound Library Identifies Mycobacterial Lipoamide Dehydrogenase Inhibitors

Lpd (lipoamide dehydrogenase) in Mycobacterium tuberculosis (Mtb) is required for virulence and is a genetically validated tuberculosis (TB) target. Numerous screens have been performed over the last decade, yet only two inhibitor series have been identified. Recent advances in large-scale virtual screening methods combined with make-on-demand compound libraries have shown the potential for finding novel hits. In this study, the Enamine REAL library consisting of ∼1.12 billion compounds was efficiently screened using the GPU Shape screen method against Mtb Lpd to find additional chemical matter that would expand on the known sulfonamide inhibitor series. We identified six new inhibitors with IC50 in the range of 5-100 μM. While these compounds remained chemically close to the already known sulfonamide series inhibitors, some diversity was found in the cores of the hits. The two most potent hits were further validated by one-step potency optimization to submicromolar levels. The co-crystal structure of optimized analogue TDI-13537 provided new insights into the potency determinants of the series.

Enhancing Hit Discovery in Virtual Screening through Absolute Protein-Ligand Binding Free-Energy Calculations

In the hit identification stage of drug discovery, a diverse chemical space needs to be explored to identify initial hits. Contrary to empirical scoring functions, absolute protein-ligand binding free-energy perturbation (ABFEP) provides a theoretically more rigorous and accurate description of protein-ligand binding thermodynamics and could, in principle, greatly improve the hit rates in virtual screening. In this work, we describe an implementation of an accurate and reliable ABFEP method in FEP+. We validated the ABFEP method on eight congeneric compound series binding to eight protein receptors including both neutral and charged ligands. For ligands with net charges, the alchemical ion approach is adopted to avoid artifacts in electrostatic potential energy calculations. The calculated binding free energies correlate with experimental results with a weighted average of R² = 0.55 for the entire dataset. We also observe an overall root-mean-square error (RMSE) of 1.1 kcal/mol after shifting the zero-point of the simulation data to match the average experimental values. Through ABFEP calculations using apo versus holo protein structures, we demonstrated that the protein conformational and protonation state changes between the apo and holo proteins are the main physical factors contributing to the protein reorganization free energy manifested by the overestimation of raw ABFEP calculated binding free energies using the holo structures of the proteins. Furthermore, we performed ABFEP calculations in three virtual screening applications for hit enrichment. ABFEP greatly improves the hit rates as compared to docking scores or other methods like metadynamics. The good performance of ABFEP in rank ordering compounds demonstrated in this work confirms it as a useful tool to improve the hit rates in virtual screening, thus facilitating hit discovery.

Induced-Fit Docking Enables Accurate Free Energy Perturbation Calculations in Homology Models

Homology models have been used for virtual screening and to understand the binding mode of a known active compound; however, rarely have the models been shown to be of sufficient accuracy, comparable to crystal structures, to support free-energy perturbation (FEP) calculations. We demonstrate here that the use of an advanced induced-fit docking methodology reliably enables predictive FEP calculations on congeneric series across homology models ≥30% sequence identity. Furthermore, we show that retrospective FEP calculations on a congeneric series of drug-like ligands are sufficient to discriminate between predicted binding modes. Results are presented for a total of 29 homology models for 14 protein targets, showing FEP results comparable to those obtained using experimentally determined crystal structures for 86% of homology models with template structure sequence identities ranging from 30 to 50%. Implications for the use and validation of homology models in drug discovery projects are discussed.

Discovery of a dual Ras and ARF6 inhibitor from a GPCR endocytosis screen

Internalization and intracellular trafficking of G protein-coupled receptors (GPCRs) play pivotal roles in cell responsiveness. Dysregulation in receptor trafficking can lead to aberrant signaling and cell behavior. Here, using an endosomal BRET-based assay in a high-throughput screen with the prototypical GPCR angiotensin II type 1 receptor (AT1R), we sought to identify receptor trafficking inhibitors from a library of ~115,000 small molecules. We identified a novel dual Ras and ARF6 inhibitor, which we named Rasarfin, that blocks agonist-mediated internalization of AT1R and other GPCRs. Rasarfin also potently inhibits agonist-induced ERK1/2 signaling by GPCRs, and MAPK and Akt signaling by EGFR, as well as prevents cancer cell proliferation. In silico modeling and in vitro studies reveal a unique binding modality of Rasarfin within the SOS-binding domain of Ras. Our findings unveil a class of dual small G protein inhibitors for receptor trafficking and signaling, useful for the inhibition of oncogenic cellular responses.

While Ras is a promising target for cancer therapy, development of inhibitors targeting Ras signaling has proven challenging. Here, the authors report the discovery of Rasarfin, a small molecule from a phenotypic screen on G protein-coupled receptor (GPCR) endocytosis that acts as a dual Ras and ARF6 inhibitor.

Beta-arrestins operate an on/off control switch for focal adhesion kinase activity

Focal adhesion kinase (FAK) regulates key biological processes downstream of G protein-coupled receptors (GPCRs) in normal and cancer cells, but the modes of kinase activation by these receptors remain unclear. We report that after GPCR stimulation, FAK activation is controlled by a sequence of events depending on the scaffolding proteins β-arrestins and G proteins. Depletion of β-arrestins results in a marked increase in FAK autophosphorylation and focal adhesion number. We demonstrate that β-arrestins interact directly with FAK and inhibit its autophosphorylation in resting cells. Both FAK-β-arrestin interaction and FAK inhibition require the FERM domain of FAK. Following the stimulation of the angiotensin receptor AT1AR and subsequent translocation of the FAK-β-arrestin complex to the plasma membrane, β-arrestin interaction with the adaptor AP-2 releases inactive FAK from the inhibitory complex, allowing its activation by receptor-stimulated G proteins and activation of downstream FAK effectors. Release and activation of FAK in response to angiotensin are prevented by an AP-2-binding deficient β-arrestin and by a specific inhibitor of β-arrestin/AP-2 interaction; this inhibitor also prevents FAK activation in response to vasopressin. This previously unrecognized mechanism of FAK regulation involving a dual role of β-arrestins, which inhibit FAK in resting cells while driving its activation at the plasma membrane by GPCR-stimulated G proteins, opens new potential therapeutic perspectives in cancers with up-regulated FAK.

Identification of Allosteric Inhibitors against Active Caspase-6

Caspase-6 is a cysteine protease that plays essential roles in programmed cell death, axonal degeneration, and development. The excess neuronal activity of Caspase-6 is associated with Alzheimer disease neuropathology and age-dependent cognitive impairment. Caspase-6 inhibition is a promising strategy to stop early stage neurodegenerative events, yet finding potent and selective Caspase-6 inhibitors has been a challenging task due to the overlapping structural and functional similarities between caspase family members. Here, we investigated how four rare non-synonymous missense single-nucleotide polymorphisms (SNPs), resulting in amino acid substitutions outside human Caspase-6 active site, affect enzyme structure and catalytic efficiency. Three investigated SNPs were found to align with a putative allosteric pocket with low sequence conservation among human caspases. Virtual screening of 57,700 compounds against the putative Caspase-6 allosteric pocket, followed by in vitro testing of the best virtual hits in recombinant human Caspase-6 activity assays identified novel allosteric Caspase-6 inhibitors with IC50 and Ki values ranging from ~2 to 13 µM. This report may pave the way towards the development and optimisation of novel small molecule allosteric Caspase-6 inhibitors and illustrates that functional characterisation of rare natural variants holds promise for the identification of allosteric sites on other therapeutic targets in drug discovery.

Emerging Paradigm of Intracellular Targeting of G Protein-Coupled Receptors

G protein-coupled receptors (GPCRs) recognize a diverse array of extracellular stimuli, and they mediate a broad repertoire of signaling events involved in human physiology. Although the major effort on targeting GPCRs has typically been focused on their extracellular surface, a series of recent developments now unfold the possibility of targeting them from the intracellular side as well. Allosteric modulators binding to the cytoplasmic surface of GPCRs have now been described, and their structural mechanisms are elucidated by high-resolution crystal structures. Furthermore, pepducins, aptamers, and intrabodies targeting the intracellular face of GPCRs have also been successfully utilized to modulate receptor signaling. Moreover, small molecule compounds, aptamers, and synthetic intrabodies targeting β-arrestins have also been discovered to modulate GPCR endocytosis and signaling. Here, we discuss the emerging paradigm of intracellular targeting of GPCRs, and outline the current challenges, potential opportunities, and future outlook in this particular area of GPCR biology.

Identification of Polo-like kinase 1 interaction inhibitors using a novel cell-based assay

Polo-like kinase 1 (Plk1) plays several roles in cell division and it is a recognized cancer drug target. Plk1 levels are elevated in cancer and several types of cancer cells are hypersensitive to Plk1 inhibition. Small molecule inhibitors of the kinase domain (KD) of Plk1 have been developed. Their selectivity is limited, which likely contributes to their toxicity. Polo-like kinases are characterized by a Polo-Box Domain (PBD), which mediates interactions with phosphorylation substrates or regulators. Inhibition of the PBD could allow better selectivity or result in different effects than inhibition of the KD. In vitro screens have been used to identify PBD inhibitors with mixed results. We developed the first cell-based assay to screen for PBD inhibitors, using Bioluminescence Resonance Energy Transfer (BRET). We screened through 112 983 compounds and characterized hits in secondary biochemical and biological assays. Subsequent Structure-Activity Relationship (SAR) analysis on our most promising hit revealed that it requires an alkylating function for its activity. In addition, we show that the previously reported PBD inhibitors thymoquinone and Poloxin are also alkylating agents. Our cell-based assay is a promising tool for the identification of new PBD inhibitors with more drug-like profiles using larger and more diverse chemical libraries.

Mapping the putative G protein-coupled receptor (GPCR) docking site on GPCR kinase 2: insights from intact cell phosphorylation and recruitment assays

G protein-coupled receptor kinases (GRKs) phosphorylate agonist-occupied receptors initiating the processes of desensitization and β-arrestin-dependent signaling. Interaction of GRKs with activated receptors serves to stimulate their kinase activity. The extreme N-terminal helix (αN), the kinase small lobe, and the active site tether (AST) of the AGC kinase domain have previously been implicated in mediating the allosteric activation. Expanded mutagenesis of the αN and AST allowed us to further assess the role of these two regions in kinase activation and receptor phosphorylation in vitro and in intact cells. We also developed a bioluminescence resonance energy transfer-based assay to monitor the recruitment of GRK2 to activated α_2A-adrenergic receptors (α_2AARs) in living cells. The bioluminescence resonance energy transfer signal exhibited a biphasic response to norepinephrine concentration, suggesting that GRK2 is recruited to Gβγ and α_2AAR with EC₅₀ values of 15 nm and 8 μm, respectively. We show that mutations in αN (L4A, V7E, L8E, V11A, S12A, Y13A, and M17A) and AST (G475I, V477D, and I485A) regions impair or potentiate receptor phosphorylation and/or recruitment. We suggest that a surface of GRK2, including Leu⁴, Val⁷, Leu⁸, Val¹¹, and Ser¹², directly interacts with receptors, whereas residues such as Asp¹⁰, Tyr¹³, Ala¹⁶, Met¹⁷, Gly⁴⁷⁵, Val⁴⁷⁷, and Ile⁴⁸⁵ are more important for kinase domain closure and activation. Taken together with data on GRK1 and GRK6, our data suggest that all three GRK subfamilies make conserved interactions with G protein-coupled receptors, but there may be unique interactions that influence selectivity.

Do Natural Silks Make Good Engineering Materials?

Fast relaxation of stresses lower than the yield stress is demonstrated in Bombyx mori(silkworm) cocoon silk and Nephila clavipes (spider) major ampullate silk (MAS; dragline). Stress relaxation and creep make natural silk unsuitable as a long-term load-bearing material. Instead, silk-like materials are better suited to applications in which energy dissipation is important, and in which high loads need to be withstood on a once-off basis for only very short periods of time. Examples might include use as a ballistic material that arrests the penetration of fragments from the explosion of a pressure vessel, an aircraft luggage container, or a tyre. Treatment in a domestic microwave oven is shown to significantly reduce the rate of stress relaxation in both silkworm cocoon and spider MAS. Except for ductility, the tensile properties of cocoon silk measured in constant strain rate experiments are enhanced by this treatment. Initial experiments on MAS suggest that the tensile properties of this material also are enhanced by exposure to microwaves, in this case with the exception of initial modulus.

SHEF: a vHTS geometrical filter using coefficients of spherical harmonic molecular surfaces

SHEF (spherical harmonic coefficient filter), a geometrical matching procedure constituting a preliminary step in the virtual high throughput screening of large databases of small drug-like molecules, is demonstrated. This filter uses a description of both the binding site of the target and the ligand surfaces using spherical harmonic polynomial expansions. Using this representation, which is based on limited sets of spherical harmonic coefficients, considerably reduces the complexity of surface complementarity calculation. As a first test, 188 known protein-ligand complexes were used, and the results of docking the abstracted ligands into the bare proteins using SHEF were compared to the original X-ray structures. The ability of SHEF to retrieve known ligands "hidden" in a virtual library of 1,000 randomly selected drug-like compounds is also demonstrated.

A fast surface-matching procedure for protein–ligand docking

A very simple, fast, and efficient scheme is proposed for performing preliminary protein-ligand docking as the first step of intensive high-throughput virtual screening. The procedure acts as a surface-complementarity filter that first calculates the 2D-contour maps of both the protein cavity and of the ligands using a spherical harmonics description of the associated molecular surfaces. Next, the obtained 2D-fingerprint images are compared to detect their complementarity. This scheme was tested on three typical cases of protein cavities, namely, a well-closed pocket, a small open pocket, and a large open one. For that purpose, for each case, a sample of 101 ligand conformers was generated (the X-ray one and 100 different conformers generated using simulated annealing), and these conformational samples were ranked according to the complementarity with the protein cavity surface. Compared to traditional docking procedures such as FRED (considered as typical of a very fast rigid body docking algorithms) and GOLD (considered as typical of the more accurate flexible docking algorithms), our procedure was much faster and more successful in detecting the right X-ray conformation. We did, however, identify a certain weakness in the case of the very large pocket where results were not as expected. In general, our method could be used for incorporating indirectly flexibility in protein-ligand docking calculations as such a scheme can easily handle several conformational states of both the protein and the ligand.