CENsible: Interpretable Insights into Small-Molecule Binding with Context Explanation Networks.

We present a novel and interpretable approach for assessing small-molecule binding using context explanation networks. Given the specific structure of a protein/ligand complex, our CENsible scoring function uses a deep convolutional neural network to predict the contributions of precalculated terms to the overall binding affinity. We show that CENsible can effectively distinguish active vs inactive compounds for many systems. Its primary benefit over related machine-learning scoring functions, however, is that it retains interpretability, allowing researchers to identify the contribution of each precalculated term to the final affinity prediction, with implications for subsequent lead optimization.

Protocol for the comprehensive biochemical and cellular profiling of small-molecule degraders using CoraFluor TR-FRET technology.

Comprehensive characterization of small-molecule degraders, including binary and ternary complex formation and degradation efficiency, is critical for bifunctional ligand development and understanding structure-activity relationships. Here, we present a protocol for the biochemical and cellular profiling of small-molecule degraders based on CoraFluor time-resolved fluorescence resonance energy transfer (TR-FRET) technology. We describe steps for labeling antibodies and proteins, tracer saturation binding, binary target engagement, ternary complex profiling, and off-rate determination. We then detail procedures for the quantification of endogenous and GFP fusion proteins in cell lysates. For complete details on the use and execution of this protocol, please refer to Ichikawa et al.1.

Molecular glues for protein-protein interactions: Progressing toward a new dream.

The modulation of protein-protein interactions with small molecules is one of the most rapidly developing areas in drug discovery. In this review, we discuss advances over the past decade (2014-2023) focusing on molecular glues (MGs)-monovalent small molecules that induce proximity, either by stabilizing native interactions or by inducing neomorphic interactions. We include both serendipitous and rational discoveries and describe the different approaches that were used to identify them. We classify the compounds in three main categories: degradative MGs, non-degradative MGs or PPI stabilizers, and MGs that induce self-association. Diverse, illustrative examples with structural data are described in detail, emphasizing the elements of molecular recognition and cooperative binding at the interface that are fundamental for a MG mechanism of action.

Hybrid Small-Molecule/Protein Fluorescent Probes.

Hybrid small-molecule/protein fluorescent probes are powerful tools for visualizing protein localization and function in living cells. These hybrid probes are constructed by diverse site-specific chemical protein labeling approaches through chemical reactions to exogenous peptide/small protein tags, enzymatic post-translational modifications, bioorthogonal reactions for genetically incorporated unnatural amino acids, and ligand-directed chemical reactions. The hybrid small-molecule/protein fluorescent probes are employed for imaging protein trafficking, conformational changes, and bioanalytes surrounding proteins. In addition, fluorescent hybrid probes facilitate visualization of protein dynamics at the single-molecule level and the defined structure with super-resolution imaging. In this review, we discuss development and the bioimaging applications of fluorescent probes based on small-molecule/protein hybrids.

Discovery of a Plant 14-3-3 Inhibitor Possessing Isoform Selectivity and In Planta Activity.

Synthetic modulators of plant 14-3-3s are promising chemical tools both for understanding the 14-3-3-related signaling pathways and controlling plant physiology. Herein, we describe a novel small-molecule inhibitor for 14-3-3 proteins of Arabidopsis thaliana. The inhibitor was identified from unexpected products in a stock solution in dimethyl sulfoxide (DMSO) of an in-house chemical library. Mass spectroscopy, mutant-based analyses, fluorescence polarization assays, and thermal shift assays revealed that the inhibitor covalently binds to an allosteric site of 14-3-3 with isoform selectivity. Moreover, infiltration of the inhibitor to Arabidopsis leaves suppressed the stomatal aperture. The inhibitor should provide new insight into the design of potent and isoform-selective 14-3-3 modulators.

Large Libraries of Structurally Diverse Macrocycles Suitable for Membrane Permeation.

Macrocycles offer an attractive format for drug development due to their good binding properties and potential to cross cell membranes. To efficiently identify macrocyclic ligands for new targets, methods for the synthesis and screening of large combinatorial libraries of small cyclic peptides were developed, many of them using thiol groups for efficient peptide macrocyclization. However, a weakness of these libraries is that invariant thiol-containing building blocks such as cysteine are used, resulting in a region that does not contribute to library diversity but increases molecule size. Herein, we synthesized a series of structurally diverse thiol-containing elements and used them for the combinatorial synthesis of a 2,688-member library of small, structurally diverse peptidic macrocycles with unprecedented skeletal complexity. We then used this library to discover potent thrombin and plasma kallikrein inhibitors, some also demonstrating favorable membrane permeability. X-ray structure analysis of macrocycle-target complexes showed that the size and shape of the newly developed thiol elements are key for binding. The strategy and library format presented in this work significantly enhance structural diversity by allowing combinatorial modifications to a previously invariant region of peptide macrocycles, which may be broadly applied in the development of membrane permeable therapeutics.

Emerging strategies for prospective discovery of molecular glue degraders.

Molecular glue (MG) degraders are monovalent small molecule compounds that co-opt E3 ubiquitin ligases to target neo-substrates for proteasomal degradation. Here, we provide a concise review of recent advances in rational MG discovery, which are categorized into two major strategies, ligand modification and de novo discovery. We also highlight the structural mechanisms underlying the formation of MG-enabled ternary complexes and their thermodynamic properties. Finally, we summarize the broader category of proximity inducers including MGs, proteolysis-targeting chimeras (PROTACs), peptides, and viral proteins. MGs are specified as a unique class of proximity inducers with chemical simplicity and a requirement of pre-existing weak protein-protein interactions. We propose that leveraging the weak basal interaction provides a starting point to prospectively develop MGs to degrade high-value therapeutic targets.

Expanding the ligand spaces for E3 ligases for the design of protein degraders.

Targeted protein degradation (TPD) has recently emerged as an exciting new drug modality. However, the strategy of developing small molecule-based protein degraders has evolved over the past two decades and has now established molecular tags that are already in clinical use, as well as chimeric molecules, PROteolysis TArgeting Chimeras (PROTACs), based mainly on ligand systems developed for the two E3 ligases CRBN and VHL. The large size of the human E3 ligase family suggests that PROTACs can be developed by targeting a large diversity of E3 ligases, some of which have restricted expression patterns with the potential to design disease- or tissue-specific degraders. Indeed, many new E3 ligands have been published recently, confirming the druggability of E3 ligases. This review summarises recent data on E3 ligases and highlights the challenges in developing these molecules into efficient PROTACs rivalling the established degrader systems.

Small-Molecule Fluorescent Probes for Binding- and Activity-Based Sensing of Redox-Active Biological Metals.

Although transition metals constitute less than 0.1% of the total mass within a human body, they have a substantial impact on fundamental biological processes across all kingdoms of life. Indeed, these nutrients play crucial roles in the physiological functions of enzymes, with the redox properties of many of these metals being essential to their activity. At the same time, imbalances in transition metal pools can be detrimental to health. Modern analytical techniques are helping to illuminate the workings of metal homeostasis at a molecular and atomic level, their spatial localization in real time, and the implications of metal dysregulation in disease pathogenesis. Fluorescence microscopy has proven to be one of the most promising non-invasive methods for studying metal pools in biological samples. The accuracy and sensitivity of bioimaging experiments are predominantly determined by the fluorescent metal-responsive sensor, highlighting the importance of rational probe design for such measurements. This review covers activity- and binding-based fluorescent metal sensors that have been applied to cellular studies. We focus on the essential redox-active metals: iron, copper, manganese, cobalt, chromium, and nickel. We aim to encourage further targeted efforts in developing innovative approaches to understanding the biological chemistry of redox-active metals.

DeepP450: Predicting Human P450 Activities of Small Molecules by Integrating Pretrained Protein Language Model and Molecular Representation.

Cytochrome P450 enzymes (CYPs) play a crucial role in Phase I drug metabolism in the human body, and CYP activity toward compounds can significantly affect druggability, making early prediction of CYP activity and substrate identification essential for therapeutic development. Here, we established a deep learning model for assessing potential CYP substrates, DeepP450, by fine-tuning protein and molecule pretrained models through feature integration with cross-attention and self-attention layers. This model exhibited high prediction accuracy (0.92) on the test set, with area under the receiver operating characteristic curve (AUROC) values ranging from 0.89 to 0.98 in substrate/nonsubstrate predictions across the nine major human CYPs, surpassing current benchmarks for CYP activity prediction. Notably, DeepP450 uses only one model to predict substrates/nonsubstrates for any of the nine CYPs and exhibits certain generalizability on novel compounds and different categories of human CYPs, which could greatly facilitate early stage drug design by avoiding CYP-reactive compounds.

Advancing ASMS with LC-MS/MS for the discovery of novel PDCL2 ligands from DNA-encoded chemical library selections.

BACKGROUND: A safe, effective, and reversible nonhormonal male contraceptive drug is greatly needed for male contraception as well as for circumventing the side effects of female hormonal contraceptives. Phosducin-like 2 (PDCL2) is a testis-specific phosphoprotein in mice and humans. We recently found that male PDCL2 knockout mice are sterile due to globozoospermia caused by impaired sperm head formation, indicating that PDCL2 is a potential target for male contraception. Herein, our study for the first time developed a biophysical assay for PDCL2 allowing us to screen a series of small molecules, to study structure-activity relationships, and to discover two PDCL2 binders with novel chemical structure. OBJECTIVE: To identify a PDCL2 ligand for therapeutic male contraception, we performed DNA-encoded chemical library (DECL) screening and off-DNA hit validation using a unique affinity selection mass spectrometry (ASMS) biophysical profiling strategy. MATERIALS AND METHODS: We employed the screening process of DECL, which contains billions of chemically unique DNA-barcoded compounds generated through individual sequences of reactions and different combinations of functionalized building blocks. The structures of the PDCL2 binders are proposed based on the sequencing analysis of the DNA barcode attached to each individual DECL compound. The proposed structure is synthesized through multistep reactions. To confirm and determine binding affinity between the DECL identified molecules and PDCL2, we developed an ASMS assay that incorporates liquid chromatography with tandem mass spectrometry (LC-MS/MS). RESULTS: After a screening process of PDCL2 with DECLs containing >440 billion compounds, we identified a series of hits. The selected compounds were synthesized as off-DNA small molecules, characterized by spectroscopy data, and subjected to our ASMS/LC-MS/MS binding assay. By this assay, we discovered two novel compounds, which showed good binding affinity for PDCL2 in comparison to other molecules generated in our laboratory and which were further confirmed by a thermal shift assay. DISCUSSION AND CONCLUSION AND RELEVANCE: With the ASMS/LC-MS/MS assay developed in this paper, we successfully discovered a PDCL2 ligand that warrants further development as a male contraceptive.

Enhancing interaction of actin and actin-binding domain 1 of dystrophin with modulators: Toward improved gene therapy for Duchenne muscular dystrophy.

Duchenne muscular dystrophy is a lethal muscle disease, caused by mutations in the gene encoding dystrophin, an actin-binding cytoskeletal protein. Absence of functional dystrophin results in muscle weakness and degeneration, eventually leading to cardiac and respiratory failure. Strategies to replace the missing dystrophin via gene therapy have been intensively pursued. However, the dystrophin gene is too large for current gene therapy approaches. Currently available micro-dystrophin constructs lack the actin-binding domain 2 and show decreased actin-binding affinity in vitro compared to full-length dystrophin. Thus, increasing the actin-binding affinity of micro-dystrophin, using small molecules, could be a beneficial therapeutic approach. Here, we have developed and validated a novel high-throughput screening (HTS) assay to discover small molecules that increase the binding affinity of dystrophin's actin-binding domain 1 (ABD1). We engineered a novel FRET biosensor, consisting of the mClover3, fluorescent protein (donor) attached to the C-terminus of dystrophin ABD1, and Alexa Fluor 568 (acceptor) attached to the C-terminal cysteine of actin. We used this biosensor in small-molecule screening, using a unique high-precision, HTS fluorescence lifetime assay, identifying several compounds from an FDA-approved library that significantly increase the binding between actin and ABD1. This HTS assay establishes feasibility for the discovery of small-molecule modulators of the actin-dystrophin interaction, with the ultimate goal of developing therapies for muscular dystrophy.

DEL Selections Against a Soluble Protein Target.

Affinity-based DNA-encoded library (DEL) selection is considered a powerful tool for small molecule drug discovery. Such selections are a multi-round process that involves incubation of a target protein with the DEL, capture of the protein and associated DEL compounds on a solid support, separation of bound molecules from the bulk DEL that is unbound, and recovery of bound DEL molecules. Each step is of great importance in order to achieve successful selections. Here we describe the selection process against a soluble target protein in both the immobilized and in-solution modes.

Isolation of a Natural Killer Group 2D Small-Molecule Ligand from DNA-Encoded Chemical Libraries.

Natural Killer Group 2D (NKG2D) is a homo-dimeric transmembrane protein which is typically expressed on the surface of natural killer (NK) cells, natural killer T (NKT) cells, gamma delta T (γδT) cells, activated CD8 positive T-cells and activated macrophages. Bispecific molecules, capable of bridging NKG2D with a target protein expressed on the surface of tumor cells, may be used to redirect the cytotoxic activity of NK-cells towards antigen-positive malignant T-cells. In this work, we report the discovery of a novel NKG2D small molecule binder [KD =(410±60) nM], isolated from a DNA-Encoded Chemical Library (DEL). The discovery of small organic NKG2D ligands may facilitate the generation of fully synthetic bispecific adaptors, which may serve as an alternative to bispecific antibody products and which may benefit from better tumor targeting properties.

ATP synthase FOF1 structure, function, and structure-based drug design.

ATP synthases are unique rotatory molecular machines that supply biochemical reactions with adenosine triphosphate (ATP)-the universal "currency", which cells use for synthesis of vital molecules and sustaining life. ATP synthases of F-type (FOF1) are found embedded in bacterial cellular membrane, in thylakoid membranes of chloroplasts, and in mitochondrial inner membranes in eukaryotes. The main functions of ATP synthases are control of the ATP synthesis and transmembrane potential. Although the key subunits of the enzyme remain highly conserved, subunit composition and structural organization of ATP synthases and their assemblies are significantly different. In addition, there are hypotheses that the enzyme might be involved in the formation of the mitochondrial permeability transition pore and play a role in regulation of the cell death processes. Dysfunctions of this enzyme lead to numerous severe disorders with high fatality levels. In our review, we focus on FOF1-structure-based approach towards development of new therapies by using FOF1 structural features inherited by the representatives of this enzyme family from different taxonomy groups. We analyzed and systematized the most relevant information about the structural organization of FOF1 to discuss how this approach might help in the development of new therapies targeting ATP synthases and design tools for cellular bioenergetics control.

Switchable assembly and function of antibody complexes in vivo using a small molecule.

The antigen specificity and long serum half-life of monoclonal antibodies have made them a critical part of modern therapeutics. These properties have been coopted in a number of synthetic formats, such as antibody-drug conjugates, bispecific antibodies, or Fc-fusion proteins to generate novel biologic drug modalities. Historically, these new therapies have been generated by covalently linking multiple molecular moieties through chemical or genetic methods. This irreversible fusion of different components means that the function of the molecule is static, as determined by the structure. Here, we report the development of a technology for switchable assembly of functional antibody complexes using chemically induced dimerization domains. This approach enables control of the antibody's intended function in vivo by modulating the dose of a small molecule. We demonstrate this switchable assembly across three therapeutically relevant functionalities in vivo, including localization of a radionuclide-conjugated antibody to an antigen-positive tumor, extension of a cytokine's half-life, and activation of bispecific, T cell-engaging antibodies.

Ultralarge Virtual Screening Identifies SARS-CoV-2 Main Protease Inhibitors with Broad-Spectrum Activity against Coronaviruses.

Drugs targeting SARS-CoV-2 could have saved millions of lives during the COVID-19 pandemic, and it is now crucial to develop inhibitors of coronavirus replication in preparation for future outbreaks. We explored two virtual screening strategies to find inhibitors of the SARS-CoV-2 main protease in ultralarge chemical libraries. First, structure-based docking was used to screen a diverse library of 235 million virtual compounds against the active site. One hundred top-ranked compounds were tested in binding and enzymatic assays. Second, a fragment discovered by crystallographic screening was optimized guided by docking of millions of elaborated molecules and experimental testing of 93 compounds. Three inhibitors were identified in the first library screen, and five of the selected fragment elaborations showed inhibitory effects. Crystal structures of target-inhibitor complexes confirmed docking predictions and guided hit-to-lead optimization, resulting in a noncovalent main protease inhibitor with nanomolar affinity, a promising in vitro pharmacokinetic profile, and broad-spectrum antiviral effect in infected cells.

Molecular Basis of Small-Molecule Binding to α-Synuclein.

Intrinsically disordered proteins (IDPs) are implicated in many human diseases. They have generally not been amenable to conventional structure-based drug design, however, because their intrinsic conformational variability has precluded an atomic-level understanding of their binding to small molecules. Here we present long-time-scale, atomic-level molecular dynamics (MD) simulations of monomeric α-synuclein (an IDP whose aggregation is associated with Parkinson's disease) binding the small-molecule drug fasudil in which the observed protein-ligand interactions were found to be in good agreement with previously reported NMR chemical shift data. In our simulations, fasudil, when bound, favored certain charge-charge and π-stacking interactions near the C terminus of α-synuclein but tended not to form these interactions simultaneously, rather breaking one of these interactions and forming another nearby (a mechanism we term dynamic shuttling). Further simulations with small molecules chosen to modify these interactions yielded binding affinities and key structural features of binding consistent with subsequent NMR experiments, suggesting the potential for MD-based strategies to facilitate the rational design of small molecules that bind with disordered proteins.

Harnessing an emissive guanine surrogate to design small-molecule fluorescent chemosensors of O6-methylguanine-DNA-methyltransferase (MGMT).

The fluorescence properties of an emissive guanine surrogate, thienoguanine (thGN, 2-aminothieno[3,4-d]pyrimidin-4(3H)-one), were exploited to design two real-time chemosensors of O6-methylguanine-DNA-methyltransferase (MGMT), a key DNA repair enzyme involved in the resistance to DNA-alkylating anti-cancer drugs though direct reversal of O6-alkylated guanine adducts.

High-throughput metabolomics predicts drug-target relationships for eukaryotic proteins.

Chemical probes are important tools for understanding biological systems. However, because of the huge combinatorial space of targets and potential compounds, traditional chemical screens cannot be applied systematically to find probes for all possible druggable targets. Here, we demonstrate a novel concept for overcoming this challenge by leveraging high-throughput metabolomics and overexpression to predict drug-target interactions. The metabolome profiles of yeast treated with 1,280 compounds from a chemical library were collected and compared with those of inducible yeast membrane protein overexpression strains. By matching metabolome profiles, we predicted which small molecules targeted which signaling systems and recovered known interactions. Drug-target predictions were generated across the 86 genes studied, including for difficult to study membrane proteins. A subset of those predictions were tested and validated, including the novel targeting of GPR1 signaling by ibuprofen. These results demonstrate the feasibility of predicting drug-target relationships for eukaryotic proteins using high-throughput metabolomics.