AI-driven molecular generation of not-patented pharmaceutical compounds using world open patent data.

Developing compounds with novel structures is important for the production of new drugs. From an intellectual perspective, confirming the patent status of newly developed compounds is essential, particularly for pharmaceutical companies. The generation of a large number of compounds has been made possible because of the recent advances in artificial intelligence (AI). However, confirming the patent status of these generated molecules has been a challenge because there are no free and easy-to-use tools that can be used to determine the novelty of the generated compounds in terms of patents in a timely manner; additionally, there are no appropriate reference databases for pharmaceutical patents in the world. In this study, two public databases, SureChEMBL and Google Patents Public Datasets, were used to create a reference database of drug-related patented compounds using international patent classification. An exact structure search system was constructed using InChIKey and a relational database system to rapidly search for compounds in the reference database. Because drug-related patented compounds are a good source for generative AI to learn useful chemical structures, they were used as the training data. Furthermore, molecule generation was successfully directed by increasing and decreasing the number of generated patented compounds through incorporation of patent status (i.e., patented or not) into learning. The use of patent status enabled generation of novel molecules with high drug-likeness. The generation using generative AI with patent information would help efficiently propose novel compounds in terms of pharmaceutical patents. Scientific contribution: In this study, a new molecule-generation method that takes into account the patent status of molecules, which has rarely been considered but is an important feature in drug discovery, was developed. The method enables the generation of novel molecules based on pharmaceutical patents with high drug-likeness and will help in the efficient development of effective drug compounds.

PoSSuM v.3: A Major Expansion of the PoSSuM Database for Finding Similar Binding Sites of Proteins.

Information on structures of protein-ligand complexes, including comparisons of known and putative protein-ligand-binding pockets, is valuable for protein annotation and drug discovery and development. To facilitate biomedical and pharmaceutical research, we developed PoSSuM (https://possum.cbrc.pj.aist.go.jp/PoSSuM/), a database for identifying similar binding pockets in proteins. The current PoSSuM database includes 191 million similar pairs among almost 10 million identified pockets. PoSSuM drug search (PoSSuMds) is a resource for investigating ligand and receptor diversity among a set of pockets that can bind to an approved drug compound. The enhanced PoSSuMds covers pockets associated with both approved drugs and drug candidates in clinical trials from the latest release of ChEMBL. Additionally, we developed two new databases: PoSSuMAg for investigating antibody-antigen interactions and PoSSuMAF to simplify exploring putative pockets in AlphaFold human protein models.

Applying deep learning to iterative screening of medium-sized molecules for protein-protein interaction-targeted drug discovery.

We combined a library of medium-sized molecules with iterative screening using multiple machine learning algorithms that were ligand-based, which resulted in a large increase of the hit rate against a protein-protein interaction target. This was demonstrated by inhibition assays using a PPI target, Kelch-like ECH-associated protein 1/nuclear factor erythroid 2-related factor 2 (Keap1/Nrf2), and a deep neural network model based on the first-round assay data showed a highest hit rate of 27.3%. Using the models, we identified novel active and non-flat compounds far from public datasets, expanding the chemical space.

TDP-43 N-terminal domain dimerisation or spatial separation by RNA binding decreases its propensity to aggregate.

Aggregation of the 43 kDa TAR DNA-binding protein (TDP-43) is a pathological hallmark of amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD). RNA binding and TDP-43 N-terminal domain dimerisation has been suggested to ameliorate TDP-43 aggregation. However, the relationship between these factors and the solubility of TDP-43 is largely unknown. Therefore, we developed new oligonucleotides that can recruit two TDP-43 molecules and interfere with their intermolecular interactions via spatial separation. Using these oligonucleotides and TDP-43-preferable UG-repeats, we uncovered two distinct mechanisms for modulating TDP-43 solubility by RNA binding: One is N-terminal domain dimerisation, and the other is the spatial separation of two TDP-43 molecules. This study provides new molecular insights into the regulation of TDP-43 solubility.

Paip2A inhibits translation by competitively binding to the RNA recognition motifs of PABPC1 and promoting its dissociation from the poly(A) tail.

Eukaryotic mRNAs possess a poly(A) tail at their 3'-end, to which poly(A)-binding protein C1 (PABPC1) binds and recruits other proteins that regulate translation. Enhanced poly(A)-dependent translation, which is also PABPC1 dependent, promotes cellular and viral proliferation. PABP-interacting protein 2A (Paip2A) effectively represses poly(A)-dependent translation by causing the dissociation of PABPC1 from the poly(A) tail; however, the underlying mechanism remains unknown. This study was conducted to investigate the functional mechanisms of Paip2A action by characterizing the PABPC1-poly(A) and PABPC1-Paip2A interactions. Isothermal titration calorimetry and NMR analyses indicated that both interactions predominantly occurred at the RNA recognition motif (RRM)2-RRM3 regions of PABPC1, which have comparable affinities for poly(A) and Paip2A (dissociation constant, Kd = 1 nM). However, the Kd values of isolated RRM2 were 200 and 4 µM in their interactions with poly(A) and Paip2A, respectively; Kd values of 5 and 1 µM were observed for the interactions of isolated RRM3 with poly(A) and Paip2A, respectively. NMR analyses also revealed that Paip2A can bind to the poly(A)-binding interfaces of the RRM2 and RRM3 regions of PABPC1. Based on these results, we propose the following functional mechanism for Paip2A: Paip2A initially binds to the RRM2 region of poly(A)-bound PABPC1, and RRM2-anchored Paip2A effectively displaces the RRM3 region from poly(A), resulting in dissociation of the whole PABPC1 molecule. Together, our findings provide insight into the translation repression effect of Paip2A and may aid in the development of novel anticancer and/or antiviral drugs.

DLiP-PPI library: An integrated chemical database of small-to-medium-sized molecules targeting protein-protein interactions.

Protein-protein interactions (PPIs) are recognized as important targets in drug discovery. The characteristics of molecules that inhibit PPIs differ from those of small-molecule compounds. We developed a novel chemical library database system (DLiP) to design PPI inhibitors. A total of 32,647 PPI-related compounds are registered in the DLiP. It contains 15,214 newly synthesized compounds, with molecular weight ranging from 450 to 650, and 17,433 active and inactive compounds registered by extracting and integrating known compound data related to 105 PPI targets from public databases and published literature. Our analysis revealed that the compounds in this database contain unique chemical structures and have physicochemical properties suitable for binding to the protein-protein interface. In addition, advanced functions have been integrated with the web interface, which allows users to search for potential PPI inhibitor compounds based on types of protein-protein interfaces, filter results by drug-likeness indicators important for PPI targeting such as rule-of-4, and display known active and inactive compounds for each PPI target. The DLiP aids the search for new candidate molecules for PPI drug discovery and is available online (https://skb-insilico.com/dlip).

Comprehensive Approach of 19F Nuclear Magnetic Resonance, Enzymatic, and In Silico Methods for Site-Specific Hit Selection and Validation of Fragment Molecules that Inhibit Methionine γ-Lyase Activity.

Fragment-based screening using 19F NMR (19F-FS) is an efficient method for exploring seed and lead compounds for drug discovery. Here, we demonstrate the utility and merits of using 19F-FS for methionine γ-lyase-binding fragments, together with a 19F NMR-based competition and mutation assay, as well as enzymatic and in silico methods. 19F NMR-based assays provided useful information on binding between 19F-FS hit fragments and target proteins. Although the 19F-FS and enzymatic assay were weakly correlated, they show that the 19F-FS hit fragments contained compounds with inhibitory activity. Furthermore, we found that in silico calculations partially account for the differences in activity levels between the 19F-FS hits as per NMR analysis. A comprehensive approach combining the 19F-FS and other methods not only identified fragment hits but also distinguished structural differences in chemical groups with diverse activity levels.

Peptide Toxins Targeting KV Channels.

A number of peptide toxins isolated from animals target potassium ion (K+) channels. Many of them are particularly known to inhibit voltage-gated K+ (KV) channels and are mainly classified into pore-blocking toxins or gating-modifier toxins. Pore-blocking toxins directly bind to the ion permeation pores of KV channels, thereby physically occluding them. In contrast, gating-modifier toxins bind to the voltage-sensor domains of KV channels, modulating their voltage-dependent conformational changes. These peptide toxins are useful molecular tools in revealing the structure-function relationship of KV channels and have potential for novel treatments for diseases related to KV channels. This review focuses on the inhibition mechanism of pore-blocking and gating-modifier toxins that target KV channels.

Identification of novel inhibitors of Keap1/Nrf2 by a promising method combining protein-protein interaction-oriented library and machine learning.

Protein-protein interactions (PPIs) are prospective but challenging targets for drug discovery, because screening using traditional small-molecule libraries often fails to identify hits. Recently, we developed a PPI-oriented library comprising 12,593 small-to-medium-sized newly synthesized molecules. This study validates a promising combined method using PPI-oriented library and ligand-based virtual screening (LBVS) to discover novel PPI inhibitory compounds for Kelch-like ECH-associated protein 1 (Keap1) and nuclear factor erythroid 2-related factor 2 (Nrf2). We performed LBVS with two random forest models against our PPI library and the following time-resolved fluorescence resonance energy transfer (TR-FRET) assays of 620 compounds identified 15 specific hit compounds. The high hit rates for the entire PPI library (estimated 0.56-1.3%) and the LBVS (maximum 5.4%) compared to a conventional screening library showed the utility of the library and the efficiency of LBVS. All the hit compounds possessed novel structures with Tanimoto similarity ≤ 0.26 to known Keap1/Nrf2 inhibitors and aqueous solubility (AlogP < 5). Reasonable binding modes were predicted using 3D alignment of five hit compounds and a Keap1/Nrf2 peptide crystal structure. Our results represent a new, efficient method combining the PPI library and LBVS to identify novel PPI inhibitory ligands with expanded chemical space.

Mechanism of hERG inhibition by gating-modifier toxin, APETx1, deduced by functional characterization.

BACKGROUND: Human ether-à-go-go-related gene potassium channel 1 (hERG) is a voltage-gated potassium channel, the voltage-sensing domain (VSD) of which is targeted by a gating-modifier toxin, APETx1. APETx1 is a 42-residue peptide toxin of sea anemone Anthopleura elegantissima and inhibits hERG by stabilizing the resting state. A previous study that conducted cysteine-scanning analysis of hERG identified two residues in the S3-S4 region of the VSD that play important roles in hERG inhibition by APETx1. However, mutational analysis of APETx1 could not be conducted as only natural resources have been available until now. Therefore, it remains unclear where and how APETx1 interacts with the VSD in the resting state. RESULTS: We established a method for preparing recombinant APETx1 and determined the NMR structure of the recombinant APETx1, which is structurally equivalent to the natural product. Electrophysiological analyses using wild type and mutants of APETx1 and hERG revealed that their hydrophobic residues, F15, Y32, F33, and L34, in APETx1, and F508 and I521 in hERG, in addition to a previously reported acidic hERG residue, E518, play key roles in the inhibition of hERG by APETx1. Our hypothetical docking models of the APETx1-VSD complex satisfied the results of mutational analysis. CONCLUSIONS: The present study identified the key residues of APETx1 and hERG that are involved in hERG inhibition by APETx1. These results would help advance understanding of the inhibitory mechanism of APETx1, which could provide a structural basis for designing novel ligands targeting the VSDs of KV channels.

GZD824 Inhibits GCN2 and Sensitizes Cancer Cells to Amino Acid Starvation Stress.

Eukaryotic initiation factor 2α (eIF2α) kinase general control nonderepressible 2 (GCN2) drives cellular adaptation to amino acid limitation by activating the integrated stress response that induces activating transcription factor 4 (ATF4). Here, we found that a multikinase inhibitor, GZD824, which we identified using a cell-based assay with ATF4 immunostaining, inhibited the GCN2 pathway in cancer cells. Indeed, GZD824 suppressed GCN2 activation, eIF2α phosphorylation, and ATF4 induction during amino acid starvation stress. However, at lower nonsuppressive concentrations, GZD824 paradoxically stimulated eIF2α phosphorylation and ATF4 expression in a GCN2-dependent manner under unstressed conditions. Such dual properties conceivably arose from a direct effect on GCN2, as also observed in a cell-free GCN2 kinase assay and shared by a selective GCN2 inhibitor. Consistent with the GCN2 pathway inhibition, GZD824 sensitized certain cancer cells to amino acid starvation stress similarly to ATF4 knockdown. These results establish GZD824 as a multikinase GCN2 inhibitor and may enhance its utility as a drug under development. SIGNIFICANCE STATEMENT: GZD824, as a direct general control nonderepressible 2 (GCN2) inhibitor, suppresses activation of the integrated stress response during amino acid limitation, whereas it paradoxically stimulates this stress-signaling pathway at lower nonsuppressive concentrations. The pharmacological activity we identify herein will provide the basis for the use of GZD824 to elucidate the regulatory mechanisms of GCN2 and to evaluate the potential of the GCN2-activating transcription factor 4 pathway as a target for cancer therapy.

Conformational equilibrium shift underlies altered K+ channel gating as revealed by NMR.

The potassium ion (K+) channel plays a fundamental role in controlling K+ permeation across the cell membrane and regulating cellular excitabilities. Mutations in the transmembrane pore reportedly affect the gating transitions of K+ channels, and are associated with the onset of neural disorders. However, due to the lack of structural and dynamic insights into the functions of K+ channels, the structural mechanism by which these mutations cause K+ channel dysfunctions remains elusive. Here, we used nuclear magnetic resonance spectroscopy to investigate the structural mechanism underlying the decreased K+-permeation caused by disease-related mutations, using the prokaryotic K+ channel KcsA. We demonstrated that the conformational equilibrium in the transmembrane region is shifted toward the non-conductive state with the closed intracellular K+-gate in the disease-related mutant. We also demonstrated that this equilibrium shift is attributable to the additional steric contacts in the open-conductive structure, which are evoked by the increased side-chain bulkiness of the residues lining the transmembrane helix. Our results suggest that the alteration in the conformational equilibrium of the intracellular K+-gate is one of the fundamental mechanisms underlying the dysfunctions of K+ channels caused by disease-related mutations.

Structural mechanism underlying G protein family-specific regulation of G protein-gated inwardly rectifying potassium channel.

G protein-gated inwardly rectifying potassium channel (GIRK) plays a key role in regulating neurotransmission. GIRK is opened by the direct binding of the G protein ßγ subunit (Gßγ), which is released from the heterotrimeric G protein (Gαßγ) upon the activation of G protein-coupled receptors (GPCRs). GIRK contributes to precise cellular responses by specifically and efficiently responding to the Gi/o-coupled GPCRs. However, the detailed mechanisms underlying this family-specific and efficient activation are largely unknown. Here, we investigate the structural mechanism underlying the Gi/o family-specific activation of GIRK, by combining cell-based BRET experiments and NMR analyses in a reconstituted membrane environment. We show that the interaction formed by the αA helix of Gαi/o mediates the formation of the Gαi/oßγ-GIRK complex, which is responsible for the family-specific activation of GIRK. We also present a model structure of the Gαi/oßγ-GIRK complex, which provides the molecular basis underlying the specific and efficient regulation of GIRK.

Nanodiscs for Structural Biology in a Membranous Environment.

The structures of many membrane proteins have been analyzed in detergent micelles. However, the environment of detergent micelles differs somewhat from that of the lipid bilayer, where membrane proteins exhibit physiological functions. Therefore, a more membrane-like environment has been awaited for structural analysis of membrane proteins. Nanodiscs are "hockey-puck"-shaped lipid bilayer particles that distribute in a monodispersed manner in aqueous solution. We review how nanodiscs or protein-reconstituted nanodiscs are prepared and how they are utilized to analyze protein structure, dynamics, and interactions with lipid molecules using solution NMR and cryo-electron microscopy.

Functional roles of Mg2+ binding sites in ion-dependent gating of a Mg2+ channel, MgtE, revealed by solution NMR.

Magnesium ions (Mg2+) are divalent cations essential for various cellular functions. Mg2+ homeostasis is maintained through Mg2+ channels such as MgtE, a prokaryotic Mg2+ channel whose gating is regulated by intracellular Mg2+ levels. Our previous crystal structure of MgtE in the Mg2+-bound, closed state revealed the existence of seven crystallographically-independent Mg2+-binding sites, Mg1-Mg7. The role of Mg2+-binding to each site in channel closure remains unknown. Here, we investigated Mg2+-dependent changes in the structure and dynamics of MgtE using nuclear magnetic resonance spectroscopy. Mg2+-titration experiments, using wild-type and mutant forms of MgtE, revealed that the Mg2+ binding sites Mg1, Mg2, Mg3, and Mg6, exhibited cooperativity and a higher affinity for Mg2+, enabling the remaining Mg2+ binding sites, Mg4, Mg5, and Mg7, to play important roles in channel closure. This study revealed the role of each Mg2+-binding site in MgtE gating, underlying the mechanism of cellular Mg2+ homeostasis.

Structural basis for the ethanol action on G-protein-activated inwardly rectifying potassium channel 1 revealed by NMR spectroscopy.

Ethanol consumption leads to a wide range of pharmacological effects by acting on the signaling proteins in the human nervous system, such as ion channels. Despite its familiarity and biological importance, very little is known about the molecular mechanisms underlying the ethanol action, due to extremely weak binding affinity and the dynamic nature of the ethanol interaction. In this research, we focused on the primary in vivo target of ethanol, G-protein-activated inwardly rectifying potassium channel (GIRK), which is responsible for the ethanol-induced analgesia. By utilizing solution NMR spectroscopy, we characterized the changes in the structure and dynamics of GIRK induced by ethanol binding. We demonstrated here that ethanol binds to GIRK with an apparent dissociation constant of 1.0 M and that the actual physiological binding site of ethanol is located on the cavity formed between the neighboring cytoplasmic regions of the GIRK tetramer. From the methyl-based NMR relaxation analyses, we revealed that ethanol activates GIRK by shifting the conformational equilibrium processes, which are responsible for the gating of GIRK, to stabilize an open conformation of the cytoplasmic ion gate. We suggest that the dynamic molecular mechanism of the ethanol-induced activation of GIRK represents a general model of the ethanol action on signaling proteins in the human nervous system.

Characterization of the multimeric structure of poly(A)-binding protein on a poly(A) tail.

Eukaryotic mature mRNAs possess a poly adenylate tail (poly(A)), to which multiple molecules of poly(A)-binding protein C1 (PABPC1) bind. PABPC1 regulates translation and mRNA metabolism by binding to regulatory proteins. To understand functional mechanism of the regulatory proteins, it is necessary to reveal how multiple molecules of PABPC1 exist on poly(A). Here, we characterize the structure of the multiple molecules of PABPC1 on poly(A), by using transmission electron microscopy (TEM), chemical cross-linking, and NMR spectroscopy. The TEM images and chemical cross-linking results indicate that multiple PABPC1 molecules form a wormlike structure in the PABPC1-poly(A) complex, in which the PABPC1 molecules are linearly arrayed. NMR and cross-linking analyses indicate that PABPC1 forms a multimer by binding to the neighbouring PABPC1 molecules via interactions between the RNA recognition motif (RRM) 2 in one molecule and the middle portion of the linker region of another molecule. A PABPC1 mutant lacking the interaction site in the linker, which possesses an impaired ability to form the multimer, reduced the in vitro translation activity, suggesting the importance of PABPC1 multimer formation in the translation process. We therefore propose a model of the PABPC1 multimer that provides clues to comprehensively understand the regulation mechanism of mRNA translation.

Nuclear Magnetic Resonance Approaches for Characterizing Protein-Protein Interactions.

The gating of potassium ion (K+) channels is regulated by various kinds of protein-protein interactions (PPIs). Structural investigations of these PPIs provide useful information not only for understanding the gating mechanisms of K+ channels, but also for developing the pharmaceutical compounds targeting K+ channels. Here, we describe a nuclear magnetic resonance spectroscopic method, termed the cross saturation (CS) method, to accurately determine the binding surfaces of protein complexes, and its application to the investigation of the interaction between a G protein-coupled inwardly rectifying K+ channel and a G protein α subunit.

ATP-dependent modulation of MgtE in Mg2+ homeostasis.

Magnesium is an essential ion for numerous physiological processes. MgtE is a Mg2+ selective channel involved in the maintenance of intracellular Mg2+ homeostasis, whose gating is regulated by intracellular Mg2+ levels. Here, we report that ATP binds to MgtE, regulating its Mg2+-dependent gating. Crystal structures of MgtE-ATP complex show that ATP binds to the intracellular CBS domain of MgtE. Functional studies support that ATP binding to MgtE enhances the intracellular domain affinity for Mg2+ within physiological concentrations of this divalent cation, enabling MgtE to function as an in vivo Mg2+ sensor. ATP dissociation from MgtE upregulates Mg2+ influx at both high and low intracellular Mg2+ concentrations. Using site-directed mutagenesis and structure based-electrophysiological and biochemical analyses, we identify key residues and main structural changes involved in the process. This work provides the molecular basis of ATP-dependent modulation of MgtE in Mg2+ homeostasis.MgtE is an Mg2+ transporter involved in Mg2+ homeostasis. Here, the authors report that ATP regulates the Mg+2-dependent gating of MgtE and use X-ray crystallography combined with functional studies to propose the molecular mechanisms involved in this process.

Dynamic regulation of GDP binding to G proteins revealed by magnetic field-dependent NMR relaxation analyses.

Heterotrimeric guanine-nucleotide-binding proteins (G proteins) serve as molecular switches in signalling pathways, by coupling the activation of cell surface receptors to intracellular responses. Mutations in the G protein α-subunit (Gα) that accelerate guanosine diphosphate (GDP) dissociation cause hyperactivation of the downstream effector proteins, leading to oncogenesis. However, the structural mechanism of the accelerated GDP dissociation has remained unclear. Here, we use magnetic field-dependent nuclear magnetic resonance relaxation analyses to investigate the structural and dynamic properties of GDP bound Gα on a microsecond timescale. We show that Gα rapidly exchanges between a ground-state conformation, which tightly binds to GDP and an excited conformation with reduced GDP affinity. The oncogenic D150N mutation accelerates GDP dissociation by shifting the equilibrium towards the excited conformation.