Explore the vast range of titles available.

Dysregulation of peptide-activated pathways causes a range of diseases, fostering the discovery and clinical development of peptide drugs. Many endogenous peptides activate G protein-coupled receptors (GPCRs) — nearly 50 GPCR peptide drugs have been approved to date, most of them for metabolic disease or oncology, and more than 10 potentially first-in-class peptide therapeutics are in the pipeline. The majority of existing peptide therapeutics are agonists, which reflects the currently dominant strategy of modifying the endogenous peptide sequence of ligands for peptide-binding GPCRs. Increasingly, novel strategies are being employed to develop both agonists and antagonists, to both introduce chemical novelty and improve drug-like properties. Pharmacodynamic improvements are evolving to allow biasing ligands to activate specific downstream signalling pathways, in order to optimize efficacy and reduce side effects. In pharmacokinetics, modifications that increase plasma half-life have been revolutionary. Here, we discuss the current status of the peptide drugs targeting GPCRs, with a focus on evolving strategies to improve pharmacokinetic and pharmacodynamic properties.Many G protein-coupled receptors (GPCRs) have endogenous peptide agonists, and modifying the sequence of these peptides has led to some successful therapeutics. In this Review, Davenport and colleagues discuss strategies to generate effective GPCR-targeted peptide therapeutics by introducing chemical novelty, extending plasma half-life, improving a therapeutic’s drug-like properties or generating biased ligands. These approaches could overcome some of the challenges in developing peptide therapeutics.

Many therapies in clinical trials are based on single drug-single target relationships. To further extend this concept to multi-target approaches using multi-targeted drugs, we developed a machine learning pipeline to unravel the target landscape of kinase inhibitors. This pipeline, which we call 3D-KINEssence, uses a new type of protein fingerprints (3D FP) based on the structure of kinases generated through a 3D convolutional neural network (3D-CNN). These 3D-CNN kinase fingerprints were matched to molecular Morgan fingerprints to predict the targets of each respective kinase inhibitor based on available bioactivity data. The performance of the pipeline was evaluated on two test sets: a sparse drug-target set where each drug is matched in most cases to a single target and also on a densely-covered drug-target set where each drug is matched to most if not all targets. This latter set is more challenging to train, given its non-exclusive character. Our model’s root-mean-square error (RMSE) based on the two datasets was 0.68 and 0.8, respectively. These results indicate that 3D FP can predict the target landscape of kinase inhibitors at around 0.8 log units of bioactivity. Our strategy can be utilized in proteochemometric or chemogenomic workflows by consolidating the target landscape of kinase inhibitors.

Deep generative models have shown the ability to devise both valid and novel chemistry, which could significantly accelerate the identification of bioactive compounds. Many current models, however, use molecular descriptors or ligand-based predictive methods to guide molecule generation towards a desirable property space. This restricts their application to relatively data-rich targets, neglecting those where little data is available to sufficiently train a predictor. Moreover, ligand-based approaches often bias molecule generation towards previously established chemical space, thereby limiting their ability to identify truly novel chemotypes. In this work, we assess the ability of using molecular docking via Glide—a structure-based approach—as a scoring function to guide the deep generative model REINVENT and compare model performance and behaviour to a ligand-based scoring function. Additionally, we modify the previously published MOSES benchmarking dataset to remove any induced bias towards non-protonatable groups. We also propose a new metric to measure dataset diversity, which is less confounded by the distribution of heavy atom count than the commonly used internal diversity metric. With respect to the main findings, we found that when optimizing the docking score against DRD2, the model improves predicted ligand affinity beyond that of known DRD2 active molecules. In addition, generated molecules occupy complementary chemical and physicochemical space compared to the ligand-based approach, and novel physicochemical space compared to known DRD2 active molecules. Furthermore, the structure-based approach learns to generate molecules that satisfy crucial residue interactions, which is information only available when taking protein structure into account. Overall, this work demonstrates the advantage of using molecular docking to guide de novo molecule generation over ligand-based predictors with respect to predicted affinity, novelty, and the ability to identify key interactions between ligand and protein target. Practically, this approach has applications in early hit generation campaigns to enrich a virtual library towards a particular target, and also in novelty-focused projects, where de novo molecule generation either has no prior ligand knowledge available or should not be biased by it.

G protein-coupled receptors (GPCRs) are the largest class of membrane receptors, playing a key role in the regulation of processes as varied as neurotransmission and immune response. Evidence for GPCR oligomerisation has been accumulating that challenges the idea that GPCRs function solely as monomeric receptors; however, GPCR oligomerisation remains controversial primarily due to the difficulties in comparing evidence from very different types of structural and dynamic data. Using a combination of single-molecule and ensemble FRET, double electron–electron resonance spectroscopy, and simulations, we show that dimerisation of the GPCR neurotensin receptor 1 is regulated by receptor density and is dynamically tuneable over the physiological range. We propose a “rolling dimer” interface model in which multiple dimer conformations co-exist and interconvert. These findings unite previous seemingly conflicting observations, provide a compelling mechanism for regulating receptor signalling, and act as a guide for future physiological studies. Evidence suggests oligomerisation of G protein-coupled receptors in membranes, but this is controversial. Here, authors use single-molecule and ensemble FRET, and spectroscopy to show that the neurotensin receptor 1 forms multiple dimer conformations that interconvert - “rolling” interfaces.

Although class A G protein–coupled receptors (GPCRs) can function as monomers, many of them form dimers and oligomers, but the mechanisms and functional relevance of such oligomerization is ill understood. Here, we investigate this problem for the CXC chemokine receptor 4 (CXCR4), a GPCR that regulates immune and hematopoietic cell trafficking, and a major drug target in cancer therapy. We combine single-molecule microscopy and fluorescence fluctuation spectroscopy to investigate CXCR4 membrane organization in living cells at densities ranging from a few molecules to hundreds of molecules per square micrometer of the plasma membrane. We observe that CXCR4 forms dynamic, transient homodimers, and that the monomer–dimer equilibrium is governed by receptor density. CXCR4 inverse agonists that bind to the receptor minor pocket inhibit CXCR4 constitutive activity and abolish receptor dimerization. A mutation in the minor binding pocket reduced the dimer-disrupting ability of these ligands. In addition, mutating critical residues in the sixth transmembrane helix of CXCR4 markedly diminished both basal activity and dimerization, supporting the notion that CXCR4 basal activity is required for dimer formation. Together, these results link CXCR4 dimerization to its density and to its activity. They further suggest that inverse agonists binding to the minor pocket suppress both dimerization and constitutive activity and may represent a specific strategy to target CXCR4.

Binding of the glucagon peptide to the glucagon receptor (GCGR) triggers the release of glucose from the liver during fasting; thus GCGR plays an important role in glucose homeostasis. Here we report the crystal structure of the seven transmembrane helical domain of human GCGR at 3.4 Å resolution, complemented by extensive site-specific mutagenesis, and a hybrid model of glucagon bound to GCGR to understand the molecular recognition of the receptor for its native ligand. Beyond the shared seven transmembrane fold, the GCGR transmembrane domain deviates from class A G-protein-coupled receptors with a large ligand-binding pocket and the first transmembrane helix having a ‘stalk’ region that extends three alpha-helical turns above the plane of the membrane. The stalk positions the extracellular domain (∼12 kilodaltons) relative to the membrane to form the glucagon-binding site that captures the peptide and facilitates the insertion of glucagon’s amino terminus into the seven transmembrane domain. The X-ray crystal structure of the human glucagon receptor, a potential drug target for type 2 diabetes, offers a structural basis for molecular recognition by class B G-protein-coupled receptors. Two class B human GPCR receptors G-protein-coupled receptors (GPCRs) are membrane proteins that act as sensors for a broad range of extracellular signals, including photons, ions, small organic molecules and even entire proteins. Approximately a third of known drugs target GPCRs. Until now all the published structures of GPCRs have been from class A GPCRs. In this issue of Nature two groups independently report the crystal structures of two receptors of the B family, the second largest of four family divisions based on primary sequence and pharmacology. Hollenstein et al . solved the structure of human corticotropin-releasing factor receptor 1. This GPCR binds to corticotropin-releasing hormone, a potent mediator of endocrine, autonomic, behavioral and immune responses to stress. In all known class A GPCRs, the ligand-binding sites are close to the extracellular boundaries of the receptors; in this GPCR, the antagonist (CP-376395) binds in a hydrophobic pocket located in the cytoplasmic half of the V-shaped receptor. Siu et al . solved the X-ray crystal structure of the human glucagon receptor. This GPCR binds to the glucagon peptide, which triggers the release of glucose from the liver, making it a potential drug target for type 2 diabetes. The structure reveals a larger ligand-binding pocket than that seen in class A GPCRs.

A plethora of AI-based techniques now exists to conduct de novo molecule generation that can devise molecules conditioned towards a particular endpoint in the context of drug design. One popular approach is using reinforcement learning to update a recurrent neural network or language-based de novo molecule generator. However, reinforcement learning can be inefficient, sometimes requiring up to 10 5 molecules to be sampled to optimize more complex objectives, which poses a limitation when using computationally expensive scoring functions like docking or computer-aided synthesis planning models. In this work, we propose a reinforcement learning strategy called Augmented Hill-Climb based on a simple, hypothesis-driven hybrid between REINVENT and Hill-Climb that improves sample-efficiency by addressing the limitations of both currently used strategies. We compare its ability to optimize several docking tasks with REINVENT and benchmark this strategy against other commonly used reinforcement learning strategies including REINFORCE, REINVENT (version 1 and 2), Hill-Climb and best agent reminder. We find that optimization ability is improved ~ 1.5-fold and sample-efficiency is improved ~ 45-fold compared to REINVENT while still delivering appealing chemistry as output. Diversity filters were used, and their parameters were tuned to overcome observed failure modes that take advantage of certain diversity filter configurations. We find that Augmented Hill-Climb outperforms the other reinforcement learning strategies used on six tasks, especially in the early stages of training or for more difficult objectives. Lastly, we show improved performance not only on recurrent neural networks but also on a reinforcement learning stabilized transformer architecture. Overall, we show that Augmented Hill-Climb improves sample-efficiency for language-based de novo molecule generation conditioning via reinforcement learning, compared to the current state-of-the-art. This makes more computationally expensive scoring functions, such as docking, more accessible on a relevant timescale.

Generative models are undergoing rapid research and application to de novo drug design. To facilitate their application and evaluation, we present MolScore. MolScore already contains many drug-design-relevant scoring functions commonly used in benchmarks such as, molecular similarity, molecular docking, predictive models, synthesizability, and more. In addition, providing performance metrics to evaluate generative model performance based on the chemistry generated. With this unification of functionality, MolScore re-implements commonly used benchmarks in the field (such as GuacaMol, MOSES, and MolOpt). Moreover, new benchmarks can be created trivially. We demonstrate this by testing a chemical language model with reinforcement learning on three new tasks of increasing complexity related to the design of 5-HT 2a ligands that utilise either molecular descriptors, 266 pre-trained QSAR models, or dual molecular docking. Lastly, MolScore can be integrated into an existing Python script with just three lines of code. This framework is a step towards unifying generative model application and evaluation as applied to drug design for both practitioners and researchers. The framework can be found on GitHub and downloaded directly from the Python Package Index. Scientific Contribution MolScore is an open-source platform to facilitate generative molecular design and evaluation thereof for application in drug design. This platform takes important steps towards unifying existing benchmarks, providing a platform to share new benchmarks, and improves customisation, flexibility and usability for practitioners over existing solutions. Graphical Abstract

The human glucagon receptor, GCGR, belongs to the class B G-protein-coupled receptor family and plays a key role in glucose homeostasis and the pathophysiology of type 2 diabetes. Here we report the 3.0 Å crystal structure of full-length GCGR containing both the extracellular domain and transmembrane domain in an inactive conformation. The two domains are connected by a 12-residue segment termed the stalk, which adopts a β-strand conformation, instead of forming an α-helix as observed in the previously solved structure of the GCGR transmembrane domain. The first extracellular loop exhibits a β-hairpin conformation and interacts with the stalk to form a compact β-sheet structure. Hydrogen–deuterium exchange, disulfide crosslinking and molecular dynamics studies suggest that the stalk and the first extracellular loop have critical roles in modulating peptide ligand binding and receptor activation. These insights into the full-length GCGR structure deepen our understanding of the signalling mechanisms of class B G-protein-coupled receptors. The crystal structure of the full-length human glucagon receptor reveals the essential role of the 12-residue ‘stalk’ segment and an extracellular loop in the regulation of ligand binding and receptor activation. Full-length class B GPCR structures The glucagon-like peptide-1 receptor (GLP-1R) and the glucagon receptor (GCGR) belong to the class B G-protein-coupled receptor family and have opposing physiological roles in glucose homeostasis and insulin release. As such, they are important in regulating metabolism and appetite and offer significant treatment possibilities for type 2 diabetes. However, as yet, no full-length structures of these receptors have been solved. Three papers in this issue of Nature report the structure of GLP-1R. Ray Stevens and colleagues describe the crystal structure of the human GLP-1R transmembrane domain in an inactive state in complex with negative allosteric modulators. Fiona Marshall and colleagues describe the active-state full-length receptor in complex with truncated peptide agonists, which have potent activity in mice on oral administration. Georgios Skiniotis, Brian Kobilka and colleagues describe the cryo-electron microscopy structure of an unmodified GLP-1R in complex with its endogenous peptide ligand, GLP-1, and the heterotrimeric G protein. Finally, in a fourth paper in this week's issue of Nature , Beili Wu and colleagues report the crystal structure of the full-length GCGR in an inactive conformation. Taken together, these studies provide key insights into the activation and signalling mechanisms of class B receptors and provide therapeutic opportunities for targeting this receptor family.

Generative chemical language models (CLMs) have demonstrated success in learning language-based molecular representations for de novo drug design. Here, we integrate structure-based drug design (SBDD) principles with CLMs to go from protein structure to novel small-molecule ligands, without a priori knowledge of ligand chemistry. Using Augmented Hill-Climb, we successfully optimise multiple objectives within a practical timeframe, including protein-ligand complementarity. Resulting de novo molecules contain known or promising adenosine A 2A receptor ligand chemistry that is not available in commercial vendor libraries, accessing commercially novel areas of chemical space. Experimental validation demonstrates a binding hit rate of 88%, with 50% having confirmed functional activity, including three nanomolar ligands and two novel chemotypes. The two strongest binders are co-crystallised with the A 2A receptor, revealing their binding mechanisms that can be used to inform future iterations of structure-based de novo design, closing the AI SBDD loop. Here the authors combine a deep generative model with structure-based drug design and prospectively validate functionally active, nanomolar, A 2A adenosine receptor ligands and solve their crystal structures to close the Artificial Intelligence Structure-Based Drug Design loop.