G Protein-Coupled Receptor-Ligand Pose and Functional Class Prediction.

G protein-coupled receptor (GPCR) transmembrane protein family members play essential roles in physiology. Numerous pharmaceuticals target GPCRs, and many drug discovery programs utilize virtual screening (VS) against GPCR targets. Improvements in the accuracy of predicting new molecules that bind to and either activate or inhibit GPCR function would accelerate such drug discovery programs. This work addresses two significant research questions. First, do ligand interaction fingerprints provide a substantial advantage over automated methods of binding site selection for classical docking? Second, can the functional status of prospective screening candidates be predicted from ligand interaction fingerprints using a random forest classifier? Ligand interaction fingerprints were found to offer modest advantages in sampling accurate poses, but no substantial advantage in the final set of top-ranked poses after scoring, and, thus, were not used in the generation of the ligand-receptor complexes used to train and test the random forest classifier. A binary classifier which treated agonists, antagonists, and inverse agonists as active and all other ligands as inactive proved highly effective in ligand function prediction in an external test set of GPR31 and TAAR2 candidate ligands with a hit rate of 82.6% actual actives within the set of predicted actives.

Structure-based pharmacophore modeling 2. Developing a novel framework for structure-based pharmacophore model generation and selection.

Pharmacophore models are three-dimensional arrangements of molecular features required for biological activity that are used in ligand identification efforts for many biological targets, including G protein-coupled receptors (GPCR). Though GPCR are integral membrane proteins of considerable interest as targets for drug development, many of these receptors lack known ligands or experimentally determined structures necessary for ligand- or structure-based pharmacophore model generation, respectively. Thus, we here present a structure-based pharmacophore modeling approach that uses fragments placed with Multiple Copy Simultaneous Search (MCSS) to generate high-performing pharmacophore models in the context of experimentally determined, as well as modeled GPCR structures. Moreover, we have addressed the oft-neglected topic of pharmacophore model selection via development of a cluster-then-predict machine learning workflow. Herein score-based pharmacophore models were generated in experimentally determined and modeled structures of 13 class A GPCR and resulted in pharmacophore models exhibiting high enrichment factors when used to search a database containing 569 class A GPCR ligands. In addition, classification of pharmacophore models with the best performing cluster-then-predict logistic regression classifier resulted in positive predictive values (PPV) of 0.88 and 0.76 for selecting high enrichment pharmacophore models from among those generated in experimentally determined and modeled structures, respectively.

Structure-based pharmacophore modeling 1. Automated random pharmacophore model generation.

Pharmacophores are three-dimensional arrangements of molecular features required for biological activity that are often used in virtual screening efforts to prioritize ligands for experimental testing. G protein-coupled receptors (GPCR) are integral membrane proteins of considerable interest as targets for ligand discovery and drug development. Ligand-based pharmacophore models can be constructed to identify structural commonalities between known bioactive ligands for targets including GPCR. However, structure-based pharmacophores (which only require an experimentally determined or modeled structure for a protein target) have gained more attention to aid in virtual screening efforts as the number of publicly available experimentally determined GPCR structures have increased (140 unique GPCR represented as of October 24, 2022). Thus, the goal of this study was to develop a method of structure-based pharmacophore model generation applicable to ligand discovery for GPCR that have few known ligands. Pharmacophore models were generated within the active sites of 8 class A GPCR crystal structures via automated annotation of 5 randomly selected functional group fragments to sample diverse combinations of pharmacophore features. Each of the 5000 generated pharmacophores was then used to search a database containing active and decoy/inactive compounds for 30 class A GPCR and scored using enrichment factor and goodness-of-hit metrics to assess performance. Application of this method to the set of 8 class A GPCR produced pharmacophore models possessing the theoretical maximum enrichment factor value in both resolved structures (8 of 8 cases) and homology models (7 of 8 cases), indicating that generated pharmacophore models can prove useful in the context of virtual screening.

Application of computational methods for class A GPCR Ligand discovery.

G protein-coupled receptors (GPCR) are integral membrane proteins of considerable interest as targets for drug development due to their role in transmitting cellular signals in a multitude of biological processes. Of the six classes categorizing GPCR (A, B, C, D, E, and F), class A contains the largest number of therapeutically relevant GPCR. Despite their importance as drug targets, many challenges exist for the discovery of novel class A GPCR ligands serving as drug precursors. Though knowledge of the structural and functional characteristics of GPCR has grown significantly over the past 20 years, a large portion of GPCR lack reported, experimentally determined structures. Furthermore, many GPCR have no known endogenous and/or synthetic ligands, limiting further exploration of their biochemical, cellular, and physiological roles. While many successes in GPCR ligand discovery have resulted from experimental high-throughput screening, computational methods have played an increasingly important role in GPCR ligand identification in the past decade. Here we discuss computational techniques applied to GPCR ligand discovery. This review summarizes class A GPCR structure/function and provides an overview of many obstacles currently faced in GPCR ligand discovery. Furthermore, we discuss applications and recent successes of computational techniques used to predict GPCR structure as well as present a summary of ligand- and structure-based methods used to identify potential GPCR ligands. Finally, we discuss computational hit list generation and refinement and provide comprehensive workflows for GPCR ligand identification.

Discovery of agonist-antagonist pairs for the modulation of Ca [2]+ and voltage-gated K+ channels of large conductance that contain beta1 subunits.

Large conductance, calcium/voltage-gated potassium channels (BK) regulate critical body processes, including neuronal, secretory and smooth muscle (SM) function. While BK-forming alpha subunits are ubiquitous, accessory beta1 subunits are highly expressed in SM. This makes beta1 an attractive target for pharmaceutical development to treat SM disorders, such as hypertension or cerebrovascular spasm. Compounds activating BK via beta1 have been identified, yet they exhibit low potency and off-target effects while antagonists that limit agonist activity via beta 1 remain unexplored. Beta1-dependent BK ligand-based pharmacophore modeling and ZINC database searches identified 15 commercially available hits. Concentration-response curves on BK alpha + beta1 subunit-mediated currents were obtained in CHO cells. One potent (EC50 = 20 nM) and highly efficacious activator (maximal activation = ×10.3 of control) was identified along with a potent antagonist (KB = 3.02 nM), both of which were dependent on beta1. Our study provides the first proof-of-principle that an agonist/antagonist pair can be used to control beta1-containing BK activity.

Self-docking and cross-docking simulations of G protein-coupled receptor-ligand complexes: Impact of ligand type and receptor activation state.

G protein-coupled receptors (GPCR) are the largest family of cell surface receptors in vertebrates. Their abundance and role in nearly all physiological systems make GPCR the largest protein family targeted for development of pharmaceuticals. Ligand discovery aimed at identification of chemical tools and drug leads is aided by molecular docking simulations that allow critical analysis of the potential interactions between small molecules and proteins in resulting complexes. However, blind assessments of ligand pose quality and affinity prediction have thus far not provided broadly generalizable performance expectations for docking into experimentally-characterized GPCR targets. Likewise, the relative importance of receptor activation state and ligand function differences have also not been systematically assessed. This study compares performance when docking ligands of varied function into varied GPCR activation states in the absence of extensive resampling of the input GPCR structure, and only limited sidechain flexibility after ligand placement. Simulations were performed using 37 experimental structures of 11 Class A GPCR crystallized in multiple activation states (giving rise to 37 self-docking and 68 cross docking simulations). Our results show that one specific subset of cross-docking simulations gave results of similar quality to self-docking. Median ligand RMSD values for top-scored poses were 1.2 Å and 2.0 Å for self-docking and StateMatch/FunctionMatch cross-docking, respectively. The distributions of ligand RMSD values were not statistically different for these two conditions, according to a Kolmogorov-Smirnov test. Therefore, docking performance against GPCR targets can be estimated in advance based on docking target structure activation states, with higher accuracy expected when docking agonists into active state structures and inverse agonists or antagonists into inactive state structures. Receptor conformational sampling in advance of docking or receptor conformational adjustment after docking are more likely to produce substantial improvements for other pairings of receptor activation state and ligand function.

Molecular modelling guided design, synthesis and QSAR analysis of new small molecule non-lipid autotaxin inhibitors.

The lysophospholipase D autotaxin (ATX) generates lysophosphatidic acid (LPA) that activates six cognate G-protein coupled receptors (GPCR) in cancerous cells, promoting their motility and invasion. Four novel compounds were generated aided by molecular docking guided design and synthesis techniques to obtain new dual inhibitors of ATX and the lysophosphatidic acid receptor subtype 1 (LPAR1). Biological evaluation of these compounds revealed two compounds, 10 and 11, as new ATX enzyme inhibitors with potencies in the range of 218-220 nM and water solubility (>100 µg/mL), but with no LPAR1 inhibitory activity. A QSAR model was generated that included four newly designed compounds and twenty-one additional compounds that we have reported previously. The QSAR model provided excellent predictability of the pharmacological activity and potency among structurally related drug candidates. This model will be highly useful in guiding the synthesis of new ATX inhibitors in the future.

Benchmarking GPCR homology model template selection in combination with de novo loop generation.

G protein-coupled receptors (GPCR) comprise the largest family of membrane proteins and are of considerable interest as targets for drug development. However, many GPCR structures remain unsolved. To address the structural ambiguity of these receptors, computational tools such as homology modeling and loop modeling are often employed to generate predictive receptor structures. Here we combined both methods to benchmark a protocol incorporating homology modeling based on a locally selected template and extracellular loop modeling that additionally evaluates the presence of template ligands during these modeling steps. Ligands were also docked using three docking methods and two pose selection methods to elucidate an optimal ligand pose selection method. Results suggest that local template-based homology models followed by loop modeling produce more accurate and predictive receptor models than models produced without loop modeling, with decreases in average receptor and ligand RMSD of 0.54 Å and 2.91 Å, respectively. Ligand docking results showcased the ability of MOE induced fit docking to produce ligand poses with atom root-mean-square deviation (RMSD) values at least 0.20 Å lower (on average) than the other two methods benchmarked in this study. In addition, pose selection methods (software-based scoring, ligand complementation) selected lower RMSD poses with MOE induced fit docking than either of the other methods (averaging at least 1.57 Å lower), indicating that MOE induced fit docking is most suited for docking into GPCR homology models in our hands. In addition, target receptor models produced with a template ligand present throughout the modeling process most often produced target ligand poses with RMSD values ≤ 4.5 Å and Tanimoto coefficients > 0.6 after selection based on ligand complementation than target receptor models produced in the absence of template ligands. Overall, the findings produced by this study support the use of local template homology modeling in combination with de novo ECL2 modeling in the presence of a ligand from the template crystal structure to generate GPCR models intended to study ligand binding interactions.

Optical Control of Lysophosphatidic Acid Signaling.

Lysophosphatidic acid (LPA) is a phospholipid that acts as an extracellular signaling molecule and activates the family of lysophosphatidic acid receptors (LPA1-6). These G protein-coupled receptors (GPCRs) are broadly expressed and are particularly important in development as well as in the nervous, cardiovascular, reproductive, gastrointestinal, and pulmonary systems. Here, we report on a photoswitchable analogue of LPA, termed AzoLPA, which contains an azobenzene photoswitch embedded in the acyl chain. AzoLPA enables optical control of LPA receptor activation, shown through its ability to rapidly control LPA-evoked increases in intracellular Ca2+ levels. AzoLPA shows greater activation of LPA receptors in its light-induced cis-form than its dark-adapted (or 460 nm light-induced) trans-form. AzoLPA enabled the optical control of neurite retraction through its activation of the LPA2 receptor.

Optical control of sphingosine-1-phosphate formation and function.

Sphingosine-1-phosphate (S1P) plays important roles as a signaling lipid in a variety of physiological and pathophysiological processes. S1P signals via a family of G-protein-coupled receptors (GPCRs) (S1P1-5) and intracellular targets. Here, we report on photoswitchable analogs of S1P and its precursor sphingosine, respectively termed PhotoS1P and PhotoSph. PhotoS1P enables optical control of S1P1-3, shown through electrophysiology and Ca2+ mobilization assays. We evaluated PhotoS1P in vivo, where it reversibly controlled S1P3-dependent pain hypersensitivity in mice. The hypersensitivity induced by PhotoS1P is comparable to that induced by S1P. PhotoS1P is uniquely suited for the study of S1P biology in cultured cells and in vivo because it exhibits prolonged metabolic stability compared to the rapidly metabolized S1P. Using lipid mass spectrometry analysis, we constructed a metabolic map of PhotoS1P and PhotoSph. The formation of these photoswitchable lipids was found to be light dependent, providing a novel approach to optically probe sphingolipid biology.

A benchmark study of loop modeling methods applied to G protein-coupled receptors.

G protein-coupled receptors (GPCR) are important drug discovery targets. Despite progress, many GPCR structures have not yet been solved. For these targets, comparative modeling is used in virtual ligand screening to prioritize experimental efforts. However, the structure of extracellular loop 2 (ECL2) is often poorly predicted. This is significant due to involvement of ECL2 in ligand binding for many Class A GPCR. Here we examine the performance of loop modeling protocols available in the Rosetta (cyclic coordinate descent [CCD], KIC with fragments [KICF] and next generation KIC [NGK]) and Molecular Operating Environment (MOE) software suites (de novo search). ECL2 from GPCR crystal structures served as the structure prediction targets and were divided into four sets depending on loop length. Results suggest that KICF and NGK sampled and scored more loop models with sub-angstrom and near-atomic accuracy than CCD or de novo search for loops of 24 or fewer residues. None of the methods were able to sample loop conformations with near-atomic accuracy for the longest targets ranging from 25 to 32 residues based on 1000 models generated. For these long loop targets, increased conformational sampling is necessary. The strongly conserved disulfide bond between Cys3.25 and Cys45.50 in ECL2 proved an effective filter. Setting an upper limit of 5.1 Å on the S-S distance improved the lowest RMSD model included in the top 10 scored structures in Groups 1-4 on average between 0.33 and 1.27 Å. Disulfide bond formation and geometry optimization of ECL2 provided an additional incremental benefit in structure quality.

GPCR homology model template selection benchmarking: Global versus local similarity measures.

G protein-coupled receptors (GPCR) are integral membrane proteins of considerable interest as targets for drug development. GPCR ligand interaction studies often have a starting point with either crystal structures or comparative models. The majority of GPCR do not have experimentally-characterized 3-dimensional structures, so comparative modeling, also called homology modeling, is a good structure-based starting point. Comparative modeling is a widely used method for generating models of proteins with unknown structures by analogy to crystallized proteins that are expected to exhibit structural conservation. Traditionally, comparative modeling template selection is based on global sequence identity and shared function. However high sequence identity localized to the ligand binding pocket may produce better models to examine protein-ligand interactions. This in silico benchmark study examined the performance of a global versus local similarity measure applied to comparative modeling template selection for 6 previously crystallized, class A GCPR (CXCR4, FFAR1, NOP, P2Y12, OPRK, and M1) with the long-term goal of optimizing GPCR ligand identification efforts. Comparative models were generated from templates selected using both global and local similarity measures. Similarity to reference crystal structures was reflected in RMSD values between atom positions throughout the structure or localized to the ligand binding pocket. Overall, models deviated from the reference crystal structure to a similar degree regardless of whether the template was selected using a global or local similarity measure. Ligand docking simulations were performed to assess relative performance in predicting protein-ligand complex structures and interaction networks. Calculated RMSD values between ligand poses from docking simulations and crystal structures indicate that models based on locally selected templates give docked poses that better mimic crystallographic ligand positions than those based on globally-selected templates in five of the six benchmark cases. However, protein model refinement strategies in advance of ligand docking applications are clearly essential as the average RMSD between crystallographic poses and poses docked into local template models was 9.7 Šand typically less than half of the ligand interaction sites are shared between the docked and crystallographic poses. These data support the utilization of local similarity measures to guide template selection in protocols using comparative models to investigate ligand-receptor interactions.

Highly Potent Non-Carboxylic Acid Autotaxin Inhibitors Reduce Melanoma Metastasis and Chemotherapeutic Resistance of Breast Cancer Stem Cells.

Autotaxin (ATX, aka. ENPP2) is the main source of the lipid mediator lysophosphatidic acid (LPA) in biological fluids. This study reports on inhibitors of ATX derived by lead optimization of the benzene-sulfonamide in silico hit compound 3. The new analogues provide a comprehensive structure-activity relationship of the benzene-sulfonamide scaffold that yielded a series of highly potent ATX inhibitors. The three most potent analogues (3a, IC50 ∼ 32 nM; 3b, IC50 ∼ 9 nM; and 14, IC50 ∼ 35 nM) inhibit ATX-dependent invasion of A2058 human melanoma cells in vitro. Two of the most potent compounds, 3b and 3f (IC50 ∼ 84 nM), lack inhibitory action on ENPP6 and ENPP7 but possess weak antagonist action specific to the LPA1 G protein-coupled receptor. In particular, compound 3b potently reduced in vitro chemotherapeutic resistance of 4T1 breast cancer stem-like cells to paclitaxel and significantly reduced B16 melanoma metastasis in vivo.

G protein-coupled receptors: the evolution of structural insight.

G protein-coupled receptors (GPCR) comprise a diverse superfamily of over 800 proteins that have gained relevance as biological targets for pharmaceutical drug design. Although these receptors have been investigated for decades, three-dimensional structures of GPCR have only recently become available. In this review, we focus on the technological advancements that have facilitated efforts to gain insights into GPCR structure. Progress in these efforts began with the initial crystal structure determination of rhodopsin (PDB: 1F88) in 2000 and has continued to the most recently published structure of the A1AR (PDB: 5UEN) in 2017. Numerous experimental developments over the past two decades have opened the door for widespread GPCR structural characterization. These efforts have resulted in the determination of three-dimensional structures for over 40 individual GPCR family members. Herein we present a comprehensive list and comparative analysis of over 180 individual GPCR structures. This includes a summary of different GPCR functional states crystallized with agonists, dual agonists, partial agonists, inverse agonists, antagonists, and allosteric modulators.

Discovery and synthetic optimization of a novel scaffold for hydrophobic tunnel-targeted autotaxin inhibition.

Autotaxin (ATX) is a ubiquitous ectoenzyme that hydrolyzes lysophosphatidylcholine (LPC) to form the bioactive lipid mediator lysophosphatidic acid (LPA). LPA activates specific G-protein coupled receptors to elicit downstream effects leading to cellular motility, survival, and invasion. Through these pathways, upregulation of ATX is linked to diseases such as cancer and cardiovascular disease. Recent crystal structures confirm that the catalytic domain of ATX contains multiple binding regions including a polar active site, hydrophobic tunnel, and a hydrophobic pocket. This finding is consistent with the promiscuous nature of ATX hydrolysis of multiple and diverse substrates and prior investigations of inhibitor impacts on ATX enzyme kinetics. The current study used virtual screening methods to guide experimental identification and characterization of inhibitors targeting the hydrophobic region of ATX. An initially discovered inhibitor, GRI392104 (IC50 4µM) was used as a lead for synthetic optimization. In total twelve newly synthesized inhibitors of ATX were more potent than GRI392104 and were selective for ATX as they had no effect on other LPC-specific NPP family members or on LPA1-5 GPCR.

Reprint of: "Synthetic lipids and their role in defining macromolecular assemblies".

Lipids have a variety of physiological roles, ranging from structural and biophysical contributions to membrane functions to signaling contributions in normal and abnormal physiology. This review highlights some of the contributions made by Robert Bittman to our understanding of lipid assemblies through the production of synthetic lipid analogs in the sterol, sphingolipid, and glycolipid classes. His contributions have included the development of a fluorescent cholesterol analog that shows strong functional analogies to cholesterol that has allowed live imaging of cholesterol distribution in living systems, to stereospecific synthetic approaches to both sphingolipid and glycolipid analogs crucial in defining the structure-activity relationships of lipid biological targets.

Synthetic lipids and their role in defining macromolecular assemblies.

Lipids have a variety of physiological roles, ranging from structural and biophysical contributions to membrane functions to signaling contributions in normal and abnormal physiology. This review highlights some of the contributions made by Robert Bittman to our understanding of lipid assemblies through the production of synthetic lipid analogs in the sterol, sphingolipid, and glycolipid classes. His contributions have included the development of a fluorescent cholesterol analog that shows strong functional analogies to cholesterol that has allowed live imaging of cholesterol distribution in living systems, to stereospecific synthetic approaches to both sphingolipid and glycolipid analogs crucial in defining the structure-activity relationships of lipid biological targets.

Multi-generational pharmacophore modeling for ligands to the cholane steroid-recognition site in the ß1 modulatory subunit of the BKCa channel.

Large conductance, voltage- and Ca(2+)-gated K(+) (BKCa) channels play a critical role in smooth muscle contractility and thus represent an emerging therapeutic target for drug development to treat vascular disease, gastrointestinal, bladder and uterine disorders. Several compounds are known to target the ubiquitously expressed BKCa channel-forming α subunit. In contrast, just a few are known to target the BKCa modulatory ß1 subunit, which is highly expressed in smooth muscle and scarce in most other tissues. Lack of available high-resolution structural data makes structure-based pharmacophore modeling of ß1 subunit-dependent BKCa channel activators a major challenge. Following recent discoveries of novel BKCa channel activators that act via ß1 subunit recognition, we performed ligand-based pharmacophore modeling that led to the successful creation and fine-tuning of a pharmacophore over several generations. Initial models were developed using physiologically active cholane steroids (bile acids) as template. However, as more compounds that act on BKCa ß1 have been discovered, our model has been refined to improve accuracy. Database searching with our best-performing model has uncovered several novel compounds as candidate BKCa ß1 subunit ligands. Eight of the identified compounds were experimentally screened and two proved to be activators of recombinant BKCa ß1 complexes. One of these activators, sobetirome, differs substantially in structure from any previously reported activator.

Activation of calcium- and voltage-gated potassium channels of large conductance by leukotriene B4.

Calcium/voltage-gated, large conductance potassium (BK) channels control numerous physiological processes, including myogenic tone. BK channel regulation by direct interaction between lipid and channel protein sites has received increasing attention. Leukotrienes (LTA4, LTB4, LTC4, LTD4, and LTE4) are inflammatory lipid mediators. We performed patch clamp studies in Xenopus oocytes that co-expressed BK channel-forming (cbv1) and accessory ß1 subunits cloned from rat cerebral artery myocytes. Leukotrienes were applied at 0.1 nm-10 µm to either leaflet of cell-free membranes at a wide range of [Ca(2+)]i and voltages. Only LTB4 reversibly increased BK steady-state activity (EC50 = 1 nm; Emax reached at 10 nm), with physiological [Ca(2+)]i and voltages favoring this activation. Homomeric cbv1 or cbv1-ß2 channels were LTB4-resistant. Computational modeling predicted that LTB4 docked onto the cholane steroid-sensing site in the BK ß1 transmembrane domain 2 (TM2). Co-application of LTB4 and cholane steroid did not further increase LTB4-induced activation. LTB4 failed to activate ß1 subunit-containing channels when ß1 carried T169A, A176S, or K179I within the docking site. Co-application of LTB4 with LTA4, LTC4, LTD4, or LTE4 suppressed LTB4-induced activation. Inactive leukotrienes docked onto a portion of the site, probably preventing tight docking of LTB4. In summary, we document the ability of two endogenous lipids from different chemical families to share their site of action on a channel accessory subunit. Thus, cross-talk between leukotrienes and cholane steroids might converge on regulation of smooth muscle contractility via BK ß1. Moreover, the identification of LTB4 as a highly potent ligand for BK channels is critical for the future development of ß1-specific BK channel activators.

Design of anticancer lysophosphatidic acid agonists and antagonists.

Lysophosphatidic acid (LPA) and its receptors, LPA1-6, are integral parts of signaling pathways involved in cellular proliferation, migration and survival. These signaling pathways are of therapeutic interest for the treatment of multiple types of cancer and to reduce cancer metastasis and side effects. Validated therapeutic potential of key receptors, as well as recent structure-activity relationships yielding compounds with low nanomolar potencies are exciting recent advances in the field. Some compounds have proven efficacious in vivo against tumor proliferation and metastasis, bone cancer pain and the pulmonary fibrosis that can result as a side effect of pulmonary cancer radiation treatment. However, recent studies have identified that LPA contributes through multiple pathways to the development of chemotherapeutic resistance suggesting new applications for LPA antagonists in cancer treatment. This review summarizes the roles of LPA signaling in cancer pathophysiology and recent progress in the design and evaluation of LPA agonists and antagonists.