When Cofactors Aren't X Factors: Functional Groups That Are Labile in Human Liver Microsomes in the Absence of NADPH.

The metabolic stability of compounds is often assessed at an early stage in drug discovery programs by profiling with hepatic microsomes. Exclusion of the reduced form of nicotinamide adenine dinucleotide phosphate (NADPH) in these assays provides insight into non-cytochrome P450 (CYP)-mediated metabolism. This report uses a matched molecular pair (MMP) application to assess which chemical substituents are commonly susceptible to non-NADPH-mediated metabolism by microsomes. The analysis found the overall prevalence of metabolism in the absence of NADPH to be low, with esters, amides, aldehydes, and oxetanes being among the most commonly susceptible functional groups. Given that non-CYP enzymes, such as esterases, may be expressed extrahepatically and lead to lower confidence in predicted pharmacokinetic profiles, an awareness of the functional groups that commonly undergo non-NADPH-mediated metabolism-as well as options for their replacement based on experimental MMP data-may help researchers derisk metabolic stability issues at an earlier stage in drug discovery.

Benchmarking Accuracy and Generalizability of Four Graph Neural Networks Using Large In Vitro ADME Datasets from Different Chemical Spaces.

In this work, we benchmark a variety of single- and multi-task graph neural network (GNN) models against lower-bar and higher-bar traditional machine learning approaches employing human engineered molecular features. We consider four GNN variants - Graph Convolutional Network (GCN), Graph Attention Network (GAT), Message Passing Neural Network (MPNN), and Attentive Fingerprint (AttentiveFP). So far deep learning models have been primarily benchmarked using lower-bar traditional models solely based on fingerprints, while more realistic benchmarks employing fingerprints, whole-molecule descriptors and predictions from other related endpoints (e. g., LogD7.4) appear to be scarce for industrial ADME datasets. In addition to time-split test sets based on Genentech data, this study benefits from the availability of measurements from an external chemical space (Roche data). We identify GAT as a promising approach to implementing deep learning models. While all the deep learning models significantly outperform lower-bar benchmark traditional models solely based on fingerprints, only GATs seem to offer a small but consistent improvement over higher-bar benchmark traditional models. Finally, the accuracy of in vitro assays from different laboratories predicting the same experimental endpoints appears to be comparable with the accuracy of GAT single-task models, suggesting that most of the observed error from the models is a function of the experimental error propagation.

Docking optimization, variance and promiscuity for large-scale drug-like chemical space using high performance computing architectures.

There is a continuing need to hasten and improve protein-ligand docking to facilitate the next generation of drug discovery. As the drug-like chemical space reaches into the billions of molecules, increasingly powerful computer systems are required to probe, as well as tackle, the software engineering challenges needed to adapt existing docking programs to use next-generation massively parallel processing systems. We demonstrate docking setup using the wrapper code approach to optimize the DOCK program for large-scale computation as well as docking analysis using variance and promiscuity as examples. Wrappers provide faster docking speeds when compared with the naive multi-threading system MPI-DOCK, making future endeavors in large-scale docking more feasible; in addition, eliminating highly variant or promiscuous compounds will make databases more useful.

ß2-Adrenoceptor agonists in the regulation of mitochondrial biogenesis.

The stimulation of mitochondrial biogenesis (MB) via cell surface G-protein coupled receptors is a promising strategy for cell repair and regeneration. Here we report the specificity and chemical rationale of a panel of ß2-adrenoceptor agonists with regards to MB. Using primary cultures of renal cells, a diverse panel of ß2-adrenoceptor agonists elicited three distinct phenotypes: full MB, partial MB, and non-MB. Full MB compounds had efficacy in the low nanomolar range and represent two chemical scaffolds containing three distinct chemical clusters. Interestingly, the MB phenotype did not correlate with reported receptor affinity or chemical similarity. Chemical clusters were then subjected to pharmacophore modeling creating two models with unique and distinct features, consisting of five conserved amongst full MB compounds were identified. The two discrete pharmacophore models were coalesced into a consensus pharmacophore with four unique features elucidating the spatial and chemical characteristics required to stimulate MB.

High-throughput respirometric assay identifies predictive toxicophore of mitochondrial injury.

Many environmental chemicals and drugs negatively affect human health through deleterious effects on mitochondrial function. Currently there is no chemical library of mitochondrial toxicants, and no reliable methods for predicting mitochondrial toxicity. We hypothesized that discrete toxicophores defined by distinct chemical entities can identify previously unidentified mitochondrial toxicants. We used a respirometric assay to screen 1760 compounds (5 µM) from the LOPAC and ChemBridge DIVERSet libraries. Thirty-one of the assayed compounds decreased uncoupled respiration, a stress test for mitochondrial dysfunction, prior to a decrease in cell viability and reduced the oxygen consumption rate in isolated mitochondria. The mitochondrial toxicants were grouped by chemical similarity and two clusters containing four compounds each were identified. Cheminformatic analysis of one of the clusters identified previously uncharacterized mitochondrial toxicants from the ChemBridge DIVERSet. This approach will enable the identification of mitochondrial toxicants and advance the prediction of mitochondrial toxicity for both drug discovery and risk assessment.

The ß2-adrenoceptor agonist formoterol stimulates mitochondrial biogenesis.

Mitochondrial dysfunction is a common mediator of disease and organ injury. Although recent studies show that inducing mitochondrial biogenesis (MB) stimulates cell repair and regeneration, only a limited number of chemicals are known to induce MB. To examine the impact of the ß-adrenoceptor (ß-AR) signaling pathway on MB, primary renal proximal tubule cells (RPTC) and adult feline cardiomyocytes were exposed for 24 h to multiple ß-AR agonists: isoproterenol (nonselective ß-AR agonist), (±)-(R*,R*)-[4-[2-[[2-(3-chlorophenyl)-2-hydroxyethyl]amino]propyl]phenoxy] acetic acid sodium hydrate (BRL 37344) (selective ß(3)-AR agonist), and formoterol (selective ß(2)-AR agonist). The Seahorse Biosciences (North Billerica, MA) extracellular flux analyzer was used to quantify carbonylcyanide p-trifluoromethoxyphenylhydrazone (FCCP)-uncoupled oxygen consumption rate (OCR), a marker of maximal electron transport chain activity. Isoproterenol and BRL 37244 did not alter mitochondrial respiration at any of the concentrations examined. Formoterol exposure resulted in increases in both FCCP-uncoupled OCR and mitochondrial DNA (mtDNA) copy number. The effect of formoterol on OCR in RPTC was inhibited by the ß-AR antagonist propranolol and the ß(2)-AR inverse agonist 3-(isopropylamino)-1-[(7-methyl-4-indanyl)oxy]butan-2-ol hydrochloride (ICI-118,551). Mice exposed to formoterol for 24 or 72 h exhibited increases in kidney and heart mtDNA copy number, peroxisome proliferator-activated receptor γ coactivator 1α, and multiple genes involved in the mitochondrial electron transport chain (F0 subunit 6 of transmembrane F-type ATP synthase, NADH dehydrogenase subunit 1, NADH dehydrogenase subunit 6, and NADH dehydrogenase [ubiquinone] 1ß subcomplex subunit 8). Cheminformatic modeling, virtual chemical library screening, and experimental validation identified nisoxetine from the Sigma Library of Pharmacologically Active Compounds and two compounds from the ChemBridge DIVERSet that increased mitochondrial respiratory capacity. These data provide compelling evidence for the use and development of ß(2)-AR ligands for therapeutic MB.

Time-resolved fluorescence resonance energy transfer and surface plasmon resonance-based assays for retinoid and transthyretin binding to retinol-binding protein 4.

Retinol-binding protein-4 (RBP4) is an emerging candidate drug target for type 2 diabetes and lipofuscin-mediated macular degeneration. The retinoic acid derivative fenretinide (N-(4-hydroxyphenyl) retinamide; HPR) exerts therapeutic effects in mouse models of obesity, diabetes, and Stargardt's disease by targeting RBP4. Fenretinide competes with retinoids for RBP4 binding, disrupts RBP4-transthyretin (TTR) complexes, and results in urinary secretion of RBP4 and systemic depletion of retinol. To enable the search for nonretinoid molecules with fenretinide-like activities we developed a HTS-compatible homogeneous TR-FRET assay monitoring the displacement of retinoic acid derivatives from RBP4 in high-density 384-well and 1536-well microtiter plate formats. The retinoid displacement assay proved to be highly sensitive and robust after miniaturization with IC(50)s for fenretinide and retinol ranging around 50 and 100 nM, respectively, and Z'-factors around 0.7. In addition, a surface plasmon resonance (SPR)-based secondary assay was developed to interrogate small molecule RBP4 binders for their ability to modulate the RBP4-TTR interaction. Finally, a 1.6 x 10(6) compound library was screened against the retinoid displacement assay. Several potent retinoid competitors were identified that also appeared to disrupt RBP4-TTR complexes. Some of these compounds could potentially serve as valuable tools to further probe RBP4 biology in the future.

In silico activity profiling reveals the mechanism of action of antimalarials discovered in a high-throughput screen.

The growing resistance to current first-line antimalarial drugs represents a major health challenge. To facilitate the discovery of new antimalarials, we have implemented an efficient and robust high-throughput cell-based screen (1,536-well format) based on proliferation of Plasmodium falciparum (Pf) in erythrocytes. From a screen of approximately 1.7 million compounds, we identified a diverse collection of approximately 6,000 small molecules comprised of >530 distinct scaffolds, all of which show potent antimalarial activity (<1.25 microM). Most known antimalarials were identified in this screen, thus validating our approach. In addition, we identified many novel chemical scaffolds, which likely act through both known and novel pathways. We further show that in some cases the mechanism of action of these antimalarials can be determined by in silico compound activity profiling. This method uses large datasets from unrelated cellular and biochemical screens and the guilt-by-association principle to predict which cellular pathway and/or protein target is being inhibited by select compounds. In addition, the screening method has the potential to provide the malaria community with many new starting points for the development of biological probes and drugs with novel antiparasitic activities.