Prediction Accuracy of Production ADMET Models as a Function of Version: Activity Cliffs Rule.

As with many other institutions, our company maintains many quantitative structure-activity relationship (QSAR) models of absorption, distribution, metabolism, excretion, and toxicity (ADMET) end points and updates the models regularly. We recently examined version-to-version predictivity for these models over a period of 10 years. In this approach we monitor the goodness of prediction of new molecules relative to the training set of model version V before they are incorporated in the updated model V+1. Using a cell-based permeability assay (Papp) as an example, we illustrate how the QSAR models made from this data are generally predictive and can be utilized to enrich chemical designs and synthesis. Despite the obvious utility of these models, we turned up unexpected behavior in Papp and other ADMET activities for which the explanation is not obvious. One such behavior is that the apparent predictivity of the models as measured by root-mean-square-error can vary greatly from version to version and is sometimes very poor. One intuitively appealing explanation is that the observed activities of the new molecules fall outside the bulk of activities in the training set. Alternatively, one may think that the new molecules are exploring different regions of chemical space than the training set. However, the real explanation has to do with activity cliffs. If the observed activities of the new molecules are different than expected based on similar molecules in the training set, the predictions will be less accurate. This is true for all our ADMET end points.

Repression of Organic Anion Transporting Polypeptide (OATP) 1B Expression and Increase of Plasma Coproporphyrin Level as Evidence for OATP1B Downregulation in Cynomolgus Monkeys Treated with Chenodeoxycholic Acid.

Farnesoid X receptor (FXR) is a nuclear receptor known to markedly alter expression of major transporters and enzymes in the liver. However, its effects toward organic anion transporting polypeptides (OATP) 1B1 and 1B3 remain poorly characterized. Therefore, the present study was aimed at determining the effects of chenodeoxycholic acid (CDCA), a naturally occurring FXR agonist, on OATP1B expression in cynomolgus monkeys. Multiple administrations of 50 and 100 mg/kg of CDCA were first shown to significantly repress mRNA expression of SLCO1B1/3 approximately 60% to 80% in monkey livers. It also suppressed cytochrome P450 (CYP)7A1-mRNA and induced OSTα/ß-mRNA, which are well known targets of FXR and determinants of bile acid homeostasis. CDCA concomitantly decreased OATP1B protein abundance by approximately 60% in monkey liver. In contrast, multiple doses of 15 mg/kg rifampin (RIF), a pregnane X receptor agonist, had no effect on hepatic OATP1B protein, although it induced the intestinal P-glycoprotein and MR2 proteins by ∼2-fold. Moreover, multiple doses of CDCA resulted in a steady ∼2- to 10-fold increase of the OATP1B biomarkers coproporphyrins (CPs) in the plasma samples collected prior to each CDCA dose. Additionally, 3.4- to 11.2-fold increases of CPI and CPIII areas under the curve were observed after multiple administrations compared with the single dose and vehicle administration dosing groups. Taken together, these data suggest that CDCA represses the expression of OATP1B1 and OATP1B3 in monkeys. Further investigation of OATP1B downregulation by FXR in humans is warranted, as such downregulation effects may be involved in bile acid homeostasis and potential drug interactions in man. SIGNIFICANCE STATEMENT: Using gene expression and proteomics tools, as well as endogenous biomarker data, for the first time, we have demonstrated that OATP1B expression was suppressed and its activity was reduced in the cynomolgus monkeys following oral administration of 50 and 100 mg/kg/day of chenodeoxycholic acid (CDCA), a Farnesoid X receptor agonist, for 8 days. These results lead to a better understanding of OATP1B downregulation by CDCA and its role on bile acid and drug disposition.

Strategic Incorporation of Polarity in Heme-Displacing Inhibitors of Indoleamine-2,3-dioxygenase-1 (IDO1).

Indoleamine-2,3-dioxygenase-1 (IDO1) has emerged as a target of significant interest to the field of cancer immunotherapy, as the upregulation of IDO1 in certain cancers has been linked to host immune evasion and poor prognosis for patients. In particular, IDO1 inhibition is of interest as a combination therapy with immune checkpoint inhibition. Through an Automated Ligand Identification System (ALIS) screen, a diamide class of compounds was identified as a promising lead for the inhibition of IDO1. While hit 1 possessed attractive cell-based potency, it suffered from a significant right-shift in a whole blood assay, poor solubility, and poor pharmacokinetic properties. Through a physicochemical property-based approach, including a focus on lowering AlogP98 via the strategic introduction of polar substitution, compound 13 was identified bearing a pyridyl oxetane core. Compound 13 demonstrated improved whole blood potency and solubility, and an improved pharmacokinetic profile resulting in a low predicted human dose.

Improvement in ADMET Prediction with Multitask Deep Featurization.

The absorption, distribution, metabolism, elimination, and toxicity (ADMET) properties of drug candidates are important for their efficacy and safety as therapeutics. Predicting ADMET properties has therefore been of great interest to the computational chemistry and medicinal chemistry communities in recent decades. Traditional cheminformatics approaches, using learners such as random forests and deep neural networks, leverage fingerprint feature representations of molecules. Here, we learn the features most relevant to each chemical task at hand by representing each molecule explicitly as a graph. By applying graph convolutions to this explicit molecular representation, we achieve, to our knowledge, unprecedented accuracy in prediction of ADMET properties. By challenging our methodology with rigorous cross-validation procedures and prognostic analyses, we show that deep featurization better enables molecular predictors to not only interpolate but also extrapolate to new regions of chemical space.

Experimental Error, Kurtosis, Activity Cliffs, and Methodology: What Limits the Predictivity of Quantitative Structure-Activity Relationship Models?

Given a particular descriptor/method combination, some quantitative structure-activity relationship (QSAR) datasets are very predictive by random-split cross-validation while others are not. Recent literature in modelability suggests that the limiting issue for predictivity is in the data, not the QSAR methodology, and the limits are due to activity cliffs. Here, we investigate, on in-house data, the relative usefulness of experimental error, distribution of the activities, and activity cliff metrics in determining how predictive a dataset is likely to be. We include unmodified in-house datasets, datasets that should be perfectly predictive based only on the chemical structure, datasets where the distribution of activities is manipulated, and datasets that include a known amount of added noise. We find that activity cliff metrics determine predictivity better than the other metrics we investigated, whatever the type of dataset, consistent with the modelability literature. However, such metrics cannot distinguish real activity cliffs due to large uncertainties in the activities. We also show that a number of modern QSAR methods, and some alternative descriptors, are equally bad at predicting the activities of compounds on activity cliffs, consistent with the assumptions behind "modelability." Finally, we relate time-split predictivity with random-split predictivity and show that different coverages of chemical space are at least as important as uncertainty in activity and/or activity cliffs in limiting predictivity.

Informing the Selection of Screening Hit Series with in Silico Absorption, Distribution, Metabolism, Excretion, and Toxicity Profiles.

High-throughput screening (HTS) has enabled millions of compounds to be assessed for biological activity, but challenges remain in the prioritization of hit series. While biological, absorption, distribution, metabolism, excretion, and toxicity (ADMET), purity, and structural data are routinely used to select chemical matter for further follow-up, the scarcity of historical ADMET data for screening hits limits our understanding of early hit compounds. Herein, we describe a process that utilizes a battery of in-house quantitative structure-activity relationship (QSAR) models to generate in silico ADMET profiles for hit series to enable more complete characterizations of HTS chemical matter. These profiles allow teams to quickly assess hit series for desirable ADMET properties or suspected liabilities that may require significant optimization. Accordingly, these in silico data can direct ADMET experimentation and profoundly impact the progression of hit series. Several prospective examples are presented to substantiate the value of this approach.

The design and synthesis of novel spirocyclic heterocyclic sulfone ROMK inhibitors as diuretics.

A spirocyclic class of ROMK inhibitors was developed containing a structurally diverse heterocyclic sulfone moiety and spirocyclic core starting from lead 1. These compounds not only displayed exquisite ROMK potency but significantly improved selectivity over hERG. The lead compounds were found to have favorable pharmacokinetic properties and displayed robust diuretic, natriuretic and blood pressure lowering effects in spontaneously hypertensive rats.

Preclinical Characterization of 18F-MK-6240, a Promising PET Tracer for In Vivo Quantification of Human Neurofibrillary Tangles.

A PET tracer is desired to help guide the discovery and development of disease-modifying therapeutics for neurodegenerative diseases characterized by neurofibrillary tangles (NFTs), the predominant tau pathology in Alzheimer disease (AD). We describe the preclinical characterization of the NFT PET tracer 18F-MK-6240. METHODS: In vitro binding studies were conducted with 3H-MK-6240 in tissue slices and homogenates from cognitively normal and AD human brain donors to evaluate tracer affinity and selectivity for NFTs. Immunohistochemistry for phosphorylated tau was performed on human brain slices for comparison with 3H-MK-6240 binding patterns on adjacent brain slices. PET studies were performed with 18F-MK-6240 in monkeys to evaluate tracer kinetics and distribution in the brain. 18F-MK-6240 monkey PET studies were conducted after dosing with unlabeled MK-6240 to evaluate tracer binding selectivity in vivo. RESULTS: The 3H-MK-6240 binding pattern was consistent with the distribution of phosphorylated tau in human AD brain slices. 3H-MK-6240 bound with high affinity to human AD brain cortex homogenates containing abundant NFTs but bound poorly to amyloid plaque-rich, NFT-poor AD brain homogenates. 3H-MK-6240 showed no displaceable binding in the subcortical regions of human AD brain slices and in the hippocampus/entorhinal cortex of non-AD human brain homogenates. In monkey PET studies, 18F-MK-6240 displayed rapid and homogeneous distribution in the brain. The 18F-MK-6240 volume of distribution stabilized rapidly, indicating favorable tracer kinetics. No displaceable binding was observed in self-block studies in rhesus monkeys, which do not natively express NFTs. Moderate defluorination was observed as skull uptake. CONCLUSION: 18F-MK-6240 is a promising PET tracer for the in vivo quantification of NFTs in AD patients.

Discovery of 6-(Fluoro-(18)F)-3-(1H-pyrrolo[2,3-c]pyridin-1-yl)isoquinolin-5-amine ([(18)F]-MK-6240): A Positron Emission Tomography (PET) Imaging Agent for Quantification of Neurofibrillary Tangles (NFTs).

Neurofibrillary tangles (NFTs) made up of aggregated tau protein have been identified as the pathologic hallmark of several neurodegenerative diseases including Alzheimer's disease. In vivo detection of NFTs using PET imaging represents a unique opportunity to develop a pharmacodynamic tool to accelerate the discovery of new disease modifying therapeutics targeting tau pathology. Herein, we present the discovery of 6-(fluoro-(18)F)-3-(1H-pyrrolo[2,3-c]pyridin-1-yl)isoquinolin-5-amine, 6 ([(18)F]-MK-6240), as a novel PET tracer for detecting NFTs. 6 exhibits high specificity and selectivity for binding to NFTs, with suitable physicochemical properties and in vivo pharmacokinetics.

Efficiency gains in tracer identification for nuclear imaging: can in vivo LC-MS/MS evaluation of small molecules screen for successful PET tracers?

Positron emission tomography (PET) imaging has become a useful noninvasive technique to explore molecular biology within living systems; however, the utility of this method is limited by the availability of suitable radiotracers to probe specific targets and disease biology. Methods to identify potential areas of improvement in the ability to predict small molecule performance as tracers prior to radiolabeling would speed the discovery of novel tracers. In this retrospective analysis, we characterized the brain penetration or peak SUV (standardized uptake value), binding potential (BP), and brain exposure kinetics across a series of known, nonradiolabeled PET ligands using in vivo LC-MS/MS (liquid chromatography coupled to mass spectrometry) and correlated these parameters with the reported PET ligand performance in nonhuman primates and humans available in the literature. The PET tracers studied included those reported to label G protein-coupled receptors (GPCRs), intracellular enzymes, and transporters. Additionally, data for each tracer was obtained from a mouse brain uptake assay (MBUA), previously published, where blood-brain barrier (BBB) penetration and clearance parameters were assessed and compared against similar data collected on a broad compound set of central nervous system (CNS) therapeutic compounds. The BP and SUV identified via nonradiolabeled LC-MS/MS, while different from the published values observed in the literature PET tracer data, allowed for an identification of initial criteria values we sought to facilitate increased potential for success from our early discovery screening paradigm. Our analysis showed that successful, as well as novel, clinical PET tracers exhibited BP of greater than 1.5 and peak SUVs greater than approximately 150% at 5 min post dose in rodents. The brain kinetics appeared similar between both techniques despite differences in tracer dose, suggesting linearity across these dose ranges. The assessment of tracers in a CNS exposure model, the mouse brain uptake assessment (MBUA), showed that those compound with initial brain-to-plasma ratios >2 and unbound fraction in brain homogenate >0.01 were more likely to be clinically successful PET ligands. Taken together, early incorporation of a LC/MS/MS cold tracer discovery assay and a parallel MBUA can be an useful screening paradigm to prioritize and rank order potential novel PET radioligands during early tracer discovery efforts. Compounds considered for continued in vivo PET assessments can be identified quickly by leveraging in vitro affinity and selectivity measures, coupled with data from a MBUA, primarily the 5 min brain-to-plasma ratio and unbound fraction data. Coupled utilization of these data creates a strategy to efficiently screen for the identification of appropriate chemical space to invest in for radiotracer discovery.

Discovery of a novel series of orally active nociceptin/orphanin FQ (NOP) receptor antagonists based on a dihydrospiro(piperidine-4,7'-thieno[2,3-c]pyran) scaffold.

Nociceptin/OFQ (N/OFQ) is a 17 amino acid peptide that is the endogenous ligand for the ORL1/NOP receptor. Nociceptin appears to regulate a host of physiological functions such as biological reactions to stress, anxiety, mood, and drug abuse, in addition to feeding behaviors. To develop tools to study the function of nociceptin and NOP receptor, our research effort sought to identify orally available NOP antagonists. Our effort led to the discovery of a novel chemical series based on the dihydrospiro(piperidine-4,7'-thieno[2,3-c]pyran) scaffold. Herein we show that dihydrospiro(piperidine-4,7'-thieno[2,3-c]pyran)-derived compounds are potent NOP antagonists with high selectivity versus classical opioid receptors (µ, δ, and κ). Moreover, these compounds exhibit sufficient bioavailability to produce a high level of NOP receptor occupancy in the brain following oral administration in rats.

Evaluation in vitro and in animals of a new 11C-labeled PET radioligand for metabotropic glutamate receptors 1 in brain.

PURPOSE: Two allosteric modulators of the group I metabotropic glutamate receptors (mGluR1 and mGluR5) were evaluated as positron emission tomography (PET) radioligands for mGluR1. METHODS: LY2428703, a full mGluR1 antagonist (IC(50) 8.9 nM) and partial mGluR5 antagonist (IC(50) 118 nM), and LSN2606428, a full mGluR1 and mGluR5 antagonist (IC(50) 35.3 nM and 10.2 nM, respectively) were successfully labeled with (11)C and evaluated as radioligands for mGluR1. The pharmacology of LY2428703 was comprehensively assessed in vitro and in vivo, and its biodistribution was investigated by liquid chromatography-mass spectrometry/mass spectrometry, and by PET imaging in the rat. In contrast, LSN2606428 was only evaluated in vitro; further evaluation was stopped due to its unfavorable pharmacological properties and binding affinity. RESULTS: (11)C-LY2428703 showed promising characteristics, including: (1) high potency for binding to human mGluR1 (IC(50) 8.9 nM) with no significant affinity for other human mGlu receptors (mGluR2 through mGluR8); (2) binding to brain displaceable by administration of an mGluR1 antagonist; (3) only one major radiometabolite in both plasma and brain, with a negligible brain concentration (with 3.5 % of the total radioactivity in cerebellum) and no receptor affinity; (4) a large specific and displaceable signal in the mGluR1-rich cerebellum with no significant in vivo affinity for mGluR5, as shown by PET studies in rats; and (5) lack of substrate behavior for efflux transporters at the blood-brain barrier, as shown by PET studies conducted in wild-type and knockout mice. CONCLUSION: (11)C-LY2428703, a new PET radioligand for mGluR1 quantification, displayed promising characteristics both in vitro and in vivo in rodents.

Discovery of a potent, dual serotonin and norepinephrine reuptake inhibitor.

The objective of the described research effort was to identify a novel serotonin and norepinephrine reuptake inhibitor (SNRI) with improved norepinephrine transporter activity and acceptable metabolic stability and exhibiting minimal drug-drug interaction. We describe herein the discovery of a series of 3-substituted pyrrolidines, exemplified by compound 1. Compound 1 is a selective SNRI in vitro and in vivo, has favorable ADME properties, and retains inhibitory activity in the formalin model of pain behavior. Compound 1 thus represents a potential new probe to explore utility of SNRIs in central nervous system disorders, including chronic pain conditions.

Development of LC-MS/MS-based receptor occupancy tracers and positron emission tomography radioligands for the nociceptin/orphanin FQ (NOP) receptor.

Currently, a lack of sufficient tools has limited the understanding of the relationship between neuropsychiatric disorders and the nociceptin/orphanin FQ (N/OFQ) peptide (NOP) receptor. Herein, we describe the discovery and development of an antagonist NOP receptor occupancy (RO) tracer and a novel positron emission tomography (PET) radioligand suitable to probe the NOP receptor in human clinical studies. A thorough structure-activity relationship (SAR) around the high-affinity 3-(2'-fluoro-4',5'-dihydrospiro[piperidine-4,7'-thieno[2,3-c]pyran]-1-yl)-2-(2-halobenzyl)-N-alkylpropanamide scaffold identified a series of subnanomolar, highly selective NOP antagonists. Subsequently, these unlabeled NOP ligands were evaluated in vivo by liquid chromatography-tandem mass spectrometry (LC-MS/MS) in rat to determine brain uptake, kinetics and specific binding. (S)-27 was identified as a suitable unlabeled preclinical RO tracer to accurately quantify NOP receptor engagement in rat brain. Three compounds were selected for evaluation in nonhuman primates as PET tracers: (-)-26, (-)-30, and (-)-33. Carbon-11 labeling of (+)-31 yielded [(11)C]-(S)-30, which exhibited minimal generation of central nervous system (CNS) penetrant radiometabolites, improved brain uptake, and was an excellent PET radioligand in both rat and monkey. Currently [(11)C]-(S)-30 is being evaluated as a PET radiotracer for the NOP receptor in human subjects.

Synthesis of diketopiperazine-based carboline homodimers and in vitro growth inhibition of human carcinomas.

Starting from d- or l-tryptophan, we have synthesized and characterized six compounds 2.29-2.31a and b that belong to a class of nitrogen heterocycles: the carboline-based homodimers. Each individual homodimer features a 1,3-trans relationship on each side of the central diketopiperazine core, but differs in absolute stereochemistry and also in substitution on the 4' and 4'' oxygens (-Bn, -CH(3), or -H). The in vitro cytotoxicity of the six compounds was evaluated by measuring the growth inhibition in NCI-H520 and PC-3 human carcinoma cells. Phenol 2.30a inhibited cancer cell growth approximately three times better than its enantiomer 2.30b and possessed a GI(50) comparable to the clinically used agent etoposide in both cell lines. We have concluded that both the stereochemistry imparted by l-tryptophan and the presence of hydroxy substituents at the 4' and 4'' positions are necessary to generate cytotoxic properties in the homodimer class. We are now employing 2.30a as a new lead compound in our efforts to discover improved indole-based cancer chemotherapeutics.

Regioselective oxidation of 2-amino-3-aroyl-4,5-dialkylthiophenes by DMSO.

Solutions of 2-amino-3-aroyl-4,5-dialkylthiophenes in DMSO (dimethyl sulfoxide) undergo regioselective oxidation of benzylic carbon under mild conditions. We describe three examples and propose a mechanism for oxidation.

In vitro metabolism of 2-acetylbenzothiophene: relevance to zileuton hepatotoxicity.

Zileuton, an inhibitor of 5-lipooxygenase, the initial enzyme in the leukotriene pathway, was marketed as a new treatment for asthma. This drug has been associated with liver toxicity, which has limited its clinical usefulness. We provide evidence here that the liver toxicity likely involves a sequence of biotransformations leading to 2-acetylbenzothiophene (2-ABT), which is subsequently metabolized to give a reactive intermediate(s). In vitro experiments with the human lymphoblast MCL5 cell line demonstrated that 2-ABT is cytotoxic in a P450-dependent manner. Human liver microsome (HLM) incubations with 2-ABT revealed the formation of two short-lived oxidized species, "M + 16" and "M + 32". Both of these metabolites formed adducts in the presence of GSH or NAC. Singly oxidized M + 16 adducts, from either GSH or NAC, appeared to be unstable in acidic medium and eliminated water readily to form a new compound. Authentic synthetic standards demonstrated that 2-ABT-S-oxide M1 corresponded to the M + 16 metabolite and that the S-oxide underwent nucleophilic addition with GSH and NAC to produce the singly oxidized adducts observed in HLM. The S-oxide adducts readily eliminated water to form a rearomatized 2-ABT-GSH adduct or 2-ABT-NAC adduct. Coelution experiments with the synthetic standard confirmed the structure of the eliminated 2-ABT-NAC adduct C1. LC/MS analyses of urine samples collected from rats dosed with zileuton indicate that C1 is a metabolite of zileuton formed in vivo. The in vitro and in vivo data presented here demonstrate the formation of 2-ABT from zileuton and its further bioactivation to a potentially toxic metabolite.