Collaborations Pharmaceuticals (United States)

There is an urgent need to develop additional treatments for tuberculosis (TB) to complement the small panel of approved drugs and to devise shorter treatment regimens. Within the last 20 years, we have seen an increased focus on using cheminformatics-based approaches to understand the properties of Mycobacterium tuberculosis (Mtb) active molecules and machine learning algorithms to subsequently learn from public data sets. We now demonstrate how we have continually used many machine learning approaches that have enabled us to select or synthesize new compounds for testing in vitro to validate our models and to identify new chemical matter. We now put our results into context with studies from other groups to make the case for using machine learning models more widely to aid in finding new Mtb inhibitors. TB research has been slow to adapt to these approaches to increase drug discovery efficiency, but it is better late than never.

CYP2B6 is an important enzyme in Phase I metabolism of key pharmaceuticals, and inhibition of this enzyme could lead to adverse drug events. Machine learning models can potentially predict interactions with CYP2B6; however, there is limited data with which to train these models in the public domain. We proposed to enhance the applicability domain and improve the predictive capability of our CYP2B6 inhibition model by selecting a small diverse set of compounds to test in vitro and adding the results to our model training set. We used a distance-based approach to define the applicability domain of the model and then measured the chemical diversity by creating t-Distributed stochastic neighbor embedding plots to represent the chemical space of our model. After comparing this chemical space to a 49-plate drug-repurposing library, we were able to identify a plate with the highest average minimum Euclidean distance from the model training set. We then performed in vitro testing of this plate for CYP2B6 inhibition activity at 10 μM and added this new data to our machine learning model. A one-class classification approach was used to evaluate the efficacy of our applicability domain-expansion technique. These results show that this method did not appreciably increase the performance of the model or the applicability domain, but we did increase the diversity of the training set. Additionally, the in vitro experiments identified vilanterol and allylestrenol as inhibitors of CYP2B6 with IC50 values in the sub-low micromolar range.SignificanceCYP2B6 inhibition can affect the metabolism of important drugs, like methadone and propofol, and result in variability that would lead to adverse events. Machine learning models can help uncover new molecules with inhibitory potential of this enzyme, but only if the predictions from these models are reliable. This study illustrates how the intentional expansion of a machine learning model’s applicability domain is neither a simple nor straightforward task, but even a conservative effort can reveal new molecules with CYP2B6 inhibition activity.

Chemokine receptors CCR3, CCR4, and CCR5 are G protein-coupled receptors implicated in diseases like cancer, Alzheimer's, asthma, human immunodeficiency virus (HIV), and macular degeneration. Recently, CCR3 and CCR4 have emerged as potential stroke targets. Although only the CCR5 antagonist maraviroc is US Food and Drug Administration-approved (for HIV), we curated data on CCR3, CCR4, and CCR5 antagonists from ChEMBL to develop and validate machine learning models. The top 5-fold cross-validation statistics for these models were high for both classification and regression models for CCR3 (receiver operating characteristic [ROC], 0.94; R2= 0.8), CCR4 (ROC, 0.98; R2= 0.57), and CCR5 (ROC, 0.96; R2= 0.78). The models for CCR3/4 were used to screen a small library of US Food and Drug Administration-approved drugs and 17 were initially tested invitro against both CCR3/4 receptors. A promising compound lapatinib, a dual tyrosine kinase inhibitor, was identified as an antagonist for CCR3 (IC50, 0.7 μM) and CCR4 (IC50, 1.8μM). Additional testing also identified it as an CCR5 antagonist (IC50, 0.9 μM), and it showed moderate invitro HIV I inhibition. We demonstrated how machine learning can be used to identify molecules for repurposing as antagonists for G protein-coupled receptors such as CCR3, CCR4, and CCR5. Lapatinib may represent a new orally available chemical probe for these 3 receptors, and it provides a starting point for further chemical optimization for multiple diseases impacting human health. SIGNIFICANCE STATEMENT: We describe the building of machine learning models for the chemokine receptors CCR3, CCR4, and CCR5 trained on data from the ChEMBL database. Using these models, we identified lapatinib as a potent inhibitor of CCR3, CCR4, and CCR5. Our study illustrates the potential of machine learning in identifying molecules for repurposing as antagonists for G protein-coupled receptors, including CCR3, CCR4, and CCR5, which have various therapeutic applications.

The development of new drugs addressing serious mental health and other disorders should avoid the psychedelic experience. Analogs of psychedelic drugs can have clinical utility and are termed "psychoplastogens". These represent promising candidates for treating opioid use disorder to reduce drug dependence, with rarely reported serious adverse effects. This drug abuse cessation is linked to the induction of neuritogenesis and increased neuroplasticity, a hallmark of psychedelic molecules, such as lysergic acid diethylamine. Some, but not all psychoplastogens may act through the G-protein coupled receptor (GPCR) 5HT2A whereas others may display very different polypharmacology making prediction of hallucinogenic potential challenging. In the process of developing tools to help design new psychoplastogens, we have used artificial intelligence in the form of machine learning classification models for predicting psychedelic effects using a published in vitro data set from PsychLight (support vector classification (SVC), area under the curve (AUC) 0.74) and in vivo human data derived from books from Shulgin and Shulgin (SVC, AUC, 0.72) with nested five-fold cross validation. We have also explored conformal predictors with ECFP6 and electrostatic descriptors in an effort to optimize them. These models have been used to predict known 5HT2A agonists to assess their potential to act as psychedelics and induce hallucinations for PsychLight (SVC, AUC 0.97) and Shulgin and Shulgin (random forest, AUC 0.71). We have tested these models with head twitch data from the mouse. This predictive capability is desirable to reliably design new psychoplastogens that lack in vivo hallucinogenic potential and help assess existing and future molecules for this potential. These efforts also provide useful insights into understanding the psychedelic structure activity relationship.

Venomous agent X (VX) is an organophosphate acetylcholinesterase (AChE) inhibitor, and although it is one of the most toxic AChE inhibitors known, the extent of metabolism in humans is not currently well understood. The known metabolism in humans is limited to the metabolite identification from a single victim of the Osaka poisoning in 1994, which allowed for the identification of several metabolic products. VX has been reported to be metabolized in vitro by paraoxonase-1 and phosphotriesterase, although their binding constants are many orders of magnitude above the LD50, suggesting limited physiologic relevance. Using incubation with human liver microsomes (HLMs), we have now characterized the metabolism of VX and the formation of multiple metabolites as well as identified a Food and Drug Administration-approved drug [ethylenediaminetetraacetic acid (EDTA)] that enhances the metabolic rate. HLM incubation alone shows a pronounced increase in the metabolism of VX compared with buffer, suggesting that cytochrome P450-mediated metabolism of VX is occurring. We identified a biphasic decay with two distinct rates of metabolism. The enhancement of VX metabolism in multiple buffers was assessed to attempt to mitigate the effect of hydrolysis rates. The formation of VX metabolites was shown to be shifted with HLMs, suggesting a pathway enhancement over simple hydrolysis. Additionally, our investigation of hydrolysis rates in various common buffers used in biologic assays discovered dramatic differences in VX stability. The new human in vitro VX metabolic data reported points to a potential in vivo treatment strategy (EDTA) for rescue in individuals that are poisoned though enhancement of metabolism alongside existing treatments. SIGNIFICANCE STATEMENT: Venomous agent X (VX) is a potent acetylcholinesterase inhibitor and chemical weapon. To date, we do not possess a clear understanding of its metabolism in humans that would assist us in treating those exposed to it. This study now describes the human liver microsomal metabolism of VX and identifies ethylenediaminetetraacetic acid, which appears to enhance the rate of metabolism. This may provide a potential treatment option for human VX poisoning.

Hepatitis B virus (HBV) is a hepatotropic DNA virus that replicates by reverse transcription. It chronically infects >296 million people worldwide, including ∼850,000 in the USA, and kills 820,000 annually worldwide. Current nucleos(t)ide analogue (NA) or pegylated interferon α therapies do not eradicate the virus and would benefit from a complementary antiviral drug. We performed a preliminary screen of 28 dispirotripiperazines against HBV, identifying 9 hits with EC50 of 0.7-25 μM. Compound 11826096 displays the most potent activity and represents a promising lead for future optimization. While the mechanism of action is unknown, preliminary assays limit possible targets to activities involved in RNA accumulation, translation, capsid assembly, and/or capsid stability. In addition, we built machine learning models to determine if they were able to predict the activity of this series of compounds. The novelty of these molecules indicated they were outside of the applicability domain of these models.

There are very fewsmall-molecule antivirals for SARS-CoV-2thatare either currently approved (or emergency authorized) in the USor globally, including remdesivir, molnupiravir, and paxlovid. Theincreasing number of SARS-CoV-2 variants that have appeared sincethe outbreak began over three years ago raises the need for continualdevelopment of updated vaccines and orally available antivirals inorder to fully protect or treat the population. The viral main protease(Mpro) and the papain-like protease (PLpro)are key for viral replication; therefore, they represent valuabletargets for antiviral therapy. We herein describe an in vitro screenperformed using the 2560 compounds from the Microsource Spectrum libraryagainst Mpro and PLpro in an attempt to identifyadditional small-molecule hits that could be repurposed for SARS-CoV-2.We subsequently identified 2 hits for Mpro and 8 hits forPLpro. One of these hits was the quaternary ammonium compoundcetylpyridinium chloride with dual activity (IC50 = 2.72± 0.09 μM for PLpro and IC50 = 7.25± 0.15 μM for Mpro). A second inhibitor of PLpro was the selective estrogen receptor modulator raloxifene(IC50 = 3.28 ± 0.29 μM for PLpro andIC50 = 42.8 ± 6.7 μM for Mpro). Weadditionally tested several kinase inhibitors and identified olmutinib(IC50 = 0.54 ± 0.04 μM), bosutinib (IC50 = 4.23 ± 0.28 μM), crizotinib (IC50 = 3.81± 0.04 μM), and dacominitinib (IC50 = IC50 3.33 ± 0.06 μM) as PLpro inhibitorsfor the first time. In some cases, these molecules have also beentested by others for antiviral activity for this virus, or we haveused Calu-3 cells infected with SARS-CoV-2. The results suggest thatapproved drugs can be identified with promising activity against theseproteases, and in several cases we or others have validated theirantiviral activity. The additional identification of known kinaseinhibitors as molecules targeting PLpro may provide newrepurposing opportunities or starting points for chemical optimization.