Benchmarking causal reasoning algorithms for gene expression-based compound mechanism of action analysis.

BACKGROUND: Elucidating compound mechanism of action (MoA) is beneficial to drug discovery, but in practice often represents a significant challenge. Causal Reasoning approaches aim to address this situation by inferring dysregulated signalling proteins using transcriptomics data and biological networks; however, a comprehensive benchmarking of such approaches has not yet been reported. Here we benchmarked four causal reasoning algorithms (SigNet, CausalR, CausalR ScanR and CARNIVAL) with four networks (the smaller Omnipath network vs. 3 larger MetaBase™ networks), using LINCS L1000 and CMap microarray data, and assessed to what extent each factor dictated the successful recovery of direct targets and compound-associated signalling pathways in a benchmark dataset comprising 269 compounds. We additionally examined impact on performance in terms of the functions and roles of protein targets and their connectivity bias in the prior knowledge networks. RESULTS: According to statistical analysis (negative binomial model), the combination of algorithm and network most significantly dictated the performance of causal reasoning algorithms, with the SigNet recovering the greatest number of direct targets. With respect to the recovery of signalling pathways, CARNIVAL with the Omnipath network was able to recover the most informative pathways containing compound targets, based on the Reactome pathway hierarchy. Additionally, CARNIVAL, SigNet and CausalR ScanR all outperformed baseline gene expression pathway enrichment results. We found no significant difference in performance between L1000 data or microarray data, even when limited to just 978 'landmark' genes. Notably, all causal reasoning algorithms also outperformed pathway recovery based on input DEGs, despite these often being used for pathway enrichment. Causal reasoning methods performance was somewhat correlated with connectivity and biological role of the targets. CONCLUSIONS: Overall, we conclude that causal reasoning performs well at recovering signalling proteins related to compound MoA upstream from gene expression changes by leveraging prior knowledge networks, and that the choice of network and algorithm has a profound impact on the performance of causal reasoning algorithms. Based on the analyses presented here this is true for both microarray-based gene expression data as well as those based on the L1000 platform.

MAVEN: compound mechanism of action analysis and visualisation using transcriptomics and compound structure data in R/Shiny.

BACKGROUND: Understanding the Mechanism of Action (MoA) of a compound is an often challenging but equally crucial aspect of drug discovery that can help improve both its efficacy and safety. Computational methods to aid MoA elucidation usually either aim to predict direct drug targets, or attempt to understand modulated downstream pathways or signalling proteins. Such methods usually require extensive coding experience and results are often optimised for further computational processing, making them difficult for wet-lab scientists to perform, interpret and draw hypotheses from. RESULTS: To address this issue, we in this work present MAVEN (Mechanism of Action Visualisation and Enrichment), an R/Shiny app which allows for GUI-based prediction of drug targets based on chemical structure, combined with causal reasoning based on causal protein-protein interactions and transcriptomic perturbation signatures. The app computes a systems-level view of the mechanism of action of the input compound. This is visualised as a sub-network linking predicted or known targets to modulated transcription factors via inferred signalling proteins. The tool includes a selection of MSigDB gene set collections to perform pathway enrichment on the resulting network, and also allows for custom gene sets to be uploaded by the researcher. MAVEN is hence a user-friendly, flexible tool for researchers without extensive bioinformatics or cheminformatics knowledge to generate interpretable hypotheses of compound Mechanism of Action. CONCLUSIONS: MAVEN is available as a fully open-source tool at https://github.com/laylagerami/MAVEN with options to install in a Docker or Singularity container. Full documentation, including a tutorial on example data, is available at https://laylagerami.github.io/MAVEN .

Two-Site Evaluation of the Repeatability and Precision of an Automated Dual-Column Hydrogen/Deuterium Exchange Mass Spectrometry Platform.

Hydrogen/deuterium exchange coupled with mass spectrometry (HDX-MS) is an information-rich biophysical method for the characterization of protein dynamics. Successful applications of differential HDX-MS include the characterization of protein-ligand binding. A single differential HDX-MS data set (protein ± ligand) is often comprised of more than 40 individual HDX-MS experiments. To eliminate laborious manual processing of samples, and to minimize random and gross errors, automated systems for HDX-MS analysis have become routine in many laboratories. However, an automated system, while less prone to random errors introduced by human operators, may have systematic errors that go unnoticed without proper detection. Although the application of automated (and manual) HDX-MS has become common, there are only a handful of studies reporting the systematic evaluation of the performance of HDX-MS experiments, and no reports have been published describing a cross-site comparison of HDX-MS experiments. Here, we describe an automated HDX-MS platform that operates with a parallel, two-trap, two-column configuration that has been installed in two remote laboratories. To understand the performance of the system both within and between laboratories, we have designed and completed a test-retest repeatability study for differential HDX-MS experiments implemented at each of two laboratories, one in Florida and the other in Spain. This study provided sufficient data to do both within and between laboratory variability assessments. Initial results revealed a systematic run-order effect within one of the two systems. Therefore, the study was repeated, and this time the conclusion was that the experimental conditions were successfully replicated with minimal systematic error.

Conformational features of secondary N-cyclopropyl amides.

NMR studies in conjunction with ab initio calculations revealed unexpected conformational behavior of N-cyclopropylacetamide (1). This secondary amide displays 16-19% E-rotamer (cis) around the carbonyl-nitrogen bond in apolar solvents, in contrast to other aliphatic secondary acetamides in which significant E-rotamer populations are rare due to steric contacts between the substituents on the amide bond. In addition, 1 adopts an ortho conformation around the N-cPr bond instead of the anti conformation generally preferred by secondary acetamides. This distinct conformational behavior was also observed for other secondary N-cyclopropyl amides.

The chemical basis of serine palmitoyltransferase inhibition by myriocin.

Sphingolipids (SLs) are essential components of cellular membranes formed from the condensation of L-serine and a long-chain acyl thioester. This first step is catalyzed by the pyridoxal-5'-phosphate (PLP)-dependent enzyme serine palmitoyltransferase (SPT) which is a promising therapeutic target. The fungal natural product myriocin is a potent inhibitor of SPT and is widely used to block SL biosynthesis despite a lack of a detailed understanding of its molecular mechanism. By combining spectroscopy, mass spectrometry, X-ray crystallography, and kinetics, we have characterized the molecular details of SPT inhibition by myriocin. Myriocin initially forms an external aldimine with PLP at the active site, and a structure of the resulting co-complex explains its nanomolar affinity for the enzyme. This co-complex then catalytically degrades via an unexpected 'retro-aldol-like' cleavage mechanism to a C18 aldehyde which in turn acts as a suicide inhibitor of SPT by covalent modification of the essential catalytic lysine. This surprising dual mechanism of inhibition rationalizes the extraordinary potency and longevity of myriocin inhibition.

Mechanism of action deconvolution of the small-molecule pathological tau aggregation inhibitor Anle138b.

BACKGROUND: A key histopathological hallmark of Alzheimer's disease (AD) is the presence of neurofibrillary tangles of aggregated microtubule-associated protein tau in neurons. Anle138b is a small molecule which has previously shown efficacy in mice in reducing tau aggregates and rescuing AD disease phenotypes. METHODS: In this work, we employed bioinformatics analysis-including pathway enrichment and causal reasoning-of an in vitro tauopathy model. The model consisted of cultured rat cortical neurons either unseeded or seeded with tau aggregates derived from human AD patients, both of which were treated with Anle138b to generate hypotheses for its mode of action. In parallel, we used a collection of human target prediction models to predict direct targets of Anle138b based on its chemical structure. RESULTS: Combining the different approaches, we found evidence supporting the hypothesis that the action of Anle138b involves several processes which are key to AD progression, including cholesterol homeostasis and neuroinflammation. On the pathway level, we found significantly enriched pathways related to these two processes including those entitled "Superpathway of cholesterol biosynthesis" and "Granulocyte adhesion and diapedesis". With causal reasoning, we inferred differential activity of SREBF1/2 (involved in cholesterol regulation) and mediators of the inflammatory response such as NFKB1 and RELA. Notably, our findings were also observed in Anle138b-treated unseeded neurons, meaning that the inferred processes are independent of tau pathology and thus represent the direct action of the compound in the cellular system. Through structure-based ligand-target prediction, we predicted the intracellular cholesterol carrier NPC1 as well as NF-κB subunits as potential targets of Anle138b, with structurally similar compounds in the model training set known to target the same proteins. CONCLUSIONS: This study has generated feasible hypotheses for the potential mechanism of action of Anle138b, which will enable the development of future molecular interventions aiming to reduce tau pathology in AD patients.

Crystal structure of active CDK4-cyclin D and mechanistic basis for abemaciclib efficacy.

Despite the biological and therapeutic relevance of CDK4/6 for the treatment of HR+, HER2- advanced breast cancer, the detailed mode of action of CDK4/6 inhibitors is not completely understood. Of particular interest, phosphorylation of CDK4 at T172 (pT172) is critical for generating the active conformation, yet no such crystal structure has been reported to date. We describe here the x-ray structure of active CDK4-cyclin D3 bound to the CDK4/6 inhibitor abemaciclib and discuss the key aspects of the catalytically-competent complex. Furthermore, the effect of CDK4/6 inhibitors on CDK4 T172 phosphorylation has not been explored, despite its role as a potential biomarker of CDK4/6 inhibitor response. We show mechanistically that CDK4/6i stabilize primed (pT172) CDK4-cyclin D complex and selectively displace p21 in responsive tumor cells. Stabilization of active CDK4-cyclin D1 complex can lead to pathway reactivation following alternate dosing regimen. Consequently, sustained binding of abemaciclib to CDK4 leads to potent cell cycle inhibition in breast cancer cell lines and prevents rebound activation of downstream signaling. Overall, our study provides key insights demonstrating that prolonged treatment with CDK4/6 inhibitors and composition of the CDK4/6-cyclin D complex are both critical determinants of abemaciclib efficacy, with implications for this class of anticancer therapy.

The Angiopoietin-Like Protein 3 and 8 Complex Interacts with Lipoprotein Lipase and Induces LPL Cleavage.

Lipoprotein lipase (LPL) is the key enzyme that hydrolyzes triglycerides from triglyceride-rich lipoproteins. Angiopoietin-like proteins (ANGPTL) 3, 4, and 8 are well-characterized protein inhibitors of LPL. ANGPTL8 forms a complex with ANGPTL3, and the complex is a potent endogenous inhibitor of LPL. However, the nature of the structural interaction between ANGPTL3/8 and LPL is unknown. To probe the conformational changes in LPL induced by ANGPTL3/8, we found that HDX-MS detected significantly altered deuteration in the lid region, ApoC2 binding site, and furin cleavage region of LPL in the presence of ANGPTL3/8. Supporting this HDX structural evidence, we found that ANGPTL3/8 inhibits LPL enzymatic activities and increases LPL cleavage. ANGPTL3/8-induced effects on LPL activity and LPL cleavage are much stronger than those of ANGPTL3 or ANGPTL8 alone. ANGPTL3/8-mediated LPL cleavage is blocked by both an ANGPTL3 antibody and a furin inhibitor. Knock-down of furin expression by siRNA significantly reduced ANGPT3/8-induced cleavage of LPL. Our data suggest ANGPTL3/8 promotes furin-mediated LPL cleavage.

Defining Target Engagement Required for Efficacy In Vivo at the Retinoic Acid Receptor-Related Orphan Receptor C2 (RORγt).

The Th17 pathway has been implicated in autoimmune diseases. The retinoic acid receptor-related orphan receptor C2 (RORγt) is a master regulator of Th17 cells and controls the expression of IL-17A. RORγt is expressed primarily in IL-17A-producing lymphoid cells. Here we describe a virtual screen of the ligand-binding pocket and subsequent screen in a binding assay that identified the 1-benzyl-4',5'-dihydrospiro[piperidine-4,7'-thieno[2,3-c]pyran]-2'-carboxamide scaffold as a starting point for optimization of binding affinity and functional activity guided by structure-based design. Compound 12 demonstrated activity in a mouse PK/PD model and efficacy in an inflammatory arthritis mouse model that were used to define the level and duration of target engagement required for efficacy in vivo. Further optimization to improve ADME and physicochemical properties with guidance from simulations and modeling provided compound 22, which is projected to achieve the level and duration of target engagement required for efficacy in the clinic.

A novel phage display vector for selection of target-specific peptides.

Intrinsic low display level of polypeptides on phage is a fundamental and limiting hurdle in successful isolation of target-specific binders by phage display technology. To circumvent this challenge, we optimized the copy number of peptides displayed on the phage surface using type 33 phage vector. We randomized the first 67 amino acids of the wild type PIII to identify mutants that would result in its reduced expression. Consequently, the display level was improved by 30-fold due to higher incorporation of the synthetic PIII-peptide fusion protein on the phage surface. Utilization of this novel phage vector should provide a solid basis for the discovery of therapeutic peptides.

Integration of phage and yeast display platforms: A reliable and cost effective approach for binning of peptides as displayed on-phage.

Hundreds of target specific peptides are routinely discovered by peptide display platforms. However, due to the high cost of peptide synthesis only a limited number of peptides are chemically made for further analysis. Here we describe an accurate and cost effective method to bin peptides on-phage based on binding region(s), without any requirement for peptide or protein synthesis. This approach, which integrates phage and yeast display platforms, requires display of target and its alanine variants on yeast. Flow cytometry was used to detect binding of peptides on-phage to the target on yeast. Once hits were identified, they were synthesized to confirm their binding region(s) by HDX (Hydrogen deuterium exchange) and crystallography. Moreover, we have successfully shown that this approach can be implemented as part of a panning process to deplete non-functional peptides. This technique can be applied to any target that can be successfully displayed on yeast; it narrows down the number of peptides requiring synthesis; and its utilization during selection results in enrichment of peptide population against defined binding regions on the target.

High-throughput hydrogen bond strength calculation and its applications in optimizing drug ADME properties.

AIM: Modifying the molecule's intrinsic hydrogen bond strength (HBS) is a useful approach in optimizing its permeability and P-glycoprotein (P-gp) efflux. Quantum mechanics (QM) based computation has been utilized to estimate the molecular intrinsic HBS. Despite its usefulness, the computation is time consuming for a large set of molecules. METHODOLOGY/RESULTS: We introduced a fragment-based high-throughput HBS calculation method and validated it with internal and external datasets. Examples have been presented where the P-gp efflux and permeability can be optimized by modulating calculated HBS. CONCLUSION: The results will enable medicinal chemists to calculate HBS in a high-throughput manner while optimizing permeability and P-gp efflux. This will further improve the efficiency of balancing multiple properties during drug discovery process.

A Decoupled Automation Platform for Hydrogen/Deuterium Exchange Mass Spectrometry Experiments.

Hydrogen/deuterium exchange mass spectrometry (HDX-MS) is a biophysical technique well suited to the characterization of protein dynamics and protein-ligand interactions. In order to accurately define the rate of exchange, HDX experiments require the repeated measure of deuterium incorporation into the target protein across a range of time points. Accordingly, the HDX-MS experiment is well suited to automation, and a number of automated systems for HDX-MS have been developed. The most widely utilized platforms all operate an integrated design, where robotic liquid handling is interfaced directly with a mass spectrometer. With integrated designs, the exchange samples are prepared and injected into the LC-MS following a "real-time" serial workflow. Here we describe a new HDX-MS platform that is comprised of two complementary pieces of automation that disconnect the sample preparation from the LC-MS analysis. For preparation, a plate-based automation system is used to prepare samples in parallel, followed by immediate freezing and storage. A second piece of automation has been constructed to perform the thawing and LC-MS analysis of frozen samples in a serial mode and has been optimized to maximize the duty cycle of the mass spectrometer. The decoupled configuration described here reduces experiment time, significantly improves capacity, and improves the flexibility of the platform when compared with a fully integrated system.

The synthesis of azadirachtin: a potent insect antifeedant.

We describe in full the first synthesis of the potent insect antifeedant azadirachtin through a highly convergent approach. An O-alkylation reaction is used to unite decalin ketone and propargylic mesylate fragments, after which a Claisen rearrangement constructs the central C8-C14 bond in a stereoselective fashion. The allene which results from this sequence then enables a second critical carbon-carbon bond forming event whereby the [3.2.1] bicyclic system, present in the natural product, is generated via a 5-exo-radical cyclisation process. Finally, using knowledge gained through our early studies into the reactivity of the natural product, a series of carefully designed steps completes the synthesis of this challenging molecule.

Deciphering Binding Interactions of IL-23R with HDX-MS: Mapping Protein and Macrocyclic Dodecapeptide Ligands.

Molecular characterization of the binding epitope of IL-23R and its cognate cytokine IL-23 is paramount to understand the role in autoimmune diseases and to support the discovery of new inhibitors of this protein-protein interaction. Our results revealed that HDX-MS was able to identify the binding epitope of IL-23R:IL-23, which opened the way to evaluate a peptide macrocycle described in the literature as disrupter of this autoimmune target. Thus, the characterization of the interactions of this chemotype by HDX-MS in combination with computational approaches was achieved. To our knowledge, this is the first reported structural evidence regarding the site where a small compound binds to IL-23R.

Utilization of peptide phage display to investigate hotspots on IL-17A and what it means for drug discovery.

To date, IL-17A antibodies remain the only therapeutic approach to correct the abnormal activation of the IL-17A/IL-17R signaling complex. Why is it that despite the remarkable success of IL-17 antibodies, there is no small molecule antagonist of IL-17A in the clinic? Here we offer a unique approach to address this question. In order to understand the interaction of IL-17A with its receptor, we combined peptide discovery using phage display with HDX, crystallography, and functional assays to map and characterize hot regions that contribute to most of the energetics of the IL-17A/IL-17R interaction. These functional maps are proposed to serve as a guide to aid in the development of small molecules that bind to IL-17A and block its interaction with IL-17RA.

Assignment of absolute configuration on the basis of the conformational effects induced by chiral derivatizing agents: the 2-arylpyrrolidine case.

A novel approach for determining the absolute configuration of a chiral compound is proposed. The methodology is based on the distinct conformational effects imposed on a chiral substrate by each enantiomer of a chiral derivatizing agent. As a proof of concept, it is shown that the absolute configuration of 2-arylpyrrolidines can easily be determined by inspection of the multiplicity of the NMR signal of the methine proton of the pyrrolidine ring in the corresponding Mosher's amides.

A Residue-Resolved Bayesian Approach to Quantitative Interpretation of Hydrogen-Deuterium Exchange from Mass Spectrometry: Application to Characterizing Protein-Ligand Interactions.

Characterization of interactions between proteins and other molecules is crucial for understanding the mechanisms of action of biological systems and, thus, drug discovery. An increasingly useful approach to mapping these interactions is measurement of hydrogen/deuterium exchange (HDX) using mass spectrometry (HDX-MS), which measures the time-resolved deuterium incorporation of peptides obtained by enzymatic digestion of the protein. Comparison of exchange rates between apo- and ligand-bound conditions results in a mapping of the differential HDX (ΔHDX) of the ligand. Residue-level analysis of these data, however, must account for experimental error, sparseness, and ambiguity due to overlapping peptides. Here, we propose a Bayesian method consisting of a forward model, noise model, prior probabilities, and a Monte Carlo sampling scheme. This method exploits a residue-resolved exponential rate model of HDX-MS data obtained from all peptides simultaneously, and explicitly models experimental error. The result is the best possible estimate of ΔHDX magnitude and significance for each residue given the data. We demonstrate the method by revealing richer structural interpretation of ΔHDX data on two nuclear receptors: vitamin D-receptor (VDR) and retinoic acid receptor gamma (RORγ). The method is implemented in HDX Workbench and as a standalone module of the open source Integrative Modeling Platform.

Optimization of Hydroxyethylamine Transition State Isosteres as Aspartic Protease Inhibitors by Exploiting Conformational Preferences.

NMR conformational analysis of a hydroxyethylamine peptide isostere developed as an aspartic protease inhibitor shows that it is a flexible architecture. Cyclization to form pyrrolidines, piperidines, or morpholines results in a preorganization of the whole system in solution. The resulting conformation is similar to the conformation of the inhibitor in the active site of BACE-1. This entropic gain results in increased affinity for the enzyme when compared with the acyclic system. For morpholines 27 and 29, the combination of steric and electronic factors is exploited to orient substituents toward S1, S1', and S2' pockets both in the solution and in the bound states. These highly preorganized molecules proved to be the most potent compounds of the series. Additionally, the morpholines, unlike the pyrrolidine and piperidine analogues, have been found to be brain penetrant BACE-1 inhibitors.