Discovery of non-boronic acid Arginase 1 inhibitors through virtual screening and biophysical methods.

Inhibiting Arginase 1 (ARG1), a metalloenzyme that hydrolyzes l-arginine in the urea cycle, has been demonstrated as a promising therapeutic avenue in immuno-oncology through the restoration of suppressed immune response in several types of cancers. Most of the currently reported small molecule inhibitors are boronic acid based. Herein, we report the discovery of non-boronic acid ARG1 inhibitors through virtual screening. Biophysical and biochemical methods were used to experimentally profile the hits while X-ray crystallography confirmed a class of trisubstituted pyrrolidine derivatives as optimizable alternatives for the development of novel classes of immuno-oncology agents targeting this enzyme.

Development of a robust crystallization platform for immune receptor TREM2 using a crystallization chaperone strategy.

TREM2 has been identified by genomic analysis as a potential and novel target for the treatment of Alzheimer's disease. To enable structure-based screening of potential small molecule therapeutics, we sought to develop a robust crystallization platform for the TREM2 Ig-like domain. A systematic set of constructs containing the structural chaperone, maltose binding protein (MBP), fused to the Ig domain of TREM2, were evaluated in parallel expression and purification, followed by crystallization studies. Using protein crystallization and high-resolution diffraction as a readout, a MBP-TREM2 Ig fusion construct was identified that generates reproducible protein crystals diffracting at 2.0 Å, which makes it suitable for soaking of potential ligands. Importantly, analysis of crystal packing interfaces indicates that most of the surface of the TREM2 Ig domain is available for small molecule binding. A proof of concept co-crystallization study with a small library of fragments validated potential utility of this system for the discovery of new TREM2 therapeutics.

A potent seven-membered cyclic BTK (Bruton's tyrosine Kinase) chiral inhibitor conceived by structure-based drug design to lock its bioactive conformation.

A seven-membered cyclic chiral analog of potent lead BTK inhibitor 1 was envisioned by structure-based design to lock the molecule into its bioactive conformation. For the elaboration of the seven-membered ring, compound 1 pyridone 6-position was substituted with the purpose to prevent formation of reactive metabolites. Eventually, the cyclic chiral compound 3 maintained the high potency of 1, and most importantly showed no activity at either GSH or TDI assays suggesting no formation of reactive metabolites. The anticipated bound conformation of 3 to BTK was confirmed by X-ray crystallography. Synthetically, the crucial seven-membered ring formation was obtained by using TosMIC as a connective reagent.

Design and synthesis of selective, dual fatty acid binding protein 4 and 5 inhibitors.

Dual inhibition of fatty acid binding proteins 4 and 5 (FABP4 and FABP5) is expected to provide beneficial effects on a number of metabolic parameters such as insulin sensitivity and blood glucose levels and should protect against atherosclerosis. Starting from a FABP4 selective focused screening hit, biostructure information was used to modulate the selectivity profile in the desired way and to design potent dual FABP4/5 inhibitors with good selectivity against FABP3. With very good pharmacokinetic properties and no major safety alerts, compound 12 was identified as a suitable tool compound for further in vivo investigations.

In vivo emergence of a novel mutant L159F/L320F in the NS5B polymerase confers low-level resistance to the HCV polymerase inhibitors mericitabine and sofosbuvir.

BACKGROUND: Resistance to mericitabine (prodrug of HCV NS5B polymerase inhibitor PSI-6130) is rare and conferred by the NS5B S282T mutation. METHODS: Serum HCV RNA from patients who experienced viral breakthrough, partial response, or nonresponse in 2 clinical trials in which patients received mericitabine plus peginterferon alfa-2a (40KD)/ribavirin were analyzed by population and clonal sequence analysis as well as phenotypic assay for assessment of in vivo mericitabine resistance. RESULTS: Among 405 patients treated with mericitabine plus peginterferon alfa-2a/ribavirin in PROPEL and JUMP-C, virologic breakthrough or nonresponse were not observed; 12 patients experienced a partial response. The NS5B S282T resistance mutation was not observed in any patient. A number of treatment-associated NS5B changes were observed and characterized. A novel double mutant (L159F/L320F) with impaired replication capacity was detected in one HCV genotype 1b-infected patient. Introduction of double mutant L159F/L320F into genotype 1a (H77) and 1b (Con-1) replicons, respectively, increased the EC50 for mericitabine by 3.1- and 5.5-fold and the EC90 by 3.1- and 8.9-fold. The double mutant also decreased susceptibility to sofosbuvir (GS-7977) and GS-938 but not setrobuvir, relative to wild-type. CONCLUSIONS: A novel and replication-deficient double mutation (L159F/L320F) confers low-level resistance to mericitabine and cross-resistance to both sofosbuvir and GS-938. CLINICAL TRIALS REGISTRATION: NCT00869661, NCT01057667.

Discovery of pyrazolo[3,4-d]pyrimidine derivatives as GPR119 agonists.

We designed and synthesized a novel series of phenylamino- and phenoxy-substituted pyrazolo[3,4-d]pyrimidine derivatives as GPR119 agonists. SAR studies indicated that electron-withdrawing substituents on the phenyl ring are important for potency and full efficacy. Compound 26 combined good potency with a promising pharmacokinetic profile in mice, and lowered the glucose excursion in mice in an oral glucose-tolerance test.

Discovery of camphor-derived pyrazolones as 11ß-hydroxysteroid dehydrogenase type 1 inhibitors.

Starting from screening hit, (4S,7R)-1,7,8,8-tetramethyl-2-phenyl-1,2,4,5,6,7-hexahydro-4,7-methano-indazol-3-one (7), we optimized the potency and pharmacokinetic properties. This led to the identification of compounds with good in vivo activity in a mouse pharmacodynamic model of inhibition of 11ßHSD1.

A simple statistical test to infer the causality of target/phenotype correlation from small molecule phenotypic screens.

MOTIVATION: Cell-based phenotypic screens using small molecule inhibitors is an important technology for early drug discovery if the relationship between the disease-related cellular phenotype and inhibitors' biological targets can be determined. However, chemical inhibitors are rightfully believed to be less specific than perturbation by biological agents, such as antibody and small inference RNA. Therefore, it is often a challenge in small molecule phenotypic screening to infer the causality between a particular cellular phenotype and the inactivation of the responsible protein due to the off-target effect of the inhibitors. RESULTS: In this article, we present a Roche in-house effort of screening 746 structurally diverse compounds for their cytotoxicity in HeLa cells measured by high content imaging technology. These compounds were also systematically profiled for the targeted and off-target binding affinity to a panel of 25 pre-selected protein kinases in a cell-free system. In an effort to search for the kinases whose activities are crucial for cell survival, we found that the simple association method such as the chi-square test yields a large number of false positives because the observed cytotoxic phenotype is likely to be the result of promiscuous action of less specific inhibitors instead of true consequence of inactivation of single relevant target. We demonstrated that a stratified categorical data analysis technique such as the Cochran-Mantel-Haenszel test is an effective approach to extract the meaningful biological connection from the spurious correlation resulted from confounding covariates. This study indicates that, empowered by appropriate statistical adjustment, small molecule inhibitor perturbation remains a powerful tool to pin down the relevant biomarker for drug safety and efficacy research. CONTACT: xin.wei@roche.com SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Potent MCH-1 receptor antagonists from cis-1,4-diaminocyclohexane-derived indane analogs.

Benzimidazole and indane are the two key fragments in our potent and selective MCH-1 receptor (MCHR1) antagonists. To identify novel linkers connecting the two fragments, we investigated diamino-cycloalkane-derived analogs and discovered highly potent antagonists with cis-1,4-diaminocyclohexane as a unique spacer in this chemical class. Structural overlay suggested that cis-1-substituted-4-aminocyclohexane functions as a bioisostere of 4-substituted-piperidine and that the active conformation adopts a U-shaped orientation.

Discovery of 1-arylcarbonyl-6,7-dimethoxyisoquinoline derivatives as glutamine fructose-6-phosphate amidotransferase (GFAT) inhibitors.

Through high throughput screening and subsequent hit identification and optimization, we synthesized a series of 1-arylcarbonyl-6,7-dimethoxyisoquinoline derivatives as the first reported potent and reversible GFAT inhibitors. SAR studies of this class of compounds indicated significant impact on GFAT inhibition potency by substitutions on the A-ring and C-ring. The ketone group was found to be necessary for high potency. Compound 28 (RO0509347) demonstrated potent GFAT inhibition (IC(50)=1µM) with a desirable pharmacokinetic profile in rats, and showed significant efficacy in reducing the glucose excursion in an OGTT test in ob/ob mice.

Overview of artificial neural networks.

The artificial neural network (ANN), or simply neural network, is a machine learning method evolved from the idea of simulating the human brain. The data explosion in modem drug discovery research requires sophisticated analysis methods to uncover the hidden causal relationships between single or multiple responses and a large set of properties. The ANN is one of many versatile tools to meet the demand in drug discovery modeling. Compared to a traditional regression approach, the ANN is capable of modeling complex nonlinear relationships. The ANN also has excellent fault tolerance and is fast and highly scalable with parallel processing. This chapter introduces the background of ANN development and outlines the basic concepts crucially important for understanding more sophisticated ANN. Several commonly used learning methods and network setups are discussed briefly at the end of the chapter.

Discovery of [4-Amino-2-(1-methanesulfonylpiperidin-4-ylamino)pyrimidin-5-yl](2,3-difluoro-6- methoxyphenyl)methanone (R547), a potent and selective cyclin-dependent kinase inhibitor with significant in vivo antitumor activity.

The cyclin-dependent kinases (CDKs) and their cyclin partners are key regulators of the cell cycle. Since deregulation of CDKs is found with high frequency in many human cancer cells, pharmacological inhibition of CDKs with small molecules has the potential to provide an effective strategy for the treatment of cancer. The 2,4-diamino-5-ketopyrimidines 6 reported here represent a novel class of potent and ATP-competitive inhibitors that selectively target the cyclin-dependent kinase family. This diaminopyrimidine core with a substituted 4-piperidine moiety on the C2-amino position and 2-methoxybenzoyl at the C5 position has been identified as the critical structure responsible for the CDK inhibitory activity. Further optimization has led to a good number of analogues that show potent inhibitory activities against CDK1, CDK2, and CDK4 but are inactive against a large panel of serine/threonine and tyrosine kinases (K(i) > 10 microM). As one of these representative analogues, compound 39 (R547) has the best CDK inhibitory activities (K(i) = 0.001, 0.003, and 0.001 microM for CDK1, CDK2, and CDK4, respectively) and excellent in vitro cellular potency, inhibiting the growth of various human tumor cell lines including an HCT116 cell line (IC(50) = 0.08 microM). An X-ray crystal structure of 39 bound to CDK2 has been determined in this study, revealing a binding mode that is consistent with our SAR. Compound 39 demonstrates significant in vivo efficacy in the HCT116 human colorectal tumor xenograft model in nude mice with up to 95% tumor growth inhibition. On the basis of its superior overall profile, 39 was chosen for further evaluation and has progressed into Phase I clinical trial for the treatment of cancer.

Fragment library design: using cheminformatics and expert chemists to fill gaps in existing fragment libraries.

Fragment based screening (FBS) has emerged as a mainstream lead discovery strategy in academia, biotechnology start-ups, and large pharma. As a prerequisite of FBS, a structurally diverse library of fragments is desirable in order to identify chemical matter that will interact with the range of diverse target classes that are prosecuted in contemporary screening campaigns. In addition, it is also desirable to offer synthetically amenable starting points to increase the probability of a successful fragment evolution through medicinal chemistry. Herein we describe a method to identify biologically relevant chemical substructures that are missing from an existing fragment library (chemical gaps), and organize these chemical gaps hierarchically so that medicinal chemists can efficiently navigate the prioritized chemical space and subsequently select purchasable fragments for inclusion in an enhanced fragment library.

Discovery of Potent and Orally Active p53-MDM2 Inhibitors RO5353 and RO2468 for Potential Clinical Development.

The development of small-molecule MDM2 inhibitors to restore dysfunctional p53 activities represents a novel approach for cancer treatment. In a previous communication, the efforts leading to the identification of a non-imidazoline MDM2 inhibitor, RG7388, was disclosed and revealed the desirable in vitro and in vivo pharmacological properties that this class of pyrrolidine-based inhibitors possesses. Given this richness and the critical need for a wide variety of chemical structures to ensure success in the clinic, research was expanded to evaluate additional derivatives. Here we report two new potent, selective, and orally active p53-MDM2 antagonists, RO5353 and RO2468, as follow-ups with promising potential for clinical development.

Discovery of 2-[3,5-dichloro-4-(5-isopropyl-6-oxo-1,6-dihydropyridazin-3-yloxy)phenyl]-3,5-dioxo-2,3,4,5-tetrahydro[1,2,4]triazine-6-carbonitrile (MGL-3196), a Highly Selective Thyroid Hormone Receptor ß agonist in clinical trials for the treatment of dyslipidemia.

The beneficial effects of thyroid hormone (TH) on lipid levels are primarily due to its action at the thyroid hormone receptor ß (THR-ß) in the liver, while adverse effects, including cardiac effects, are mediated by thyroid hormone receptor α (THR-α). A pyridazinone series has been identified that is significantly more THR-ß selective than earlier analogues. Optimization of this series by the addition of a cyanoazauracil substituent improved both the potency and selectivity and led to MGL-3196 (53), which is 28-fold selective for THR-ß over THR-α in a functional assay. Compound 53 showed outstanding safety in a rat heart model and was efficacious in a preclinical model at doses that showed no impact on the central thyroid axis. In reported studies in healthy volunteers, 53 exhibited an excellent safety profile and decreased LDL cholesterol (LDL-C) and triglycerides (TG) at once daily oral doses of 50 mg or higher given for 2 weeks.

Design of libraries targeting protein-protein interfaces.

TARGETING PPIS: A novel strategy for designing libraries targeting protein-protein interfaces enabled us to identify diverse chemical entry points to interact with therapeutic targets for which conventional screening libraries delivered no or only few hit structures. The concept was experimentally validated by early hit evaluation in biochemical screens and early ADMET profiling.

Automated pharmacophore query optimization with genetic algorithms - a case study using the MC4R system.

Due to the recent availability of high quality small molecule databases, such as ZINC and PubChem,1,2 virtual screening is playing an even more important role in identifying biologically relevant molecules in drug discovery campaigns. The success of pharmacophore-based virtual screening (PBVS) relies largely on the accuracy and specificity of the pharmacophore query employed. Deriving a pharmacophore query from a single structure inevitably introduces uncertainty, and the derived query is unlikely to be optimal against every collection of input compounds, especially when it is desired to discriminate among compounds with similar chemical structures. In this study, we present an optimization approach empowered by genetic algorithms (GA) to enhance the accuracy and specificity of a primary pharmacophore query. The example utilized is the human melanocortin type 4 receptor (hMC4R), for which the pharmacophore query was built on the basis of the structure of a rigid cyclic peptide agonist.(3) The optimized query is shown to be capable of identifying 37 positive hMC4R agonists with no false positives from a training set containing 55 agonists and 51 nonagonists. This represents a significant improvement from the initial query which exhibited a 37/32 hit rate. The final, optimized query is challenged with a testing set comprising of 55 hMC4R agonists and 50 nonagonists and achieves a hit rate of 33/8, that improved from 40/31. The impact of GA controlling parameters, including mutation rate, crossover rate, fitness function, population size, and convergence criterion, on performance of optimization are examined and discussed.

Enhancing specificity and sensitivity of pharmacophore-based virtual screening by incorporating chemical and shape features--a case study of HIV protease inhibitors.

Virtual screening (VS), if applied appropriately, could significantly shorten the hit identification and hit-to-lead processes in drug discovery. Recently, the version of VS that is based upon similarity to a pharmacophore has received increased attention. This is due to two major factors: first, the public availability of the ZINC1 conformational database has provided a large selection pool with high-quality and purchasable small molecules; second, new technology has enabled a more accurate and flexible definition of pharmacophore models coupled with an efficient search speed. The major goal of this study was to achieve improved specificity and sensitivity of pharmacophore-based VS by optimizing the variables used to generate conformations of small molecules and those used to construct pharmacophore models from known inhibitors or from inhibitor-protein complex structures. By using human immunodeficiency virus protease and its inhibitors (PIs) as a case study, the impact of the key variables, including the selection of chemical features, involvement of excluded volumes (EV), the tolerance radius of excluded volumes, energy windows, and the maximum number of conformers in conformation generation, was explored. Protein flexibility was simulated by adjusting the sizes of EV. Our best pharmacophore model, combining both chemical features and excluded volumes, was able to correctly identify 60 out of 75 structurally diverse known PIs, while misclassifying only 5 out of 75 similar compounds that are not inhibitors. To evaluate the specificity of the model, 1193 oral drugs on the market were screened, and 25 original hits were identified, including 5 out of 6 known PI drugs.

Development of neural network QSPR models for Hansch substituent constants. 1. Method and validations.

In an attempt to develop predictive models for Hansch substituent constants for less common substituents, neural network QSPR (Quantitative Structure-Property Relationship) studies were conducted to correlate Hansch substituent constants for hundreds of chemically diverse functional groups with two different molecular descriptor sets. The Hansch substituent constants under study were pi, MR, F and R, describing the hydrophobic, steric/polarizability, and electronic (field and resonance) characteristics of the substituents, respectively. E-state descriptors were used for pi and MR, while the molecular descriptor set based upon the approach of Kvasnicka, Sklenak, and Pospichal (J. Am. Chem. Soc. 1993, 115, 1495-1500) was adopted for F and R. Both QSPR models demonstrated good predictivity in test sets.