Assessment of Reversibility for Covalent Cysteine Protease Inhibitors Using Quantum Mechanics/Molecular Mechanics Free Energy Surfaces.

We have used molecular dynamics (MD) simulations with hybrid quantum mechanics/molecular mechanics (QM/MM) potentials to investigate the reaction mechanism for covalent inhibition of cathepsin K and assess the reversibility of inhibition. The computed free energy profiles suggest that a nucleophilic attack by the catalytic cysteine on the inhibitor warhead and proton transfer from the catalytic histidine occur in a concerted manner. The results indicate that the reaction is more strongly exergonic for the alkyne-based inhibitors, which bind irreversibly to cathepsin K, than for the nitrile-based inhibitor odanacatib, which binds reversibly. Gas-phase energies were also calculated for the addition of methanethiol to structural prototypes for a number of warheads of interest in cysteine protease inhibitor design in order to assess electrophilicity. The approaches presented in this study are particularly applicable to assessment of novel warheads, and computed transition state geometries can be incorporated into molecular models for covalent docking.

Crystal structure of Leishmania mexicana cysteine protease B in complex with a high-affinity azadipeptide nitrile inhibitor.

Leishmania mexicana is an obligate intracellular protozoan parasite that causes the cutaneous form of leishmaniasis affecting South America and Mexico. The cysteine protease LmCPB is essential for the virulence of the parasite and therefore, it is an appealing target for antiparasitic therapy. A library of nitrile-based cysteine protease inhibitors was screened against LmCPB to develop a treatment of cutaneous leishmaniasis. Several compounds are sufficiently high-affinity LmCPB inhibitors to serve both as starting points for drug discovery projects and as probes for target validation. A 1.4 Å X ray crystal structure, the first to be reported for LmCPB, was determined for the complex of this enzyme covalently bound to an azadipeptide nitrile ligand. Mapping the structure-activity relationships for LmCPB inhibition revealed superadditive effects for two pairs of structural transformations. Therefore, this work advances our understanding of azadipeptidyl and dipeptidyl nitrile structure-activity relationships for LmCPB structure-based inhibitor design. We also tested the same series of inhibitors on related cysteine proteases cathepsin L and Trypanosoma cruzi cruzain. The modulation of these mammalian and protozoan proteases represents a new framework for targeting papain-like cysteine proteases.

Synthesis and structure-activity relationship of nitrile-based cruzain inhibitors incorporating a trifluoroethylamine-based P2 amide replacement.

The structure-activity relationship for nitrile-based cruzain inhibitors incorporating a P2 amide replacement based on trifluoroethylamine was explored by deconstruction of a published series of inhibitors. It was demonstrated that the P3 biphenyl substituent present in the published inhibitor structures could be truncated to phenyl with only a small loss of affinity. The effects of inverting the configuration of the P2 amide replacement and linking a benzyl substituent at P1 were observed to be strongly nonadditive. We show that plotting affinity against molecular size provides a means to visualize both the molecular size efficiency of structural transformations and the nonadditivity in the structure-activity relationship. We also show how the relationship between affinity and lipophilicity, measured by high-performance liquid chromatography with an immobilized artificial membrane stationary phase, may be used to normalize affinity with respect to lipophilicity.

A comparative study of warheads for design of cysteine protease inhibitors.

The effects on potency of cruzain inhibition of replacing a nitrile group with alternative warheads were explored. The oxime was almost an order of magnitude more potent than the corresponding nitrile and has the potential to provide access to the prime side of the catalytic site. Dipeptide aldehydes and azadipeptide nitriles were found to be two orders of magnitude more potent cruzain inhibitors than the corresponding dipeptide nitriles although potency differences were modulated by substitution at P1 and P3. Replacement of the α methylene of a dipeptide aldehyde with cyclopropane led to a loss of potency of almost three orders of magnitude. The vinyl esters and amides that were characterized as reversible inhibitors were less potent than the corresponding nitrile by between one and two orders of magnitude.

Comment on The Ecstasy and Agony of Assay Interference Compounds.

A recent editorial (Aldrich et al. The Ecstasy and Agony of Assay Interference Compounds . J. Chem. Inf. MODEL: 2017 , 57 , 387 - 390 ) is examined critically. When assessing assay hits from screening, it is important to draw a distinction between false positives, that have no effect on target function, and compounds that affect target function through an undesirable mechanism of action. Observation of frequent-hitter behavior for a compound should be regarded as circumstantial evidence, rather than definitive proof, that the compound has interfered with assay readouts or acted through an undesirable mechanism of action. The applicability domain of published (Baell and Holloway J. Med. Chem. 2010 , 53 , 2719 - 2740 ) Pan Assay INterference compoundS (PAINS) filters is limited by the narrow scope of the proprietary data used to derive them. It is suggested that journal guidelines for authors should not prescribe, as those for the Journal of Medicinal Chemistry appear to do, that activity in assays reported for compounds that match PAINS filters be treated any differently from that for compounds that do not match PAINS filters. It is argued that use of models based on proprietary data in the evaluation of manuscripts would contradict the editorial policy of any journal that deemed the use of proprietary data to be unacceptable in modeling studies.

The influence of hydrogen bonding on partition coefficients.

This Perspective explores how consideration of hydrogen bonding can be used to both predict and better understand partition coefficients. It is shown how polarity of both compounds and substructures can be estimated from measured alkane/water partition coefficients. When polarity is defined in this manner, hydrogen bond donors are typically less polar than hydrogen bond acceptors. Analysis of alkane/water partition coefficients in conjunction with molecular electrostatic potential calculations suggests that aromatic chloro substituents may be less lipophilic than is generally believed and that some of the effect of chloro-substitution stems from making the aromatic π-cloud less available to hydrogen bond donors. Relationships between polarity and calculated hydrogen bond basicity are derived for aromatic nitrogen and carbonyl oxygen. Aligned hydrogen bond acceptors appear to present special challenges for prediction of alkane/water partition coefficients and this may reflect 'frustration' of solvation resulting from overlapping hydration spheres. It is also shown how calculated hydrogen bond basicity can be used to model the effect of aromatic aza-substitution on octanol/water partition coefficients.

Nitrate in the active site of protein tyrosine phosphatase 1B is a putative mimetic of the transition state.

The X-ray crystal structure of the complex of protein tyrosine phosphatase 1B with nitrate anion has been determined and modelled quantum-mechanically. Two protomers were present in the structure, one with the mechanistically important WPD loop closed and the other with this loop open. Nitrate was observed bound to each protomer, making close contacts with the S atom of the catalytic cysteine and a tyrosine residue from a crystallographically related protomer.

Ligand efficiency metrics considered harmful.

Ligand efficiency metrics are used in drug discovery to normalize biological activity or affinity with respect to physicochemical properties such as lipophilicity and molecular size. This Perspective provides an overview of ligand efficiency metrics and summarizes thermodynamics of protein-ligand binding. Different classes of ligand efficiency metric are critically examined and the study concludes with suggestions for alternative ways to account for physicochemical properties when prioritizing and optimizing leads.

Inflation of correlation in the pursuit of drug-likeness.

Drug-likeness is a frequently invoked, although not always precisely defined, concept in drug discovery. Opinions on drug-likeness are to a large extent shaped by the relationships that are observed between surrogate measures of drug-likeness (e.g. aqueous solubility; permeability; pharmacological promiscuity) and fundamental physicochemical properties (e.g. lipophilicity; molecular size). This article draws on examples from the literature to highlight approaches to data analysis that exaggerate trends in data and the term correlation inflation is introduced in the context of drug discovery. Averaging groups of data points prior to analysis is a common cause of correlation inflation and results from analysis of binned continuous data should always be treated with caution.

ClogP(alk): a method for predicting alkane/water partition coefficient.

Alkane/water partition coefficients (P(alk)) are less familiar to the molecular design community than their 1-octanol/water equivalents and access to both data and prediction tools is much more limited. A method for predicting alkane/water partition coefficient from molecular structure is introduced. The basis for the ClogP(alk) model is the strong (R² = 0.987) relationship between alkane/water partition coefficient and molecular surface area (MSA) that was observed for saturated hydrocarbons. The model treats a molecule as a perturbation of a saturated hydrocarbon molecule with the same MSA and uses increments defined for functional groups to quantify the extent to which logP(alk) is perturbed by the introduction each functional group. Interactions between functional groups, such as intramolecular hydrogen bonds are also parameterized within a perturbation framework. The functional groups and interactions between them are specified substructurally in a transparent and reproducible manner using SMARTS notation. The ClogP(alk) model was parameterized using data measured for structurally prototypical compounds that dominate the literature on alkane/water partition coefficients and then validated using an external test set of 100 alkane/water logP measurements, the majority of which were for drugs.

Automated molecule editing in molecular design.

The ability to modify chemical structures in an automated and controlled manner is useful in molecular design. This Perspective introduces the MUDO molecule editor and shows how automated molecule editing can be used to standardize structures, enumerate tautomeric and ionization states, identify matched molecular pairs. Unlike its predecessor Leatherface, MUDO can also process 3D structures and this capability can be used to link non-covalently docked ligands to proteins.

Hydrogen-Bond Donors in Drug Design.

Hydrogen-bond donors are seen to cause more problems for drug designers than hydrogen-bond acceptors. Most of the polarity in drug-like compounds comes from hydrogen-bond acceptors since they typically exceed the hydrogen-bond donors in number and are more heavily solvated on an individual basis. The implications of this polarity imbalance for optimization of permeability and aqueous solubility are discussed. A factor that should be considered in optimization of ligand recognition by targets is that the presence of a hydrogen-bond donor generally implies that a hydrogen-bond acceptor is also present (but not vice versa). Frustrated solvation and secondary electrostatic interactions result from aligned hydrogen-bond donors and acceptors, and the design opportunities presented by these phenomena are discussed. Hydrogen-bond donors based on oxygen, nitrogen and carbon are compared as target recognition elements, and halogen- and chalcogen-bond donors are discussed as hydrogen-bond donor equivalents.

Discovery of novel imidazo[1,2-a]pyridines as inhibitors of the insulin-like growth factor-1 receptor tyrosine kinase.

We disclose a novel series of insulin-like growth factor-1 receptor kinase inhibitors based on the 3-(pyrimidin-4-yl)-imidazo[1,2-a]pyridine scaffold. The influence on the inhibitory activity of substitution on the imidazopyridine and at the C5 position of the pyrimidine is discussed. In the course of this optimization, we discovered a potent and selective inhibitor with suitable pharmacokinetics for oral administration.

Discovery of a series of indan carboxylic acid glycogen phosphorylase inhibitors.

A series of carboxylic acid glycogen phosphorylase inhibitors, which have potential as oral antidiabetic agents, is described. Defining and applying simple physicochemical design criteria was used to assess the opportunity and to focus synthetic efforts on compounds with the greatest probability of success. The study led to compound 17, which exhibits a good balance of properties including potent inhibition of recombinant human liver glycogen phosphorylase in vitro, a good DMPK profile including excellent bioavailability and low clearance and good in vivo activity in a glucagon challenge model of diabetes in Zucker rats.

Matched molecular pair analysis of activity and properties of glycogen phosphorylase inhibitors.

Matched molecular pair analysis has been used in design of inhibitors of glycogen phosphorylase.

5-Aminopyrimidin-2-ylnitriles as cathepsin K inhibitors.

A series of pyrimidine nitrile inhibitors of Cathepsin K with reduced glutathione reactivity has been identified and Molecular Core Matching (MoCoM) has been used to quantify the effect of an amino substituent at C5.

Design of selective Cathepsin inhibitors.

A number of molecular recognition features have been exploited in structure-based design of selective Cathepsin inhibitors.

Dipeptidyl nitrile inhibitors of Cathepsin L.

A series of potent Cathepsin L inhibitors with good selectivity with respect to other cysteine Cathepsins is described and SAR is discussed with reference to the crystal structure of a protein-ligand complex.

Design of compound libraries for fragment screening.

Approaches to the design of libraries for fragment screening are illustrated with reference to a 20 k generic fragment screening library and a 1.2 k generic NMR screening library. Tools and methods for library design that have been developed within AstraZeneca are described, including Foyfi fingerprints and the Flush program for neighborhood characterization. It will be shown how Flush and the BigPicker, which selects maximally diverse sets of compounds, are used to apply the Core and Layer method for library design. Approaches to partitioning libraries into cocktails are also described.

The nature of ligand efficiency.

Ligand efficiency is a widely used design parameter in drug discovery. It is calculated by scaling affinity by molecular size and has a nontrivial dependency on the concentration unit used to express affinity that stems from the inability of the logarithm function to take dimensioned arguments. Consequently, perception of efficiency varies with the choice of concentration unit and it is argued that the ligand efficiency metric is not physically meaningful nor should it be considered to be a metric. The dependence of ligand efficiency on the concentration unit can be eliminated by defining efficiency in terms of sensitivity of affinity to molecular size and this is illustrated with reference to fragment-to-lead optimizations. Group efficiency and fit quality are also examined in detail from a physicochemical perspective. The importance of examining relationships between affinity and molecular size directly is stressed throughout this study and an alternative to ligand efficiency for normalization of affinity with respect to molecular size is presented.