Dual Orexin Receptor Antagonists (DORAs) bind to both the Orexin 1 and 2 receptors and there is one marketed DORA, Belsomra®, which treats insomnia. This presentation will highlight ligand-based computational methods that were used in the DORA lead identification and optimization process at Merck, ultimately resulting in this medicine. I will also highlight where experimental methods (e.g. xray, NMR) were used to validate the small molecule computational approaches. Additionally, given that Belsomra® does activate the PXR receptor, leading to potential drug-drug interactions (DDI), I will introduce a new metric hitherto unpublished, which we have successfully applied to multiple projects within Vertex. This metric computes the maximum interatomic distance from an ensemble of molecular conformations and does not require knowledge of any protein structure to rapidly predict PXR activation.

The sampling of smaller drug-like compounds has been an active area of research. However, few studies have tested in details the sampling of larger more flexible compounds, which are also relevant to drug discovery. Here, we investigate extensively mainstream conformational sampling methods on three carefully curated compound sets, namely the ‘Drug-like’, larger ‘Flexible’, and ‘Macrocycle’ compounds. These test molecules are chemically diverse with reliable X-ray protein-bound bioactive structures. The compared sampling methods include Stochastic Search and LowModeMD from MOE, all the low-mode based approaches from MacroModel, and MD/LLMOD developed for macrocycles. The performance of the computational protocols was assessed via (i) the reproduction of the X-ray bioactive structures, (ii) the size, coverage and diversity of the output conformational ensembles, (iii) the compactness/extendedness of the conformers, and (iv) the ability to locate the global energy minimum. In MOE, the recent LowModeMD emerged as the method of choice. Mixed torsional/low-mode from MacroModel performed as well as LowModeMD, and MD/LLMOD performed well for macrocycles. The low-mode based approaches yielded very encouraging results with the flexible and macrocycle sets. Thus, one can tackle larger flexible open-chain and macrocyclic compounds for drug discovery more productively with computational tools.

Targeted covalent inhibition is an established approach for increasing the potency and selectivity of potential drug candidates, and has also been used to identify tool compounds for target validation studies. It is evident that binding to reversible recognition elements within the pocket is essential for selective covalent inhibition, but this must also be achieved with the appropriate level of inherent reactivity of the covalent functionality and compatible geometry and reactivity of the target amino acid. We have assessed the ability of several experimental and computational approaches to predict the intrinsic reactivity of a range of cysteine targeting warheads, some of which may be used in prospective compound design. In addition, molecular dynamics (MD) simulations in combination with pKa prediction methods have been used to investigate the accessibility and potential reactivity of target cysteines towards covalent inhibition. As part of this work a number of case studies are reported including retrospective examples from covalent drug discovery efforts along with prospective use towards a novel target.

With the increasing size of data sets and the parallel development of multiple structural series in medicinal chemistry projects, managing and analyzing structure activity/property relationship data is becoming ever more challenging. Tools and methods for the efficient visualization, analysis and profiling of compounds therefore remain of deep interest. Here, we describe a new web-based application, MOEsaic, which enhances typical medicinal chemistry workflows aimed at interrogating the SAR/SPR data through the application of interactive matched molecular pairs (MMP) analysis and R-group profiling.

Dr. Graeme Robb
Chemistry, Oncology, IMED Biotech Unit, AstraZeneca, Cambridge, UK

Inhibition of the interaction between B-cell lymphoma 6 (BCL6) and co-repressors has been implicated as a therapeutic target in diffuse large B-cell lymphoma (DLBCL) cancers. Profiling of potent and selective BCL6 inhibitors are critical to test this hypothesis, however inhibition of a protein-protein interaction without prior small-molecule precedent is challenging. We used existing X-ray protein crystal structures to scan the co-repressor sequence for hot-spots and thus identify potential binding-sites for shape-based virtual screening. The results, together with a parallel fragment screen led to the identification of a series of pyrazolo[1,5-a]pyrimidines with weak binding affinity. Early crystallography was critical in enabling structure-based design. The understanding of key affinity interactions within the binding site and the modelling of the dynamic behaviour and free energy of ligand, pocket and solvent facilitated the identification of regions containing unstable solvent molecules, the displacement of which increased affinity 1000-fold. Inhibitor conformations were studied with NMR and we developed a predictive model using mixed Monte Carlo / molecular dynamics simulation, that gave good correlation with observed affinity. We subsequently designed macrocycles to restrict conformational freedom and favour the conformation required for binding, which gave a further 100-fold increase in potency. In an attempt to increase affinity further, we designed several analogues to irreversibly bind to a cysteine residue in the binding site. Despite demonstrating covalent bond formation, these compounds were less effective and we have been able to show, through long time-scale molecular dynamics, that the geometry of the cysteine is poorly available for Michael addition. The use of protein crystal structure, the measurement and prediction of ligand conformation and the use of modern structure-based design techniques has enabled rapid progress from fragment hit to potent BCL6 inhibitor.

It is without question that the successful execution of a drug discovery project is a collaborative effort, combining the specific skills and expertise of a number of different interested parties. What is the measure of success and how do we define ‘impact’? How can a computational chemist ‘influence’ the decision making process from the initial conception of an idea to compound synthesis?

This presentation will focus specifically on the interface between computational chemistry and medicinal chemistry exploring the different tools and methods employed to share and transfer knowledge for the best possible outcome on a project.

For years protein-protein interactions were considered as undruggable protein targets. However, during the last decade several successes in the design of small molecule protein protein interactions inhibitors (SMPPII) have proven this prejudice wrong. On the other side, the successes in the design of small molecule protein protein interaction stabilizers (SMPPIS) has remained very low.

During this presentation I will discuss the pharmacophore based workflows that were successful for the design of SMPPII targeting the HIV-1 integrase:LEDGF/p75 and the SUMO:SBM interaction.

Next, I will also discuss our pharmacophore based strategy that led to the successful design of SMPPIS for two different targets: a first class targets the salmonella PhoP:PhoQ two component signaling pathway and may be a promising new antibacterial compounds. The second class of compounds stabilize the interaction of the PKD2 protein kinase domain and a regulatory protein domain locking the complex in an inactivated conformation.

PI3Kdelta is a lipid kinase, primarily expressed in leukocytes, that plays a key role in immune cell signalling. It is activated across a broad range of cell types in response to multiple triggers relevant to respiratory disease, including allergens, smoke, viruses and inflammatory cytokines. We previously disclosed¹ nemiralisib (GSK2269557) and related PI3Kdelta selective inhibitors that were designed for inhaled delivery as potential therapeutic treatments to modulate the inflammatory processes leading to exacerbations associated with both asthma and COPD. Here we present hypotheses for the high PI3K isoform selectivity achieved for nemiralisib, discuss how modelling techniques and crystallography enhanced understanding and show how learnings were applied to subsequent chemical templates.

1: Down et. al, J. Med. Chem., 2015, 58 (18), pp 7381–7399.