MICA: Design, synthesis, and pharmacological profiling of agonists of the human orexin receptors

Structure-based design of G-protein-coupled receptor (GPCR) ligands has recently become possible as experimental 3D structures of several GPCRs are now available. However, the design of small-molecule agonists (activators) rather than antagonists (blockers) of such receptors, and especially of peptide-activated GPCRs, remains challenging. Many of these receptors are of great interest as potential drug targets, and in many cases good chemical biology tools are missing.

Together with our industrial partner Heptares, a pharmaceutical company with considerable expertise in rational GPCR ligand design, we will develop drug-like agonists for the orexin receptors (OXRs), whose cognate ligands are orexin peptides, using an integrated approach based on our combined expertise in pharmacology, structural biology, and medicinal chemistry. We chose the OXR system because of the important medical potential of oral OXR agonists in narcolepsy, obesity, and hypophagia, as well as attention deficit hyperactivity disorder, bipolar disorders, Parkinson's disease, and colon cancer. Furthermore, non-peptide OXR agonists are currently also missing as permeable tool compounds to elucidate poorly understood OXR biology. The OXR system provides an ideal test bed for the structure-based design of peptide-activated GPCR agonists due to the fact that we have at our disposal an unprecedented structural understanding and experimental tools, including for the first time OXR crystal structures and a full panel of OXR mutants for every residue in the ligand binding region.

We will use three strategies for the structure-based design of OXR agonists. We expect that the most important of these in terms of providing medicinal chemistry starting points will be virtual in silico screening of large databases of drug-like and commercially available compounds against 3D models of the active, agonist-form of the OXRs, which will be derived from the experimental structures of the antagonist-forms. Alternative strategies, at least one of which will also be explored, depending on the success of the virtual screening approach, are peptidomimetic conversion of the OX peptides into permeable compounds, and structure-based redesign of small-molecule OXR antagonists, many of which are known, including compounds currently under clinical evaluation.

Optimisation of hit compounds will be carried out using our established medicinal chemistry approaches that we have used in other GPCR-targeted chemical biology projects. These strategies will be aimed at establishing structure-activity relationships with respect to OXR affinity, potency, agonism versus antagonism activity, and physicochemical properties known to govern bioavailability. Importantly, however, here compound optimisation will be underpinned by a structural understanding of how compounds bind to the OXRs. This understanding will be established through a combination of molecular modelling, biophysical analysis of the interaction of test compounds with members of the panel of OXRs mutants, and, if feasible, X-ray crystallography.

We have several OXR assays already available and will develop a full screening cascade to assess the pharmacological activity and mode of action of active compounds from the various design strands. The primary screens will be an agonism-sensitive reporter gene assay using mammalian cells expressing human OXRs and containing a reporter gene whose product can be measured, as well as an assay measuring activation of a relevant cellular pathway downstream of OXRs. Secondary assays will include a range of functional assays to assess signal transduction mode and efficiency. Finally, we will assess promising lead compounds for brain bioavailability and OXR activity using a rat telemetry model in which circadian variation in core temperature, blood pressure, heart rate, and locomotor activity, all of which are associated with OXR activation, will be observed.

Several crystal structures of ligand-bound GPCRs have been solved but the structural basis of receptor agonism remains poorly understood, especially for peptide-activated GPCRs, for which drug-like agonists would be therapeutically desirable. We have chosen the two orexin receptors (OXRs), for which our industrial partner Heptares has generated for the first time crystal structures of ligand complexes, in order to develop strategies for the structure-based design of small-molecule agonists of peptide-activated GPCRs. OXR agonists are relevant in narcolepsy, Parkinson's disease, cancer, and other indications. We will use a combination of homology modelling and torsional space interrogation techniques to generate agonist-form OXR models. These will form the basis for virtual screening of large compound databases. Predicted hits from this strategy, as well as alternative peptidomimetic and structure-based antagonist-to-agonist conversion design strands, will be assessed for OXR agonist activity in a primary screen using highly-coupled functional and reporter gene assays. Confirmed actives will be subjected to an extensive secondary assay battery comprising functional assays capable of distinguishing between different coupling pathways and potential biased signalling. Confirmed hit series will form starting points for structural elaboration and optimisation in terms of affinity, potency, selectivity, mode of agonist action, and bioavailability through medicinal chemistry. Structural hypotheses for compound design and optimisation will be greatly assisted by probing an existing set of receptor variants, in which all the residues in the ligand-binding domain have been mutated systematically, with the compounds we design and make. Once new small-molecule agonists with sufficient potency have been identified, these will be examined for CNS bioavailability and in vivo pharmacological activity as OXR agonists in a rat telemetry model.

This project builds upon recent advances in X-ray crystallographic structures of GPCRs to generate understanding of the molecular interactions of cognate peptide agonists and small-molecule ligands as they bind into their receptors. It uses the orexin OX1R and OX2R receptors, implicated as potential therapeutic targets in sleep disorders including narcolepsy, eating disorders including obesity and hypophagia, as well as Parkinson's disease, and colorectal cancer, as a model system to extend knowledge of ligand-receptor interactions. In silico modelling of orexin receptors and their interactions with peptide agonists orexin-A and -B, and small-molecule antagonists is used as the starting point for rational small-molecule agonist compound screening and drug design. The outcome of this research will be small-molecule agonists of OX1R and OX2R with good oral and CNS bioavailability. These compounds will benefit the research community, the pharmaceutical industry, the healthcare profession, and patients and their families in the following ways:

Researchers
At present, there are no examples of rationally-designed CNS-bioavailable small-molecule agonists of neuropeptide-activated GPCRs. This project will use novel in silico structure-based design, coupled with traditional synthetic and medicinal chemistry and supported by in vitro and in vivo pharmacology, to provide the first examples of such small-molecule peptide receptor agonists. The principles developed by this project should be applicable to a whole range of peptide-activated GPCRs, providing researchers worldwide with a tractable approach to develop synthetic small-molecule agonist therapeutics for other disease-associated peptide-activated receptors. Furthermore, the postdoctoral researcher on the project, as well as postgraduate researchers undertaking similarly aligned projects, will benefit from the multidisciplinary approach and training this project offers.

Pharmaceutical industry
The availability of small-molecule agonists at orexin receptors at the end of this project (and/or following subsequent preclinical development) provides an in-licensing opportunity for the pharmaceutical sector to bolster its failing, high-risk, early phase drug discovery pipeline with a mature, de-risked lead compound for the rare narcolepsy therapeutic indication. However, the therapeutic potential of such compounds extends beyond narcolepsy to treatment of obesity and hypophagia, as well as attention deficit hyperactivity disorder, and depression, and related bipolar disorders, and the even larger indications of Parkinson's disease and certain cancers. The potential benefit to the pharmaceutical partner is therefore considerable.

Healthcare profession, patients and families
Narcolepsy is a debilitating sleep disorder characterised by excessive daytime sleepiness (EDS), loss of muscle tone in response to emotional stimuli (cataplexy), and intrusive rapid eye movement (REM) sleep. Current drug treatment with modafinil and methylphenidate aim to treat only the sleepiness associated with narcolepsy and have no effect on other debilitating symptoms, e.g. cataplexy, sleep paralysis, and frequent nocturnal wakening. Overwhelming evidence suggests that pharmacological restoration of orexin receptor function would overcome these deficiencies in narcolepsy, offering far better therapy. The small-molecule therapeutics that will progress through pre-clinical and clinical development over a 7-12 year timeframe, will overcome the poor bioavailability and high cost of goods of orexin peptides, to provide a cost-effective and well-tolerated new drug of real benefit to patients.

Conclusion
The outcome of this project will provide significant impact in terms of scientific novelty and understanding, as well as providing valuable starting points and tools for further small molecule peptide GPCR agonist drug development in other areas of significant unmet medical need.