Protease-activated receptors (PARs) are a unique family of G protein-coupled receptors with an unusual activation mechanism. Endogenously, the N terminus of these receptors is cleaved by extracellular proteases to reveal a tethered ligand sequence which can intramolecularly bind and activate the receptor. Protease-activated receptor-2 (PAR2) is predominately activated by serine proteases, such as trypsin and tryptase, which cleave between R 35–S 36 thereby exposing a new N-terminal sequence SLIGKV–, that acts as a tethered agonist for intramolecular PAR2 activation. The activated receptor stimulates multiple different G protein-dependent and independent signalling pathways. PAR2 has been shown to signal through Gα proteins (G q/11, G s and G 12/13) across different cell environments, as well as through G protein-independent proteins including β-arrestins 1/2. In addition, synthetic peptide agonists corresponding to the tethered ligand sequence, either human SLIGKV-NH 2 or the rodent sequence SLIGRL-NH 2, can activate these signalling pathways via PAR2.
PAR2 has been shown to have roles in pain and migraine, cancer, metabolic disease, both obesity and cardiovascular, as well as in inflammation and inflammatory diseases. The wide range of effects of PAR2 underscores the importance of this receptor in human physiology and disease. As a consequence, this target is highly sought after in drug discovery and for a number of years has been a focus of major pharmaceutical endeavours. However, as yet the discovery of an effective drug has proved challenging for PAR2. The only marketed drug for a PAR is vorapaxar, a selective antagonist of PAR1 that is an antiplatelet treatment for improving restricted blood flow. Even though PAR2 has close sequence homology with PAR1, the discovery of PAR2 antagonists has been less successful, with weak antagonists that only inhibit selected signalling pathways (e.g. GB88), or that show agonist properties in some cell types (e.g. C391). Development of tool compounds is important for better understanding of the mechanisms of PAR2 activation on different cell and tissue types and in diseases where PAR2 is a key mediator and potential therapeutic target.
In 2017, we reported the crystal structures of PAR2 bound to AZ8838 and AZ3451. AZ8838 was derived from an initial high-throughput screen hit and AZ3451 was obtained from a DNA-encoded library technology screen, which also found an agonist, ‘compound 1’. AZ8838 binds in an occluded pocket made up of transmembrane helices (TM) 1–3, 7 and extracellular loop 2 (ECL2), whereas AZ3451 occupies a pocket that faces the lipid bilayer and is formed by TM 2, 3 and 4. However, the definition of the orthosteric site of the tethered ligand remains elusive due to the lack of an agonist-bound PAR2 crystal structure. More recently, with the AZ8838-bound structure as a starting point, we used an extensive combinatorial approach of site-directed mutagenesis and computational modelling to propose the putative orthosteric site.
Here, we present the pharmacological characterisation of a novel agonist of PAR2 (AZ2429) and two novel antagonists (AZ8838 and AZ3451). Agonist AZ2429 is proposed to bind at the same site in PAR2 as the activating peptide SLIGKV-NH 2 but is a more potent activator of multiple PAR2-dependent signalling pathways. Antagonists AZ8838 and AZ3451 can inhibit both G protein-dependent and independent pathways via PAR2 in vitro and exert anti-inflammatory effects in vivo in a rat model of PAR2 agonist-induced paw oedema. Our approach to interrogate ligand-receptor interactions, using molecular modelling and surface plasmon resonance in combination with functional assays, characterises AZ8838 as a competitive antagonist, whilst AZ3451 is a negative allosteric modulator. Our findings highlight coupling between ligand binding sites in the PAR2 receptor and illustrate opportunities for both orthosteric and allosteric inhibitors of PAR2 functions in vitro and in vivo.
Using the stabilised (StaR) PAR2 receptor, two novel antagonist series were discovered. For the first series, the initial hit compound of the benzimidazole series, AZ8935, was identified in a DNA-encoded library technology screen. Chemical expansion of this series was carried out to provide analogues with improved potency across different PAR2-mediated signalling pathways. Substitution at the para-position of the phenyl ring with a nitrile group (AZ3451) led to the biggest increase in potency over the unsubstituted analogue AZ8935. Heterocycles with a heteroatom in the para-position (AZ2623) also had a gain in potency relative to AZ8935. Modification of the cyclohexyl ring to a tetrahydropyran (AZ7126) was tolerated with only a modest decrease in potency in Ca 2+ and IP1 (inositol-1-phosphate)signalling, relative to the most potent analogue, AZ3451. The imidazole series was developed from an initial weak high-throughput screen hit. Optimisation resulted in compound AZ8838, although a constrained heterocycle (AZ0107) also maintained similar potency.
