Blog | October 18, 2021

Mechanistic Understanding of a Difficult GPCR Aided by DEL and Medicinal Chemistry

By Shawn Johnstone, Director of Chemistry

Protease-activated receptor-2 ligands reveal orthosteric and allosteric mechanisms of receptor inhibition

GPCRs are important targets to investigate, not just for their high therapeutic potential but also mechanistically for their complex downstream signaling and diverse modes of modulation. PAR2 is a particularly interesting GPCR because its activation is unusual. After an extracellular protease cleaves PAR2’s N-terminus, the truncated peptide retracts, binds and activates itself. PAR2 also occupies an important historical place in our developing understanding of the relationship between biased signaling and therapeutic outcome. This Nature Communication research article highlights several of these exciting mechanistic phenomena and then describes how modern drug discovery tools can be used to define and detail complex GPCR-ligand interactions, pointing to new directions to control therapeutic outcome.

The previous X-ray crystallography of PAR2 combined with newly identified ligands revealed distinct mechanisms of inhibition

Prior to the report, the exact location of PAR2’s orthosteric site was unknown. A team of experts from Heptares and AstraZeneca crystallized PAR2 with three new ligands and identified two separate binding sites but did not definitively identify the orthosteric site. In this report, the Heptares and AstraZeneca team used a radioligand displacement assay, Schild analysis (with endogenous agonist and increasing concentrations of ligand) and binding kinetics from SPR experiments to elucidate mechanisms of antagonism. One PAR2 ligand, AZ8838, blocked endogenous agonist binding in SPR and right shifted concentration-response curves without changes in maximum response (Schild slope of 1), indicating that the binding site for AZ8838 (identified by crystallography) and PAR2’s orthosteric site overlap.

An advantage of DNA libraries in the small molecule drug discovery process is greater exploration of chemical space for identifying unrelated novel candidates with distinct mechanisms

A second ligand, AZ3451, discovered by X-Chem from a DEL screen on PAR2, bound PAR2 non-competitively with saturable antagonism and a Schild slope <1, exhibiting a completely different chemical profile than AZ8838. “Probe dependency,” in which allosteric and orthosteric ligands reciprocally influence potency, signal bias and binding kinetics, was evident from binding, cellular and biophysical assays. This included an unusual probe-dependent, biphasic response and an increased k_on in the presence of an agonist. Allosteric cooperativity was quantified by fitting Schild data to the Operational Allosteric Model. The allosteric cooperativity of AZ3451 leaned more heavily toward affinity than efficacy. As the probe changed, cooperativity changed but remained affinity weighted. Combined, these data firmly establish that AZ3451 is a negative allosteric modulator (NAM) of PAR2.

A successful demonstration of DEL complementing the drug discovery process

Both ligands were effective in an in vivo anti-inflammatory model, despite their differences in mechanisms of antagonism. They reduced PAR2 agonist-induced paw inflammation at 10 mg/kg doses and inhibited mast cell and neutrophil activation and degranulation in a rat model.

In summary, two distinct mechanisms of PAR2 inhibition were identified, revealing different methods of potential modulation. PAR2 modulation will continue to evolve as a therapeutic strategy to treat pain and inflammation, and this Nature Communication article exemplifies methods and strategies that can be used to investigate the effects of signal bias, binding kinetics and allosteric cooperativity on therapeutic outcome.

Back to Blog