Protease-Activated Receptors: The Receptor That Carries Its Own Ligand

Receptor pharmacology

Protease-activated receptors break the rule that every other peptide receptor obeys. There is no circulating peptide hormone that switches them on. The peptide agonist is already part of the receptor, folded into its own N-terminal tail, and it stays inactive until an enzyme cuts the tail and exposes it. Understanding that single structural quirk explains why thrombin can act like a hormone, why one drug in this class carries a boxed warning, and why almost every mouse experiment on the topic pointed the wrong way.

A receptor that is its own agonist

The discovery is precisely dated. In 1991, Vu, Hung, Wheaton and Coughlin cloned the thrombin receptor and found something nobody had described before: a thrombin cleavage site 41 amino acids from the receptor's start methionine. A synthetic peptide mimicking the new N-terminus created by that cut was a potent agonist on its own. Mutant receptors that could not be cleaved failed to respond to thrombin, but responded perfectly to the free peptide.

The conclusion, in the paper's own words, was “a novel signaling mechanism in which thrombin cleaves its receptor's amino-terminal extension to create a new receptor amino terminus that functions as a tethered ligand and activates the receptor.”

Three consequences follow immediately. Activation is irreversible at the level of the individual receptor — you cannot un-cut a protein, so signal termination has to happen by internalisation and degradation rather than by the ligand letting go. Activation is proportional to enzyme activity rather than to hormone concentration, which makes PARs sensors of proteolysis: coagulation, inflammation, tissue injury, allergen exposure. And, as we will see, activation is very hard to block with a conventional drug.

Four protease-activated receptors, one mechanism

Humans have four. PAR1, PAR2 and PAR3 sit together on chromosome 5q13.3; PAR4 is elsewhere entirely, on 19p13.11, so the familiar phrase “the PAR gene cluster” only really applies to the first three. All four are class A G protein-coupled receptors.

PAR3 is the odd one out and deserves care, because three separate claims about it often get flattened into one. Human PAR3 does confer thrombin responsiveness when expressed in a test system. The synthetic TFRGAP peptide, unlike the other three, is not an agonist. And in the mouse, PAR3 does not signal at all — it functions as a cofactor that presents thrombin to PAR4, which the authors described as a G protein-coupled receptor acting as an accessory molecule for another receptor.

Where the enzyme cuts changes what the cell does

The most interesting property of this system is that a single receptor can be cut in more than one place, and the resulting tethered ligands are not equivalent. PAR1 has two coagulation-protease cleavage sites: Arg41 and Arg46.

Thrombin strongly prefers Arg41, partly because a hirudin-like sequence just downstream of that bond docks onto thrombin's exosite, and partly because the residue two positions before Arg41 is a proline, which thrombin likes. Cutting there produces pro-inflammatory endothelial signalling. Cutting at Arg46 instead produces a different tethered ligand and cytoprotective, anti-inflammatory signalling — but neither activated protein C nor thrombin can reach Arg46 without a co-receptor first holding it in place (the endothelial protein C receptor for APC, thrombomodulin for thrombin). That is a form of signalling bias arriving by an unusual route: the difference is generated by where the scissors land, not by how a ligand fits the pocket — a useful contrast with the ordinary agonist and antagonist vocabulary.

Some proteases do the opposite. Neutrophil elastase, cathepsin G and chymotrypsin can chop the exodomain without revealing a functional tethered ligand, which leaves a receptor that can no longer be activated by thrombin but still responds to a synthetic agonist peptide. On PAR1, cleavage at Phe55–Trp56 by cathepsin G is documented as inhibitory while cleavage at Arg41 by the same enzyme is activating — a good reminder that “protease X activates PAR Y” is usually a statement about a particular concentration in a particular tissue.

TRAP peptides, and why antagonists are so difficult

Because the tethered ligand works as a free peptide too, synthetic agonists were available from the very first paper. SFLLRN — usually as its C-terminal amide, and known in the literature as TRAP-6 or thrombin receptor activating peptide — became the standard way to poke PAR1 without adding thrombin and setting off the whole clotting cascade. It is written into the approved US labelling of the one marketed PAR1 drug as the pharmacodynamic assay.

Antagonism is a different matter. A conventional competitive antagonist wins by mass action: flood the system and the antagonist occupies the site more often than the agonist. That argument collapses when the agonist is covalently attached to the receptor and never diffuses away. The 2.2 Ångström crystal structure of human PAR1 bound to vorapaxar made the point structurally: the drug's binding pocket is shallow but almost entirely shielded from solvent, and the authors describe “an unusual mode of drug binding that explains how a small molecule binds virtually irreversibly to inhibit receptor activation by the tethered ligand.” To beat a ligand that never leaves, the blocker has to be one that effectively never leaves either.

A second strategy attacks from inside. Pepducins are lipidated peptides modelled on a receptor's own intracellular loops; the lipid anchors them in the membrane and flips them to the cytoplasmic face, where they interfere with the receptor–G protein interface rather than the ligand pocket. PZ-128, a PAR1 pepducin, reached a phase 2 trial in patients undergoing cardiac catheterisation: 100 patients, no significant difference in bleeding, no significant difference in major adverse coronary events, and an exploratory signal in the subgroup with raised baseline troponin. It remains an interesting proof of concept rather than a treatment.

Vorapaxar: what a PAR1 blocker actually did

Vorapaxar (Zontivity) is the only PAR1 antagonist ever approved. The FDA approved it on 8 May 2014 to reduce thrombotic cardiovascular events in patients with a history of myocardial infarction or peripheral arterial disease. Its label describes it as a reversible antagonist “but its long half-life makes it effectively irreversible” — an effective half-life of three to four days, an apparent terminal half-life around eight days, and significant platelet inhibition still present four weeks after the last dose.

The two big trials tell the whole story of the class.

• TRACER tested vorapaxar on top of standard therapy in 12,944 patients with acute coronary syndrome. Follow-up was terminated early after a safety review. The primary composite endpoint missed significance. Moderate and severe bleeding rose from 5.2% to 7.2%, and intracranial haemorrhage from 0.2% to 1.1% — a hazard ratio of 3.39.
• TRA 2°P–TIMI 50 tested secondary prevention in 26,449 patients. Here it worked: cardiovascular death, myocardial infarction or stroke fell from 10.5% to 9.3% at three years. But after two years the data and safety monitoring board recommended stopping treatment in patients with a history of stroke because of intracranial haemorrhage, and overall moderate or severe bleeding rose from 2.5% to 4.2%.

Atopaxar, a reversible PAR1 antagonist from a different company, completed a phase 2 programme showing reduced ischaemia on Holter monitoring without a significant bleeding excess, but with dose-dependent transaminase elevation and QTc prolongation at the highest doses. No phase 3 study of it was ever registered.

The species trap

Anyone reading the preclinical literature on this class needs to know one thing before anything else.

In humans, PAR1 and PAR4 together account for essentially all thrombin signalling in platelets: blocking either alone leaves a substantial response at high thrombin concentrations, while blocking both virtually abolishes secretion and aggregation. In mice, PAR1-activating peptides do not activate platelets at all, and PAR1-deficient mice have normal platelet responses to thrombin. A drug screened in mice for antiplatelet activity through PAR1 would look like a complete failure. This is one of the cleanest published examples of why animal models do not transfer automatically in peptide and receptor pharmacology.

Beyond platelets

PAR2 is the receptor drawing most current interest, largely outside coagulation. It is activated by trypsin, mast cell tryptase, matriptase and neutrophil elastase, and it links proteolysis to pain, itch and inflammation — partly through release of substance P and sensitisation of TRPV1. PAR2-deficient mice are protected against allergic sensitisation to house dust mite and cockroach allergens, which is not a coincidence: several common allergens are themselves proteases. Structural work has produced PAR2 complexes with both orthosteric and negative allosteric modulators, and one pH-dependent monoclonal antibody against PAR2 is in a phase 2 migraine trial. Nothing in this class is approved.

PAR1 also has roles well away from platelets: endothelial barrier function, cytokine induction, and the cytoprotective arm of the protein C pathway. That last one is why the Arg41-versus-Arg46 distinction is more than a curiosity — the same receptor sits on both sides of the inflammation ledger, and which side it takes depends on which enzyme, with which co-receptor, got to it first.

Frequently asked questions

Is a PAR a peptide receptor?

Yes, but an unusual one. It is a class A G protein-coupled receptor activated by a peptide agonist, exactly like the receptors for many hormones — the difference is that the peptide is manufactured as part of the receptor rather than delivered to it.

Why can PARs not be switched off by an ordinary antagonist?

Because competitive antagonism relies on the agonist being free to dissociate, and a tethered ligand never dissociates. The only marketed PAR1 blocker works by binding a shielded pocket so tightly and for so long that it is functionally irreversible — which is also why its bleeding risk cannot be reversed by stopping the drug.

What is TRAP-6?

A synthetic hexapeptide, SFLLRN, matching the first six residues of the tethered ligand unmasked when thrombin cuts PAR1. It activates PAR1 without any proteolysis, which makes it a standard laboratory probe and a pharmacodynamic assay in clinical trials. It is a research reagent, not a therapeutic.

Do PARs have anything to do with the peptides discussed elsewhere on this site?

Only structurally. PARs are receptors, not administered compounds, and no PAR agonist or antagonist is sold as a research peptide in the way the compounds in our library are. The reason they are worth knowing about is conceptual: they are the clearest demonstration that a peptide sequence can be a signal without ever existing as a separate molecule.

References

1. Vu TKH, Hung DT, Wheaton VI, Coughlin SR. Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell 1991;64:1057–1068. pubmed.ncbi.nlm.nih.gov
2. IUPHAR/BPS Guide to Pharmacology. Proteinase-activated receptors (family 59) — tethered ligand sequences, agonist proteases and disarming cleavage. guidetopharmacology.org
3. Zhang C, Srinivasan Y, Arlow DH, et al. High-resolution crystal structure of human protease-activated receptor 1. Nature 2012;492:387–392. pubmed.ncbi.nlm.nih.gov
4. Morrow DA, Braunwald E, Bonaca MP, et al. Vorapaxar in the secondary prevention of atherothrombotic events (TRA 2°P–TIMI 50). N Engl J Med 2012;366:1404–1413. pubmed.ncbi.nlm.nih.gov
5. Tricoci P, Huang Z, Held C, et al. Thrombin-receptor antagonist vorapaxar in acute coronary syndromes (TRACER). N Engl J Med 2012;366:20–33. pubmed.ncbi.nlm.nih.gov
6. US FDA. Zontivity (vorapaxar) prescribing information, NDA 204886, revised 11/2019 — boxed warning, contraindications and pharmacokinetics. accessdata.fda.gov
7. Kahn ML, Nakanishi-Matsui M, Shapiro MJ, Ishihara H, Coughlin SR. Protease-activated receptors 1 and 4 mediate activation of human platelets by thrombin. J Clin Invest 1999;103:879–887. jci.org
8. Nakanishi-Matsui M, Zheng YW, Sulciner DJ, et al. PAR3 is a cofactor for PAR4 activation by thrombin. Nature 2000;404:609–613. pubmed.ncbi.nlm.nih.gov

Informational only — not medical advice. VialHelp does not sell peptides and does not recommend any product, treatment or dose. Antiplatelet drugs are prescribed and monitored by clinicians; nothing above is guidance on their use. Discuss any medical question with a qualified healthcare professional. Intended for readers 21+.