Cathepsin S signals via PAR2 and generates a novel tethered ligand receptor agonist

Abstract

Protease-activated receptor-2 is widely expressed in mammalian epithelial, immune and neural tissues. Cleavage of PAR2 by serine proteases leads to self-activation of the receptor by the tethered ligand SLIGRL. The contribution of other classes of proteases to PAR activation has not been studied in detail. Cathepsin S is a widely expressed cysteine protease that is upregulated in inflammatory conditions. It has been suggested that cathepsin S activates PAR2. However, cathepsin S activation of PAR2 has not been demonstrated directly nor has the potential mechanism of activation been identified. We show that cathepsin S cleaves near the N-terminus of PAR2 to expose a novel tethered ligand, KVDGTS. The hexapeptide KVDGTS generates downstream signaling events specific to PAR2 but is weaker than SLIGRL. Mutation of the cathepsin S cleavage site prevents receptor activation by the protease while KVDGTS retains activity. In conclusion, the range of actions previously ascribed to cysteine cathepsins in general, and cathepsin S in particular, should be expanded to include molecular signaling. Such signaling may link together observations that had been attributed previously to PAR2 or cathepsin S individually. These interactions may contribute to inflammation.

Figure 1. Concentration-effect curves for PAR2 derived peptides.

Calcium-dependent responses were determined in PAR2-transfected HeLa cells following incubation with the indicated peptides. The curves for activating peptides are similar in shape although that for KVDGTS is shifted to the right and has a lower peak calcium response as compared to SLIGRL and SLIGKVDGTS. The inverse hexapeptides, STGDVK and LRGILS were not active.

Figure 2. PAR2 N-terminal hexapeptides are capable of activating the receptor.

a) Single-cell calcium imaging in HeLa cells transfected with PAR2 cDNA demonstrated a similar response to KVDGTS (thick solid line) and SLIGRL (thick dotted line). A weaker response was observed in response to IGKVDG (thin solid line). HeLa cells transfected with salmon sperm DNA and stimulated with KVDGTS did not respond (horizontal dotted line). b) Cathepsin S (solid line) and KVDGTS (100 µM) (dotted line) elicit similar calcium responses in NHEKs. KVDGTS (100 µM) and SLIGRL (10 µM), IGKVDG (100 µM) and cathepsin S (2 µM).

Figure 3. KVDGTS alters PAR2 response to cathepsin S.

a) HeLa cells transfected with PAR2 cDNA were treated with KVDGTS and subsequently with cathepsin S after a 2 minute interval (dotted line). Pre-treatment with KVDGTS attenuates the response to cathepsin S. HeLa cells treated with cathepsin S failed to respond subsequently to KVDGTS (100 µM) (solid line). b) KVDGTS (100 µM) elicted calcium responses in NHEKs and the subsequent response to cathepsin S (2 µM) was attenuated. Treatment with cathepsin S abolishes the response to the subsequent addtion of KVDGTS. The second agent was delivered at the 300 second timepoint. KVDGTS (100 µM), cathepsin S (2 µM).

Figure 4. Cathepsin S and hexapeptide agonists increase the concentration of inositol phosphate.

IP1 is a downstreatm metabolite of IP3. IP1 concentrations were measured following treatment with cathepsin S, KVDGTS and SLIGRL in HeLa cells that had been transfected with PAR2 cDNA. Hexapeptide agonists activated this signaling cascade downstream of PAR2, but to a lesser extent than cathepsin S. Cathepsin S had no effect above baseline IP1 levels on salmon sperm transfected cells. Cathepsin S (1 µM), KVDGTS (100 µM) and SLIGRL (10 µM).

Figure 5. Cathepsin S and PAR2 peptide agonists induce PKC Ser660 phosphorylation in HeLa cells.

The Western blot was performed on HeLa cells that had been transfected with PAR2 cDNA. Treatments included SLIGRL (lane 1), SLIGKVDGTS (lane 2), KVDGTS (lane 3) and cathepsin S (lane 4). Non-transfected HeLa cells treated with SLIGRL (10 µM) (lane 5) and PAR2-transfected but untreated HeLa cells (lane 6) served as controls. SLIGRL (10 µM), SLIGKVDGTS (10 µM), KVDGTS (100 µM) and cathepsin S (1 µM).

Figure 6. Cathepsin S cleavage of the PAR2 N-terminus requires glycine⁴⁰.

a) Luciferase luminescence was measured after cathepsin S (2 µM) treatment of HeLa cells expressing luciferase-PAR2 (bar 1), PAR2M1 (bar 2), PAR2M2 (bar 3), and PAR2M3 (bar 4). Luminescence of the transfected cells without cathepsin S treatment was subtracted from each of the measurements. b) Western blots were performed on supernatants of luciferase-PAR2 and PAR2 mutant-transfected HeLa cells following treatment with cathepsin S. Luciferase-PAR2, cathepsin S(−) (lane 1); luciferase-PAR2, cathepsin S (+) (lane 2); PAR2M1, cathepsin S (−) (Lane 3); PAR2M1 cathepsin S (+) (lane 4); PAR2M2, cathepsin S (−) (lane 5); PAR2M2, cathepsin S (+) (lane 6); PAR2M3, cathepsin S (−) (lane 7); PAR2M3, cathepsin S (+) (lane 8) and non-transfected HeLa cells, cathepsin S (+) (lane 9). Molecular weight markers on the left side of the blot are in kDa.

Figure 7. Cathepsin S fails to activate mutant PAR2 but KVDGTS retains activity on mutant PAR2.

a) Calcium imaging was performed in HeLa cells transfected with native or mutant PAR2 receptors following treatment with cathepsin S. Cathepsin S (2 µM) was able to stimulate native PAR2 and PAR2M1 but not PAR2M2 or PAR2M3. b) In contrast, KVDGTS activated PAR2 and all three substitution mutants. Cathepsin S (2 µM), KVDGTS (100 µM).