Recent evidence from our laboratory indicates that the angiotensin converting enzyme breakdown product of bradykinin, Arg-Pro-Pro-Gly-Phe (RPPGF), binds to the active site of thrombin and binds to amino acids LDPR in the P4-P1 position on PAR1 (

Am J Physiol
). Further, RPPGF and, an RPPGF analog, rOicPGF, individually inhibit the active site of both thrombin and factor VIIa (JPET In Press, 2004). Additional studies show that RPPGF at 2 mg/kg IP or rOicPGF at 0.44 mg/kg IP significantly delays carotid artery thrombosis in the Rose Bengal Model in the 129SF2 and C57BL mice (JPET, In Press, 2004). Both RPPGF and rOicPGF inhibit gamma thrombin-induced mouse platelet aggregation in PRP. Since mouse platelets express PAR4 and not PAR1, RPPGF or rOicPGF also interacts with mouse PAR4. Studies determined if RPPGF and rOicPGF bind human PAR4. Biotinylated-S41ILPAPRGYPGQ52 (biotin-SIL12), a peptide that encompasses the thrombin binding and cleavage site on human PAR4, specifically binds to peptide RPPGC linked to microtiter plates. Soluble RPPGF, rOicPGF, or unlabeled SIL12 inhibits biotin-SIL12 binding to RPPGC with an IC50 at about 1 mM. Scrambled RPPGF, FPRPG, does not block binding. CDNA of human PAR4 from Gly19-Arg78 was prepared from HEL cell mRNA by RT-PCR and cloned into Novagen pET31b vector. The recombinant PAR4 exodomain (rPAR4ec) was confirmed by N-terminal sequencing and its expected size was determined by mass spectrometry. Further, it is recognized by polyclonal antibody to PAR4, but not to antibodies to PAR1 or PAR3. RPPGFK-biotin specifically binds to rPAR4ec linked to microtiter plates. This binding is blocked by rOicPGF or RPPGF (IC50= 0.2 or 0.75 mM, respectively), but not peptide WPPGF. These data indicate that the N-terminal arginines are essential to interact with rPAR4ec. Further studies mapped the site where RPPGF binds to rPAR4ec. Alanine was substituted for proline at the P4 position (PAR4ec-P44A), the P2 position (rPAR4-P46A), or both the P4 and P2 position (rPAR4-p44A/P46A). RPPGFK-biotin binding to linked rPAR4ec is blocked by increasing concentrations (0.1–0.3 mM) soluble r PAR1ec-WT, rPAR4ec-WT, and rPAR4ec-P44A with an IC50 of 0.22 mM. In contrast, 20-fold excess (0.3 mM) rPAR4ec-P46A or rPAR4ec-P44A/P46A does not block RPPGFFK-biotin binding to rPARec linked to microtiter plates. These data indicate that the RPPGF binding site on rPAR4ec is Pro46. RPPGF, a naturally occurring biologic peptide, has the ability to bind to both PAR1 and PAR4 at the thrombin cleavage site. This information suggests that there are conformational identities among thrombin substrates. These data also suggest that the ACE breakdown product of bradykinin may be a thrombin inhibitor that contributes to the constitutive anticoagulant state of the intravascular compartment. Since RPPGF and rOicPGF are directed to 4 individual targets that contribute to thrombosis, these compounds could be used as model structures to develop single, small molecule inhibitors against multiple targets that contribute to thrombosis risk.

Author notes

Corresponding author