Two Specific Domains on the Upper Surface of the αIib β Propeller Determine the Sensitivity of αIibβ3 for RGD-Containing Peptides.

A feature of many integrins is their ability to interact with ligands containing an Arg-Gly-Asp (RGD) motif. RGD-containing peptides and peptidomimetics can inhibit ligand binding to these integrins, including the binding of fibrinogen and von Willebrand Factor (VWF) to the major platelet integrin αIIbβ3. Despite substantial sequence conservation, RGD-containing peptides interact poorly with αIIbβ3 on rat and mouse platelets compared to human platelets, a phenomenon we previously found to be due to sequence differences in the third and fourth N-terminal repeats of human and rat αIIb. To localize the specific residues that control RGD sensitivity and determine whether these residue differences determine sensitivity for all RGD-based ligands, we constructed a series of inter-species chimeric αIIbβ3 molecules in which non-conserved residues in the loops connecting the 2nd and 3rd and the 3rd and 4th blades of the amino-terminal αIIb β propeller were exchanged between the rat and human proteins. We then measured the inhibitory effect of the tetrapeptide RGDS and the tyrosine-based RGD peptidomimetic tirofiban on fibrinogen to the αIIbβ3 chimeras expressed in Chinese hamster ovary (CHO) cells. We chose to study tirofiban because it is a more extended molecule than RGDS with a greater distance between its positively-charged arginine and negatively-charged aspartate moieties. We found that specific substitutions in either the 2–3 loop (a combination of V¹⁵⁶Δ, E¹⁵⁷S, D¹⁵⁹G, and W¹⁶²G (mature human αIIb numbering)) or in the D²³²H 3–4 loop affected the sensitivity of rat αIIbβ3 to the inhibitory effects of RGDS alone or in combination. To alter tirofiban sensitivity, only a simultaneous substitution in both loops was effective. Mouse αIIbβ3 is also resistant to RGDS, and this resistance is localized to the same region of the 2–3 loop. There is no D²³²H in mouse αIIb, and murine αIIbβ3 is as expected as sensitive to tirofiban as its human counterpart. It is of interest that in the published structure of an RGD pentapeptide soaked into the crystal of the related integrin αvβ3 two residues complex with the Arg. These residues are homologous to αIIb D¹⁶³ and D²³², the former located immediately adjacent to the substituted residues in the 2–3 loop that affect RGD resistance and the latter being the important difference between human and rat in the 3–4 loop. Whether D¹⁶³ and D²³² complex to RGDS on αIIbβ3 is unknown. In the published crystal structure of the extracellular portion of human αIIbβ3 co-crystallized with either of several extended RGD-like compounds, the basic residue of the RGD-like compounds was found to hydrogen bond to αIIb D²²⁴, a residue located at the base of a pocket that is formed by the 2–3 and 3–4 loops and that is highly conserved among different αIIb species. Our results indicate that sensitivity to such an extended RGD-containing compounds is determined by the composition of the walls of this pocket, thereby facilitating or impeding access to D²²⁴. Thus these differences in αIIb composition and function between human and rodent αIIbβ3, which may have been an evolutionary response to venomous exposure by rodents, may provide insights into the intermediate steps by which the αIIbβ3 receptor binds to its natural ligands.