Regulation of the Function of Αvβ3 in Platelets by a Designed Peptide Targeting the Αv Transmembrane Domain.
Hang Yin,Joanna S. Slusky,Bryan W. Berger,Rustent I. Litvinov,Gaston Vilaire,James D. Lear,William F. DeGrado,Joel S. Bennett
DOI: https://doi.org/10.1182/blood.v108.11.1504.1504
IF: 20.3
2006-01-01
Blood
Abstract:Integrins reside on cell surfaces in an equilibrium between inactive and active conformations. When inactive, the transmembrane (TM) domains of integrin α and β subunits interact, but the domains separate when integrins assume their active conformation. Although this conformational change has not been shown for αvβ3, we hypothesized that a peptide designed to bind to the αv TM domain might activate αvβ3 in platelets by disrupting the TM domain heterodimer of the inactive molecule. To design such a peptide, we used CHAMP (Computed Helical Anti-Membrane Protein) methodology. In the CHAMP method, the αv TM helix was scanned for motifs likely to mediate αvβ3 TM domain interactions. Next, the backbone conformation for an αv-binding peptide was selected based on known structural preferences for the motifs identified in the αv helix. Finally, the sequence for the peptide was designed computationally using a side-chain repacking algorithm. The CHAMP peptide, anti-αv, and its TM helix target, αv-TM, were synthesized by solid phase synthesis. We also synthesized anti-αvmut in which the putative anti-αv-TM binding motif, GXXXG, was mutated to LXXXL. Lys2 dipeptides and short polyethylene glycol sequences were appended to the C- and N-termini of the peptides, respectively, to facilitate their solubility and insertion into membranes. CD spectroscopy revealed that both anti-αv and αv-TM were helical in micelles and phospholipid vesicles. Measurement of Trp fluorescence intensity revealed that anti-αv rapidly inserted into unilamellar POPC/POPG vesicles. Analytical ultracentrifuge and fluorescence resonance energy transfer experiments demonstrated that anti-αv bound to the αv-TM, but not to a homologous αIIb-TM domain peptide. In addition, there was negligible interaction between αv-TM and anti-αvmut. Anti-αv also formed heteromeric complexes with the αv TM domain in bacterial membranes but not with the TM domains of αIIb, α2, β1, or β3. αvβ3 mediates the adhesion of agonist-stimulated platelets to the matrix protein osteopontin (OPN). We found that anti-αV at μM concentrations induced platelet adhesion to OPN. Adhesion was prevented by the divalent cation chelator EDTA, consistent with an integrin-mediated process, but was only minimally affected by pre-incubating the platelets with PGE1, implying that anti-αv-induced adhesion by interacting directly with the αv TM domain. Force spectroscopy using laser tweezers confirmed the specificity of the interaction of anti-αv with platelets: anti-αv induced specific rupture forces between platelets and OPN-coated beads, but not between platelets and fibrinogen-coated beads. Moreover, only non-specific rupture forces were detected between OPN-coated beads and platelets incubated with anti-αvmut. These results demonstrate the successful application of computational methods to design a soluble peptide that specifically recognizes the TM domain of αv in platelets membranes, even when a 400-fold excess of the homologous integrin subunit αIIb is present. Further, because the peptide binds to the site on the αv TM domain that interacts with the β3 TM helix and activates αvβ3, these results strongly support the hypothesis that separation of the αv and β3 TM domains regulates the function of this integrin.