Search

An antibody engineering strategy was employed to build high affinity ligands and antagonists of integrins alpha v beta 3 and alpha IIb beta 3. Previously, we inserted the integrin recognition motif, RGD, into the antigen binding site of a human antibody and selected the optimal flanking sequences from a phage-display library (Barbas, C. F., Languino, L. R., and Smith, J. W. (1993) Proc. Natl. Acad. Sci. U.S.A. 90, 10003-10007). The resulting antibody, Fab-9, blocked the function of integrin alpha v beta 3 but also bound to the ligand binding site of platelet integrin alpha IIb beta 3. In this report, the antibody engineering effort has been extended by 1) redesigning Fab-9 to achieve specificity for platelet integrin alpha IIb beta 3, 2) building non-RGD-containing antibodies that bind the ligand binding site of both beta 3-integrins, and 3) testing the hypothesis that peptides derived from complementarity determining regions (CDR) can be used to emulate the activity of the parent synthetic antibody. These goals were accomplished by subjecting the original antibody, Fab-9, to a "motif optimization" (MTF). A phage library was constructed in which the residues flanking the RGD motif in Fab-9 were maintained, but the RGDX sequence was randomized. This library was panned on purified alpha IIb beta 3 to identify high affinity binders. Four function-blocking antibodies lacking RGD, but with specificity for alpha IIb beta 3, were characterized. The antibody with the highest preference for alpha IIb beta 3, MTF-10, had an adhesion sequence of KGDN. This sequence is similar in primary structure to the active sequence within the disintegrin barbourin, which also antagonizes alpha IIb beta 3 (Scarborough, R. M., Rose, J. W., Hsu, M. A., Phillips, D. R., Fried, V. A., Campbell, A. M., Nannizzi, L., and Charo, I. F. (1991) J. Biol. Chem. 266, 9359-9362). MTF-10 had a 70-fold higher affinity for alpha IIb beta 3 than alpha v beta 3. Through our selection strategy, we also identified several antibodies that lack RGD but still blocked ligand binding to both integrins with high affinity. Therefore, the RGD sequence is not necessary for a high affinity interaction with the ligand binding site of beta 3-integrins. Further investigation showed that the activity of inhibitory antibodies could be emulated by synthetic peptides derived from the protein sequences of the antibody's HCDR3. CDR-derived peptides blocked ligand binding to integrins and maintained essentially the same specificity as the parent antibody.

An antibody engineering strategy was employed to build high affinity ligands and antagonists of integrins alpha v beta 3 and alpha IIb beta 3. Previously, we inserted the integrin recognition motif, RGD, into the antigen binding site of a human antibody and selected the optimal flanking sequences from a phage-display library (Barbas, C. F., Languino, L. R., and Smith, J. W. (1993) Proc. Natl. Acad. Sci. U.S.A. 90, 10003-10007). The resulting antibody, Fab-9, blocked the function of integrin alpha v beta 3 but also bound to the ligand binding site of platelet integrin alpha IIb beta 3. In this report, the antibody engineering effort has been extended by 1) redesigning Fab-9 to achieve specificity for platelet integrin alpha IIb beta 3, 2) building non-RGD-containing antibodies that bind the ligand binding site of both beta 3-integrins, and 3) testing the hypothesis that peptides derived from complementarity determining regions (CDR) can be used to emulate the activity of the parent synthetic antibody. These goals were accomplished by subjecting the original antibody, Fab-9, to a "motif optimization" (MTF). A phage library was constructed in which the residues flanking the RGD motif in Fab-9 were maintained, but the RGDX sequence was randomized. This library was panned on purified alpha IIb beta 3 to identify high affinity binders. Four function-blocking antibodies lacking RGD, but with specificity for alpha IIb beta 3, were characterized. The antibody with the highest preference for alpha IIb beta 3, MTF-10, had an adhesion sequence of KGDN. This sequence is similar in primary structure to the active sequence within the disintegrin barbourin, which also antagonizes alpha IIb beta 3 (Scarborough, R. M., Rose, J. W., Hsu, M. A., Phillips, D. R., Fried, V. A., Campbell, A. M., Nannizzi, L., and Charo, I. F. (1991) J. Biol. Chem. 266, 9359-9362). MTF-10 had a 70-fold higher affinity for alpha IIb beta 3 than alpha v beta 3. Through our selection strategy, we also identified several antibodies that lack RGD but still blocked ligand binding to both integrins with high affinity. Therefore, the RGD sequence is not necessary for a high affinity interaction with the ligand binding site of beta 3-integrins. Further investigation showed that the activity of inhibitory antibodies could be emulated by synthetic peptides derived from the protein sequences of the antibody's HCDR3. CDR-derived peptides blocked ligand binding to integrins and maintained essentially the same specificity as the parent antibody.