Enhancing peptide ligand binding to vascular endothelial growth factor by covalent bond formation.

Formation of a stable covalent bond between a synthetic probe molecule and a specific site on a target protein has many potential applications in biomedical science. For example, the properties of probes used as receptor-imaging ligands may be improved by increasing their residence time on the targeted receptor. Among the more interesting cases are peptide ligands, the strongest of which typically bind to receptors with micromolar dissociation constants, and which may depend on processes other than simple binding to provide images. The side chains of cysteine, histidine, or lysine are attractive for chemical attachment to improve binding to a receptor protein, and a system based on acryloyl probes attaching to engineered cysteine provides excellent positron emission tomographic images in animal models (Wei et al. (2008) J. Nucl. Med. 49, 1828-1835). In nature, lysine is a more common but less reactive residue than cysteine, making it an interesting challenge to modify. To seek practically useful cross-linking yields with naturally occurring lysine side chains, we have explored not only acryloyl but also other reactive linkers with different chemical properties. We employed a peptide-VEGF model system to discover that a 19mer peptide ligand, which carried a lysine-tagged dinitrofluorobenzene group, became attached stably and with good yield to a unique lysine residue on human vascular endothelial growth factor (VEGF), even in the presence of 70% fetal bovine serum. The same peptide carrying acryloyl and related Michael acceptors gave low yields of attachment to VEGF, as did the chloroacetyl peptide.

Chemically modified antibodies as diagnostic imaging agents.

Notable new applications of antibodies for imaging involve genetically extracting the essential molecular recognition properties of an antibody, and in some cases enhancing them by mutation, before protein expression. The classic paradigm of intravenous administration of a labeled antibody to image not only its target but also its metabolism can be improved on. Protocols involving molecular targeting with an engineered unlabeled protein derived from an antibody, followed by capture of a small probe molecule that provides a signal, are being developed to a high level of utility. This is accompanied by new strategies for probe capture such as irreversible binding, incorporation of engineered enzyme active sites, and antibody-ligand systems that generate a signal only upon binding or uptake.