James P. Beltzer

Isolation, characterization, and recovery of small peptide phage display epitopes selected against viable malignant glioma cells.

Phage display techniques rely on nearly random oligonucleotide sequences inserted into the protein III filament binding protein of an Escherichia coli filamentous phage M13 to generate a library of phage that express more than 10 7 different peptides. Phage that expresses a sequence having high affinity for a specific molecule, cell, or tissue can then be isolated through selective binding and recovery. Selected phage cannot only be used as gene transfer vectors in themselves, but the small peptide epitopes can be sequenced and potentially recombined into the attachment proteins of viral vectors, or used by themselves to target other therapeutic agents and diagnostic imaging radiolabels. Most phage display selections are carried out against purified and/or fixed protein targets, raising concerns as to the relevance of the selected epitopes. We have selected phage from the CMTI library against viable U87-MG human malignant glioma cells using a derivation of biopanning. The library, which initially contained phage expressing 2×10 7 different epitope sequences, collapsed after four rounds of selection such that 42% of recovered clones expressed a consensus sequence. Selective binding to viable adherent U87-MG cells was subsequently demonstrated under physiologic conditions at 167% (±27%) unselected phage using a novel, viable enzyme-linked immunosorbent assay technique. In comparison, there was no difference in binding to control 9L rat gliosarcoma, PANC-1 human pancreatic adenocarcinoma, T98-MG human malignant glioma, or AST-4 human malignant glioma cells of selected compared to unselected phage. Using polymerase chain reaction, the epitope was recovered with flanking unique restriction sites for recombination into a herpes simplex virus type-1 vector. This study demonstrates and discusses optimized methodologies for using phage display to target viable cells. Cancer Gene Therapy (2001) 8, 506–511