Characterization of an engineered human purine nucleoside phosphorylase fused to an anti-her2/neu single chain Fv for use in ADEPT

BACKGROUND: Antibody Directed Enzyme Prodrug Therapy (ADEPT) can be used to generate cytotoxic agents at the tumor site. To date non-human enzymes have mainly been utilized in ADEPT. However, these non-human enzymes are immunogenic limiting the number of times that ADEPT can be administered. To overcome the problem of immunogenicity, a fully human enzyme, capable of converting a non-toxic prodrug to cytotoxic drug was developed and joined to a human tumor specific scFv yielding a fully human targeting agent. METHODS: A double mutant of human purine nucleoside phosphorylase (hDM) was developed which unlike the human enzyme can cleave adenosine-based prodrugs. For tumor-specific targeting, hDM was fused to the human anti-HER2/neu single chain Fv (scFv), C6 MH3B1. Enzymatic activity of hDM with its natural substrates and prodrugs was determined using spectrophotomeric approaches. A cell proliferation assay was used to assess the cytotoxicity generated following conversion of prodrug to drug as a result of enzymatic activity of hDM. Affinity of the targeting scFv, C6 MH3B1 fused to hDM to Her2/neu was confirmed using affinity chromatography, surface plasmon resonance, and flow-cytometry. RESULTS: In vitro hDM-C6 MH3B1 binds specifically to HER2/neu expressing tumor cells and localizes hDM to tumor cells, where the enzymatic activity of hDM-C6 MH3B1, but not the wild type enzyme, results in phosphorolysis of the prodrug, 2-fluoro-2'-deoxyadenosine to the cytotoxic drug 2-fluoroadenine (F-Ade) causing inhibition of tumor cell proliferation. Significantly, the toxic small drug diffuses through the cell membrane of HER2/neu expressing cells as well as cells that lack the expression of HER2/neu, causing a bystander effect. F-Ade is toxic to cells irrespective of their growth rate; therefore, both the slowly dividing tumor cells and the non-dividing neighboring stromal cells that support tumor growth should be killed. Analysis of potential novel MHCII binding peptides resulting from fusion of hDM to C6 MH3B1 and the two mutations in hDM, and of the structure of hDM compared to the wild-type enzyme suggests that hDM-C6 MH3B1 should exhibit minimal immunogenicity in humans. CONCLUSION: hDM-C6 MH3B1 constitutes a novel human based protein that addresses some of the limitations of ADEPT that currently preclude its successful use in the clinic.