Immunogenicity caused by the use of non-human components in Antibody Directed Enzyme Prodrug Therapy (ADEPT) has limited its clinical application. To overcome this problem, we introduced rationally designed mutations in the substrate binding pocket of human purine nucleoside phosphorylase (hPNP) to change its substrate specificity. Our goal was to introduce minimal changes in the amino acid and structure of hPNP to develop a mutant enzyme, that unlike the wild type enzyme, accepts (deoxy)adenosine-based prodrugs as substrate. Amongst the different mutants of human PNP tested, a double mutant with amino acid substitutions Glu201Gln:Asn243Asp (hDM) is most efficient in cleaving the adenosine-based prodrugs, such as 2-fluoro-2'-deoxyadenosine (F-dAdo) to their corresponding cytotoxic drugs.

To target hDM to the tumor site, the enzyme was fused to either a human Anti-HER2/neu Peptide mimetic (AHNP) or to a human anti-HER2/ neu single chain Fv (C6MH3B1). Incubation of both fusion proteins, followed by wash to remove the unbound fraction resulted in a dose-dependent cytotoxicity with HER2/neu expressing cells in the presence of F-dAdo. Consistent with the greater affinity of C6MH3B1 for ECD ^HER2, cytotoxicity was seen at lower concentrations of hDM-C6MH3B1 than hDM-AHNP, making C6MH3B1 a better candidate for delivering hDM to tumor cells. The toxic drug is cytotoxic to tumor cells irrespective of their expression of tumor associated antigen or growth rate. This cytotoxicity should be restricted to the tumor microenvironment, since the endogenously expressed wild type enzyme does not use adenosine-based prodrugs as substrates.

We have successfully developed an ADEPT that is comprised of all human components (hDM-C6MH3B1). hDM, unlike the wild type enzyme is capable of cleaving the relatively non-toxic F-dAdo to its cytotoxic metabolite F-Ade, causing tumor cell apoptosis in vitro. The similarity of the structure of hDM with wild type hPNP, as well as our evaluation of possible MHCII binding peptides that result from the two introduced mutations or the fusion of the two proteins suggest that the likelihood that hDM-C6MH3B1 will elicit an immune response in human is small. We anticipate that targeting hDM to tumors followed by administration of F-dAdo will be cytotoxic to tumors, with minimal systemic toxicity and immunogenicity.