Inefficiency of gene delivery, together with inadequate bystander killing represent two major hurdles in the development of toxin-mediated gene transfer for treatment of human malignancy. The E.coli Deo D gene (purine nucleoside phosphorylase (PNP)) is a well characterized enzyme that is capable of catalyzing the conversion of several nontoxic deoxyadenosine analogs to highly toxic adenine analogs. We have shown that expression of PNP in less that 1% of human cancer cells leads to death of virtually all bystander cells after treatment with the prodrug 6-methylpurine-2'- deoxyriboside (MeP-dR), a deoxyadenosine substrate for the E. coli PNP but not the human PNP. Prodrug activation as part of a gene therapy-based strategy offers substantial advantages over the expression of directly toxic genes, such as ricin, diphtheria toxin, or pseudomonas exotoxin. These advantages include the capability to 1) titrate cell killing, 2) optimize therapeutic index by adjusting either levels of prodrug or of recombinant enzyme expression, and 3) interrupt toxicity by omitting administration of the prodrug. This proposal intends to use recombinant DNA technology to evaluate E.coli PNP mediated tumor cell killing in vitro, to characterize bystander killing using this gene, and to develop animal models for the delivery of E.coli PNP to established tumors in vivo. Two specific model systems for use of E.coli PNP during in vivo treatment of tumors are proposed: 1) retroviral based treatment of intracranial glioblastoma; and 2) cationic liposomal gene transfer for treatment of metastatic solid tumors such as melanoma.
Showing the most recent 10 out of 29 publications