Chemotherapy utilizing cytotoxic drugs is currently the most commonly used anticancer treatment. The utility of chemotherapeutic drugs is severely hampered, however, by their limited selectivity for tumor cells over normal cells. Enzyme prodrug therapy (EPT) attempts to circumvent this obstacle by using a nontoxic prodrug which is converted into a cytotoxic product only in the vicinity of the tumor (see above figure). This strategy exploits the targeting of an activating enzyme to the vicinity of the tumor using monoclonal antibodies specific to particular cancer types covalently attached to the activating enzyme (Antibody-Directed Enzyme Prodrug Therapy, or ADEPT). As of April 2005, there were 74 EPT cancer treatments in clinical trials.

There are some important problems with this therapy, however. Principle among these is the fact that prodrugs tend to be activated elsewhere in the body. In the case of the cytochrome P450- catalyzed activation of the cyclophosphamide prodrug CPA, for example, this is unsurprising as the enzyme conjugate is a human liver P450, CYP2B6. CPA:CYP2B6 treatment thus reduces tumor size at the cost of significant liver damage. We are attempting to circumvent this problem by designing new prodrug activating enzymes which bear no resemblance, in structure, sequence or reactivity, to human P450s. This will in turn allow the redesign of the prodrug to be more orthogonal to human metabolism. A related project involves the design of novel flavoproteins that activate nitroaromatic prodrugs. We are starting with the redesign of the flavin cofactor into a series of safranin-like coenzymes, which have differing reduction potentials. Again, this is intended to allow the redesign of the nitroaromatic prodrugs to reduce systemic toxiciy.