Antiviral prodrug activation and targeting viral protease.

Abstract

Prodrug approaches have been widely used to overcome pharmaceutical and pharmacokinetics barriers. However, knowledge about site(s) of prodrug activation is frequently limited. Our results have shown that an activity-based protein profiling agent, biotinylated fluorophosphonate (FP-PEG-Biotin) probe, will bind to active serine hydrolases irreversibly and compete with prodrugs, for labeling and identifying potential prodrug activating enzymes. Our results show that FP-PEG-Biotin competes with the prodrug oseltamivir for at least two protein bands from Caco-2 cell homogenates, in a concentration-dependent manner. Neither of these bands may be carboxylesterase and thus may be additional activating enzymes for oseltamivir. The novel idea of targeting prodrug activation to specific sites offers the potential to both increase drug efficacy and reduce side effects. We have identified human Cytomegalovirus protease as a potential target. Human cytomegalovirus is a highly prevalent human pathogen. It encodes a serine protease that is responsible for capsid assembly. The crystal structure of this enzyme has revealed a unique serine protease that shares little similarity with its mammalian counterparts. Therefore, we investigated HCMV protease as potential targeting site for prodrug activation. For the first time, we demonstrated that HCMV protease not only can hydrolyze small ester compounds, but the hydrolysis also involves substrate specificity. Ganciclovir is a currently marketed anti-HCMV drug. Its long-term administration often results in side effects that interrupt the treatment. We have synthesized prodrugs of acyclovir and ganciclovir and shown that the amino acid ester prodrugs are substrates of HCMV protease. alpha-amino protection of alanyl ganciclovir prodrugs, especially with large hydrophobic protection groups, not only improved hydrolysis by HCMV protease, but they also improved the prodrug stability in Caco-2 homogenate, rat and human plasma. Further, we have shown that these prodrugs exhibit a cellular uptake that is comparable to that of ganciclovir. Finally a dipeptide prodrug of ganciclovir showed an increase in both HCMV protease hydrolysis and Caco-2 cell homogenate or plasma stability. Thus this research has shown that prodrug activation may involve several enzymes and the targeting prodrugs to protease/esterase enzymes encoded by infectious agents may be a very viable approach to improved specificity and optimization of drug therapy.