Proteolytic Profiling with Imaging Agents for Rational Design of Target-Activated Peptide Prodrugs.

Abstract

Peptide prodrugs can be used to alter pharmacokinetic properties of drugs such improved oral bioavailability and site-specific delivery. This thesis focuses on peptide prodrug activating enzymes and selection of peptide promoieties to achieve targeted activation. The peptide prodrug L-Val-L-Ser-cyclic cidofovir (Val-Ser-cHPMPC) was previously shown to improve the oral bioavailability of the poorly absorbed antiviral cHPMPC. However, this prodrug must be efficiently and predictably hydrolyzed in vivo to the parent compound cHPMPC to exhibit antiviral activity. Herein, we describe the identification and characterization of puromycin-sensitive aminopeptidase (APP-S) as the primary activator of Val-Ser-cHMPC. For orally absorbed prodrugs, it is often desirable to achieve immediate activation upon absorption. Conversely, to achieve site-specific delivery it is desirable to select a promoiety that is preferentially cleaved by a protease that is overexpressed or uniquely expressed in diseased tissue. One such protease that is overexpressed in tumors is the cysteine endopeptidase legumain. To measure protease activity in a minimally invasive manner, we synthesized a peptide conjugate of the MRI contrast agent Gd-DTPA. The peptide effectively blocked the ninth coordination site of gadolinium, which resulted in a slower relaxitivity (R1) than the parent compound Gd-DTPA or the proposed single amino acid metabolite, which should result in enhanced signal intensity in vivo. However, the Gd-DTPA-tetrapeptide analogue was not significantly hydrolyzed by purified recombinant legumain. The purified legumain was shown to be active using the model substrate Z-Ala-Ala-Asn-AMC, but there was minimal hydrolysis of the model compound when incubated with HEK-293 cells overexpressing legumain. This difference in enzymatic activity between the purified protease and protease expressed in whole cells suggested a whole-cell system would be more physiologically relevant for peptide promoiety screening. We selected amino acid and peptide conjugates of the fluorescent compounds AMC and ACC to screen whole-cells and were able to identify several promoieties that were hydrolyzed significantly faster by BT-549 breast cancer cells compared to MCF7 or MRC-5 cells. One of these dipeptide promoieties, L-Lys-L-Ala, was selected to make a doxorubicin prodrug. There was differential activation of L-Lys-L-Ala-Dox by BT-549 cells, suggesting that our peptide promoiety screening system can be applied to rational prodrug design.