In this study, we determined the ability of recombinant human liver prolidase to hydrolyze nerve agents in vitro and its ability to afford protection in vivo in mice. Using adenovirus containing the human liver prolidase gene, the enzyme was over expressed by 200- to 300-fold in mouse liver and purified to homogeneity by affinity and gel filtration chromatography. The purified enzyme hydrolyzed sarin, cyclosarin and soman with varying rates of hydrolysis. The most efficient hydrolysis was with sarin, followed by soman and by cyclosarin {apparent kcat/Km [(1.9 ± 0.3), (1.7 ± 0.2), and (0.45 ± 0.04)] × 10(5 )M(-1 )min(-1), respectively}; VX and tabun were not hydrolyzed by the recombinant enzyme. The enzyme hydrolyzed P (+) isomers faster than the P (-) isomers. The ability of recombinant human liver prolidase to afford 24 hour survival against a cumulative dose of 2 × LD50 of each nerve agent was investigated in mice. Compared to mice injected with a control virus, mice injected with the prolidase expressing virus contained (29 ± 7)-fold higher levels of the enzyme in their blood on day 5. Challenging these mice with two consecutive 1 × LD50 doses of sarin, cyclosarin, and soman resulted in the death of all animals within 5 to 8 min from nerve agent toxicity. In contrast, mice injected with the adenovirus expressing mouse butyrylcholinesterase, an enzyme which is known to afford protection in vivo, survived multiple 1 × LD50 challenges of these nerve agents and displayed no signs of toxicity. These results suggest that, while prolidase can hydrolyze certain G-type nerve agents in vitro, the enzyme does not offer 24 hour protection against a cumulative dose of 2 × LD50 of G-agents in mice in vivo.
