OP nerve agents serve as hemisubstrates of the wild-type AChE, BChE, and hCE1 enzymes because these chemicals fail to complete the typical two-step serine hydrolase mechanism. Rather, OP nerve agents produce relatively long-lived covalently phosphonylated active site serines 
. Histidine (H468 in wt hCE1; see ) activates the serine (S221 in hCE1) for nucleophilic attack on the OP phosphonyl center. The resultant pentahedral intermediate is stabilized by hCE1 backbone nitrogen atoms in the oxyanion hole (G142 and G143); the fluoride leaving group dissociates upon collapse of the transition state, resulting in a covalently modified enzyme 
. The resulting tetrahedral phosphonyl adduct adversely affects the ability of H468 to facilitate base mediated C-O bond cleavage 
, which leaves this residue in a state where it cannot activate the water molecule required for hydrolytic desphosphonylation 
. Thus, for serine hydrolases like AChE and hCE1 that are inhibited by OP hemisubstrates, the addition of a strong oxime is often required to complete the catalytic cycle and regenerate active enzyme. Current treatments following OP exposure include administration of strongly-nucleophilic oximes to dephosphonylate AChE. However, such compounds, such as 2-pralidoxime obidoxime, do not offer broad-spectrum protection against all agents and must be administered quickly 
Based on previous crystal structures of hCE1 in covalent complexes with nerve agents 
, we hypothesized that introducing a general base catalyst would facilitate the activation of a properly located water molecule to hydrolyze an OP adduct. The V146H/L363E amino acid substitutions dramatically increased the spontaneous rates of enzyme reactivation following sarin, soman, and cyclosarin inhibition compared to native enzyme (). A panel of single and double mutant controls indicated that V146H or L363Q facilitated increased sarin hydrolysis, while L363E acted as general base catalyst for cyclosarin and soman hydrolysis. Because V146H, L363Q, or their combination, resulted in similar rates of reactivation, these mutations may enhance dephosphonylation by interacting with the sarin O
-isopropyl group, thereby affording deprotonation of the catalytic H468 for water activation and hydrolysis. A G117H mutation in BChE resulted in increased rates of dephosphonylation through an analogous mechanism 
In contrast, the L363E mutation positions an anionic carboxylate adjacent to the phosphonyl-enzyme intermediate to activate a water molecule for nucleophilic attack. Glutamic acid and aspartic acid residues act as general acid-base catalysts in several established enzyme mechanisms, including those of lysozyme and protein tyrosine phosphatases 
, and apparently support the same role in hCE1. The V146H addition synergistically increased base activity of L363E (). The pH profile for cyclosarin hydrolysis by V146H/L363E showed that the fastest rates of hCE1 reactivation was when H146 was likely to be in a charged state, below pH 6.2 and where a cationic H146 can stabilize the anionic base (). Based on these data, we postulate that E363 Oε1 is stabilized by H146 and Oε2 on E363 is available to deprotonate a water molecule to hydrolyze the covalent cyclohexyl phosphonyl group (). This hypothesis is supported by the observation that the V146Q/L363E variant also exhibited a significant increase in cyclosarin hydrolysis ().
Previous attempts to introduce general base catalysts in OP-inhibited serine hydrolases have resulted in enzymes that also showed enhanced rates of reactivations against soman, sarin, and VX, but the ability of these mutant enzymes to bind the respective OPs was greatly diminished 
. For example a G117H/E197Q BChE mutant enhanced the rate of soman hydrolysis up to 2,500-fold faster than wt BChE, but the soman bimolecular rate of inhibition for this mutant was decreased by a similar magnitude 
. In contrast to the G117H/E197Q BChE mutant, V146H/L363E hCE1 exhibits a greater second-order rate of inhibition (ki
) than the wt hCE1 enzyme (). Further, the Km
values for the cyclosarin model compounds reported in demonstrate that the affinity for V146H/L363E hCE1 is greater than the wt enzyme for these analogs. Taken together, these data show that V146H/L363E hCE1 retains greater than wt affinity for cyclosarin, and that the engineered hCE1 mutants enhance rates of reactivation via dephosphonylation rather than decreasing inhibition.
In conclusion, we have rationally designed a variant form of the liver detoxifying enzyme hCE1 that spontaneously dephosphonylates after inhibition by sarin, soman, and cyclosarin up to 33,000-fold faster than wt enzyme. Wild type hCE1 has a catalytic efficiency (kcat
towards the standard esterase substrate para-nitrophenylbutyrate (pNPB) of 1.2×105
. AChE, one of the most efficient enzymes known, exhibits a kcat
close to the diffusion control limit (108
) for acetylcholine hydrolysis 
. Combining the nanomolar Km
of V146H/L363E towards the PR
cyclosarin-like compound along with rate of reactivation after inhibition by this compound, this double mutant shows a catalytic efficiency of 8.8×102
. Thus, the redesigned hCE1 compares favorably to other mammalian enzymes that have been rationally engineered to improve hemi-substrate metabolism (). This enzyme will likely require substantial increases in catalytic efficiency for OP compounds in order to provide in vivo
protection, but nonetheless can serve as a lead candidate for further development of novel countermeasures for nerve agent or pesticide poisoning.
Catalytic efficiencies (kcat/Km) of engineered enzymes towards hemisubstrates.