Reaction Products of Acetylcholinesterase and VX Reveal a Mobile Histidine in the Catalytic Triad

ReceiVed July 30, 1999 The presence of a precisely aligned active-site triad (Ser-HisAsp/Glu) in the three-dimensional structures of widely different hydrolytic enzymes has generated intense interest in the chemical modus operandi of this catalytic motif.1 One hypothesis, which has not received wide acceptance, proposes that the imidazole of the catalytic His is mobile during enzyme function.2 We solved the structures of the phosphonylation and dealkylation (“aging”) reaction products of acetylcholinesterase (AChE; EC 3.1.1.7) and an organophosphorus (OP) inhibitor, O-ethyl-S-[2-[bis(1-methylethyl)amino]ethyl] methylphosphonothioate (VX) by X-ray crystallography. The structures clearly demonstrate reversible movement of the catalytic His. Moreover, the conformational change apparently involves a hydrogen (H-) bond with a glutamate (E199) which had been implicated previously in OP and substrate reactions. Most serine hydrolases, including AChE, use a catalytic triad and a dipolar oxyanion hole in tandem to catalyze substrate hydrolysis via an acylation-deacylation mechanism.3 This twostep mechanism is also a weakness, however, because it renders the enzyme susceptible to stoichiometric inhibition by “hemisubstrates” which react to form stable analogues of natural reaction intermediates. Following phosphonylation of the active-site Ser Oγ, some OP-enzyme adducts undergo further post-inhibitory reactions, including dealkylation, which result in truly irreversible enzyme inhibition (collectively called “aging”) (Scheme 1). Structures of the reaction products of Torpedo californica (Tc) AChE with DFP, sarin or soman4 after aging reveal that the OP undergoes dealkylation to yield a stable anionic phosphonyl adduct.5 As reported previously for aged OP-serine proteases,6 the catalytic imidazole (H440) of TcAChE is positioned to form H-bonds with its normal carboxylic acid partner (E327), and with one oxygen of the negatively charged phosphonyl moiety. Such structures are limited, however, because they reveal only the final product (II) of the OP reaction. To overcome this limitation, we employed VX. Although phosphonylation of AChE with VX is rapid (>107 M-1 min-1), the ethyl group of VX dealkylates slowly, thus allowing us to solve the structures of both (I) and (II) by conventional X-ray crystallography to 2.2 and 2.4 A resolution, respectively.7 The most striking feature of the pro-aged VX-AChE structure (I) was disruption of the catalytic triad due to movement of H440. The H440 N∂ was 4.5 A away from its resting state partner, E327 O , and within H-bond distance (2.7 A) of E199 O (Figure 1). The observed uncoupling of the catalytic triad offers a new explanation for the slow, often negligible spontaneous reactivation of the VX-AChE adduct. Movement of the imidazole was reversible, however, because the catalytic triad could be restored by either: (1) dephosphonylation with a nucleophile (reactivation) or (2) dealkylation of the VX O-ethyl group (aging). To confirm that active enzyme could be regenerated from the alternate conformation, VX-TcAChE crystals were dissolved in phosphate buffer (pH 7.5), incubated for 20 h with 10 mM pralidoxime, and AChE activity then measured.8 Oxime-reactivated VXTcAChE was indistinguishable from native enzyme with respect to substrate kinetics, corroborating that the H440 movement caused by phosphonylation was reversed upon dephosphonylation. Alternatively, if crystalline VX-TcAChE was allowed to age completely (II), the H440 imidazole moved to a position close to that found in native TcAChE. The aged adduct was essentially identical to those solved previously for sarin or soman,5 and the catalytic H440-E327 pair had reverted to its native conformation (Figure 1). Functional Significance for the Glu327-His440-Glu199 Array in AChE. Site-specific replacement of E199 with Q reduces the rate constant for aging approximately 100-fold.9 Furthermore, the pH dependence of OP-AChE aging follows an asymmetric bell curve with a maximum rate at pH 6, and three apparent pKa’s which implicate two carboxylic acids (4.0-4.9) and an imidazole (5.2-6.6).10 Mechanistic hypotheses of the aging reaction, therefore, have centered upon the role of E199.11 The VX-TcAChE structures, obtained near the optimal pH for aging, provide the first evidence that H440 is mobile in * Author for correspondence. E-mail: limill@sgjs8.weizmann.ac.il. Present address: U.S. Army Medical Research Institute of Infectious Diseases, 1425 Porter Street, Fort Detrick, MD 21702-5011. † Present address: Free University of Berlin, Takustrasse 6, D-14195, Berlin, Germany. ‡ Israel Institute for Biological Research. (1) Reviewed in Dodson, G.; Wlodawer, A. Trends Biochem. Sci. 1998, 23, 347-352. (2) (a) Satterthwait, A. C.; Jencks, W. P. J. Am. Chem. Soc. 1974, 96, 70187031. (b) Bachovchin, W. W. Biochemistry 1986, 25, 7751-7759. (3) Kraut, J. Annu. ReV. Biochem. 1977, 46, 331-358. (4) DFP, diisopropylphosphorofluoridate; sarin, O-isopropylmethylphosphonofluoridate; soman, O-pinacolyl methylphosphonofluoridate. (5) Millard, C. B.; Kryger, G.; Ordentlich, A.; Greenblatt, H. M.; Harel, M.; Raves, M. L.; Segall, Y.; Barak, D.; Shafferman, A.; Silman, I.; Sussman, J. Biochemistry 1999, 38, 7032-7039. (6) Kossiakoff, A. A.; Spencer, S. A. Nature 1980, 288, 414-416. (7) Coordinates and details of data collection and refinement are available from the Protein Databank under accession codes 1vxr (pro-aged adduct) and 1vxo (aged adduct). Briefly, crystals of space group P3121 were prepared in PEG-200, pH 6.0, and data were collected at 100 K with synchrotron radiation at Elletra, Trieste (1vxr) or Broohaven National Laboratory, NY (1vxo). All crystals were isomorphous with native TcAChE (2ace); see ref 5 and Sussman, J. L.; Harel, M.; Frolow, F.; Oefner, C.; Goldman, A.; Toker, L.; and Silman, I. Science 1991, 253, 872-879. The final R-factors were 18.9% (1vxr) and 19.4% (1vxo), and the free-R-factors were 23.0% (1vxr) and 23.4% (1vxo). (8) The substrate was acetythiocholine and enzyme activity was measured by the method of Ellman, G. L.; Courtney, K. D.; Andres, V.; Featherstone, R. M. Biochem. Pharmacol. 1961, 7, 88-95. (9) (a) Ordentlich, A.; Kronman, C.; Barak, D.; Stein, D.; Ariel, N.; Marcus, D.; Velan, B.; Shafferman, A. FEBS Lett. 1993, 334, 215-220. (b) Saxena, A.; Doctor, B. P.; Maxwell, D. M.; Lenz, D. E.; Radic, Z.; Taylor, P. Biochem. Biophys. Res. Commun. 1993, 197, 343-349. Scheme 1. OP Reaction with Serine Hydrolasesa