Structures of recombinant wild-type human acetylcholinesterase and of its E202Q mutant as complexes with fasciculin-II, a 'three-finger' polypeptide toxin purified from the venom of the eastern green mamba (Dendroaspis angusticeps), are reported. The structure of the complex of the wild-type enzyme was solved to 2.8 A resolution by molecular replacement starting from the structure of the complex of Torpedo californica acetylcholinesterase with fasciculin-II and verified by starting from a similar complex with mouse acetylcholinesterase. The overall structure is surprisingly similar to that of the T. californica enzyme with fasciculin-II and, as expected, to that of the mouse acetylcholinesterase complex. The structure of the E202Q mutant complex was refined starting from the corresponding wild-type human acetylcholinesterase structure, using the 2.7 A resolution data set collected. Comparison of the two structures shows that removal of the charged group from the protein core and its substitution by a neutral isosteric moiety does not disrupt the functional architecture of the active centre. One of the elements of this architecture is thought to be a hydrogen-bond network including residues Glu202, Glu450, Tyr133 and two bridging molecules of water, which is conserved in other vertebrate acetylcholinesterases as well as in the human enzyme. The present findings are consistent with the notion that the main role of this network is the proper positioning of the Glu202 carboxylate relative to the catalytic triad, thus defining its functional role in the interaction of acetylcholinesterase with substrates and inhibitors.

We have examined the effects of 11 substitutions of active centre gorge residues of human acetylcholinesterase (HuAChE) on the rates of phosphonylation by 1,2,2-trimethylpropyl methyl-phosphonofluoridate (soman) and the aging of the resulting conjugates. The rates of phosphonylation were reduced to as little as one-seventieth, mainly in mutants of the hydrogen-bond network (Glu-202, Glu-450, Tyr-133). These recombinant enzymes as well as the F338A, W86A, W86F and D74N mutant HuAChEs varied in their resistance to aging (15-3300-fold relative to the wild type). The most dramatic resistance to aging was observed for the phosphonyl conjugate of the mutant W86A enzyme (1850-3300-fold relative to the wild type). It is proposed that Trp-86 contributes to the aging process by stabilizing the evolving carbonium ion on the 1,2,2-trimethylpropyl moiety, via charge-pi interaction. The rate-enhancing effect of Trp-86 provides a rationale for the unique facility of aging in soman-inhibited cholinesterases, compared with the corresponding conjugates in other serine hydrolases. Replacements of Glu-202 by aspartic acid, glutamine or alanine residues resulted in a similar (1/130-1/300) decrease of the rates of aging. A comparable decrease was also observed for the conjugate of the F338A mutant. These results, and the similar pH dependence of aging rates for the wild-type and E202Q and F338A mutant HuAChEs, indicate that Glu-202 is not involved in proton transfer to the phosphonyl moiety. On the basis of these findings and of molecular modelling we suggest that Glu-202 and Phe-338 contribute to the aging process by stabilizing the imidazolium of the catalytic triad His-447 via charge-charge and charge-pi interactions respectively, thereby facilitating an oxonium formation on the phosphonyl moiety.

Vaccination by anthrax protective antigen (PA)-based vaccines requires multiple immunization, underlying the need to develop more efficacious vaccines or alternative vaccination regimens. In spite of the vast use of PA-based vaccines, the definition of a marker for protective immunity is still lacking. Here we describe studies designed to help define such markers. To this end we have immunized guinea pigs by different methods and monitored the immune response and the corresponding extent of protection against a lethal challenge with anthrax spores. Active immunization was performed by a single injection using one of two methods: (i) vaccination with decreasing amounts of PA and (ii) vaccination with constant amounts of PA that had been thermally inactivated for increasing periods. In both studies a direct correlation between survival and neutralizing-antibody titer was found (r 2 ‫؍‬ 0.92 and 0.95, respectively). Most significantly, in the two protocols a similar neutralizing-antibody titer range provided 50% protection. Furthermore, in a complementary study involving passive transfer of PA hyperimmune sera to naive animals, a similar correlation between neutralizing-antibody titers and protection was found. In all three immunization studies, neutralization titers of at least 300 were sufficient to confer protection against a dose of 40 50% lethal doses (LD 50 ) of virulent anthrax spores of the Vollum strain. Such consistency in the correlation of protective immunity with anti-PA antibody titers was not observed for antibody titers determined by an enzyme-linked immunosorbent assay. Taken together, these results clearly demonstrate that neutralizing antibodies to PA constitute a major component of the protective immunity against anthrax and suggest that this parameter could be used as a surrogate marker for protection.

Amino acids located within and around the ‘active site gorge’ of human acetylcholinesterase (AChE) were substituted. Replacement of W86 yielded inactive enzyme molecules, consistent with its proposed involvement in binding of the choline moiety in the active center. A decrease in affinity to propidium and a concomitant loss of substrate inhibition was observed in D74G, D74N, D74K and W286A mutants, supporting the idea that the site for substrate inhibition and the peripheral anionic site overlap. Mutations of amino acids neighboring the active center (E202, Y337 and F338) resulted in a decrease in the catalytic and the apparent bimolecular rate constants. A decrease in affinity to edrophonium was observed in D74, E202, Y337 and to a lesser extent in F338 and Y341 mutants. E202, Y337 and Y341 mutants were not inhibited efficiently by high substrate concentrations. We propose that binding of acetylcholine, on the surface of AChE, may trigger sequence of conformational changes extending from the peripheral anionic site through W286 to D74, at the entrance of the ‘gorge’, and down to the catalytic center (through Y341 to F338 and Y337). These changes, especially in Y337, could block the entrance/exit of the catalytic center and reduce the catalytic efficiency of AChE.

Several highly attenuated spore-forming nontoxinogenic and nonencapsulated Bacillus anthracis vaccines differing in levels of expression of recombinant protective antigen (rPA) were constructed. Biochemical analyses (including electrospray mass spectroscopy and N terminus amino acid sequencing) as well as biological and immunological tests demonstrated that the rPA retains the characteristics of native PA. A single immunization of guinea pigs with 5 ؋ 10 7 spores of one of these recombinant strains, MASC-10, expressing high levels of rPA (>100 g/ml) from a constitutive heterologous promoter induced high titers of neutralizing anti-PA antibodies. This immune response was long lasting (at least 12 months) and provided protection against a lethal challenge of virulent (Vollum) anthrax spores. The recombinant B. anthracis spore vaccine appears to be more efficacious than the vegetative cell vaccine. Furthermore, while results clearly suggest a direct correlation between the level of expression of PA and the potency of the vaccine, they also suggest that some B. anthracis spore-associated antigen(s) may contribute in a significant manner to protective immunity.The etiological agent of anthrax disease in animals and humans is the spore-forming bacterium Bacillus anthracis. The major factors of virulence of B. anthracis are located on two plasmids, pXO1 and pXO2. pXO2 encodes a poly-D-glutamic acid capsule (19, 41), while pXO1 encodes two binary exotoxins, the lethal toxin (LT) and the edema toxin (ET) (43,46,61). These two toxins are composed of three different proteins: protective antigen (PA), edema factor (EF), and lethal factor (LF) (for a review, see reference 36). PA is the common receptor binding domain of the toxins and can interact with the two different effector domains, EF and LF, to mediate their entry into target cells (14). EF is a calmodulin-dependent adenylate cyclase (37) responsible for the edema seen at the site of infection in experimental animals (17). The LF is a metalloprotease (34) recently shown to cleave the amino termini of the mitogen-activated protein kinase kinases 1 and 2, which results in their inactivation (13). It remains to be determined whether these are the main physiological substrates for the LT activity in vivo (5,22).Two types of anthrax vaccines are licensed for use in humans: the spores of the toxigenic, nonencapsulated B. anthracis STI-1 strain (55) and the cell-free PA-based vaccines consisting of aluminum hydroxide-adsorbed supernatant material from cultures of the toxigenic, nonencapsulated B. anthracis strain V770-NPI-R (49) or alum-precipitated culture filtrate from the Sterne strain (6). The use of the live attenuated STI-1 occasionally results in general and local adverse responses, observed both after primary application and revaccination, and the frequency of responses increases with the number of vaccinations (58). Furthermore, it was reported that the STI-1 vaccine has a relatively low immunogenicity (reviewed by Stepanov et al. in reference 58). To increase the i...

No abstract

The possible role of post-translational modifications such as subunit oligomerization, protein glycosylation and oligosaccharide processing on the circulatory life-time of proteins was studied using recombinant human acetylcholinesterase (rHuAChE). Different preparations of rHuAChE containing various amounts of tetramers, dimers and monomers are cleared at similar rates from the circulation, suggesting that oligomerization does not play an important role in determining the rate of clearance. An engineered rHuAChE mutant containing only one N-glycosylation site was cleared from the circulation more rapidly than the wild-type triglycosylated enzyme. On the other hand, hyperglycosylated mutants containing either four or five occupied N-glycosylation sites, analagous to those present on the slowly cleared fetal bovine serum acetylcholinesterase (FBS-AChE), were also cleared more rapidly from the bloodstream than the wild-type species. Furthermore, the two different tetraglycosylated mutants were cleared at different rates while the pentaglycosylated mutant exhibited the most rapid clearance profile. These results imply that though the number of N-glycosylation sites plays a role in the circulatory life-time of the enzyme, the number of N-glycan units in itself does not determine the rate of clearance. When saturating amounts of asialofetuin were administered together with rHuAChE, the circulatory half-life of the enzyme was dramatically increased (from 80 min to 19 h) and was found to be similar to that displayed by plasma-derived cholinesterases while desialylation of these enzymes caused a sharp decrease in the circulatory half-life to approximately 3-5 min. Determination of the average number of sialic acid residues per enzyme subunit of the five different N-glycosylation species generated, revealed that the rate of clearance is not a function of the absolute number of appended sialic acid moieties but rather of the number of unoccupied sialic acid attachment sites per enzyme molecule. Specifically, we demonstrate an inverse-linear relationship between the number of vacant sialic acid attachment sites and the values of the enzyme residence time within the bloodstream.

Substitution of Trp-86, in the active center of human acetylcholinesterase (HuAChE), by aliphatic but not by aromatic residues resulted in a several thousandfold decrease in reactivity toward charged substrate and inhibitors but only a severalfold decrease for noncharged substrate and inhibitors. The W86A and W86E HuAChE enzymes exhibit at least a 100-fold increase in the Michaelis-Menten constant or 100-10,000-fold increase in inhibition constants toward various charged inhibitors, as compared to W86F HuAChE or the wild type enzyme. On the other hand, replacement of Glu-202, the only acidic residue proximal to the catalytic site, by glutamine resulted in a nonselective decrease in reactivity toward charged and noncharged substrates or inhibitors. Thus, the quaternary nitrogen groups of substrates and other active center ligands, are stabilized by cation-aromatic interaction with Trp-86 rather than by ionic interactions, while noncharged ligands appear to bind to distinct site(s) in HuAChE. Analysis of the Y133F and Y133A HuAChE mutated enzymes suggests that the highly conserved Tyr-133 plays a dual role in the active center: (a) its hydroxyl appears to maintain the functional orientation of Glu-202 by hydrogen bonding and (b) its aromatic moiety maintains the functional orientation of the anionic subsite Trp-86. In the absence of aromatic interactions between Tyr-133 and Trp-86, the tryptophan acquires a conformation that obstructs the active site leading, in the Y133A enzyme, to several hundredfold decrease in rates of catalysis, phosphorylation, or in affinity to reversible active site inhibitors. It is proposed that allosteric modulation of acetylcholinesterase activity, induced by binding to the peripheral anionic sites, proceeds through such conformational change of Trp-86 from a functional anionic subsite state to one that restricts access of substrates to the active center.