Mechanism-based glycoside hydrolase inhibitors are carbohydrate analogs that mimic the natural substrate’s structure. Their covalent bond formation with the glycoside hydrolase makes these compounds excellent tools for chemical biology and potential drug candidates. Here we report the synthesis of cyclohexene-based α-galactopyranoside mimics and the kinetic and structural characterization of their inhibitory activity toward an α-galactosidase from Thermotoga maritima (TmGalA). By solving the structures of several enzyme-bound species during mechanism-based covalent inhibition of TmGalA, we show that the Michaelis complexes for intact inhibitor and product have half-chair (2H3) conformations for the cyclohexene fragment, while the covalently linked intermediate adopts a flattened half-chair (2H3) conformation. Hybrid QM/MM calculations confirm the structural and electronic properties of the enzyme-bound species and provide insight into key interactions in the enzyme-active site. These insights should stimulate the design of mechanism-based glycoside hydrolase inhibitors with tailored chemical properties.

Glycoside hydrolases (GHs) catalyze hydrolyses of glycoconjugates in which the enzyme choreographs a series of conformational changes during the catalytic cycle. As a result, some GH families, including α-amylases (GH13), have their chemical steps concealed kinetically. To address this issue for a GH13 enzyme, we prepared seven cyclohexenyl-based carbasugars of α-d-glucopyranoside that we show are good covalent inhibitors of a GH13 yeast α-glucosidase. The linear free energy relationships between rate constants and pK a of the leaving group are curved upward, which is indicative of a change in mechanism, with the better leaving groups reacting by an SN1 mechanism, while reaction rates for the worse leaving groups are limited by a conformational change of the Michaelis complex prior to a rapid SN2 reaction with the enzymatic nucleophile. Five bicyclo[4.1.0]­heptyl-based carbaglucoses were tested with this enzyme, and our results are consistent with pseudoglycosidic bond cleavage that occurs via SN1 transition states that include nonproductive binding of the leaving group to the enzyme. In total, we show that the conformationally orthogonal reactions of these two carbasugars reveal mechanistic details hidden by conformational changes that the Michaelis complex of the enzyme and natural substrate undergoes which align the nucleophile for efficient catalysis.

Positive charge stabilized on remote C5-allylic center with catalysis occurring via a loose SN2 transition state.

No abstract

Hydrolyses of cyclohexenyl-based carbasugars that mimic either α-Dglucose or α-D-galactose were explored with two Bacteroides thetaiotaomicron enzymes from glycoside hydrolase family 97: an inverting α-glucosidase (BtGH97a) and a retaining αgalactosidase (BtGH97b). Both enzymes yield nucleophilic substitutions at the pseudoanomeric center of the carbasugar substrates, giving significantly different linear energy relationships for the catalytic rate constant as a function of the leaving group ability. Specifically, the kinetic data for the inverting α-glucosidase is consistent with the reaction giving a hydrolyzed inverted carbaglucose product by a mechanism that proceeds with little nucleophilic participation by the bound water molecule at the reaction transition state. In contrast, the reaction of carbagalactose substrates with the retaining GH97 enzyme involves a rate-limiting nonchemical step, likely a conformational change, followed by rapid substitution involving a nucleophilic amino acid residue to give a covalently bound intermediate. Considering the structural similarities between these two GH97 enzymes, the kinetic data nonetheless reveal a significant (>10 6 ) difference in the rates of nucleophilic attack between the unique enzymatic nucleophileswith the less nucleophilic species being H 2 O in the inverting α-glucosidase and the better nucleophile being a carboxylate in the retaining αgalactosidase. The enzymatic rate constant ratio for the phenyl carbasugars contrasts with the corresponding kinetic data obtained using natural substrate phenyl glycopyranosides. Last, for the galactocarbasugar with a phenol leaving group, the second-order rate constant for alkylation of the GH97 α-galactosidase is only ∼10-fold lower than that for glycosylation of this enzyme by the parent carbohydrate phenyl α-D-galactopyranoside. This modest difference in rate constants underscores our conclusion that retaining glycoside hydrolases may not have optimized the nucleophilicity of their active site nucleophiles with the result that the transition state free energies for formation and hydrolysis of the covalent enzyme intermediate are matched.

No abstract

In this study, we look at how a catalytically efficient α-galactosidase stabilizes transition state (TS) charge delocalization for substrate hydrolysis. We then assess whether covalent inhibition of the enzyme by three types of mechanism-based covalent inhibitors occurs via similar modes of TS stabilization. We show, using Bartlett-type linear free energy relationships, that good correlations are obtained between the catalytic efficiencies (k cat /K m and/or k inact /K i ) for enzyme-catalyzed reactions of natural and activated galactoside substrates and of representatives of three families of classical mechanism-based inhibitors: a 2-deoxy-2-fluoroglycoside, allylic carbasugars, and an epoxy carbasugar. Of note, we show that glycoside natural substrates and allylic carbasugars display log(rate)−log(rate) correlations that are unity (slope ≈ 1), an observation consistent with them having identical positive charge stabilization at the S N 1-like glycosylation and pseudo-glycosylation TSs, respectively. In contrast, 2-deoxy-2-fluoroglycoside mechanism-based inhibitors react via a different enzyme-catalyzed mechanism (S N 2), while the strained epoxy carbasugar inactivates the α-galactosidase by traversing a TS in which the glycoside hydrolase stabilizes the inactivation TS that has a significantly lower degree of charge stabilization to those for the natural glycoside substrates. To add weight to these conclusions, we computed free energy landscapes and their associated galactosylation and pseudogalactosylation TSs using QM/MM molecular dynamics methods with the whole solvated enzyme.

No abstract