Development of Tunable Mechanism-Based Carbasugar Ligands that Stabilize Glycoside Hydrolases through the Formation of Transient Covalent Intermediates

Abstract

Mutations in many members of the set of human lysosomal glycoside hydrolases cause a wide range of lysosomal storage diseases. As a result, much effort has been directed toward identifying pharmacological chaperones of these lysosomal enzymes. The majority of the candidate chaperones are active site-directed competitive iminosugar inhibitors but these have met with limited success. As a first step toward an alternative class of pharmacological chaperones we explored the potential of small molecule mechanism-based reversible covalent inhibitors to form transient enzyme-inhibitor adducts. By serial synthesis and kinetic analysis of candidate molecules, we show that rational tuning of the chemical reactivity of glucose-configured carbasugars delivers cyclohexenyl-based allylic carbasugar that react with the lysosomal enzyme β-glucocerebrosidase (GCase) to form covalent enzyme-adducts with different half-lives. X-ray structural analysis of these compounds bound noncovalently to GCase, along with the structures of the covalent adducts of compounds that reacted with the catalytic nucleophile of GCase, reveal unexpected reactivities of these compounds. Using differential scanning fluorimetry, we show that formation of a transient covalent intermediate stabilizes the folded enzyme against thermal denaturation. In addition, these covalent adducts break down to liberate the active enzyme and a product that is no longer inhibitory. We further show that the one compound, which reacts through an unprecedented S_N1'-like mechanism, exhibits exceptional reactivity-illustrated by this compound also covalently labeling an α-glucosidase. We anticipate that such carbasugar-based single turnover covalent ligands may serve as pharmacological chaperones for lysosomal glycoside hydrolases and other disease-associated retaining glycosidases. The unusual reactivity of these molecules should also open the door to creation of new chemical biology probes to explore the biology of this important superfamily of glycoside hydrolases.

Figure 1

Small molecule covalent inhibitors and the mechanism of action of retaining glycosidases. (a) Mechanism used by glucocerebrosidase for hydrolysis of its natural β-glucoside substrate. (b) Kinetic scheme for GCase-catalyzed turnover of β-d-glucopyranoside substrates. (c) Mechanism of action for carbasugar covalent inhibitors. (d) Kinetic scheme for covalent carbasugar inhibitors.

Figure 2

Structures of carbasugars used in this study and their in vitro characterization. Carbasugar covalent inhibitors used in the current study. The pseudoanomeric carbons (shown by the asterisk on 1) are drawn in identical positions for all compounds.

Scheme 1. Synthesis of d- and l-Carbaxylopyranosyl Halides; (a) Synthesis of Intermediates 14 and 15, (b) Synthesis of Intermediates 16 and 23, (c) Synthesis of Target Carbasugars

Reagents: (i) NaIO₄, H₂O, rt; (ii) Ph₃P⁺CH₃, BuLi, 0 °C; (iii) dioxane/water, cat H⁺, 90 °C; (iv) CH₂CHMgBr, THF, 0 °C; (v) Grubbs’ 2nd generation, CH₂Cl₂, heat; (vi) CH₃COCOCH₃, HC(OCH₃)₃, cat H⁺; (vii) DAST, diglyme, 0 °C; (viii) TFA, CH₂Cl₂ (for 4 and 5 from 20 and 21, respectively); (ix) Dess-Martin periodinane, CH₂Cl₂; (x) NaBD₄ (MeOD); (xi) TFA, CH₂Cl₂, H₂O; (xii) BCl₃, CH₂Cl₂, −78 °C (for 6, 7, 6-D, and 7-D from 26 to 29, respectively). Abbreviations: DAST, (diethylamino)sulfur trifluoride; TFA, trifluoroacetic acid.

Figure 3

In vitro characterization of carbasugars used in this study. (a) Nonlinear least-squares fit of k_app values to a modified Michaelis–Menten expression (eq S3) for the approach to steady-state GCase activity on incubation with d-carbaxylosyl fluoride 4. (b) Linear fit of k_app values for the approach to steady state for covalent inhibition of GCase by d-carbaxylosyl chloride 6 (blue) and l-carbaxylosyl chloride 7 (red). (c) Stabilization of folded GCase measured using differential scanning fluorimetry (DSF) in the presence of d-carbaxylosyl chloride 6 (blue), l-carbaxylosyl chloride 7 (red), or vehicle along (black). All lines are the best nonlinear least-squares fit to the appropriate equation. All experiments were carried out at T = 25 °C in McIlvaine buffer pH 5.2 containing 0.1% Triton X-100, 0.24% sodium taurodeoxycholate, and 0.1% BSA. (d) Mechanisms for the GCase pseudoglycosylation step by single turnover chaperones 6 and 7, which after diffusion of chloride out of the active site give identical allylic cation intermediates that are trapped by the enzymatic nucleophile (Glu340) to give the same covalent intermediate (E-αCarb).

Figure 4

Structural characterization of covalent intermediates and Michaelis complex between GCase and carbasugars. (a) Crystal structure of the GCase-6 complex, showing two orientations of the covalent complex in which 6 adopts an envelope (²E) conformation. The maximum-likelihood σ_A-weighted 2F_o – F_c electron density map is contoured to 1 σ (0.37 e/ų). (b) Crystal structure of the GCase-7 complex, showing two orientations of the covalent complex in which, the carbasugar adopts an envelope (²E) conformation, numbering of the ring starts at the covalent attachment to the enzyme. The maximum-likelihood σ_A-weighted 2F_o – F_c electron density map is contoured to 1 σ (0.36 e/ų). Carbasugar average b-factor = 20 Ų. (c) Crystal structure of the GCase-5 complex, showing three orientations of l-carbasugar fluoride bound noncovalently in the active site in a half-chair (⁴H₃) conformation. Green atom indicates fluorine. The maximum-likelihood σ_A-weighted 2F_o – F_c electron density map is contoured to 1 σ (0.37 e/ų).

Figure 5

pH-rate profile for the logarithm of the inactivation second-order rate constants (k_inact/K_i) for yeast α-glucosidase by 7. All experiments were performed at T = 37 °C in 50 mM buffer (acetic acid-sodium acetate buffer for pH 5.0–5.5, sodium phosphate buffer for pH 6.0–7.5, TAPS buffer for pH 7.9) containing 0.1% BSA. The line through the points is the best fit to a model where two protonation events (low pH) lead to an inactive enzyme, while a single deprotonation at higher pH values gives inactive enzyme.