Acid-base catalysis in Leuconostoc mesenteroides sucrose phosphorylase probed by site-directed mutagenesis and detailed kinetic comparison of wild-type and Glu237-->Gln mutant enzymes

Abstract

The role of acid-base catalysis in the two-step enzymatic mechanism of alpha-retaining glucosyl transfer by Leuconostoc mesenteroides sucrose phosphorylase has been examined through site-directed replacement of the putative catalytic Glu237 and detailed comparison of purified wild-type and Glu237-->Gln mutant enzymes using steady-state kinetics. Reactions with substrates requiring Brønsted catalytic assistance for glucosylation or deglucosylation were selectively slowed at the respective step, about 10(5)-fold, in E237Q. Azide, acetate and formate but not halides restored catalytic activity up to 300-fold in E237Q under conditions in which the deglucosylation step was rate-determining, and promoted production of the corresponding alpha-glucosides. In situ proton NMR studies of the chemical rescue of E237Q by acetate and formate revealed that enzymatically formed alpha-glucose 1-esters decomposed spontaneously via acyl group migration and hydrolysis. Using pH profiles of kcat/K(m), the pH dependences of kinetically isolated glucosylation and deglucosylation steps were analysed for wild-type and E237Q. Glucosylation of the wild-type proceeded optimally above and below apparent pK(a) values of about 5.6 and 7.2 respectively whereas deglucosylation was dependent on the apparent single ionization of a group of pK(a) approximately 5.8 that must be deprotonated for reaction. Glucosylation of E237Q was slowed below apparent pK(a) approximately 6.0 but had lost the high pH dependence of the wild-type. Deglucosylation of E237Q was pH-independent. The results allow unequivocal assignment of Glu237 as the catalytic acid-base of sucrose phosphorylase. They support a mechanism in which the pK(a) of Glu237 cycles between approximately 7.2 in free enzyme and approximately 5.8 in glucosyl enzyme intermediate, ensuring optimal participation of the glutamate residue side chain at each step in catalysis. Enzyme deglucosylation to an anionic nucleophile took place with Glu237 protonated or unprotonated. The results delineate how conserved active-site groups of retaining glycoside hydrolases can accommodate enzymatic function of a phosphorylase.

Figure 1. Reaction mechanisms of retaining (trans)glycosidases and catalytic mutants thereof

(A) Double-displacement mechanism of an α-retaining enzyme, which interconverts glycosides with the leaving groups R₁-OH and R₂-OH. In the case of SPase R₁-OH is D-fructose and R₂-OH is phosphate. (B) Two-step (Ping Pong Bi Bi) kinetic mechanism of SPase. (C) Functional complementation of an otherwise inactive mutant in which the acid–base catalyst (an aspartate or glutamate residue) was replaced by a non-functional residue (in the present study, a glutamine residue). Nu is an external nucleophile such as azide, formate, acetate and halides.

Figure 2. (A) Comparison of the region containing the catalytic acid–base of different members of family GH-13 and (B) SDS/PAGE of purified wild-type LmSPase and E237Q

(A) The sequences shown are LmSPase (D90314), BaSPase (AF543301), amylosucrase from Neisseria polysaccharea (NpASase, AJ011781) and 4-α-glucantransferase from Thermotoga maritima (TmGTase, AE000512). The alignment was performed with the Vector NTI program using the AlignX-modul with the PAM250 scoring matrix. Conserved amino acids are shaded in black (100% identity), dark grey (80–100% identity) and light grey (60–80% identity). (B) Lane 1, wild-type; lane 2, E237Q; lane 3, molecular mass standard. The staining of protein bands was performed with Coomassie Blue.

Figure 3. ¹H NMR monitoring of product formation during E237Q-catalysed αG1P conversion in the presence of sodium acetate (A) and the corresponding reaction scheme (B)

The primary product is α-glucose 1-acetic acid ester [δ_H 6.05 (1 H, d, J 3.6 Hz, H-1), 2.14 (3 H, s, CO-CH₃)]. Migration of the acetyl group and mutarotation lead to accumulation of 2-O-acetyl-α-D-glucose [δ_H 5.32 (1 H, d, J 3.8 Hz, H-1), 2.118 (3 H, s, CO-CH₃)] and 2-O-acteyl-β-D-glucose [δ_H 4.76 (1 H, d, J 8.1 Hz, H-1), 2.121 (3 H, s, CO-CH₃) not shown]. Consecutively, the α- and β-anomers of 3-O-acetyl-D-glucose and 6-O-acetyl-D-glucose are also formed in smaller amounts. Signals of their acetate groups appear in the range of 2.14 p.p.m. to 2.05 p.p.m. and the α-anomeric protons cause signals between 5.25 p.p.m. and 5.20 p.p.m.. All β-anomeric proton signals are partly overlapped by the overwhelming HDO signal and not shown. 4-O-Acetyl-D-glucose has not been formed in detectable amounts after 16 h.

Figure 4. ¹H NMR monitoring of product formation during E237Q-catalysed αG1P conversion in the presence of sodium formate (A) and the corresponding reaction scheme (B)

αG1P [δ_H 5.41 (1 H, dd, J 7.2, 3.4 Hz, H-1)]. The primary product is α-glucose 1-formic acid ester [δ_H 8.21 (1 H, s, CHO), 5.36 (1H, d, J 3.8 Hz, H-1)]. Migration of the formyl group and mutarotation consequently lead to formation of small amounts of α- and β-anomers of 2-,3- or 4-O-acetyl-D-glucose. The signals of the corresponding formyl groups are in the range of 8.31 p.p.m. to 8.25 p.p.m., and one α-anomeric proton signal appears at 5.25 p.p.m. (indicated by asterisks). The further α-anomeric and all β-anomeric proton signals are overlapped by other signals. 6-O-Formyl α-D-glucose [δ_H 8.15 (1 H, s, CHO), 5.18 (1 H, d, J 3.8 Hz, H-1), 4.42 (dd, 12.7, 2.4 Hz, 1H, H-6a), 4.38 (dd, 12.7, 4.7 Hz, 1H, H-6b)] and 6-O-formyl β-D-glucose [δ_H 8.15 (1 H, s, CHO), 4.63 (1 H, d, J 8.0 Hz, H-1), 4.47 (1 H, dd, J 12.4, 1.9 Hz, H-6a), 4.32 (1 H, dd, J 12.7, 5.3 Hz, H-6b)] have been accumulated in detectable amounts prior to hydrolysis to formate and free glucose. The signal at 8.24 p.p.m. is a ¹³C-satellite of the formate signal at 8.41 p.p.m..

Figure 5. pH profiles for k_cat (A) and k_cat/K_m (B) of the phosphorolysis (●) and synthesis (○) reaction catalysed by wild-type LmSPase

Initial rates were obtained in 20 mM Mes buffer at 30 °C. Control experiments for the phosphorolysis reaction were performed in 20 mM Mes/sodium acetate (pH 4.5–6.0) and 50 mM Tes (pH 6.5–8.0) buffer (■). Solid lines are fits of eqn (2) or eqn (3) to the data as described in the Experimental section.

Figure 6. pH profiles for k_cat (A) and k_cat/K_m (B and C) of the arsenolysis of αG1P catalysed by the wild-type (●) and the E237Q mutant (○)

Initial rates were obtained in 20 mM Mes buffer at 30 °C. Control experiments for wild-type LmSPase were performed in 20 mM Mes/sodium acetate (pH 4.5–6.0) and 50 mM Tes (pH 6.5–8.0) buffer (■). k_cat/K_m values were corrected for the fraction of protonated αG1P (B) or arsenate (C) respectively. Solid lines are fits of eqn (2) or eqn (3) to the data as described in the Experimental section. Dotted lines show the trend of the uncorrected data of k_cat/K_m for the wild-type (▼) and the E237Q mutant (▽).