ABSTRACT
Glu-237 (homologous to Glu-232) was replaced by Gln, so that the carboxamide group of the Gln cannot fulfill the acid-base catalytic function proposed for the carboxylic group of the original Glu (Scheme 5.3). Detailed kinetic characterization of E237Q showed that catalytic steps involving a poor leaving group or nucleophile were slowed by five orders of magnitude in the mutein as compared to the wild-type enzyme whereas other steps expected to proceed without support from an acid-base catalyst were not strongly affected by the site-directed substitution (Schwarz et al., 2007b). The activity of E237Q (measured as conversion of a-d-glucose 1-phosphate) was stimulated in the presence of anionic nucleophiles such as azide, acetate and formate. It was shown that this “chemical rescue” of E237Q resulted because external nucleophiles reacted more rapidly than water with the glucosylated mutein. Formation of the resulting a-glucosyl products (a-glucose 1-azide, a-glucose 1-acetate) was confirmed by NMR. Results of pH studies suggested a role for Glu-237 in which the carboxylic acid/carboxylate side chain of the Glu alternates between catalytic functions of a general acid and base in glucosylation and deglucosylation half-reactions of sucrose phosphorylase, respectively. It was proposed that the pKa of the Glu-237 side chain cycles between 7.2 in the free enzyme and 5.8 in the glucosyl-enzyme intermediate (Schwarz et al., 2007b).During analysis of the participation of Glu-237 in each step of catalysis the question arose as to how glucosylated sucrose phosphorylase manages to accommodate a phosphate ion in the active site, while the unprotonated side chain of the Glu remains in place. An observation often made with retaining glycoside hydrolases and transglycosidases is that nucleophilic anions are excluded from reactions with glycosyl-enzyme intermediates, probably because of unfavorable electrostatic interactions with the negative charge on the Brønsted catalytic group of the enzyme. Recent work has shown that Asp-338 and Tyr-340 have key roles in a differential binding mechanism utilized by L. mesenteroides sucrose phosphorylase to bring about specificity for the fructose leaving group and the phosphate nucleophile, respectively, during the catalytic steps of phosphorolysis of sucrose (Mueller and Nidetzky, 2007a). The results support a suggestion for the reaction cycle of sucrose phosphorylase derived from the crystal structure of the enzyme from B. adolescentis, where movements of active site entrance loops cause
corresponding Asp-342 and Tyr-344to alternate between positions in and out of the catalytic centre, allowing the phosphorylase to switch between accommodation of fructose and phosphate during enzyme glucosylation and deglucosylation, respectively (Mirza et al., 2006).An auxiliary catalytic function was proposed for Asp-295. The carboxylate side chain of this Asp was suggested to facilitate the steps of glucosylation and deglucosylation of Asp-196through a strong hydrogen bond (≥ 23 kJ/mol) with the equatorial 2-hydroxyl of the glucosyl oxocarbenium ion-like species, thought to be formed in the transition states flanking the b-glucosyl enzyme intermediate (Mueller and Nidetzky, 2007b). Crystal structures of B. adolescentis sucrose phosphorylase portray a partial conformational itinerary for the glucosyl moiety along the reaction pathway, revealing relaxed 4C1 chair conformations in enzyme-bound sucrose (substrate) and glucose (hydrolysis product) as well as a 1,4B boat conformation for the covalent intermediate (Mirza et al., 2006). The immediate vicinity of 4C1 and 1,4B conformations to a 4H3 conformation in a pseudo-rotational cycle of glucopyranose strongly suggests that the half-chair conformation (4H3) is used for the oxocarbenium ion-like transition state of sucrose phosphorylase (Davies et al., 2003; Grindley, 2001). A tentative explanation for why the stabilization provided by Asp-295 is apparently selective for the transition state of the reaction is that the conformational change 4C1→4H3 in the glucosylation step and 1,4B→4H3 (probably via a 1S3 skew boat) in the deglucosylation step contributes to significant shortening, hence strengthening of the proposed hydrogen bond between the side chain of Asp-295 and the sugar 2-hydroxyl. Electronic effects such as gradual depression of the pKa of the 2-hydroxyl in response to progressive generation of partial positive charge at the anomeric carbon as heterolysis of the exocyclic C-1-O bond takes place could also be responsible for optimization of this hydrogen bond in the transition state.The key catalytic amino acids in L. mesenteroides sucrose phosphorylase (Asp-196, Glu-237) were replaced by the corresponding carboxamide residues, and it was shown that the D196N E237Q double mutein was capable of promoting slow phosphorolysis of a-glucose 1-fluoride with retention of configuration to yield a-d-glucose 1-phosphate as the product (Goedl and Nidetzky, 2009). a-Glucose 1-fluoride is an excellent
alternative glucosyl donor substrate for the enzyme that is utilized by the wild-type phosphorylase with a catalytic efficiency exceeding that for reaction with the natural substrate sucrose. Unlike sucrose, a-glucose 1-fluoride does not rely on participation from a Brønsted catalytic acid for its conversion and was therefore a suitable choice of substrate for assaying muteins that have Glu-237 replaced. The stereochemical course of the reaction of D196N E237Q was remarkable in consideration of the absence of an enzyme catalytic nucleophile. Interestingly, therefore, the double mutein was shown to react kinetically through a ternary complex of phosphorylase, a-glucose 1-fluoride and phosphate, in stark contrast to the Ping-Pong kinetic mechanism of the wild-type enzyme that reflects reaction in two catalytic steps via enzyme glucosylation and deglucosylation (see Scheme 5.3). D196N E237Q displayed a very low level of activity for phosphorolysis of a-methyld-glucoside, indicating that the mutein retained the ability to convert an O-glucosidic substrate. Using remodeling of the active site of sucrose phosphorylase in the form of D196N E237Q, an enzymatic reaction coordinate for a-retaining glucosyl transfer via a covalent intermediate was changed for the first time into one where this intermediate was lacking and configurational retention appeared to be achieved through a direct front-side nucleophilic displacement reaction.Note: retentive phosphorolysis of b-glucosidic substrates by C. fimi NagZ is expected to also proceed by a double displacement-like catalytic mechanism where a fully conserved Asp would serve as the catalytic nucleophile (Macdonald et al., 2015). The catalytic acid-base is a His (Bacik et al., 2012). 5.3.3 Retaining a,a-Trehalose PhosphorylaseAccording to Table 5.1, retaining a,a-trehalose phosphorylase (family GT-4) must be distinguished clearly from “counterpart” inverting trehalose phosphorylase (family GH-65) (Van Hoorebeke et al., 2010b). No crystal structure of a retaining trehalose phosphorylase has been determined so far. However, the structures of several glycosyltransferases related to the trehalose phosphorylase by common membership to family GT-4 have been solved (Chua et al., 2008; Guerin et al., 2007; Martinez-Fleites et al., 2006; Ruane et al.,
2008; Steiner et al., 2010; Vetting et al., 2008). A three-dimensional model of trehalose phosphorylase from Schizophyllum commune has thus been created based on this structural basis. In spite of its possible ambiguity, the model is clear in showing that the active site of the trehalose phosphorylase (Fig. 5.3A) is very different from the active site of sucrose phosphorylase (Fig. 5.2).However, active site similarity between trehalose phosphorylase and glycogen phosphorylase is suggested (Fig. 5.3). Mutagenesis data confirm the importance of Lys-512 and Arg-507 as key phosphate-binding residues for the activity of trehalose phosphorylase (Goedl and Nidetzky, 2008). Both trehalose and glycogen phosphorylase lack a suitable amino acid in their respective active site whose side chain would be positioned appropriately to be a candidate catalytic nucleophile. A direct front-side nucleophilic displacement mechanism was therefore suggested for both enzymes. In literature, this mechanism is sometimes referred to as “internal return-like” (SNi-like) and is considered to be a strong proposal in an ongoing debate about the catalytic mechanism of retaining glycosyltransferases (Lairson et al., 2008; Ardèvol and Rovira, 2015). Note: a recent summary of mechanistic considerations for enzymatic glycosyl transfer with retention of configuration is given in Ardèvol and Rovira (2015).
