Glycosyl transfer: a history of the concept's development and view of its major contributions to biochemistry.

Humans have fewer protein-coding genes than expected for all the inherent complexities of development. Supplementary factors include the post-translational modification of proteins by glycosylation. The latter term plus transglycosylation, glycosyltransferases, and 'to glycosylate' are used in biochemistry as though they always existed. Instead they have a history that can bring new insights in this science area to younger investigators. The present account describes five decades of findings and ideas on enzymic saccharide synthesis leading, finally to a rational theory that will surely continue to serve the biosciences well in the future.

Transition state structures for the hydrolysis of alpha-D-glucopyranosyl fluoride by retaining and inverting reactions of glycosylases.

Secondary tritium and primary 14C kinetic isotope effects were measured for the hydrolysis of alpha-D-glucopyranosyl fluoride catalyzed by sugar beet seed alpha-D-glucosidase, forming alpha-D-glucose, and by Rhizopus niveus glucoamylase forming beta-D-glucose. The data provided a novel opportunity to model and directly compare the transition state structures for the hydrolysis of a substrate promoted with retention or inversion of configuration according to the enzyme catalyst. The isotope effects for the reaction catalyzed by each enzyme are most consistent with an SN1 rather than an SN2 mechanism. The modeled transition state structures for the hydrolysis promoted by the alpha-glucosidase and the glucoamylase both bear significant oxocarbonium ion character, with the D-glucosyl residue having a flattened 4C1 conformation and a C-1-O-5 bond order of 1.92, even though opposite D-glucose anomers were produced from the substrate. The transition states show some modest differences, but their general similarity strongly suggests that the stereochemical outcome of glycosylase reactions does not predict the transition state structure, nor does the transition state structure of such reactions predict the stereochemical outcome. The results support previously reported evidence for the separate topological control of product configuration by protein structures in these and other glycosylases.

Crystal structures of soybean beta-amylase reacted with beta-maltose and maltal: active site components and their apparent roles in catalysis.

The crystal structures of catalytically competent soybean beta-amylase, unliganded and bathed with small substrates (beta-maltose, maltal), were determined at 1.9-2.2-A resolution. Two molecules of beta-maltose substrate bind to the protein in tandem, with some maltotetraose enzymic condensation product sharing the same binding sites. The beta-amylase soaked with maltal shows a similar arrangement of two bound molecules of 2-deoxymaltose, the enzymic hydration product. In each case the nonreducing ends of the saccharide ligands are oriented toward the base of the protein's active site pocket. The catalytic center, located between the bound disaccharides and found deeper in the pocket than where the inhibitor alpha-cyclodextrin binds, is characterized by the presence of oppositely disposed carboxyl groups of two conserved glutamic acid residues. The OE2 carboxyl of Glu 186 is below the plane of the penultimate glucose residue (Glc 2) of bound maltotetraose, 2.6 A from the oxygen atom of that ligand's penultimate alpha-1,4-glucosidic linkage. The OE2 carboxyl of Glu 380 lies above the plane of Glc 2, 2.8 A from the O-1 atom of the more deeply bound beta-maltose. Saccharide binding does not alter the spatial coordinates of these two carboxyl groups or the overall conformation of the 57-kDa protein. However, the saccharide complexes of the active enzyme are associated with a significant (10 A) local conformational change in a peptide segment of a loop (L3) that borders the active site pocket.(ABSTRACT TRUNCATED AT 250 WORDS)

The 2.0-A resolution structure of soybean beta-amylase complexed with alpha-cyclodextrin.

New crystallographic findings are presented which offer a deeper understanding of the structure and functioning of beta-amylase, the first known exo-type starch-hydrolyzing enzyme. A refined three-dimensional structure of soybean beta-amylase, complexed with the inhibitor alpha-cyclodextrin, has been determined at 2.0-A resolution with a conventional R-value of 17.5%. The model contains 491 amino acid residues, 319 water molecules, 1 sulfate ion, and 1 alpha-cyclodextrin molecule. The protein consists of a core with an (alpha/beta)8 supersecondary structure, plus a smaller globular region formed by long loops (L3, L4, and L5) extending from beta-strands beta 3, beta 4, and beta 5. Between the two regions is a cleft that opens into a pocket whose floor contains the postulated catalytic center near the carboxyl group of Glu 186. The annular alpha-cyclodextrin binds in (and partly projects from) the cleft with its glucosyl O-2/O-3 face abutting the (alpha/beta)8 side and with its alpha-D(1 --> 4) glucosidic linkage progression running clockwise as viewed from that side. The ligand does not bind deeply enough to interact with the carboxyl group of Glu 186. Rather, it occupies most of the cleft entrance, strongly suggesting that alpha-cyclodextrin inhibits catalysis by blocking substrate access to the more deeply located reaction center. Of the various alpha-cyclodextrin interactions with protein residues in loops L4, L5, L6, and L7, most notable is the shallow inclusion complex formed with Leu 383 (in L7, on the core side of the cleft) through contacts of its methyl groups with the C-3 atoms of four of the ligand's D-glucopyranosyl residues. All six residues of the bound alpha-cyclodextrin are of 4C1 conformation and are joined by alpha-1,4 linkages with similar torsional angles to form a nearly symmetrical torus as reported for crystalline inclusion complexes with alpha-cyclodextrin. We envision a significant role for the methyl groups of Leu 383 at the cleft entrance with respect to the productive binding of the outer chains of starch.

Stereochemistry of D-galactal and D-galacto-octenitol hydration by coffee bean alpha-galactosidase: insight into catalytic functioning of the enzyme.

Green coffee bean alpha-galactosidase was found to catalyze the hydration of D-galactal and (Z)-3,7-anhydro-1,2-dideoxy-D-galacto-oct-2-enitol (D-galacto-octenitol), each a known substrate for beta-galactosidase. The hydration of D-galactal by the alpha-galactosidase in D2O yielded 2-deoxy-2(S)-D-[2-2H]galactose; the hydration of D-[2-2H]galacto-octenitol in H2O yielded 1,2-dideoxy-2(R)-D-[2-2H]galactooct-3-ulose. Thus, the enzyme protonated each substrate from beneath the plane of the ring, as assumed for alpha-D-galactosides. These results provide an unequivocal assignment of the orientation of an acidic catalytic group to the alpha-galactosidase reaction center. In addition, they reveal a pattern of glycal/exocyclic enitol/glycoside protonation by the enzyme that differs from the pattern reported for beta-galactosidase and from that reported for alpha-glucosidases. Further findings show that D-galacto-octenitol is hydrated by the coffee bean alpha-galactosidase to form the alpha-anomer of 1,2-dideoxy-D-galactooctulose and by Escherichia coli beta-galactosidase to form the beta-anomer. That each enzyme converts this enolic substrate to a product whose de novo anomeric configuration matches that formed from its D-galactosidic substrates provides new evidence for the role of protein structure in controlling the steric outcome of reactions catalyzed by these and other glycosylases. The findings are discussed in light of the concept that catalysis by glycosidases involves a "plastic" protonation phase and a "conserved" product configuration phase.

Mechanism of maltal hydration catalyzed by beta-amylase: role of protein structure in controlling the steric outcome of reactions catalyzed by a glycosylase.

Crystalline (monomeric) soybean and (tetrameric) sweet potato beta-amylase were shown to catalyze the cis hydration of maltal (alpha-D-glucopyranosyl-2-deoxy-D-arabino-hex-1-enitol) to form beta-2-deoxymaltose. As reported earlier with the sweet potato enzyme, maltal hydration in D2O by soybean beta-amylase was found to exhibit an unusually large solvent deuterium kinetic isotope effect (VH/VD = 6.5), a reaction rate linearly dependent on the mole fraction of deuterium, and 2-deoxy-[2(a)-2H]maltose as product. These results indicate (for each beta-amylase) that protonation is the rate-limiting step in a reaction involving a nearly symmetric one-proton transition state and that maltal is specifically protonated from above the double bond. This is a different stereochemistry than reported for starch hydrolysis. With the hydration catalyzed in H2O and analyzed by gas-liquid chromatography, both sweet potato and soybean beta-amylase were found to convert maltal to the beta-anomer of 2-deoxymaltose. That maltal undergoes cis hydration provides evidence in support of a general-acid-catalyzed, carbonium ion mediated reaction. Of fundamental significance is that beta-amylase protonates maltal from a direction opposite that assumed for protonating starch, yet creates products of the same anomeric configuration from both. Such stereochemical dichotomy argues for the overriding role of protein structures in dictating the steric outcome of reactions catalyzed by a glycosylase, by limiting the approach and orientation of water or other acceptors to the reaction center.

Substrate-induced activation of maltose phosphorylase: interaction with the anomeric hydroxyl group of alpha-maltose and alpha-D-glucose controls the enzyme's glucosyltransferase activity.

Maltose phosphorylase, long considered strictly specific for beta-D-glucopyranosyl phosphate (beta-D-glucose 1-P), was found to catalyze the reaction beta-D-glucosyl fluoride + alpha-D-glucose----alpha-maltose + HF, at a rapid rate, V = 11.2 +/- 1.2 mumol/(min.mg), and K = 13.1 +/- 4.4 mM with alpha-D-glucose saturating, at 0 degrees C. This reaction is analogous to the synthesis of maltose from beta-D-glucose 1-P + D-glucose (the reverse of maltose phosphorolysis). In acting upon beta-D-glucosyl fluoride, maltose phosphorylase was found to use alpha-D-glucose as a cosubstrate but not beta-D-glucose or other close analogs (e.g., alpha-D-glucosyl fluoride) lacking an axial 1-OH group. Similarly, the enzyme was shown to use alpha-maltose as a substrate but not beta-maltose or close analogs (e.g., alpha-maltosyl fluoride) lacking an axial 1-OH group. These results indicate that interaction of the axial 1-OH group of the disaccharide donor or sugar acceptor with a particular protein group near the reaction center is required for effective catalysis. This interaction appears to be the means that leads maltose phosphorylase to promote a narrowly defined set of glucosyl transfer reactions with little hydrolysis, in contrast to other glycosylases that catalyze both hydrolytic and nonhydrolytic reactions.

Hydrolysis of beta-D-glucopyranosyl fluoride to alpha-D-glucose catalyzed by Aspergillus niger alpha-D-glucosidase.

Aspergillus niger alpha-D-glucosidase, crystallized and free of detectable activity for beta-D-glucosides, catalyzes the slow hydrolysis of beta-D-glucopyranosyl fluoride to form alpha-D-glucose. Maximal initial rates, V, for the hydrolysis of beta-D-glucosyl fluoride, p-nitrophenyl alpha-D-glucopyranoside, and alpha-D-glucopyranosyl fluoride are 0.27, 0.75, and 78.5 mumol.min-1.mg-1, respectively, with corresponding V/K constants of 0.0068, 1.44, and 41.3. Independent lines of evidence make clear that the reaction stems from beta-D-glucosyl fluoride and not from a contaminating trace of alpha-D-glucosyl fluoride, and is catalyzed by the alpha-D-glucosidase and not by an accompanying trace of beta-D-glucosidase or glucoamylase. Maltotriose competitively inhibits the hydrolysis, and beta-D-glucosyl fluoride in turn competitively inhibits the hydrolysis of p-nitrophenyl alpha-D-glucopyranoside, indicating that beta-D-glucosyl fluoride is bound at the same site as known substrates for the alpha-glucosidase. Present findings provide new evidence that alpha-glucosidases are not restricted to alpha-D-glucosylic substrates or to reactions providing retention of configuration. They strongly support the concept that product configuration in glycosylase-catalyzed reactions is primarily determined by enzyme structures controlling the direction of approach of acceptor molecules to the reaction center rather than by the anomeric configuration of the substrate.

Stereochemical course of hydrolysis and hydration reactions catalysed by cellobiohydrolases I and II from Trichoderma reesei.

Cellobiohydrolase I from Trichoderma reesei catalyzes the hydrolysis of methyl beta-D-cellotrioside (Km = 48 microM, kcat = 0.7 min-1) with release of the beta-cellobiose (retention of configuration). The same enzyme catalyzes the trans-hydration of cellobial (Km = 116 microM, kcat = 1.16 min-1) and lactal (Km = 135 microM, kcat = 1.35 min-1), presumably with glycosyl oxo-carbonium ion mediation. Protonation of the double bond is from the direction opposite that assumed for methyl beta-cellotrioside, but products formed from these prochiral substrates are again of beta configuration. Cellobiohydrolase II from the same microorganism hydrolyzes methyl beta-D-cellotetraoside (Km = 4 microM, kcat = 112 min-1) with inversion of configuration to produce alpha-cellobiose. The other reaction product, methyl beta-cellobioside, is in turn partly hydrolysed by cellobiohydrolase II to form methyl beta-D-glucoside and D-glucose, presumably the alpha-anomer. Reaction with cellobial is too slow to permit unequivocal determination of product configuration, but clear evidence is obtained that protonation occurs from the si-direction, again opposite that assumed for protonating glycosidic substrates. These results add substantially to the growing evidence that individual glycosidases create the anomeric configuration of their reaction products by means that are independent of substrate configuration.

Steric course of the hydrolysis of alpha,alpha-trehalose and alpha-D-glucosyl fluoride catalyzed by pig kidney trehalase.

We are unable to confirm the report of Labat et al.3 that pig kidney trehalase hydrolyzes alpha,alpha-trehalose to form solely alpha-D-glucose. Highly purified trehalase from pig renal cortex was found, in reactions monitored by 1H-n.m.r. spectra, to hydrolyze alpha,alpha-trehalose with the formation of both alpha- and beta-D-glucose. That the beta anomer constitutes the enzymically mobilized glucosyl residue is indicated by the further finding that beta-D-glucose is the product formed on hydrolysis of alpha-D-glucosyl fluoride by the enzyme. Present results show the stereochemical behavior of pig kidney trehalase in hydrolyzing alpha,alpha-trehalose to be indistinguishable from that reported by ourselves and others for trehalase preparations from a range of biological sources including rabbit renal cortex.

Alpha-secondary tritium kinetic isotope effects for the hydrolysis of alpha-D-glucopyranosyl fluoride by exo-alpha-glucanases.

alpha-Secondary tritium kinetic isotope effects ranging from 1.17 to 1.26 were measured for the hydrolysis of alpha-D-glucopyranosyl fluoride (forming beta-D-glucose) catalyzed by several glucoamylases and a glucodextranase. These results indicate that cleavage of the C-F bond is slow and that the enzymic transition state has significant oxo-carbonium ion character. Strong support for this conclusion is provided by the agreement found in the case of Rhizopus niveus glucoamylase (alpha-TV/K 1.26; Km 26 mM) between measured values of the alpha-secondary deuterium kinetic isotope effects (alpha-DV/K 1.16; alpha-DV 1.20) and those calculated from the tritium isotope effect. The data are consistent with the promotion of an intramolecular elimination of fluoride by the present exo-alpha-glucanases based on their ability to stabilize, perhaps with a counter ion, the development of a carbonium ion-like transition state. Although the oxo-carbonium ion is formally denoted as an intermediate it could represent a transition state along a reaction pathway to a covalent glucosyl intermediate.

Synthesis of (Z)-3,7-anhydro-1,2-dideoxy-2-deuterio-D-gluco-oct-2- enitol, a prochiral substrate for probing the catalytic functioning of of glucosylases.

Synthesis of the title compound provides a prochiral, glycosyl-donor substrate well suited for use as a probe of the catalytic functioning of D-glucosyl-mobilizing enzymes, because the full stereochemistry of enzymic reactions at its double bond may be unambiguously determined by examining the reaction products. The starting material for the synthesis was 2,6-anhydro-D-glycero-D-gulo-heptonic acid, from which 3,7-anhydro-4,5,6,8-tetra-O-benzyl-1-deoxy-D-glycero-D-gulo-2- octulose was prepared in eight steps. Reduction with lithium aluminum deuteride, and conversion of the resulting diastereomeric alcohols into (Z)-3,7-anhydro-4,5,6,8-tetra-O-benzyl-1,2-dideoxy-2-deuterio-D- gluco-oct-2-enitol (11) and 3,7-anhydro-4,5,6,8-tetra-O-benzyl-1,2-dideoxy-2-deuterio-D- glycero-D-gulo-oct-1-enitol (16), was carried out. By-products were 3,7-anhydro-2-O-benzoyl-4,5,6,8-tetra-O-benzyl-1,2-dideoxy-2-deuterio -D-erythro-L-galacto-octitol and 3,7-anhydro-2-O-benzoyl-4,5,6,8-tetra-O-benzyl-1,2-dideoxy-2-deuterio -D-erythro-L-talo-octitol, which could, like compound 16, be recycled. On debenzylation the oct-2-enitol 11 yielded (Z)-3,7-anhydro-1,2-dideoxy-2-deuterio-D-gluco-oct-2-enitol.

Steric course of the hydration of D-gluco-octenitol catalyzed by alpha-glucosidases and by trehalase.

Crystalline Aspergillus niger alpha-glucosidase and highly purified preparations of rice alpha-glucosidase II and Trichoderma reesei trehalase were found to catalyze the hydration of [2-(2)H]-D-gluco-octenitol, i.e., (Z)-3,7-anhydro-1,2-dideoxy-[2-2H]-D-gluco-oct-2-enitol, to yield 1,2-dideoxy-[2-2H]-D-gluco-octulose. In each case, the stereochemistry of the reaction was elucidated by examining the newly formed centers of asymmetry at C-2 and C-3 of the hydration product. The C-1 to C-3 fragment of each isolated [2-2H]-D-gluco-octulose product was recovered as [2-2H]propionic acid and identified by its positive optical rotatory dispersion as the S isomer, showing that each enzyme had protonated the octenitol (at C-2) from above its re face. 1H NMR spectra of enzyme/D-gluco-octenitol digests in D2O showed that the alpha-anomer of [2-2H]-D-gluco-octulose was exclusively produced by each alpha-glucosidase, whereas the beta-anomer was formed by action of the trehalase. The trans hydration catalyzed by the alpha-glucosidases was found to be very strongly inhibited by the substrate; the cis hydration reaction catalyzed by the trehalase showed no such inhibition. Special importance is attached to the finding that in hydrating octenitol each enzyme creates a product of the same anomeric form as in hydrolyzing an alpha-D-glucosidic substrate. This result adds substantially to the growing evidence that individual glycosylases create the configuration of their reaction products by a means that is independent of donor substrate configuration, that is, by a means other than "retaining" or "inverting" substrate configuration.

Stereochemical studies of D-glucal hydration by alpha-glucosidases and exo-alpha-glucanases: indications of plastic and conserved phases in catalysis by glycosylases.

Alpha-Glucosidases from Aspergillus niger, pig serum, ungerminated rice, buckwheat, and sugar beet seeds (but not from brewers' yeast or honeybee) were found to catalyze the hydration of D-glucal. Each reactive alpha-glucosidase, incubated with D-glucal in D2O, was shown to protonate (deuteriate) this prochiral substrate from above its re face, i.e., from a direction opposite that assumed for protonating alpha-D-glucosidic substrates. At the same time, D-glucal hydration catalyzed by three of the alpha-glucosidases that acted rapidly enough in D2O to determine product configuration was found to yield 2-deoxy-D-glucose of the same specific (alpha-) configuration as the D-glucose produced from alpha-D-glucosidic substrates. These findings substantially extend those reported earlier for the hydration of D-glucal by one (Candida tropicalis) alpha-glucosidase preparation. Together with other recent results, they suggest that the process of catalysis by alpha-glucosidases (and perhaps glycosylases in general) may comprise two separate and separately controlled parts, namely, a "plastic" phase concerned with substrate protonation and a substrate-unrelated "conserved" phase concerned with the creation of product configuration. In contrast to the alpha-glucosidases, three "inverting" exo-alpha-glucanases (Arthrobacter globiformis glucodextranase; Rhizopus niveus and Paecilomyces varioti glucoamylase) were found to protonate D-glucal from below its si face. Further, whereas the catalysis of D-glucal hydration by the alpha-glucosidases was intensively inhibited by excess substrate, that promoted by the exo-glucanases showed no detectable substrate inhibition.

Catalytic versatility of Bacillus pumilus beta-xylosidase: glycosyl transfer and hydrolysis promoted with alpha- and beta-D-xylosyl fluoride.

Bacillus pumilus beta-xylosidase, an enzyme considered restricted to hydrolyzing a narrow range of beta-D-xylosidic substrates with inversion of configuration, was found to catalyze different stereochemical, essentially irreversible, glycosylation reactions with alpha- and beta-D-xylopyranosyl fluoride. The enzyme promoted the hydrolysis of beta-D-xylopyranosyl fluoride at a high rate, V = 6.25 mumol min-1 mg-1 at 0 degrees C, in a reaction that obeyed Michaelis-Menten kinetics. In contrast, its action upon alpha-D-xylopyranosyl fluoride was slow and characterized by an unusual relation between the rate of fluoride release and the substrate concentration, suggesting the possible need for two substrate molecules to be bound at the active center in order for reaction to occur. Moreover, 1H NMR spectra of a digest of alpha-D-xylosyl fluoride showed the substrate to be specifically converted to alpha-D-xylose by the enzyme. The observed retention of configuration is not consistent with direct hydrolysis by this "inverting" enzyme but is strongly indicative of the occurrence of two successive inverting reactions: xylosyl transfer from alpha-D-xylosyl fluoride to form a beta-D-xylosidic product, followed by hydrolysis of the latter to produce alpha-D-xylose. The transient intermediate product formed enzymically from alpha-D-xylosyl fluoride in the presence of [14C]xylose was isolated and shown by its specific radioactivity and 1H NMR spectrum as well as by methylation and enzymic analyses to be 4-O-beta-D-xylopyranosyl-D-xylopyranose containing one [14C]xylose residue.(ABSTRACT TRUNCATED AT 250 WORDS)

Hydration of cellobial by exo- and endo-type cellulases: evidence for catalytic flexibility of glycosylases.

New insight has been obtained into the catalytic capabilities of cellulase. Essentially homogeneous preparations of exo- (or Avicelase-) type and endo- (or CMCase-) type cellulases from Irpex lacteus and Aspergillus niger, respectively, were shown to hydrate the enolic bond of cellobial to form 2-deoxycellobiose. The A. niger enzyme also synthesized a small amount of a 2-deoxycellobiosyl-transfer product from cellobial. By use of digests conducted in deuterated buffer and 1H NMR spectra for product analysis, both cellulases were found to protonate (deuterate) the double bond of cellobial from below the si face of the D-glucal moiety, i.e., from a direction opposite that assumed for protonation of the beta-D-glycosidic linkages of cellulose and cellodextrins. The exo enzyme, which hydrolyzes the latter substrates primarily to cellobiose, rapidly catalyzed cellobial hydration to produce the beta-anomer of beta-D-glucopyranosyl(1----4)-2-deoxy-D-glucose-2(e)-d. The A. niger cellulase produced the same 2-deoxycellobiose-d from cellobial, though too slowly for its configuration to be determined. However, evidence was obtained for the formation of a beta-2-deoxycellobiosyl-d-D-glucose-transfer product by the enzyme. Thus, it is likely that all of the observed reactions with cellobial represent trans additions at the double bond. In any case, the anomeric configuration of products is created de novo. Separate mechanisms are described for the reaction of cellobial hydration and for the stereochemically different reaction of cellulose hydrolysis catalyzed by the present enzymes, assuming an arrangement of their catalytic groups analogous to that found in lysozyme.(ABSTRACT TRUNCATED AT 250 WORDS)

Catalytic flexibility of glycosylases. The hydration of maltal by beta-amylase to form 2-deoxymaltose.

Crystalline, alpha-glucosidase-free sweet potato beta-amylase was found to catalyze hydration of the enolic bond of maltal (alpha-D-glucopyranosyl-(1----4)-2-deoxy-D-glucal) to form 2-deoxymaltose (alpha-D-glucopyranosyl-(1----4)-2-deoxy-D-glucose). The reaction at pH 5.0 showed Vmax 0.082 mumol/min/mg and km 94.5 mM. An exceptionally large solvent deuterium isotope effect, VH/VD = 8, was observed from pH(pD) 4.2 to 5.4; and at pH(pD) 5.0 the effect was found to be directly related to the mole fraction of 2H. The hydration product, isolated from a beta-amylase/maltal digest in acetate-d4/D2O buffer (pD 5.4) was identified through its 1H NMR spectrum as alpha-D-glucopyranosyl-(1----4)-2-deoxy-D-[2(a)-2H]glucose. beta-Amylase in 2H2O thus catalyzes deuteration of the double bond of maltal from a direction opposite that assumed for protonation of the glycosidic oxygen atoms of starch chains and maltosaccharides. This finding confirms the functional flexibility of the enzyme's catalytic groups first demonstrated in studies of the reactions catalyzed with alpha- and beta-maltosyl fluoride (Hehre, E. J., Brewer, C. F., and Genghof, D. S. (1979) J. Biol. Chem. 254, 5942-5950). A possible mechanism of the maltal hydration by beta-amylase involves protonation of substrate from above as the first and rate-limiting step, followed by formation of a transient carbonium ion-enzyme intermediate. Although other possible mechanisms cannot be ruled out, it is clear that this hydration reaction differs from reactions catalyzed with amylaceous substrates and with alpha- and beta-maltosyl fluoride. The ability of beta-amylase to catalyze different types of reactions with different substrates is discussed with respect to observations with other enzymes that, likewise, strongly support the view (Hehre et al.) that the catalytic groups of glycosylases in general may be functionally flexible beyond requirements of the principle of microscopic reversibility.

Catalytic versatility of trehalase: synthesis of alpha-D-glucopyranosyl alpha-D-xylopyranoside from beta-D-glucosyl fluoride and alpha-D-xylose.

Trehalase was previously shown (see ref. 5) to hydrolyze alpha-D-glucosyl fluoride, forming beta-D-glucose, and to synthesize alpha, alpha-trehalose from beta-D-glucosyl fluoride plus alpha-D-glucose. Present observations further define the enzyme's separate cosubstrate requirements in utilizing these nonglycosidic substrates. alpha-D-Glucopyranose and alpha-D-xylopyranose were found to be uniquely effective in enabling Trichoderma reesei trehalase to catalyze reactions with beta-D-glucosyl fluoride. As little as 0.2mM added alpha-D-glucose (0.4mM alpha-D-xylose) substantially increased the rate of enzymically catalyzed release of fluoride from 25mM beta-D-glucosyl fluoride at 0 degrees. Digests of beta-D-glucosyl fluoride plus alpha-D-xylose yielded the alpha, alpha-trehalose analog, alpha-D-glucopyranosyl alpha-D-xylopyranoside, as a transient (i.e., subsequently hydrolyzed) transfer-product. The need for an aldopyranose acceptor having an axial 1-OH group when beta-D-glucosyl fluoride is the donor, and for water when alpha-D-glucosyl fluoride is the substrate, indicates that the catalytic groups of trehalose have the flexibility to catalyze different stereochemical reactions.

Factors determining steric course of enzymic glycosylation reactions: glycosyl transfer products formed from 2,6-anhydro-1-deoxy-D-gluco-hept-1-enitol by alpha-glucosidases and an inverting exo-alpha-glucanase.

Glycosyl transfer products were formed from 2,6-anhydro-1-deoxy-D-gluco-hept-1-enitol (heptenitol) by purified alpha-glucosidases from Candida tropicalis and rice and by an inverting exo-alpha-glucanase (glucodextranase) from Arthrobacter globiformis. The products were structurally defined through 1H and 13C NMR (nuclear magnetic resonance) spectra of their crystalline per-O-acetates in comparison with those of authentic methyl 1-deoxy-alpha- and methyl 1-deoxy-beta-D-gluco-heptuloside. 1-Deoxy-alpha-D-gluco-heptulosyl-(2 leads to 7)-heptenitol and 1-deoxy-alpha-D-gluco-heptulosyl-(2 leads to 7)-D-gluco-heptulose were produced by both the Candida alpha-glucosidase and the glucodextranase; 1-deoxy-alpha-D-gluco-heptulosyl-(2 leads to 5)- and 1-deoxy-alpha-D-gluco-heptulosyl-(2 leads to 7)-D-gluco-heptuloses by the rice alpha-glucosidase. These results, together with our earlier findings of sterospecific hydration of heptenitol catalyzed by the same enzymes [Hehre, E. J., Brewer, C. F., Uchiyama, T., Schlesselmann, P., & Lehmann, J. (1980) Biochemistry 19, 3557-3564], show the inadequacy of the long-accepted notion that carbohydrase-catalyzed reactions always lead to retention (or always lead to inversion) of substrate configuration. In particular, the finding that glucodextranase forms transfer products of alpha configuration and a hydration product of beta configuration from the same substrate provides a clear example of the functioning of acceptors rather than donor substrates in selecting the steric course of reactions catalyzed by a glycosylase. The circumstances under which acceptor cosubstrates might be expected to show this significant effect are discussed. The opportunity presumably would exist whenever carbonium ion mediated reactions are catalyzed by glycosylases that provide oppositely oriented approaches of different acceptors to the catalytic center.