INTRODUCTION

Yeast Saccharomyces cerevisiae expresses invertase (- fructofuranoside hydrolase, EC) as a major external form and a minor internal form (1). Both enzymes are produced from the same SUC2 gene, but are translated from different start codons (2, 3, 4) resulting in the presence of two additional amino acids on the amino end of external invertase (5). Unlike internal invertase, which remains in the cytosol, the external enzyme gains entry into the endoplasmic reticulum owing to the presence of a signal sequence. During secretion and transport to the periplasmic space via the endoplasmic reticulum, the enzyme acquires an average of ten mannose oligosaccharides to account for 50% (w/w) of its molecular mass. The mannose oligosaccharides appear to aid in the self-association of the functionally active homodimer into a tetramer, hexamer, and octamer, all with the same specific activity (6, 7, 8).

Invertase readily hydrolyzes sucrose and, to a lesser extent, inulin and the raffinose family of substrates. Koshland and Stein (9) proposed a double displacement catalytic mechanism for invertase, which suggests the involvement of a nucleophile and a proton donor in a two-step reaction. Attempts to identify those amino acids essential for invertase activity led to the implication of a carboxylate group, in addition to histidine, as well as methionine and tryptophan residues (10, 11, 12). Braun (13) showed that a carboxylate group forms an ester linkage with conduritol B epoxide, resulting in the irreversible inactivation of the enzyme. We have identified this nucleophile as Asp-23 by characterization of conduritol B epoxide-labeled peptides and confirmed the essential nature of this amino acid by converting it to an asparagine residue by site-directed mutagenesis (14). The present paper extends this study by clarifying the proton donor or acid/base catalyst involved in hydrolyzing the substrate.

MATERIALS AND METHODS

The assay mixture (500 µl) containing 0.05 M sodium acetate, pH 5.0, and 200 mM sucrose was incubated with 0.1-0.5 unit of enzyme at 37 °C. After 10 min, 150 µl of 0.5 M sodium phosphate, pH 7.1, was added and the reaction was terminated by heating in a boiling water bath for 5 min. An aliquot was analyzed for glucose by a glucose oxidase-peroxidase-coupled colorimetric reaction (15). Enzyme units are expressed in micromoles of sucrose hydrolyzed per minute and are based on initial velocity measurements. k_cat was determined using 120,000 as the M_r of a functional unit of invertase, which in this case is a dimer. The kinetics of native and mutant invertases were carried out using sucrose concentrations from 5 to 200 mM. The kinetic parameters were determined by nonlinear regression analysis of the substrate versus velocity plot using the ENZFIT program.

External invertase was purchased from Boehringer and Mannheim and purified (16). Iodine monochloride, prepared according to the method of Izzo et al. (17), was used to inactivate invertase. ICl (0.1 mM) was prepared by diluting a 20 mM stock with 50 mM sodium acetate buffer, pH 5.5. Invertase monomer (13 µM, in a final reaction volume of 50 µl) was taken up in 10 mM sodium acetate, pH 5.5, and chilled on ice. A 1-8-fold molar excess of freshly diluted chilled ICl was added to the enzyme. The reaction was quenched after 5 min by adding 2 µl of 0.5 M sodium thiosulfate.

Varying concentrations of I₂ in KI were used to inactivate larger amounts of invertase than used above. The reaction mixture consisted of 10 mM sodium acetate, pH 5.5, 67 µM invertase monomer, and I₂ in KI ranging from 1- to 4-fold molar excess over invertase monomer in a total volume of 500 µl. Stock iodine reagent (20 mM I₂ in 100 mM KI) was diluted with an equal volume of 50 mM sodium acetate, pH 5.5, for use in the reaction. Following incubation for 5 min on ice, 20 µl of 0.5 M sodium thiosulfate was added to quench traces of unused iodine reagent. An aliquot of the enzyme was diluted 200-fold with 50 mM sodium acetate, pH 5.0, and assayed for activity to follow the extent of inactivation. The modified enzyme was dialyzed in 50 mM Tris-Cl, pH 7.4, and concentrated in a Centricon 30 (Millipore). The enzyme concentration was determined from A₂₈₀ (1 mg/ml) = 2.25, with a monomer molecular weight = 60,000.

The cysteine content of invertase was determined under denaturing conditions with DTNB¹ (Sigma), added to an invertase solution (~20 µM) in a final volume of 1 ml of 0.1 M Tris-Cl, pH 7.4, containing 8 M urea to a concentration of 1 mM. The resulting yellow color was read periodically at 412 nm against a control containing only buffer and the thiol reagent. The number of SH groups was determined using a molar absorbance coefficient of 13,600 (18).

Gly mutations were introduced into the four His sites in invertase using the single-stranded uracil-containing pYES-SUC DNA as template, according to the procedures described in the Mutagene kit (Bio-Rad). The shuttle vector pYES-SUC was constructed by inserting the SUC2 coding sequence into the HindIII site of pYES2.0 (Invitrogen). The single-stranded uracil-containing pYES-SUC DNA was prepared from Escherichia coli CJ 236 (dut ung, thi, relA; pCJ105(Cmr)). The following oligonucleotides (mismatches underlined) were used in mutagenesis: H13G, CCTTTGGTCCTTCACACCC; H37G, GCCAAATGGTCTGTACTTTC; H57G, GTTTTGGGGCTGCTACTTCCG; and H250G, TCAATGGTACTTTTTGAAGCGT. Following complementary strand synthesis and ligation using T7 DNA polymerase and T4 DNA ligase, respectively, the DNA was transformed into MV1190 cells. The histidine mutant invertases were expressed in Y334 yeast cells (MAT pep4-3 prb1-1122 ura3-52 leu2-3 112reg1-501 gal1).

The yeast shuttle vector pEI13, which differs from pEI23 (14) by a single base change at position 124 of the SUC2 coding sequence, expresses fully active invertase. For mutagenesis of the two cysteine residues, the SUC2 coding sequence was excised from pEI13 as a 2-kilobase pair HindIII fragment and subcloned into M13mp19.

The reverse oligonucleotides (base changes underlined) CAAATCGCAACTCTTTGTCTTGGA, CAAACCTGGATTCGTATTGGTAGCC, and ACCTGGACACGTATTGGTAGCCT were used to mutagenize wild type invertase to C108A, C205A, and E204A, respectively. The mutagenesis protocol employed is described in the Sculptor kit (Amersham Corp.). Following mutagenesis the SUC2 coding sequence was excised with HindIII and subcloned back into the HindIII site of pEI13. The plasmid was transformed into HB101 bacterial cells, and the colonies containing the insert in translational orientation were selected by hybridization with a ³²P-labeled overlapping oligonucleotide. For expression, pEI13 DNA was transformed into yeast 275 (MATa ura3-52 leu2-3, 112ade2-101 suc2 del9) using a CaCl₂-lithium acetate-dithiothreitol protocol (19).

The yeast 334 histidine mutant invertase clones, selected from ura selective medium plates, were grown in 5 ml of synthetic dextrose medium (2% dextrose, 0.67% yeast nitrogen base without amino acids) supplemented with amino acids and adenine (uracil omitted) at 30 °C overnight on a shaker. A 0.5-ml aliquot of each of the overnight cultures was added to 5 ml of YPD medium and grown until A₆₀₀ = 1.0-1.5, at which time 1% galactose was added to induce invertase; growth was continued for another 20 h. Controls were maintained by growing 334 cells (without plasmid) in YPD medium. For comparison of activities about 3 ml of the cultures containing equivalent absorbance units were harvested. The cells were washed twice with 0.05 M sodium acetate, pH 5.0, and suspended in 500 µl of 0.5 M sodium acetate, pH 5.0, 5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and ruptured in a miniBead Beater (BioSpec) for 2 min with 0.5-mm beads. The extracts were microcentrifuged, and the activity in the supernatants was expressed as percent of the wild type enzyme.

The cysteine and glutamate mutant invertase clones were gown initially in leucine-selective synthetic dextrose medium (100 ml). These cultures were inoculated into 2 liters of YPD medium and grown to saturation at 30 °C. The cells were harvested and suspended in 1 liter of half-strength YPD containing a 0.2% sucrose and shaken at 30 °C for a day to induce invertase. The harvested cells (60 g) were suspended in 90 ml of 0.05 M Tris-HCl, pH 7.4, 5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride and disrupted for 3 min with an equal volume of 0.5-mm glass beads (prewetted with buffer above) in the Bead Beater with a cooling interval after each one minute of use. The nucleoproteins in the crude extract were precipitated with 10 ml of 6% streptomycin sulfate solution. The clear solution was then heated at 50 °C for 20 min, and the precipitated protein was removed. Ammonium sulfate was added to 50% saturation to remove additional unwanted protein. The supernatant was then dialyzed in 0.01 M potassium phosphate, pH 7.5, and loaded onto a DE-52 column (2.5 × 25 cm), which was eluted with 600 ml of a 0-0.2 M NaCl linear gradient. Invertase fractions were concentrated by ultrafiltration on a PM30 membrane and purified on a Sephacryl S-500 column (1.6 × 60 cm) using 0.02 M sodium acetate, pH 5.0, 0.5 M NaCl as the eluant. The Sephacryl S-500 column was coupled to a Beckman HPLC system, and the eluant was monitored at 229 nm. In the presence of chloride ions, dimeric invertase associates into higher oligomers and elutes earlier than the contaminating proteins. Since E204A invertase has little activity, the enzyme was monitored through the purification steps by an immunoblot procedure using the ProtoBlot kit (Promega). All the purification steps, with the exception of heat treatment and Sephacryl S-500 column chromatography, were carried out at 4 °C. A typical yield was about 11 mg.

Amino acid analyses were performed using a Beckman System Gold analyzer. Protein sequences homologous to invertase were retrieved using the Nentrez program. Amino acid maximum homology comparisons of invertase with other hydrolases were performed using Prosis (Pharmacia Biotech Inc.).

RESULTS

The general glycosylhydrolase reaction mechanism as proposed originally by Koshland and Stein (9) involves the protonation of the glycosidic oxygen by a proton donor followed by a nucleophilic attack on the anomeric carbon of the sugar substrate by a carboxylate group. Several studies suggested that one of the four histidines in invertase could be the suggested proton donor. The inactivation of invertase by diazonium tetrazole was interpreted as indicative of histidine or tyrosine involvement (11), but since nitration experiments showed that tyrosines were nonessential, histidine appeared as the most likely proton donor. Inactivation by methyl mercuric iodide coupled to other observations led to the proposal that histidine and methionine were chelated by this reagent (12). The pH activity curve in the presence and absence of the affinity label conduritol B epoxide yielded pK_a values 3.8 and 6.8, with the latter value suggesting again the involvement of a histidine (13). In view of these observations, we investigated the potential role of histidine using site-directed mutagenesis. Each of the four invertase histidines (His-13, His-37, His-57, His-250) was mutated to a glycine and the mutant enzymes were expressed from the Gal1 promoter of pYES2.0 (see ``Materials and Methods''). Since the activities of the mutant invertases ranged from 92 to 100% of the native enzyme (data not shown), a direct role for histidine in the catalytic process appears to be ruled out.

Earlier studies reported that iodine inactivated approximately 55% of invertase (20, 21). A later study by Waheed and Shall (11) supported this observation and, in addition, showed that -mercaptoethanol fully restored the activity of iodine-inactivated invertase. Based on the inactivation of invertase by cyanogen bromide and the reactivation of iodine-treated invertase with -mercaptoethanol, these authors suggested that iodine might reversibly oxidize an essential methionine residue. To investigate this possibility, we hydrolyzed invertase and iodine-modified invertase in 15% NaOH and analyzed the hydrolyzate in each case for methionine sulfoxide. Neither sample showed an appreciable amount of methionine sulfoxide, eliminating methionine as a modified residue. Also the absence of absorbance changes at 280 nm, or fluorescence emission changes of the iodine-modified invertase excluded oxidation of tryptophan residues.

Since the only residues not clearly tested for their effect on invertase activity were the two free cysteines (Cys-108 and Cys-205), the effect of sulfhydryl reagents on enzyme activity was assessed. At concentrations of 5 mM HgCl₂ and pCMS, invertase was inactivated by 90% and 63%, respectively, suggesting the possible involvement of SH groups in the catalytic mechanism. Because the previously reported (11, 20, 21) inactivation of invertase by iodine could be due to modification of the cysteine sulfhydryls, we investigated whether there was a correlation between loss in activity and loss in SH content on treatment with iodine. Fig. 1 shows that invertase was 94% inactivated at a 4-fold molar excess of ICl in sodium acetate buffer, pH 5.5, at 4 °C. At an 8-fold molar excess, invertase was completely inactivated (data not shown), which differs from the earlier reports that iodine inactivated only about half of the enzyme activity (20, 21). The iodination reaction was rapid, taking less than 1 min either at 4 °C or 23 °C. The inactivation of invertase by I₂ in KI closely followed the loss of two cysteinyl residues as determined by DTNB analysis. The almost complete inactivation of invertase at a 2-fold excess of I₂ in KI suggests that the only amino acid modified is cysteine. To determine whether the iodine reagent iodinates the cysteine residues or oxidizes them, invertase was reacted with ¹²⁵ICl. The lack of incorporation of ¹²⁵I into the enzyme (data not shown) eliminated cysteinyl iodide formation as a product of the reaction. It is also evident from Fig. 1 that ICI is less effective than I₂ in inactivating the enzyme at comparable concentrations (61% versus 95% at a 2-fold molar excess of each).

To determine whether one or both cysteinyl residues were essential for activity, each cysteine was mutagenized to an alanyl reside. Table I shows the activity of various invertases in relation to their sulfhydryl content. Iodination of C108A invertase resulted in the complete loss of enzyme activity, which correlated with the loss of a single sulfhydryl group. The iodine inactivation of native invertase and its C108A variant was completely reversed upon addition of dithiothreitol. This reversal was accompanied by the restoration of about two SH groups in native invertase and one in the C108A mutant, that involving Cys-205. The C205A mutant, however, retained only about 30% the activity of native invertase, which was unaffected by iodination (Table I), although Cys-108 was completely oxidized. The fact that C108A retains all of its activity while C205A retains only 30% suggests the greater involvement of the latter in either the catalytic process or more likely in a structural alteration of the protein affecting substrate binding. This thesis is borne out by the reduced catalytic efficiency of C205A (Table II) relative to C108A. It is interesting to note that, although Cys-205 is unlikely to be involved directly in catalysis, this residue is conserved in the glycosylhydrolase sequences listed in Table III.

Table I. Effect of cysteine modification on the activity of native and mutant invertases Enzyme treatmenta SH residues/monomerb Activity % Invertase 1.90 100 Iodinated invertase 0.10 1 Iodinated invertase + dithiothreitol 1.94 96 C108A invertase 1.04 100 Iodinated C108A invertase 0.00 0 Iodinated C108A invertase + dithiothreitol 1.18 94 C205A invertase 0.97 32 Iodinated C205A invertase 0.00 30 a Invertase was reacted with I2 in KI as described under ``Materials and Methods.'' The iodine-inactivated enzyme was reactivated by incubation in 0.05 M Tris-HCl, pH 7.6 + 20 mM dithiothreitol for 3 h at 37 °C. b The SH groups were estimated by DTNB (see ``Materials and Methods'') in the presence of 8 M urea, since these groups are buried and unreactive in the native enzyme.

Table II. Comparison of the catalytic properties of mutant invertases with native invertase Invertase and its variants kcata Km kcat/Km min-1 mm (m-1min-1) Native 5.66  ± 0.50  × 105 26.1  ± 2.0 2.16  × 107 C108A 5.49  ± 0.51  × 105 29.3  ± 2.2 1.87  × 107 C205A 1.54  ± 0.54  × 105 43.8  ± 0.7 3.51  × 106 E204A 1.90  ± 0.20  × 102 27.5  ± 0.6 6.90  × 103 a The kinetics were carried out at pH 5.0 with purified enzymes as described under ``Materials and Methods.'' The values represent an average of two experiments.

Table III. Homology comparison of the catalytic residues of invertase with other fructosylhydrolases Enzyme Source Sequence comparison NCBI Invertase S. cerevisiae GWM NPN GLWYa GYQY C PGLIEb 3834c Inulinase Kluyveromyces marxianus GWM NDPN GLWY GTQY EC PGLVK 322904 Invertase Pichia anomala GWM NDPN GMFY GFQY EC PGLFK 562772 Invertase Zymomonas mobilis GWM NDPN GLIF AFMW EC PDFFS 217200 Invertase Potato NWI NDPN APMY TGNW EC PDFFP 313129 Invertase Maize NWI NDPN APLY TGMW EC PDFFP 736359 Invertase Tomato NWM NDPN GPLY TGMW EC VDFYP 124701 Invertase Fava beans NWI NDPN GPMY TGMW EC PDFYP 861155 Invertase Carrot NWM NDPN GPLF TGMW EC VDFYP 480998 Invertase Mung bean NWM NDPN GPMY TGMW EC VDFFP 218326 Invertase Arabidopsis thaliana NWM NDPN GPMI SGMW EC PDFFP 899153 Invertase Escherichia coli GNM NDPN GLIY SYMW EC PDFFR 131824 Sucrose-6-PO4-Dd Vibrio cholerae GLL NDPN GFIY GYMW EC PDWFE 155269 Sucrase Klebsiella pneumoniae GLL NDPN GFCQ GYMW EC PDLFP 134320 Sucrase Bacillus subtilis GLL NDPN GVIY GYMW EC PDLFS 143482 Sucrase Pediococcus pentosaceus GLL NDPN GFSY GYMI EC PKSGL 475113 Sucrase Vibrio alginolyticus GLL NDPN GLCY GYMW EC PDFFE (Ref. 41) Sucrase Lactobacillus lactis GLL NDPN GFSY GYMI EC PNLIF 417755 Sucrase Streptococcus mutans GLL NDPN GFSY EYMI EC PNLVF (Ref. 42) Fructanase Streptococcus mutans GWA NDPN GLVY DLHT EC PDMYP 477562 Levanase Bacteroides fragilis GWM NDAN GLVY PSQR EC PDMVE 143972 Levanase Bacillus subtilis NWM NDPN GMVY GGVW EC PDLFE 134175 Fructosyl-Td Bacillus polymyxa NWM NDPN GLVY RGIF EC PDIFR 480970 a Residue underlined represents Asp-23. b Residue underlined represents Glu-204. c National Center for Biotechnology Information sequence identification number. d Abbreviations: sucrose-6-phosphate dehydrogenase; this enzyme has a 25% homology to yeast invertase, which is in the same range as the plant fructosylhydrolases. Whether this enzyme has hydrolase activity in view of the homology of its putative active site region to the other hydrolases is not known.

Glycosylhydrolases are classified into more than 45 families based on their sequence homology (22, 23). Family 32 includes fructosylhydrolases such as invertase (sucrase), levanase, fructanase, and inulinase. The sequence surrounding the catalytic Asp-23 residue of yeast invertase is conserved in family 32 enzymes. In fact the sequence N-D-P-N is invariant in all but levanase from Bacteroides fragilis, where Pro is replaced by Ala (Table III). Interestingly, residues comparable to Glu-204/Cys-205 are conserved in all of the enzymes, while Glu-204/Cys-205/Pro-206 is conserved in all but higher plant invertases. A single Glu has been established as a proton donor in lysozyme (24) and in other glycosylhydrolases (25). The pK_a of this Glu is shown to be abnormally high, ranging from 6 to 7 in lysozyme, cyanogenic -glucosidase and xylanase (26, 27, 28). Since one of the catalytic residues in yeast invertase has been reported to have a pK_a of 6.8 (12), and since C205A mutagenesis failed to profoundly affect the activity, we investigated the possible involvement of Glu-204 as a proton donor by mutagenizing it to Ala.

The kinetic parameters of the mutant invertases shown in Table II reveal that of the mutants examined, E204A shows the most dramatic change, with about a 3000-fold reduction in k_cat. Since its K_m is similar to that of the native enzyme, it is obvious that Glu-204 plays a key role in the catalytic step. There was little change in the k_cat/K_m of C108A relative to native invertase and the 6-fold reduction in the k_cat/K_m of C205A is still too small, relative to E204A, to implicate Cys-205 in the catalytic process, except perhaps in an ancillary role.

A study of the pH dependence of inactivation of invertase by conduritol B epoxide suggests that two ionizable groups with pK_a 3.05 and 6.8, a nucleophile and a proton donor, respectively, are involved in catalysis (12). We have investigated the pH dependence of the specificity constant k_cat/K_m of native and mutant invertases to gain an insight into the nature of the catalytic residues associated with the free enzyme (Fig. 2). The native enzyme and C205A yielded broad bell-shaped curves from which pK_a values of 3.7 and 3.5, respectively, could be calculated. These values most likely represent the ionizable side chains of the nucleophile, Asp-23. The upper limb of the curves yielded a pK_a of 6.2 for both enzymes, the ionizable pH range wherein a carboxyl from Glu-204 could act as a general base. Attempts to verify this with E240A proved difficult because of the rather low activity of this mutant, although the apparently higher pK_a of about 6.8 reflects a change in the active site geometry. Conceivably the small, but still measurable activity of E204A is the result of vicinal residues compensating for the loss of Glu-204. A plot of log k_cat versus pH (not presented) of each protein revealed a slightly lower pK_a for the basic limb of the enzyme-substrate complex of 5.8-6.0, which in effect could be due to an alteration in the conformation of the protein in the presence of the substrate.

DISCUSSION

Based on previous chemical modification experiments with invertase, histidine and methionine (10, 11, 12) were suggested as participating in the catalytic process as a proton donor and a catalytic residue, respectively. In the present study, these residues were systematically eliminated as participants in catalysis. Asp-23 was shown by us in an earlier study (14) to be involved in the catalytic reaction as a nucleophile based on its irreversible reaction with the affinity label conduritol B epoxide, and mutagenesis to Asn, which resulted in the complete inactivation of invertase. We have now found that Glu-204 is also essential for the hydrolysis of sucrose by replacing it with an alanyl residue, which in effect led to a 3,000-fold reduction in k_cat. Iodine inactivation experiments suggested that one or both cysteines might be involved in the reaction, but the C108A mutation revealed no loss in activity and the C205A mutation was associated with a 70% decrease in activity (Table I). This result suggests that Cys-205 is in some way involved in the catalytic reaction, which is not surprising in view of its proximity to Glu-204. Possibly Cys-205 plays a supportive role in catalysis by maintaining a suitable microenvironment in the active site, or perhaps in aiding substrate binding.

Why iodine totally inactivates invertase at about equimolar concentrations, whereas the C205A mutation abolishes only 70% of the activity, is not entirely clear. Iodine, depending upon its concentration, could oxidize cysteine to cysteinesulfenic acid or cysteinesulfinic acid. The restoration of both sulfhydryl groups and full invertase activity by dithiothreitol suggests oxidation of cysteine to cysteinesulfenic acid rather than cysteinesulfinic acid, since the latter is resistant to reduction back to cysteine. In addition, since a 2-fold excess of iodine is sufficient to oxidize almost two cysteine residues, the oxidation could not have proceeded to a sulfinic acid. Cysteinesulfenic acid, being unstable, usually forms a disulfide with a neighboring cysteine residue, or condenses with a second sulfenic acid to yield a thiosulfinate (29). However, since the reaction of either cysteine mutant of invertase with iodine results in the modification of the remaining single cysteine residue, both of the above possibilities would appear to be excluded. Conceivably the adjacent cysteinesulfenic acid residue impairs the proton donating capacity of Glu-204 through electrostatic interactions resulting in the observed loss in enzyme activity. It is of interest to note that the abnormally high pK_a of 6.2 for Glu-204 (Fig. 2) is not unusual. Thus, in lysozyme and xylanase (26, 28), the high pK_a of Glu is due to electrostatic interactions with other carboxylates within the enzyme structure, while in -glucosidase the high pK_a of Glu is attributed to the hydrophobic microenvironment of this enzyme (27).

A comparison of yeast invertase with the fructosylhydrolases in Table III shows significant sequence homologies around the catalytic residues corresponding to Asp-23 and Glu-204. The Glu-Cys-Pro sequence is conserved in all but three plant enzymes, which have invariant Glu-Cys residues, again suggesting a minor role for Cys in the catalytic process. Inulinase from Kluyveromyces marxianus and invertase from Pichia anomala share 53% and 44% homology, respectively, and contain a single cysteine residue that corresponds to Cys-205 in invertase. Surprisingly, yeast invertase is closer to levanases and fructosyltransferase from bacteria (34-35% homology) than bacterial sucrases (30-31% homology) and plant invertases (27-29% homology). There are no conserved regions comparable to yeast invertase in levansucrases from Acetobacter diazotrophicus (NCBI 1050187), Bacillus subtilis (NCBI 732568), Erwinia amylovora (NCBI 433559), and Zymomonas mobilis (NCBI 809530), although levansucrase from Bacillus amyloliquefaciens (30) has been included in family 32 glycosylhydrolases.

The catalytic mechanism of glycosylhydrolases in most cases involves two acidic residues, one acting as a nucleophile and the other as a proton donor (31, 32, 33). Those hydrolases that retain the anomeric configuration follow a double displacement mechanism involving an enzyme-covalent intermediate, whereas with inverting enzymes a water molecule activated by the nucleophile is involved in the release of products in one step. The distance between the proton donor and nucleophile is 5.5 Å in the enzyme that retains configuration, while this distance is increased to 10 Å in the case of the inverting enzyme, reflecting the accommodation of a water molecule. A probable reaction mechanism for invertase, a retaining enzyme (34), is depicted in Fig. 3. Glu-204 serves as a proton donor to the glycosidic oxygen of sucrose, while the nucleophile Asp-23 attacks C-2 of the fructose moiety, resulting in the formation of a covalent fructosyl-enzyme complex associated with release of the glucose. At high concentrations of substrate, the fructosyl-enzyme can transfer the fructosyl moiety to primary alcohols such as methanol and ethanol and to monosaccharides (21). Support for the proposed mechanism is provided by the finding that protonation of the epoxide oxygen of conduritol B epoxide, followed by nucleophilic attack on C-2 by Asp-23, results in the formation of a covalent enzyme-inositol complex via an ester linkage. The linkage between Asp-23 and conduritol B epoxide was established by sequencing the purified affinity-labeled peptide (14). In the second step Glu-204 acts as a general base by abstracting a proton from the water molecule, which enables the hydroxyl group to displace fructose and to restore the original active center residues. Recent studies supporting this type of mechanism have been provided by Wang and Withers (35), where removal of the proton donor Glu from a -glucosidase by mutagenesis results in greatly reduced activity, which can be restored by supplying the missing Glu carboxyl group on the substrate. The crystal structure of a similar retaining enzyme, -1,4-glycanase, was presented recently (36), with its attendant covalent glycosyl intermediate and associated nucleophile and acid/base catalyst.

It is of interest to note that endoglycosidases, such as endo--N-acetylglucosaminidase H and peptide N-glycosidase F, also employ glutamyl carboxyls as general bases in the hydrolysis of their respective substrates (37, 38). Similarly, glutamyl residues have been implicated in the activation of water in the hydrolyses affected by adenosine deaminase (39) and cytidine deaminase (40).