Identification of Glu-330 as the catalytic nucleophile of Candida albicans exo-beta-(1,3)-glucanase.

The exo-β-(1,3)-glucanase from Candida albicans hydrolyzes cell wall β-glucans via a double-displacement mechanism involving a glycosyl enzyme intermediate. Reaction of the enzyme with 2′,4′-dinitrophenyl-2-deoxy-2-fluoro-β-D-glucopyranoside resulted in the time-dependent inactivation of this enzyme via the accumulation of a 2-deoxy-2-fluoro-glycosyl-enzyme intermediate as monitored also by electrospray mass spectrometry. The catalytic competence of this intermediate is demonstrated by its reactivation through hydrolysis (kreact = 0.0019 min−1) and by transglycosylation to benzyl thio-β-D-glucopyranoside (kreact = 0.024 min−1; Kreact = 56 mM). Peptic digestion of the labeled enzyme followed by tandem mass spectrometric analysis in the neutral loss mode allowed detection of two glycosylated active site peptides, the sequences of which were identified as NVAGEW and NVAGEWSAA. A crucial role for Glu-330 is confirmed by site-directed mutagenesis at this site and kinetic analysis of the resultant mutant. The activity of the Glu-330 → Gln mutant is reduced over 50,000-fold compared to the wild type enzyme. The glutamic acid, identified in the exoglucanase as Glu-330, is completely conserved in this family of enzymes and is hereby identified as the catalytic nucleophile.

The physiological role of Exg has not been clearly established, but it is undoubtedly involved in metabolism of cell wall ␤-1,3-glucan. It catalyzes the exolytic hydrolysis of polymers such as laminarin (␤-1,3 linkages) and to a lesser extent pustulan (␤-1,6 linkages) (5). Recently it has been shown that deletion of Saccharomyces cerevisiae EXG1, which encodes the major exoglucanase secreted by the yeast resulted in a marked resistance to killer toxin (6). This indicates that Exg1 is involved in the metabolism of ␤-1,6-linked glucan, the cell wall receptor for the toxin. In addition to catalyzing hydrolysis reactions, Exg of C. albicans also catalyzes the rapid transfer of glucosyl residues between suitable donor and acceptor molecules (7).
EXG, the gene for the C. albicans, encodes a protein consisting of a pre-pro sequence of 38 residues and a mature enzyme of 400 residues (8). It has 58% identity at the deduced amino acid sequence level with EXG1 of S. cerevisiae. A homology comparison of the deduced amino acid sequence of Exg with those of other related glycosidases reveals two absolutely conserved acidic residues, Glu-230 and Glu-330. Since the generally accepted mechanism involves two carboxyl-containing amino acid residues acting as an acid/base catalyst and a nucleophile (3,4,9), it is possible that these residues represent the catalytic pair of Exg. Indeed, in a previous study we showed that the point mutations (E230D and E230Q) decreased the catalytic activity 15,000-and 400-fold, respectively (10). However, an independent identification of the catalytic nucleophile by labeling studies is needed to distinguish the roles of these residues.
The application of 2-deoxy-2-fluoro-glycosides in identifying active site nucleophiles has proven successful in a number of retaining glycosidases and glycanases (11). These compounds function as mechanism-based inactivators, trapping the enzyme by the accumulation of a relatively stable but viable 2-deoxy-2-fluoro-glycosyl-enzyme intermediate. This labeling technique has been made more powerful with the advent of electrospray tandem mass spectrometry, circumventing the need for radiolabels in the identification and isolation of the labeled residue. Analysis of the proteolytic digest of a labeled enzyme by mass spectrometry has allowed for the identification of the catalytic nucleophiles of Bacillus subtilis xylanase (9) and human glucocerebrosidase (12).
In this study we report the use of 2Ј,4Ј-dinitrophenyl-2-deoxy-2-fluoro-␤-D-glucopyranoside to inactivate the exo-(1,3)-␤glucanase and the use of this reagent to identify the catalytic nucleophile. In addition we describe the mutation of this residue and report the kinetic studies that confirm its key role in catalysis.
Organisms and Plasmids-Escherichia coli strains DH5␣ and TG1 were used as the hosts for plasmids and bacteriophage M13mp19, respectively. Plasmids pUC19 and pBluescript KS ϩ/Ϫ were used for subcloning experiments. Bacteriophage M13mp19 was used as the vector with the mutagenesis kit and pBluescript KS ϩ/Ϫ was used as the vector for PCR mutagenesis. An EXG1-deficient mutant of S. cerevisiae, AWY-1 (␣ his3 1 leu2-3 leu2-112 ura3-52 trp1-289 gal can1 exg1:ura3) (7) was used as the host organism for expression of recombinant C. albicans EXG species with expression vectors derived from plasmid pEMBLyex4 (19).
Media-E. coli cultures carrying plasmids were grown in Luria or Terrific broth supplemented with 75 g/ml ampicillin, and strains carrying M13mp19 were grown in 2 ϫ YT broth (20). S. cerevisiae strains carrying plasmids were grown in the minimal medium of Wickerham (21), plus 20 g/liter glucose, and 0.03 mg/ml each of L-histidine and L-tryptophan. The expression medium was the same except that the carbon source was 30 g/liter glycerol, 20 g/liter sodium lactate, and 0.5 g/liter glucose.
Preparation of a Mutagenesis System-The 1314-base pair open reading frame of the C. albicans EXG gene has previously been cloned into M13mp19 as a 1.4-kilobase pair EcoRI-HindIII fragment (22). Singlestranded DNA was prepared, and unique SacI and XbaI restriction sites were introduced into the gene at positions 521 and 904, respectively, using the SacI and XbaI oligonucleotides in sequential rounds of mutagenesis with the oligonucleotide-directed in vitro mutagenesis kit. The mutated gene was then transferred to pUC19 as a 1.4-kilobase pair fragment, and restriction sites for PstI and SalI (3Ј to the open reading frame and from a previous cloning cassette) were deleted by digestion with these enzymes, end-filling, and then ligation of the blunted ends. The EcoRI-HindIII fragment was then cloned into pBluescript KS. Prior to inserting EXG into the yeast expression plasmid pEMBLyex4, it was modified by digesting with SalI and SacI, end-filling and ligating the blunted ends. The EXG gene was then liberated from pBluescript KS as a 1.4-kilobase pair PstI-HindIII fragment and cloned into the pEM-BLyex derivative described above, yielding pGB4. SalI-SacI, SacI-XbaI, and XbaI-HindIII fragments of 557, 383, and 443 base pairs from pGB4 were ligated to pBS KS to produce plasmids pFOX1, pFOX2, and pFOX3, and the fragments were sequenced fully in both directions to confirm the integrity of EXG inserted into pGB4.
PCR Mutagenesis-Primary PCR was performed in a 20-l reaction mixture containing 1.5 mM MgCl 2 , 50 M each of dATP, dCTP, dGTP, dTTP, 2.5 units of Taq DNA polymerase, 50 mM KCl, 10 mM Tris-HCl, pH 9.0, 0.1% Triton X-100, 20 pmol each of the appropriate mutagenic oligonucleotides, and either pFOX2 or pFOX3 as the template. The T7 and T3 oligonucleotide primers for pBluescript were used as the outside primers in both the primary and secondary PCR reactions. Reaction mixtures were subjected to an initial cycle of 95°C for 60 s, 55°C for 30 s, and 72°C for 90 s. Thirty cycles of PCR were then performed. The PCR cycle was 15 s at 95°C, 30 s at 55°C, and 90 s at 72°C. Following the final cycle, the tubes were held at 72°C for 5 min. The reaction products were then size fractionated by agarose gel electrophoresis, and desired bands were band-stabbed into secondary reaction mixtures using sterile needles. Secondary PCR was performed under the same conditions except that the pFOX templates were omitted. Products of correct size were recovered using the Geneclean II kit, digested with SacI and XbaI or XbalII and HindIII restriction enzymes, fractionated on agarose gels and recovered using the Geneclean II kit. The digested products were then ligated into pBluescript KS and cloned into E. coli DH5␣. Clones were analyzed by restriction mapping and sequenced fully to confirm the desired base changes and to check for the absence of nonspecific mutations.
Expression of Exoglucanase in S. cerevisiae-Wild type fragments of SCHEME I. Proposed retaining ␤-glycosidase mechanism.
the EXG gene in pGB4 were replaced by mutated fragments, generating plasmids pE230Q2, pE330Q2, pE330D2, and pQ330E2, and S. cerevisiae strain AWY-1 was transformed with these plasmids. Enzyme Purification-Recombinant wild type Exg was purified from 10-liter batch fermentor cultures grown at 28°C with aeration at 10 liters/min and agitation at 200 rpm, and mutant exoglucanase species were purified from 1-liter cultures grown in minimal medium in shake flasks at 400 rpm, 28°C. Cultures were inoculated to an A 600 ϭ 0.15-0.20, grown to A 600 ϭ 0.35-0.4, and galactose was added to a final concentration of 20 g/liter to induce expression. Growth was continued for an additional 24 -72 h. The cells were harvested by centrifugation (4000 ϫ g, 15 min), and the supernatant was further clarified by passage through 1.2-m filters. The medium was concentrated by ultrafiltration (M r ϭ 10,000 cut-off filter) and dialyzed against 5 mM sodium acetate buffer, pH 5.6. With large scale preparations of the native recombinant enzyme, the concentrate was then applied to a DEAE-Sepharose column (5 ϫ 15 cm) equilibrated with the same buffer, and the column was washed until the E 280 was 0. The enzyme was then eluted in a 1.6-liter gradient from 50 mM to 150 mM sodium acetate, pH 5.6. Active fractions (at the end of the gradient) were pooled and concentrated by ultrafiltration, and the buffer exchanged with 5 mM sodium acetate, pH 5.6. (NH 4 ) 2 SO 4 was added to 0.6 M, and the solution was passed through a 0.22-m filter and loaded onto a FPLC Phenyl-Superose HR 5/5 column (Pharmacia Biotech Inc.) pre-equilibrated with 0.6 M (NH 4 ) 2 SO 4 in 50 mM sodium phosphate buffer, pH 7.0. The column was developed at a flow rate of 0.5 ml/min with a reverse gradient of 0.6 -0.0 M (NH 4 ) 2 SO 4 in phosphate buffer over 30 min. Fractions (1 ml) containing the enzyme were pooled, and the solvent was changed to 5 mM sodium acetate buffer, pH 5.6, using a Centricon-10 microconcentrator (Amicon). The enzyme preparations were finally concentrated to 0.5-1.0 ml and stored at 4°C with sodium azide added to a final concentration of 0.02%. With 1-liter cultures of mutant derivatives, the ion-exchange step was omitted, and the enzyme preparations were purified in one step on the Phenyl-Superose column.
Enzyme and Protein Analysis-Exg activity was determined with laminarin, 0.54 -7.8 mg/ml, in 100 mM sodium acetate buffer, pH 5.6. Assays (125 l) were incubated for 30 min at 37°C, and the reactions were stopped by heating to 100°C for 10 min. Glucose formation was measured by the glucose oxidase method (23), and 1 unit of enzymic activity produced 1 mol of glucose/min. Alternatively, enzyme activity was determined fluorimetrically with 4-methylumbelliferyl-␤-D-glucopyranoside, 8 mM, in 20 mM sodium acetate buffer, pH 5.6. Assays (250 l) were incubated for 30 min at 37°C and the reactions were stopped with 500 l of 0.1 M glycine-NaOH buffer, pH 10.3. The formation of the methylumbelliferone was measured by emission at 450 nm with excitation at 384 nm. One unit of enzyme activity produces 1 mol of methylumbelliferone/min. The assay with methylumbelliferyl-glucoside was approximately 20 times more sensitive than the laminarin assay. Protein was estimated by a modified Lowry method (24) using bovine serum albumin as a standard. SDS-polyacrylamide gel electrophoresis was carried out with 10% acrylamide gels (25).
Enzyme Inactivation Kinetics-Kinetic studies were performed at 37°C, in 100 mM sodium acetate buffer, pH 5.6, containing 0.1% BSA. A continuous spectrophotometric assay based on the hydrolysis of DNPGlc was used to monitor enzyme activity by measurement of the rate of 2,4-dinitrophenolate release ( ϭ 400 nm, ⑀ ϭ 1071 M Ϫ1 cm Ϫ1 ), using a Unicam UV4 UV-visible spectrophotometer equipped with a circulating water bath. The inactivation of Exg by 2FDNPGlc was monitored by incubation of the enzyme (0.02 mg/ml) under the above conditions in the presence of various concentrations of the inactivators. Residual enzyme activity was determined at appropriate time intervals by addition of an aliquot (10 l) of the inactivation mixture to a solution of DNPGlc (2.0 mM, 600 l) in the above buffer and measurement of 2,4-dinitrophenolate release. Pseudo-first order rate constants at each inactivator concentration (k obs ) were determined by fitting each curve to a first order rate equation. Values for the inactivation rate constant (k i ) and the dissociation constant for the inactivator (K i ) were determined by fitting to Equation 1.
Reactivation of the 2FGlc-inactivated enzyme was studied as follows. Enzyme (200 l, 0.02 mg/ml inactivated as above, but in the absence of BSA) was concentrated using 30-kDa nominal cut-off centrifugal concentrators (Amicon Corp., Danvers, MA) to a volume of approximately 50 l, then diluted with 200 l of buffer without BSA. This was repeated twice, and the retentate was diluted to a final volume of 200 l with buffer containing 1 mg/ml BSA. Samples of the inactivated enzyme were then incubated at 37°C in the presence of various concentrations of acceptor, BTGlc (0 -88 mM). Reactivation was monitored by removal of aliquots (10 l) at appropriate time intervals and assaying as described above. Any activity loss due to denaturation of the enzyme was corrected for by a control experiment involving incubation of enzyme with no inactivator. The observed reactivation rate constant, k obs , for each reaction was determined from the slope of the plots of ln(full rate minus observed rate) versus time. The rate constant (k react ) and dissociation constant (K react ) for the reactivation were determined from a reciprocal replot of k obs versus acceptor concentrations.
Labeling of Exo-1,3-␤-glucanase-Exg (20 l, 2.0 mg/ml, in 100 mM sodium acetate, pH 5.6, buffer) was incubated with 2FDNPGlc (2 l, 20 mM) at 37°C for 2 h. Complete inactivation was confirmed by assaying with DNPGlc as described earlier. This mixture was immediately used for intact protein experiments or digested with pepsin as described below.
Mass Spectrometry-Mass spectra were recorded using PE-Sciex API III (neutral loss experiments) and PE-Sciex API 300 (LC/MS and MS/MS experiments) triple quadrupole mass spectrometers (Sciex, Thornhill, Ontario, Canada), each equipped with an ion spray ion source. Peptides were separated by reverse phase HPLC on an Ultrafast Microprotein Analyzer (Michrom BioResources Inc., Pleasanton, CA) directly interfaced with the mass spectrometer. In each of the MS experiments, the proteolytic digest was loaded onto a C18 column (Reliasil, 1 ϫ 150 mm), then eluted with a gradient of 0 -60% solvent B over 20 min, followed by 100% B over 2 min at a flow rate of 50 l/min (solvent A: 0.05% trifluoroacetic acid, 2% acetonitrile in water; solvent B: 0.045% trifluoroacetic acid, 80% acetonitrile in water). A post-column splitter was present in all experiments, splitting off 80% of the sample into a fraction collector and sending 20% into the mass spectrometer. Spectra were obtained in either the single-quadrupole scan mode (LC/ MS), the tandem MS neutral loss mode, or the tandem MS daughter scan mode (LC/MS/MS) using the settings described in each experiment.
Stoichiometry of Incorporation of Inactivator by ESMS-Exg (4 g, native or 2FGlc-labeled) was introduced into the mass spectrometer through a microbore PLRP-5 column (1 ϫ 50 mm) on a Michrom HPLC system. The quadrupole mass analyzer of an API 300 Triple Quadrupole Mass Spectrometer was scanned over a m/z range of 300-2500 Da, with a step size of 0.5 Da and a dwell time of 1 ms/step. The ion source voltage was set at 4.8 kV, and the orifice energy was 50 V. The molecular weights of the glucanase species were determined by using the deconvolution software, Multiview 1.1, supplied by Sciex.
Proteolysis-Exg (20 l, native or labeled, 2.0 mg/ml) was mixed with 100 mM phosphate buffer, pH 2 (40 l), and pepsin (20 l, 0.2 mg/ml in 100 mM phosphate buffer, pH 2). Mixtures were incubated at room temperature for 90 min, and then immediately analyzed by ESMS or frozen until required.
ESMS Analysis of the Proteolytic Digest-The single-quadrupole mode (LC/MS) conditions used were identical to those described earlier for analysis of the intact protein. In the neutral loss scanning mode, MS/MS spectra were obtained on the API III by searching for the mass loss of m/z 165, corresponding to the loss of the 2FGlc label from a peptide ion in the singly charged state. Thus, scan range: m/z 300-1800; step size: 0.5; dwell time: 1 ms/step; ion source voltage: 5 kV; orifice energy: 80; RE1 ϭ 117; DM1 ϭ 0.05; R1 ϭ 10 V; R2 ϭ Ϫ40 V; RE3 ϭ 120; DM3 ϭ 0.10; collision gas (90% argon, 10% N 2 ) thickness (CGT): 3.0 ϫ 10 14 molecules/cm 2 (CGT ϭ 300 -304). To maximize the sensitivity of neutral loss detection, the resolution (RE and DM) is normally compromised without generating artifact neutral loss peaks.
Aminolysis of the 2FGlc-labeled Peptide Digest-To a sample (20 l, 0.5 mg/ml) of 2FGlu-labeled digest was added concentrated ammonium hydroxide (5 l). The mixture was incubated for 15 min at 50°C, acidified with 50% trifluoroacetic acid, and analyzed by ESMS using the same conditions as for the intact protein.

RESULTS AND DISCUSSION
Inactivation Kinetics-Incubation of C. albicans Exg with 2FDNPGlc resulted in time-dependent inactivation of the enzyme. Data were consistent with the simple kinetic scheme shown below in which inactivation is a consequence of the accumulation of the glycosyl-enzyme intermediate E-I (Scheme II).
Thus, the inactivation followed the expected pseudo-first order kinetics as seen in Fig. 1a. Values for the inactivation rate constant (k i ϭ 0.042 Ϯ 0.002 min Ϫ1 ) and the dissociation rate constant (K i ϭ 1.49 Ϯ 0.14 mM) were determined by direct fit to the kinetic expression as described under "Experimental Procedures." A double-reciprocal plot of the pseudo-first order rate constants versus inhibitor concentration is shown in Fig.  1c. Comparison of the Michaelis-Menten kinetic parameter for the substrate DNPGlc (K m ϭ 2.2 Ϯ 0.2 mM) with the K i value for 2FDNPGlc indicates that replacement of the 2-hydroxyl by fluorine does not significantly affect ground state binding of the inactivator to the enzyme.
Interestingly, no inactivation was observed for the 2-chloro analog of the inhibitor (2C1DNPG1c), as previously demonstrated with other glycosyl hydrolases (26,27). Further, 2ClD-NPGlc did not function as a substrate, no detectable hydrolysis over background being observed. 2ClDNPGlc was then tested as a competitive inhibitor against DNPGlc hydrolysis, and an approximate K i of 0.7 Ϯ 0.2 mM was determined. These results indicated that the 2ClDNPGlc can clearly bind, but the 2-chlorine prohibits not only turnover, but even formation of the glycosyl-enzyme intermediate.
Incubation of Exg with 2FDNPGlc in the presence of a competitive inhibitor, BTGlc, resulted in reduction of the apparent inactivation rate constant (k obs ) from 0.025 Ϯ 0.001 min Ϫ1 to 0.0053 Ϯ 0.0006 min Ϫ1 (Fig. 1b). These results indicate that the inactivator is active site-directed, and by analogy with results from ␤-glucosidase (28,29), this indicates that inactivation is a consequence of the stabilization and trapping of the normal intermediate in catalysis.
Reactivation of Inactivated Exg-Evidence for the catalytic competence of the 2-deoxy-2-fluoro-glycosyl-enzyme intermediate was obtained by measuring the rates of spontaneous reactivation upon incubation in buffer at 37°C after removal of excess inactivator. The regain of activity (due to regeneration of free enzyme) followed a first order process, giving a reactivation rate constant of k react ϭ 0.0019 min Ϫ1 , corresponding to a t 1/2 of 526 min. The rate of reactivation was increased upon addition of BTGlc (a non-hydrolyzable glucoside) to the reaction mixture. This suggested that, as seen previously (9,12,30), the sugar presumably facilitates turnover of the intermediate via transglycosylation. The reactivation process followed pseudo-first order kinetics, and the rate was dependent on the BTGlc concentration in a saturable manner (Fig. 2a). A doublereciprocal replot of k obs versus reactivator concentration yielded the kinetic parameters for the reactivation process of k react (0.024 Ϯ 0.002 min Ϫ1 ) and K react (56.3 Ϯ 9.5 mM) (Fig. 2b).
Stoichiometry of Incorporation of Inactivator-The mass of native Exg was determined by ESMS to be 45,758 Ϯ 6 Da (expected mass of 45,755 Da from the amino acid sequence). After inactivation of the enzyme with 2FDNPGlc, a sample was analyzed by ESMS, and a new peak at 45,921 Ϯ 4 Da was observed in the mass spectrum. The mass difference between the labeled and unlabeled glucanase is therefore 163 Ϯ 10 Da, equal within experimental error to that of the 2FGlc label (165 Da), indicating a 1:1 stoichiometry of inactivation.

Identification of the Labeled Active Site Peptide by ESMS-
Peptic digestion of the 2FGlc-labeled enzyme resulted in a mixture of peptides that was separated by reverse phase HPLC, using the mass spectrometer as a detector. When scanned in LC/MS mode, a large number of peaks, arising from every peptide in the digestion, was observed (Fig. 3a). The peptides bearing the 2FGlc label were identified in a second experiment by using the tandem mass spectrometer in neutral loss mode. In this mode the ions are subjected to limited fragmentation by an inert gas in a collision cell. The ester linkage between the inactivator and the peptide is one of the more labile bonds present and would be expected to undergo facile homolytic cleavage. The loss of a neutral sugar of known mass (165 Da) allows for the quadrupoles Q1-Q3 to be scanned in a linked mode, detecting only ions differing in mass by loss of the label. When the spectrometer was scanned in the neutral loss tandem MS/MS mode, a search for the mass loss m/z 165 (corresponding to the loss of 2FGlc label) resulted in the detection of two peaks at 16.4 and 16.8 min not present in the unlabeled, control digest MS/MS experiment (Fig. 3b). The background signals that are present in the profile are due to unlabeled peptides undergoing equivalent neutral losses to that of the label (165 Da), most likely in this case loss of a phenylalanine residue as is shown in the HPLC profile of Fig.  3c, where a sample of unlabeled enzyme was subjected to an identical analysis. The two peptides of mass 839 Da and 1068.5 Da correspond to unlabeled peptide fragments of 674 Da (839 Ϫ 165) and 903.5 Da (1068.5 Ϫ 165). Computer analysis of the amino acid sequence of EXG (22) revealed 9 and 11 respective candidate peptides for the masses 674 and 903.5 Ϯ 1 Da. However, since the suspect peptide sequences should contain either an aspartate or glutamate residue (known nucleophiles of retaining glycosyl hydrolases) and since this residue should be present in both mass groups, elimination of all but two candidate peptides for each mass was possible. Only a single glutamic acid residue, Glu-330, is present in the following possible peptides: WNVAGE (325-330) or NVAGEW (326 -331) for mass 674 Ϯ 1 Da and NVAGEWSAA (326 -334) or AGEWSAALT (328 -336) for mass 903.5 Ϯ 1 Da. Since all the possible sequences are overlapping, Glu-330 can be tentatively assigned as the active site nucleophile for Exg.
Further evidence was obtained by the determination of the complete amino acid sequence of the labeled peptides by ESMS/MS experiments. Daughter ion scans (Fig. 4) of the peptide of 839 Da subjected to collision-induced fragmentation resulted in the loss of the covalent inactivator, generating the unlabeled peptide (675 Da). Further fragmentation of this peptide was observed, giving rise to peaks at 471, 342, 285, and 214 Da corresponding to respective losses of W, EW, GEW, and AGEW. A parallel series of "y" (labeled) fragments are also observed. The sequence NVAGEW is therefore predicted. To confirm this, analysis of the predicted "b" and "y" fragmentation pattern for the other plausible sequence, WNVAGE, showed poor correlation to the experimental results. Similar MS/MS experiments were carried out with peptide of mass 1068.5 Da (data not shown), predicting that the corresponding sequence is NVAGEWSAA.
Covalent attachment of 2FGlc to Glu-330 through an ester linkage was confirmed by treatment of the labeled peptide digest with ammonium hydroxide. After treatment, the labeled peptide (839 Da) was replaced by two new peptides having molecular weights of 674 (m/z 675, MHϩ) and 673 (m/z 674, MHϩ). This would indicate that the labeled glutamate reacted by both aminolysis and hydrolysis, generating Gln-330 and the free acid Glu-330, thereby implicating the catalytic nucleophile to be Glu-330.
Sequence Analysis-As reported earlier (30), family 5 enzymes can be further divided into five "subfamilies" within which alignments are considerably improved. Several new enzymes have since been added to this family, including a number of exo-1,3-␤-glucanases, which form another subfamily within which there is a high degree (Ͼ50%) of homology across the entire protein. The N-terminal region of this subfamily shows reasonable similarity to those of other family 5 members, but the C-terminal region is much more poorly aligned. However, even within this diverse family of glycosyl hydrolases, seven residues are conserved in most of the exo-1,3-␤-glucanases, these being Arg-130, His-173, Asn-229, Glu-230, His-291, Tyr-293, and Glu-330 in Exg. Bortoli-German et al. (31), using EGZ from Erwinia chrysanthemi, have demonstrated that mutagenesis of any of the seven conserved residues resulted in reductions in cellulase activity.
Alignment of the peptide region around Glu-330 reveals that this glutamic acid residue is completely conserved within a region of strong sequence preservation in the exo-(1,3)-␤-glucanase family (Fig. 5). This residue is also conserved within all family 5 enzymes and had previously been identified as the nucleophile in the Clostridium thermocellum endoglucanase by active site labeling experiments (30). Our labeling of Glu-330 in C. albicans exo-(1,3)-glucanase therefore confirms this assignment within this diverse family.
Structural Context-On the basis of sequence comparisons, a number of the glycosidase families, including family 5, are believed to have the same fold and to have evolved from a common ancestor (32,33). Three-dimensional structures of several of these enzymes are now available (see Refs. [33][34][35], including some from family 5 (36 -38), although none in the same subfamily as the C. albicans enzyme. These structures share a common (␣/␤) 8 barrel fold with the nucleophile glutamic acid located immediately after a hydrophobic cluster within a ␤ strand. Indeed, in the recent structure of a complex of endoglucanase E1 from Acidothermus cellulolyticus with cellotetraose (38) the OE1 atom of the carboxylate of Glu-282, the corresponding nucleophilic residue, is located within 3.5 Å of the anomeric carbon of the scissile bond, and is appropriately positioned for nucleophilic attack. Analysis of the sequence alignments of Exg with the C. cellolyticum enzyme, two enzymes that share only 15% sequence identity, shows a similar scaffold and a conserved active site. Exg must therefore possess a quite different arrangement of loops to operate as an exoglucanase and on 1,3-linked substrates.
Site-directed Mutagenesis-PCR-based site-directed mutagenesis at Glu-330 was carried out as described under "Experimental Procedures" and the expressed mutant protein isolated according to previously described protocols (7,10). The same procedure was used to replace Glu-230 in the conserved NEP motif with Gln, the other conserved carboxylic acid that had previously been suggested as the acid catalyst in this (30) and other (39) family 5 enzymes. The results of kinetic analysis of these mutants and of wild type enzyme are shown in Table I. As can be seen, using either the natural substrate laminarin, or the artificial substrate MUG, very similar activities were observed for the wild type native and wild type recombinant Exg, confirming that the recombinant DNA manipulation and expression in S. cerevisiae has no effect upon activity. The specific activity of the acid/base mutant Glu-230 3 Gln is approximately 280-fold lower than that of the wild type enzyme with both substrates, confirming earlier observations regarding the importance of this residue (10). This reduction in activity due to removal of an acid/base catalyst in a retaining glycosidase is somewhat less than that seen with a number of other glycosidases (40,41) but is similar to that seen upon mutation of the equivalent residue in other family 5 glucanases (39,42,43) where activities of less than 1% were observed.
Mutation of Glu-330 to Gln results in a much greater decrease in activity of at least 50,000-fold, consistent with the essential role of the catalytic nucleophile. This decrease in activity agrees well with results in other glycosidases in which the catalytic nucleophile has been mutated to a non-carboxylic acid residue, decreases of 10 3 to 10 7 having generally been seen (12, 44 -46). The reverse mutation Gln-330 3 Glu restored full activity, confirming that no deleterious secondary mutations had been introduced into the gene during the manipulations.
These results therefore confirm the identity of Glu-330 in C. albicans exo-␤-(1,3)-glucanase as the catalytic nucleophile, thereby solidifying the assignment of this residue within the relatively diverse glycosidase family 5.