Genes malh and pagl of Clostridium acetobutylicum ATCC 824 encode NAD+- and Mn2+-dependent phospho-alpha-glucosidase(s).

The genome of Clostridium acetobutylicum 824 contains two genes encoding NAD+, Mn2+, and dithiothreitol-dependent phospho-alpha-glucosidases that can be assigned to family 4 of the glycosylhydrolase superfamily. The two genes, designated malh (maltose 6-phosphate hydrolase) and pagl (phospho-alpha-glucosidase), respectively, reside in separate operons that also encode proteins of the phosphoenolpyruvate-dependent:sugar phosphotransferase system. C. acetobutylicum grows on a variety of alpha-linked glucosides, including maltose, methyl-alpha-d-glucoside, and the five isomers of sucrose. In the presence of the requisite cofactors, extracts of these cells readily hydrolyzed the chromogenic substrate p-nitrophenyl-alpha-d-glucopyranoside 6-phosphate, but whether hydrolysis reflected expression of enzymes encoded by the malh or pagl genes was not discernible by spectrophotometric analysis or polyacrylamide gel electrophoresis. Resolution of this question required the cloning of the malh and pagl genes, and subsequent high expression, purification, and characterization of maltose-6'-phosphate hydrolase (MalH) and phospho-alpha-glucosidase (PagL), respectively. MalH and PagL exhibit 50% residue identity, and in solution are tetramers comprising similar sized ( approximately 50 kDa) subunits. The two proteins cross-react with polyclonal rabbit antibody against phospho-alpha-glucosidase from Fusobacterium mortiferum. Purified MalH and PagL cleaved p-nitrophenyl-alpha-d-glucopyranoside 6-phosphate with comparable efficiency, but only MalH catalyzed the hydrolysis of disaccharide 6'-phosphates formed via the phosphoenolpyruvate-dependent:sugar phosphotransferase system. Importantly, analysis of the proteome of C. acetobutylicum 824 by electrospray ionization-mass spectrometry confirmed expression of MalH during growth on many alpha-glucosides tested. Site-directed changes C169S and D170N yielded full-length, but catalytically inactive MalH. Of the two putative operons, our findings suggest that only proteins encoded by the mal operon participate in the dissimilation of maltose and related O-alpha-linked glucosides by C. acetobutylicum 824.

For almost a century, and to varying extent, Clostridium acetobutylicum has been used for the industrial production of the organic solvents acetone and butanol (for historical reviews and current developments, see Refs. [1][2][3]. The genome of C. acetobutylicum 824 has now been sequenced (4), and, in conjunction with recent advances in molecular biology and biotechnology, this knowledge has stimulated renewed interest in the genetic manipulation of the pathways and products of solventogenesis. C. acetobutylicum has the capacity to metabolize a variety of carbohydrates, but starch, being inexpensive and readily available, is an industrially attractive source for microbial fermentation (3,5). Prior to its metabolism by C. acetobutylicum, starch is first hydrolyzed by amylolytic enzymes (3) to yield primarily maltose (4-O-␣-D-glucopyranosyl-D-glucopyranose). However, as emphasized by Tangney et al. (6), surprisingly little is known of the mechanism(s) by which this O-␣linked disaccharide is transported into, and subsequently hydrolyzed within, the cell.
Hydrolysis of the glycosidic bond is catalyzed by a vast array of O-glycosyl hydrolases (EC 3.2.1.-), and substrate cleavage may be accompanied by either retention or inversion of configuration at the anomeric (C1) center (7,8). An extremely helpful classification of these diverse and widespread enzymes has been developed during the past decade by Henrissat and associates (9,10). This classification scheme is based on amino acid sequence similarity, and presently includes 91 distinct families that comprise the glycosylhydrolase superfamily of enzymes. 1 Glycosylhydrolase family 4 (GHF4) 2 includes 6-phospho-␣-glucosidases (13-15), 6-phospho-␤-glucosidases (16), ␣-glucosidases (17)(18)(19), and ␣-galactosidases (20 -22). The GHF4 enzymes have received considerable attention, in part, because members of this unique family are distinguished from all others by their requirements for reducing agent, divalent metal ion (Mn 2ϩ , Co 2ϩ , Ni 2ϩ , or Fe 2ϩ ) and dinucleotide (NAD ϩ ) for activity. Whether these cofactors function in a catalytic or structural capacity in GHF4 enzymes has yet to be established (14,18,19).
Our interest in maltose dissimilation by C. acetobutylicum 824 began in 1999 (see Fig. 2 of Ref. 16), by the finding that the (then partially sequenced) genome of this organism contained * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Although indicative of phospho-␣-glucosidase and PEP-PTS activities, it was unclear from the preceding observations whether the products of only one or both operons participated in the dissimilation of individual ␣-glucosides by C. acetobutylicum. Two approaches were adopted in an attempt to resolve these questions. First, malh and pagl genes were cloned in high expression vectors, the enzymes were purified, and their substrate specificities were determined. As we show in this report, the similarity of the physical properties of MalH and PagL (e.g. molecular size, oligomeric structure, pI, and antibody crossreactivity) precluded the separation and independent assay of the two enzymes in cell extracts, by either spectrophotometric or electrophoretic methods. Consequently, in a second approach, we used ESI/MS analysis of the microbial proteome to unequivocally identify the phospho-␣-glucosidase species induced by growth of C. acetobutylicum on different ␣-glucosides.
Growth of C. acetobutylicum ATCC 824 C. acetobutylicum 824 was obtained from the American Type Culture Collection (Manassas, VA), and the organism was maintained anaerobically (GasPaK, BBL) at 37°C by serial transfer in reinforced clostridial medium (RCM, Oxoid, Ltd.). For growth of the organism in batch culture (1 liter volume), we used the synthetic medium of Monot et al. (31) supplemented with 1% (w/v) desired carbohydrate. The medium (pH 6.8) was filter-sterilized and transferred immediately to an anaerobic chamber for 24 h prior to inoculation with 0.05% (v/v) freshly grown RCM culture. After growth to stationary phase, the cells were harvested by centrifugation (13,000 ϫ g for 10 min at 5°C) and washed twice by resuspension, and centrifugation, from 25 mM Tris-HCl buffer (pH 7.5) containing 1 mM MnCl 2 , 0.1 mM NAD ϩ , and 1 mM dithiothreitol (designated TMND buffer). The average yield was ϳ2 g wet weight of cells/liter of culture.

Preparation of Cell-free Extracts of C. acetobutylicum 824
Washed cells were resuspended in 2.5 volumes of TMND buffer to yield a homogeneous thick suspension of cells. Lysozyme (0.6 mg) was added to 3-ml volumes of each cell suspension, and the preparations were incubated for 30 min at 37°C. The organisms were disrupted (six 15-s periods of sonic oscillation, at 0°C) in a Branson model 185 fine probe sonifier, and the samples were clarified by centrifugation (14,000 rpm for 30 min at 5°C) in an Eppendorf bench top centrifuge model 5417R). The supernatant fluids, containing 25-30 mg of protein ml Ϫ1 , were removed for enzyme assays and for the SDS-PAGE, Western blot, and activity stain experiments presented in Fig. 6.

Cloning of malh and pagl Genes from C. acetobutylicum 824
Based on the complete genome sequence of C. acetobutylicum 824 (4), the following pairs of primers were designed to amplify the genes malh and pagl, respectively. For amplification of malh, the forward primer was 5Ј-GGGCCCCATGGGCATGAAAAAATTTTCAGTTGTAATAG-CAGG-3Ј (the malh sequence is in boldface, and the NcoI site is underlined) and reverse primer was 5Ј-CCCGGGCTGCAGTTATTAA TT-TAATTCTGGCCAGTAATC-3Ј (the sequence complimentary to the downstream region of malh is in bold face, and the PstI site is underlined). For amplification of pagl, the forward primer was 5Ј-GGGC-CCCATGGGCATGAAAAAGTATTCAATTTGTATTGTTGG-3Ј (the pagl sequence is in boldface, and the NcoI site is underlined) and reverse primer was 5Ј-CCCGGGGAATTCTTATTAATGAAGT-TCATTCCAATAACC-3Ј (the sequence complimentary to the downstream region of pagl is in boldface, and the EcoRI site is underlined). The malh and pagl genes were amplified by the high fidelity Pfu polymerase (Stratagene). The components of the amplification mixtures (100 l) were as follows: 5 units of Pfu DNA polymerase, 1ϫ reaction buffer provided by the manufacturer, 20 mM each of the four DNTPs, 250 ng of each primer, and 100 ng of DNA. Amplifications were carried out in a thermal cycler (GeneAmp PCR System 9700, PE Applied Biosystems). After an initial 2-min incubation at 95°C, the mixtures were subjected to 30 cycles of amplification under the following conditions: denaturation at 95°C, 1 min; annealing at 50°C, 1 min; and extension at 72°C for 2 min/kilobase of insert. These procedures were followed by a 10-min run-off at 72°C. The amplicons were digested with restriction endonucleases (malh was digested with NcoI and PstI; pagl with NcoI and EcoRI), electrophoresed in 1% agarose, and purified (QIAquick gel extraction kit). The purified ϳ1.3-kb DNA fragments were ligated to the similarly digested (and purified) high expression vector pTrcHis2B to form the two recombinant plasmids pTrcHis2Bmalh and pTrcHis2Bpagl, respectively. In these constructs, malh and pagl are under control of the trc promoter, which is also regulated by the lacO operator and the product of the lacl q gene. Expression of the products of malh and pgll are induced in the presence of isopropyl-␤-D-thiogalactopyranoside (IPTG). Recombinant plasmids were introduced into Escherichia coli PEP 43 by electroporation, and transformants were selected on LB agar plates supplemented with 150 g/ml ampicillin. (Note that the mutant strain E. coli PEP 43 lacks all phospho-␤-glucosidase activities normally resident in this organism, and this strain was generously provided by Dr. B. G. Hall, Biology Department, University of Rochester, Rochester, NY.).

Site-directed Mutagenesis of malh
The role of individual amino acid residues in the CDMP motif of MalH was assessed by site-directed mutagenesis of the malh gene. This method requires the use of PfuTurbo DNA polymerase and a temperature cycler, and the appropriate reagents were obtained as the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA).
Plasmids of pTrcHis2Bmalh containing the desired mutations (C169S, D170N, M171V, and P172A) were introduced into cells of E. coli PEP 43 by electroporation. Verification of the relevant base changes was provided by DNA sequence analysis. Sequencing was performed by the dideoxynucleotide chain termination method using modified T7 DNA polymerase and [␣- 35 S]dATP labeling. The sequencing reagents were obtained as a kit (Sequenase version 2.0, U. S. Biochemical Corp., Cleveland, OH). The MacVector sequence analysis package (version 7.0, Genetics Computer Group, Madison, WI) was used to assemble, edit, and analyze the results.

Growth of Recombinant Cells and Preparation of Extracts Containing MalH and PagL
Cells of E. coli PEP 43 containing plasmids encoding either malh or pagl genes were grown on a rotary shaker at 37°C in LB medium supplemented with ampicillin (150 g/ml). At A 600 nm ϭ 0.5 units, a solution of IPTG was added to a final concentration of 1 mM, and growth was continued for ϳ4 h. The cells were harvested by centrifugation (13,000 ϫ g for 10 min at 5°C), and the organisms (ϳ3.1 g/liter) were washed by resuspension and centrifugation from TMND buffer.

Purification of MalH and PagL
Because of their similarity in physical properties, the procedure described here for MalH was also used for the purification of PagL.
Washed cells (20 g wet weight) of E. coli PEP 43, containing highly expressed MalH, were homogenized with 50 ml of TMND buffer. The organisms were disrupted (at 0°C) by two 1.5-min periods of sonic oscillation in a Branson model 350 sonifier operating at ϳ75% of maximum power, and the cell extract was clarified by ultracentrifugation (180,000 ϫ g for 2 h at 5°C). The high speed supernatant fluid was dialyzed against 4 liters of TMND buffer in a cold room overnight. MalH was purified by chromatography in three stages by low pressure chromatography. Column flow rates were maintained by use of a P-1 peristaltic pump interfaced with a Frac-100 collector. Elution of proteins was monitored at A 280 nm by a UV-1 optical control unit connected to a single-channel chart recorder.
Step 1: DEAE-TrisAcryl M (Anion Exchange) Chromatography-The dialyzed solution (ϳ 56 ml) was transferred at a flow rate of 0.6 ml/min to a column (2.6 ϫ 14 cm) previously equilibrated with TMND buffer. Nonadsorbed materials were removed by washing with buffer, and MalH was eluted with 500 ml of a linear, increasing concentration gradient of NaCl (0 -0.2 M) in TMND buffer. Fractions of 5 ml were collected, and MalH activity was detected by the intense yellow color formed upon addition of fraction samples (5 l) to microtiter wells containing 100 l of pNP␣Glc6P assay solution. Fractions with highest activity (fractions 37-44) were pooled and concentrated to 5.7 ml by filtration in an Amicon pressure unit (PM-10 membrane; 45 p.s.i.).
Step The sequences of the two GHF4 phospho-␣-glucosidases were aligned using the ClustalW algorithm (33). Identical amino acids are highlighted (black background), and numbers at left and right denote residue positions. The arrows above the sequences at the N terminus indicate a portion of the nucleotide-binding domain, including the conserved G(X)GS motif of GHF4 proteins (14,16,18,19). Arrows beneath the sequences indicate residues that constitute the signature pattern of proteins in this family. 1 Glutamyl residues that may participate in catalysis (14,19) are indicated by asterisks. The filled circle shows the cysteine residue that is positionally conserved in all GHF4 proteins, and that occurs in the CDMP motif of the phospho-␣-glucosidase subgroup of enzymes. of concentrate from step 1 was applied at 0.3 ml/min to a column of Ultrogel AcA-54 (2.6 ϫ 94 cm), previously equilibrated with TMND buffer containing 0.1 M NaCl. Fractions of 3 ml were collected, and those fractions with greatest activity (fractions 56 -62) were pooled and concentrated to 2 ml in a 10-ml Amicon filtration cell. Analysis of this preparation by SDS-PAGE revealed one major polypeptide (mass ϳ 50 kDa) and trace amounts of three other polypeptides (mass ϳ 70 -100 kDa).
Step 3: Ultrogel AcA-44 (Molecular Sieve) Chromatography-The 2-ml concentrate from step 2 was applied (flow rate, 0.15 ml/min) to a column of Ultrogel AcA-44 (1.6 ϫ 94 cm) equilibrated with the buffer solution used in step 2. Fractions of 2 ml were collected, and a single peak of enzyme activity was eluted at approximately the void volume of the column. These fractions (fractions 55-59) were concentrated to a volume of 3.8 ml containing 5.9 mg of protein/ml (specific activity 4.9 mol of pNP␣Glc6P hydrolyzed min Ϫ1 mg of protein Ϫ1 ). SDS-PAGE analysis of this preparation revealed a single polypeptide (MalH, mass ϳ 50 kDa). The same procedures were used for the purification of PagL.

Electrophoresis Procedures
Native (nondenaturing) electrophoresis and SDS-PAGE were performed in the Novex X-Cell mini system, using reagents and conditions according to the instructions from the manufacturer (Invitrogen). In SDS-PAGE experiments, denatured samples, and Mark12 TM wide range protein standards were electophoresed in precast NuPage (4 -12%) BisTris gels with MES-SDS (pH 7.3) as the running buffer. Proteins were visualized by staining with Coomassie Brilliant Blue R-250. For Western (immuno) blots, the separated proteins and SeeBlue TM molecular weight markers were transferred to nitrocellulose membranes using NuPage transfer buffer (pH 7.2). Immunodetection of phospho-␣-glucosidase (MalH or PagL) was achieved by sequential incubation of the membrane with polyclonal antibody to MalH from F. mortiferum and goat anti-rabbit horseradish peroxidase-conjugated antibody, as described previously (13). Electrophoresis of cell extracts for in situ detection of enzyme activity was carried out under native (nonreducing) conditions (at ϳ 10°C) in precast Tris-glycine (4 -20%) gels. Tris-glycine (pH 8.

Enzyme Activity Assays
The activities of MalH and PagL in cell extracts, and in column fractions during enzyme purification, were measured (at 37°C) in a discontinuous assay that contained (in 2 ml): 50 mM Tris-HCl buffer (pH 7.5), 1 mM MnCl 2 , 1 mM NAD ϩ , 1 mM DTT, and 0.5 mM pNP␣Glc6P. After enzyme addition, samples of 0.25 ml were withdrawn (20-s intervals over a 2-min period) and added immediately to 0.75 ml of a solution containing 0.5 M Na 2 CO 3 and 0.1 M EDTA to stop the reaction. The A 400 nm of the solution was measured, and the rates of substrate hydrolysis (pNP formation) were calculated from the progress plots, assuming a molar extinction coefficient for the (yellow) p-nitrophenoxide anion ⑀ ϭ 18,300 M Ϫ1 cm Ϫ1 . One unit of enzyme activity is the amount of MalH or PagL that catalyzes the formation of 1 mol of pNP min Ϫ1 at 37°C. A continuous spectrophotometric assay was used in kinetic analyses, and for determination of substrate specificity of MalH and PagL with respect to hydrolysis of the 6-phospho-␣-and 6-phospho-␤-disaccharides. This G6PDH/NADP ϩ -coupled assay monitors the formation of Glc6P produced upon hydrolysis of the glucoside-6P substrate(s). The standard assay contained (in 1 ml): 0.1 M HEPES buffer (pH 7.5), 1 mM MgCl 2 , 1 mM MnCl 2 , 1 mM NAD ϩ , 1 mM NADP ϩ , 10 mM DTT, 3 units of G6PDH, 0.2-3 mM disaccharide-6P, and 50 -150 g of purified enzyme. Reactions were initiated by enzyme addition, and the increase in A 340 nm was followed in a Beckman DU 640 recording spectrophotometer. Initial rates of reaction were determined using the kinetics program of the instrument, and a molar extinction coefficient ⑀ ϭ 6,220 M Ϫ1 cm Ϫ1 was assumed for calculation of NADPH (Glc6P) formed. Kinetic parameters were determined from Eadie-Hofstee plots generated by the dogStar software (D. G. Gilbert) version 1.0c kinetics program.

Analytical Methods
Protein concentrations were either measured by the BCA protein assay kit (Pierce) or calculated using the theoretical molar absorption . Protein (monomer) mass was determined by electrospray in an HP1100 mass spectrometer, and the sequence of residues at the N termini of MalH and PagL was obtained using an ABI 477A protein sequencer (Applied Biosystems, Inc.) connected in line to an ABI 120A phenylthiohydantoin analyzer. The molecular weights of MalH and PagL in their native states were estimated by gel permeation high performance liquid chromatography (HPLC) on a Phenomenex BioSep-SEC-S2000 column (7.8 ϫ 300 mm) equilibrated with 50 mM phosphate buffer (pH 7.0). The column was calibrated with ϳ30 g of the following standard proteins: carbonic anhydrase (31 kDa); ovalbumin (44 kDa); BSA (monomer, 66 kDa; dimer, 132 kDa) and aldolase (158 kDa). Migration of the proteins was monitored at 214 nm. The approximate molecular weights of MalH and PagL (ϳ 40 g each, applied separately) were derived from the plot of log M r versus retention time (min).

Liquid Chromatography-Electrospray Ionization Mass Spectrometry (LC-ESI/MS)
A Hewlett Packard LC-MS system (Agilent Technologies, Palo Alto, CA) consisting of a binary pump, degasser, autosampler, and an HP1100 LC-mass selective detector was used for these studies. Data were acquired on the HP ChemStation data system. Calibration of the instrument was performed with a standard Agilent ES tuning mix. The mass spectrometer was scanned from m/z 500 to 1700 every 4 s. Nitro-gen gas was used to assist the nebulization and desolvation process. Cell-free extracts (5 l) containing approximately 100 -150 g of total protein was used for each injection. Zorbax SB-C3 reversed phase columns (150 mm ϫ 2.1 mm, inner diameter) equipped with guard columns (C3 reversed phase) were used for the separations. Solvent A of the mobile phase was 5% (v/v) acetic acid (pH 2.6), and solvent B was acetonitrile. The flow rate was 200 l min Ϫ1 . A shallow gradient (5-80% acetonitrile over 60 min) was used for the chromatographic separation. The column temperature was maintained at 40°C, and, after a 5-min delay, the column effluent was directed into the mass spectrometer.

RESULTS
The Two Putative ␣-Glucoside:PTS Operons of C. acetobutylicum 824 -The genetic and organizational similarities of the two putative ␣-glucoside:PTS operons are illustrated in Fig. 1. For convenience, they are designated the maltose (mal) and phospho-␣-glucoside (pagl) operons, respectively. Genes in each operon encode proteins representing: (i) a helix-turn-helix protein from the RpiR family of transcriptional regulators, (ii) a fused EII(CB) membrane-localized transporter of the PEP-PTS, and (iii) putative phospho-␣-glucosylhydrolase(s), designated maltose 6Ј-phosphate hydrolase (MalH) and phospho-␣glucosidase (PagL), respectively. It is of interest that the pagl operon also includes a gene that encodes an EIIA component of the PTS. A corresponding gene is not present in the mal operon    of C. acetobutylicum, nor is a comparable EIIA gene found in any of the ␣-glucoside:PTS operons described recently in other species (15,24,32).
Sequence Alignment of MalH and PagL from C. acetobutylicum 824 -Comparative sequence alignment reveals 50% residue identity between MalH and PagL (Fig. 2). Significantly, both proteins contain an N-terminal NAD ϩ -binding domain (14,15,18,19), and they exhibit the conserved motif (G(X)NH) and signature sequence (P-X-  1 All family 4 phospho-␣-glucosidases described thus far contain the amino acid motif CDMP (see Fig. 2). The occurrence of this unique motif in MalH and PagL suggested to us that both proteins might exhibit phospho-␣-glucosidase activity.
Cloning, Expression, and Purification of MalH and PagL-Genes malh and pagl were cloned into the vector pTrcHis2B, and the recombinant plasmids were transformed into cells of E. coli PEP 43 for IPTG-induced expression of the two proteins. The same three-stage procedure was used for the chromatographic purification of MalH and PagL (see "Experimental Procedures"). Analysis of denatured samples of the purified proteins by SDS-PAGE, revealed a single polypeptide of ϳ50 kDa for both MalH and PagL (Fig. 3, A and B, respectively). The two proteins were indistinguishable by their migration positions, and both MalH and PagL cross-reacted strongly with polyclonal antibody raised against phospho-␣-glucosidase (MalH) from F. mortiferum (Fig. 3C).
Physical Properties of PagL and MalH-In contrast with the molecular mass estimates obtained by SDS-PAGE, HPLC gel filtration yielded molecular mass estimates of ϳ200 kDa for both MalH and PagL (Fig. 4). The failure to detect species of intermediate molecular mass (ϳ50 or ϳ100 kDa) suggests that, in their native solution states, both MalH and PagL exist in tetrameric form. The homogeneity of purified PagL was confirmed by microsequence analysis that provided the unambiguous determination of 28 of the first 30 residues from the N terminus of the protein: GMKKYSI(X)IVGGGSRYTPDM-LAML(X)NQKER. A glycine residue derived from the vector DNA precedes the start Met residue, and amino acids were not recovered in cycles 8 and 25. With the exception of the initial glycine and the two "missing" residues (Cys), the sequence obtained was in agreement with that deduced by translation of the pagl gene. The results of ESI/MS yielded a molecular mass for PagL of 51,205 Da, which is greater than the theoretical value (51,155 Da) by the mass of a single glycyl residue. Microsequence analysis also revealed an additional glycine residue at the N terminus of MalH, but otherwise the sequence of the next 29 amino acids agreed precisely with that expected from translation of the malh gene: GMKKFSVVIAGGGST-FTPGIVLMLLDNMDK. Consistent with these data, the mass of the protein determined experimentally by ESI/MS (50,038 Da) was greater than the theoretically predicted mass of 49,972. The mass difference was attributed to an additional Gly residue and an oxidation.
Cofactor Requirements and Substrate Specificity of PagL and MalH-Previous studies of proteins assigned to GHF4 (14, 16, 17, 20 -22), have shown that NAD ϩ , Mn 2ϩ ion, and DTT are prerequisites for enzyme activity. Data presented in Table I show that both MalH and PagL are dependent upon these same cofactors for hydrolysis of the chromogenic substrate, pNP␣Glc6P. In studies of nucleotide specificity (Table II), it was found that only ␤-NAD ϩ (and, to a lesser degree, its deamino derivative) served as activator(s) of MalH and PagL. Interestingly, in the case of the latter enzyme, considerable activation was elicited by the guanine analog, ␤-NGD ϩ . In the presence of saturating concentrations of cofactors (1 mM Mn 2ϩ , 1 mM NAD ϩ , and 20 mM DTT), kinetic analyses yielded comparable parameters for the hydrolysis of pNP␣Glc6P by the two enzymes: MalH, K m ϭ 28.4 Ϯ 5.9 M and V max ϭ 5.2 Ϯ 0.4 mol hydrolyzed min Ϫ1 mg of protein Ϫ1 ; and PagL, K m ϭ 58.3 Ϯ 5.6 M and V max ϭ 4.4 Ϯ 0.2 mol hydrolyzed min Ϫ1 mg of protein Ϫ1 . However, the two enzymes differed markedly in their capacity to hydrolyze the "natural" phosphorylated ␣-glucoside products of the PEP-dependent PTS (Table III). With the exception of sucrose-6P, MalH catalyzed the hydrolysis of all O-␣-linked disaccharide 6-phosphates tested, including maltose-6ЈP and all five phosphorylated isomers of sucrose. Surprisingly, under the same conditions, PagL failed to cleave any of these compounds. None of the ␤-linked disaccharide 6-phosphates tested (cellobiose-6ЈP, gentiobiose-6ЈP, and methyl-␤-Glc6P) were substrates for either MalH or PagL.
Requirement(s) for DTT and a Conserved Cysteine Residue for Enzyme Activity-In the presence of the requisite NAD ϩ and Mn 2ϩ cofactors, almost all GHF4 enzymes require a comparatively high concentration of DTT for optimum activity (17,20,21), but the role(s) of this reducing agent have yet to be defined (for discussion, see Ref. 19). For both MalH and PagL, 5-10 mM DTT elicited maximum rates of pNP␣Glc6P hydrolysis (data not shown). All GHF4 proteins contain a conserved cysteine residue centered around amino acids 160 -180 of their sequences. Crystallographic data obtained during the structural determination of ␣-glucosidase (AglA) from Thermotoga maritima (19) provide evidence that loss of activity of this GHF4 enzyme may result from oxidation of the conserved Cys-SH-174 residue to the sulfinic acid (Cys-SO 2 H) form. Of the five GHF4 phospho-␣-glucosidases that have now been characterized (including MalH and PagL), this important Cys residue occurs in a distinctive motif comprising the four amino acids CDMP. Site-directed mutagenesis experiments confirmed Cys-169 and the adjacent residue Asp-170, as essential residues for MalH activity in C. acetobutylicum. Changes C169S, D170N, M171V, and P172A resulted in the expression of a polypeptide of the mass (ϳ50 kDa) and immunoreactivity expected for complete MalH, but the conversions C169S and D170N yielded catalytically inactive proteins (Fig. 5).
Expression of Phospho-␣-glucosidase Activity in Cells of C. acetobutylicum 824 -Cells of C. acetobutylicum were grown on a variety of carbohydrates, including glucose, ␤-glucosides (cellobiose, gentiobiose), and ␣-glucosides (maltose, maltitol, methyl-␣-glucoside, sucrose, and its five O-␣-linked isomers). Cellfree extracts were prepared and assayed for phospho-␣glucosidase activity using pNP␣Glc6P as substrate (Table IV). No enzyme activity was detectable in cells grown previously on glucose, sucrose, or ␤-glucosides. However, high levels of phospho-␣-glucosidase activity were induced during growth of the organism on all other ␣-glucosides tested.
Analysis of Cell Extracts by PAGE Procedures-In an attempt to determine whether the phospho-␣-glucosidase activity was the result of the presence of MalH or PagL, (or both enzymes), samples of the extracts were examined by three different PAGE procedures. SDS-PAGE showed that extracts containing phospho-␣-glucosidase activity were induced for high level expression of a polypeptide of ϳ50 kDa (Fig. 6A), which also cross-reacted specifically with antibody prepared against phospho-␣-glucosidase (MalH) from F. mortiferum (Fig.  6B). Finally, native PAGE and in situ staining for enzyme activity revealed a single zone of fluorescence formed by the hydrolysis of the fluorogenic substrate, 4MU␣Glc6P (Fig. 6C). Although these findings confirmed the presence of phospho-␣glucosidase activity, the similarity in molecular masses of MalH and PagL (49,972 and 51,155 Da, respectively), precluded identification of the catalytically active species in the extracts. We therefore resorted to electrospray ionization-mass spectrometry (ESI/MS) in an attempt to answer these questions.
Identification of MalH in the Proteome of C. acetobutylicum by ESI/MS-An extract prepared from maltulose-grown cells of C. acetobutylicum was used in our initial ESI/MS analysis. An excellent separation of components of the microbial proteome was achieved by reverse phase (C3) chromatography using a shallow gradient of 5% acetic acid/acetonitrile. The mass selective detection spectrum revealed a prominent polypeptide at an elution time of 31.5 min (Fig. 7A), and the ESI spectrum exhibited an envelope of multiply charged molecules ranging from ϩ30 to ϩ72 charge states (Fig. 7B). After deconvolution (Fig. 7C), the average molecular mass of the protein (49,973 Da) was in perfect agreement with the theoretical value calculated from the amino acid composition of MalH. Subsequently, ESI/MS was used for the proteomic analysis of cells of C. acetobutylicum grown previously on 10 different sugars. Consistent with the enzymatic analyses (Table IV), the phospho-␣glucosidase was not detectable in the proteome of glucose or sucrose-grown organisms (Table V). However, extracts from cells grown previously on all ␣-glucosides, including the five separate isomers of sucrose, contained MalH. (It may be noted parenthetically that further ESI/MS studies also provided evidence for expression of PagL during growth of C. acetobutylicum on maltose and methyl-␣-D-glucoside. The significance of these findings is presently unclear.) DISCUSSION The aims of our investigation were twofold: first, to determine the physical and enzymatic properties of the proteins encoded by the malh and pagl genes; and second, to study the expression of the two gene products during growth of C. acetobutylicum on maltose and related ␣-linked glucosides.