Cloning, Sequencing, and Characterization of a Heat- and Alkali-Stable Type I Pullulanase from Anaerobranca gottschalkii

Abstract

The gene encoding a type I pullulanase was identified from the genome sequence of the anaerobic thermoalkaliphilic bacterium Anaerobranca gottschalkii. In addition, the homologous gene was isolated from a gene library of Anaerobranca horikoshii and sequenced. The proteins encoded by these two genes showed 39% amino acid sequence identity to the pullulanases from the thermophilic anaerobic bacteria Fervidobacterium pennivorans and Thermotoga maritima. The pullulanase gene from A. gottschalkii (encoding 865 amino acids with a predicted molecular mass of 98 kDa) was cloned and expressed in Escherichia coli strain BL21(DE3) so that the protein did not have the signal peptide. Accordingly, the molecular mass of the purified recombinant pullulanase (rPulAg) was 96 kDa. Pullulan hydrolysis activity was optimal at pH 8.0 and 70°C, and under these physicochemical conditions the half-life of rPulAg was 22 h. By using an alternative expression strategy in E. coli Tuner(DE3)(pLysS), the pullulanase gene from A. gottschalkii, including its signal peptide-encoding sequence, was cloned. In this case, the purified recombinant enzyme was a truncated 70-kDa form (rPulAg′). The N-terminal sequence of purified rPulAg′ was found 252 amino acids downstream from the start site, presumably indicating that there was alternative translation initiation or N-terminal protease cleavage by E. coli. Interestingly, most of the physicochemical properties of rPulAg′ were identical to those of rPulAg. Both enzymes degraded pullulan via an endo-type mechanism, yielding maltotriose as the final product, and hydrolytic activity was also detected with amylopectin, starch, β-limited dextrins, and glycogen but not with amylose. This substrate specificity is typical of type I pullulanases. rPulAg was inhibited by cyclodextrins, whereas addition of mono- or bivalent cations did not have a stimulating effect. In addition, rPulAg′ was stable in the presence of 0.5% sodium dodecyl sulfate, 20% Tween, and 50% Triton X-100. The pullulanase from A. gottschalkii is the first thermoalkalistable type I pullulanase that has been described.

Pullulanases (pullulan 6-glucanohydrolase; EC 3.2.1.41) are debranching enzymes that are able to hydrolyze the α-1,6 bonds of the linear α-glucan pullulan, producing maltotriose as the final product. Pullulanases are divided into two classes based on substrate specificity: (i) type I pullulanases or debranching enzymes, which specifically hydrolyze the α-1,6 linkages in branched oligosaccharides, such as starch, amylopectin, and glycogen, forming linear α-1,4-linked oligomers; and (ii) type II pullulanases or amylopullulanases, which cleave α-1,6 glycosidic linkages in pullulan and branched substrates in addition to the α-1,4 glycosidic linkages in polysaccharides other than pullulan (7).

A number of pullulanases have been purified from different bacterial and archaeal sources and characterized. Most enzymes from thermophilic and hyperthermophilic microorganisms are type II pullulanases (35, 42, 50, 52) and have been isolated mainly from thermophilic archaea (11, 14, 17, 32, 40, 49). Thermostable type I pullulanases have been characterized from the aerobic moderately thermophilic bacteria Bacillus acidopullulolyticus (25, 29, 36) and Bacillus flavocaldarius KP 1228 (53), the thermophilic bacteria Thermus aquaticus YT-1 (47), Thermus caldophilus GK-24 (31), and Bacillus thermoleovorans (43), and the extreme anaerobic thermophilic bacteria Caldocellulosiruptor saccharolyticus (1), Fervidobacterium pennivorans (8, 12, 33), and Thermotoga maritima (9, 34). Sequence information has revealed that there is low overall conservation among type I enzymes; however, a highly conserved region consisting of seven amino acids, YNWGYDP, is found in all type I pullulanases that have been described to date (7, 8). Interestingly, data concerning the physicochemical properties of all debranching enzymes described so far show that they are active mostly in the acidic or neutral pH range. Alkalistable glycoside hydrolases have been found in several mesophilic bacteria, such as Bacillus sp. strain 202-1 (45), Bacillus sp. strain TS-23 (41), Bacillus sp. strain KSM-1876 (4, 21), Bacillus sp. strain KSM-1378 (5, 20), and Bacillus sp. strain XAL601 (39, 54). All these enzymes, however, have molecular masses that are greater than the average molecular mass for the known pullulanases and are not classified as type I pullulanases because they have amylase and pullulanase activities located in two different active sites. Production of an alkalistable type I pullulanase has been reported for the mesophilic bacteria Bacillus sp. strain S-1 (30, 38) and Micrococcus sp. strain Y-1 (30). There have been no reports, however, that thermophilic microorganisms are able to produce heat- and alkaline-stable type I pullulanases.

Anaerobranca gottschalkii DSM 13577 is an obligate anaerobic thermoalkaliphilic bacterium that was isolated from a hot inlet of Lake Bogoria in Kenya (48). Since this strain grows heterotrophically on a variety of polysaccharides at pH 9.5 and 55°C, it is a suitable source for extracellular thermoalkalistable starch-hydrolyzing enzymes. However, although this bacterium is able to grow on pullulan, pullulanase activity was not detected in the culture supernatant (48). The closely related organism Anaerobranca horikoshii DSM 9786, isolated from Yellowstone National Park, is an obligate anaerobic thermoalkaliphilic bacterium whose optimum growth conditions are pH 8.5 and 57°C (15). A. horikoshii has been reported to be proteolytic, but there has been no report that it produces amylolytic enzymes. Recently, the whole genome of A. gottschalkii has been sequenced (H. P. Klenk, A. Ruepp, M. Stark, A. Zibat, and I. Becker, Int. Congr. Biocatalysis, p. 23, 2002), and an open reading frame (ORF) with high pairwise similarity to the genes encoding pullulanases in the thermophilic anaerobic bacteria F. pennivorans and T. maritima was identified. In this paper we describe cloning and sequencing of a pullulanase gene from A. gottschalkii and its heterologous expression in Escherichia coli, as well as purification of a recombinant enzyme and its biochemical properties. This is the first report of a type I pullulanase in thermophilic anaerobic bacteria that grow optimally under alkaline conditions. Interestingly, two different expression experiments revealed two active stable forms of the pullulanase, an enzyme with the expected molecular mass (96 kDa) and a 70-kDa truncated form.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

A. gottschalkii DSM 13577 and A. horikoshii DSM 9786 were grown anaerobically as previously described (15, 48). Fermentation of A. gottschalkii was carried out in a 20-liter bioreactor with a continuous flow of N₂-CO₂ (80:20, vol/vol) (48). E. coli electrocompetent Novablue, Bl21(DE3), and Tuner(DE3)(pLysS) (Novagen, Schwalbach, Germany) cells expressing the gene encoding the recombinant pullulanase from A. gottschalkii (rPulAg) were grown aerobically at 37°C in Luria-Bertani (LB) medium (51) containing 50 μg of ampicillin per ml and were induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) when the absorbance at 600 nm reached 0.6 to 0.8.

DNA manipulation.

Chromosomal DNA from A. horikoshii and A. gottschalkii, as well as plasmid DNA, were prepared by using kits obtained from QIAGEN GmbH (Hilden, Germany) and following the manufacturer's instructions. The genomic sequence encoding the pullulanase from A. gottschalkii was obtained from the genome sequencing project (Klenk et al., Int. Congr. Biocatalysis, 2002). Other molecular procedures were performed as described by Sambrook and Russell (51).

Cloning of the pullulanase gene from A. horikoshii.

Genomic DNA was partially digested with Sau3AI and fractionated on an agarose gel. Fractions between 1.5 and 8 kb were used to prepare a DNA library in pZAP express (Stratagene, La Jolla, Calif.). Phagemids were excised as described in the instructions of the manufacturer. The mass excised library (>10,000 clones) was grown on LB agar plates containing 50 μg of kanamycin per ml, AZCL-amylose (0.05%, wt/vol), and AZCL-pullulan (0.05%, wt/vol) (Megazyme, Wicklow, Ireland) that were inoculated with 250 CFU/plate. After growth for 2 days at 37°C, positive clones were restreaked onto LB agar plates containing kanamycin and the single substrates AZCL-amylose and AZCL-pullulan.

The insert from E. coli PUL1012 was tagged by using a primer island transposition kit (Perkin-Elmer Applied Biosystems), and pullulanase-negative clones that were resistant to trimethoprim and kanamycin were selected. The clones were sequenced with primers specific for the transposon. Primer walking was used to fill any gaps.

Expression of the pullulanase gene from A. gottschalkii.

Recombinant DNA techniques and expression of the recombinant gene were performed in E. coli XL1-Blue and BL21(DE3), respectively. The A. gottschalkii pullulanase gene was cloned from a plasmid gene library (Epidauros AG, Bernried, Germany) containing genomic DNA from A. gottschalkii. Two overlapping clones from the library were used to isolate the complete approximately 2.6-kb pullulanase ORF by PCR-aided amplification of two fragments, referred to as F2pulAg and F1pulAg, containing about 1.0 and 1.6 kb of the ORF, respectively, by using the following primers (restriction sites used for cloning are underlined, and the restriction enzymes are indicated in parentheses): F2rPulAg-forward, 5′-ACCTTTTTAGCTAGCATGGGAAGTAAGGATGTA-3′ (NheI); F2rPulAg-reverse, 5′-CTGACGTATGGATCCACTGCTTCATTAACT-3′ (BamHI); F1rPulAg-forward, 5′-AGTTAATGAAGCAGTGGATCCATACGTCAG-3′ (BamHI); and F1rPulAg-reverse, 5′-CAAAAAGATAAGCCTTTAAAGCTTATCTTTTTGGTTTAAG-3′ (HindIII).

The PCR mixtures (100 μl) contained 100 pmol of the forward primer, 100 pmol of the reverse primer, each deoxynucleoside triphosphate at a concentration of 200 μM, 0.1 μg of template DNA, and 9 U of Pfu DNA polymerase (Promega). The temperature profiles for amplification with the F1pulAg and F2PulAg primers after denaturation for 10 min at 94°C were 30 cycles of 94°C for 1 min, 63°C for 1 min, and 72°C for 3 min and 30 cycles of 94°C for 1 min, 60°C for 1 min, and 72°C for 3 min, respectively. In both cases, a final extension reaction at 72°C for 5 min was carried out after the last amplification cycle. For intracellular recombinant expression, the A. gottschalkii pullulanase gene was inserted into the expression vector pET21a without the region encoding the enzyme's putative N-terminal signal peptide by using the following cloning strategy. The 1.6-kb PCR product F1pulAg was digested with BamHI-HindIII and cloned in the vector pET21a. The resulting construct was designated pet21aF1. The 1.0-kb PCR product F2pulAg and pet21aF1 were digested with NheI-BamHI and then ligated. The resulting construct, pET21a-pulAG, carried the A. gottschalkii pullulanase gene under control of the T7 promoter. rPulAg was produced in E. coli strain BL21(DE3). For overexpression, cells were grown at 37°C in 2 liters of LB medium containing 100 μg of ampicillin per ml. At an optical density at 600 nm of 0.6 to 0.8, induction of pullulanase gene expression was started by adding 1 mM IPTG, and the preparation was incubated for 12 h.

In an alternative expression experiment, the pullulanase gene, including its signal peptide-encoding sequence, was amplified by PCR by using the Expand long template PCR system (Boehringer Mannheim) and primers pETBlueForward (5′-TTGTTTAAACTTAGGAAATTA-3) and pETBlueReverse (5′-GGGTGTCCCGGGTTATTTTTGGTAAAGTACCATCATAC-3′). The temperature profile after denaturation for 3 min at 94°C was 25 cycles of 94°C for 1 min, 45°C for 3 min, and 72°C for 3 min, followed by a final extension step at 72°C for 7 min. Cloning of the amplified fragment, which contained a single 3′ deoxyribosyladenine, was carried out by using an Introductory pETBlue-1 Novagen AccepTor vector kit according to the instructions of the manufacturer. The single 3′ deoxyribosyladenine ends of pullulanase PCR products generated by the activity of nonproofreading Taq DNA polymerase were compatible with the single 3′ deoxyribosyluridine overhangs of the linearized pETBlue-1 AccepTor vector as determined by direct ligation. The NovaBlue host was used for initial cloning and verification of the constructs in the vector. Recombinant plasmid pETBlue-1 containing the pullulanase gene was used to transform the IPTG-inducible T7 promoter E. coli strain BL21(DE3)Lys-Tuner. For expression, an overnight culture of the recombinant strain in LB medium containing 50 μg of ampicillin per ml and 34 μg of chloramphenicol per ml was inoculated into LB medium with the same antibiotic content and incubated with shaking at 37°C. The cultures were induced with 1 mM IPTG when an optical density at 600 nm of 0.6 to 08 was reached. The cells were harvested after 18 h. The cell pellets were suspended in 50 mM Tris-HCl buffer (pH 8.5) (5 ml per g [wet weight] of cells) containing the protease inhibitor 4-(2-aminoethyl)-benzenesulfonylfluoride hydrochloride (Pefabloc SC PLUS; Roche, Mannheim, Germany) at a concentration of 4 mM and sonicated for 15 min. Following centrifugation the supernatant containing the pullulanase was assayed to determined the activity and protein concentration as described below.

The correct cloning of the pullulanase ORF was confirmed in all cases both by sequencing and by restriction enzyme analysis. Positive clones producing thermostable pullulanase were detected on pullulan-red agar plates as described previously (8, 14). The plates were incubated at 37°C overnight and then at 55°C until halos were observed.

Enzyme assay.

Pullulanase activity was determined by measuring the enzymatic release of reducing groups from pullulan. Unless indicated otherwise, the assay mixtures (total volume, 500 μl) contained 250 μl of 1% (wt/vol) pullulan (molecular weight, 200,000; ICN, Meckenheim, Germany), 100 μl of McIlvaine buffer (pH 8.0), and appropriately diluted enzyme. After incubation at 60°C for 10 to 15 min, the reaction was stopped by addition of 750 μl of cold dinitrosalicylic acid reagent (6), followed by boiling for 15 min and determination of the absorption at 575 nm. Sample blanks were used to correct for nonenzymatic release of reducing sugars. One unit of pullulanase activity was defined as the amount of enzyme that released 1 μmol of reducing sugars (with maltose as the standard) per min under the assay conditions used.

Purification of recombinant pullulanase from E. coli.

All purification steps were performed at room temperature. For purification of the recombinant full-length pullulanase (rPulAg), cells of E. coli BL21(DE3)(pET21a-pulAg) were harvested, washed with 20 mM Tricine-HCl (pH 8.0), resuspended in a small volume of the same buffer, and disrupted by repeated passage through a French pressure cell (American Instrument Company, Silver Spring, Md.) at 6.9 MPa. After centrifugation (40,000 × g, 30 min, 4°C) the cleared homogenate was incubated at 70°C for 15 min and cooled on ice. Precipitated proteins were sedimented by centrifugation (40,000 × g, 30 min, 4°C), and the supernatant was subjected to anion-exchange chromatography on a 50-ml-bed-volume XK26 Source 15Q column (Pharmacia, Freiburg, Germany). Elution was carried out in 20 mM Tricine-HCl (pH 8.0) with the following NaCl gradient program: 10 column volumes of 0 to 0.4 M NaCl, 1 column volume of 0.4 to 1 M NaCl, and 2 column volumes of 1 M NaCl. Fractions containing pullulytic activity were pooled, the NaCl concentration was adjusted to 5 M, and the pooled fractions were subjected to hydrophobic interaction chromatography on a 20-ml-bed-volume XK16 Phenyl-Sepharose HP column (Pharmacia). Elution was carried out in 20 mM Tricine-HCl (pH 8.0) as follows: 2 column volumes of 5 M NaCl, 1 column volume of 5 to 3 M NaCl, 9 column volumes of 3 to 0 M NaCl, and 2 column volumes of 0 mM NaCl. The flow rate was 5 ml per min. Fractions with pullulanase activity were pooled and concentrated to a final volume of 10 ml by ultrafiltration (10-kDa cutoff).

The following method was used for purification of the N-terminally truncated pullulanase (rPulAg′) from E. coli Tuner(DE3)(pLysS)(pET-Blue1-pulAg) (designated clone CBpulAg). Ten grams of cells was washed with 50 mM Tris-HCl (pH 8.5) and then suspended in 50 ml of buffer containing 4 mM Pefabloc SC PLUS protease inhibitor (Roche). The cells were disrupted by sonication, the cellular debris was removed by centrifugation for 20 min at 20,000 × g, and the supernatant was dialyzed against 100 volumes of 50 mM Tris-HCl (pH 8.5). The supernatant was then loaded onto a Q-Sepharose column (2.5 by 20 cm) equilibrated in the same buffer until no absorbance at 280 nm was detectable. Proteins were eluted with a linear 0 to 0.5 M NaCl gradient in 50 mM Tris-HCl (pH 8.5). Active fractions were pooled, concentrated by ultrafiltration, and dialyzed overnight against 100 volumes of 50 mM Tris-HCl (pH 8.5). The supernatant was heat treated at 65°C for 60 min, and the denatured host proteins were collected by centrifugation for 15 min at 30,000 × g. The pullulanase remained in the clear supernatant. The enzyme solution was adjusted with ammonium sulfate to a final concentration of 1 M and applied at a flow rate of 20 ml/h to a Phenyl-Sepharose column (2.5 by 6 cm) that was equilibrated with 50 mM Tris-HCl (pH 8.5) containing 1 M ammonium sulfate. After the column was washed with 50 ml of equilibration buffer, the pullulanase was eluted with a linear 1 to 0 M ammonium sulfate reverse gradient in the same buffer. Fractions containing high levels of pullulanase activity were pooled and dialyzed against 50 mM Tris-HCl (pH 8.5) containing 0.15 M NaCl. The concentrated sample was applied to a Superdex S-200 prep column (2.6 by 96 cm; Amersham Biotech Inc., Freiburg, Germany) that was equilibrated with buffer B. Elution was performed with the same buffer, and 2-ml fractions were collected at a flow rate of 2 ml/min. Fractions containing enzyme were analyzed on a sodium dodecyl sulfate (SDS)-12% polyacrylamide gel electrophoresis (PAGE) gel.

Purification of the native pullulanase from A. gottschalkii cells.

Ten grams of A. gottschalkii cells obtained by large-scale fermentation was washed with 50 mM Tris-HCl (pH 8.5) and then resuspended in 50 ml of the same buffer containing 4 mM Pefabloc SC PLUS. The cells were disrupted by sonication, the cell debris was removed by centrifugation for 20 min at 20,000 × g, and the supernatant was dialyzed against 100 volumes of 50 mM Tris-HCl (pH 8.5). The supernatant was then loaded onto a Q-Sepharose column (2.5 by 20 cm) equilibrated with 50 mM Tris-HCl (pH 8.5) until no absorbance at 280 nm was detectable. The column was washed with the same buffer, and proteins were eluted with a linear 0 to 0.5 M NaCl gradient. Active fractions were pooled, concentrated by ultrafiltration, and dialyzed overnight against 100 volumes of 50 mM Tris-HCl (pH 8.5).

Gel electrophoresis.

SDS high-molecular-weight marker (Sigma) was used in order to determine the apparent molecular weights of the samples. SDS-PAGE as described by Laemmli (37) was routinely performed under reducing conditions with 12% polyacrylamide gels by using a Bio-Rad Mini-PROTEAN II electrophoresis unit (Bio-Rad, Munich, Germany). Samples were treated in denaturing buffer at 95°C for 5 min before loading. After electrophoresis the gels were rinsed in water or in 0.25% (vol/vol) Triton X-100 at 4°C for 1 h to remove the SDS and incubated for 10 to 30 min under optimal assay conditions to detect the pullulanase activity. Zymogram staining for pullulytic activity was performed as described by Furegon et al. (16). Native PAGE was performed with 4 to 20% polyacrylamide gradient gels purchased from Novex (MBI Fermentas, St. Leon Rot, Germany). The gels were run at 120 V for 16 h at 4°C. High-molecular-weight marker proteins (Pharmacia Biotech) were used as standards. Gels that were not used for activity staining were soaked in Coomassie blue staining solution (0.15% Coomassie brilliant blue R-250 in 45.5% methanol-8% acetic acid) for 30 min; this was followed by destaining overnight in 25% methanol-35% acetic acid. Protein content was determined by the Bradford method (10).

Influence of pH and temperature.

The pH of the buffers was generally adjusted at the temperature used for the experiment. Experiments were carried out with the purified recombinant enzyme. The influence of pH on pullulanase activity was determined at 60°C by using the assay protocol described above. To determine the influence of temperature on the enzymatic activity, standard assay reaction mixtures at pH 8.0 were incubated at temperatures ranging from 40 to 80°C for 15 min. The kinetics of thermoinactivation was monitored at pH 8.0. During incubation of the enzyme at different temperatures, samples were withdrawn after various times and clarified by centrifugation, and the residual activity was measured.

Characterization of hydrolysis products.

Hydrolysis products arising from the activity of pullulanase with various linear and branched polysaccharides were analyzed by high-performance liquid chromatography (HPLC) with an Aminex HPX-42A column (300 by 78 mm; Bio-Rad). Double-distilled water was used as the mobile phase at a flow rate of 0.3 ml/min (49). The purified pullulanase was incubated at 65°C with pullulan, starch, glycogen, amylopectin, dextran, cyclodextrins, panose, maltoheptaose, maltohexaose, maltopentaose, maltotetraose (all at a concentration of 0.5% [wt/vol]), as well as with amylose (0.2%, wt/vol).

Samples were withdrawn at different times, and the reactions were stopped by incubation on ice. In order to distinguish between maltotriose (only α-1,4 bonds) and panose or isopanose (α-1,4 and β-1,6 bonds), incubation was performed with α-glucosidase from yeast in 50 mM Tris-HCl (pH 8.5) at 37°C.

Thin-layer chromatography (TLC) of mono- and oligosaccharides was performed on 0.2-mm silica gel-coated aluminum sheets (type 60; Merck, Darmstadt, Germany) with a solvent system consisting of 1-propanol, nitromethane, and water (5:3:2, vol/vol/vol). Carbohydrate spots were visualized by spraying the chromatograms with aniline-diphenylamine reagent (1% [wt/vol] diphenylamine and 1% [vol/vol] aniline in acetone mixed with 0.1 volume of 85% phosphoric acid just before use) and incubating the preparations at 160°C for 10 min.

Effects of metal ions and other reagents on pullulanase (rPulAg′) activity.

The effects of various substances on pullulanase (rPulAg′) activity were examined after incubation of the purified and extensively dialyzed enzyme (final concentration, 0.2 U/ml) with metal ions and other reagents at various concentrations at 65°C for 10 min. Samples were withdrawn, cooled on ice, and tested for pullulanase activity.

N-terminal polypeptide sequence analysis.

The N-terminal sequence of the purified truncated pullulanase was determined by automated Edman degradation by using a pulsed liquid sequencer (model 473A; Applied Biosystems, Foster City, Calif.) connected online to an HPLC apparatus for phenylthiohydantoin derivative identification; the procedures suggested by the manufacturer were used. DNA sequencing was performed with an ABI automatic DNA sequencer by using primer extension in both directions.

Sequence analysis.

A DNA sequence analysis was carried out by using the Lasergene program for Windows (DNAStar Inc., Madison, Wis.). Multiple alignment was performed by using the ClustalW algorithm (55). BLAST and FASTA algorithms were used to search the databases (2). Signal sequence prediction was carried out by using the SignalP program for UNIX (46).

Nucleotide sequence accession numbers.

The nucleotide sequences of the A. horikoshii and A. gottschalkii pullulanases have been deposited in the GenBank database under accession numbers AY217725 and AY541591, respectively.

RESULTS

Sequence analysis of the pullulanases from A. horikoshii and A. gottschalkii.

A library of A. horikoshii DSM 9786 genomic DNA was constructed in λZAP, and the mass-excised λ phage library was screened with AZCL-pullulan and amylose on LB agar plates at 37°C. The complete plasmid insert (3,054 bp) of the pullulanase-positive E. coli clone PULL1012 was sequenced. The pullulanase gene (A. horikoshii pul) has a putative TTG translational start site, is 2,598 bp long, has a G+C content of 37.6%, and encodes an enzyme with a predicted molecular mass of 98.8 kDa. The postulated Shine-Dalgarno sequence (AGGAGG) lies 9 to 14 bp upstream of the predicted translational start codon. In addition, a signal sequence consisting of 23 amino acids with predicted (43) cleavage taking place between a glycine residue and a cysteine residue was found, indicating that there is lipoprotein signal peptide processing. Enzyme assays carried out with E. coli PULL1012 cell lysate indicated that there was high activity with pullulan and no significant activity with soluble amylose, which was expected for a type I pullulanase (data not shown).

The A. gottschalkii pullulanase gene, whose sequence is 86% identical to the sequence of the corresponding gene from A. horikoshii, is also 2,598 bp long and encodes a protein consisting of 865 amino acids with a predicted molecular mass of 98.8 kDa before processing. The G+C content of rPulAg is 36.5%. A 23-amino-acid lipoprotein signal peptide with cleavage taking place between cysteine and glycine residues is predicted. The primary sequences of the pullulanases from A. gottschalkii and A. horikoshii are 94% identical, and both of these proteins exhibit 39% pairwise amino acid identity with the type I pullulanases from the anaerobic extreme thermophiles T. maritima (5) (EMBL accession number AJ001087) and F. pennivorans (EMBL accession number AF096862) and 40% amino acid identity with the type I pullulanase from B. acidopullulyticus (EMBL accession number AX203843). The highly conserved seven-residue motif typical of all type I pullulanases (8), YNWGYDP, was detected at positions 446 to 452 on the N-terminal side of region I (Table 1). The four regions conserved in all amylolytic enzymes belonging to glycoside hydrolase family 13 were identified in the PulA sequence, and the conserved amino acid region YNWGYDP, which was previously found to be conserved among all type I pullulanases (7, 8), was also present (Table 1). Construction of a phylogenetic tree based on pullulanase sequences from different bacterial sources showed that both Anaerobranca enzymes may be evolutionarily closely related to the type I pullulanases from F. pennivorans (8), T. maritima (9), and B. acidopullulyticus (29) and that these enzymes clearly comprise a separate cluster. Nevertheless, some homology to mesophilic alkaliphilic pullulanases from Bacillus species was observed (Fig. 1).

TABLE 1.

Regions conserved among type I pullulanases^a

Source	YNWGYDP		Region I		Region II		Region III		Region IV
	Position	Sequence	Position	Sequence	Position	Sequence	Position	Sequence	Position	Sequence
Anaerobranca gottschalkii	446	YNWGYDP	487	GIRVIKDMVYNHT	559	DGFRFDLMAL	591	YGEPWQA	666	PTESIVYVSCHDNLTLWD
Bacillus acidopullulyticus	436	YNWGYDP	476	RIAINMDVVYNHT	549	DGFRFDLMAL	580	YGEPWTG	656	PSETINYVTSHDNMTLWD
Fervidobacterium pennivorans	424	YNWGYDP	465	GIRVILDMVFPHT	538	DGFRFDQMGL	570	YGEPWGG	648	PQETINYVEVHDNHTLWD
Thermotoga maritima	430	YNWGYDP	473	FTGVIMDMVFPHT	548	DGFRFDQMGL	580	YGEPWGG	638	PEETINYAACHDNHTLWD
Bacteroides thetaiotaomicron	234	YNWGYDP	275	GIRVIMDVVYNHT	347	DGFRFDLMGI	379	YGEGWAA	670	PVQMISYVSCHDGLCLVD
Thermus sp.	272	YNWGYNP	333	GLRVVMDAVYNHV	405	DGFRFDLMGV	437	YGQGWDL	512	PRQSINYVECHDNHTFWD
Bacillus sp. strain KSM-1876b	1349	YNWGYNP	1490	DMGVVLDVVFNHT	1459	DGFRFDMMGD	1491	IGEGWVF	1470	PGDVVQYIEAHDNLTLYD
Caldicellululosiruptor saccharolyticus	405	YNWGYDP	445	GIGVVMDVVFNHT	520	DGFRFDLMGL	552	YGEGWVM	629	PDECVNYVSCHDNLTLFD
Klebsiella aerogenes	574	YNWGYDP	615	GMNVIMDVVYNHT	691	DGFRFDLMGY	723	FGEGWDS	842	PTEVVNYVSKHDNQTLWD
Klebsiella pneumoniae	562	YNWGYDP	603	GMNVIMDVVYNHT	679	DGFRFDLMGY	711	FGEGWDS	830	PTEVVNYVSKHDNQTLWD
Bacillus stearothermophilus	572	YNWGYNP	613	GMNVIMDVVYNHT	689	DGFRFDLMGY	721	FGEGWDS	456	PTEVVNYVSKHDNQTLWD

^aThe accession numbers of the sequences are listed in the legend to Fig. 1. The underlined amino acid residues are the residues identified by Nakajima (44) as being highly conserved. Q in region III of the Thermus pullulanase may be a G/C sequencing error.

^bThe enzyme can be considered to have a domain with type I pullulanase activity.

FIG. 1.

Relatedness tree for pullulanases that hydrolyze α-1,6 glycosidic linkages in pullulan, as inferred from an amino acid sequence alignment. The sequences of enzymes from the following organisms were aligned: A. gottschalkii (this study), A. horikoshii (this study) (accession number AY217725), T. maritima (accession number AJ001087), F. pennivorans (accession number AF096862), Clostridium perfringens (accession numbers NC-003366.1 and NP-562468), Bacteroides thetaiotaomicron (accession number U67061), Bacillus halodurans (accession number NC-002570), Bacillus stearothermophilus (accession number E03513) (patent JP 1992099489-A), B. acidopullulyticus (EMBL accession number AX203843), C. saccharolyticus (accession number L39876), Bacillus subtilis (accession number NC-000964), Thermus thermophilus (accession number AB054527), Clostridium acetobutylicum (accession number 003030), Deinococcus radiodurans (accession number NC-001263), Streptococcus pneumoniae (accession number NC-003098), Bacillus sp. (accession number D78258), Bacillus sp. strain KSM-1876 (accession number AB49812), and K. pneumoniae (accession number X52181). Scale bar = 10 amino acid changes per 100 amino acids.

Purification of rPulAg from E. coli.

The recombinant pullulanase from A. gottschalkii without the signal peptide sequence was expressed in strain E. coli BL21(DE3). The pullulanase was purified 5.3-fold by heat treatment of the crude cellular extract for 20 min at 70°C, followed by two chromatographic separation steps (see Materials and Methods). The purification results are summarized in Table 2. When this method was used, about 10 mg of >95% pure pullulanase (Fig. 2A) with a specific activity of 56 U/mg was obtained, which corresponded to a yield of 10% of the total activity in the crude extract. As determined by SDS-PAGE analysis, the purified protein had a molecular mass of about 96 kDa, which is close to the expected molecular mass deduced from the amino acid sequence of A. gottschalkii pullulanase without the signal peptide. Native gel electrophoresis and gel filtration analysis indicated that the full-length rPulAg is a monomer (data not shown).

TABLE 2.

Purification of rPulAg and rPulAg′ after expression in E. coli

Enzyme	Prepn	Total protein (mg)	Total activity (U)a	Sp act (U/mg)a	Recovery (%)	Purifi- cation (fold)
rPulAg	Crude extract	296	3,240	11	100	1
	Heat treatment	98	1,914	19	59	2
	Source 15Q	26	1,238	47	38	4.7
	Phenyl-Sepharose	10	560	56	17	5.3
rPulAg′	Crude extract	640	841	1.3	100	1
	Q-Sepharose	4.3	42	9.8	5.0	7.5
	Phenyl-Sepharose	2.2	35	16	4.2	12
	Superdex S-200	0.85	28	43	3.3	33

^aOne unit of pullulanase catalyzed the formation of 1 μmol of reducing sugar per min under the defined conditions.

FIG. 2.

Purification of rPulAg. Samples of full-length (PulAg) and truncated pullulanase (PulAg′) from various purification steps were subjected to SDS-PAGE analysis. The sizes of molecular mass markers are indicated on the left. (A) Purification of rPulAg from E. coli BL21(DE3)(pET21a-pulAg). Lane 1, marker proteins; lane 2, crude extract (25 μg); lane 3, heat-treated crude extract (20 μg); lane 4, Source 15Q pool (5 μg); lane 5, Phenyl-Sepharose pool (5 μg). (B) Purification of rPulAg′ from E. coli Tuner(DE3)(pLysS)(pET-Blue1-pulAg). Lane 1, crude extract (15 μg); lane 2, Q-Sepharose pool (15 μg); lane 3, Q-Sepharose pool after heat treatment (10 μg); lane 4, Phenyl-Sepharose pool (6 μg); lane 5, gel filtration pool (0.5 μg).

Purification of an N-terminally truncated form of pullulanase, rPulAg′.

The A. gottschalkii pullulanase gene, including its signal peptide-encoding sequence, was cloned into the pET-Blue-1 expression vector as described in Materials and Methods, yielding the E. coli clone CBpulAg. The entire 2.6-kb insert was sequenced in both directions, which demonstrated that no substitutions or mutations occurred during the cloning procedure. During growth of the strain in LB broth, samples were collected every 4 h to determine optimal pullulanase expression. The total enzymatic activity was found in the soluble fraction of the sonicated cells and not in the pellet, demonstrating that the enzyme did not precipitate as insoluble aggregates. Furthermore, pullulanase activity was absent in the culture supernatant. The crude extract of E. coli cells grown to the early stationary phase showed the highest activity and was used for purification of the pullulanase. The pullulanase activity was purified 33-fold from the crude extract of E. coli CBpulAg by anion-exchange chromatography, heat treatment at 65°C for 60 min, hydrophobic interaction chromatography, and size exclusion chromatography, which yielded a >95% pure enzyme with a specific activity of about 43 U/mg (Table 2). Proteins from the purification steps were separated by SDS—12% PAGE, which revealed an apparent molecular mass of 70 kDa for the purified enzyme (Fig. 2B). The molecular mass of the native protein calculated by gel filtration on Superdex S-200 (69 ± 3 kDa) indicated that rPulAg′ was a monomer.

Since this size differed dramatically from the size predicted from the gene sequence (96 kDa after removal of the signal peptide), Edman degradation analysis of rPulAg′ was carried out, which revealed the sequence EFYYPGDDLGNTYTRNSTK. The sequence contained 251 amino acid residues from the start of the predicted A. gottschalkii pullulanase primary structure. Intriguingly, only two codons upstream of the sequence encoding the N terminus of the truncated pullulanase (rPulAg′), we found a second putative TTG translational start codon within the A. gottschalkii pullulanase ORF and a possible Shine-Dalgarno region (5′-AGGAGGG-3′) 9 bp upstream of this point. The translation product obtained from this start site yielded a smaller protein (617 amino acids with a predicted molecular mass of 70 kDa). Production of the truncated form of the pullulanase may have been due either to translation initiation at this alternative start site or to posttranslational cleavage by E. coli proteases. The formation of the 70-kDa pullulanase could not be prevented by addition of the potent protease inhibitor Pefabloc (Roche) during preparation of the crude extract.

rPulAg′ was found to be catalytically active. This fact demonstrates that the first 251 amino acids of the pro-pullulanase (or 229 residues from the predicted N terminus of the mature protein) are not necessary for catalytic activity.

Purification of the native pullulanase from A. gottschalkii.

Analysis of the authentic pullulanase produced by A. gottschalkii cells shed light on the true size of the enzyme as it occurs in its natural producer organism. The culture supernatant, concentrated by ultrafiltration, of a large-scale fermentation of A. gottschalkii yielded very little pullulanase activity (data not shown). Therefore, cells of A. gottschalkii containing cell-associated enzyme were used as a source for the wild-type pullulanase. The specific activity of the pullulanase in the cell sonicate was 0.022 U/mg. Partial purification of the pullulanase by Q-Sepharose chromatography resulted in a 15-fold increase in specific activity, to 0.33 U/mg. The activity at this stage of purification was sufficient to perform a partial characterization of the native enzyme. The zymogram pattern on SDS-PAGE gels had an activity band at 96 kDa (data not shown), which matched the molecular mass predicted from the sequence and confirmed that the first TTG is the correct translational start site. The was no activity band at 70 kDa, indicating that the second putative translational start site (745 bp into the ORF) is not used in A. gottschalkii. Nevertheless, the possibility that under certain growth conditions a truncated form of the enzyme may be released into the culture supernatant of the anaerobic thermoalkaliphile cannot be excluded.

Comparative characterization of rPulAg and rPulAg′.

The ranges of temperature and pH over which the A. gottschalkii pullulanases (both rPulAg and rPulAg′, as well as the partially purified authentic enzyme) were active and stable were determined with pullulan as the substrate. All pullulanase forms were active at pH values ranging from 5.0 to 10.0, and the optimum pH was 8.0 to 8.5 (Fig. 3A); all of the forms retained at least 60% of the maximal activity at pH 9.0. All enzyme forms were active at temperatures between 40 and 80°C. By using the standard 10-min assay, the highest pullulan hydrolysis rate was observed at 65 to 70°C for all enzymes (Fig. 3B). The enzyme was stable after several days of incubation at 4°C at pH values around 8.5, but it was unstable at pH values below 4.0 and above 10.0 (data not shown).

FIG. 3.

(A and B) Influence of pH on the activity and influence of temperature on the activity and stability of the pullulanase from A. gottschalkii. The effects of pH (A) and temperature (B) were determined for rPulAg (▴), rPulAg′ (▪), and a partially purified authentic enzyme from A. gottschalkii cells (○). (C) Effect of long-term incubation of rPulAg at elevated temperatures. Dilute enzyme (20 μg/ml) in 40 mM sodium phosphate-citrate buffer (pH 8.0) was incubated at 50°C (○), 65°C (□), 70°C (▴), 75°C (▾), and 80°C (•). After various periods of time, samples were withdrawn, and the residual activity was measured with the standard assay.

The resistance of rPulAg and rPulAg′ to thermal denaturation was evaluated by short-term and prolonged incubation over a temperature range from 50 to 80°C. No loss of enzyme activity was observed after incubation of purified samples of rPulAg or rPulAg′ at 50 or 65°C for several hours (Fig. 3C and data not shown), and 80% of the initial activity was detected after 6 h of incubation at 70°C. Both enzymes, however, were rapidly inactivated during incubation at 75°C (the half-lives were less than 10 min) or at a higher temperature (Fig. 3C). A long-term thermal inactivation experiment with rPulAg revealed a half-life of 22 h at 70°C (data not shown), demonstrating the remarkable thermal resistance at a temperature above the upper growth temperature limit of the natural producer, A. gottschalkii. rPulAg and rPulAg′ do not require Ca²⁺ ions or any additional stabilizing compounds for stability. The apparent kinetic parameters for Michaelis-Menten-like pullulan (0.1 to 5 mg/ml) hydrolysis were determined at pH 8 at 60°C, which resulted in K_m values for rPulAg and rPulAg′ of 0.75 and 0.83 mg/ml, respectively, and V_max values of 61 and 70 U/mg, respectively (data not shown).

Substrate specificity of rPulAg and rPulAg′ and analysis of hydrolysis products.

Purified rPulAg was examined to determine its ability to hydrolyze various carbohydrates in 40 mM sodium phosphate-citrate buffer (pH 8.0) by using final concentrations of 0.2% (wt/vol) for amylose and 0.5% (wt/vol) for the other polysaccharides. The enzyme was added at a concentration of 20 μg/ml, and the preparations were incubated at 60°C for 0.5 to 16 h before the products were subjected to TLC and/or HPLC analysis. For substrates identified in this way, the specific activity was also measured with the standard dinitrosalicylic acid assay. The pullulanase from A. gottschalkii preferentially hydrolyzed pullulan (Fig. 4A) and other carbohydrates containing α-1,6 glycosidic linkages, including amylopectin (14% activity relative to the pullulan-cleaving activity) (Fig. 4B), soluble starch (15%), β-limit dextrin (24%) (Fig. 4C), and glycogen (3%), while amylose was not significantly hydrolyzed. rPulAg′ had a similar substrate profile (β-limit dextrin was not tested) (data not shown). No reaction was observed with various other polysaccharides (dextran, cellulose, laminarin, mannan), α-, β-, and γ-cyclodetrins, various di- and trisaccharides (isomaltose, trehalose, turanose, cellobiose, melibiose, sucrose, panose, melizitose, raffinose), and linear small α-1,4-linked maltodextrins (data not shown).

FIG. 4.

TLC analysis of product formation with pullulan, amylopectin, and β-limit dextrin. The reaction mixtures (200 μl) contained 0.5% pullulan (A), 0.25% amylopectin (B), or 0.25% β-limit dextrin (C) dissolved in 40 mM sodium phosphate-citrate buffer (pH 8.0) and 2 μg (pullulan digestion) or 5 μg (amylopectin and β-limit dextrin digestion) of purified rPulAg and were incubated at 60°C. During pullulan hydrolysis (A), samples were withdrawn after 0, 30, 60, 120, and 180 min (lanes 2 to 6) and subjected to TLC analysis by using 1-propanol-nitromethane-H₂O (5:3:2, vol/vol/vol) as the solvent system. In the amylopectin (B) and β-limit dextrin (C) degradation experiments samples were withdrawn after 0, 10, 20, 30, 40, 60, 90, and 120 min (lanes 1 to 8) and separated on TLC plates with 1-propanol-ethyl acetate-H₂O (6:1:3, vol/vol/vol). G1 to G6, glucose through maltohexaose standard oligosaccharides.

HPLC analysis showed that both recombinant enzymes completely hydrolyzed pullulan to maltotriose (data not shown). In order to confirm that the hydrolysis product from pullulan was maltotriose (three glucose residues linked by two α-1,4 glycosidic linkages) and not panose or isopanose (both possible glucose trisaccharides with one α-1,4 and one α-1,6 glycosidic linkage), the products of pullulan hydrolysis were incubated with α-glycosidase from yeast. The formation of glucose as the main product confirmed the formation of maltotriose (and not panose) from pullulan (data not shown).

The substrate and product profiles described above indicate that rPulAg specifically attacks α-1,6 linkages of branched oligosaccharides and therefore can be classified as a type I pullulanase. Hence, the pullulanase from A. gottschalkii is the first thermoalkalistable type I pullulanase from an anaerobic thermoalkaliphilic microorganism that has been described.

Effects of metal ions and other reagents.

The effects of metal ions and different reagents on A. gottschalkii pullulanase were tested. The presence of NaCl or KCl at a concentration of 50 to 250 mM did not affect activity. Also, addition of divalent cations, such as Ba²⁺, Sr²⁺, Ca²⁺, Mg²⁺, Mn²⁺, Ni²⁺, or Fe²⁺, all as chloride salts at a concentration of 1 mM, had no significant effect on activity (with all ions the activity was >90% of the activity of the control without ions). On the other hand, addition of 1 mM Co²⁺and addition of 1 mM Cu²⁺ reduced the pullulanase activity by 30 and 43%, respectively, and addition of Zn²⁺, Cd²⁺, or Hg²⁺ completely inactivated the enzyme. None of the compounds tested had a positive effect on pullulanase activity. EDTA was not inhibitory even at high concentrations (0.5 M), indicating that divalent metal ions are not required for enzyme function. In agreement, the pullulanase primary structure lacks typical Ca²⁺ binding residues found in other members of glycoside hydrolase family 13. Truncated A. gottschalkii pullulanase was quite stable during incubation at 50°C for 1 h in the presence of different surfactants, such as 0.1% SDS, 20% Triton X-100, 20% polyoxyethylenesorbitan monolaurate (Tween 20), and 20% polyoxyethylenesorbitan monooleate (Tween 80), which led to no loss of activity. As observed for most pullulanases, rPulAg′ activity was completely eliminated by low concentrations of N-bromosuccinimide (0.01%), which oxidizes tryptophan residues. The rPulAg was also very susceptible to alkylating, mercaptide-forming, or oxidizing reagents, such as iodoacetamide, iodoacetic acid, and p-chloromercuribenzoate (data not shown).

rPulAg was partially inhibited by α-, β-, and γ-cyclodextrins. Added to standard reaction mixtures at a concentration of 1%, these cyclic oligosaccharides inhibited the pullulan-cleaving activity of rPulAg by 37, 56, and 26%, respectively. The K_i for β-cyclodextrin was determined to be 6 mM (data not shown). Inhibition by β-cyclodextrin, but with a much higher affinity for the inhibitor (K_i, 0.075 mM), has also been reported for the type I pullulanase PulA from T. maritima (34).

DISCUSSION

The industrial applications of alkaliphilic microorganisms have been recognized for a long time, and enzymes such as alkaline proteases, alkaline amylases, and alkaline cellulases are on the market (22, 24). The main application of alkalistable enzymes is in the detergent industry. Detergent enzymes account for approximately 30% of the total enzyme production worldwide (22, 23, 24). The essential requirements for optimal performance of a detergent enzyme are high activity and stability in the temperature range from 40 to 60°C under alkaline pH conditions. The use of pullulanase in detergents has not been thoroughly investigated yet (24). The enzyme from A. gottschalkii could support amylases as detergent additives by concerted action in removal of starch spots in the presence of detergents. Until very recently, little attention has been paid to isolating microorganisms that are able to grow at high pH values and elevated temperatures. One of the reasons for this is the difficulty of isolating microorganisms that prefer this combination of extreme conditions. The strategies that have developed to stabilize cell walls, membranes, and proteins in order to cope with alkalinity at a high temperature are not well understood. The only bacteria identified to date that are able to grow at a pH above 9.5 and a temperature between 55 and 70°C are the recently described species A. horikoshii (15) and A. gottschalkii (48). These organisms represent a promising source for production of extracellular biocatalysts with properties that are ideal for application in the detergent industry. In particular, A. gottschalkii is interesting in this respect because of its ability to grow on various different polysaccharides and proteinaceous substrates at pH 9.5 and temperatures up to 65°C (48).

Analysis of the amino acid sequences derived from the A. horikoshii and A. gottschalkii pullulanase genes revealed the YNWGYDP motif that is common to type I pullulanases and the four conserved regions (regions I, II, III, and IV) which are typical of all amylolytic enzymes that belong to glycoside hydrolase family 13 (the so-called α-amylase family) (Table 1). These regions are involved in the active site architecture within the catalytic domain of glycoside hydrolase family 13 enzymes and encompass the catalytic triad of acidic residues (20, 40, 44), which consists of residues Asp493, Glu593, and Asp677 for the A. gottschalkii pullulanase. The pattern of enzyme action against the substrate for pullulanases differs from that for α-amylases, indicating that other amino acids in addition to the three residues mentioned above are involved in the catalytic activity (40, 42, 57). Other amino, acids such as the Trp595 and Trp682 residues in regions III and IV, respectively, may be involved in substrate binding or catalysis. The Trp595 residue corresponds to Trp497 in region III of isoamylase of Pseudomonas amyloderamosa (3, 27), which specifically hydrolyzes α-1,6 glycosidic linkages in starch. Moreover, the histidine residues, which have been found to participate specifically in catalysis of α-1,6 glycosidic linkages by Klebsiella aerogenes pullulanase (57), are also present in A. gottschalkii pullulanase in regions II and IV at positions His498 and His676, respectively.

The pullulanase described in this study is the first enzyme which has been cloned and expressed at a high level from A. gottschalkii. Expression of the entire prepro-pullulanase coding sequence in E. coli yielded the truncated enzyme rPulAg′, which was 26 kDa smaller than the size expected on the basis of the sequence of the A. gottschalkii pullulanase ORF. In contrast, the molecular mass of the enzyme produced from an expression construct lacking the signal peptide-encoding region of the gene, rPulAg, was 96 kDa, which corresponds to the molecular mass of the authentic enzyme isolated from A. gottschalkii cells. The truncation of rPulAg′ may have been due either to translation initiation at the false TTG start site 745 bp within the ORF or to the action of E. coli proteases. Since the 70-kDa enzyme form was produced despite addition of the protease inhibitor Pefabloc during purification, it seems more likely that use of the alternative translation initiation site by E. coli is the reason for production of the truncated protein. A second potential translational start site was also found in the pullulanase gene of F. pennivorans (8). Mutation of this alternative translation start site in the gene led to production of the 93-kDa full-size F. pennivorans pullulanase in E. coli without the smaller 83-kDa enzyme (8). The in silico domain analysis of the pullulanase suggested that there may be an N-terminal PUD domain which might be involved in substrate binding. In the case of the truncated pullulanase, N-terminal deletion that includes this possible PUD domain may be the reason for the lower specific activity of the truncated pullulanase (43 U/mg) than of the full-length enzyme (56 U/mg). It should be noted that most of the properties of the pullulanase were not affected by the truncation.

The size of the authentic A. gottschalkii pullulanase and the full-length recombinant form (rPulAg) (about 96 kDa) is common among the type I pullulanases, Also, the monomeric quaternary structure resembles that of the other debranching enzymes described so far (1, 9, 13, 25, 34, 36, 43, 47, 53) with the exception of F. pennivorans pullulanase, which has a dimeric structure (8). The temperature and pH profiles of rPulAg and its remarkable resistance to thermal inactivation at temperatures up to 70°C (half-life at this temperature, 22 h) are in agreement with the optimal growth conditions of A. gottschalkii (48). To our knowledge, this enzyme is the first purified thermoalkalistable type I pullulanase from an anaerobic microorganism that has been described.

Interestingly, the A. gottschalkii and A. horikoshii pullulanases appear to be lipoproteins in their natural hosts. According to the primary structures deduced from the gene sequences, these enzymes are produced as precursors with typical lipoprotein signal peptides, which are characterized by a so-called lipobox with the consensus sequence L-(A/S)-(A/G)-C at positions −3 to 1 (56). The invariable cysteine residue at position 1 is the target for diacylglycerol modification and becomes the first residue of the mature enzyme after signal peptide removal. It is noteworthy that a well-studied pullulanase from Klebsiella oxytoca (Klebsiella pneumoniae), which was instrumental in determining the mechanism of type II protein secretion in gram-negative bacteria, is also a lipoprotein. Assignment of the Anaerobranca pullulanases to the lipoprotein group and thus their probable cell membrane association explain the absence of pullulanase activity in the culture supernatant of A. gottschalkii, whereas it was possible to partially purify the enzyme from cells of this organism. In addition, this may be the reason for the failure to obtain the full-length 96 kDa pullulanase by expression of the gene, including its lipoprotein signal peptide-encoding sequence, as high-level production of a foreign lipoprotein in E. coli may be problematic.

The full-length and truncated forms of the recombinant pullulanase from A. gottschalkii are among the few enzymes that are active at an alkaline pH (pH 8.0 to 9.0) and a relatively high temperature (65 to 70°C). As reported for the pullulanase from C. yticus (1), the F. pennivorans 83-kDa polypeptide (8), and the truncated forms of the alkaline pullulanase from Bacillus sp. strain 1 (38), removal of a long N-terminal peptide does not compromise the catalytic activity and the substrate binding and does not have a great effect on the physicochemical characteristics (specificity and sensitivity) of the enzymes, which was also observed in this work for rPulAg and rPulAg′.

Many starch-hydrolyzing enzymes are strongly inhibited by N-bromosuccinimide, which oxidizes tryptophan residues. This observation suggests that a Trp residue(s) may play an important role in substrate binding or catalysis in rPulAg, as has been observed for an α-amylase (40), a pullulanase (4), and cellulases (28, 59). The sensitivity of rPulAg to thiol group-reactive agents indicates that cysteinyl groups are involved in the active site or are important for the enzyme structure. Three cysteine residues are indeed present in the amino acid sequence of the pullulanase. As has been reported previously for an alkaline amylase (18, 19) and a pullulanase (38, 39), the A. gottschalkii pullulanase is very stable in response to a high concentration (0.5 M) of EDTA. The resistance of rPulAg to chelating agents is interesting since these reagents are, in general, indispensable ingredients in detergent formulations (18). In addition, the stability of the recombinant A. gottschalkii enzyme in the presence of detergents makes it suitable as an additive for use in laundry and dishwashing detergents.