|Home | About | Journals | Submit | Contact Us | Français|
The naturally occurring structural isomer of sucrose, trehalulose, is produced by sucrose isomerase (SI). Screening of chromosomal DNA from “Pseudomonas mesoacidophila” MX-45 with an SI-specific probe facilitated the cloning of two adjacent gene homologs, mutA and mutB. Both genes were expressed separately in Escherichia coli, and their enzyme products were characterized. MutA hydrolyzed the substrates trehalulose, isomaltulose, and sucrose into glucose and fructose. Due to its highest activity on trehalulose, MutA was referred to as trehalulase. mutB encodes the SI (trehalulose synthase) and catalyzes the isomerization of sucrose to mainly trehalulose. From Northern blot analysis it is apparent that the mutB gene is not transcribed as part of an operon and was transcriptionally upregulated when P. mesoacidophila MX-45 cells were grown in sucrose medium, whereas under investigated conditions no transcript for mutA was detected. Mutants of mutB were created by a random mutagenesis approach in order to alter the product specificity of MutB. Two types of mutants have emerged, one type that prefers the hydrolytic reaction on sucrose and another type that still acts as an SI but with a significant shift in the product from trehalulose to isomaltulose. The hydrolytic character of MutB R311C was demonstrated through its higher catalytic efficiency for glucose production over trehalulose production. MutB D442N favored the transfer reaction, with an isomer preference for isomaltulose.
Trehalulose (α-d-glucopyranosyl-1,1-d-fructose) and isomaltulose (α-d-glucopyranosyl-1,6-d-fructofuranose, commonly referred to as palatinose) are naturally occurring isomers of sucrose (α-d-glucopyranosyl-1,2-β-d-fructofuranoside). Some beneficial properties over sucrose make them valuable substitutes for use in food industries. Since they are acariogenic (11, 23), digested and absorbed more slowly than sucrose (22, 38), and attenuate insulin levels in the bloodstream (15), they can be applied in diabetic and sports food and drinks. Besides these advantages as a food ingredient, their reducing properties make them attractive precursors for chemical modifications, and they are used to manufacture sugar alcohols consumed as low-caloric sweeteners (35) and biocompatible polymers (21).
Commercial isomaltulose and trehalulose are produced from sucrose by enzymatic conversion. Sucrose isomerase (SI) enzymes (EC 184.108.40.206) catalyze the isomerization of sucrose into trehalulose and isomaltulose as main products. In addition, glucose and fructose are produced as by-products from the hydrolytic reaction of SIs. The product ratio depends mainly on the enzyme origin but also on reaction conditions, particularly temperature and pH (42). Some purified enzymes produce predominantly isomaltulose (75 to 80%), and others produce predominantly trehalulose (90%). As a consequence of their product specificity, the former enzymes are known as isomaltulose synthases, whereas the latter enzymes are referred to as trehalulose synthases. The formation of multiple products lowers the yield and complicates the recovery of the desired isomer. An ideal SI for industrial application would exhibit complete conversion and high specificity at high substrate concentrations. To improve enzyme properties, SI genes might serve as templates for screening or rational design.
The presence of SIs has been reported in numerous microorganisms including Erwinia, Klebsiella, Pantoea, Protaminobacter, Pseudomonas, and Serratia (1, 7, 8, 16, 20, 25, 26, 41, 45, 46). While several isomaltulose-specific SI enzymes from bacteria have been characterized and some of their genes have been cloned (3, 24, 47, 50), only a few SI enzymes with trehalulose specificity are known (26-28, 33). A trehalulose synthase from “Pseudomonas mesoacidophila” MX-45 has already been purified by Nagai et al. (27).
In this paper we report the cloning of the trehalulose synthase gene mutB from P. mesoacidophila MX-45 and demonstrate the existence of an adjacent SI-homologous gene (mutA), which encodes a hydrolytic enzyme acting preferentially on trehalulose. In order to determine whether mutA and mutB are cotranscribed in P. mesoacidophila, Northern blot experiments were performed.
The heterologously produced MutA and MutB proteins were purified and biochemically characterized. MutB was analyzed in more detail since this protein was used for crystallization trials, as we published recently (29, 31). During this study active-site mutants of MutB were used to resolve the crystal structure in complex with a substrate. The construction of these mutants is described hereafter. Further structural studies of SI have been reported for the isomaltulose synthase PalI from Klebsiella sp. strain LX3 (48) as well as a proposed unique RLDRD motif that might control the product specificity of SIs. Recently, we succeeded in the elucidation of the crystal structure of SmuA from Protaminobacter rubrum (30, 32). By comparison of the crystal structure of the isomaltulose synthases PalI, SmuA, and MutB (30), it was still not clear what structural features are responsible for their different product specificities. To elucidate the structural requirements for different product specificities, we attempted to create mutB mutants through random mutagenesis. Identification of MutB mutants with single amino acid alterations may be of special interest as they can be used as templates for structure determination in order to decipher features involved in different product specificities.
Trehalulose (purity 80%), isomaltulose, an authentic sample of trehalulose, and P. mesoacidophila MX-45 were obtained from Südzucker AG. P. mesoacidophila MX-45 was grown at 30°C in Pseudomonas medium (Difco Medium P) supplemented with peptone (2%), K2SO4 (1%), MgCl2 (0.14%), and sucrose (4%) or in M9 minimal medium supplemented with yeast extract (0.05%), MgSO4 (1 mM), CaCl2 (0.1 mM), and a carbon source (0.2%). For standard molecular biology techniques like DNA cloning, plasmid preparation, and DNA sequencing, Escherichia coli JM109 was used as a host. Competent E. coli DH5α cells were used for electroporation. E. coli strains were grown in double yeast tryptone (dYT) medium (16g of Bacto tryptone, 10 g of yeast extract, 5 g of NaCl, and 1 liter of deionized water) at a relevant temperature. When necessary, a selective antibiotic was added (100 μg ml−1 ampicillin).
Recombinant DNA techniques were performed by conventional protocols (34). PCRs were performed in 100-μl reaction mixtures using Taq DNA polymerase (Biomaster GmbH) or an Expand High Fidelity PCR kit (Roche Applied Science). Plasmids were prepared with a QIAprep spin miniprep kit (Qiagen GmbH). Genomic DNA was prepared by CsCl dye buoyant density centrifugation according to Sedlmeier and Altenbuchner (36). Total RNA was prepared with an RNeasy minikit (Qiagen GmbH). Cells of P. mesoacidophila MX-45 and E. coli harboring plasmids were harvested during log-phase growth. RNA stabilization of approximately 6 × 108 cells was done with RNAprotect bacteria reagent (Qiagen GmbH), and cells were stored at −20°C. For subsequent RNA isolation cell pellets were thawed and lysed, and total RNA was isolated according to the RNeasy minikit protocol for bacteria. E. coli transformation was performed according to the transformation and storage solution method (9).
The SI screening primers S633 (5′-TGGTGGAARGARGCTGT-3′) and S634 (5′-TCCCAGTTCARGTCAGGCTG-3′) were used in a PCR with chromosomal DNA from P. mesoacidophila MX-45. The obtained 500-bp PCR fragment was inserted into the positive selection vector pJOE4786 (2), which was cut with EcoRV to create plasmid pHWG705. The inserts from pHWG705.1 and pHWG705.6 were cut out with BamHI, digoxigenin (DIG) labeled by random-primed DNA labeling (DIG High Prime; Roche Applied Science), and used as probes mutA and mutB, respectively.
Five separate restriction enzyme digests were performed in duplicate on genomic DNA from P. mesoacidophila MX-45 and size fractionated on 0.8% agarose gels. The fragments were transferred on nylon membrane, using standard techniques, and probed with DIG-labeled mutA and mutB probes separately. Detection was done using a DIG detection kit (Roche Applied Science).
DNA sequences were obtained using a Pharmacia automated laser fluorescence sequencer, an AutoRead Sequencing kit, and a Repro Gel Long system (Amersham Pharmacia Biotech/GE Healthcare), and oligonucleotides were obtained from MWG Biotech. The DNA insert in plasmid pHWG693.11 was digested with various restriction enzymes. The fragments were inserted into pUCBM20, and a universal primer (CAGGAAACAGCTATGAC) and reverse primer (CGACGTTGTAAAACGACG) were used for sequencing. The gaps were closed with primers designed from obtained sequences. Adjacent and overlapping sequence information was obtained from a 4.5-kb HindIII genomic DNA fragment, which hybridized with the mutA probe and was cloned in pHWS173.1. The plasmid pHWG692.1 contains a 3.4 kb chromosomal BglII restriction fragment and was found after cloning and colony hybridization with the mutB probe. The sequences were aligned with Seqman software (Dnastar). Annotations were done with the BLAST programs at the electronic mail server from the National Center of Biotechnology Information, Bethesda, MD.
For expression in E. coli, the coding region of the trehalulase mutA ranging from codon 2132 to codon 3835 was amplified by PCR, with genomic DNA from P. mesoacidophila MX-45 serving as a template. The gene-specific primers S3883 (5′-AAAACATATGACTGAAAAGTTATCC-3′) and S3884 (5′-GCAAGCTTCATGCGCAGTGGTAG-3′) were used in this reaction, which introduced an NdeI restriction site at the putative N-terminal mutA sequence and a HindIII restriction site just behind the putative stop codon. The fragment was inserted into pJOE2702 (43) cut with the appropriate restriction enzymes, creating the plasmid pHWG653. For construction of MutA with a His6 tag (MutA-His6), the mutA gene was amplified without a stop codon using the N-terminal primer S3883 and S4055 (5′-ATGGGATCCTGCGCAGTGGTAGAC-3′), which introduced a BamHI site for ligation into pJOE4036 (37). In this way, the gene was C-terminally fused in frame to six histidine codons present in the vector to create the plasmid pHWG680.
The entire coding region of the trehalulose synthase mutB, encoded from 4067 to 5821, was amplified by PCR as above by using the oligonucleotide primers S985 (5′-CGGAATTCCATATGCTTATGAAGAGATTATC-3′) and S986 (5′-CGGGATCCTTACTTCACCTTGTAGATG-3′) introducing NdeI and EcoRI restriction sites before the ATG start codon and a BamHI site just behind the translational stop codon, respectively. The obtained PCR fragment was inserted as an NdeI/BamHI fragment into pJOE2702 (43), creating the rhamnose vector derivative pHWG315. The mutB gene without the coding sequence for the first 23-amino-acid signal peptide was amplified by PCR using the primer S4059 (5′-GCCATATGGAGGAGGCCGTAAAG-3′), introducing an NdeI restriction site and an ATG start codon, and the primer S986. After digestion with NdeI and BamHI, the fragment was inserted into pJOE2702 vector as described above to create the plasmid pHWG682.3. A mutB gene variant without the coding sequence for the first 27-amino-acid signal peptide was constructed in a similar manner using the N-terminal primer S4060 (5′-GCCATATGAAGCCGGGCGCGC-3′).
The active-site mutant MutB D354Q was generated by two synthetic oligonucleotides (5′-CCTCGGCAATCACCAGAATCCCCGTGCGG-3′ and 5′-CCGCACGGGGATTCTGGTGATTGCCGAGG-3′; mutations are in boldface) using pHWG315 as a template, following the guidelines of QuikChange site-directed mutagenesis protocol (Stratagene). The mutants MutB D227A and MutB E281Q were constructed in a similar manner and were described previously (29). Transformation of the linear amplified mutant plasmid was done with electrocompetent E. coli DH5α cells and yielded pHWG678.6 (D227A), pHWG679.9 (E281Q), and pHWG684.2 (D354Q). For expression of the genes the plasmids were transformed in E. coli JM109.
Site-specific mutagenesis followed the guidelines of the QuikChange site-directed mutagenesis protocol (Stratagene). For a random mutagenic approach on the mutB gene, N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) was added to an exponential culture (optical density at 600 nm [OD600] of 0.5) of E. coli JM109 harboring pHWG315 to final concentrations of 0.1 to 0.3 mg ml−1. Following the growth overnight at 34°C, plasmids were isolated and transformed into E. coli JM109 and spread on McConkey agar plates supplemented with 1% sucrose, 0.1% rhamnose, and ampicillin.
E. coli JM109 carrying the respective recombinant plasmid was grown on 5 ml of dYT medium containing ampicillin (100 μg ml−1). For expression of the gene, the culture was grown to a cell density (OD600) of 0.3 and then induced by the addition of rhamnose (0.1%) and grown for a further 4 h at 30°C. The cells were harvested by centrifugation (5,000 × g for 10 min at 4°C), resuspended, and concentrated to an OD600 of 20 in 100 mM K-phosphate buffer, pH 6.5, for the MutA protein and in 10 mM Ca-acetate buffer, pH 5.5, for MutB and MutB variants. The enzymes were extracted by sonication. After centrifugation (10,000 × g for 10 min at 4°C), the supernatants of the crude cell extracts were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).
An overnight culture of E. coli JM109/pHWG680 was diluted 100-fold in 50 ml of dYT medium supplemented with ampicillin. After 2 h of growth at 37°C, the culture was shifted to 30°C, and the cells were induced by the addition of 0.2% rhamnose. After further growth for 4 h, the cells were harvested by centrifugation (Sorvall SS34; 5,000 × g for 10 min), washed with 0.1 M K-phosphate buffer, pH 6.5, and disrupted in a high-pressure homogenizer (EmulsiFlex-C5; Avestin, Canada). Cell debris was removed by centrifugation (Sorvall SS34; 20,000 × g for 10 min), and the supernatant (15 ml) was applied onto 1 ml of Ni-nitrilotriacetic acid agarose resin (Qiagen GmbH) in a column using gravity flow. The resin was washed with 10 ml of washing buffer (50 mM K-phosphate, 200 mM NaCl, pH 6.5), and bound protein was eluted with 2 ml of elution buffer (50 mM K-phosphate, 200 mM imidazole, pH 6.5). MutA-containing fractions were combined and applied onto an NAP10 column (GE Healthcare) to remove the imidazole. MutA enzyme was stored in 0.1 M K-phosphate buffer (pH 6.5)-0.02% Na-azide at 4°C.
The purification of MutB wild-type proceeded from induced E. coli JM109/pHWG315 cells following a procedure described elsewhere (31). The mutant enzymes were expressed and purified as described for the native enzyme. Briefly, the proteins obtained from crude extracts of the recombinantly produced enzyme were separated by fast-protein liquid chromatography (GE-Healthcare) through anion exchange chromatography columns by applying a linear NaCl gradient buffer. Eluted protein fractions were tested in the case of the MutB mutants for activity following the procedure described for the native protein (31). The separation of active-site mutants was followed by SDS-PAGE analysis of the proteins eluted under a gradient concentration similar to that used for the wild-type MutB eluted from the column. Pure enzyme fractions were combined, concentrated, and dialyzed for 10 mM Ca-acetate buffer, pH 6.5, by ultrafiltration (Amicon).
SDS-PAGE was done according to the methods of Laemmli (17). Protein concentrations were determined by the method of Bradford (6) using a Bio-Rad protein assay dye reagent and bovine serum albumin as a standard.
Trehalulose synthase MutB, purified from E. coli, was spotted to a polyvinylidene difluoride membrane (Bio-Rad). The N-terminal amino acid sequence was analyzed by TOPLAB (Martinsried). Matrix-assisted laser desorption (MALDI)-mass spectrometry was performed by TOPLAB (Martinsried) with lyophilized MutB.
MutA activity was tested for the substrates sucrose, isomaltulose, and trehalulose by incubation of 45 μl of enzyme solution (crude cell extracts or purified enzyme) with 5 μl of a substrate solution at a final concentration of 100 mM in 0.1 M K-phosphate buffer, pH 6.5, at 25°C. After the reaction was terminated by boiling for 5 min, the sugar composition in the reaction mixtures was analyzed by high-performance anion-exchange chromatography with electrochemical detection, as described previously (29). The standard activity assay for MutA was performed with isomaltulose as a substrate. MutA activity was calculated according to the release of glucose, which was determined by an Ecoline S+ glucose hexokinase test (DiaSys GmbH, Holzheim) and quantified according to a glucose standard curve. One unit of MutA activity was defined as the amount of enzyme that released 1 μmol of glucose in 1 min in the initial stage of the reaction.
Kinetic parameters for MutA were obtained by measuring the reaction products (glucose and trehalulose) during incubation of the purified enzyme with various concentrations of isomaltulose as a substrate. Reaction products were quantified by high-performance liquid chromatography (HPLC) and used to calculate Km and Vmax values according to Lineweaver Burk plots.
SI activity was determined as described previously (31). MutB mutants were analyzed by their ability to act on sucrose and the amount of the released products. The product composition of the reaction mixture following 100 mM sucrose conversion was measured after approximately 90% sucrose consumption, and the individual sugars were determined by HPLC as described previously (29). Effects of temperatures on MutB activity and product composition were determined after the enzyme was equilibrated in reaction buffer at the respective temperature for 15 min before the addition of the substrate sucrose. To calculate the kinetic parameters, SI activity was measured by incubating the purified enzymes with various sucrose concentrations (10 to 200 mM) under standard assay conditions. The data for the respective products that were formed during the enzyme reaction were analyzed by using double-reciprocal plots according Lineweaver Burk to calculate Km and Vmax values.
Five micrograms of total RNA from P. mesoacidophila MX-45 cells, grown under various conditions, or 0.05 μg of total E. coli RNA containing plasmids for mutA expression was separated on 1.2% formaldehyde gels and blotted by following the manual of the DIG Northern Starter Kit (Roche Applied Science).
Briefly, following transfer, the nylon membrane was UV cross-linked and hybridized at 68°C with a denatured DIG-labeled RNA probe (approximately 100 ng ml−1) per 10 ml of Easy Hyb solution (Roche Applied Science). Detection was performed using a DIG luminescence detection kit (Roche Applied Science).
DIG-labeled single-stranded RNA probes were generated according to in vitro transcription labeling technique provided in a DIG Northern Starter Kit (Roche Applied Science). A 630-bp EcoRV mutA gene fragment from pHWG653 was cloned in antisense orientation downstream of the T7 promoter in pJOE4786 (2). Five micrograms of the resulting plasmid pHWG741 was linearized with NsiI and transcribed with T7 RNA polymerase in the presence of DIG-UTP to create a 730-bp mutA probe, which was used for hybridization in Northern blotting. A 1,695-bp EcoRI mutB gene fragment from pHWG272.5, which contains the mutB gene downstream of the tac promoter, was inserted in pJOE4786 to give the plasmid pHWG717.1 with antisense orientation of the mutB gene. After linearization with BglII and transcription with T7 RNA polymerase in the presence of DIG-UTP, a 1,160-bp DIG-labeled RNA transcript was obtained and used as a mutB probe for hybridization in Northern blot analysis.
The DNA sequence of mutA and mutB, along with their genomic context, was deposited in the GenBank database under the accession number FJ792641.
For cloning of the trehalulose synthase gene from P. mesoacidophila MX-45, a probe deduced from conserved regions of known sucrose SIs was constructed. After alignment of different SI sequences, highly conserved amino acid regions were examined for low variability of the nucleotide sequence after retrotranslation; hence, conserved regions with portions of high tryptophan content were chosen (see Fig. S1 in the supplemental material). The utilization of the degenerate oligonucleotide primers S633 and S634 resulted in the amplification of DNA fragments with sizes of about 0.5 kb from the chromosomal DNA from P. mesoacidophila MX-45. The amplified fragments were cloned into the positive selection vector pJOE4786 (2), and the inserts of various transformants were sequenced. Although the inserts uniformly revealed a size of 509 bp, they surprisingly were substantially different in their sequences. Two types of sequences were identified, both of which showed homology to known SIs (see Fig. S1 in the supplemental material). The inserts in two representative clones (pHWG705.1 and pHWG705.6) exhibited 69% identity at the nucleotide level and 65% identity at their deduced amino acid sequence level. Hence, two different classes of amplicons were obtained, and these are referred to as the mutA probe in pHWG705.1 and the mutB probe in pHWG705.6, respectively.
The genomic DNA of P. mesoacidophila MX-45 was digested with different restriction enzymes and hybridized in Southern blot experiments against the DIG-labeled mutA and mutB probes (Fig. (Fig.1).1). This resulted in hybridization signals for both probes at about 10 kb (BamHI) or 6.2 kb (NotI). BglII- and HindIII-digested DNA displayed two hybridization signals with the mutB probe and confirmed the already known restriction sites of the mutB probe. The genomic DNA was cut again with NotI, and fragments with sizes of about 6.2 kb were pooled and cloned into the positive selection vector pJOE4786 (2). Cells of E. coli JM109 were transformed with the resulting recombinant plasmids. The transformants obtained were probed by colony hybridization with the mutA probe. Out of 175 colonies four clones showed positive hybridization signals. These clones were identical, as judged by restriction enzyme mapping. Therefore, only one of the strains (E. coli JM109/pHWG693.11) was further studied. The plasmid was analyzed by Southern blot analysis, and positive hybridization signals were found with the mutA and the mutB probes. E. coli JM109/pHWG693.11 was cultivated in dYT medium, and crude cell extracts converted sucrose into trehalulose, indicating the expression of a trehalulose synthase gene. The trehalulose synthase activity of the recombinant strain was 5.3 U mg−1 and was not dependent on sucrose supplementation of the medium. Its activity was approximately fivefold more than the activity measured in the P. mesoacidophila MX-45 strain (Table (Table1)1) .
The DNA insert in plasmid pHWG693.11 was sequenced by creating various subclones. Adjacent and overlapping chromosomal sequence information was obtained from further chromosomal DNA fragments clones. Altogether, a region of 8,070 bp was sequenced. The results of the DNA sequence analysis of the chromosomal fragment from P. mesoacidophila MX-45 are displayed in Fig. S2 in the supplemental material.
The sequenced chromosomal fragment contained seven putative open reading frames (ORFs). ORF 1 and ORF 2 showed 77 to 90% sequence identities to putative transposon-related proteins from Rhizobium leguminosa (IS66-like). The third ORF (1,704 bp) was identified as mutA because it encompassed the previously obtained sequence of the mutA probe. The predicted translation product of the mutA gene is a protein of 567 amino acid residues. The start codon of the following ORF (ORF 4, 1,755 bp) was identified after an intergenic region of 232 bp. ORF 4 included the above-described sequence of the mutB probe, and sequence comparison demonstrated its identity to the SI precursor from P. mesoacidophila deposited in the GenBank (DQ304536). MutB encoded a protein of 584 amino acid residues and included a proposed signal peptide of 23 amino acid residues, as predicted by the SignalP program (4). We assumed the ATG at position 4067 to be the start codon, as a predicted ribosomal binding site is located immediately upstream of the putative translation initiation codon. A stem-loop structure detected immediately downstream of the mutB gene (5884 to 5877 bp) lends support to a rho-independent termination of RNA synthesis.
The mature MutB and the MutA protein demonstrated a sequence identity of 52%. Both were homologous to members of the family 13 glycoside hydrolases which also includes SIs. The MutB protein shared 69% identity with isomaltulose synthases from Klebsiella sp. strain LX3 (50) and Erwinia rhapontici (5) and 44% with the oligo-1,6-glucosidase from Bacillus cereus (44). MutA displayed 52% sequence identity to isomaltulose synthases mentioned above and 44% to the oligo-1,6-glucosidase. Three further putative ORFs were identified downstream of mutB. The deduced proteins exhibited high sequence similarities to putative proteins from the plasmid pSymB from Sinorhizobium meliloti and encoded a putative acetyltransferase, a transcriptional regulator protein of the LysR family (69% identity), and a putative arginosuccinate lyase (83% identity). The order of the genes was identical in P. mesoacidophila MX-45 and pSymB.
Plasmid pHWS173.1 encoded the full-length mutA gene and a truncated mutB gene (see Fig. S2 in the supplemental material). Crude cell extracts of E. coli JM109/pHWS173.1 were tested for their activity on sucrose and revealed its hydrolysis into glucose and fructose. To confirm the nature of the mutA product, the protein was expressed in E. coli under the control of the rhamnose promoter. After induction of E. coli JM109/pHWG653, the MutA production in crude cell extracts was confirmed by SDS-PAGE. Protein staining revealed a prominent band exhibiting the expected size of 64.8 kDa for the MutA protein (Fig. (Fig.2A).2A). The crude cell extract released glucose hydrolytically from the substrates sucrose, trehalulose, and isomaltulose. The highest activity (0.4 U mg−1) was obtained with trehalulose as a substrate (Table (Table1).1). The activities for the substrates isomaltulose and sucrose were only 80% and 10%, respectively, of this value. Therefore, MutA was referred to as trehalulose hydrolase. Due to the limited availability and detectable impurities of the used trehalulose substrate, the standard activity assay was performed with isomaltulose as a substrate.
Purified MutA protein was used to test its substrate specificity in more detail. A carboxy-terminal His tag was added to the MutA protein in order to allow purification by IMAC. Its production was achieved from induced E. coli JM109/pHWG680 cells. Crude extracts were prepared and were used to isolate the MutA protein by IMAC. The purified MutA was enriched 26-fold to a specific activity of 45 U mg−1 (for the substrate isomaltulose) and was more than 90% pure according to SDS-PAGE (Fig. (Fig.2B).2B). Its action on the substrates sucrose, trehalulose, and isomaltulose was analyzed by subsequent sugar analysis via HPLC. Chromatograms indicated, besides the presence of glucose and fructose from the hydrolytic reaction, additional products in the reaction mixture when trehalulose and isomaltulose were used as substrates. By comparison to standards, the additional peaks could be assigned to isomaltulose, trehalulose, and isomaltose. Quantification of the product composition by sugar analysis revealed that, with trehalulose as a substrate, 35.7% glucose, 42% fructose, 8.5% isomaltulose, and 13.8% isomaltose were formed as products. Isomaltulose was converted to 30.6% glucose, 35.5% fructose, 22% trehalulose, and 11.7% isomaltose. Sucrose was hydrolyzed only to glucose and fructose. Thus, for all the tested substrates, glucose and fructose were the main products. This confirmed the hydrolytic activity of MutA on the respective disaccharides. The isomerization of the disaccharides trehalulose and isomaltulose indicated a minor transferase activity of MutA. Kinetic parameters were determined for isomaltulose, the substrate that was used in the standard MutA assay (Table (Table2).2). Its catalytic efficiency for the hydrolytic reaction (formation of glucose) was more then 10 times higher than that for the transfer reaction (formation of trehalulose).
The ability of E. coli JM109/pHWG693.11 to convert sucrose into trehalulose confirmed the presence of the coding region of a trehalulose synthase within the plasmid. To confirm the nature of the mutB product, subcloning was performed in the l-rhamnose-inducible expression vector pJOE2702 (43). Expression of mutB was carried out by rhamnose induction of E. coli JM109/pHWG315. Enzymatic activity was assayed by incubation of crude cell extract with sucrose and by subsequent product analysis by HPLC. The formation of trehalulose as a main product demonstrates the trehalulose synthase activity of the recombinant MutB protein (Table (Table3).3). The production of MutB yielded 100 U mg−1 in crude cell extracts. The expression of mutB was confirmed by SDS-PAGE analysis (Fig. (Fig.2C),2C), which demonstrated in crude cell extracts of the induced cells a prominent protein band of 55 kDa, which was absent from the extracts prepared from uninduced cells. Notably, whole cells of induced E. coli JM109/pHWG315 were able to convert sucrose into trehalulose in contrast to cells from strains recombinantly producing intracellular MutB variants (pHWG682.3 or pHWG683.8, described later). This indicated that the mature MutB protein is located in the periplasmic space.
The purification of MutB from crude cell extracts of E. coli JM109/pHWG315 has already been described (31). Briefly, after two purification steps by fast-protein liquid chromatography through anion-exchange chromatography columns, the final product was enriched ninefold and exhibited a specific activity of 900 U mg−1. Its homogeneity was confirmed by SDS-PAGE displaying a single protein band at 55 kDa (Fig. (Fig.2C,2C, lane 3). The observed molecular mass differed from the calculated molecular mass of 64,291 Da for the mature MutB. Detailed molecule properties were obtained from MALDI-time of flight mass spectrometry analysis and amino-terminal sequencing. Thus, the molecular mass of the purified MutB was 63,750 ± 100 Da. Automated Edman amino-terminal sequencing revealed KPGAP as the predominant sequence and indicated that in comparison to the predicted amino-terminal sequence 24EEAVKPGAP32 of the mature protein, four amino-terminal residues had been cleaved. The resulting protein mass is calculated as 63,863 Da and is in agreement with the results obtained by MALDI-time of flight analysis.
To confirm the nature of different processed mature MutB proteins, we attempted intracellular production of two MutB variants, one according to the predicted mature protein sequence (M)-24EEAVK, encoded in plasmid pHWG682.3 and another exhibiting the N-terminal sequence (M)-28KPGAP (pHWG683.8). After expression in E. coli JM109, trehalulose synthase activities were measured in crude cell extracts and revealed for both of the MutB variants activity of 130 U mg−1. From SDS-PAGE analysis it was estimated that both enzyme variants were expressed at the same levels (data not shown). This indicated that variation at the N-terminal region did not affect the activity or the expression level of the obtained MutB enzymes. Whole cells of the recombinantly produced MutB variants were not able to convert sucrose into trehalulose and justify the intracellular localization of these MutB variants.
The product composition of the MutB reaction was measured after 90% sucrose conversion with 100 mM sucrose, and the reaction products were analyzed by HPLC. Purified MutB converted sucrose mainly to trehalulose, and minor amounts of products were contributed by isomaltulose and the monosaccharides glucose and fructose. At a reaction temperature of 25°C, trehalulose accounted for 85%, isomaltulose for 14%, and glucose and fructose for the remaining 1% of the sucrose consumed. Changing the temperature affected the ratio of monosaccharides and isomaltulose, as has been shown for isomaltulose synthase PalI from Klebsiella sp. (50). At 20°C monosaccharides were barely detectable, and at this temperature the highest amount of trehalulose was formed (90%). Higher temperatures promoted both the monosaccharide and isomaltulose release and led to lower yields of trehalulose (Fig. (Fig.3).3). Raising the reaction temperature from 20 to 50°C caused a fourfold increase in the isomaltulose amount. At 50°C the trehalulose/isomaltulose ratio was nearly 1:1. MutB was most active at 40°C for sucrose conversion, and at above 50°C the MutB activity quickly decreased.
The purified MutB was used to determine the enzyme kinetics for the substrate sucrose. The apparent Michaelis constant (Km) and the maximum velocity Vmax were calculated for each product formed (Table (Table2).2). With respect to the main product, trehalulose, the Km value for sucrose was 30 mM, which is similar to the values obtained for isomaltulose synthase enzymes from P. rubrum and Pantoea dispersa UQ68J (47). The calculated Vmax value for the product trehalulose was 1,020 U mg−1, which is in agreement with the determined maximum specific activity of 900 U mg−1 for the purified MutB. The catalytic efficiency of MutB for trehalulose production amounted to 3.6 × 104 M−1 s−1, and this is the highest rate among the known SIs. Its value is two times higher than the value for the P. dispersa UQ68J enzyme, which is considered to be a very efficient enzyme (47). Among the SIs, MutB also exhibited the highest ratio for the catalytic efficiency of the transfer reaction versus hydrolytic reaction and is thus considered to be a very efficient alpha-glucosyltransferase. The calculation of the kinetic parameters for the by-products formed by MutB revealed similar Km values for the product isomaltulose and lower affinity for the product glucose.
Cells of P. mesoacidophila MX-45 were cultivated in minimal medium supplemented with either 0.2% sucrose or 0.2% glycerol. The total RNA was prepared from exponentially growing cells, and DIG-labeled RNA-probes of mutA and mutB were used to probe RNA blots. Northern blot analysis with the mutB probe revealed the existence of an mRNA transcript with a size of approximately 2 kb (Fig. (Fig.4).4). The level of the mRNA was higher when cultures were grown on sucrose-supplemented medium. This was in agreement with the increase of trehalulose synthase activity found in P. mesoacidophila MX-45 when the medium was supplemented with sucrose.
In contrast, no mutA transcript was detected under identical conditions when a DIG-labeled mutA RNA probe was used. In order to validate the mutA probe used, control experiments were performed with RNA isolated from E. coli strains harboring the expression plasmid for the mutA gene (pHWG653). A transcript of 1,800 bp was detected in rhamnose induced cells, whereas no message was detectable with uninduced cells (Fig. (Fig.44).
To test the potential catalytic triad in MutB, sequence alignments of isomaltulose synthases, oligo-1,6-glucosidases, and amylosucrases were done; the results suggested Glu 281 as a candidate for the general acid catalyst, Asp 227 as the attacking nucleophile, and Asp 354 as a candidate for the formation of hydrogen bonds to O2 and O3. The role of these potential key amino acids was examined by creating mutants and analyzing their activity. The MutB variants D227A, E281Q, and D354N were obtained by site-directed mutagenesis of the mutB gene. The expression of the gene variants with E. coli JM109 cells harboring the plasmids pHWG678.6 for D227A, pHWG679.9 for E281Q, and pHWG684.2 for D354N was confirmed by SDS-PAGE analysis (data not shown). Their crude cell extracts were tested for their activity on sucrose and were found to be inactive.
In order to examine the structural features that are involved in the hydrolytic reaction of SIs, we attempted a mutagenic approach to create hydrolytic variants of MutB. The idea for screening such mutants was based on the formation of detectable amounts of monosaccharides. Glucose and fructose can be metabolized by the host strain, and such cells should display a red phenotype on McConkey agar plates supplemented with sucrose. MutB enzyme formed only 2% of the monosaccharides during conversion of sucrose at 37°C (Fig. (Fig.3),3), and E. coli JM109/pHWG315 displayed a white phenotype on these indicator plates.
To obtain mutants, mutB was subjected to random mutagenesis using treatment with MNNG. The mutant library was plated on McConkey agar plates supplemented with 1% sucrose, 0.1% rhamnose, and ampicillin and screened for red colonies. The highest number of mutants was obtained with a dose of 0.2 mg ml−1 MNNG. Out of about 5,000 colonies, 7 colonies were found to display a red phenotype on McConkey agar plates. These colonies were isolated and purified for single colonies on the indicator plates.
To characterize the mutants, colonies exhibiting a red phenotype on McConkey sucrose agar plates were cultivated and induced for protein production. A standard trehalulose synthase activity assay was performed with crude cell extracts, and the product profiles were analyzed by HPLC and are summarized in Table Table3.3. As expected, all the mutants exhibited an increased amount of monosaccharides in comparison to the wild-type MutB enzyme. This validated the screening idea. The monosaccharide composition ranged from 7 to 80% of the sucrose consumed and displayed a great variety for the respective mutants. The mutant 659.3 mainly formed monosaccharides (80%) and was referred to as a hydrolytic variant. The mutants 660.2, 660.7, and 660.17 still favored the transfer reaction by producing 80 to 90% of sucrose isomers. Interestingly, for the mutants 660.2, 660.3, 660.17, and 660.18, the product specificity was shifted from trehalulose to isomaltulose. These mutants exhibited the characteristics of an isomaltulose synthase. Mutants exhibiting different properties were chosen to determine their respective nucleic acid sequences. Single or double amino acid replacements were detected by their deduced peptide sequences (Table (Table3).3). The double mutant 660.18 (MutB P29S A230V) displayed the same characteristics as the single mutant 660.3 (MutB A230V). We assume that the exchange A230V is responsible for the properties of the strain 660.18.
To characterize the hydrolytic mutant 659.3 (MutB R311C) and the isomaltulose synthase-like mutant 660.2 (MutB D442N) in more detail, the proteins were produced and purified according to the procedure described for MutB. The mutations did not appear to affect the expression level, as visualized by SDS-PAGE (data not shown), but resulted in changes in enzyme kinetics (Table (Table2).2). In comparison to the wild-type MutB, both the mutants displayed reduced trehalulose synthase activity due to decreased Kcat values. We determined the catalytic efficiency, Kcat/Km, for each product formed to characterize the enzyme properties regarding the hydrolytic (formation of glucose) or transfer (formation of sucrose isomers) reaction. For comparison, the kinetic data for the recently described isomaltulose synthase SmuA from P. rubrum (32) were determined in similar experiments. The ratio of the catalytic efficiency for the transfer reaction versus the hydrolytic reaction (Kcat/Km transfer/hydrolysis) was calculated, and values above 1 were assigned to a transferase while values below 1 were assigned to a hydrolytic enzyme. Further, the ratio of the catalytic efficiency for the production of sucrose isomers (Kcat/Km trehalulose/isomaltulose) was used to examine the product specificity of SIs.
The catalytic efficiency for trehalulose or isomaltulose production of MutB R311C was significantly lower than that for the wild-type MutB; however, its value for glucose production increased 12 times. Its value below 1 for Kcat/Km transfer/hydrolysis clearly demonstrated its hydrolytic character. For MutB D442N the Kcat value for isomaltulose production was eight times higher than that of the wild type enzyme; however, due to a drastic increase in Km, the catalytic efficiency was two times lower. MutB D442N was assigned to a transferase due to its value for Kcat/Km transfer/hydrolysis, with a preference for the product isomaltulose, like the SmuA enzyme.
A DNA probe for SI yielded two distinct amplicons from P. mesoacidophila, both sharing sequence homology to SIs. Their use for Southern blot experiments facilitated the cloning of mutA, a trehalulose hydrolase gene, and mutB, the SI gene. Because the SIs share substantially conserved regions with family 13 glucosidases, the primers used for PCR yielded not only SI-specific sequences but also oligo-1,6-glucosidase-homologous sequences. In order to obtain a more SI-specific probe, sequence elements of motifs conserved among SIs but not among glucosidases, as proposed by Wu et al. (47), may appear to be more diagnostic for SIs. Actually, the sequences of the obtained SI and the hydrolytic enzyme from P. mesoacidophila MX-45 are not conserved within these regions.
Despite the genetic localization of mutA and mutB, it is apparent that both genes are not transcribed as bicistronic mRNA. The length of the detected mRNA suggests that only mutB is encoded by this transcript. The mRNA level for mutB was upregulated when cells were grown on medium supplemented with sucrose. Since maximum trehalulose synthase activity was detected with P. mesoacidophila MX-45 cells grown on sucrose medium, it could be possible that sucrose or reaction products thereof may be involved in transcriptional regulation. This supports the notion that under conditions of excess sucrose availability, trehalulose is formed as a reserve material, as reported for SIs from E. rhapontici (5). According to this study isomaltulose is produced from sucrose by SI in the presence of sucrose, whereas the metabolizing enzymes are expressed only in the presence of isomaltulose and not of sucrose. On the other hand, P. mesoacidophila MX-45 is not able to grow on minimal medium supplemented with trehalulose or isomaltulose (data not shown), and under the experimental conditions used for Northern blot analysis, no mRNA transcript of mutA was detected. In E. rhapontici the isomaltulose pathway is organized in an operon consisting of gene products involved in uptake and metabolism, and isomaltulose is cleaved hydrolytically by PalQ (5). MutA exhibits 45% sequence identity to PalQ, but homologous gene products responsible for uptake were not found in the genomic context of mutA. In addition, gene products for a metabolic route via a phosphotransferase system, as described for various organisms capable of metabolizing sucrose isomers (39, 40), were not found in the vicinity of mutA. Further, the role of disaccharides (e.g., isomaltulose) as osmoprotectants was reported for some bacteria (10). In the case of the accumulation of trehalulose in P. mesoacidophila in response to osmotic stress, hydrolysis may not need to be rapid. To date, no functional proof of mutA transcription in P. mesoacidophila can be given, and it remains to be elucidated whether the transcription of mutA undergoes further regulation, e.g., osmoregulation, and/or whether mutA is expressed in the presence of different carbohydrates.
Functional expression of mutA in E. coli showed significant activity for trehalulose hydrolysis. To our knowledge MutA is the first enzyme described that is able to hydrolyze trehalulose. This enzyme might be useful for some industrial processes, e.g., treatment of sticky cotton fiber. The “stickiness” of cotton is associated with trehalulose-rich honeydew, which is produced by the silverleaf whitefly Bemisia argentifolii (33), and causes serious problems in spinning mills (13, 14). Certain carbohydrate-degrading enzymes like glucose oxidase (12) and a transglucosidase (18) have been used to hydrolyze the carbohydrate components of honeydew to reduce the stickiness of cotton. With the MutA enzyme a specific hydrolysis of trehalulose is conceivable.
MutB contains a putative N-terminal signal peptide, making it most probably located in the cells' periplasmic space. It was expected that the overexpression of mutB would require tight control of gene expression. This was achieved by the application of the l-rhamnose-inducible expression vector in E. coli. The expression of mutB was strictly dependent on induction with rhamnose, and after induction of E. coli JM109/pHWG315 cultures, about 10% of soluble MutB was produced. Notably, no reproducible production of MutB was obtained during expression of mutB with a tac vector derivative (data not shown). Wu et al. (47) used the T7 expression system (pET24b) for intracellular production of various SIs in E. coli. Expression of their respective genes without signal sequences yielded for some SIs up to 10% of the total E. coli protein; however, the authors noted substantial amounts of insoluble protein, and the P. rubrum SI was totally inactive due to misfolding. Our attempt to express the mature mutB gene without signal sequence yielded MutB as approximately 15% of the total cell protein without any detectable insoluble MutB protein (data not shown).
The MutB protein was finally purified up to 10-fold from the recombinant E. coli. Purification of PalI from Klebsiella sp. strain LX3 required 350-fold enrichment after production of a glutathione S-transferase-PalI fusion protein in E. coli (50). This reflects the high efficiency of the applied l-rhamnose-inducible expression system. A trehalulose synthase enzyme had already been isolated from P. mesoacidophila MX-45 (27). We noted striking differences regarding the specific activity of the purified enzymes. The recombinantly produced MutB exhibited a specific activity of 900 U mg−1, whereas the purified enzyme, as described by Nagai et al. (27), exhibited a specific activity of only 14 U mg−1. In order to verify the enzyme's nature, we examined product composition during sucrose conversion at various temperatures. MutB formed smaller amounts of trehalulose at higher temperatures, and the ratio of trehalulose to isomaltulose amounted to 1 at 50°C, like the isolated enzyme from P. mesoacidophila MX-45; hence, the enzymes are similar.
MutB is the only characterized member of the SIs which produces mainly trehalulose with only trace amounts of monosaccharides, originating from the hydrolytic reaction (1.2% at 37°C). The other known SIs convert sucrose mainly to isomaltulose and generate higher amounts of monosaccharides (10 to 25% at 37°C) (48, 50). In order to study specific structural requirements for the hydrolytic reaction of SIs, we performed a mutagenic approach to convert MutB to a “hydrolytic” enzyme. Among the MutB mutants obtained, the variant 659.3 (MutB R311C) preferentially performed the hydrolytic reaction, which is shown by the low catalytic efficiency (Kcat/Km) for the transfer versus the hydrolytic reaction. Interestingly, some of the MutB variants showed a significant shift in their main product from trehalulose to isomaltulose and formed medium amounts (10 to 30%) of monosaccharides. These mutants can be considered as isomaltulose synthases. Among the known SIs UQ68J from P. dispersa (47) showed the highest isomaltulose-producing efficiency. Notably, the Kcat value of the isomaltulose synthase MutB D442N is higher than that of UQ68J. But due to the drastic increase in Km, the catalytic efficiency for isomaltulose production is lower than that of UQ68J. Industrial processes for isomaltulose production operate with sucrose concentrations up to 1.2 M. Under these conditions the unfavorable Km value of MutB D442N is negligible but may lead to an incomplete conversion of sucrose.
The observed differences in reaction kinetics and product specificities of the mutant enzymes most probably reflect modifications in the active-site architecture of SIs. For PalI it was reported that a unique 325RLDRD329 sequence, located in a loop region, might control the product specificity of SIs (49). MutB possesses a different corresponding sequence (311RYDRA315). Surprisingly, Aroonnual et al. (3) reported that the replacement of RLDRD by RYDRA in PalI from K. pneumoniae NK33 had no effect on the ratio of isomaltulose to trehalulose formation. Recently, the importance of the Arg residues of this motif has been demonstrated in an SI from P. rubrum, and mutations at these sites led to enzymes with altered product compositions but with decreased sucrose conversion rates and some synthesis potential for isomaltose (19). Interestingly, the hydrolytic mutant MutB R311C has its amino acid alteration at one of these conserved Arg residues within the so-called “isomerization” motif. However, the other MutB mutants demonstrated that another region or regions contribute to the product specificity of MutB, too. Further insights about the active-site structure of MutB were obtained from the crystal structure of MutB (29). We analyzed the localization of their mutations in respect to residues defining the active-site of MutB (see Fig. S4 in the supplemental material). It is striking that the MutB variants exhibited mutations in the close neighborhood to the active-site residues. For instance, Ala 230, which is altered in the mutants 660.3 (A230V) and 660.18 (P29S A230V), is located in the immediate vicinity of the catalytic triad residue Asp 227. Val 285 of the mutant 660.7 is proximate to the catalytic triad residue Glu 281 and to Phe 283, which is part of the aromatic clamp at the entrance of the catalytic pocket and is supposed to play an essential role in controlling the reaction specificity of MutB (29). In addition, Val 285 is notable. All the known isomaltulose-specific SIs (24, 47, 50) possess a conserved Val at the corresponding site. Since most of the SIs are similar in structure and function, amino acid residues that are not conserved in this class of enzymes could be considered as candidates responsible for substrate specificity. In the mutant 660.12, Phe 191 is replaced by Leu. Phe 191 mediates the binding of the glucose moiety of the substrate to the enzyme by a stacking mechanism (29). The perturbed interaction by the mutated residue seems to have altered the product species formed. Asn 442 of the mutant 660.2 (D442N) and Ile 439 of the double mutant 660.17 (T439I P80S) are proximate to Arg 441, which is part of both substrate binding subsites and is suggested to play an important role in substrate binding and positioning (29). The second mutation of the double mutant 660.17, Ser 80, is proximate to Asp 88, which is involved in binding of the glucose moiety of the substrate. These results suggest that mutations in the near vicinity of residues involved in substrate binding may disturb the interaction network. In order to investigate whether one or both of the altered residues of 660.17 is responsible for the enzyme property, the creation of single mutants and analysis of their properties are necessary.
All the mutations of the MutB variants obtained are localized near to or in the substrate pocket. They contain alterations of residues that are in the proximity of residues defining the active site and, therefore, have a high probability of changing substrate preference or product specificity. Thus, product specificity and catalytic reactions seem to be guided by a set of very subtle features rather than by a single one. The MutB mutants presented in this paper can serve as templates for elucidation of their crystal structures in order to understand the specific structural features required for product specificity. Crystallization trials with MutB R311C and MutB D442N are in progress. The construction of the general acid/base knockout mutants and the elucidation of their structures in complex with sucrose, trehalulose, or isomaltulose in comparison to the already resolved structure of the MutB E281Q-sucrose complex might allow the identification of structural features involved in the formation of different product species by SIs.
Published ahead of print on 25 September 2009.
†Supplemental material for this article may be found at http://aem.asm.org/.