![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Previous Article | Next Article 
Applied and Environmental Microbiology, January 2006, p. 368-377, Vol. 72, No. 1
0099-2240/06/$08.00+0 doi:10.1128/AEM.72.1.368-377.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Division of Microbiology and Epidemiology, Department of Basic Veterinary Sciences, Faculty of Veterinary Medicine, P.O. Box 66, FIN-00014 University of Helsinki, Finland,1 Laboratory of Bioprocess Engineering, Department of Chemical Technology, Helsinki University of Technology, P.O. Box 6100, FIN-02015 Espoo, Finland2
Received 22 July 2005/ Accepted 3 October 2005
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Some bacteria have been reported to metabolize xylitol by pathways in which xylitol is oxidized to the unphosphorylated form of L- or D-xylulose. Growth in xylitol-containing medium has been shown to induce an NAD+-dependent xylitol-2-dehydrogenase and a D-xylulokinase in some strains of enteric bacteria. In these strains, xylitol is catabolized by the oxidation of xylitol to D-xylulose, which is then phosphorylated to D-xylulose-5-phosphate (6). The bacterial plant pathogen Pantoea ananatis (formerly named Erwinia uredovora) can acquire the ability to utilize xylitol as a growth substrate by a spontaneous mutation. The mutant synthesizes two enzymes involved in xylitol catabolism, namely, xylitol-4-dehydrogenase (XDH) and L-xylulokinase. The xylitol-4-dehydrogenase, which catalyzes the oxidation of xylitol to L-xylulose, apparently has no significant activity with other naturally occurring simple pentitols (ribitol or D-arabitol). An L-xylulokinase-deficient Pantoea ananatis strain can be obtained by transposon mutagenesis of the spontaneous mutant and used for L-xylulose production from xylitol in nongrowing cells (7, 8). The production of L-xylulose from xylitol using nongrowing cells has also been demonstrated for a natural isolate, Alcaligenes sp. strain 701B (14).
Fungi have been reported to catabolize L-arabinose via L-xylulose (4, 26, 35). The pathway includes a reaction in the direction opposite to that for the oxidation of xylitol. In fungi, the reduction of L-xylulose to xylitol is catalyzed by an L-xylulose reductase (a synonym for xylitol-4-dehydrogenase). The gene for this enzyme has been identified in Hypocrea jecorina and expressed in Saccharomyces cerevisiae (26). In addition, an L-xylulose reductase has been purified to homogeneity from Aspergillus niger and then characterized (35). Both fungal enzymes have a strict requirement for NADPH. The L-arabinose pathway has also been reported recently for the yeast Ambrosiozyma monospora, with the exception that the L-xylulose reductase of this strain is strictly NADH dependent (33).
Enzymes having L-xylulose reduction activity have also been identified in mammalian tissues such as hamster and guinea pig livers and kidneys. In mammals, the L-xylulose reductase is a part of the uronate cycle of glucose metabolism, in which the enzyme catalyzes the conversion of L-xylulose into xylitol, with NADPH as the cofactor. The L-xylulose reductase has been reported to be identical to diacetyl reductase, and it has therefore been suggested that this enzyme also has a role in detoxification of reactive dicarbonyl compounds in mammals (2, 11, 20).
To our knowledge, no bacterial xylitol-4-dehydrogenases have been purified to homogeneity, and no corresponding gene sequences are publicly available. In this work, we report for the first time the gene cloning and sequencing of a bacterial xylitol-4-dehydrogenase. In addition, the purification and characterization of some of the enzymatic properties of this novel enzyme are presented.
Biotechnological methods have been shown to be well-suited for production of rare and unnatural sugars (1). The markets and number of applications for these sugars have been increasing over the past few years (1, 30). In the present work, we show that E. coli expressing the xylitol-4-dehydrogenase gene can be used for production of L-xylulose from xylitol with a high yield.
L-Xylulose, which is currently very expensive, can be used as a raw material for the production of other rare sugars. It has been shown previously that L-xylulose can be easily isomerized to L-xylose either enzymatically (12) or under alkaline conditions (5). L-Xylose is a rare sugar with a high market value that is used as a raw material for the production of antiviral drugs (19). In addition, the production of another rare sugar, L-lyxose, by enzymatic isomerization of L-xylulose has been described (3).
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DNA techniques.
Basic molecular biological techniques were used as described by Sambrook and Russell (28). Chromosomal DNAs from P. ananatis ATCC 19321, P. ananatis ATCC 43072, and P. ananatis ATCC 43073 were isolated with a Nucleobond AX kit (Macherey-Nagel GmbH & Co.). Plasmids were isolated with the Wizard Plus DNA purification system (Promega).
Cloning and sequencing of xylitol dehydrogenase gene, xdh.
The Vectorette II system (Sigma Genosys Ltd., Cambridge, United Kingdom) was used to construct genomic libraries of P. ananatis ATCC 43072 and Pantoea ananatis ATCC 43073. Bacterial chromosomal DNAs were separately digested with BamHI, ClaI, EcoRI, or HindIII. The resulting fragments were ligated with the respective Vectorette units and amplified by PCR. The PCR products were cloned into pCR2.1-TOPO vector, using E. coli TOP10 as the host, and the resulting clones were sequenced by the dideoxy chain termination method of Sanger et al. (29) by using an ABI PRISM 310 genetic analyzer (Applied Biosystems). Homology searches of the sequences in the databases were carried out using the WU-BLAST2 program (EMBL-EBI [http://www.ebi.ac.uk/blast2/]). The BPROM program (SoftBerry Inc.) was used to localize putative promoter regions in the sequences. The protein sequences were aligned using Clustal W 1.82 (EMBL-EBI [http://www.ebi.ac.uk/clustalw/]), and to predict the protein secondary structure, the PHD program (PBIL [http://npsa-pbil.ibcp.fr/]) was used. For cloning of the xylitol dehydrogenase, the xdh gene with its translation stop codon (TAA) was amplified by PCR from P. ananatis ATCC 43072 genomic DNA, using oligonucleotides 1502 and 1503 (Table 1) as primers. The amplified 816-bp xdh fragment was cloned into the expression vector pKK223-3 as an EcoRI/HindIII fragment, and E. coli M15 was transformed with the resulting plasmid.
|
-33P]dATP using the Megaprime DNA labeling system (Amersham Biosciences). Quantification of RNA on the membrane was performed by hybridization of a 16S rRNA-specific [-33P]dATP-labeled PCR probe (primers 1360 and 1361) (34) (Table 1). The hybridization signal was detected using a Molecular Imager GS-525 system (Bio-Rad). For reverse transcription-PCR (RT-PCR), RNA samples were treated with DNase (Promega), purified with RNeasy columns (QIAGEN), and reverse transcribed using an iScript cDNA synthesis kit (Bio-Rad) according to the instructions of the manufacturers. The amounts of RT-PCR products were estimated visually after image enhancing using a Bio-Rad integration control unit.
Enzyme activity assays.
Xylitol-4-dehydrogenase activity was assayed at 30°C by measuring the change in absorbance at 340 nm. The reaction mixture typically contained 50 mM Tris-50 mM glycine-NaOH (pH 9.0), 1 mM NAD+, 500 mM xylitol, and 10 mM MgCl2. The enzyme activities are calculated as nanokatals (nmoles of NADH generated per second). Protein concentrations were determined using the Bio-Rad protein assay reagent, with bovine serum albumin as the protein standard.
The pH optimum of the purified enzyme was determined using the following three buffer systems: 50 mM potassium phosphate (pH 6.1 to 7.5), 50 mM Tris-HCl (pH 6.9 to 9.1), and 50 mM Tris-50 mM glycine-NaOH (pH 8.7 to 10.2). The pH values were measured at 30°C in solutions similar to the final reaction mixtures, with the exception that the enzyme and NAD+ were omitted.
The effects of divalent metal cations on the xylitol-4-dehydrogenase activity were determined as follows. MgCl2 was removed from the purified protein sample by diluting 50 µl of sample with 450 µl of 20 mM Tris-HCl, pH 9.0, containing 11.1 mM EDTA. The solution was incubated for 10 min at 30°C and concentrated by ultracentrifugation (Ultrafree MC 50,000 NMWL [Millipore]) to <50 µl. The sample was diluted to approximately 500 µl with 100 mM Tris-HCl, pH 9.0, concentrated again, and diluted to 1.5 ml with 100 mM Tris-HCl, pH 9.0. The samples were incubated for 15 min in the presence of 10 mM Mg2+, Ca2+, Cu2+, or Ni2+ at 30°C in 50 mM Tris-HCl, pH 9.0. Enzymatic activities of the enzyme samples were determined by adding an equal volume of the substrate solution. The final concentrations in the reaction mixtures used in the activity assays were as described above, with the exception that 50 mM Tris-HCl, pH 9.0, was used as the buffer.
The effect of p-chloromercuribenzoic acid on enzymatic activity was tested by incubating the enzyme for 15 min at 30°C with 1 mM p-chloromercuribenzoic acid in the presence and absence of 5 mM ß-mercaptoethanol or 5 mM dithiothreitol. After incubation, enzymatic assays were performed by adding an equal volume of the substrate solution at the final concentrations described above for the activity assay.
The substrate specificity was investigated under the same conditions as those described above. The polyol concentration in the reaction mixtures was 500 mM. D- and L-xylulose were used as the substrates at a concentration of 50 mM, with 0.2 mM NADH as the cofactor.
The apparent Km and Vmax values were determined for xylitol and NAD+, with the other substrate present in excess (94 mM xylitol and 1.8 mM NAD+). Initial velocity data were analyzed by nonlinear least-square regression using the program DYNAFIT (15).
Purification of xylitol dehydrogenase from Pantoea ananatis ATCC 43072.
Cells were grown to mid-exponential phase and separated from 5 liters of culture by centrifugation at 4,500 x g for 10 min. All buffers used in the purification procedure were supplemented with 10 mM MgCl2, and all ammonium sulfate solutions were buffered with 50 mM Tris-HCl, pH 7.5.
The cells were washed once in extraction buffer containing 50 mM Tris-HCl, pH 7.5, 1 mM phenylmethylsulfonyl fluoride, and 0.2 mM ß-mercaptoethanol. Washed cells were suspended in 45 ml of extraction buffer and disrupted by sonication as described previously (22). The cell debris was removed by centrifugation at 31,000 x g for 30 min at 4°C. Ammonium sulfate was added to the supernatant to achieve 20% saturation. The suspension was incubated for 1 h on ice and centrifuged at 31,000 x g for 20 min at 2°C. The ammonium sulfate saturation of the supernatant was adjusted to 60%, and the suspension was incubated overnight on ice. The suspension was centrifuged at 20,000 x g for 20 min at 4°C, and the precipitate was dissolved into 20% saturated ammonium sulfate.
In order to maximize the purity of the enzyme, only fractions with the highest specific activities in the chromatographic purification steps were selected for further purification. The enzyme solution from ammonium sulfate fractionation was applied to a phenyl Sepharose (highly substituted) (Amersham Biosciences) column (1.6 by 12 cm), and the column was washed with 20% saturated ammonium sulfate. The column was eluted with a linear gradient of 25 to 12.5% saturated ammonium sulfate (100 ml), washed with 12.5% saturated ammonium sulfate (100 ml), and eluted with a linear gradient of 12.5 to 0% saturated ammonium sulfate (100 ml). The selected active fractions were pooled, and ammonium sulfate was added at 75% saturation to the pool. The suspension was incubated for 1 h on ice, and the precipitate was separated by centrifugation as described above. The precipitate was dissolved in gel filtration buffer containing 25 mM Tris-HCl, pH 8.0, and 150 mM NaCl, and the solution was applied to a Sephacryl S-300 (Amersham Biosciences) gel filtration column (2.6 by 85 cm) and eluted. The sample buffer of the pooled active fractions was changed by concentrating the protein solution by ultrafiltration (PM10; Millipore), followed by dilution with 10 mM potassium phosphate buffer, pH 7.0. The sample was applied to a hydroxyapatite Bio-Gel HTP (Bio-Rad) column (1.6 by 9 cm) and eluted with a linear gradient of 10 mM to 400 mM potassium phosphate buffer, pH 7.0 (200 ml). The selected active fractions were pooled, and the sample buffer was changed to 10 mM Tris-HCl, pH 8.0, by concentration and dilution as described above. The pool of selected active fractions was applied to a DEAE-Memsep 1000 HP (Millipore) column and eluted with a linear gradient of 0 to 0.6 M NaCl (200 ml). The selected active fractions were concentrated by ultrafiltration (Ultrafree MC 10,000 NMWL [Millipore]). The concentrated enzyme sample was subjected to native polyacrylamide gel electrophoresis in a 7.5% gel (Bio-Rad). The gel was stained for xylitol-4-dehydrogenase activity using a previously described Nitro Blue Tetrazolium-phenazine methosulfate system (21). The band showing xylitol-4-dehydrogenase activity was cut from the gel and incubated for 30 min at 37°C in Laemmli sample buffer. The sample buffer and the band were loaded into a 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel and electrophoresed (16). The proteins were electroblotted from the SDS-PAGE gel onto a Sequi-Blot polyvinylidene difluoride membrane (Bio-Rad) (18). The N-terminal amino acid sequence of the protein sample was determined at the Protein Chemistry Unit of the Biomedicum Center (Helsinki, Finland).
Purification of xylitol-4-dehydrogenase produced in Escherichia coli.
Seven hundred fifty milliliters of medium was inoculated with 250 ml of overnight culture of E. coli M15 derivative ERF2157 cells. The culture was grown for 30 min, and 0.25 mM IPTG (isopropyl-ß-D-thiogalactopyranoside) was added to induce the synthesis of xylitol-4-dehydrogenase. The culture was grown for another 2.5 h, and the cells were harvested by centrifugation as described above. All buffers used in the purification procedure were supplemented with 15 mM MgCl2, and all ammonium sulfate solutions were buffered with 50 mM Tris-HCl, pH 7.5.
The cells were washed with extraction buffer containing 50 mM Tris-HCl, pH 8.0, and 0.2 mM ß-mercaptoethanol. The washed cells were suspended in 30 ml of extraction buffer supplemented with one Complete mini EDTA-free (Roche) protease inhibitor tablet. The cells were disrupted by sonication. The cell debris was removed by centrifugation at 31,000 x g for 20 min at 4°C, and the resulting cell extract was subjected to ammonium sulfate fractionation as described above for the native enzyme.
In order to achieve maximum purity in the chromatographic purification steps, the active fractions were analyzed by SDS-PAGE, and only the purest fractions were combined and purified further.
The supernatant from ammonium sulfate fractionation was applied to a butyl Sepharose 4 FF (Amersham Biosciences) column (1.6 by 15 cm). The column was washed with 20% saturated ammonium sulfate and eluted with a linear gradient of ammonium sulfate from 20% saturation to 0% (250 ml). The fractions were analyzed for xylitol-4-dehydrogenase activity, and the selected active fractions were combined. The pool of active fractions was concentrated by ultrafiltration (PM10 [Millipore]; Centiplus YM-30 [Amicon]). The concentrated sample was applied to a HiLoad 26/60 Superdex 200 preparation-grade (Amersham Biosciences) gel filtration column (2.6 by 32 cm) and eluted with 50 mM Tris-HCl, pH 8.0, containing 150 mM NaCl. The selected active fractions were combined, and the buffer was changed by ultrafiltration and dilution with 10 mM Tris-HCl, pH 8.0, as described above. The sample was applied to a Biosepra Q Ceramic HyperD 20 (Ciphergen, Cergy-Saint-Christophe, France) ion-exchange column (1 by 7 cm) and eluted with a linear gradient of 0 to 0.7 M NaCl in 10 mM Tris-HCl, pH 8.0 (200 ml). The active fractions were pooled, concentrated by ultrafiltration, and stored at –20°C in the presence of 50% (wt/vol) glycerol.
Production of L-xylulose from xylitol by recombinant E. coli.
The recombinant E. coli strain ERF2157 was cultivated and induced, and the cells were harvested as described above. The cells were washed and suspended in conversion buffer containing 20 mM potassium phosphate buffer, pH 7.0, 5 mM MgCl2, 100 µg/ml ampicillin, and 25 µg/ml kanamycin. Xylitol was added to the cell suspensions at a final concentration of 5 g/liter or 10 g/liter. E. coli carrying the expression plasmid pKK223-3 without the gene for xylulose-4-dehydrogenase was used as the control strain. The cell suspensions were dispensed in a volume of 50 ml into 250-ml Erlenmeyer flasks and incubated at 37°C at 200 rpm. The cell densities of the control strain and the production strain were 5.5 g/liter and 5.2 g/liter (cell dry weight per volume), respectively. The samples taken from the cell suspensions were centrifuged at 16,000 x g, and the supernatants were filtered through 0.2-µm membranes. Xylitol and D/L-xylulose were analyzed in the supernatants by high-performance liquid chromatography (HPLC) as described below.
Purification of xylulose for polarimetric analysis.
The recombinant E. coli strain ERF2157 was grown, induced, and harvested as described above. The cells were suspended at a cell density of 5.1 g/liter (cell dry weight per volume) in 1 liter of conversion buffer (see above) supplemented with 10 g/liter xylitol. The cell suspensions (2 x 500 ml) were incubated for 22 h in 2-liter Erlenmeyer flasks at 37°C at 200 rpm. The cell suspensions were centrifuged for 20 min at 8,500 x g, and the supernatants were collected and filtered through a layer of activated carbon followed by filtration through a 0.2-µm membrane. The filtered solution was evaporated to 50 ml, and 100 ml of ethanol was added. The resulting solution was centrifuged for 20 min at 23,000 x g. The supernatant was concentrated by evaporation to approximately 50 ml, and ionic compounds were removed from it by anion exchange (Amberlite IRA-958) followed by cation exchange (Finex CS 16 G PE [Finex, Kotka, Finland]). The resulting solution was concentrated by evaporation to 10 ml and subjected to hydroxyl affinity chromatography. The concentrated sample was applied to a column (2.6 by 78 cm) of Dowex 50WX4-400 (200- to 400-mesh) in the Ca2+ form. The sample was eluted with deionized water, and xylulose was analyzed in the collected fractions by HPLC as described below. The fractions containing xylulose were combined, concentrated by evaporation, and subjected to polarimetric analysis (Perkin-Elmer polarimeter 343) at a concentration of 10 g/liter at 589 nm and 20°C.
Other methods.
The molecular mass of recombinant xylitol-4-dehydrogenase was estimated by analytical gel filtration using a HiLoad 26/60 Superdex 200 preparation-grade (Amersham Biosciences) column (2.6 by 32 cm) according to the manufacturer's instructions. Xylitol and D/L-xylulose (D- and L-forms of xylulose cannot be distinguished by HPLC) were analyzed in the samples at 70°C using an HPLC system equipped with a deashing Micro-Guard precolumn (Bio-Rad), a Bio-Rad HPX-87P Aminex column, and a Waters 410 RI refractive index detector. Deionized water was used as the eluent at a flow rate of 0.6 ml/min. The chemicals used as standards were purchased from Sigma-Aldrich and were of analytical grade.
Nucleotide sequence accession number.
The nucleotide sequence of the Pantoea ananatis ATCC 43072 xdh gene has been deposited in GenBank under accession number AY894680.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Sequence analysis of P. ananatis ATCC 43072 xdh gene.
Sequence analysis of the 2.1-kb gene region isolated revealed the putative P. ananatis xdh gene as an open reading frame (ORF1) of 795 bp with an encoding capacity for a protein of 264 amino acids. The amino acid sequence determined from the N terminus of the intact 30-kDa protein was found to be the predicted amino acid sequence of ORF1. In addition, three sequences, NTLSPTVVLTP, YGVDTTTNLSGK, and AVNLTGPFL, derived from electrospray ionization-tandem mass spectrometry analysis (Institute of Biotechnology, Helsinki, Finland) of XDH fragments, corresponded to the amino acid sequence deduced from ORF1 (data not shown).
The 5'-end region of another open reading frame (ORF2) encoding an unknown protein was found 54 nucleotides downstream of the putative xdh gene (ORF1). Two putative promoters with –35 and –10 regions (TTGTTC-15 bp-AAATAAAAT and TTGCTT-18 bp-TCTTTTAAT) could be identified 18 and 302 bp upstream of the ATG start codon of xdh. A conserved region (AGGACA) for a putative ribosome binding site (31) was localized 6 bp upstream of the start codon of xdh (data not shown). No transcription terminator sequence could be recognized within the available sequence downstream of the translation stop codon of the xdh gene (Fig. 1).
|
|
Analysis of P. ananatis xdh transcripts.
Northern blot analysis of total RNA isolated from the xylitol-positive mutant P. ananatis ATCC 43072 revealed repeatedly in independent experiments the same cluster of transcripts of approximately 1, 2.2, and 3 kb when using an xdh-specific probe (Fig. 3), suggesting that xdh is transcribed as part of a polycistronic mRNA. These transcripts were found only when P. ananatis ATCC 43072 was grown with xylitol as the sole carbon source (Fig. 3, lane A). If the cells were grown either with glucose or with a mixture of glucose and xylitol as the carbon source, no xdh-specific transcripts were obtained (Fig. 3, lanes B and C). Northern blot analysis of total RNA isolated from the wild-type strain P. ananatis ATCC 19321 grown on glucose revealed no xdh transcripts (Fig. 3, lane D). Quantification of total RNAs from these strains on the hybridization membrane using a 16S rRNA-specific probe did not show differences in the quantities of the RNA samples (data not shown).
|
|
|
The concentration of Mg2+ needed for maximal reactivation was studied by incubating the enzyme in the presence of different concentrations of Mg2+. The results indicated that 10 mM Mg2+ was sufficient for maximal reactivation (data not shown).
Effect of p-chloromercuribenzoic acid on enzymatic activity.
p-Chloromercuribenzoic acid (1 mM) inhibited >95% of the xylitol-4-dehydrogenase activity. The inhibition was completely counteracted by the addition of ß-mercaptoethanol (5 mM). The addition of 5 mM dithiothreitol restored 50% of the activity in the presence of p-chloromercuribenzoate. However, when incubated in the presence of dithiothreitol alone, the enzymatic activity was only 40% of the activity without additions. The results suggest that SH groups play an important role in the enzyme reaction catalyzed by xylitol-4-dehydrogenase. According to the amino acid sequence, xylitol-4-dehydrogenase has eight cysteine residues.
Effect of pH on enzymatic activity.
The effect of pH on the enzyme's activity is shown in Fig. 6. The results indicate that maximal activity is achieved under alkaline conditions (above 10.2). However, NAD+ appeared to be unstable at pH values above 10.5 during extended incubation. For this reason, enzymatic activities in this study were routinely determined at pH 9.0.
|
|
Kinetic parameters.
The purified recombinant xylitol-4-dehydrogenase displayed Michaelis-Menten kinetics for xylitol and NAD+ under the conditions used. The apparent Km and Vmax values determined are presented in Table 3. For comparison, the apparent Km values of the partially purified native enzyme were also determined for xylitol and for NAD+. Native xylitol-4-dehydrogenase was purified approximately 30-fold by ammonium sulfate fractionation, hydrophobic interaction chromatography, and ion-exchange chromatography as described in Materials and Methods.
|
Production of xylulose and its polarimetric analysis.
Xylulose was produced by resting recombinant E. coli ERF2157 cells as described in Materials and Methods. The time course of xylulose production and its excretion into the medium is presented in Fig. 7. Xylitol was efficiently converted to xylulose, and no major by-products could be detected in the incubation medium by HPLC. In the cell suspension initially containing 5 g/liter xylitol, 94% (mol/mol) of the xylitol consumed was converted to xylulose after 19 h of incubation, and the yield of xylulose from the xylitol initially present was 82% (mol/mol) at this time point. At an initial xylitol concentration of 10 g/liter, 89% (mol/mol) of the xylitol consumed was converted to xylulose at 19 h, and the yield from total xylitol was 70%. With a control strain transformed with the expression plasmid without the xylulose-4-dehydrogenase gene, no xylitol consumption or xylulose production could be detected during 30 h of incubation of cell suspensions.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Sequencing of the xdh gene region from the wild-type XDH-negative strain P. ananatis ATCC 19321 and its XDH-positive derivative ATCC 43072 revealed no mutations or rearrangements either in the 350-bp upstream region, including the two putative xdh promoters, or within the structural gene. Instead, a putative mutation site could be localized by PCR analysis upstream of the 350-bp region, as there was a lack of PCR products with the wild-type DNA when the upstream mutant-specific primers were used (Fig. 1), clearly suggesting that the mutation site is located outside the xdh operon and is of a regulatory nature. However, without further characterization we cannot exclude the possibility that the regulatory mutation affecting the XDH phenotype may be located somewhere else in the ATCC 43072 genome and that the observed difference in the sequences has no relevance to the XDH-positive phenotype.
The analysis of xdh-derived transcripts was hampered by the high instabilities of these mRNAs. Thus, only degradation products could be detected. However, RT-PCR and Northern analyses with various probes repeatedly showed that xdh and ORF2 form an operon structure, which is also in accordance with the DNA sequence data. RT-PCR also suggested that xdh is the first gene of this operon. The forward primers 1534 and 1533 (Fig. 1) used for RT-PCR bind between the two putative promoter regions but clearly downstream of the theoretical +1 position of the transcripts derived from the upper promoter. Primer 1533 gave a clear RT-PCR product, whereas that with primer 1534 was much weaker (data not shown). This may imply a rapid degradation of xdh transcripts from the 5' end and also explain the unsuccessful attempts at 5'-end mapping. As indicated in the results, the gene product of ORF2 has significant similarity to the ribose/sugar-binding protein of ABC transporters. If the 3.0-kb xdh-ORF2 transcripts do indeed represent the entire operon, then the size of this mRNA excludes the possibility of it encoding other putative components of the expected ABC transporter. Whether the ORF2-encoded protein and XDH are functionally related remains to be established.
A previous report has indicated that xdh is constitutively expressed in the ATCC 43072 mutant strain (8). This is in contrast to our results, which show that xdh transcription is strictly repressed in the presence of glucose, regardless of the presence of xylitol in the growth medium. No cre-like consensus sequence (10) was found, however, within the promoter region sequenced. It remains to be tested whether other carbon sources besides glucose could allow transcription of the xdh operon without the presence of xylitol.
The amino acid sequence of the P. ananatis xdh gene shares significant similarity with a number of members of the SDR superfamily. The SDR superfamily forms a large and functionally heterogeneous protein family, for which three-dimensional structures display highly similar /ß folding patterns (13, 24). SDR has five families, of which the classical and extended families are the largest ones. The classical family has been further divided into seven subfamilies (13, 25). The similar secondary structure elements and sequence motifs common to all classical short-chain dehydrogenases were found in P. ananatis XDH. However, the conserved amino acid residue (D/E) at the end of the second ß-strand at the beginning of the Rossmann fold, which determines different coenzyme-binding subfamilies (13, 25), could not be found in XDH, thus preventing further determination of the subfamily of P. ananatis XDH. Surprisingly, in spite of the conserved overall structure, the cysteine residues, which have a strong impact on the XDH activity, show no conservation regarding the primary sequence or secondary structures.
Since it proved to be very difficult to purify sufficient amounts of xylitol-4-dehydrogenase from the xylitol-positive P. ananatis mutant ATCC 43072, we were not able to compare the pure native enzyme with the purified recombinant enzyme. However, the apparent Km values for xylitol and NAD+ determined for the partially purified native enzyme corresponded well with those for the purified recombinant enzyme (Table 3).
Partial purification of the native xylulose-4-dehydrogenase from P. ananatis ATCC 43072 has also been reported previously by Doten and Mortlock (6). The substrate specificity of the partially purified native enzyme and the results presented in the present work for the recombinant enzyme are roughly in agreement. However, there are several discrepancies between the earlier report and the results of the present work. Doten and Mortlock did not report the use of Mg2+ as a cofactor in activity assays or in any of the buffers used in the purification of the enzyme from P. ananatis. In contrast, our attempts to isolate the native enzyme in the absence of Mg2+ led to a total loss of enzymatic activity at early stages of the purification. The molecular mass of the nondenatured native enzyme was estimated earlier to be 136 kDa by nondenaturing polyacrylamide gel electrophoresis, whereas a value of 100 kDa was obtained for the recombinant enzyme, using analytical gel filtration, in the present study. The apparent Km value of 48 mM reported previously for the partially purified native enzyme is fivefold higher than the one determined in the present work for the recombinant xylitol-4-dehydrogenase. However, in this case different constant concentrations of the cosubstrate NAD+ were used (0.83 mM previously and 1.8 mM in the present work). In addition, our results indicate that the recombinant enzyme has significant side activity for the D-form of xylulose (20% of that for L-xylulose), which was not reported by Doten and Mortlock for the partially purified native enzyme. These differences between the two reports remain unexplained at the present time.
NADPH-linked L-xylulose reductases (synonymous to xylitol-4-dehydrogenases) have been identified from different mammalian tissues. The apparent Km values for xylitol, ranging from 10 mM to 40 mM (2, 11, 20), for these enzymes are in the same order of magnitude as the value (9.4 mM) for the P. ananatis xylitol-4-dhydrogenase. A comparison of the previously reported NADP+-linked xylitol-4-dehydrogenase from guinea pig liver and the NAD+-linked enzyme from P. ananatis revealed similarities in their properties. Both enzymes are stimulated by Mg2+ and inhibited by dithiothreitol and have an alkaline pH optimum for activity.
The L-xylulose reductases (or xylitol-4-dehydrogenases) of the molds Penicillium chrysogenum (4), A. niger (35), and H. jecorina (26) have been reported to be NADPH specific. In these microorganisms, the L-xylulose reductase-catalyzed reaction is part of the pathway of L-arabinose catabolism. The L-xylulose reductase from H. jecorina has a requirement for Mg2+ and a side activity for the D-form of xylulose, as does the P. ananatis enzyme. No kinetic parameters for xylitol oxidation are available for the H. jecorina enzyme. The A. niger enzyme has a very high apparent Km for xylitol (925 mM), which reflects its role in the catabolism of L-arabinose, where the reaction in the opposite direction, the reduction of L-xylulose to xylitol, is an intermediate step in the pathway (35).
The only known microbial NADH-specific L-xylulose reductase has been identified from the yeast A. monospora. The A. monospora enzyme has kinetic parameters similar to those of the recombinant P. ananatis xylitol-4-dehydrogenase, with an apparent Km of 7.2 mM for xylitol and an apparent Vmax of 630 nanokatals, whereas these values were 9.4 mM and 695 nanokatals, respectively, for the recombinant P. ananatis enzyme.
Although the results suggest that the recombinant xylitol-4-dehydrogenase also has activity for D-xylulose, we found no evidence that D-xylulose was present, at least in any significant amounts, in the xylulose produced by recombinant E. coli from xylitol. A similar production experiment has been previously carried out using resting cells of the P. ananatis ATCC 43074 mutant (8). In this case, also, analysis of the ketose produced from xylitol suggested that the D-form was absent in the xylulose excreted into the medium. It was difficult in practice to reliably determine the Km values for D- and L-xylulose because of their extremely high prices. It would, however, seem plausible that the affinity of the enzyme is far greater for L-xylulose than for D-xylulose in the concentrations that these ketoses are present in the cell. In addition, E. coli is known to metabolize D-xylose by isomerizing it to D-xylulose, which is then phosphorylated to D-xylulose-5-phosphate (9). It is also possible that some of the D-xylulose formed could be metabolized via this pathway.
Microbial production of L-xylulose appears to be superior to the complicated chemical methods reported previously (32). The reported strategies for the production of L-xylulose have been based on using resting cells. The advantages of this production mode include simple purification of the product, since no major by-products are formed and complex medium components are not used during the production phase.
High L-xylulose yields were obtained in the present study. However, the thermodynamic equilibrium between xylulose and xylitol has been shown to be strongly on the side of xylitol (27). Possibly, the NAD+/NADH ratio of the cells is high under the aerobic conditions used, which would therefore result in the efficient oxidation of xylitol. The current production strain compares favorably with the natural L-xylulose producers reported in the literature. The highest L-xylulose yields from xylitol of 80% for Alcaligenes sp. strain 701B (14) and 70% for P. ananatis ATCC 43074 (8) were obtained at a xylitol concentration of 5 g/liter in 24 h and 18 h, respectively. With the recombinant cells used in the present study, a high yield (70%) could also be achieved in 19 h with an initial xylitol concentration of 10 g/liter, whereas in the previous reports with Alcaligenes sp. strain 701B and P. ananatis, only 50% and 48% yields, respectively, could be achieved at this xylitol concentration. The possibility of using higher xylitol concentrations enables higher volumetric productivities. Furthermore, it is most likely that optimization of the production conditions (pH, aeration, and initial xylitol concentration) would result in even more efficient conversion of xylitol to L-xylulose.
The production of rare and unnatural sugars appears at first hand to be a suitable target for the development of microbial production processes, since the prices of these compounds are high. Although L-xylulose is very expensive, the markets for it are small. However, an efficient production method for L-xylulose would, in principle, enable the utilization of this rare sugar in new applications of commercial value, such as in the production of L-xylose and other L-sugars. The first available bacterial gene sequence for the P. ananatis xylitol-4-dehydrogenase opens new possibilities for metabolic engineering of rare sugar production into bacterial hosts.
|
ACKNOWLEDGMENTS |
|---|
This research was funded by an ABS graduate school scholarship to Johannes Aarnikunnas and by the Academy of Finland (210778).
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pst size marker; lanes 2 to 4, cDNA, RNA, and DNA samples, respectively, PCR amplified by using the xdh-specific primers 1532 and 1501; lanes 5 to 7, cDNA, RNA, and DNA samples, respectively, PCR amplified by using the xdh- and ORF2-specific primers 1428 and 1498; lanes 8 to 10, cDNA, RNA, and DNA samples, respectively, PCR amplified by using the ORF2-specific primers 1524 and 1521 (see Fig. 
), 50 mM Tris-HCl (pH 6.9 to 9.1) (
), and 50 mM Tris-50 mM glycine-NaOH (pH 8.7 to 10.2) (
). Values are presented relative to the maximum activity obtained. The errors in the assays were <10%.
) or 10 g/liter (•) xylitol as presented in Materials and Methods. The initial cell density was 5.5 g/liter (cell dry weight per volume).