Identification in the yeast Pichia stipitis of the first L-rhamnose-1-dehydrogenase gene Outi M. Koivistoinen, Satu Hilditch, Sanni P. Voutilainen, Harry Boer, Merja Penttila ¨ and Peter Richard VTT Technical Research Centre of Finland, Espoo, Finland l-Rhamnose (l-6-deoxy-mannose) is a C6 sugar that is enriched in some fractions of plant biomass, such as hemicellulose and pectin. Several microorganisms liv- ing on decaying plant material are able to use l-rham- nose as a source of carbon and energy. There are at least two pathways for the catabolism of l-rhamnose, one with phosphorylated intermediates and one with- out. The path with the phosphorylated intermediates was described in bacteria, and involves the following intermediates: l-rhamnulose, l-rhamnulose 1-phos- phate, dihydroxyacetone phosphate, and l-lactalde- hyde. The corresponding enzymes are l-rhamnose isomerase (EC 5.3.1.14) [1,2], rhamnulokinase (EC 2.7.1.5) [3,4] and rhamnulose-1-phosphate aldolase (EC 4.1.2.19) [5]. l-Lactaldehyde can then be reduced to 1,2-propenediol or oxidized to lactate by lactalde- hyde reductase (EC 1.1.1.77) or lactaldehyde dehydro- genase (EC 1.2.1.22) respectively, depending on the redox conditions [6]. Gene sequences for all these enzymes have been described [7]. The pathway without phosphorylated intermediates is distinctly different. It has been described in different yeast species [8,9]. The intermediates in this pathway are l-rhamnono-1,4-lactone, l-rhamnonate, l-erythro- 3,6-dideoxyhexulosonate, pyruvate and l-lactaldehyde. The corresponding enzymes are NAD-utilizing l-rhamnose-1-dehydrogenase (EC 1.1.1.173) [10,11], l-rhamnono-1,4-lactonase (EC 3.1.1.65), l-rhamnonate dehydratase (EC 4.2.1.90), and l-erythro-3,6-dide- oxyhexulosonate aldolase (EC 4.1.2 ) (Fig. 1). The Keywords L-rhamnonate; L-rhamnose catabolism; L-rhamnose dehydrogenase; MALDI- TOF MS; Pichia stipitis Correspondence P. Richard, Technical Research Centre of Finland, Tietotie 2, Espoo, PO Box 1000, 02044 VTT, Finland Fax: +358 20 722 7071 Tel: +358 20 722 7190 E-mail: peter.richard@vtt.fi (Received 4 January 2008, revised 10 March 2008, accepted 11 March 2008) doi:10.1111/j.1742-4658.2008.06392.x There are two distinctly different pathways for the catabolism of l-rham- nose in microorganisms. One pathway with phosphorylated intermediates was described in bacteria; here the enzymes and the corresponding gene sequences are known. The other pathway has no phosphorylated intermedi- ates and has only been described in eukaryotic microorganisms. For this pathway, the enzyme activities have been described but not the correspond- ing gene sequences. The first enzyme in this catabolic pathway is the NAD- utilizing l-rhamnose 1-dehydrogenase. The enzyme was purified from the yeast Pichia stipitis , and the mass of its tryptic peptides was determined using MALDI-TOF MS. This enabled the identification of the correspond- ing gene, RHA1. It codes for a protein with 258 amino acids belonging to the protein family of short-chain alcohol dehydrogenases. The ORF was expressed in Saccharomyces cerevisiae. As the gene contained a CUG codon that codes for serine in P. stipitis but for leucine in S. cerevisiae, this codon has changed so that the same amino acid was expressed in S. cerevisiae. The heterologous protein showed the highest activity and affinity with l-rhamnose and a lower activity and affinity with l-mannose and l-lyxose. The enzyme was specific for NAD. A northern blot analysis revealed that transcription in P. stipitis is induced during growth on l-rhamnose but not on other carbon sources. Abbreviation YNB, yeast nitrogen base. 2482 FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS l-lactaldehyde is oxidized to l-lactate in an NAD-cou- pled reaction, as in the pathway with the phosphory- lated intermediates. For this pathway, only the enzyme activities have been described; the corresponding genes have not been identified in any yeast or in any other organism. l-Rhamnose dehydrogenase activity was described in Pichia stipitis NRC5568. This enzyme used NAD as a cofactor. The enzyme activity was l-rhamnose- induced and d-glucose-repressed. With the crude cell extract, an activity of about 0.1 lmolÆmin )1 Æmg )1 was observed. It was suggested that the reaction product was the l-rhamnono-d-lactone and not the more stable l-rhamnono-c-lactone, as the c-lactone could not be identified as a reaction product [9]. In the present work, we identified the gene coding for the l-rhamnose dehydrogenase in P. stipitis.We expressed it in the heterologous host Saccharomy- ces cerevisiae and characterized the enzyme kinetic properties. Results The P. stipitis strain CBS 6054 was grown on yeast nitrogen base (YNB) supplemented with 2% l-rham- nose, 1% d-glucose and 1% l-rhamnose or 2% d-glucose. The cells were harvested before the sugars were utilized, and the crude cell extract was analyzed for l-rhamnose dehydrogenase activity. Cells grown on l-rhamnose as a sole carbon source had an l-rhamnose dehydrogenase activity of 14 nkatÆmg )1 of extracted protein. Cells grown on the d-glucose ⁄ l-rhamnose mixture had an activity of 2 nkatÆmg )1 , and the cells grown on d-glucose did not show any l-rhamnose dehydrogenase activity. We used the cell extract of the l-rhamnose-grown cells to purify the protein. The purification included three steps: a DEAE column, native PAGE, and SDS ⁄ PAGE. From the DEAE column, which was eluted with a salt gradient, the activity eluted as a sin- gle peak with a specific activity of 10 nkatÆmg )1 . The fractions around this activity peak were analyzed by SDS ⁄ PAGE, and showed about 20 different proteins (Fig. 2). The active fraction was then concentrated and separated by native PAGE, and a single band with l-rhamnose dehydrogenase activity was identified using zymogram staining. This active band from the native PAGE was cut out from the gel, and the partially puri- fied protein was eluted. It was then applied to an SDS ⁄ PAGE gel, and this revealed four proteins with estimated sizes of 30, 35, 52 and 70 kDa (Fig. 2). The 30 kDa protein was preliminarily identified as the COOH C O CH 3 HC C H CH 3 HO O pyruvic acid L-lactaldehyde L-rhamnose L-rhamnose dehydrogenase EC 1.1.1.173 L-erythro-3,6-dideoxy hexulosonic acid aldolase EC 4.1.2. L-lactaldehyde dehydrogenase EC 1.2.1.22 L-lactic acid COOH C OH C C H OH C HO H HO H CH 3 H COOH C O C C H H C HO H HO H CH 3 L-rhamnonic acid L-rhamnonic ac id dehydratase EC 4.2.1.90 L-erythro-3,6-dideoxy hexulosonic acid COOH C H CH 3 HO NAD H 2 O NADH NADH NAD OH H H HO OH H H O H 3 C HO H (OH) (H) H HO OH H H O H 3 C HO H O L-rhamnonic acid-1,4-lactone L-rhamnonic acid lactonase EC 3.1.1.65 Fig. 1. Fungal path for L-rhamnose catabolism. The enzyme activities but not the corresponding gene sequences of this pathway have been described previously. The identification of a gene coding for the L-rhamnose dehydrogenase is the subject of the present article. O. M. Koivistoinen et al. Identification of an L-rhamnose dehydrogenase gene FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS 2483 l-rhamnose dehydrogenase. This was done by compar- ing the results of the SDS ⁄ PAGE that was done after the zymogram staining with those of the SDS ⁄ PAGE of the active fractions after the DEAE column and correlating them with the enzyme activity of that frac- tion. The 30 kDa protein matched with l-rhamnose dehydrogenase activity. After trypsination, the peptide masses of the 30 kDa protein were determined by MALDI-TOF MS. As the genome sequence of P. stipitis is known [12], these masses allowed the identification of the protein on the basis of matching peptide sequences. The masses 555.247, 900.475, 1199.639, 1761.782, 1872.708 and 2552.586 were identified as tryptic peptides of a protein with the GenBank identifier ABN68405. This protein had been annotated as a putative d-glucose-1-dehydro- genase II. The protein has 258 amino acids and a calculated molecular mass of 27.102 Da, and belongs to the family of short-chain alcohol dehydrogenases. We called the gene for the first gene in the l-rhamnose catabolic pathway RHA1. To verify that we had indeed identified the l-rham- nose dehydrogenase, we expressed the protein in S. ce- revisiae. P. stipitis is known to translate CTG to serine and not to leucine [13]. The l-rhamnose dehydrogenase contained one such codon at bp 166–168 of the ORF, which we changed to TCG. In this way, we ensured that a protein with the same amino acid sequence was expressed in S. cerevisiae.InS. cerevisiae, the l-rham- nose dehydrogenase was expressed from a multicopy plasmid with the S. cerevisiae PGK1 promoter, which is a strong and constitutive promoter. In the crude extract of S. cerevisiae, we found an l-rhamnose dehy- drogenase activity of about 200 nkatÆmg )1 of protein. In the control strain, which contained the empty vec- tor, no activity was observed. In order to facilitate the purification, we expressed the Rha1 protein in S. cerevisiae with an N-terminal or with a C-terminal histidine-tag. The histidine-tags were introduced by adding the additional nucleotide sequence by PCR as specified in Experimental proce- dures. Both constructs were expressed with the same vector in the same yeast strain. When testing the two modified proteins in the crude extract of S. cerevisiae, we observed that the N-terminally tagged enzyme did not exhibit any activity. The C-terminally tagged enzyme showed activity in the crude extract; however, the activity was reduced by about 80% when compared to the activity of the nontagged enzyme in the crude extract. As the tag- ging of the enzyme had such a strong effect on the activity, we did not proceed to purify the enzyme. Instead, we used the crude cell extract of the S. cere- visiae strain expressing the untagged enzyme for the kinetic characterization. We observed activity with l-rhamnose, l-lyxose, and l-mannose. No activities were observed with d-eryth- rose, d-allose, d-ribose, d-arabinose, d-tagatose, d-glu- cose, d-galactose, d-xylose and l-arabinose, and none of the sugars showed activity in the control strain with the empty vector. The highest activity, V max about 200 ± 20 nkatÆmg )1 of protein in the crude extract, was observed with l-rhamnose. With l-lyxose and l-mannose, the activities were lower; the V max values were 170 ± 20 nkatÆmg )1 and 75 ± 10 nkatÆmg )1 respectively. The highest affinity was towards l-rham- nose, the K m being 1.5 ± 0.025 mm. Lower affinities were obtained with l-lyxose and l-mannose; here, the K m values were 5 ± 0.5 mm and 25 ± 5 mm respec- tively. The enzyme showed activity with NAD as a cofactor; the V max was 200 ± 20 nkatÆmg )1 , and the K m was 0.2 ± 0.03 mm (Fig. 3). No activity was observed with NADP as a cofactor (Fig. 4). The activity was pH-dependent. At pH 6.8, the activity was about 100 ± 10 nkatÆmg )1 , at pH 8.0 it was 200 ± 20 nkatÆmg )1 , and at pH 9.5 it was 240 ± 25 nkatÆmg )1 . To test the activity in the reverse direction, we incubated the enzyme preparation with l-rhamnonate and NADH at pH 8.0. No activity was observed under these conditions. At this pH, l-rhamn- onate is expected to be in the linear and not in the c-lactone or d-lactone form. AB C Fig. 2. Coomassie-stained SDS ⁄ PAGE gel of the protein fractions after the different purification steps. Lane A contains the molecular mass markers (masses in kDa are indicated). Lane B contains the combined active fractions after the DEAE column separation. Lane C shows the protein eluted from the excised band of the native PAGE gel after zymogram staining. Identification of an L-rhamnose dehydrogenase gene O. M. Koivistoinen et al. 2484 FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS Northern analysis To study the role of the l-rhamnonate dehydrogenase in P. stipitis, the transcription of RHA1 with different carbon sources was studied by northern analysis. The P. stipitis strain CBS 6054 was grown on l-rhamnose, d-glucose, maltose, d-galactose, d-xylose and a glyc- erol ⁄ ethanol mixture as carbon sources. The results are shown in Fig. 5. We observed transcription only on l-rhamnose, suggesting that RHA1 expression is l-rhamnose-induced. Discussion There are at least two different catabolic pathways for l-rhamnose. One pathway has phosphorylated inter- mediates; the enzyme activities and the corresponding genes have been well described, and it has only been observed in prokaryotic microorganisms. The other pathway has no phosphorylated intermediates, and has so far been described only in yeast. For this pathway, the enzyme activities have been described; however, none of the corresponding genes had been identified. In this work, we identified the gene coding for the first enzyme in this pathway, an NAD:l-rhamnose-1-dehy- drogenase. Twerdochlib et al. [9] had reported previously that l-rhamnose dehydrogenase activity was present in P. stipitis when the yeast was grown on a mixture of d-glucose and l-rhamnose, and absent when grown on d-glucose as a sole carbon source. We confirmed this, and also noticed that the activity was increased sev- eral-fold when the yeast was grown on l-rhamnose as a sole carbon source. The enzymatic activity that was induced in this way was then purified. During the puri- fication, the activity always appeared as a single peak, indicating that only one enzyme is responsible for this activity. The purified enzyme was then digested with trypsin, and the masses of the peptides were identified using MALDI-TOF MS. The genome sequence of P. stipitis is available, and this enabled the identifica- tion of the corresponding ORF, which was then called RHA1. RHA1 was induced when the yeast was grown on l-rhamnose, but not when it was grown on any other carbon sources, as shown by the northern blot analysis (Fig. 5). This suggests that the induction of the l-rhamnose dehydrogenase activity is the result of induction of transcription of RHA1. To characterize the enzyme’s kinetic properties, we expressed RHA1 in the heterologous host S. cerevisiae. This resulted in l-rhamnose-1-dehydrogenase activity, showing that this gene does indeed code for a protein with this activity. We also added a histidine-tag to the N-terminus or to the C-terminus of the protein in order to facilitate the purification. However, the tagged proteins showed no or very much reduced activity in the crude extract. As this might be an indication that the tag is interfering with the catalytic activity, we did not use any of the tagged proteins for the kinetic char- acterization, but used the crude cell extract of the S. cerevisiae strain expressing RHA1. The Rha1 protein was specific for NAD as a cofac- tor, which is in agreement with earlier observations [9]. The sugars that were accepted in the catalytic reaction were l-rhamnose, l-lyxose, and l-mannose (Figs 3 and 4). C1 to C4 in these sugars share the same configura- tion. When the hydroxyl group at C4 was in the oppo- site configuration, as in d-ribose, no activity was observed. Also, no activity was found with any other C1–C4 configuration, and when the C5 was missing, as 0 10 20 30 40 50 60 0 50 100 150 200 A B L-rhamnose L-lyxose L-mannose Activity (nkat·mg –1 ) Sugar concentration (mM) 0.0 0.5 1.0 1.5 2.0 0 50 100 150 200 NA D Activity (nkat·mg –1 ) NAD (mM) Fig. 3. Kinetic properties of the L-rhamnose-1-dehydrogenase. The heterologously expressed protein was analyzed in a crude cell extract at pH 8.0. (A) The NAD concentration is 1.5 m M. The curves are calculated assuming a Michaelis–Menten kinetic model: L-rham- nose, K m = 1.5 mM, V max = 200 nkatÆmg )1 ; L-lyxose, K m =5mM, V max = 170 nkatÆmg )1 ; L-mannose, K m =25mM, V max = 75 nkatÆ mg )1 . (B) The L-rhamnose concentration is 60 mM. The curve is cal- culated assuming a Michaelis–Menten kinetic model: K m = 0.2 mM, V max = 200 nkatÆmg )1 . O. M. Koivistoinen et al. Identification of an L-rhamnose dehydrogenase gene FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS 2485 in d-erythrose. This indicates that the C1–C4 stereo- chemical configuration is essential for recognition by the enzyme, and that an additional carbon atom must be attached to C4. We cannot say whether a hydroxyl group on C5 is required, as a sugar without a hydroxyl at C5 was not tested. Among the three sugars that showed activity, the highest activity and affinity were observed with l-rhamnose, indicating that this enzyme is indeed an l-rhamnose dehydrogenase. The possibil- ity that the enzyme is a glucose-1-dehydrogenase, as suggested in the first annotation based on sequence similarity, can be excluded, as no activity was observed with glucose. Twerdochlib reported that the l-rhamnonse dehy- drogenase from P. stipitis did not produce any detect- able amounts of l-rhamnono-c-lactone, suggesting that this enzyme produced the more unstable l-rhamnono- d-lactone [9]. If the l-rhamnono-d-lactone was in a rapid equilibrium with the l-rhamnonic acid at neutral pH, one might expect to see some reverse activity with l-rhamnonic acid and NADH. We tested the reverse reaction but could not observe any, indicating that the intermediate is a lactone that, at neutral pH, is present in too low concentrations for the reverse reaction to occur. Sugar dehydrogenases that oxidize the sugar to a sugar acid are not very common in eukaryotic micro- organisms. S. cerevisiae has NADP-requiring [14] and NAD-requiring d-arabinose dehydrogenases [15], ARA1 and ARA2, which contribute to erythroascorbic acid production. These proteins belong to the family of aldo ⁄ keto reductases. In the mold Hypocrea jecori- na, an NADP-requiring d-xylose dehydrogenase was described that belonged to the GFO ⁄ IDH ⁄ MOCA protein family [16]. There are also other reports of eukaryotic sugar dehydrogenases, such as an NADP-utilizing d-glucose dehydrogenase in Schizosaccharomyces pombe [17], an NAD-utilizing d-glucose dehydrogenase in Aspergil- lus niger [18], and an NADP-utilizing d-xylose dehy- drogenase in Pichia quercuum [19]. However, for these proteins, the corresponding sequences are not known, and it is not clear to what protein family they belong. The protein described in this article belongs to the pro- tein family of short-chain dehydrogenases, and has the conserved domain of a fabG [3-ketoacyl-(acyl-carrier protein) reductase]. The sugar dehydrogenases in eukary- otic microorganisms belong to very different protein families, although the catalytic reaction is very similar. Experimental procedures Enzyme assays If not otherwise specified, the enzyme activity was measured in a reaction mixture containing 100 mm Tris ⁄ HCl (pH 8.0), 1 mm NAD, and the crude cell extract or a pro- tein preparation. The reaction was started by the addition of 10 mml-rhamnose or other sugars when specified. The formation of NADH was followed by measuring the absorbance at 340 nm. To assay the enzyme activity in the reverse direction, the crude cell extract was incubated in CHO C OH C C H OH C HO H HO H CH 3 H CHO C OH C C H OH C HO H HO H CH 2 OH H CHO C OH C C H OH CH 2 OH HO H H L-rhamnose L-mannose L-lyxose Fig. 4. Fischer projection of the sugars that showed activity with the L-rhamnose dehydrogenase. ABCDE F Fig. 5. Northern blot analysis of RHA1 expression. The expression of RHA1 in P. stipitis on L-rhamnose (lane A), D-glucose (lane B), maltose (lane C), D-galactose (lane D), D-xylose (lane E) and a glyc- erol ⁄ ethanol mixture (lane F). The lower panel shows the total RNA in the gel after staining with the SYBR Green II RNA gel stain. Identification of an L-rhamnose dehydrogenase gene O. M. Koivistoinen et al. 2486 FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS 100 mm Tris ⁄ HCl (pH 8.0), 200 lm NADH and 100 mm l-rhamnonate. The disappearance of NADH was followed by measuring the absorbance at 340 nm. l-Rhamnonate was synthesized from l-rhamnose by oxidation with bromine, and purified by ion exchange chromatography as described by Yew et al. [20]. The analysis was done in a Cobas Mira automated analyzer (Roche, Basel, Switzerland) at 30 °C. Enzyme purification The P. stipitis strain CBS 6054 was grown in 500 mL of medium containing YNB without amino acids (BD, Rock- ville, MD, USA) and 2% l-rhamnose as a carbon source in shake flasks under aerobic conditions for about 2 days. The yeast was collected by centrifugation at 3000 g for 15 mins, washed, and resuspended in 40 mL of 5 mm sodium phos- phate (pH 7.0) supplemented with Complete medium with- out EDTA (Roche) protease inhibitor. Equal amounts of glass beads (0.4 mm diameter), fresh cell cake and resuspen- sion buffer were extracted in a Mini-Bead Beater (Biospec Products, Bartlesville, OK, USA) two times for 1 min each. The mixture was then centrifuged in an Eppendorf micro- centrifuge at full speed for 20 min at 4 °C. The supernatant was desalted with a PD10 column (GE Healthcare, Amer- sham, UK) equilibrated with 5 mm sodium phosphate (pH 7.0) and subsequently loaded onto a 10 mL DEAE column (Merck, Darmstadt, Germany). The protein amount loaded onto the column was about 16 mg. The column was then eluted with 200 mL of a linear gradient from 0 to 200 mm NaCl in the same buffer. Fractions of 2.5 mL were collected and analyzed for l-rhamnose dehydrogenase activ- ity. The fractions in which activity was observed were then concentrated using Vivaspin 2 10 000 MWCO PES centrifu- gation columns (Vivascience Satorius group) and analyzed by SDS ⁄ PAGE. The concentrated protein was then sepa- rated by native PAGE (12% acrylamide). The gel was then stained in a zymogram staining solution similar to what has been described previously [21]. The zymogram staining solu- tion contained 200 mm Tris ⁄ HCl (pH 8.0), 100 mml-rham- nose, 0.25 mm nitroblue tetrazolium, 0.06 mm phenazine methosulfate, and 0.5 mm NAD. The only band that appeared was cut out and eluted by overnight incubation in 100 mm Tris ⁄ HCl (pH 8.0) and 0.1% SDS. The protein was again concentrated, and separated by SDS ⁄ PAGE. Of the four proteins that were detected, the 30 kDa protein coin- cided with the l-rhamnose dehydrogenase activity as judged by the previous SDS ⁄ PAGE gel of the active fractions. In-gel digestion and MALDI-TOF MS The 30 kDa protein observed in the SDS ⁄ PAGE gel was in-gel digested with trypsin, and the peptides were extracted essentially according to the method of Rosenfeld et al. [22]. The samples were desalted using a C-18 matrix (Eppendorf Perfect Pure C-18 Tip). The saturated matrix solution was prepared by dissolving recrystallized a-cyano-4-hydroxycin- namic acid (CCA; Bruker Daltonics, Bremen, Germany) in a 50% acetonitrile ⁄ 0.1% trifluoroacetic acid solution. Equal volumes of purified peptide sample or calibration standard (peptide calibration mixture II; Bruker Daltonics) were mixed with the saturated matrix solution. One microliter of this matrix ⁄ sample mixture was applied to the target (target plate ground steel TF; Bruker Daltonics) and allowed to dry at room temperature. The peptide masses were then determined by MALDI-TOF MS using a Bruker Auto- flex II mass spectrometer. flexanalysis software (Bruker Daltronics) was used for the data analysis. Cloning of the ORF, site-directed mutagenesis and heterologous expression The ORF was amplified from the genomic DNA by PCR using primers 5¢- GGATCCATCATGACTGGATTGTTGA ATGG-3¢ and 5¢- GGATCCCTATTGTAAATTGACGAA CAATCCTC-3¢, and the DynazymeEXT DNA polymerase (Finnzymes, Espoo, Finland). The primers contained BamHI restriction sites (underlined) to facilitate further plasmid constructions. The PCR product was then ligated to the pCR2.1 TOPO vector (Invitrogen, Carlsbad, CA, USA) and cloned. Nucleotides 166–168 of the ORF were changed from CTG to TCG with the QuikChange site- directed mutagenesis kit (Stratagene, La Jolla, CA, USA). After the site-directed mutagenesis, the ORF was released as a BamHI fragment and ligated to the BglII site of p1181, which is a multicopy yeast expression vector based on YEplac195, containing URA3 for selection, where the PGK1 promoter and terminator were introduced [23]. The S. cerevisiae strain CEN.PK2-1D was then transformed with the resulting plasmid and grown on selective medium. A control strain contained p1181. S. cerevisiae strains expressing C-terminally or N-terminally histidine-tagged enzymes were generated in a similar way. The coding sequences for six histidines were introduced by PCR either at the N-terminus or at the C-terminus. To introduce the histidine-tag at the N-terminus, we introduced a coding sequence for MHHHHHHGG before the original start codon. To introduce the histidine-tag at the C-terminus, we introduced the coding sequence for GGHHHHHH before the stop codon. The template for the PCR was the vector where the CTG was changed to TCG. For the expression of the histidine-tagged proteins, the same plasmid was used. A crude cell extract was made by vortexing with glass beads as described above for P. stipitis. Northern analysis The P. stipitis strain CBS 6054 was grown in YNB medium supplemented with 20 gÆL )1 of the following carbon sources: l-rhamnose, d-glucose, maltose, d-galactose, d-xylose, and an ethanol ⁄ glycerol mixture. The RNA was O. M. Koivistoinen et al. Identification of an L-rhamnose dehydrogenase gene FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS 2487 extracted from the yeast cells with the Trizol reagent kit (Life Technologies Inc.); about 5 lg of the total RNA per sample was used in the analysis. The RNA amount was checked by staining (Fig. 4, lower panel) with the SYBR Green II RNA gel stain (Lonza, Rockland, ME, USA). Northern hybridization was carried out using standard methods. As a probe for the hybridization, the ORF, released as a BamHI fragment from the TOPO vector, was used. The probe was labeled with [ 32 P]dCTP[aP] (GE Healthcare) using the randomly primed DNA labeling kit (Roche). Acknowledgements This research was supported by an Academy Research Fellowship for P. Richard from the Academy of Finland. We thank Outi Ko ¨ no ¨ nen for excellent techni- cal assistance. References 1 Wilson DM & Ajl S (1957) Metabolism of L-rhamnose by Escherichia coli. I. L-rhamnose isomerase. J Bacteriol 73, 410–414. 2 Takagi Y & Sawada H (1964) The metabolism of L-rhamnose in Escherichia coli. I. L-Rhamnose isomer- ase. Biochim Biophys Acta 92, 10–17. 3 Takagi Y & Sawada H (1964) The metabolism of L-rhamnose in Escherichia coli. II. L-Rhamnulose kinase. Biochim Biophys Acta 92, 18–25. 4 Wilson DM & Ajl S (1957) Metabolism of L-rhamnose by Escherichia coli. II. The phosphorylation of L-rhamnulose. J Bacteriol 73, 415–420. 5 Sawada H & Takagi Y (1964) The metabolism of L-rhamnose in Escherichia coli. 3. L-Rhamulose- phosphate aldolase. Biochim Biophys Acta 92, 26–32. 6 Baldoma L & Aguilar J (1988) Metabolism of L-fucose and L-rhamnose in Escherichia coli: aerobic–anaerobic regulation of L-lactaldehyde dissimilation. J Bacteriol 170, 416–421. 7 Moralejo P, Egan SM, Hidalgo E & Aguilar J (1993) Sequencing and characterization of a gene cluster encoding the enzymes for L-rhamnose metabolism in Escherichia coli. J Bacteriol 175 , 5585–5594. 8 Rigo LU, Marechal LR, Vieira MM & Veiga LA (1985) Oxidative pathway for L-rhamnose degradation in Pullularia pullulans. Can J Microbiol 31, 817–822. 9 Twerdochlib AL, Pedrosa FO, Funayama S & Rigo LU (1994) L-Rhamnose metabolism in Pichia stipitis and Debaryomyces polymorphus. Can J Microbiol 40, 896–902. 10 Rigo LU, Nakano M, Veiga LA & Feingold DS (1976) L-Rhamnose dehydrogenase of Pullularia pullulans. Biochim Biophys Acta 445, 286–293. 11 Vieira MM, Rigo LU, Marechal LR & Veiga LA (1979) Induction and catabolite repression of L-rham- nose dehydrogenase in Pullularia pullulans. J Bacteriol 138, 55–59. 12 Jeffries TW, Grigoriev IV, Grimwood J, Laplaza JM, Aerts A, Salamov A, Schmutz J, Lindquist E, Dehal P, Shapiro H et al. (2007) Genome sequence of the ligno- cellulose-bioconverting and xylose-fermenting yeast Pichia stipitis. Nat Biotechnol 25, 319–326. 13 Laplaza JM, Rivas Torres B, Jin YS & Jeffries TW (2006) Sh ble and Cre adapted for functional genomics and metabolic engineering of Pichia stipitis. Enzyme Microb Technol 38, 741–747. 14 Kim ST, Huh WK, Lee BH & Kang SO (1998) D-Arabinose dehydrogenase and its gene from Saccha- romyces cerevisiae. Biochim Biophys Acta 1429, 29–39. 15 Amako K, Fujita K, Shimohata TA, Hasegawa E, Kishimoto R & Goda K (2006) NAD + -specific D-arab- inose dehydrogenase and its contribution to erythroasc- orbic acid production in Saccharomyces cerevisiae. FEBS Lett 580, 6428–6434. 16 Bergha ¨ ll S, Hilditch S, Penttila ¨ M & Richard P (2007) Identification in the mould Hypocrea jecorina of a gene encoding an NADP + :D-xylose dehydrogenase. FEMS Microbiol Lett 277, 249–253. 17 Tsai CS, Shi JL & Ye HG (1995) Kinetic studies of gluconate pathway enzymes from Schizosaccharomyces pombe. Arch Biochem Biophys 316 , 163–168. 18 Elshafei AM & Abdel-Fatah OM (2001) Evidence for a non-phosphorylated route of galactose breakdown in cell-free extracts of Aspergillus niger. Enzyme Microb Technol 29, 76–83. 19 Suzuki T & Onishi H (1973) Oxidation and reduction of D-xylose by cell-free extract of Pichia quercuum. Appl Microbiol 25 , 850–852. 20 Yew WS, Fedorov AA, Fedorov EV, Rakus JF, Pierce RW, Almo SC & Gerlt JA (2006) Evolution of enzy- matic activities in the enolase superfamily: L-fuconate dehydratase from Xanthomonas campestris. Biochemistry 45, 14582–14597. 21 Richard P, Londesborough J, Putkonen M, Kalkkinen N & Penttila ¨ M (2001) Cloning and expression of a fun- gal L-arabinitol 4-dehydrogenase gene. J Biol Chem 276, 40631–40637. 22 Rosenfeld J, Capdevielle J, Guillemot JC & Ferrara P (1992) In-gel digestion of proteins for internal sequence analysis after one- or two-dimensional gel electrophore- sis. Anal Biochem 203, 173–179. 23 Verho R, Richard P, Jonson PH, Sundqvist L, Londes- borough J & Penttila ¨ M (2002) Identification of the first fungal NADP-GAPDH from Kluyveromyces lactis. Biochemistry 41, 13833–13838. Identification of an L-rhamnose dehydrogenase gene O. M. Koivistoinen et al. 2488 FEBS Journal 275 (2008) 2482–2488 ª 2008 The Authors Journal compilation ª 2008 FEBS . either at the N-terminus or at the C-terminus. To introduce the histidine-tag at the N-terminus, we introduced a coding sequence for MHHHHHHGG before the original start codon. To introduce the. compar- ing the results of the SDS ⁄ PAGE that was done after the zymogram staining with those of the SDS ⁄ PAGE of the active fractions after the DEAE column and correlating them with the enzyme. not the correspond- ing gene sequences. The first enzyme in this catabolic pathway is the NAD- utilizing l-rhamnose 1-dehydrogenase. The enzyme was purified from the yeast Pichia stipitis , and the