Yeast strains H. polymorpha CBS4732 (leu2-2) 27, Δxyl1 23 and transformants listed in Table 1 were grown on YPD (0.5% yeast extract, 1% peptone, 2% glucose) or minimal medium (0.67% YNB without amino acids, 4% xylose or 2% glucose) at 37°C. For the CBS4732s strain leucine (40 mg L^-1) was supplemented into the medium. For selection of yeast transformants on YPD, 130 – 150 mg L^-1 of zeocin or 0.5 – 0.6 mg L^-1 of G418 were added.

The E. coli DH5α strain (Φ80d lacZΔM15, recA1, endA1, gyrA96, thi-1, hsdR17(r_K^-, m_K⁺), supE44, relA1, deoR, Δ(lacZYA-argF)U169) was used as a host for propagation of plasmids. Strain DH5α was grown at 37°C in LB medium as described previously 28. Transformed E. coli cells were maintained on a medium containing 100 mg L^-1 of ampicillin.

Standard cloning techniques were used as described 28. Genomic DNA of H. polymorpha was isolated using the Wizard^® Genomic DNA Purification Kit (Promega, Madison, WI, USA). Restriction endonucleases and DNA ligase (Fermentas, Vilnius, Lithuania) were used according to the manufacturer specifications. Plasmid isolation from E. coli was performed with the Wizard^® Plus SV Minipreps DNA Purification System (Promega, Madison, WI, USA). PCR-amplification of the fragments of interest was done with Platinum^® Taq DNA Polymerase High Fidelity (Invitrogen, Carlsbad, CA, USA) according to the manufacturer specification. PCRs were performed in GeneAmp^® PCR System 9700 thermocycler (Applied Biosystems, Foster City, CA, USA). Transformation of the yeast H. polymorpha by electroporation was carried out as described previously 29.

Recombinant plasmids pX1N-Z and pX1M-Z bearing native and modified version of XR, respectively, were constructed on the basis of the plasmid pUC57 (Fermentas, Vilnius, Lithuania). BamHI/SacI fragment with the HpGAP promoter and HpAOX terminator from the plasmid pKO8-GAPpr 23 was cloned into the BamHI-SacI digested plasmid pUC57 after elimination of restriction sites for NdeI and HindIII. The resulting plasmid restriction sites NdeI and NotI located between the HpGAP promoter and HpAOX terminator were removed, and a unique restriction site for HindIII produced. The ORF of XYL1 was PCR-amplified from genomic DNA of CBS4732 using pair of primers HpX1for (CCC

AAG CTT

AAG CTT

GGT ACC TG CAG

GGT ACC TG CAG

The H. polymorpha XYL2 gene with own terminator and the HpGAP promoter were amplified from the genomic DNA of CBS4732 using the corresponding pairs of primers Ko134 (GAT AAT ATA GAA ACA AAA TGA ACA ATC CTT CTG CTG)/Ko133 (ACA

GGA TCC

GGA TCC

The expression cassette containing prGAP-XYL3-trAOX was obtained as SacII restriction fragment from the plasmid pKO8/GAP/HpXYL3 24 and cloned into the SacII linearized plasmids pGLG61 30. The resulting plasmid was designated pGLG61/HpXYL3 (Figure 1). The accuracy of constructed plasmids was verified by sequencing. Constructed plasmids are presented in Table 1.

The XR activity in cell extracts was determined spectrophotometrically at 37°C. The XR assay mixture contained: Tris-HCl buffer (pH 7.0) 100 mM, NADPH 0.15 mM and xylose 350 mM. The reaction was started with cell extract addition 3. To evaluate K_M towards NADPH or NADH, XR activities were measured with four different concentrations of cofactors 20, 50, 100 and 150 μM in triplicate.

The XDH activity in cell extracts was determined spectrophotometrically at 37°C. The XDH assay mixture contained: Tris-HCl buffer (pH 8.8) 100 mM, MgCl₂ 10 mM, NAD 3 mM and xylitol 300 mM. The reaction was started with cell extract addition 3.

The XK activity in cell extracts was determined spectrophotometrically at 37°C as described before 31, with some modifications. The XK assay mixture contained: Tris-HCl buffer (pH 7.8) 50 mM, MgCl₂ 5 mM, NADH 0.2 mM, phosphoenolpyruvate 1 mM, D-xylulose 8.5 mM, lactate dehydrogenase (EC 1.1.1.27) (Fluka, St. Louis, MO, USA) 10 U, pyruvate kinase (EC 2.7.1.40) (Fluka, St. Louis, MO, USA) 0.05 U, and ATP 2 mM. The reaction was started with addition of cell extract. For the XK assay, a blank without pyruvate kinase and lactate dehydrogenase was used to minimize interference of XDH activity in H. polymorpha.

All assay experiments were repeated at least twice.

Cells of transformants were grown in 100 ml of YPX medium (1% yeast extract, 2% peptone, 4% xylose) in Erlenmeyer flasks (bottle size – 300 ml) for 2 days and then inoculated into the 40 ml of YNB medium with 12% xylose in 100 ml Erlenmeyer flasks. Fermentation was carried out at the temperature of 48°C with limited aeration (140 revolutions × min^-1). Concentrations of ethanol in medium were determined using alcohol oxidase/peroxidase-based enzymatic kit "Alcotest" 32. Concentrations of xylitol in medium were determined enzymaticaly as described earlier (Enzymatic determination of D-sorbitol and xylitol, R-Biopharm GmbH, Darmstadt, Germany) with slight modifications. Nitrotetrazolium Blue (NTB) 12 mM and phenazine methosulfate 15 mM were used instead iodonitrotetrazolium chloride and diaphorase, respectively. The absorbance of the reduced NTB was measured at 570 nm. Concentrations of xylose from fermentation in mineral medium were analyzed by chemical method as described before 33. The biomass was determined turbidimetrically in a FEK-56M photoelectric colorimeter (cuvette, 3 mm, light filter no. 6) with gravimetric calibration.

Experiments were performed at least twice.

To improve alcoholic fermentation of xylose and decrease xylitol formation, XR of H. polymorpha was subjected to site-specific mutagenesis to reduce its affinity for NADPH. The amino acid sequence of cofactor binding site of H. polymorpha XR shows strict homology to the corresponding site of other xylose-utilizing yeasts (Figure 2). Our aim was to substitute lysine and asparagine for arginine and aspartic acid at amino-acid positions 341 and 343, respectively. These substitutions resemble those developed for successful modification of XR cofactor specificity in Candida tenuis 34.

To generate strains with overexpression of native or modified versions of XR or strains with simultaneous overexpression of native or modified XR together with XDH, the H. polymorpha Δxyl1 23 strain was transformed with SacI linearized plasmids pX1N-Z and pX1M-Z or pX1N-Z-X2 and pX1M-Z-X2, respectively. The transformants were grown on YPD medium supplemented with zeocin. The presence of expression cassettes in the transformants was examined by PCR using corresponding primers. To express the XK, the recombinant plasmid pGLG61/HpXYL3 was transformed into the recipient strain H. polymorpha overexpressing native or modified versions of XR and XDH. The transformants were grown on YPD medium in the presence of increasing concentrations of G418 addition. The highest concentration of G418, which allows the transformants to grow, was 0.4 mg × ml^-1. Colonies able to grow on the selective medium appeared after 3 days of incubation with frequency of up to 20 transformants × mg^-1 DNA. The transformants were stabilized by cultivation in non-selective media for 10–12 generations with further shifting to the selective media with G418. The presence of recombinant XYL3 gene driven by the HpGAP promoter in genomic DNA of stable transformants was confirmed by PCR. As pGLG61-based plasmids promote multiple integration into the genome of recipient strains 30, constructed strains were examined by Southern hybridisation to select recombinant strains with equal amount of XK expression cassette. The strains bearing 3 copies of XK expression cassettes were selected (data not shown). Constructed yeast strains are represented in Table 1.

Biochemical properties of XR in one of constructed recombinant strains (designated XRm) were studied. Specific activities of XR (with both cofactors NADPH and NADH), XDH and XK as well as affinities of native XR (XRn) and engineered XRm were measured (Table 2). XRm was characterized by K_M of 152 μM for NADPH using xylose as a substrate, which is 17 times higher than the K_M for NADPH of the native XR (9 μM). K_M of engineered XRm for NADH remained nearly unchanged (112 μM). Specific activity of XR with NADPH in the XRm strain decreased 9.1 times when compared with the strain overexpressing native XR. Specific activity of XR with NADH in both strains remained unchanged. Strains XRn/XDH and XRm/XDH with additional overexpressed XDH possessed two-fold increase in the specific activity of XDH as compared with the wild type strain. Overexpression of XK in strains XRn/XDH/XK and XRm/XDH/XK resulted in up to 2.4-fold increase in specific activity of XK as compared to CBS4732 (Table 2).

Xylose fermentation by constructed strains was compared in batch cultures with limited aeration. A mineral medium containing xylose (12%) and an initial biomass concentration of 2 g (dry weight) × L^-1 were used. Results of ethanol and xylitol production by the constructed strains are shown in Table 2. Ethanol productivity of the XRm strain was 9.8 mg × (L × h)^-1, which is 1.5- and 1.3-fold higher than the productivity of the XRn and the wild-type strain CBS4732, respectively. Xylitol production of these strains varied insignificantly. Ethanol productivity of the strain XRm/XDH (18.4 mg × (L × h)^-1) was increased 1.5 and 2.4 times as compared to XRn/XDH and CBS4732, respectively. Strain XRm/XDH possessed 1.9- and 2.6-fold reduction in xylitol production compared with XRn/XDH and CBS4732 strains. Ethanol productivity of the strain XRm/XDH/XK (54.7 mg × (L × h)^-1) was 4- and 7.3-fold higher compared to those of the strain XRn/XDH/XK (13.8 mg × (L × h)^-1) and CBS4732 (7.5 mg × (L × h)^-1). Xylitol production of the strain XRm/XDH/XK was significantly reduced to 0.9 mg × (L × h)^-1, which is 3- and 4.7-fold lower than those of the XRn/XDH/XK and control strain, respectively. Representative fermentation profiles for the strains XRn/XDH/XK, XRm/XDH/XK and CBS4732 are shown in Figure 3. The specific ethanol and xylitol productivities per biomass at [mg (L g h)^-1] of the recombinant and control strains were also calculated and represented at the Table 2. As biomass concentrations of the strains in all fermentation experiments were very similar, the values of comparison of specific ethanol and xylitol productivities per biomass between the strains are in the same level. The consumption of xylose by H. polymorpha strains during the fermentation is low. Ethanol produced in the initial stage of xylose fermentation is reutilized after 1–2 days of fermentation.