Production of xylitol from glucose by a recombinant strain

Abstract

The present invention relates to a recombinant microbial host for the production of xylitol, the recombinant microbial host containing a nucleic acid sequence encoding an NAD+-specific D-arabitol 4-oxidoreductase (EC 1.1.1.11) using D-arabitol as substrate and producing D-xylulose as product, and a nucleic acid sequence encoding an NADPH-specific xylitol dehydrogenase using D-xylulose as substrate and producing xylitol as product.

Claims

1. A recombinant Pichia ohmeri selected from strains I-4982, I-4960 and I-4981 deposited at the National Collection of Microorganism Cultures.

2. A method for producing xylitol comprising culturing the recombinant Pichia ohmeri according to claim 1, and recovering xylitol.

Description

FIGURES AND SEQUENCES

(1) FIG. 1: 12 ABYWMP: Restriction map of the synthesized NAD.sup.+-specific D-arabitol 4-oxidoreductase from E. coli flanked by AscI and SphI restriction sites.

(2) FIG. 2A: lig7.78: Restriction map of the NADH-specific xylitol dehydrogenase from Pichia stipitis.

(3) FIG. 2B: 12AALQTP: Restriction map of the synthesized NADPH-specific xylitol dehydrogenase from Pichia stipitis flanked by HindIII and SacII restriction sites.

(4) FIG. 3: 13AAYSYP: Restriction map of the synthesized NADPH-specific xylitol dehydrogenase from Gluconobacter oxydans flanked by AscI and SphI restriction sites.

(5) FIG. 4: Construction of an expression cassette consisting of an open reading frame flanked by a poRR promoter and terminator using overlap PCR.

(6) FIG. 5: 12 AAMCJP: Restriction map of the synthesized tagatose-3-epimerase of Pseudomonas cichorii flanked by HindIII and SacII restriction sites.

(7) FIG. 6: Construction of P. ohmeri shuttle vectors with poLEU2 and poURA3 selection markers.

(8) FIG. 7: pEVE2523: Restriction map of the P. ohmeri poURA3 expression vector pEVE2523, with a cloned expression cassette containing the open reading frame of tagatose-3-epimerase of Pseudomonas cichorii flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator.

(9) FIG. 8: pEVE2560: Restriction map of the P. ohmeri poLEU2 expression vector pEVE2560, with a cloned expression cassette containing the open reading frame of tagatose-3-epimerase of Pseudomonas cichorii flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator.

(10) FIG. 9: Construction of a P. ohmeri vector for overexpression of Gluconobacter oxydans NADPH-specific xylitol dehydrogenase.

(11) FIG. 10: pEVE3284: Restriction map of the P. ohmeri pEVE3284 expression vector, with a cloned expression cassette containing the NADPH-specific xylitol dehydrogenase of Gluconobacter oxydans flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator.

(12) FIG. 11: Construction of a P. ohmeri vectors for overexpression of Pichia stipitis NADPH-specific xylitol dehydrogenase.

(13) FIG. 12: pEVE2562/pEVE2564: Restriction map of the P. ohmeri pEVE2562/pEVE2564 expression vectors, with a cloned expression cassette containing the NADPH-specific xylitol dehydrogenase of Pichia stipitis flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator with either a poURA3 or poLEU2 selection marker, respectively.

(14) FIG. 13: Construction of a P. ohmeri vector for overexpression of Pichia stipitis NADH-specific xylitol dehydrogenase.

(15) FIG. 14: pEVE2563: Restriction map of the P. ohmeri pEVE2563 expression vector, with a cloned expression cassette containing the NADH-specific xylitol dehydrogenase of Pichia stipitis flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator.

(16) FIG. 15: Construction of a P. ohmeri vector for overexpression of E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase under the control of the P. ohmeri ribulose reductase (poRR) promoter and terminator using a poURA3 selection marker.

(17) FIG. 16: pEVE2839: Restriction map of the P. ohmeri pEVE2839 expression vector, with a cloned expression cassette containing the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator.

(18) FIG. 17: Construction of a P. ohmeri vector for overexpression of E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator using a poURA3 selection marker.

(19) FIG. 18: pEVE3102: Restriction map of the P. ohmeri pEVE3102 expression vector, with a cloned expression cassette containing the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli flanked by a P. ohmeri phosphoglycerate kinase (poPGK1) promoter and ribulose reductase (poRR) terminator.

(20) FIG. 19: pEVE3123: Restriction map of the P. ohmeri pEVE3123 expression vector, with a cloned expression cassette containing the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli flanked by a P. ohmeri phosphoglycerate kinase (poPGK1) promoter and a transketolase (poTKL) terminator and a poURA3 selection marker.

(21) FIG. 20: Construction of a P. ohmeri vector for overexpression of E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator using a poLEU2 selection marker.

(22) FIG. 21: pEVE3157: Restriction map of the P. ohmeri pEVE3157 expression vector, with a cloned expression cassette containing the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli flanked by a P. ohmeri phosphoglycerate kinase (poPGK1) promoter and a transketolase (poTKL) terminator and a poLEU2 selection marker.

(23) FIG. 22: Construction of a P. ohmeri loxP vector with a poLEU2 selection marker

(24) FIG. 23: pEVE2787: Restriction map of the P. ohmeri pEVE2787 integration vector, with a cloned P. ohmeri LEU2 selection marker under the control of the endogenous promoter and terminator, flanked by two loxP sites.

(25) FIG. 24: 12ABTV4P: Restriction map of the synthesized nat1 gene from Streptomyces noursei flanked by AscI and SphI restriction sites.

(26) FIG. 25: pEVE2798: Restriction map of the P. ohmeri pEVE2798 expression vector, with a cloned nat1 marker under the control of a ribulose reductase (poRR) promoter and an orotidine-5-phosphate decarboxylase (poURA3) terminator.

(27) FIG. 26: Construction of a P. ohmeri loxP vector with a nat1 selection marker.

(28) FIG. 27: pEVE2852: Restriction map of the P. ohmeri pEVE2852 integration vector, with a cloned with a cloned nat1 marker under the control of a ribulose reductase (poRR) promoter and an orotidine-5-phosphate decarboxylase (poURA3) terminator, flanked by two loxP sites.

(29) FIG. 28: pEVE2855: Restriction map of the P. ohmeri pEVE2855 integration vector, with a cloned fragment homologous to the 5 region upstream of the LEU2 open reading frame and a nat1 selection marker flanked by two loxP sites.

(30) FIG. 29: Construction of a P. ohmeri loxP vector for the deletion of the LEU2 open reading frame.

(31) FIG. 30: pEVE2864: Restriction map of the P. ohmeri pEVE2864 integration vector, with a cloned fragment homologous to the 5 region upstream of the LEU2 open reading frame and fragment homologous to the 3 region downstream of the LEU2 open reading frame, and a nat1 selection marker flanked by two loxP sites.

(32) FIG. 31: Construction of a double expression plasmids comprising the NADPH-specific xylitol dehydrogenase of P. stipitis and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli.

(33) FIG. 32: pEVE3318: Restriction map of the P. ohmeri pEVE3318 expression vector, containing the double expression construct of the NADPH-specific xylitol dehydrogenase of P. stipitis and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli.

(34) FIG. 33: pEVE2862: Restriction map of the P. ohmeri pEVE2862 expression vector, containing the P. ohmeri LEU2 marker flanked by a P. ohmeri ribulose reductase (poRR) promoter and an orotidine-5-phosphate decarboxylase (poURA3) terminator.

(35) FIG. 34: Construction of an integrative vector for the genomic expression of the E. coli NAD+-specific D-arabitol 4-oxidoreductase gene and the P. stipitis NADPH-specific xylitol dehydrogenase gene in P. ohmeri.

(36) FIG. 35: pEVE2865: Restriction map of the P. ohmeri pEVE2865 integration vector, containing the P. ohmeri LEU2 marker flanked by two loxP sites.

(37) FIG. 36: pEVE3387: Restriction map of the P. ohmeri pEVE3387 integration vector, containing the double expression construct of the NADPH-specific xylitol dehydrogenase gene of P. stipitis and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli with a P. ohmeri LEU2 selection marker flanked by two loxP sites.

(38) FIG. 37: Construction of double/triple expression plasmids comprising the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli.

(39) FIG. 38: pEVE3322/pEVE3324: Restriction map of the P. ohmeri pEVE3322/pEVE3324 expression vectors, containing either the double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli or the triple expression construct of two NADPH-specific xylitol dehydrogenase genes of G. oxydans and one NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli.

(40) FIG. 39: Construction of an integrative vector for the genomic expression of the E. coli NAD+-specific D-arabitol 4-oxidoreductase gene and the G. oxydans NADPH-specific xylitol dehydrogenase gene in P. ohmeri.

(41) FIG. 40: pEVE3390/pEVE3392: Restriction map of the P. ohmeri pEVE3390/pEVE3392 integration vectors, containing either the double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli or the triple expression construct of two NADPH-specific xylitol dehydrogenase genes of G. oxydans and one NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli with a P. ohmeri LEU2 selection marker flanked by two loxP sites.

(42) FIG. 41: Construction of an integrative vector for the genomic expression of the E. coli NAD+-specific D-arabitol 4-oxidoreductase gene and the G. oxydans NADPH-specific xylitol dehydrogenase gene in P. ohmeri.

(43) FIG. 42: pEVE4390: Restriction map of the P. ohmeri pEVE4390 expression vector, containing the double expression construct of the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli and the NADPH-specific xylitol dehydrogenase gene of G. oxydans with a P. ohmeri LEU2 selection marker flanked by two loxP sites.

(44) FIG. 43: 13AB2EGF: Restriction map of the synthesized NAD+-specific D-arabitol 4-oxidoreductase from R. solanacearum flanked by AscI and SphI restriction sites.

(45) FIG. 44: Construction of an integrative vector for the genomic expression of the R. solanacearum NAD+-specific D-arabitol 4-oxidoreductase gene and the G. oxydans NADPH-specific xylitol dehydrogenase gene in P. ohmeri

(46) FIG. 45: pEVE3898: Restriction map of the P. ohmeri pEVE3898 expression vector, with a cloned expression cassette containing the NAD+-specific D-arabitol 4-oxidoreductase of Ralstonia solanacearum flanked by a P. ohmeri ribulose reductase (poRR) promoter and terminator.

(47) FIG. 46: pEVE4077: Restriction map of the P. ohmeri pEVE4077 expression vector, with a double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of R. solanacearum.

(48) FIG. 47: pEVE4377: Restriction map of the P. ohmeri pEVE4377 integration vector, with a double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of R. solanacearum and the poLEU2 selection marker flanked by two loxP sites.

(49) TABLE-US-00001 SEQUENCE LISTING SEQ ID No Description 1 Sequence encoding NAD.sup.+-specific D-arabitol 4-oxido- reductase from E. coli flanked by AscI and SphI restric- tion sites 2 Amino acid sequence of NAD.sup.+-specific D-arabitol 4-oxido- reductase from E. coli 3 Sequence encoding NAD.sup.+-specific D-arabitol 4-oxido- reductase from E. coli 4 Sequence encoding NADPH-specific xylitol dehydrogenase from Pichia stipitis flanked by HindIII and SacII restric- tion sites 5 Amino acid sequence of NADPH-specific xylitol dehydro- genase from P. stipitis 6 Sequence encoding NADPH-specific xylitol dehydrogenase from Pichia stipitis 7 Sequence encoding NADPH-specific xylitol dehydrogenase from Gluconobacter oxydans flanked by AscI and SphI restriction sites 8 Amino acid sequence of NADPH-specific xylitol dehydro- genase from Gluconobacter oxydans 9 Sequence encoding NADPH-specific xylitol dehydro- genase from Gluconobacter oxydans 10 Sequence encoding tagatose-3-epimerase of Pseudomonas cichorii ST24 11 Amino acid sequence of tagatose-3-epimerase of Pseudomonas cichorii ST24 28 Sequence encoding the nat1 gene of Streptomyces noursei flanked by AscI and SphI restric- tion sites 42 Sequence encoding the NAD+-specific D-arabitol 4-oxidoreductase from R. solanacearum flanked by AscI and SphI restriction sites 43 Amino acid sequence of NAD+-specific D-arabitol 4-oxidoreductase from R. solanacearum

EXAMPLES

Example 1. Choice of a Pichia ohmeri Strain as Preferred Host for Genetic Engineering

(50) As host strain of choice, Pichia ohmeri: is a producer of significant amounts of arabitol from glucose, under high osmotic pressure medium, for example medium containing 10-60% D-glucose, and preferably 25% D-glucose (Normal medium usually contains only 2-3% glucose.) has a redox balance that permits the generation of the cofactors needed.

(51) As an illustration of its performances, the following tables indicate the enzyme activities involved in the arabitol metabolic pathway of Pichia ohmeri (Sophie HUCHETTE Thesis, 1992)

(52) The Hexose Monophosphate Pathway: From Glucose-6-P to D-Ribulose-5-P and D-Xylulose-5-P

(53) The oxidative part of the PPP, also named the Hexose Monophosphate Pathway (HMP), is a NADPH-producing pathway. The two NADP.sup.+-dependent enzymes which are Glucose-6-P dehydrogenase (E.C.1.1.1.49) and 6-P-Gluconate dehydrogenase (E.C.1.1.1.44) participate to the oxidation of 1 mole of glucose-6-P in 1 mole of D-ribulose-5-P and generate 2 moles of NADPH.

(54) TABLE-US-00002 TABLE 1 Hexose Monophosphate Pathway in P. ohmeri ATCC 20209 Enzymes Specific activity U/mg NADP.sup.+ G6P dehydrogenase 1.5 NADP.sup.+ 6PG dehydrogenase 0.55 One unit of enzyme activity was defined as the consumption of 1 mole of NAD(P)H or NAD(P)+ per minute per mL of crude extract. One unit of specific activity was defined as one unit of enzyme activity per mg of proteins in crude extract.

(55) The kinetic parameters of the following enzymes were determined: D-ribulose-5-P 3-epimerase (E.C 5.1.3.1), D-ribose-5-P keto-isomerase (E.C.5.3.1.6), transketolase (E.C.2.2.1.1) and acidic phosphatases (E.C. 3.1.3.2).

(56) TABLE-US-00003 TABLE 2 Kinetic parameters of enzymes using D-Ribulose- 5-P as substrate in P. ohmeri ATCC 20209 Enzymes K.sub.M mM V.sub.M U/mg D-Ribulose-5-P 3-epimerase 6.3 3 D-Ribose-5-P keto-isomerase 0.35 1.8 Acid phosphatase 4.3 0.65 One unit of enzyme activity was defined as the consumption of 1 mole of NAD(P)H or NAD(P)+ per minute per mL of crude extract. One unit of specific activity was defined as one unit of enzyme activity per mg of proteins in crude extract.

(57) TABLE-US-00004 TABLE 3 Kinetic parameters of enzymes using D-Xylulose- 5-P as substrate in P. ohmeri ATCC 20209 Enzymes K.sub.M mM V.sub.M U/mg D-Ribulose-5-P 3-epimerase 6.6 0.7 Transketolase (D-ribose-5-P) 0.2 0.9 Transketolase (Erythrose-4-P) 0.6 1.45 Acid phosphatase 16 0.11 One unit of enzyme activity was defined as the consumption of 1 mole of NAD(P)H or NAD(P)+ per minute per mL of crude extract. One unit of specific activity was defined as one unit of enzyme activity per mg of proteins in crude extract.

(58) In vivo, D-xylulose-5-P, synthesized from the epimerization of D-ribulose-5-P, enters efficiently into the non-oxidative part of the PPP via the transketolization. Consequently, D-xylulose-5-P is not available for its dephosphorylation into D-xylulose.

(59) NADH and NADPH Specific D-Ketopentose-Oxidoreductases

(60) D-Ribulose and D-Xylulose are produced by dephosphorylization of D-Ribulose-5-P and D-Xylulose-5-P.

(61) The Michaelis-Menten constants highlight the affinities of the NADH and NADPH-D-ketopentose-oxidoreductases for each substrate and the corresponding maximum velocities.

(62) TABLE-US-00005 TABLE 4 NADH-specific D-ketopentose-oxidoreductase kinetic parameters of P. ohmeri ATCC 20209 Substrate K.sub.M mM V.sub.M U/mg D-Ribulose 90 1 Ribitol 16 0.16 D-Xylulose 5 0.6 Xylitol 7 0.2 One unit of enzyme activity was defined as the consumption of 1 mole of NAD(P)H or NAD(P)+ per minute per mL of crude extract. One unit of specific activity was defined as one unit of enzyme activity per mg of proteins in crude extract.

(63) NADH-specific D-ketopentose-oxidoreductase, forming ribitol and xylitol respectively from D-ribulose and D-xylulose shows a higher affinity for D-xylulose than D-ribulose. The reverse reaction shows a good affinity for xylitol and ribitol explaining the good growth of the host strain on these two polyols.

(64) TABLE-US-00006 TABLE 5 NADPH-specific D-ketopentose-oxidoreductase kinetic parameters of P. ohmeri ATCC 20209 Substrate K.sub.M mM V.sub.M U/mg D-Ribulose 72 3.4 D-Arabitol 1300 0.8 D-Xylulose 262 1.5 Xylitol 200 0.15 One unit of enzyme activity was defined as the consumption of 1 mole of NAD(P)H or NAD(P)+ per minute per mL of crude extract. One unit of specific activity was defined as one unit of enzyme activity per mg of proteins in crude extract.

(65) NADPH-specific D-ketopentose-oxidoreductase, forming D-arabitol from D-ribulose and forming xylitol from D-xylulose shows a higher affinity for D-ribulose than D-xylulose. The reverse reaction shows a very low affinity for D-arabitol explaining the non-growth of the host strain on this polyol.

(66) The two ketopentose-oxidoreductases from the host strain were characterized as different from the previous enzymes described in Saccharomyces rouxii by Ingram and Wood, 1965 (Journal of Bacteriology, vol. 89, no 5, 1186-1194). Indeed, in Saccharomyces rouxii, no forward reaction was detected on D-ribulose and NADH and a backward reaction was detected on D-arabitol with NADPH.

(67) The Haldane relationship predicts in vivo enzyme kinetic behaviors.

(68) TABLE-US-00007 TABLE 6 Substrat/Product K.sub.eq mM.sup.1 Haldane constants determination: NADH-specificD-Ketopentose-oxidoreductase D-Ribulose/Ribitol 78 D-Xylulose/Xylitol 104 Haldane constants determination: NADPH-specificD-Ketopentose-oxidoreductase D-Ribulose/D-Arabitol 104 D-Xylulose/Xylitol 24

(69) The two enzymes favor the forward reaction (D-ketopentose oxidation) over the backward reaction (pentitol reduction).

(70) The PPP in the host strain is extremely efficient and 2 moles of NADPH are generated from 1 mole of glucose consumed. Consequently, NADPH would be available in excess for both anabolic reactions and maintenance reactions. The host strain must produce D-arabitol from D-ribulose or xylitol from D-xylulose to balance the NADPH/NADP.sup.+ redox couple.

(71) The inhibitory effect of NADP.sup.+ on NADPH-specific D-ketopentose-oxidoreductase has been determined in vitro. The activity is 80% less when NADP.sup.+ is added in excess. Even if this concentration is not compatible with the intracellular NADP.sup.+ concentration, this result gives some overview of the role of the NADPH-specific D-ketopentose oxidoreductase into the balance of the NADPH/NADP.sup.+ redox couple.

(72) The host strain produces only D-arabitol from D-ribulose as D-xylulose is not available because of the entrance of D-xylulose-5-P into the non-oxidative part of the PPP.

(73) The link between the production of D-arabitol and the NADPH/NADP.sup.+ redox balance has been demonstrated in the host strain by evaluating the impact of the overexpression of Glucose-6-P dehydrogenase onto the D-arabitol production. So, the obtained strain harbors a G6PDH activity 1.5 times higher and produces 10% more of D-arabitol compared to the host strain (FR2772788).

Example 2. Pichia ohmeri Codon Usage

(74) The codon usage of P. ohmeri was determined from the available DNA and corresponding amino acid sequence of five P. ohmeri genes: transketolase, glucose-6-phosphate dehydrogenase (FR 2772788), ribulose reductase, beta-isopropylmalate dehydrogenase-LEU2 (Piredda and Gaillardin, Yeast, vol. 10:1601-1612 (1994) and orotidine-5-phosphate decarboxylase-URA3 (Piredda and Gaillardin, 1994, supra).

(75) Every individual gene was divided in nucleotide triplets encoding for a single amino acid. The five genes consisted of a total of 2091 codons.

(76) For each amino acid, the number of every codon present in the five genes was counted, divided by 2091 and multiplied by 1000. This way, the frequency of a specific codon in 1000 codons was estimated.

(77) The preliminary codon usage of P. ohmeri is depicted in Table 7.

(78) All heterologous genes expressed in P. ohmeri, except the xylitol dehydrogenase from P. stipitis, were codon optimized using this table and the Optimizer program (Nucleic Acids Research, 2007, 35, W126-W131).

(79) The obtained sequence was sent for gene synthesis after manual addition of recognition sites for restriction enzymes at the respective 5 and 3 ends of the sequence encoding the enzyme.

(80) TABLE-US-00008 TABLE 7 Codon usage table of P. ohmeri derived from 5 coding sequences (CDS) Pichia ohmeri [gbpln]: 5 CDS's (2091 codons) fields: [triplet] [frequency: per thousand] ([number]) TTT 10.5 (22) TCT 30.6 (64) TAT 7.7 (16) TGT 5.7 (12) TTC 30.1 (63) TCC 23.4 (49) TAC 27.3 (57) TGC 1.4 (3) TTA 5.3 (11) TCA 4.8 (10) TAA 1.4 (3) TGA 0.0 (0) TTG 64.1 (134) TCG 9.6 (20) TAG 1.0 (2) TGG 12.9 (27) CTT 10.5 (22) CCT 12.0 (25) CAT 3.3 (7) CGT 5.7 (12) CTC 12.0 (25) CCC 0.0 (0) CAC 15.8 (33) CGC 1.0 (2) CTA 0.0 (0) CCA 34.0 (71) CAA 12.4 (26) CGA 0.0 (0) CTG 2.9 (6) CCG 0.5 (1) CAG 17.7 (37) CGG 0.5 (1) ATT 27.7 (58) ACT 22.0 (46) AAT 7.7 (16) AGT 1.9 (4) ATC 30.6 (64) ACC 24.4 (51) AAC 29.2 (61) AGC 2.4 (5) ATA 2.4 (5) ACA 3.3 (7) AAA 11.0 (23) AGA 26.3 (55) ATG 14.3 (30) ACG 1.4 (3) AAG 64.1 (134) AGG 0.0 (0) GTT 27.3 (57) GCT 46.9 (98) GAT 23.0 (48) GGT 60.7 (127) GTC 19.1 (40) GCC 27.7 (58) GAC 35.9 (75) GGC 10.5 (22) GTA 1.9 (4) GCA 11.0 (23) GAA 18.7 (39) GGA 12.0 (25) GTG 21.5 (45) GCG 3.3 (7) GAG 46.9 (98) GGG 1.0 (2)

Example 3. Cloning of the E. coli Bacterial NAD+-Specific D-Arabitol 4-Oxidoreductase (D-Xylulose-Forming) Gene

(81) A DNA fragment encoding the NAD.sup.+-specific D-arabitol 4-oxidoreductase altD from E. coli was chemically synthesized (GeneArt Gene Synthesis, Life Technologies, Regensburg, Germany), according to the submitted sequence of SEQ ID NO: 1.

(82) Nucleotides 1441 to 2808 of sequence AF378082.1 (obtained from the NCBI GenBank database) coding for the altD gene were used as template and subjected to codon optimization for use in P. ohmeri ATCC 20209 according to Table 7 of example 2, using the Optimizer program.

(83) At the 5 and 3 ends of the resulting sequence, nucleotides encoding for the recognition sites of the restriction enzymes AscI (GGCGCGCC) and SphI (GCATGC) respectively, were added in order to facilitate further cloning.

(84) Additionally, an adenosine triplet was included in front of the start ATG to account for an adenosine at the 3 position in the Kozak-like sequence of yeasts.

(85) The final sequence (SEQ ID NO: 1) was then submitted for synthesis (GeneArt, Regensburg, Germany).

(86) The synthesized DNA fragment encoding the NAD.sup.+-specific D-arabitol 4-oxidoreductase from E. coli was delivered as 5 g lyophilized plasmid DNA in a pMK-RQ derived vector (12ABYWMP, FIG. 1).

(87) For further sub-cloning the gene was released by restriction cutting with AscI and SphI enzymes (New England Biolabs, Ipswich, Mass.).

Example 4. Mutagenesis and Cloning of the Pichia stipitis NADH and NADPH-Specific Xylitol Dehydrogenase

Cloning of the Pichia stipitis NADH-Specific Xylitol Dehydrogenase Gene

(88) The known nucleotide sequence of the yeast (Pichia stipitis) gene XYL2, encoding xylitol dehydrogenase (Ktter et al., Curr. Genet. 18:493-500 (1990)) was cloned in the plasmidic vector lig 7.78 following the teaching of FR 2 765 589 (see example 4 and FIG. 7 of this patent). The restriction map of the vector is presented in FIG. 2A.

(89) Mutagenesis and Cloning of the Pichia stipitis NADPH-Specific Xylitol Dehydrogenase Gene

(90) A DNA fragment encoding the NADPH-specific xylitol dehydrogenase XYL2 from Pichia stipitis was chemically synthesized (GeneArt Gene Synthesis, Life Technologies, Regensburg, Germany) according to the sequence of SEQ ID NO: 4.

(91) Nucleotides 319 to 1410 of sequence X55392.1 (obtained from the NCBI GenBank database) coding for the XYL2 gene were used as template.

(92) According to the paper from Watanabe et al. (J. Biol. Chem., 2005, 280, 10340-10345), the cofactor preference of the xylitol dehydrogenase could be changed from NADH to NADPH by introducing four published amino acid mutations: D207A/I208R/F209S/N211R (numbering based on P22144 protein sequence obtained from the UniProt database).

(93) Accordingly, the codons encoding for D207, 1208, F209 and N211 were manually replaced by GCT, AGA, TCA and AGA in the corresponding sequence, respectively.

(94) Additionally, nucleotides coding for the recognition sites of the restriction enzymes HindIII (AAGCTT) and SacII (CCGCGG) were manually included at the respective 5 and 3 ends, in order to facilitate further cloning.

(95) Furthermore, an adenosine triplet was included in front of the start ATG to account for an adenosine at the 3 position in the Kozak-like sequence of yeasts. The final sequence (SEQ ID NO: 4) was submitted for synthesis (GeneArt, Regensburg, Germany).

(96) The synthesized DNA fragment encoding the NADPH-specific xylitol dehydrogenase from P. stipitis was delivered as 5 g lyophilized plasmid DNA in a pMA-T derived vector (12AALQTP, FIG. 2B).

Example 5. Mutagenesis and Cloning of the Gluconobacter oxydans NADPH-Specific Xylitol Dehydrogenase Gene

(97) A DNA fragment encoding the NADPH-specific xylitol dehydrogenase Xdh from Gluconobacter oxydans was chemically synthesized (GeneArt Gene Synthesis, Life Technologies, Regensburg, Germany), according to the submitted sequence of SEQ ID NO: 7.

(98) Nucleotides 1063 to 1851 of sequence AB091690.1 (obtained from the NCBI GenBank database) coding for the Xdh gene were used as template and subjected to codon optimization for use in P. ohmeri ATCC 20209 according to Table 7 (Example 2) using the Optimizer program.

(99) Based on the publication by Ehrensberger et al. (Structure, 2006, 14, 567-575), the cofactor specificity of the enzyme could be changed from NADH to NADPH by introducing two published amino acid mutations: D38S/M39R (numbering based on Q8GR61 protein sequence obtained from the UniProt database).

(100) Thus, the codons encoding for D38 and M39 were manually replaced by TCT and AGA in the corresponding sequence, respectively. Additionally, nucleotides encoding for the recognition sites of the restriction enzymes AscI (GGCGCGCC) and SphI (GCATGC) were manually included at the respective 5 and 3 ends, in order to enable further cloning.

(101) Furthermore, an adenosine triplet was included in front of the start ATG to account for an adenosine at the 3 position in the Kozak-like sequence of yeasts. The final sequence (SEQ ID NO: 7), was submitted for synthesis (GeneArt, Regensburg, Germany).

(102) The synthesized DNA fragment encoding the NADPH-specific xylitol dehydrogenase from Gluconobacter oxydans was delivered as 5 g lyophilized plasmid DNA in a pMA-T derived vector (13AAYSYP, FIG. 3). For further subcloning, the gene was released by restriction cutting with AscI and SphI enzymes (New England Biolabs, Ipswich, Mass.).

Example 6. Construction of a P. ohmeri Vector for Heterologous Gene Expression Using the poURA3 Selection Marker

(103) The cloning of a vector with replaceable: promoter, open reading frame, and terminator elements

(104) was performed by two successive overlap PCRs of three individual fragments (FIG. 4).

(105) The vector was originally planned as an expression model, to test the cloning and the overexpression of the tagatose 3-epimerase gene in the recombinant Pichia ohmeri strain.

(106) As it will be described below, the tagatose 3-epimerase gene has been cloned into specific AscI-SphI restriction sites cassette, allowing the cloning of any gene of interest by using these same sites of insertion.

(107) The cloning was conceived by the following way.

(108) In a first PCR (PCR1), a 490 bp long ribulose reductase promoter fragment of P. ohmeri flanked by SpeI and AscI sites (underlined in primer sequence) was amplified using: primer EV2960:

(109) TABLE-US-00009 (SEQIDNo12) GAACTAGTGGATCCGTAGAAATCTTG
and primer EV2961:

(110) TABLE-US-00010 (SEQIDNo13) CTTTGTTCATTTTGGCGCGCCTTTTAGTTTAATAAGGGTCCGTG

(111) Additionally, at the 5 end of the reverse primer EV2961, a 13 nucleotide long fragment representing the 5 end of the tagatose-3-epimerase gene was added.

(112) This fragment together with the 8 nucleotides of the AscI site and the 10 following nucleotides of 3 end of the ribulose reductase promoter were needed as overlap for fusing the fragment of PCR1 with the fragment of PCR2 described below. Genomic DNA of P. ohmeri ATCC 20209 was used as template.

(113) For this purpose, a freshly streaked out P. ohmeri colony was resuspended in 30 l of 0.2% SDS and heated for 4 min at 95 C. After full speed centrifugation, 0.5 l of the supernatant was used for PCR.

(114) The template was amplified in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(115) The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 25 cycles with 10 sec at 98 C./20 sec at 50 C./15 sec at 72 C., and a final extension step of 10 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(116) In a second PCR (PCR2), a 911 bp long fragment of the tagatose-3-epimerase of Pseudomonas cichorii ST24 flanked by AscI and SphI sites (underlined in primer sequence) was amplified using: primer EV2962:

(117) TABLE-US-00011 (SEQIDNo14) AAACTAAAAGGCGCGCCAAAATGAACAAAGTTGGCATG
and primer EV2963:

(118) TABLE-US-00012 (SEQIDNo15) TTCTCTTCGAGAGCATGCTCAGGCCAGCTTGTCACG.

(119) The 5 end of primer EV2962 contains a 9 nucleotide long fragment representing the 3 of the ribulose reductase promoter.

(120) This fragment together with the 8 nucleotides of the AscI site and the following 12 nucleotides of the tagatose-3-epimerase open reading frame, is used for the overlap PCR to fuse the PCR2 product to the previously described PCR1 product.

(121) Additionally, the 5 end of reverse primer EV2963 contains a 12 nucleotide long fragment representing the 5 end of the ribulose reductase terminator of P. ohmeri.

(122) This fragment, together with the 6 nucleotides of the SphI site and the following 12 nucleotides of the 3 end of the tagatose-3-epimerase open reading frame, is needed as overlap for fusing PCR2 with the PCR fragment of PCR3 described below.

(123) As template 25 ng of vector 12AAMCJP (FIG. 5) (GeneArt, Regensburg, Germany) containing a synthesized copy of the tagatose-3-epimarease gene of Pseudomonas cichorii ST24 was used (nucleotide 719 to 1591 of AB000361.1, from the NCBI GenBank database)SEQ ID No: 11.

(124) The template was amplified in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(125) The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 25 cycles with 10 sec at 98 C./20 sec at 48 C./30 sec at 72 C., and a final extension step of 10 minutes at 72 C.

(126) In a third PCR (PCR3), a 380 bp long fragment of the ribulose reductase terminator of P. ohmeri flanked by SphI and SacII sites (underlined in primer sequence) was amplified using: primer EV2964

(127) TABLE-US-00013 (SEQIDNo16) AAGCTGGCCTGAGCATGCTCTCGAAGAGAATCTAG
and primer EV2965

(128) TABLE-US-00014 (SEQIDNo17) GTTCCGCGGAGAATGACACGGCCGAC

(129) The 5 end of primer EV2964 contains a 12 nucleotide long fragment of the 3 end of the tagatose-3-epimerase open reading frame that, together with the 6 nucleotides of the SphI site and the following 12 nucleotides of the ribulose reductase terminator of P. ohmeri is used for the fusion of PCR3 to the previously described PCR2.

(130) Genomic DNA of P. ohmeri ATCC 20209 was used as template. After full speed centrifugation, 0.5 l of the supernatant was used in PCR. For this purpose, a freshly streaked out P. ohmeri colony was resuspended in 30 l of 0.2% SDS and heated for 4 min at 95 C.

(131) The template was amplified in a reaction mix consisting of 200 M of each dNTP, 0.5 M of each primer and 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(132) The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 25 cycles with 10 sec at 98 C./20 sec at 50 C./15 sec at 72 C., and a final extension step of 10 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(133) Fusion of the three individual PCR fragments was performed as follows: 50 ng of each gel purified product of PCR1 and PCR2 was used as template in a PCR reaction with EV2960 and EV2963.

(134) A 30 nucleotide long homologous segment in the two fragments, resulting from the primer design described above, was used as overlap in the fusion reaction.

(135) This way, a 1.4 kb long fragment, consisting of a ribulose reductase promoter of P. ohmeri flanked by SpeI and AscI sites was fused to the open reading frame of the tagatose-3-epimerase of Pseudomonas cichorii ST24.

(136) The templates were amplified in a reaction mix consisting of 200 M of each dNTP, 0.5 M of each primer and 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(137) The PCR was performed with an initial denaturation step of 30 sec at 98 C., followed by 30 cycles with 10 sec at 98 C./20 sec at 62 C./45 sec at 72 C., and a final extension step of 10 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(138) The purified fragment was fused in a second overlap PCR to the product of PCR3. 40 ng of each fragment was used as template and amplified with EV2960 and EV2965.

(139) A 30 nucleotide long homologous segment in the two fragments, resulting from the primer design described above, was used as overlap in the fusion.

(140) This way, a 1.8 kb long fragment, consisting of a ribulose reductase promoter of P. ohmeri flanked by SpeI and AscI and the open reading frame of the tagatose-3-epimerase of Pseudomonas cichorii ST24 flanked by AscI and SphI sites was fused to the ribulose reductase terminator of P. ohmeri.

(141) The templates were amplified in a reaction mix consisting of 200 M of each dNTP, 0.5 M of each primer and 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(142) The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./20 sec at 65 C./55 sec at 72 C., and a final extension step of 10 minutes at 72 C. The PCR product was separated on an agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(143) The final PCR product consisting of a 1.7 kb long fragment of the tagatose-3-epimerase of Pseudomonas cichorii ST24 flanked by a ribulose reductase promoter and terminator was digested with restriction enzymes SpeI and SacII (New England Biolabs, Ipswich, Mass.), gel purified and ligated overnight at 16 C. with a 9.8 kb long isolated SpeI/SacII fragment of a lig7.78 vector backbone using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 6).

(144) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.). The purified plasmid DNA was used for further characterization by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(145) The newly cloned expression plasmid pEVE2523 (FIG. 7) is a shuttle E. coli-P. ohmeri vector consisting of a bacterial (E. coli) origin of replication and an ampicillin resistance gene, the yeast (P. ohmeri) autonomous replication sequence, and the poURA3 (P. ohmeri) gene for selection in yeast.

(146) Moreover, it contains an exchangeable P. ohmeri ribulose reductase promoter element (via SpeI and AscI restriction) and terminator element (via SphI and SacII) flanking an open reading frame of the tagatose-3-epimerase of Pseudomonas cichorii (exchangeable via AscI and SphI restriction).

Example 7. Construction of a P. ohmeri Vector for Heterologous Gene Expression Using the poLEU2 Selection Marker

(147) For the construction of a second P. ohmeri expression vector, the expression cassette of plasmid pEVE2523 (FIG. 7) described previously in Example 6 was cloned into a vector containing the P. ohmeri poLEU2 selection marker (FIG. 6).

(148) A blunted 1.7 kb fragment of vector pEVE2523 (FIG. 7) cut with SpeI and SacII (New England Biolabs, Ipswich, Mass.) was used as insert. Blunting was performed with the Blunting Enzyme Mix (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C.

(149) The vector backbone was obtained from a poARS vector (plig3FR 2772788) linearized with SalI (New England Biolabs, Ipswich, Mass.), blunted and dephosphorylated for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). Gel purified insert and vector backbone using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) were ligated for 1 h at RT using T4 DNA ligase (New England Biolabs, Ipswich, Mass.).

(150) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and used for further characterization by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(151) The new cloned expression plasmids pEVE2560 (FIG. 8) is a shuttle E. coli-P. ohmeri vector containing a bacterial (E. coli) origin of replication and an ampicillin resistance gene, the yeast (P. ohmeri) autonomous replication sequence, and the poLEU2 (P. ohmeri) gene for selection in yeast.

(152) Moreover, the open reading frame of the tagatose-3-epimerase of Pseudomonas cichorii flanked by a P. ohmeri ribulose reductase promoter and terminator is exchangeable via AscI and SphI restriction.

Example 8. Construction of a P. ohmeri Vector for Overexpression of Gluconobacter oxydans NADPH-Specific Xylitol Dehydrogenase

(153) A P. ohmeri vector for overexpression of Gluconobacter oxydans NADPH-specific xylitol dehydrogenase was constructed.

(154) For cloning into the expression vector, the DNA fragment encoding the Gluconobacter oxydans NADPH-specific xylitol dehydrogenase was released from vector 13AAYSYP (FIG. 3) by cutting with AscI and SphI restriction enzymes (New England Biolabs, Ipswich, Mass.).

(155) The 803 bp fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and ligated for 2 h at room temperature to the 9.8 kb AscI/SphI-digested and gel-purified vector backbone of pEVE2523 (FIG. 7) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 9).

(156) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(157) The resulting plasmid pEVE3284 (FIG. 10) contains the codon-optimized NADPH-specific xylitol dehydrogenase of Gluconobacter oxydans flanked by a ribulose reductase promoter and terminator of P. ohmeri and the poURA3 selection marker.

Example 9. Construction of a P. ohmeri Vector for Overexpression of Pichia stipitis NADPH-Specific Xylitol Dehydrogenase

(158) For sub-cloning into the expression vector, the DNA fragment encoding the NADPH-specific xylitol dehydrogenase from Pichia stipitis had to be flanked with AscI and SphI restriction sites.

(159) For this purpose: EV3101 primer

(160) TABLE-US-00015 (SEQIDNo18) AAGGCGCGCCAAAATGACTGCTAACCCTTCC containing an AscI site (underlined) and EV3102 primer

(161) TABLE-US-00016 (SEQIDNo19) GAGCATGCTTACTCAGGGCCGTCAATG
containing a SphI (underlined)

(162) were used in a PCR reaction with 30 ng of vector 12AALQTP (FIG. 2B) as template.

(163) The template was amplified in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(164) The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 25 cycles with 10 sec at 98 C./20 sec at 55 C./30 sec at 72 C., and a final extension step of 10 minutes at 72 C.

(165) The 1.1 kb PCR product was separated on a 1% agarose gel, extracted, purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and restriction digested with AscI and SphI (New England Biolabs, Ipswich, Mass.). After column purification with the DNA Clean & Concentrator-5 Kit (Zymo Research Corporation, Irvine, Calif.), it was ligated for 2 h at room temperature to the 10.6 kb AscI/SphI-digested and gel-purified vector backbone of pEVE2523 (FIG. 7) and the 11.8 kb AscI/SphI-digested and gel-purified vector backbone of pEVE2560 (FIG. 8) respectively, using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 11).

(166) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(167) The resulting plasmids pEVE2562 and pEVE2564 (FIG. 12) contain the codon optimised NADPH-specific xylitol dehydrogenase of Pichia stipitis flanked by a ribulose reductase promoter and terminator of P. ohmeri and either the poURA3 or poLEU2 selection marker, respectively.

Example 10. Construction of a P. ohmeri Vector for Overexpression of Pichia stipitis NADH-Specific Xylitol Dehydrogenase

(168) For sub-cloning into the expression vector, the DNA fragment encoding the NADH-specific xylitol dehydrogenase from Pichia stipitis had to be flanked with AscI and SphI restriction sites.

(169) For this purpose: EV3101

(170) TABLE-US-00017 (SEQIDNo18) (AAGGCGCGCCAAAATGACTGCTAACCCTTCC) containing an AscI site (underlined) and EV3102

(171) TABLE-US-00018 (SEQIDNo19) (GAGCATGCTTACTCAGGGCCGTCAATG) containing a SphI (underlined)

(172) were used in a PCR reaction with 30 ng of vector lig7.78 (FIG. 2A) as template.

(173) The template was amplified in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(174) The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 25 cycles with 10 sec at 98 C./20 sec at 55 C./30 sec at 72 C., and a final extension step of 10 minutes at 72 C.

(175) The 1.1 kb PCR product was separated on a 1% agarose gel, extracted, purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and restriction digested with AscI and SphI (New England Biolabs, Ipswich, Mass.).

(176) After column purification with the DNA Clean & Concentrator-5 Kit (Zymo Research Corporation, Irvine, Calif.), it was ligated for 2 h at room temperature to the 10.5 kb AscI/SphI-digested and gel-purified vector backbone of pEVE2560 (FIG. 8) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 13).

(177) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(178) The resulting plasmid pEVE2563 (FIG. 14) contains the codon optimised NADH-specific xylitol dehydrogenase of Pichia stipitis flanked by a ribulose reductase promoter and terminator of P. ohmeri and the poLEU2 selection marker.

Example 11. Construction of P. ohmeri Vectors for Overexpression of E. coli NAD+-Specific D-Arabitol 4-Oxidoreductase

(179) A P. ohmeri vector for overexpression of E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase was constructed.

(180) For cloning into the expression vector, the DNA fragment encoding the codon-optimised E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase was released from vector 12ABYWMP (FIG. 1) by cutting with AscI and SphI restriction enzymes (New England Biolabs, Ipswich, Mass.).

(181) The 1.4 kb fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and ligated for 2 h at room temperature to the 9.8 kb AscI/SphI-digested and gel-purified vector backbone of pEVE2523 (FIG. 7) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 15).

(182) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(183) The resulting plasmid pEVE2839 (FIG. 16) contains the codon-optimised E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase flanked by a ribulose reductase promoter and terminator of P. ohmeri and the poURA3 selection marker.

(184) In addition to the P. ohmeri ribulose reductase promoter, the NAD.sup.+-specific D-arabitol 4-oxidoreductase from E. coli was also cloned under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator.

(185) Cloning was performed in two consecutive steps, by first replacing the ribulose reductase promoter by the poPGK1 promoter, followed by an exchange of the ribulose reductase terminator for the poTKL terminator.

(186) A 611 bp long fragment of the P. ohmeri poPGK1 promoter was amplified from genomic DNA of P. ohmeri using: primer EV3177

(187) TABLE-US-00019 (SEQIDNo20) (GAAGACTAGTTCACGTGATCTC)
containing a SpeI site (underlined) and primer EV3178

(188) TABLE-US-00020 (SEQIDNo21) (CACTGGCGCGCCTTTTGTGTGGTGGTGTCC),
containing an AscI site (underlined).

(189) The genomic DNA template was prepared by resuspending a freshly streaked out P. ohmeri colony in 30 l of 0.2% SDS and heating for 4 min at 95 C. After full speed centrifugation, 0.5 l of the supernatant was used for PCR.

(190) Amplification was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(191) The PCR was accomplished with an initial denaturation step of 2 min at 96 C. followed by 25 cycles with 10 sec at 96 C./10 sec at 58 C./30 sec at 72 C., and a final extension step of 2 minutes at 72 C.

(192) The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(193) The amplified 610 bp long poPGK1 promoter fragment was restriction digested with SpeI and AscI (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 11.5 kb SpeI/AscI-digested and gel-purified vector backbone of pEVE2839 (FIG. 16) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 17).

(194) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(195) The resulting plasmid pEVE3102 (FIG. 18) contains the codon-optimised E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase flanked by a phosphoglycerate kinase (poPGK1) promoter and ribulose reductase terminator of P. ohmeri and the poURA3 selection marker. In the next step the ribulose reductase terminator of pEVE3102 was exchanged for the tranketolase (poTKL) terminator of P. ohmeri.

(196) A 213 bp long fragment of the P. ohmeri poTKL terminator was amplified from genomic DNA of P. ohmeri using: primer EV3817

(197) TABLE-US-00021 (SEQIDNo22) (TAGCAGCATGCATAGGTTAGTGAATGAGGTATG)
containing a SphI site (underlined) and primer EV3818

(198) TABLE-US-00022 (SEQIDNo23) (TAGGTCCGCGGGAGCTTCGTTAAAGGGC)
containing a SacII site (underlined).

(199) The genomic DNA template was prepared as described above.

(200) Amplification was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer.

(201) The PCR was accomplished with an initial denaturation step of 2 min at 96 C. followed by 25 cycles with 10 sec at 96 C./10 sec at 57 C./30 sec at 72 C., and a final extension step of 2 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(202) The amplified 213 bp long poTKL terminator fragment was restriction digested with SphI and SacII (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 11.5 kb SphI/SacII-digested and gel-purified vector backbone of pEVE3102 (FIG. 18) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 17).

(203) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(204) The resulting plasmid pEVE3123 (FIG. 19) contains the codon-optimised E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase flanked by a phosphoglycerate kinase (poPGK1) promoter and a transketolase (poTKL) terminator of P. ohmeri and the poURA3 selection marker.

(205) In order to be able to express the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli from a plasmid using another selection, the poURA3 marker of pEVE3123 was exchanged for the poLEU2 marker.

(206) For this purpose the poURA3 marker was released from vector pEVE3123 (FIG. 19) by restriction digestion with PsiI and AfeI (New England Biolabs, Ipswich, Mass.).

(207) The 9.1 kb vector backbone was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.), blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. and dephosphorylated for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.).

(208) As insert, a 3 kb blunted and gel-purified fragment of the poLEU2 marker released from vector pEVE2560 (FIG. 8) by AseI and AfeI restriction digestion was used. Ligation of the fragments was performed for 2 h at room temperature using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 20).

(209) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(210) The resulting plasmid pEVE3157 (FIG. 21) contains the codon-optimised E. coli NAD.sup.+-specific D-arabitol 4-oxidoreductase flanked by a phosphoglycerate kinase (poPGK1) promoter and a transketolase (poTKL) terminator of P. ohmeri and the poLEU2 selection marker.

Example 12. Expression of the Plasmidic E. coli NAD+-Specific D-Arabitol 4-Oxidoreductase and of the Plasmidic Pichia Stipitis NADPH-Specific Xylitol Dehydrogenase Gene in Pichia ohmeri Strain ATCC 20209

(211) For the biosynthetic conversion of arabitol into xylitol, the simultaneous expression of the NAD.sup.+-specific E. coli D-arabitol 4-oxidoreductase and the NADP-specific xylitol dehydrogenase of P. stipitis is necessary.

(212) The first enzyme leads to the formation of xylulose and the second ones convert xylulose into xylitol.

(213) P. ohmeri strain SRLU (MATh.sup. leu2 ura3) derived from ATCC 20209 and auxotrophic for leucine and uracil (Piredda and Gaillardin, 1994, supra) was used as host for the construction of a yeast strains secreting xylitol by transformation with plasmids: pEVE2839 (NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli) and pEVE2564 (NADPH-specific xylitol dehydrogenase of P. stipitis) leading to strain EYS2755

(214) Additionally, as a control (following the teaching of WO 94/10325) a strain expressing the NADH-specific wild type xylitol dehydrogenase of P. stipitis was also constructed by transformation with plasmids: pEVE2839 (NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli) and pEVE2563 (NADH-specific xylitol dehydrogenase of P. stipitis) into the SRLU host, leading to strain EYS2962.

(215) As control, strains transformed with the single plasmids: pEVE2839 (NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli), pEVE2563 (NADH-specific xylitol dehydrogenase of P. stipitis), and pEVE2564 (NADPH-specific xylitol dehydrogenase of P. stipitis) leading to EYS2943, EYS2696 and EYS2697 respectively, were also generated.

(216) Yeast transformation was carried out in essential by the spheroplasting method of Green et al. (Green E. D., Hieter, P., and Spencer F. A., chapter 5 in Genome Analysis: A Laboratory Manual, Vol. 3, Cloning Systems, Birren et al. (eds.), Cold Spring Harbor Press, New York, 1999) with the following modifications: Instead of Lyticase, Zymolyase 100T was used for generation of spheroplasts and the incubation with the enzyme was performed at 37 C. until the OD of the cell suspension reached 20-30% of the original OD before Zymolyase treatment.

(217) Briefly, P. ohmeri cells were grown overnight at 30 C. in YPD medium (Yeast extract 1% (w/v), Peptone 2% (w/v), Dextrose 2% (w/v)) to a final OD.sub.600 of 3-5.

(218) 200 OD.sub.600 units were harvested by centrifugation, washed once with water and 1M sorbitol, and resuspended in SCE buffer (1 M sorbitol, 100 mM citric acid trisodium salt dihydrate, 10 mM EDTA) to a final concentration of 70 ODs/ml.

(219) DTT and Zymolase (LuBio Science, Luzern, Switzerland) were added to a final concentration of 10 mM and 0.5 U/OD, respectively and the mixture incubated at 37 C. with slow shaking.

(220) The cell wall digestion was followed by measuring the optical density of the solution diluted in water. When this value dropped to 80% of the original, the digestion was terminated by careful centrifugation and washing with 1 M sorbitol and STC buffer (0.98 M sorbitol, 10 mM Tris pH 7.5, 10 mM CaCl.sub.2).

(221) Speroplasts were carefully resuspended in STC buffer containing 50 g/ml calf-thymus DNA (Calbiochem/VWR, Dietikon, Switzerland) to a final concentration of 200 OD/ml. Aliquots of 100 l were mixed with 100-200 ng of plasmid DNA and incubated for 10 min at room temperature.

(222) 1 ml PEG solution (19.6% PEG 8000 w/v, 10 mM Tris pH 7.5, 10 mM CaCl.sub.2) was added to the suspension, incubated for 10 minutes and pelleted. Spheroplasts were regenerated at 30 C. for 1-2 h in 1 ml of a 1 M sorbitol solution containing 25% YPD and 7 mM CaCl.sub.2.

(223) To the regenerated cells 7 ml of 50 C. warm top agar (0.67% yeast nitrogen base w/o amino acids, 0.13% drop-out powder without leucine/uracil/histidine/tryptophan/methionine, 0.086 w/v of required missing amino acid, 2% glucose, 1 M sorbitol, pH5.8 and 2.5% Noble agar) was added and the mixture was poured evenly onto pre-warmed, sorbitol containing selective plates (0.67% yeast nitrogen base w/o amino acids, 0.13% drop-out powder without leucine/uracil/histidine/tryptophan/methionine, 0.086 w/v of required missing amino acid, 2% glucose, 1 M sorbitol, pH5.8).

(224) Plates were incubated for 3-5 days at 30 C. Transformants were reselected on the appropriate selective plates.

(225) Each generated strain was tested in triplicates for arabitol, xylitol and ribitol production.

(226) For this purpose clones were first grown at 30 C. overnight in seed media (0.67% yeast nitrogen base without amino acids; 0.13% drop-out powder without leucine/uracil/histidine/tryptophan/methionine; 0.086 of required missing amino acid; 5% glucose; pH5.7).

(227) Out of this overnight culture a main culture in production media (0.67% yeast nitrogen base without amino acids; 0.13% drop-out powder without leucine/uracil/histidine/tryptophan/methionine; 0.086 of required missing amino acid; 15% glucose; pH5.7) at a starting OD600 of 0.2 was inoculated.

(228) This culture was grown at 37 C. for 48 hours and the arabitol, xylitol and ribitol concentrations of the supernatants were determined by HPLC/MS using a Aminex HPX-87 column (Bio-Rad, Hercules, Calif.) and a Waters TQ-Detector (Acquity UPLC linked to a triple quadrupol detector, Waters, Milford, Mass.) and isocratic conditions with 100% water as mobile phase.

(229) Polyol titers of all tested strains are depicted in Table 8.

(230) TABLE-US-00023 TABLE 8 Polyol production of P. ohmeri SRLU strains transformed with NADH- and NADPH-specific xylitol dehydrogenase of P. stipitis and/or with NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli (average of triplicates) Strain Arabitol (g/L) Xylitol (g/L) Ribitol (g/L) SRLU 32.9 2.4 nd nd EYS2943 26.4 2.8 2.3 0.1 0.7 1.2 [pEVE2839] EYS2696 36.0 2.7 nd. 0.8 0.1 [pEVE2563] EYS2697 31.1 1.6 nd 6.3 0.1 [pEVE2564] EYS2962 29.6 0.8 7.0 0.3 2.3 0.1 [pEVE2839/pEVE2563] EYS2755 16.4 2.2 19.9 0.8 10.9 0.4 [pEVE2839/pEVE2564] ndnot detected

(231) Use of the NADPH-specific xylitol dehydrogenase of P. stipitis leads to a significant increase in xylitol titers, as compared to the wild type NADH-specific enzyme.

Example 13. Expression of the Plasmidic Gluconobacter oxydans NADPH-Specific Xylitol Dehydrogenase Gene in Pichia ohmeri

(232) In addition to a xylitol producing strain using the NADP-specific xylitol dehydrogenase of P. stipitis a second strain expressing the NADP-specific xylitol dehydrogenase of G. oxydans was engineered.

(233) P. ohmeri strain SRLU (MATh.sup. leu2 ura3) derived from ATCC 20209 and auxotrophic for leucine and uracil (Piredda and Gaillardin, 1994, supra) was used as host for the construction of a yeast strains secreting xylitol by transformation with plasmids pEVE3157 (NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli) and pEVE3284 (NADPH-specific xylitol dehydrogenase of G. oxydans) leading to strain EYS3324.

(234) As control, strains transformed with the single plasmids: pEVE3157 (NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli) and pEVE3284 (NADH-specific xylitol dehydrogenase of G. oxydans), leading to EYS3067 and EYS3323 respectively, were also generated.

(235) The E. coli D-arabitol 4-oxidoreductase used for the construction of the above strains is controlled by poPGK1 promoter in contrast to the poRR promoter used in strains expressing the xylitol dehydrogenase of P. stipitis.

(236) However, to exclude a promoter influence and therefore, to be able to compare polyol levels in strains expressing the xylitol dehydrogenase from G. oxydans with those expressing the corresponding enzyme from P. stipitis, an additional strain has been generated.

(237) This strain EYS2963 was obtained by transforming the SRLU host with pEVE3123 (NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli) and pEVE2564 (NADPH-specific xylitol dehydrogenase of P. stipitis).

(238) Yeast transformation was carried out as described in Example 12. Each generated strain was tested in triplicates for arabitol, xylitol and ribitol production as described in Example 12.

(239) Polyol titers of all tested strains are depicted in Table 9.

(240) TABLE-US-00024 TABLE 9 Polyol production of P. ohmeri SRLU strains transformed with NADPH-specific xylitol dehydrogenase of G. oxydans and/or with NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli (average of triplicates) Strain Arabitol (g/L) Xylitol (g/L) Ribitol (g/L) SRLU 32.9 2.4 nd nd EYS3067 29.0 3.8 1.5 0.3 1.8 0.4 [pEVE3157] EYS3323 32.8 0.6 nd nd [pEVE3284] EYS3324 26.3 1.7 21.1 1.1 1.2 0.1 [pEVE3157/pEVE3284] EYS2963 27.3 2.5 17.7 1.7 13.9 0.7 [pEVE3123/pEVE2564] ndnot detected

(241) Xylitol titers in strains expressing the NADPH-specific xylitol dehydrogenase from G. oxydans (EYS3324) are similar to those of strains expressing the corresponding enzyme from P. stipitis (EYS2963). However, the G. oxydans enzyme leads to much lower ribitol titers, thus showing a higher substrate specificity towards xylulose.

Example 14. Generation of a Mutant P. ohmeri Strain with Increased Arabitol Secretion

(242) A higher arabitol producer mutant has been selected from an UV irradiated suspension of P. ohmeri ATCC 20209.

(243) The UV-irradiation system (Vilber Lourmat, France), was equipped with a microprocessor-controlled RMX-3 W radiometer. P. ohmeri was grown on YPD agar (Dextrose 20 g/L) at 37 C. overnight.

(244) A suspension was prepared to reach 10.sup.6 cfu/mL (OD.sub.620=0.4) and 5 mL were put into a sterile Petri dish. The suspension was irradiated after removing the cover from the dish. The UV wavelength was 254 nm and the irradiation energy was 1.8 10.sup.2 J/cm.sup.2. 90% of mortality of the yeast cells was obtained. After stopping the irradiation and replacing the lid on the dish, the suspension was transferred into a sterile tube located into an iced bath.

(245) 20 mL of YPD liquid medium was inoculated with the mutated suspension and was incubated for 12 hours at 37 C., 250 rpm.

(246) After incubation the mutated culture was diluted with sterile 40% glycerol (V/V). Aliquots were distributed into 5 mL vials and frozen at 80 C.

(247) The screening was based on the osmophilic property of Pichia ohmeri which is able to grow on very high concentrations of Dextrose (up to 600 g/L).

(248) Our goal was to select mutants able to grow faster than the mother strain on YPD agar containing Dextrose 600 g/L or 700 g/L.

(249) Defrosted aliquots were spread on YPD.sub.600 and YPD.sub.700 and the first appearing colonies were selected and tested for the production of arabitol in shake flasks.

(250) The subculture and production medium were made of glucose 50 g/L or 100 g/L respectively, yeast extract 3 g/L, MgSO.sub.4 1 g/L and KH.sub.2PO.sub.4 2 g/L, pH 5.7. The subculture (10 mL in a 100 mL flask) was incubated for 24 h at 37 C., 250 rpm. The production (40 mL in a 500 mL flask) was inoculated by 5 mL of subculture and incubated for 64 hours at 37 C., 250 rpm.

(251) TABLE-US-00025 Glucose g/L 64 h Arabitol g/L 64 h P. ohmeri ATCC 20209 6.0 52.7 P. ohmeri CNCM I-4605 0 58.6

(252) The mutant P. ohmeri strain was selected for its faster consumption of glucose and its higher production of arabitol and was deposited in France on Mar. 7, 2012, with the Collection Nationale de Cultures de Microorganismes [National Collection of Microorganism Cultures] of the Institut Pasteur (CNCM), 25 rue du Docteur Roux, 75724 PARIS Cedex 15, under number I-4605.

Example 15. Construction of a LEU2 Deletion Plasmid

(253) In order to be able to use the newly generated CNCM I-4605 strain for plasmid selection and gene integrations, a plasmid for the deletion of the LEU2 open reading frame was constructed.

(254) In a first step, a general integration vector that can be used in P. ohmeri was adapted from the S. cerevisiae CRE/loxP system. The vector backbone was isolated from pUG73 (Gueldener et al., 2002, Nucleic Acid Res, 30, e23) by restriction cutting with PstI and EcoRV enzymes (New England Biolabs, Ipswich, Mass.).

(255) As insert served a PCR fragment containing a LEU2 selection marker of P. ohmeri flanked by loxP sites, generated with primer pair: EV3043

(256) TABLE-US-00026 (SEQIDNo24) (CACTGGCGCGCCCACTGCATGCGTCGACAACCCTTAATATAACTTCGTA TAATGTATGCTATACGAAGTTATTAGGTCTAGACACATCGTGGATCCAAG CTATCAACGAGAGAGTC)
and EV3044

(257) TABLE-US-00027 (SEQIDNo25) (AGTGGCTAGCAGTGCCATGGCCTAATAACTTCGTATAGCATACATTATA CGAAGTTATATTAAGGGTTCTCGAGACGCGTCATCTAGCATCTCATCTAC CAACTC)
and poARS (plig3FR 2772788see FIG. 6) as template.

(258) The forward primer EV3043 contains an AscI (underlined) site preceding a SphI site (underlined), followed by a 48 bp long loxP fragment (bold) and a DraIII site (underlined). The 3 end of EV3043 contains an additional a 25 bp long fragment for amplification of the P. ohmeri LEU2 gene. The reversed primer EV3044 on the other hand, contains a NheI (underlined) site preceding a NcoI site (underlined), followed by a 48 bp long loxP fragment (bold) and a MluI site (underlined). The 3 end of EV3044 contains an additional a 25 bp long fragment for amplification of the P. ohmeri LEU2 gene. The template was amplified in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was performed with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 65 C./50 sec at 72 C., and a final extension step of 7 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(259) The amplified fragment was flanked by a PstI and EcoRV site in a second PCR reaction for further subcloning. Amplification was performed with: primer EV3056

(260) TABLE-US-00028 (SEQIDNo26) (CACTCTGCAGCACTGGCGCGCCCACTGCAT)
containing the PstI site (underlined) and primer EV3057

(261) TABLE-US-00029 (SEQIDNo27) (CACTGATATCAGTGGCTAGCAGTGCCATGG)
containing the EcoRV site (underlined)

(262) in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./45 sec at 72 C., and a final extension step of 7 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(263) The amplified 2.5 kb LEU2 marker was restriction digested with PstI and EcoRV enzymes (New England Biolabs, Ipswich, Mass.), gel-purified with the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and ligated for 2 h at room temperature to the 2.4 kb PstI/EcoRV (New England Biolabs, Ipswich, Mass.), gel-purified (Zymoclean Gel DNA Recovery KitZymo Research Corporation, Irvine, Calif.) backbone of vector pUG73 (Gueldener et al., 2002 Nucleic Acid Res, 30, e23) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.)(FIG. 22). After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(264) The resulting plasmid pEVE2787 (FIG. 23) contains the P. ohmeri LEU2 selection marker under the control of the endogenous promoter and terminator, flanked by two loxP sites. Additionally, a AscI and SphI site have been introduced upstream of the first loxP and a NheI and NcoI site downstream of the second loxP, in order to help in cloning of regions homologous to the integration sites in the genome.

(265) The LEU2 marker of the integration vector was then replaced by the nat1 resistance gene of Streptomyces noursei, in a second cloning step, since a deletion of the endogenous LEU2 open reading frame was aimed.

(266) A DNA fragment encoding the nat1 gene of Streptomyces noursei was chemically synthesized by GeneArt Gene Synthesis (Life Technologies, Regensburg, Germany) according to the submitted sequence of SEQ ID No 28.

(267) Nucleotides 204 to 776 of sequence 560706.1 (obtained from the NCBI GenBank database) coding for the nat1 gene were used as template and subjected to codon optimization for use in P. ohmeri ATCC 20209 according to Table 7 (above), using the Optimizer program.

(268) At the 5 and 3 ends of the resulting sequence, nucleotides encoding for the recognition sites of the restriction enzymes AscI (GGCGCGCC) and SphI (GCATGC) respectively, were manually added in the text file, in order to facilitate further cloning. Additionally, an adenosine triplet was included in front of the start ATG to account for an adenosine at the 3 position in the Kozak-like sequence of yeasts.

(269) The final sequence (SEQ ID No 28) was then submitted for synthesis to GeneArt (Regensburg, Germany). The synthesized DNA fragment encoding the nat1 gene was delivered as 5 g lyophilized plasmid DNA in a pMA-T derived vector (12ABTV4P, FIG. 24).

(270) For the cloning of the nat1 gene a vector containing a ribulose reductase (poRR) promoter and terminator was used. The terminator was exchanged by an orotidine-5-phosphate decarboxylase (poURA3) terminator and the nat1 gene was introduced between the promoter and terminator sequences.

(271) For this purpose, the orotidine-5-phosphate decarboxylase (poURA3) terminator was generated by PCR with: primer EV3393

(272) TABLE-US-00030 (SEQIDNo29) (CAAGCATGCGGGAATGATAAGAGACTTTG)
containing a SphI site (underlined) and primer EV3394

(273) TABLE-US-00031 (SEQIDNo30) (GGACCGCGGAAAGGTGAGGAAGTATATGAAC)
containing a SacII site (underlined) and pEVE2523 (FIG. 7) as template.

(274) Amplification was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 59 C./10 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The 239 bp poURA3 terminator was restriction digested with SphI and SacII enzymes (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 11 kb vector backbone of pEVE2681 linearized with SphI and SacII restriction enzymes (New England Biolabs, Ipswich, Mass.) and gel-purified with Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.). After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(275) In a second cloning step, the nat1 gene was released from 12ABTV4P (FIG. 24) by restriction cutting with SphI and AscI enzymes (New England Biolabs, Ipswich, Mass.). Additionally, a blunting of the SphI site with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. was performed in between the SphI and AscI digestion. The 587 bp gel-purified fragment (Zymoclean Gel DNA Recovery KitZymo Research Corporation, Irvine, Calif.) was than ligated to the gel-purified 10.5 kb vector backbone of the vector described above cut with SphI and AscI restriction enzymes (New England Biolabs, Ipswich, Mass.).

(276) Also the SphI site of the vector was blunted for 15 min at room temperature with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.), followed by a heat inactivation step of 10 min at 70 C. before the digestion with AscI was performed. Additionally, the vector was dephosphorylated for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). The ligation was performed for 2 h at room temperature using T4 DNA ligase (New England Biolabs, Ipswich, Mass.).

(277) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(278) The resulting plasmid pEVE2798 (FIG. 25) contains the nat1 drug resistance marker flanked by a P. ohmeri ribulose reductase (poRR) promoter and an orotidine-5-phosphate decarboxylase (poURA3) terminator.

(279) The nat1 expression cassette was used to replace the P. ohmeri LEU2 selection marker in the integrative vector. In order to facilitate further cloning the nat1 cassette had to be flanked with XbaI (underlined in primer EV3643) and MluI (underlined in primer EV3644) sites by PCR with: primer EV3643

(280) TABLE-US-00032 (SEQIDNo31) (CACTTCTAGACACT GGATCCGTAGAAATCTTG)
and primer EV3644

(281) TABLE-US-00033 (SEQIDNo32) (CACTACGCGTAAAGGTGAGGAAGTATATG).

(282) Primer EV3643 contains an additional ClaI site (dotted line) following the XbaI site. pEVE2798 served as template (FIG. 25).

(283) Amplification was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 54 C./25 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The 1.3 kb nat1 expression cassette was restriction digested with MluI and XbaI enzymes (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 2.6 kb vector backbone of pEVE2787 (FIG. 23) linearized with MluI and XbaI enzymes (New England Biolabs, Ipswich, Mass.) and gel-purified with the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 26).

(284) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(285) The resulting plasmid pEVE2852 (FIG. 27) contains the nat1 selection marker under the control of the ribulose reductase (poRR) promoter and orotidine-5-phosphate decarboxylase (poURA3) terminator and flanked by two loxP sites.

(286) The integration plasmid does not contain any P. ohmeri homologous fragments needed for site specific integration into the genome, so far. This sites were attached in the next steps.

(287) The 5 homologous region upstream of the LEU2 open reading frame was amplified from 50 ng poARS vector (FIG. 6) with: primer EV3548

(288) TABLE-US-00034 (SEQIDNo33) (CACTCTGCAGGATCCAAGCTATCAACGAGA)
containing a PstI site (underlined) and primer EV3549

(289) TABLE-US-00035 (SEQIDNo34) (CACTGCATGCGTTGCGGAAAAAACAGCC)
containing a SphI site (underlined).

(290) The PCR was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The amplification was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 61 C./15 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The 567 bp fragment was restriction digested with PstI and SphI enzymes (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 3.9 kb vector backbone of pEVE2852 (FIG. 27) linearized with PstI and SphI restriction enzymes (New England Biolabs, Ipswich, Mass.) and gel-purified with Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 29).

(291) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(292) The resulting plasmid pEVE2855 (FIG. 28) contains a fragment homologous to the 5 region upstream of the LEU2 open reading frame and a nat1 marker flanked by two loxP sites.

(293) The 3 homologous region downstream of the LEU2 open reading frame was amplified from 50 ng poARS vector (FIG. 6) with: primer EV3550

(294) TABLE-US-00036 (SEQIDNo35) (CACTCCATGGAGTAGGTATATAAAAATATAAGAG)
containing a NcoI site (underlined) and primer EV3551

(295) TABLE-US-00037 (SEQIDNo36) (CACTGCTAGCGTCGACAACAGCAACTAG)
containing a NheI site (underlined).

(296) The PCR was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The amplification was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 51 C./25 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The 1.3 kb fragment was restriction digested with NcoI and NheI enzymes (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 4.4 kb vector backbone of pEVE2855 (FIG. 28) linearized with NcoI and NheI restriction enzymes (New England Biolabs, Ipswich, Mass.) and gel-purified with Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 29).

(297) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(298) The resulting final LEU2 deletion plasmid pEVE2864 (FIG. 30) contains a fragment homologous to the 5 region upstream and a fragment homologous to the 3 region downstream of the LEU2 open reading frame and a nat1 marker flanked by two loxP sites.

Example 16. Generation of a Mutant P. ohmeri Strain Auxotrophic for Leucine

(299) Since the generated P. ohmeri CNCM I-4605 strain did not display any auxotrophy so far, a LEU2 open reading frame deletion was performed, so as to be able to use the LEU2 selection marker for gene integrations.

(300) For this purpose plasmid pEVE2864 (FIG. 30) was restriction digested with EcoRV and PstI enzymes (New England Biolabs, Ipswich, Mass.) for 2.5 h at 37 C. and the mixture used to transform the Mut165 strain according to the procedure described in Example 12.

(301) To the regenerated cells, 7 ml of 50 C. warm top agar (1% yeast extract, 2% peptone, 2% glucose, 1 M sorbitol, pH 5.8 and 2.5% Noble agar) with 25 g/ml natamycin was added and the mixture was poured evenly onto pre-warmed, sorbitol containing selection plates (1% yeast extract, 2% peptone, 2% glucose, 1 M sorbitol, pH 5.8 and 2% agar) with 25 g/ml natamycin. Plates were incubated for 4 days at 30 C. Deletion of the LEU2 open reading frame was verified by no growth on selective plates without leucine and confirmed by colony PCR using: primer EV3393

(302) TABLE-US-00038 (SEQIDNo29) (CAAGCATGCGGGAATGATAAGAGACTTTG)
and primer EV3795

(303) TABLE-US-00039 (SEQIDNo37) (CAAGTCGTGGAGATTCTGC).

(304) The 1.6 kb fragment was amplified with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 51 C./25 sec at 72 C., and a final extension step of 5 minutes at 72 C.

(305) The resulting strain contains the full open reading frame deletion of the LEU2 gene in a CNCM I-4605 background and was deposited in France on Feb. 5, 2015, with the Collection Nationale de Cultures de Microorganismes [National Collection of Microorganism Cultures] of the Institut Pasteur (CNCM), 25 rue du Docteur Roux, 75724 PARIS cedex 15, under number I-4955.

Example 17. Construction of a Double Expression Plasmids Comprising the NADPH-Specific Xylitol Dehydrogenase of P. stipitis and the NAD+-Specific D-Arabitol 4-Oxidoreductase of E. coli

(306) In order to be able to express the NADPH-specific xylitol dehydrogenase of P. stipitis and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli in the mutant P. ohmeri strain only auxotrophic for leucine, construction of a double expression plasmid was required.

(307) The expression cassette containing the NADPH-specific xylitol dehydrogenase of P. stipitis was released from pEVE2562 (FIG. 12) by restriction cutting with SpeI and SacII enzymes (New England Biolabs, Ipswich, Mass.). The 1.9 kb fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. The insert was then ligated for 2 h at room temperature to the 12.1 kb SpeI-linearized, blunted, dephosphorylated (1 h at 37 C. using Antarctic phosphataseNew England Biolabs, Ipswich, Mass.) and gel-purified pEVE3157 backbone (FIG. 21) containing the NAD+-specific D-arabitol 4-oxidoreductase of E. coli using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 31).

(308) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(309) The resulting plasmid pEVE3318 (FIG. 32) contains the double expression construct of the NADPH-specific xylitol dehydrogenase of P. stipitis flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR) and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and ribulose reductase (poRR) terminator and the poLEU2 selection marker.

Example 18. Construction of Integrative Vectors for the Expression of the E. coli NAD+-Specific D-Arabitol 4-Oxidoreductase Gene and the P. stipitis NADPH-Specific Xylitol Dehydrogenase Gene in P. ohmeri

(310) The NAD.sup.+-specific D-arabitol 4-oxidoreductase gene of E. coli and the NADPH-specific xylitol dehydrogenase gene of P. stipitis should ultimately become an integral part of the P. ohmeri genome. Therefore, an integrative vector with a LEU2 selection marker had to be constructed, by replacing the nat1 selection marker of pEVE2852 and incorporating the double expression construct of arabitol oxidoreductase and xylitol dehydrogenase.

(311) For this purpose, the P. ohmeri LEU2 open reading frame, flanked by an AscI and SphI sites, was generated by PCR with: primer EV3645

(312) TABLE-US-00040 (SEQIDNo38) (CAAGGCGCGCCAAAATGTCTACCAAAACCATTAC)
and primer EV3646

(313) TABLE-US-00041 (SEQIDNo39) (GGAGCATGCCTACTTTCCCTCAGCCAAG).

(314) Amplification was performed with 50 ng of poARS (FIG. 6) template in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 57 C./20 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The amplified LEU2 open reading frame was subsequently restriction digested with AscI and SphI enzymes (New England Biolabs, Ipswich, Mass.).

(315) Additionally, a blunting of the SphI site with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. was performed in between the SphI and AscI digestion. The 1.1 kb gel-purified fragment was than ligated to the gel-purified 11 kb vector backbone of pEVE2811 cut with SphI and AscI restriction enzymes (New England Biolabs, Ipswich, Mass.). Also the SphI site of the vector was blunted for 15 min at room temperature with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.), followed by a heat inactivation step of 10 min at 70 C. before the digestion with AscI was performed. Additionally, the vector was dephosphorylated for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). The ligation of the LEU2 open reading frame and the vector backbone was performed for 2 h at room temperature using T4 DNA ligase (New England Biolabs, Ipswich, Mass.).

(316) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(317) The resulting plasmid pEVE2862 (FIG. 33) contains the P. ohmeri LEU2 marker flanked by a P. ohmeri ribulose reductase (poRR) promoter and an orotidine-5-phosphate decarboxylase (poURA3) terminator.

(318) Subsequently, the LEU2 marker was amplified by PCR using: primer EV3643

(319) TABLE-US-00042 (SEQIDNo31) (CACTATCGATGGATCCGTAGAAATCTTG)
containing a ClaI site and primer EV3644

(320) TABLE-US-00043 (SEQIDNo32) (CACTACGCGTAAAGGTGAGGAAGTATATG)
containing a MluI site (underline) and pEVE2862 (FIG. 33) as template.

(321) Amplification was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 54 C./30 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The amplified 1.8 kb long LEU2 fragment was restriction digested with ClaI and MluI enzymes (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 2.6 kb ClaI and MluI (New England Biolabs, Ipswich, Mass.) restriction digested and gel-purified vector backbone of pEVE2852 (FIG. 27) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 34).

(322) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(323) The resulting plasmid pEVE2865 (FIG. 35) contains the P. ohmeri LEU2 marker flanked by two loxP sites.

(324) For cloning of the integration vector, pEVE2865 was restriction digested with SalI enzyme (New England Biolabs, Ipswich, Mass.), blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. and dephosphorylated dephosphorylated for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.).

(325) The 4.5 kb gel-purified fragment of the vector backbone was used for ligation. As insert served a double expression construct of the NADPH-specific xylitol dehydrogenase genes of P. stipitis and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli released from pEVE3318 (FIG. 32) by restriction cutting with NdeI and SacII enzymes (New England Biolabs, Ipswich, Mass.).

(326) The 4.4 kb fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C., followed by an additional gel purification. The vector backbone of pEVE2865 and the insert of pEVE3318 were ligated for 2 h at room temperature using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 34).

(327) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(328) The resulting plasmid pEVE3387 (FIG. 36) contains the double expression construct of the NADPH-specific xylitol dehydrogenase gene of P. stipitis flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR) and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator. As selection marker serves a P. ohmeri LEU2 gene flanked by two loxP sites.

Example 19. Construction of a First Generation Integrative P. ohmeri Strain Secreting Xylitol into the Media

(329) The previously described vector was used to randomly integrate the NAD.sup.+-specific D-arabitol 4-oxidoreductase gene of E. coli and the NADPH-specific xylitol dehydrogenase gene of P. stipitis into the genome of P. ohmeri.

(330) For this purpose strain CNCM I-4955 (Example 16) auxotrophic for leucine was transformed with pEVE3387 (FIG. 36) restriction digested with NotI (New England Biolabs, Ipswich, Mass.) for 3 h at 37 C. according to the procedure described in Example 12. Transformants were selected on sorbitol plates without any leucine.

(331) The resulting strain contains the NAD.sup.+-specific D-arabitol 4-oxidoreductase gene of E. coli and the NADPH-specific xylitol dehydrogenase gene of P. stipitis randomly integrated into the P. ohmeri genome and was deposited in France on May 20, 2015, with the Collection Nationale de Cultures de Microorganismes [National Collection of Microorganism Cultures] of the Institut Pasteur (CNCM), 25 rue du Docteur Roux, 75724 Cedex 15, under number I-4982.

Example 20. Construction of a Double/Triple Expression Plasmid Comprising the NADPH-Specific Xylitol Dehydrogenase of G. oxydans and the NAD+-Specific D-Arabitol 4-Oxidoreductase of E. coli

(332) In order to be able to express the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli in the mutant P. ohmeri strain only auxotrophic for leucine, construction of a double expression plasmid was required.

(333) The expression cassette containing the NADPH-specific xylitol dehydrogenase of G. oxydans was released from pEVE3284 (FIG. 10) by restriction cutting with SpeI and SacII enzymes (New England Biolabs, Ipswich, Mass.). The 1.6 kb fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. The vector backbone used consisted of the 12.1 kb SpeI-linearized (New England Biolabs, Ipswich, Mass.) and gel-purified (Zymoclean Gel DNA Recovery KitZymo Research Corporation, Irvine, Calif.) pEVE3157 backbone (FIG. 21) containing the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli.

(334) The backbone has additionally been blunted for 15 min at room temperature with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.), followed by heat inactivation of the enzymes for 10 min at 70 C. and dephosphorylated for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). Ligation was performed for 2 h at room temperature using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 37).

(335) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(336) The resulting plasmids pEVE3322 and pEVE3324 (FIG. 38) contain either the double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR) and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator or the triple expression construct of two NADPH-specific xylitol dehydrogenase genes of G. oxydans flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR) and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator and the poLEU2 selection marker.

Example 21. Construction of Integrative Vectors for the Expression of the E. coli NAD+-Specific D-Arabitol 4-Oxidoreductase Gene and the G. oxydans NADPH-Specific Xylitol Dehydrogenase Gene in P. ohmeri

(337) Besides the integrative vector containing the NADPH-specific xylitol dehydrogenase of P. stipitis and the NAD.sup.+-specific D-arabitol 4-oxidoreductase gene of E. coli also plasmids containing the NADPH-specific xylitol dehydrogenase of G. oxydans were generated.

(338) For this purpose, the double and triple expression cassettes containing either one or two NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli were released from pEVE3322 and pEVE3324 (FIG. 38) respectively, by restriction cutting with NdeI and SacII enzymes (New England Biolabs, Ipswich, Mass.).

(339) The 4.1 kb and 5.7 kb fragments were gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. As vector served the gel-purified (Zymoclean Gel DNA Recovery KitZymo Research Corporation, Irvine, Calif.), 5.7 kb Sail-linearized pEVE2865 (FIG. 35).

(340) The vector backbone has additionally been blunted for 15 min at room temperature with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.), followed by heat inactivation of the enzymes for 10 min at 70 C. and dephosphorylation for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). Ligation of vector and insert was performed for 2 h at room temperature to using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 39).

(341) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(342) The resulting plasmids pEVE3390 and pEVE3392 (FIG. 40) contain the double or triple expression constructs of either one or two NADPH-specific xylitol dehydrogenase genes of G. oxydans flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR) and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator. As selection marker serves a P. ohmeri LEU2 gene flanked by two loxP sites.

Example 22. Construction of Second Generation Integrative Strains Capable of Secreting More than 100 g/L Xylitol

(343) First generation strain CNCM I-4982 containing a randomly integrated copy of the NAD.sup.+-specific D-arabitol 4-oxidoreductase gene of E. coli and the NADPH-specific xylitol dehydrogenase gene of P. stipitis was used to further integrate additional copies of the two heterologous enzymes.

(344) However, in order to be able to integrate above constructs the LEU2 selection marker had to be removed. For this purpose first generation strain CNCM I-4982 was transformed with vector pEVE3163 according to the procedure described in Example 12. The vector pEVE3163 contains the CRE recombinase of bacteriophage P1 (codon optimized according to Table 7) flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR). Removal of the LEU2 selection marker was confirmed by no-growth of clones on plates without leucine.

(345) The resulting strain EYS3842 was transformed with pEVE3390 or pEVE3392 (FIG. 40) restriction digested with NotI (New England Biolabs, Ipswich, Mass.) for 3 h at 37 C. according to the procedure described in Example 12. Transformants were selected on sorbitol plates without any leucine.

(346) Resulting second generation strain EYS3929 contains two NAD.sup.+-specific D-arabitol 4-oxidoreductase genes of E. coli and two NADPH-specific xylitol dehydrogenase genes, one from G. oxydans and a second one from P. stipitis randomly integrated into the genome. Strain EYS3930, on the other hand, contains an additional NADPH-specific xylitol dehydrogenase gene of G. oxydans.

Example 23. Construction of a Further Vector Used for the Integration of Additional Gene Copies of the NAD+-Specific D-Arabitol 4-Oxidoreductase of E. coli and the NADPH-Specific Xylitol Dehydrogenase of G. oxydans

(347) In order to construct a further integration vector, a double expression cassette of the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli and the NADPH-specific xylitol dehydrogenase of G. oxydans was amplified by PCR using: primer EV4904

(348) TABLE-US-00044 (SEQIDNo40) (ATATCCCGGGCACCGTCATCACCGAAACGC)
containing a SmaI site and primer EV4905

(349) TABLE-US-00045 (SEQIDNo41) (ATATCCCGGGCACGACCACGCTGATGAGC)
containing a SmaI site (underline) and

(350) pEVE3321 as template.

(351) Amplification was performed in a reaction mix consisting of 200 M of each dNTP and 0.5 M of each primer with 0.02 U/l of iProof polymerase (BIO-RAD, Hercules, Calif.) in the appropriate 1 buffer. The PCR was accomplished with an initial denaturation step of 30 sec at 98 C. followed by 30 cycles with 10 sec at 98 C./10 sec at 68 C./75 sec at 72 C., and a final extension step of 5 minutes at 72 C. The PCR product was separated on a 1% agarose gel, extracted and purified using the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.).

(352) The amplified 3.9 kb long fragment was restriction digested with SmaI (New England Biolabs, Ipswich, Mass.) and ligated for 2 h at room temperature to the 4.4 kb PvuII (New England Biolabs, Ipswich, Mass.) linearized, Antarctic phosphatase (New England Biolabs, Ipswich, Mass.) dephosphorylated and gel-purified vector backbone of pEVE2865 (FIG. 35) using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 41).

(353) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(354) The resulting plasmids pEVE4390 (FIG. 42) contains the double expression construct of the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli under the control of the P. ohmeri phosphoglycerate kinase (poPGK1) promoter and transketolase (poTKL) terminator and the NADPH-specific xylitol dehydrogenase gene of G. oxydans flanked by a P. ohmeri ribulose reductase promoter and terminator (poRR). As selection marker serves a P. ohmeri LEU2 gene flanked by two loxP sites.

Example 24. Construction of a Vector Used for the Integration of the NADPH-Specific Xylitol Dehydrogenase of G. oxydans and the NAD+-Specific D-Arabitol 4-Oxidoreductase of R. solanacearum

(355) An additional integrative vector for the expression of the NADPH-specific xylitol dehydrogenase of G. oxydans and of the NAD+-specific D-arabitol 4-oxidoreductase of R. solanacearum was constructed as follows: In a first step a double expression vector containing the two above genes was generated. This double expression cassette was the cloned into an integrative loxP vector.

(356) A DNA fragment encoding the NAD+-specific D-arabitol 4-oxidoreductase gene of Ralstonia solanacearum was chemically synthesized by GeneArt Gene Synthesis (Life Technologies, Regensburg, Germany) according to the submitted sequence of sequence SEQ ID No 42.

(357) Nucleotides 2310548 to 2309151 of sequence AL646052.1 (obtained from the NCBI GenBank database) coding for the dalD gene were used as template and subjected to codon optimization for use in P. ohmeri ATCC 20209 according to Table 7 (above), using the Optimizer program. At the 5 and 3 ends of the resulting sequence, nucleotides encoding for the recognition sites of the restriction enzymes AscI (GGCGCGCC) and SphI (GCATGC) respectively, were manually added in the text file, in order to facilitate further cloning. Additionally, an adenosine triplet was included in front of the start ATG to account for an adenosine at the 3 position in the Kozak-like sequence of yeasts.

(358) The final sequence (SEQ ID No 42) was then submitted for synthesis to GeneArt (Regensburg, Germany). The synthesized DNA fragment encoding the dalD gene was delivered as 5 g lyophilized plasmid DNA in a pMA-RQ derived vector (13AB2EGP, FIG. 43).

(359) The 1.4 kb fragment of the D-arabitol 4-oxidoreductase from R. solanacearum was released from vector 13AB2EGP (FIG. 43) by restriction digested with AscI and SphI (New England Biolabs, Ipswich, Mass.) and gel-purified with the Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.). The insert was then ligated with the 11.8 kb backbone of pEVE2560 (FIG. 8) linearized with AscI and SphI (New England Biolabs, Ipswich, Mass.) and gel purified using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 44).

(360) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(361) The resulting plasmid pEVE3898 (FIG. 45) contains the codon-optimised R. solanacearum NAD+-specific D-arabitol 4-oxidoreductase flanked by a ribulose reductase promoter and terminator of P. ohmeri and the poLEU2 selection marker.

(362) In a next step the expression cassette containing the NADPH-specific xylitol dehydrogenase of G. oxydans flanked by a phosphoglycerate kinase promoter (poPGK) and ribulose reductase terminator (poRR) was released from pEVE3960 by restriction digest with SpeI and SacII (New England Biolabs, Ipswich, Mass.). The 1.8 kb fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. As vector served the gel-purified (Zymoclean Gel DNA Recovery KitZymo Research Corporation, Irvine, Calif.), 13.2 kb SalI-linearized pEVE3898. The vector backbone has additionally been blunted for 15 min at room temperature with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.), followed by heat inactivation of the enzymes for 10 min at 70 C. and dephosphorylation for 1 h at 37 C. using Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). Ligation of vector and insert was performed for 2 h at room temperature to using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 44).

(363) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(364) The resulting plasmid pEVE4077 (FIG. 46) contains the double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans flanked by a P. ohmeri phosphoglycerate kinase promoter (poPGK) and a ribulose reductase terminator (poRR) and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of R. solanacearum under the control of the P. ohmeri ribulose reductase promoter and (poRR) terminator and the poLEU2 selection marker.

(365) Finally, the double expression cassette of the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of R. solanacearum was released from pEVE4077 (FIG. 46) by restriction cutting with SapI (New England Biolabs, Ipswich, Mass.). The 5.9 kb fragment was gel-purified using Zymoclean Gel DNA Recovery Kit (Zymo Research Corporation, Irvine, Calif.) and blunted with the Blunting Enzyme Mix kit (New England Biolabs, Ipswich, Mass.) for 15 min at room temperature, followed by heat inactivation of the enzymes for 10 min at 70 C. As vector served the gel-purified (Zymoclean Gel DNA Recovery KitZymo Research Corporation, Irvine, Calif.), 4.4 kb EcoRV-linearized pEVE2865 (FIG. 35), dephosphorylated for 1 h at 37 C. with Antarctic phosphatase (New England Biolabs, Ipswich, Mass.). Ligation of vector and insert was performed for 2 h at room temperature to using T4 DNA ligase (New England Biolabs, Ipswich, Mass.) (FIG. 44).

(366) After transformation of XL10 Gold ultracompetent cells (Agilent Technologies, Santa Clara, Calif.) with the ligation mixture, plasmid DNA was isolated using the Zyppy Plasmid Miniprep Kit (Zymo Research Corporation, Irvine, Calif.) and further characterized by restriction digestion and sequencing (Microsynth, Balgach, Switzerland).

(367) The resulting plasmid pEVE4377 (FIG. 47) contains the double expression construct of the NADPH-specific xylitol dehydrogenase of G. oxydans and the NAD.sup.+-specific D-arabitol 4-oxidoreductase of R. solanacearum and the poLEU2 selection marker flanked by two loxP sites.

Example 25. Construction of Third Generation Integrative Strains with Increased Productivity of Xylitol

(368) The LEU2 marker of second generation strains EYS3929 and EYS3930 (Example 22) was loxed out as described in Example 18 using vector pEVE3163. The resulting strains EYS4118 and EYS4119 were transformed with pEVE4377 (FIG. 47) and pEVE4390 (FIG. 42), respectively. The vectors were restriction digested with NotI (New England Biolabs, Ipswich, Mass.) for 3 h at 37 C. according to the procedure described in Example 12. Transformants were selected on sorbitol plates without any leucine.

(369) Resulting third generation strain EYS4353 contains three NAD.sup.+-specific D-arabitol 4-oxidoreductase genes, two from E. coli and one from R. solanacearum and three NADPH-specific xylitol dehydrogenase genes, two from G. oxydans and one from P. stipitis randomly integrated into the genome.

(370) The second third generation strain, on the other hand, contains three copies of the NAD.sup.+-specific D-arabitol 4-oxidoreductase of E. coli, three copies of the NADPH-specific xylitol dehydrogenase of G. oxydans and one copy from P. stipitis, background and was deposited in France on Mar. 5, 2015, with the Collection Nationale de Cultures de Microorganismes [National Collection of Microorganism Cultures] of the Institut Pasteur (CNCM), 25 rue du Docteur Roux, 75724 PARIS Cedex 15, under number I-4960.

Example 26. Construction of Fourth Generation Integrative Strains

(371) The LEU2 marker of third generation strains CNCM I-4960 (Example 25) was loxed out as described in Example 18 using vector pEVE3163. The resulting strain EYS4955 was transformed with pEVE4377 (FIG. 47) restriction digested with NotI (New England Biolabs, Ipswich, Mass.) for 3 h at 37 C. according to the procedure described in Example 12. Transformants were selected on sorbitol plates without any leucine.

(372) Resulting fourth generation strain contains four NAD.sup.+-specific D-arabitol 4-oxidoreductase genes, three from E. coli and one from R. solanacearum and four NADPH-specific xylitol dehydrogenase genes, three from G. oxydans and one from P. stipitis randomly integrated into the genome and was deposited in France on May 20, 2015, with the Collection Nationale de Cultures de Microorganismes [National Collection of Microorganism Cultures] of the Institut Pasteur (CNCM), 25 rue du Docteur Roux, 75724 PARIS Cedex 15, under number I-4981.

Example 27. Polyol Production with Pichia ohmeri Strains (Synthetic Medium)

(373) The yeast strains CNCM I-4605, CNCM I-4982, CNCM I-4960 & CNCM I-4981 constructed as described above, were fermented according to the following protocol.

(374) The fermentation process is run under Nitrogen-limitation and can be separated into a growth phase and a production phase. During the growth phase the ammonia in the medium is completely consumed to produce biomass, once the biomass formation stops the production phase starts and Polyol levels increase. The platform used for the described fermentation process was a Multifors 2 from INFORS HT, using vessels with a working volume of 1 L. The fermenters were equipped with two Rushton six-blade disc turbines. Air was used for sparging the fermenters.

(375) Temperature, pH, agitation, and aeration rate were controlled throughout the cultivation. The temperature was maintained at 36 C. The pH was kept at 3 by automatic addition of 5 M KOH.

(376) The aeration rate was kept at 1.0 vvm and the initial stirrer speed was set to 300 rpm. In order to prevent the Dissolved Oxygen (DO) to drop below 20% an automatic stirring cascade was employed. The operating conditions used in the fermentation process are summarized in Table 10.

(377) TABLE-US-00046 TABLE 10 Operating conditions for the Polyol production fermentations Parameter Set-point Volume of liquid [L] 1 Temperature [ C.] 36 pH 3 Agitation speed [rpm] Initially 300, then DO setpoint (20%) controlled stirrer cascade Air flow rate [vvm] 1

(378) For inoculation of the fermenters a 1-stage propagation culture was used. The composition of the used propagation culture medium is described in table 11. Propagation cultures were prepared by inoculating 100 ml of medium in a 500-ml shake flask with 4 baffles (indent). The shake flasks were incubated on a shaking table at 30 C. and 150 rpm. The cells were grown for 24 hrs into mid-exponential phase.

(379) TABLE-US-00047 TABLE 11 Propagation culture medium composition. Concentration Raw material [g/L] Glucose monohydrate C.sub.6H.sub.12O.sub.6*H.sub.2O 46 Antifoam Erol 18 1 drop Potassium dihydrogen- KH.sub.2PO.sub.4 6 phosphate Magnesium sulfate MgSO.sub.4*7H.sub.2O 2.4 heptahydrate Ammonium sulfate (NH.sub.4).sub.2SO.sub.4 0.16 Iron(II) ammonium sulfate Fe(SO.sub.4).sub.2(NH.sub.4).sub.2*6H.sub.2O 0.012 hexahydrate Manganese (II) sulfate MnSO.sub.4*H.sub.2O 0.0007 monohydrate Zinc sulfate heptahydrate ZnSO.sub.4*7H.sub.2O 0.00007 Biotine C.sub.10H.sub.16N.sub.2O.sub.3S 0.0004 Sodium phosphate Na.sub.2HPO.sub.4 0.292 Citric acid monohydrate C.sub.6H.sub.8O.sub.7*H.sub.2O 0.835

(380) Prior to inoculation, an amount of the medium in the fermenter equivalent to the amount of inoculum was removed and an aliquot of the propagation culture was used for inoculation of the fermenter to a final volume of 1 L and an OD.sub.600-at-start of ca. 0.2 (CDW ca. 0.03 g/L). The composition of the medium used in the fermenter is described in table 12.

(381) TABLE-US-00048 TABLE 12 Fermentation medium composition. Concentration Raw material [g/L] Glucose monohydrate C.sub.6H.sub.12O.sub.6*H.sub.2O 250 Antifoam Erol 18 0.67 Potassium dihydrogen- KH.sub.2PO.sub.4 6 phosphate Magnesium sulfate MgSO.sub.4*7H.sub.2O 2.4 heptahydrate Ammonium sulfate (NH.sub.4).sub.2SO.sub.4 4 Iron(II) ammonium sulfate Fe(SO.sub.4).sub.2(NH.sub.4).sub.2*6H.sub.2O 0.012 hexahydrate Manganese (II) sulfate MnSO.sub.4*H.sub.2O 0.0007 monohydrate Zinc sulfate heptahydrate ZnSO.sub.4*7H.sub.2O 0.00007 Biotine C.sub.10H.sub.16N.sub.2O.sub.3S 0.0004

(382) Samples were withdrawn in regular intervals and the total fermentation broth was analyzed for Glucose consumption and extracellular Polyol (Xylitol, Arabitol and Ribitol) formation. Furthermore common fermentation metabolites (Glycerol, Acetate, Ethanol, Pyruvate, Malate, Fumarate & Succinate) were determined. The increase in biomass was on one hand followed by OD.sub.600 and on the other hand by cell dry weight (CDW) determination. The above mentioned measurements were used to determine Polyol production, Arabitol or Xylitol yield and productivity; the results are shown in table 13.

(383) TABLE-US-00049 TABLE 13 Polyol production with Pichia ohmeri strains (synthetic medium). CNCM CNCM CNCM CNCM I-4605 I-4982 I-4960 I-4981 Elapsed Fermentation 67 79 146 64 66 Time (EFT) [h] Glucose [g/L] 0 0 0 0 0 Arabitol [g/L] 118 74 0 0 0 Ribitol [g/L] 0 6 2 7 5 Xylitol [g/L] 0 28 60 110 120 Yield Arabitol [%] 52 Yield Xylitol [%] 12 26 44 48 Productivity [g/L/h] 1.76 0.35 0.41 1.71 1.81

(384) Pichia ohmeri CNCM I-4605 produces arabitol only.

(385) Pichia ohmeri CNCM I-4982 produces arabitol, xylitol and ribitol. In this strain one copy of NAD.sup.+-D-arabitol 4-oxidoreductase gene and one copy of NADPH-specific xylitol dehydrogenase gene have been integrated. The modified strain is now able to consume arabitol. Consequently, after total consumption of glucose, arabitol and ribitol are re-consumed by CNCM I-4982 to produce more xylitol.

(386) Pichia ohmeri CNCM I-4960 (third generation) and CNCM I-4981 (fourth generation) produce xylitol and ribitol but no more arabitol. The intracellular conversion of arabitol in xylulose and xylitol is efficient enough to avoid the excretion of arabitol into the broth. The more copies of the genes encoding for the NAD.sup.+-specific D-arabitol oxidoreductase and the NADPH-specific xylitol dehydrogenase have been introduced into P. ohmeri, the higher are the titer, yield and productivity of xylitol.