Production of Xylitol from D-Xylose by a Xylitol Dehydrogenase Gene-Disrupted Mutant of Candida tropicalis

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 2006, p. 4207 4213 Vol. 72, No. 6 0099-2240/06/$08.00 0 doi:10.1128/aem.02699-05 Copyright 2006, American Society for Microbiology. All Rights Reserved. Production of Xylitol from D-Xylose by a Xylitol Dehydrogenase Gene-Disrupted Mutant of Candida tropicalis Byoung Sam Ko, 1 Jinmi Kim, 2 and Jung Hoe Kim 1 * Department of Biological Sciences, Korea Advanced Institute of Science and Technology, 373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, 1 and Department of Microbiology, School of Bioscience and Biotechnology, Chungnam National University, 220 Gung-dong, Yuseong-gu, Daejeon 305-764, 2 Republic of Korea Received 15 November 2005/Accepted 7 April 2006 Xylitol dehydrogenase (XDH) is one of the key enzymes in D-xylose metabolism, catalyzing the oxidation of xylitol to D-xylulose. Two copies of the XYL2 gene encoding XDH in the diploid yeast Candida tropicalis were sequentially disrupted using the Ura-blasting method. The XYL2-disrupted mutant, BSXDH-3, did not grow on a minimal medium containing D-xylose as a sole carbon source. An enzyme assay experiment indicated that BSXDH-3 lost apparently all XDH activity. Xylitol production by BSXDH-3 was evaluated using a xylitol fermentation medium with glucose as a cosubstrate. As glucose was found to be an insufficient cosubstrate, various carbon sources were screened for efficient cofactor regeneration, and glycerol was found to be the best cosubstrate. BSXDH-3 produced xylitol with a volumetric productivity of 3.23 g liter 1 h 1, a specific productivity of 0.76 g g 1 h 1, and a xylitol yield of 98%. This is the first report of gene disruption of C. tropicalis for enhancing the efficiency of xylitol production. The asporogenic, dimorphic, diploid yeast Candida tropicalis is of both academic and commercial interest. The yeast s ability to utilize n-alkanes and fatty acids as carbon sources has been exploited to produce, -dicarboxylic acid, the starting material for the preparation of perfumes, polymers, adhesives, and macrolide antibiotics, by metabolic engineering of peroxisomal -oxidation enzymes (24, 27). Furthermore, C. tropicalis has recently attracted attention because it accumulates xylitol as a result of high xylose-assimilating activity (17, 30). Xylose is a major pentose sugar found in lignocellulose and is the second most abundant natural sugar (20). Unlike prokaryotic microorganisms, which have a xylose isomerase, most xylose-assimilating yeasts, including C. tropicalis, utilize D- xylose via two enzymatic oxidoreductive reactions with xylose reductase (XR) (EC 1.1.1.21) and xylitol dehydrogenase (XDH) (EC 1.1.1.9) (1). XR catalyzes the reduction of D- xylose to xylitol, and XDH catalyzes the oxidation of xylitol to D-xylulose. D-Xylulose is converted to D-xylulose 5-phosphate by xylulose kinase and then enters the pentose phosphate pathway. XDH requires NAD as a cofactor, whereas XR uses NAD(P)H. The overall efficiency of xylose assimilation is therefore coupled with the activities of XR and XDH. XR and XDH in another xylose-utilizing yeast, Pichia stipitis, were encoded by XYL1 and XYL2, respectively (2, 19). XYL2 genes have been cloned from some yeasts and other fungi such as Saccharomyces cerevisiae, Hypocrea jecorina, and Arxula adeninivorans (10, 25, 28). Xylitol, a five-carbon sugar alcohol, is used as a natural sweetener in the food and confectionary industries. It has an anticariogenic effect that inhibits the growth of the tooth-decaying bacterium Streptococcus mutans (21). Its sweetness level * Corresponding author. Mailing address: Department of Biological Sciences, Korea Advanced Institute of Science and Technology, 373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, Republic of Korea. Phone: 82-42-869-2614. Fax: 82-42-869-5614. E-mail: kimjh@kaist.ac.kr. is equal to that of sucrose, and it can replace sucrose on a weight-to-weight basis. When dissolved in water, xylitol has low viscosity and negative heat effects, and it does not require insulin for metabolic regulation. Owing to these benefits, the use of xylitol in the food industry is growing rapidly. Xylitol is produced by natural xylose-assimilating yeasts and fungi such as Pachysolen tannophilus, Candida guilliermondii, Candida parapsilosis, and C. tropicalis (5, 17, 22, 30). In addition, genetically engineered S. cerevisiae expressing the XYL1 gene from P. stipitis was reported to produce xylitol (11). Xylitol can be produced from D-arabitol by Gluconobacter oxydans (26). Although Candida spp. were reported to be the most active and thus potentially the most useful strains, the industrial production of xylitol has yet to be achieved because of the high production costs associated with the substrate D-xylose, an expensive raw material with a low yield of xylitol. Efforts to develop more cost-effective methods of production have included using a cosubstrate, controlling the dissolved oxygen or redox potential, and amplifying XR activity (6, 17, 18). Nonetheless, the main yield-limiting factor of xylitol is its consumption for cell growth and maintenance. Therefore, if the metabolic step from xylitol to D-xylulose could be blocked by disruption of the corresponding XDH gene, and if cosubstrates were supplied for cell growth, the yield of xylitol should reach the theoretical level of 100%. There have been several attempts to apply this strategy to yeast and other fungi. While P. stipitis is not a common xylitol-producing yeast, XDH-defective mutants of P. stipitis derived by random mutagenesis with ethylmethane sulfonate or nitrosoguanidine produced xylitol using glucose or galactose as a cosubstrate (15). A D-xylulokinase-defective mutant of P. stipitis derived by disruption of the xylulose kinase gene also produced xylitol with a low yield (13). Antisense inhibition of XDH in Trichoderma reesei (a synonym of Hypocrea jecorina) reduced XDH activity to 48% and enhanced xylitol productivity (29). However, the mutants described in these previous reports were not efficient in xylitol 4207

4208 KO ET AL. APPL. ENVIRON. MICROBIOL. TABLE 1. Candida tropicalis strains used in this study Strain Genotype a absence of XYL2 Presence or (XYL2/XYL2) b ATCC 20336 URA3/URA3 XYL2/XYL2 / ATCC 20913 ura3/ura3 XYL2/XYL2 / BSXDH-1 ura3/ura3 xyl2 ::HUH/XYL2 / BSXDH-2 ura3/ura3 xyl2 ::hisg/xyl2 / BSXDH-3 ura3/ura3 xyl2 ::hisg/xyl2 ::URA3 / a Abbreviations: XYL2, xylitol dehydrogenase gene; HUH, hisg-ura3-hisg. b, presence of XYL2;, absence of XYL2. production because they were derived from microorganisms with low xylitol-producing ability. To enhance xylitol production, it may be beneficial to block the XDH step of D-xylose processing in a high-xylitol-producing strain of yeast, such as Candida. No previous studies have constructed an XDH-defective Candida sp. to increase xylitol yield, probably owing to the lack of the corresponding gene sequence of the genus. In this study, we disrupted the XYL2 genes of C. tropicalis and then performed xylitol fermentation with the XYL2-disrupted mutant. The productivity and yield of xylitol fermentation by the XYL2-disrupted mutant were remarkably enhanced by screening suitable cosubstrates and optimizing the process. MATERIALS AND METHODS Strains and culture conditions for genetic manipulation. The C. tropicalis strains used in this study are listed in Table 1. C. tropicalis ATCC 20913 (ura3/ ura3), a uracil auxotroph derived from C. tropicalis ATCC 20336 by random mutagenesis, was used as a host strain for transformation (9). The genomic DNA of C. tropicalis ATCC 20336 was used as a source of the URA3 marker gene. Escherichia coli XL10-Gold (Stratagene, CA) was used for gene manipulation. For genetic manipulation with C. tropicalis, YM medium (3 g liter 1 yeast extract, 3 g liter 1 malt extract, 5 g liter 1 Bacto peptone, and 20 g liter 1 glucose), YNB medium (6.7 g liter 1 yeast nitrogen base without amino acids TABLE 2. Primers used in the experiments and 20 g liter 1 glucose), and YNB-5FOA medium (6.7 g liter 1 yeast nitrogen base without amino acids, 20 g liter 1 glucose, 0.1 g liter 1 uracil, 0.1 g liter 1 uridine, and 6.7 g liter 1 5-fluoroorotic acid [5-FOA]) were used. Xylose minimal medium (6.7 g liter 1 yeast nitrogen base without amino acids, 0.1 g liter 1 uracil, 0.1 g liter 1 uridine, and 20 g liter 1 xylose) was used to evaluate the ability to assimilate D-xylose as a sole carbon source. LB medium (5 g liter 1 yeast extract, 10 g liter 1 Bacto tryptone, and 10 g liter 1 glucose) was used to cultivate E. coli. Sequence analysis of XYL2. C. tropicalis ATCC 20913 genomic DNA was prepared using a total DNA extraction kit (General Bio System, Seoul, Korea). The partial XYL2 gene was amplified by PCR using an XHL PCR kit (Bioneer, Daejeon, Korea) and the primers XDH-F1 and XDH-R1, which were designed based on the sequences of the XYL2 gene in P. stipitis (GenBank accession number A16166). The resulting fragment was purified using a PCR product purification kit (Intron, Seoul, Korea) and then inserted into the pgem-t Easy vector (Promega, WI). The partial XYL2 gene was sequenced, and the complete nucleotide sequence was subsequently determined by 5 and 3 rapid amplification of cdna ends (RACE) using a 5 /3 RACE kit (Roche Diagnostics GmbH, Manheim, Germany). For 5 RACE, the gene-specific primers SP1, SP2, and SP3 were used. For 3 RACE, the gene-specific primer SP4 was used. All sequences of primers used in this study are listed in Table 2. Construction of the disruption cassettes for XYL2 loci. A 1.2-kb URA3 gene was amplified by PCR with genomic DNA of C. tropicalis ATCC 20336 using primers Ura3-F and Ura3-R, which were designed based on the URA3 sequence (GenBank accession number AB006207). The URA3 gene was inserted into the pgem-t Easy vector, and the resulting plasmid was designated pgem-ura3. Two 1.1-kb hisg fragments were amplified by PCR with plasmid pcub6 as a template DNA using two sets of primers, HisG-F1 HisG-R1 and HisG-F2 HisG-R2. First, one 1.1-kb hisg fragment was inserted into pgem-ura3 between the BamHI and XbaI sites. The second 1.1-kb hisg fragment was then inserted into the resulting plasmid between the XhoI and BglII sites. In this way, pgem-huh containing the hisg-ura3-hisg cassette was constructed. A 1,205-bp fragment of XYL2 was amplified from the genomic DNA of C. tropicalis ATCC 20913 by PCR with primers XDH-F2 and XDH-R2. The PCR product was inserted into the pgem-t Easy vector to produce pgem-xdh, from which the linear fragment was amplified by PCR with primers FBamHI and RBamHI. The PCR product was digested with BamHI and ligated with the hisg-ura3-hisg cassette digested with BamHI and BglII. The resulting plasmid, pxdh-huh, was digested with PvuII, and the resulting linear DNA, XYL2- hisg-ura3-hisg-xyl2, was used as the first disruption cassette. The second disruption cassette was made from pgem-ura3 template DNA by PCR using Primer Sequence a Restriction site(s) XDH-F1 5 -AATGGTCTTGGGTCACGAATCC XDH-R1 5 -GCTCTGACCAAGTCGTAGGCTTC SP1 5 -TCAAAGTCAATTGGAGCCTTG SP2 5 -TACCAATTTGCACAAATCGTCC SP3 5 -CAAAGATATCAACAACCATGACTC SP4 5 -CCCATGCGACTTTTTGTTCAAGTTG Ura3-F 5 -GGATCCATTCTAGATGATCTGGTTTGGATTGTTGGAG BamHI, XbaI Ura3-R 5 -AGATCTATCTCGAGTCATGAGAACTAAACTAGCAG BglII, XhoI HisG-F1 5 -GCGGATCCTTCCAGTGGTGCATGAACGC BamHI HisG-R1 5 -CCAAGCTTGCTGTTCCAGTCAATCAGGGT HindIII HisG-F2 5 -GCTCTAGATTCCAGTGGTGCATGAACGC XbaI HisG-R2 5 -CCAGATCTGCTGTTCCAGTCAATCAGGGT BglII XDH-F2 5 -CAGCTGAAGAATGTATAAATAGAACCC PvuII XDH-R2 5 -CAGCTGGACCGTCAATCAAACATTTG PvuII FBamHI 5 -GAGGATCCCGATTTGTGCAAATTGGTAACG BamHI RBamHI 5 -GCGGATCCCCGACAGCAGAAACGACACC BamHI XDH-60F 5 -CAACTTGAAAGTTGGCGACCGCGTCGCTATCGAACCAGGTGTGCCTTC CATGGGCCCGACGTCGCATGCTC XDH-60R 5 -CCACCAGCCTTCAAGATCTTGACACCCATGTAGATACATGGCTGAGCA CCACTGCATTCCGGCGGCCGCGAATTCACTAGTG XDH-F3 5 -CCCTAGTCTTTTGCCCGCCATATG XDH-R3 5-ACAAAATCAAACTTTATCTTTTTACTCG a The restriction sites introduced into the primers are underlined. Boldface type in the XDH-60F and XDH-60R sequences indicates segments that anneal to the plasmid pgem-ura3 for amplification of the second disruption cassette.

VOL. 72, 2006 ENHANCED XYLITOL PRODUCTION USING C. TROPICALIS 4209 primers XDH-60F and XDH-60R. This linear DNA contained XYL2-URA3- XYL2. Transformation of yeast. C. tropicalis was transformed using the lithium acetate (LiOAc) method with a slight modification (12): cells grown in YM medium were washed and resuspended in LiOAc solution (0.1 M LiOAc, 10 mm Tris HCl, ph 7.6, and 1 mm EDTA). A mixture of 30 l cell suspension, 50 l transforming DNA, 5 l salmon testis DNA (Sigma), and 400 l polyethylene glycol 8000 solution (50% polyethylene glycol 8000 in LiOAc solution) was incubated at 30 C for 30 min. After heat shock treatment at 42 C for 15 min, the cells were washed with sterile water and spread onto YNB plates. The cells were incubated for 3 days. To allow pop-out of the URA3 marker, Ura cells grown on YNB medium were spread onto YNB-5FOA plates. The URA3 pop-out mutants were selected from among the 5-FOA-resistant colonies using PCR. The second disruption cassette was introduced into a URA3 pop-out mutant, and the resulting cells were spread onto YNB medium. Each genetic modification was confirmed by PCR. Culture conditions for fermentation experiments. Xylitol fermentation experiments for cosubstrate screening were performed in a 250-ml Erlenmeyer flask with 50 ml xylitol fermentation medium at 200 rpm and 30 C. The fermentation medium for xylitol production consisted of 50 g liter 1 D-xylose, 10 g liter 1 cosubstrate, 10 g liter 1 yeast extract, 5 g liter 1 KH 2 PO 4, and 0.2 g liter 1 MgSO 4 7H 2 O. Batch culture was performed in a 2.5-liter jar fermenter (KoBiotech, Incheon, Korea) containing 1 liter of fermentation medium supplemented with 10 g liter 1 glucose and 15 g liter 1 glycerol with agitation at 500 rpm agitation, 1.0 vol vol 1 min 1 aeration, ph 4.0, and 30 C. Assay of XR and XDH. The XDH and XR activities were determined spectrophotometrically by monitoring the change in A 340 upon NAD reduction or NAD(P)H oxidation at 25 C, respectively (3). The cells grown on the xylitol fermentation medium were harvested by centrifugation at 10,000 g for 5 min. After the cells were washed with 50 mm potassium phosphate buffer (ph 7.0), the cells were resuspended in the same buffer and then disrupted by sonication. The cell debris was separated by centrifugation at 10,000 g for 5 min, and the supernatant was then used to measure enzyme activity. The XDH assay mixture contained 25 mm carbonate buffer (ph 9.5), 0.2 mm NAD, 20 mm xylitol, and enzyme solution. The XR assay mixture contained 50 mm potassium phosphate buffer (ph 7.0), 0.2 mm NAD(P)H, 50 mm D-xylose, and enzyme solution. The activity was expressed in units, where 1 U corresponds to the conversion of 1 mol of NAD(H) per min, and was reported as specific activity [units (milligram protein) 1 ]. Each measurement was repeated three times. Analytical methods. The concentrations of D-xylose, xylitol, and various cosubstrates were analyzed by high-pressure liquid chromatography (Waters, MA) using a Sugar-Pak I column (Waters) with water as the mobile phase at a column temperature of 90 C and a flow rate of 0.5 ml min 1. Cell growth was monitored spectrophotometrically at 600 nm. One A 600 was equivalent to 0.474 g cells liter 1. RESULTS Construction of the XYL2-disrupted mutant of C. tropicalis. The Ura-blasting method was used to disrupt a pair of genes, XYL2/XYL2, for XDH of C. tropicalis ATCC 20913 (Fig. 1A). C. tropicalis was transformed to uracil prototrophy with PvuIIdigested pxdh-huh, the first disruption cassette, and 136 Ura transformants were selected from a YNB plate. PCR confirmed that one of two XDH genes in the transformants was disrupted by the specific integration of a 4.1-kb disruption cassette (Fig. 1B). One transformant, designated BSXDH-1, was incubated on a YNB-5FOA plate to allow pop-out of the URA3 marker gene. Among 56 5-FOA-resistant mutants, four were URA3 pop-out mutants; one of these was designated BSXDH-2 (Fig. 1C). BSXDH-2 was transformed with the second disruption cassette and selected for uracil prototrophy. Twenty-three Ura transformants were obtained and tested for specific gene replacement via PCR. The disruption of both of the XYL2 genes was confirmed in the transformants (Fig. 1D). The transformants were incubated on a xylose minimal plate. The parental strain, C. tropicalis ATCC 20913, and BSXDH-2 successfully FIG. 1. Sequential gene disruption of XYL2 in C. tropicalis. (A) Physical maps of the disruption cassettes. (B) PCR confirmation of the specific integration of the first disruption cassette, where lanes 1 and 2 indicate PCR with primers XDH-F3 and HisG-F1 for amplification of 1.5 kb and lanes 3 and 4 indicate PCR with primers XDH-R3 and Ura3-R for amplification of 2.6 kb. Lanes 1 and 3 represent the host strain, C. tropicalis ATCC 20913, lanes 2 and 4 represent BSXDH-1, and lane M represents the markers. (C) PCR confirmation of URA3 marker gene pop-out, where lanes 1 and 2 indicate PCR with primers XDH-F3 and HisG-F1 for amplification of 1.5 kb and lanes 3 and 4 indicate PCR with primers XDH-R3 and Ura3-R for amplification of 2.6 kb. Lanes 1 and 3 represent BSXDH-1, lanes 2 and 4 represent BSXDH-2, and lane M represents the markers. (D) PCR confirmation of the specific integration of the second disruption cassette, where lanes 1 and 2 indicate PCR with primers XDH-R3 and Ura3-R for amplification of 1.6 kb. Lane 1 represents BSXDH-2, lane 2 represents BSXDH-3, and lane M represents the markers.

4210 KO ET AL. APPL. ENVIRON. MICROBIOL. Downloaded from http://aem.asm.org/ FIG. 2. Physiological and enzymatic confirmation of XYL2 disruption. (A) Growth of the parental strain and the XYL2-disrupted mutants on minimal medium with D-xylose as a sole carbon and energy source. Strains are indicated as follows: 1, C. tropicalis ATCC 20913; 2, BSXDH-2; 3, BSXDH-3; and 4 to 8, other transformants of the second disruption. (B) Assay of XDH and XR activities in each strain. Black bars indicate specific activity of XDH, and gray and white bars represent specific activity of XR with NADPH or NADH as the cofactor, respectively. FIG. 3. Xylitol fermentation profiles of the following cultures in a 250-ml flask: (A) C. tropicalis ATCC 20336, (B) BSXDH-1, and (C) BSXDH-3. Glucose was used as a cosubstrate. Symbols:, dry cell weight; F, glucose; E, D-xylose;, xylitol. grew on a xylose minimal plate. However, no transformants grew on the xylose minimal plate after the second transformation (Fig. 2A). This result indicates that XDH is essential for xylose metabolism in C. tropicalis. One transformant was designated BSXDH-3, and the XDH activities of C. tropicalis ATCC 20336, BSXDH-1, and BSXDH-3 grown on xylitol fermentation medium were analyzed (Fig. 2B). The results indicate that BSXDH-3 does not exhibit XDH activity and that the XYL2 gene product accounts for apparently all of the XDH activity of C. tropicalis. Xylitol production by the XYL2-disrupted mutant. The xylitol production by three strains, C. tropicalis ATCC 20336, BSXDH-1, and BSXDH-3, was evaluated in 250-ml flasks containing 50 ml xylitol fermentation medium (Fig. 3). The three strains grew similarly using glucose until the exponential growth phase. However, BSXDH-3 stopped growing when the glucose was depleted, whereas C. tropicalis ATCC 20336 and BSXDH-1 both continued to grow using D-xylose as a carbon source. BSXDH-3 also stopped producing xylitol after glucose depletion. BSXDH-3 converted 12.5 g liter 1 of 50 g liter 1 of D-xylose, and the final concentration of xylitol was 12.5 g liter 1 ; C. tropicalis ATCC 20336 and BSXDH-1 both used all of the D-xylose in the medium, and the final concentrations of xylitol were 33.6 g liter 1 and 38.4 g liter 1, respectively. However, based on the amount of D-xylose consumed, BSXDH-3 yielded 100% xylitol, while C. tropicalis ATCC 20336 and BSXDH-1 yielded 67.2% and 76.8% xylitol, respectively. This result indicates that all of the D-xylose consumed by the XYL2-disrupted mutant was converted to xylitol and that xylitol production was limited by the lack of a cofactor required by XR, because xylitol could not be further metabolized to regenerate the cofactor. The XR on May 14, 2018 by guest

VOL. 72, 2006 ENHANCED XYLITOL PRODUCTION USING C. TROPICALIS 4211 TABLE 3. Xylitol production by the XYL2-disrupted mutant BSXDH-3 in a medium with 50 g liter 1 xylose and various cosubstrates at a concentration of 10 g liter 1 Cosubstrate Dry cell wt Residual xylose Residual cosubstrate activity was apparently the same in the three strains (Fig. 2B), suggesting that the expression of XR is not affected by the disruption of XYL2. In addition, the XR of C. tropicalis required NADPH as a cofactor, rather than NADH, at a ratio of 100:27, and thus, the production of xylitol by C. tropicalis could be limited by NADPH regeneration. Screening for cosubstrates for cofactor regeneration and xylitol production. A cosubstrate was required for NADPH regeneration during xylitol production, and thus, various carbon sources were screened for efficient xylitol production. Xylitol fermentation was performed in a 250-ml flask containing 50 ml of xylitol fermentation medium and various cosubstrates at a concentration of 10 g liter 1 ; the culture was incubated for 3 days (Table 3). Although the cells grew best in a medium containing glucose as a cosubstrate, xylitol production was favorable when glycerol was the cosubstrate. All D-xylose in the medium was converted to xylitol only when glycerol was used FIG. 4. Xylitol fermentation profiles of BSXDH-3 in a 2.5-liter jar fermenter. Glucose and glycerol were used as cosubstrates for initial cell growth and cofactor regeneration, respectively. Symbols:, dry cell weight; F, glucose; E, D-xylose;, glycerol;, xylitol;, xylulose. Xylitol produced Volumetric productivity (g liter 1 h 1 ) as a cosubstrate. Hence, glycerol was selected as the best cosubstrate for xylitol production. Batch culture was conducted in a fully automated 2.5-liter jar fermenter containing 1 liter of xylitol fermentation medium supplemented with 15 g liter 1 glycerol for cofactor (NADPH) regeneration (Fig. 4). The dissolved oxygen level was kept above 20% by increasing the agitation speed from 300 to 500 rpm. The cell concentration reached 11.3 g liter 1 in 11 h using glucose and glycerol and then remained nearly constant during fermentation. After glucose depletion, glycerol began to be assimilated, and xylitol simultaneously began to accumulate. The total xylitol concentration reached 48.6 g liter 1 in 16 h, and the D-xylose was completely consumed. The xylitol yield was 98% on a weight-to-weight basis, and the volumetric productivity and the specific productivity were 3.23 g liter 1 h 1 and 0.76 g g 1 h 1, respectively. DISCUSSION D-Xylose conversion ratio (%) a Glucose 7.8 38.5 0.0 11.8 0.33 23.0 Fructose 5.3 26.6 0.0 23.6 0.33 46.8 Galactose 2.7 23.2 0.0 27.0 0.57 53.6 Mannose 2.9 13.7 0.0 36.3 0.76 72.6 Maltose 2.1 22.0 0.7 28.1 0.29 56.0 Sucrose 4.4 14.1 0.0 36.0 0.75 71.8 Lactose 0.4 47.2 9.8 2.7 0.03 5.6 Cellobiose 1.3. 35.9 0.0 14.6 0.15 28.2 Meliobiose 0.6 45.8 9.5 4.1 0.04 8.4 Sorbitol 1.4 41.3 7.9 8.9 0.09 17.4 Acetic acid 1.1 39.1 0.0 12.3 0.26 21.8 Gluconic acid 0.7 45.8 7.9 4.0 0.04 91.6 Propionic acid 0.4 48.0 9.6 1.7 0.02 4.0 Malic acid 1.8 45.7 9.6 4.1 0.04 8.6 Formic acid 0.7 44.4 8.9 5.4 0.06 11.2 Ethanol 1.3 42.0 3.6 7.9 0.08 16.0 Glycerol 3.5 0.0 0.0 49.0 1.02 100.0 a Ratio of consumed D-xylose to initial D-xylose. Since the development of the integrative DNA transformation system for C. tropicalis using URA3 complementation, several genes in C. tropicalis have been destroyed by sequential gene disruption methods. Most previous studies of gene disruption in this yeast centered on its -oxidation pathway (4, 24). In the present work, we constructed a C. tropicalis mutant lacking XDH activity. Two copies of the XYL2 gene encoding XDH were successively destroyed using the Ura-blasting method, and the disruption was confirmed by PCR with genespecific primers. This report is the first attempt to apply gene disruption techniques to analyze pentose metabolism in C. tropicalis. Glycerol was found to be the best cosubstrate. Glycerol is a by-product of C. tropicalis fermentation in oxygen-limiting conditions. The production of glycerol and ethanol in anaerobic conditions allows the regeneration of cytosolic NAD and contributes to the intracellular redox balance (8). On the other

4212 KO ET AL. APPL. ENVIRON. MICROBIOL. Strain TABLE 4. Comparison of parameters related to xylitol production by various yeasts reported in the literature Fermentation method Dry cell wt Volumetric productivity (g liter 1 h 1 ) Specific productivity (g g 1 h 1 ) Yield (%) Reference or source C. parapsilosis KFCC 10875 Fed batch 32 3.18 0.17 70 16 C. tropicalis IFO 0618 Batch 23 2.94 0.33 63 30 C. tropicalis KFCC 10960 Fed batch 19 3.79 0.31 83 17 C. tropicalis ATCC 20336 Fed batch 66 3.55 0.13 70 14 C. tropicalis BSXDH-3 Batch 11 3.23 0.76 98 This study hand, glycerol can be used as a sole carbon source for yeast under aerobic conditions, whereas the conversion of glucose and other sugars to ethanol is favored under anaerobic conditions (23). Glycerol is transported into the cytoplasm by diffusion and then phosphorylated by glycerol kinase to give glycerol-3- phosphate, which is converted to dihydroxyacetone phosphate by flavin adenine dinucleotide-linked glycerol-3-phosphate dehydrogenase in the inner membrane of mitochondria. Subsequently, dihydroxyacetone phosphate is converted via gluconeogenesis to glucose-6-phosphate, which enters the pentose phosphate pathway, generating NADPH. The additional production of reduced flavin adenine dinucleotide by glycerol-3-phosphate may increase NADPH level in cytoplasm, because glycerol is a more reduced molecule than other carbon sources. Aeration affects the redox balance of xylose-assimilating yeast (6). Oxygen limitation results in increased intracellular NADH levels, which in turn leads to increased xylitol production by C. tropicalis (7, 17). Previous studies used the two-stage batch culture technique, controlling aeration to enhance the productivity and yield of xylitol (16). However, the optimal dissolved oxygen level for maximum xylitol production was determined to be 0.8 to 1.2%, and it is very difficult to control such a low level of dissolved oxygen in practice. In the present study, it was not necessary to control aeration for xylitol production by the XYL2-disrupted mutant. BSXDH-3 exhibited a similar xylitol yield independent of aeration rate. This suggests that the high xylitol yield of the mutant resulted from the breakdown of the xylose metabolic pathway and not from a redox imbalance in the cytoplasm. Xylitol productivity, however, was dependent on aeration, and it appears that the high xylitol productivity was attributable to the rapid supply of NADPH by glycerol consumption in fully aerated conditions, because glycerol assimilation is dependent on oxygen uptake. In addition, by-products such as ethanol and glycerol did not accumulate. Xylulose, however, accumulated to a negligible level ( 1.2 g liter 1 ). Xylulose accumulation may have been the result of the unspecific activities of polyol dehydrogenases such as D-arabinitol dehydrogenase and L-arabinitol dehydrogenase. As mentioned above, several mutant strains have been investigated to improve xylitol production. An XDH-defective mutant of P. stipitis did not grow on D-xylose as a sole carbon source, because its xylose metabolic pathway was blocked. The mutant produced xylitol using galactose as a cosubstrate with a volumetric productivity of 0.42 g liter 1 h 1 and a yield of 100% (15). On the other hand, a D-xylulokinase-defective mutant of P. stipitis produced xylitol from D-xylose, because xylulose phosphorylation was bypassed via the formation of D- arabinitol and D-ribulose (13). The mutant produced xylitol with a volumetric productivity of 0.22 g liter 1 h 1 without a cosubstrate; however, the yield was only 27%, because xylitol was used to both support growth and regenerate the cofactor. The partial inhibition of XDH activity in T. reesei by antisense RNA enhanced xylitol productivity from 0.004 g liter 1 h 1 to 0.017 g liter 1 h 1 (29). Although those previous studies partially enhanced xylitol productivity and yield, they used parental microorganisms with a low capacity for xylitol production, and thus, the resulting xylitol production rates were far below those reported here, where we used an XYL2-disrupted mutant of C. tropicalis. Table 4 compares the xylitol production levels attained in this study to those of previous studies. The result from C. tropicalis KFCC 10960 is the highest reported xylitol productivity obtained in the absence of a cell concentration (17). It has recently been reported that a final xylitol concentration of 193 g liter 1 from 232.5 g liter 1 of xylose could be obtained from C. tropicalis ATCC 20336 in 32 h using fed-batch fermentation (14). However, the cells were grown at a concentration of 67.9 g liter 1 by ph-stat glucose feeding, and thus, the high density of cells achieved high volumetric productivity but a low specific productivity of only 0.13 g xylitol g cells 1 liter 1. The volumetric productivity of 3.23 g liter 1 h 1 obtained in the present study is similar to productivities reported by previous studies. However, the yield of 98% obtained in the present study was not achieved in other studies. In addition, the specific productivity of 0.76 g g 1 liter 1 was far superior to those of previous reports. Rapid regeneration of NADPH via glycerol assimilation in fully aerobic conditions may explain the high specific productivity. The xylitol production by BSXDH-3 obtained in this study was an improvement upon previous reports in many ways. First, the xylitol yield from D-xylose approached the theoretical maximum of 100%. The 98% yield is the best reported yield among xylitol production processes with considerable productivity. Second, the specific productivity of 0.76 g g 1 h 1 is also the highest reported result, while the volumetric productivity of 3.23 g liter 1 h 1 acquired via simple batch fermentation was relatively similar to other results. It can be easily improved by fed-batch fermentation. Third, BSXDH-3 did not require controlled aeration; productivity was maximized in fully aerobic conditions. Finally, this was the first attempt to apply metabolic engineering of the xylose metabolism pathway in C. tropicalis to enhance xylitol yield and productivity. This strategy for strain development will contribute greatly to other attempts to improve the metabolite production of microorganisms.

VOL. 72, 2006 ENHANCED XYLITOL PRODUCTION USING C. TROPICALIS 4213 REFERENCES 1. Alexander, M. A., T. W. Chapman, and T. W. Jeffries. 1988. Xylose metabolism by Candida shehatae in continuous culture. Appl. Microbiol. Biotechnol. 29:478 486. 2. Amore, R., P. Kötter, C. Küster, M. Ciriacy, and C. P. Hollenberg. 1991. Cloning and expression in Saccharomyces cerevisiae of the NAD(P)H-dependent xylose reductase-encoding gene (XYL1) from the xylose-assimilating yeast Pichia stipitis. Gene 109:89 97. 3. Chiang, C., and S. G. Knight. 1966. D-Xylose reductase and D-xylitol dehydrogenase from Penicillium chrysogenum. Methods Enzymol. 9:188 193. 4. Craft, D. L., K. M. Madduri, M. Eshoo, and C. R. Wilson. 2003. Identification and characterization of the CYP52 family of Candida tropicalis ATCC 20336, important for the conversion of fatty acids and alkanes to, dicarboxylic acids. Appl. Environ. Microbiol. 69:5983 5991. 5. Dahiya, J. S. 1991. Xylitol production by Petromyces albertensis grown on medium containing D-xylose. Can. J. Microbiol. 37:14 18. 6. Furlan, S. A., P. Bouilloud, P. Strehaiano, and J. P. Riba. 1991. Study of xylitol formation from xylose under oxygen limiting conditions. Biotechnol. Lett. 13:203 206. 7. Granström, T., A. Aristos, and M. Leisola. 2002. Metabolic flux analysis of Candida tropicalis growing on xylose in an oxygen-limited chemostat. Metab. Eng. 4:248 256. 8. Granström, T., and M. Leisola. 2002. Controlled transient changes reveal differences in metabolite production in two Candida yeasts. Appl. Microbiol. Biotechnol. 58:511 516. 9. Haas, L. O. C., J. M. Cregg, and A. G. Gleeson. 1990. Development of an integrative DNA transformation system for the yeast Candida tropicalis. J. Bacteriol. 172:4571 4577. 10. Habenicht, A., H. Motejadded, M. Kiess, A. Wegere, and R. Mattes. 1999. Xylose utilisation: cloning and characterisation of the xylitol dehydrogenase from Galactocandida mastotermitis. Biol. Chem. 380:1405 1411. 11. Hallborn, J., M. Walfridsson, U. Airaksinen, H. Ojamo, B. Hahn-Hägerdal, M. Penttilä, and S. Keränen. 1991. Xylitol production by recombinant Saccharomyces cerevisiae. Bio/Technology 9:1090 1095. 12. Ito, H., Y. Fukuda, K. Murata, and J. Kimura. 1983. Transformation of intact yeast cells treated with alkali cations. J. Bacteriol. 153:163 168. 13. Jin, Y. S., J. Cruz, and T. M. Jeffries. 2005. Xylitol production by a Pichia stipitis D-xylulokinase mutant. Appl. Microbiol. Biotechnol. 68:42 45. 14. Jo, Y. I., J. H. Seo, and Y. W. Ryu. 2002. Optimization of xylitol production by Candida tropicalis in two-stage fed-batch culture. Korean J. Biotechnol. Bioeng. 17:93 99. 15. Kim, M. S., Y. S. Chung, J. H. Seo, D. H. Jo, Y. H. Park, and Y. W. Ryu. 2001. High-yield production of xylitol from xylose by a xylitol dehydrogenase defective mutant of Pichia stipitis. J. Microbiol. Biotechnol. 11:564 569. 16. Kim, S. Y., J. H. Kim, and D. K. Oh. 1997. Improvement of xylitol production by controlling oxygen supply in Candida parapsilosis. J. Ferment. Bioeng. 83:267 270. 17. Kim, S. Y., J. H. Kim, and D. K. Oh. 1998. Effect of redox potential on xylitol production by Candida tropicalis. Food Sci. Biotechnol. 7:282 285. 18. Kim, Y. S., S. Y. Kim, J. H. Kim, and S. C. Kim. 1999. Xylitol production using recombinant Saccharomyces cerevisiae containing multiple xylose reductase genes at chromosomal delta-sequences. J. Biotechnol. 67:159 171. 19. Kötter, P., R. Amore, C. P. Hollenberg, and M. Ciriacy. 1990. Isolation and characterization of the Pichia stipitis xylitol dehydrogenase gene, XYL2, and construction of a xylose-utilizing Saccharomyces cerevisiae transformant. Curr. Genet. 18:493 500. 20. Ladisch, M. R., K. W. Lin, M. Voloch, and G. T. Tsao. 1983. Process considerations in the enzymatic hydrolysis of biomass. Enzyme Microb. Technol. 5:82 102. 21. Mäkinen, K. K. 1992. Dietary prevention of dental caries by xylitol clinical effectiveness and safety. J. Appl. Nutr. 44:16 28. 22. Morimoto, S., M. Matsuo, K. Azuma, and A. J. Sinskey. 1986. Purification and properties of D-xylose reductase from Pachysolen tannophilus. J. Ferment. Technol. 64:219 225. 23. Nevoigt, E., and U. Stahl. 1997. Osmoregulation and glycerol metabolism in the yeast Saccharomyces cerevisiae. FEMS Microbiol. Rev. 21:231 241. 24. Picataggio, S., T. Rohrer, K. Deanda, D. Lanning, R. Reynolds, J. Mielenz, and L. D. Eirich. 1992. Metabolic engineering of Candida tropicalis for the production of long-chain dicarboxylic acids. Bio/Technology 10:894 898. 25. Richard, P., M. H. Toivari, and M. Penttilä. 1999. Evidence that the gene YLR070c of Saccharomyces cerevisiae encodes a xylitol dehydrogenase. FEBS Lett. 457:135 138. 26. Suzuki, S., M. Sugiyama, Y. Mihara, K. Hashiguchi, and K. Yokozeki. 2002. Novel enzymatic method for the production of xylitol from D-arabitol by Gluconobacter oxydans. Biosci. Biotechnol. Biochem. 66:2614 2620. 27. Torkko, J. M., K. T. Koivuranta, A. J. Kastaniotis, T. T. Airenne, T. Glumoff, M. Ilves, A. Hartig, A. Gurvitz, and J. K. Hiltunen. 2003. Candida tropicalis expresses two mitochondrial 2-enoyl thioester reductases that are able to form both homodimers and heterodimers. J. Biol. Chem. 278:41213 41220. 28. Tran, L. H., N. Kitamoto, K. Kawai, K. Takamizawa, and T. Suzuki. 2004. Cloning and expression of a NAD -dependent xylitol dehydrogenase gene (xyla) ofaspergillus oryzae. J. Biosci. Bioeng. 97:419 422. 29. Wang, T. H., Y. H. Zhong, W. Huang, T. Liu, and Y. W. You. 2005. Antisense inhibition of xylitol dehydrogenase gene, xdh1 from Trichoderma reesei. Lett. Appl. Microbiol. 40:424 429. 30. Yahashi, Y., H. Horitsu, K. Kawai, T. Suzuki, and K. Takamizawa. 1996. Production of xylitol from D-xylose by Candida tropicalis: the effect of D- glucose feeding. J. Ferment. Bioeng. 81:148 152.