Appl Microbiol Biotechnol (2010) 87:1841–1853 DOI 10.1007/s00253-010-2629-9

APPLIED GENETICS AND MOLECULAR BIOTECHNOLOGY

Construction of a β-glucosidase expression system using the multistress-tolerant yeast Issatchenkia orientalis Takao Kitagawa & Kenro Tokuhiro & Hidehiko Sugiyama & Katsuhiro Kohda & Naoto Isono & Makoto Hisamatsu & Haruo Takahashi & Takao Imaeda

Received: 8 January 2010 / Revised: 16 April 2010 / Accepted: 18 April 2010 / Published online: 14 May 2010 # Springer-Verlag 2010

Abstract We demonstrate the value of the thermotolerant yeast Issatchenkia orientalis as a candidate microorganism for bioethanol production from lignocellulosic biomass with the goal of consolidated bioprocessing. The I. orientalis MF-121 strain is acid tolerant, ethanol tolerant, and thermotolerant, and is thus a multistress-tolerant yeast. To express heterologous proteins in I. orientalis, we constructed a transformation system for the MF-121 strain and then isolated the promoters of TDH1 and PGK1, two genes that were found to be strongly expressed during ethanol fermentation. As a result, expression of βglucosidase from Aspergillus aculeatus could be achieved

with I. orientalis, demonstrating successful heterologous gene expression in I. orientalis for the first time. The transformant could convert cellobiose to ethanol under acidic conditions and at high temperature. Simultaneous saccharification and fermentation (SSF) was performed with the transformant, which produced 29 gl−1 of ethanol in 72 h at 40°C even without addition of β-glucosidase when SSF was carried out in medium containing 100 gl−1 of microcrystalline cellulose and a commercial cellulase preparation. These results suggest that using a genetically engineered thermotolerant yeast such as I. orientalis in SSF could lead to cost reduction because less saccharification enzymes are required.

Takao Kitagawa and Kenro Tokuhiro contributed equally to this work.

Keywords Thermotolerant yeast . Issatchenkia orientalis . Microcrystalline cellulose . Ethanol . Beta-glucosidase

T. Kitagawa (*) Department of Applied Molecular Bioscience, Yamaguchi University Graduate School of Medicine, 2-16-1, Tokiwadai, Ube 755-8611, Japan e-mail: [email protected] K. Tokuhiro : K. Kohda : H. Takahashi Biotechnology Laboratory, Toyota Central R&D Labs, Inc., 41-1 Yokomichi, Nagakute, Aichi 480-1192, Japan H. Sugiyama : T. Imaeda Environmental and Applied Biotechnology Laboratory, Toyota Central R&D Labs, Inc., 41-1 Yokomichi, Nagakute, Aichi 480-1192, Japan N. Isono : M. Hisamatsu Laboratory of Food Science and Biotechnology, Department of Sustainable Resource Sciences, Graduate School of Bioresources, Mie University, 1577 Kurima, Tsu, Mie 514-8507, Japan

Introduction Fuel production from biomass is a growing area of research owing to fossil fuel depletion, the need for reduced CO2 emission, and a shift in the energy industry from oil refineries to biorefineries. Bioethanol has been produced from corn and sugar cane, which are also foods. Hence, lignocellulosic biomass is desirable as a feedstock. Production of ethanol from lignocellulosic biomass requires many steps, including biomass pretreatment, saccharification, fermentation, and purification. This multistep process is one of the reasons for the high cost of ethanol production. In order to reduce cost of ethanol produced from lignocellulosic biomass, these processes need to be consolidated. A concept for process consolidation has been suggested as follows: simultaneous saccharification and fermentation (SSF) of hexoses, simultaneous saccharification and cofer-

1842

mentation (SSCF) of hexoses and pentoses, and consolidated bioprocessing (CBP) involving self-production of saccharification enzymes and fermentation of hexoses and pentoses (Lynd et al. 2002; van Zyl et al. 2007). For CBP in the future, thermophilic or thermotolerant microorganisms are essential owing to their many advantages, including prevention of contamination, reduction of cooling costs, and improvement of the reactivity of saccharification enzymes. For example, the cellulase activities of the thermophilic bacterium Clostridium thermocellum and filamentous fungus Trichoderma reesei exhibit a 50% decrease below 40°C (Schwarz et al. 1986). For thermophilic bacteria, genetic engineering or mutagenetic approaches have been used with the goal of CBP. A genetically manipulated strain of the thermophilic bacterium Thermoanaerobacterium saccharolyticum, in which the genes for lactic acid and acetic acid syntheses have been disrupted, facilitates an improvement in ethanol production with the SSF process at high temperature (Shaw et al. 2008). Mutants of the thermophilic bacterium Clostridium thermocellum harboring a cellulosome complex have been isolated as ethanol hyperproducers by means of ultraviolet mutagenesis (Tailliez et al. 1989). On the other hand, budding yeast Saccharomyces cerevisiae is a good ethanol producer, and shows high ethanol tolerance but not thermotolerance. Screening for thermotolerance with a DNA polymerase δ mutation or ultraviolet mutation has been performed for S. cerevisiae, and resulted in mutants that grow at temperatures up to 40–42°C (Sridhar et al. 2002; Rajoka et al. 2005; Shimoda et al. 2006). Some thermotolerant and ethanologic yeasts have been isolated and modified for the production of ethanol from biomass. For example, Hansenula polymorpha produces ethanol with xylose as the carbon source through modification of the pentose phosphate pathway at high temperature (Dmytruk et al. 2008). In addition, recombinant Kluyveromyces marxianus or a mutant strain of Candida glabrata has the ability of flocculation, and allows efficient ethanol fermentation at high temperature (Nonklang et al. 2009; Watanabe et al. 2009). Moreover, a recombinant K. marxianus transformed with cellulase genes, including those encoding β-glucosidase, cellobiohydrolase, and endoglucanase, has been constructed, and its cellulase activity has been confirmed (Hong et al. 2007). Genetic engineering of a thermotolerant yeast is required to achieve CBP from lignocellulosic biomass. However, genetic engineering of a thermotolerant yeast has rarely been reported. Issatchenkia orientalis has been isolated from foods such as cheese (Prillinger et al. 1999), grape wine pomace (Seo et al. 2007), cocoa beans (Daniel et al. 2009), and rice bran (Koh and Suh. 2009). The I. orientalis MF-121 strain was isolated from a river of pH 3 flowing in a hot spring area in Japan, and facilitates ethanol fermentation in medium

Appl Microbiol Biotechnol (2010) 87:1841–1853

containing 5% sodium sulfate at pH 2 (Hisamatsu et al. 2006); thus, it is a multiple stress-tolerant yeast, being ethanol tolerant, salt tolerant, acid tolerant, and thermotolerant. Because sulfuric acid is used for the pretreatment of lignocellulose and produces undesirable by-products such furfural and acetic acid, such multiple stress tolerance is an important factor in ethanol fermentation. Indeed, the MF-121 strain can produce ethanol from saccharification products of hydrolysis of the lignocellulosic biomass by sulfuric acid (Thalagala et al. 2009). However, I. orientalis does not possess any cellulase genes. Consequently, it is necessary to establish a transformation system for I. orientalis. Here, we identified a ura3 auxotrophic mutant and isolated the URA3 gene of the MF-121 strain. As a result, we succeeded in expressing the β-glucosidase gene from Aspergillus aculeatus in MF-121, and the transformant strain succeeded in ethanol fermentation without the addition of β-glucosidase. The fermentation rate of the transformant expressing β-glucosidase on cellobiose was as fast as that on glucose at high temperature. In addition, SSF experiments with Avicel as the carbon source were performed with the transformant expressing the βglucosidase at high temperature. These results suggest that a reduction in the cost of saccharification enzymes with I. orientalis could be achieved through genetic engineering.

Materials and methods Yeast strains and media The I. orientalis MF-121 strain (FERM number P-19368) was used as the parental strain. The uracil auxotrophic mutants TTK99, TTK100 (FERM number ABP-10957), and TTK101 were isolated from a 5fluoroorotic acid (5-FOA) plate, on which the MF-121 strain had been exposed to ultraviolet light. The strain will be given a FERM number and has been deposited in the International Patent Organism Depositary (IPOD; http:// unit.aist.go.jp/pod/cie/index.html). S. cerevisiae BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0; Openbiosystems, AL, USA) was used as a control. YPD medium (10 gl−1 yeast extract, 20 gl−1 peptone, and 20 gl−1 glucose) was used as the medium for basal growth. Synthetic dextrose (SD) medium (6.7 gl−1 yeast nitrogen base without amino acids and 20 gl−1 glucose) was used for transformation of the URA3 gene, and was supplemented with 50 mg l−1 uracil to confirm uracil auxotrophy. The 5-FOA plate comprised SD medium supplemented with 1 gl−1 5-FOA (Wako, Shiga, Japan) and 50 mg l−1 uracil. YCG medium (1.7 gl−1 yeast nitrogen base without amino acids and ammonium sulfate, 10 gl−1 casamino acids (Becton, Dickinson and Co., MD, USA), and 20 gl−1 glucose) was used for liquid growth of the transformant harboring a URA3 marker. YCG5 medium (1.7 gl−1 yeast nitrogen base without amino acids and

Appl Microbiol Biotechnol (2010) 87:1841–1853

ammonium sulfate, 10 gl−1 casamino acids, and 50 gl−1 glucose), YCC5 medium (1.7 gl−1 yeast nitrogen base without amino acids and ammonium sulfate, 10 gl−1 casamino acids, and 50 gl−1 cellobiose (Sigma-Aldrich Co., MO, USA)), and YPD10 medium (10 gl−1 yeast extract, 20 gl−1 peptone, and 100 gl−1 glucose) were used for the ethanol fermentation experiments. The primers used in this study are shown in Table 1. Screen for high-expression promoters The MF-121 strain was grown under microaerophilic conditions for 20 h at 30°C in YPD10 medium. Yeast cells frozen with liquid N2 were crushed, total RNA was extracted and purified, and complementary DNA (cDNA) was synthesized. The cDNA was inserted into a pBluescript vector, resulting in a cDNA library of the MF-121 strain. The cDNA library was transformed into Escherichia coli DH5α, and 571 randomly selected clones were sequenced from the 5′ terminus DNA. The amino acid sequence predicted from the DNA sequence analysis was compared with that of known genes by means of a BLAST search. In the library, cDNA clones of the I. orientalis genes TDH1 (glyceraldehyde-3-phosphate dehydrogenase; IoTDH1) and PGK1 (3-phosphoglycerate kinase; IoPGK1) were represented at high frequency, and subsequently, the promoters of these genes were cloned. To clone the promoters of IoTDH1 and IoPGK1, genomic DNA of the MF-121 strain was purified and digested by the following procedure. Approximately 5 µg of genomic DNA was digested with an SspI restriction Table 1 Primers used in this study Primer name

Sequence (5′–3′)

Adaptor primer F

TAGCGGTAAAGACCGGTGAAACTGG AATTTGCTCACAATCGCAGGTGTG CGGGCCAAGCG CGCTTGGCCCGCACACCTGCGATTG TGAGCAAATTCCAGTTTCACCGGT CTTTAC

Adaptor primer R

AP1 primer IoTDH1+200R primer IoPGK+190R primer AP2 primer IoTDH1+170R primer IoPGK+105R primer URA3-F2 URA3-R2 IoURA3-543F IoURA3+1067R

GACCGGTGAAACTGGAATTTGC GTGATGGCATTGTCCTTACCAG AGTGGGAAGCAAGAACAACGTATC ACAATCGCAGGTGTGCGGGC CCCTTGTACTTACCGTGGGTG AACATATTCAATTGTTGGAAGAGC AG AARTTYGCHGAYATHGG CCHACHCCDGGNGTCA AAACAGGGAAGGTTGACATTGTCT AGCGGC AACACTTAGAATACGCGGAACAAT CAATCG

1843

enzyme (Takarabio Co., Shiga, Japan) and deactivated for 15 min at 70°C, and then, the digested DNA was cleaned with a DNA Clean & Concentrator-5 Kit (Zymo Research Co., CA, USA). Mixtures of 2.5 µg of the digested DNA, adaptor primer F, and adaptor primer R (Table 1) were ligated using a LigaFast DNA ligation system (Promega Co., WI, USA), and the ligated DNA of 500 bp was separated by electrophoresis, cut out of the agarose gel, and then purified with a Zymoclean Gel DNA Recovery Kit (Zymo research). The 5′ flanking DNA regions of IoTDH1 and IoPGK1 were amplified with Extaq DNA polymerase (Takarabio) using primer set 1 (AP1 primer and IoTDH1+200R primer) and primer set 2 (AP1 primer and IoPGK+190R primer; Table 1), respectively, and then, the DNA was purified. The PCR products were purified with a DNA Clean & Concentrator-5 Kit and used as a template for the nested PCR reaction. The DNA sequence of the IoTDH1 promoter or IoPGK1 promoter was amplified with primer set 3 (AP2 primer and IoTDH1+170R primer) or primer set 4 (AP2 primer and IoPGK+105R primer; Table 1), respectively. The PCR products were purified and inserted into the pCR-TOPO vector (Invitrogen Co., CA, USA), resulting in a 629-bp IoTDH1 promoter (DDBJ/EMBL/ GenBank accession number AB538278) and a 545-bp IoPGK1 promoter (DDBJ/EMBL/GenBank accession number AB538279). Cloning of a secretion signal sequence in I. orientalis From BLAST search results of the constructed cDNA library, we confirmed that the 5′ terminus DNA sequence of the SED1 (IoSED1) gene, which exhibits homology to that of S. cerevisiae SED1 (ScSED1), was present in the library. The signal sequence of the IoSED1 gene was predicted with Genetyx Win version 9.1.2 (Genetyx Co., Tokyo, Japan), and its signal sequence (Met Gln Phe Lys Tyr Leu Ala Pro Leu Ala Leu Ala Gly Ser Ala Val Ala Ala Phe corresponding to 5′-atg caa ttc aag tac tta gca cca tta gct tta gca ggt tct gct gtc gca gct ttc-3′) was determined and named secretion signal of MF-121 SED1 (SSMS). Cloning of a terminator sequence in I. orientalis In the cDNA library, we found a part of the TEF1 gene. To use it for gene expression, 333 bp of the DNA sequence of the TEF1 terminator was cloned (DDBJ/EMBL/GenBank accession number AB538281). Plasmid construction To isolate the I. orientalis URA3 (IoURA3) gene, a degenerate PCR reaction was performed in which part of IoURA3 was amplified with KOD plus DNA polymerase (TOYOBO, Osaka, Japan) using the URA3-F2 primer, the URA3-R2 primer (Table 1), and genomic DNA of the MF-121 strain as a template. Part of

1844

Appl Microbiol Biotechnol (2010) 87:1841–1853

vector (Invitrogen). The pEG-IoTDHp-SSMS-AaBGLsec (IoTEF1t) vector, in which the IoTDH1 promoter, SSMS, the bgl1 gene, and IoTEF1t were fused, was inserted into the pDONR221 vector. The pEG-IoPGK1p-SSMS-AaBGLsec(ScSAG1t) vector, in which the IoPGK1 promoter, SSMS, the bgl1 gene, and ScSAG1t were fused, was inserted into the pDONR221 vector. The pEG-ScTDH3pSSMS-AaBGLsec(IoTEF1t) vector, in which the ScTDH3 promoter, SSMS, the bgl1 gene, and IoTEF1t were fused, was inserted into the pDONR221 vector. These four entry vectors were recombined with the pHU-DEST vector in vitro using the LR clonase II enzyme mix (Invitrogen), which resulted in pHU-IoTDH1p-SSMS-AaBGLsec (ScSAG1t), pHU-IoTDH1p-SSMS-AaBGLsec(IoTEF1t), pHU-IoPGK1p-SSMS-AaBGLsec(ScSAG1t), and pHUScTDH3p-SSMS-AaBGLsec(IoTEF1t).

the IoURA3 gene was inserted into the pCR-Blunt IITOPO vector (Invitrogen Co., CA, USA), resulting in the pTOPO-MFU vector. To clone the full length IoURA3 gene, a PCR reaction was performed using the IoURA3543F primer, the IoURA3+1067R primer (Table 1), and genomic DNA of the MF-121 strain as a template, and the resulting PCR product was inserted to the pCR-Blunt IITOPO vector, resulting in pTOPO-IoURA3. Sequence analysis verified the IoURA3 DNA sequence (DDBJ/ EMBL/GenBank accession number AB538280). The pYO323-IoURA3 vector was constructed by inserting the IoURA3 fragment from the pTOPO-IoURA3 vector, which had been digested with NotI and SacI restriction enzymes (Takarabio), into the pYO323 vector (National Bioresource Project (NBRP) ID no. BYP563) harboring a 2-μm replication origin and the S. cerevisiae HIS3 marker. The pHU-DEST vector was constructed by inserting reading frame cassette B (Invitrogen) containing the ccdB survival resistant marker into pYO323-IoURA3, which had been digested with a SmaI restriction enzyme (Takarabio; Fig. 1). Plasmid vectors expressing the bgl1 gene from A. aculeatus (DDBJ/EMBL/GenBank accession number D64088; Kawaguchi et al. 1996) were constructed as follows (Fig. 1): A pEG-IoTDHp-SSMS-AaBGLsec (ScSAG1t) vector, in which the IoTDH1 promoter, SSMS, the bgl1 gene, and the S. cerevisiae SAG1 terminator (ScSAG1t) were fused, was inserted into the pDONR221

Screen for uracil auxotrophic mutants of the MF-121 strain The MF-121 strain was grown in YPD liquid medium at 30°C with shaking (120 rpm) for 24 h; 250 µl of the culture was then transferred into 10 ml of YPD medium and grown at 30°C for 5 h. Yeast cells were washed once with deionized water and then spread on a YPD plate. The plate was irradiated with ultraviolet light for 1 min and then incubated at 30°C for a day. Yeast cells on the YPD plate were replicated onto a 5-FOA plate and incubated at 30°C for 4 days. Twenty 5-FOA resistant colonies were isolated, and uracil auxotrophy was checked on SD medium.

Fig. 1 Structures of the plasmid DNAs used in this study. The pHUDEST vector was used as a destination vector. Other plasmid vectors were constructed by using LR clonase (Invitrogen) with the appropriate entry vector. Symbols labeled attR1, attR2, attB1, and attB2 show the DNA sequence sites for DNA recombination. Other symbols show the chloramphenicol resistance gene (Cmr), toxin gene (ccdB), signal sequence of MF-121 IoSED1 (SSMS), S. cerevisiae

DNA replication origin (2 micron), E. coli DNA replication origin (ori), ampicillin resistance gene (Ampr), S. cerevisiae HIS3 marker (ScHIS3), I. orientalis URA3 marker (IoURA3), bgl1 gene from A. aculeatus (AaBGL), I. orientalis TDH1 promoter (IoTDH1 pro), I. orientalis PGK1 promoter (IoPGK1 pro), S. cerevisiae TDH3 promoter (ScTDH3 pro), I. orientalis TEF1 terminator (IoTEF1t), and S. cerevisiae SAG1 terminator (ScSAG1t)

Appl Microbiol Biotechnol (2010) 87:1841–1853

Measurement of specific growth rates The MF-121 strain and BY4741 strain were each grown in 2 ml of YPD medium at 30°C with shaking (120 rpm) for 24 h. The optical density at 660 nm (OD660) of the cultures was measured, and yeast cells were inoculated to give an OD 660 value of 0.01 in 4 ml of YPD in L-tubes. Cultivation in the tubes was performed with shaking (70 rpm) at various temperatures, and OD660 values and temperatures were monitored with a Bio-photorecorder TVS062CA (ADVANTEC, Tokyo, Japan). The maximum specific growth rate (μmax) was calculated by using several adjacent points in the logarithmic growth phase (OD660 <2.5). Repeated batch fermentation The MF-121 strain or BY4741 strain was aerobically grown in 2 ml of YPD medium at 30°C with shaking (120 rpm) for 24 h. All of the fermentation tests were performed in a 100-ml reagent bottle with an airtight cap. For removal of a sample, the cap of the reagent bottle was opened, which introduced some air and resulted in a microaerobic condition. After the fermentation, yeast cells were washed twice with deionized water and inoculated to give an OD600 value of 2 in 40 ml of YPD10 medium, and then subjected to fermentation under a microaerobic condition with shaking (120 rpm) at 40°C for 24 h. Yeast cells after the fermentation were washed once with deionized water, OD600 values were measured, and then, the cells were inoculated to give an OD600 value of 2 in 40 ml of YPD10 liquid medium. These procedures were performed a total of four times. Ethanol concentrations were measured using a column Shimopack SPR-Pb (Simadzu Co., Kyoto, Japan) and a HPLC detector (model RID-10A., Shimadzu). HPLC was carried out at 80°C using deionized water, at a flow rate of 0.8 ml min−1, as the mobile phase. The theoretical ethanol yield was defined by taking the conversion of 1 g of glucose to 0.51 g of ethanol as 100%. Transformation Yeast transformation was performed according to the lithium acetate transformation protocol (Gietz and Schiestl 2007). Yeast cells were grown in YPD medium with shaking (120 rpm) for 18 to 24 h at 30°C. Cultures were transferred to 10 ml of YPD medium and then grown with shaking (120 rpm) for 5 h at 30°C. Yeast cells were precipitated by centrifugation, washed once with deionized water, and then reprecipitated. A yeast suspension was prepared from the yeast cell pellet by adding 100– 150 µl of deionized water. Next, 200-μl aliquots of transformation solution, comprising 120 µl of 60% w/v polyethylene glycol 3350 (PEG3350), 5 µl of 4 M lithium acetate, 10 µl of 1 M dithiothreitol, and 10 µl of 10 mg ml−1 deoxyribonucleic acid from salmon testes

1845

(carrier DNA) that had been boiled for 10 min at 95°C (Sigma-Aldrich) were mixed with 5–10 µl of linearized plasmid DNA or PCR product and 45–50 µl of cell suspension, and subjected to heat shock for the indicated times at 42°C or 45°C. Yeast cells in the samples were precipitated by centrifugation, mixed with 200 µl of deionized water, and then spread on SD medium. SSF experiments Transformants expressing β-glucosidase were grown in YCG medium at 30°C for 24 h. Yeast cells were washed once with deionized water, OD600 values were measured, and the cells were then inoculated into medium for the SSF process experiment. Four grams of microcrystalline cellulose Avicel PH-101 (Sigma-Aldrich Chemie GmbH, Buchs, Switzerland) was sterilized in a bottle, and then, 2 ml of 2 M sodium phosphate (pH 5), 10 ml of 4× YNB+CAA (6.8 gl−1 yeast nitrogen base without amino acids and ammonium sulfate, and 40 gl−1 casamino acids), 22.7 ml of deionized water, 5 ml of yeast cell mixture, and 327 µl (15 filter paper unit (FPU) per gram of Avicel) of a commercial preparation of cellulase from Trichoderma reesei ATCC26921 (Sigma-Aldrich) were added; 65 µl (one fifth of the volume of the commercial cellulase preparation) of a commercial β-glucosidase preparation (Novozyme 188; Sigma-Aldrich) was also added when needed. Fermentations were performed under a microaerobic condition with shaking (130 rpm) at 30°C or 40°C for 72 h. Ethanol concentrations were measured by HPLC. Enzyme assay β-Glucosidase activity was measured for both the culture supernatant and the cell surface. Transformants harboring β-glucosidase were grown in YCG medium at 30°C for 24 h. Cultures were divided into supernatants and yeast cells. The enzyme reaction for the supernatant was performed in 690 µl of deionized water, 200 µl of 250 mM sodium acetate buffer (pH 5), 100 µl of 20 mM p-nitrophenyl-β-D-glucopyranoside (pNPG; SigmaAldrich), and 10 µl of supernatant at 30°C for 60 min. Yeast cells were washed twice with deionized water and then adjusted to an OD600 value of 1. The enzyme reaction for the cell surface was performed in 675 µl of deionized water, 200 µl of 250 mM sodium acetate buffer (pH 5), 100 µl of 20 mM pNPG, and 25 µl of yeast cells (OD600 value of 1) at 30°C for 60 min. Next, 50 μl of the reaction solution was added to 100 µl of 1N sodium carbonate, and the absorbance at 400 nm was measured after centrifugation. Enzyme units were calculated from the standard curve for p-nitrophenol. One unit was defined the amount that released 1 µmol of p-nitrophenol from a substrate in 1 min. The cellulase activity of the commercially supplied enzyme preparation was measured by means of an FPU assay (Mandels et al. 1976).

1846

Appl Microbiol Biotechnol (2010) 87:1841–1853

Results Growth of the I. orientalis MF-121 strain at high temperature The thermotolerant yeast K. marxianus can grow at around 50°C (Banat et al. 1992; Hong et al. 2007; Nonklang et al. 2008). The I. orientalis MF-121 strain has been shown to grow at temperatures up to 40°C (Hisamatsu et al. 2006). We determined the restrictive temperature for growth of the MF-121 strain as compared with the S. cerevisiae parental strain BY4741 in YPD medium (Fig. 2a). The BY4741 strain could grow at temperatures up to 40–41°C, and its best specific growth rate (μmax) value was 0.37 h−1 at 38°C. On the other hand, the MF-121 strain could grow at temperatures up to 44–45°C, and its best μmax value was 1.0 h−1 at 39°C. Hence, we found that the MF-121 strain can grow healthily at temperatures higher than 40°C. This result indicates that the I. orientalis MF-121 strain is thermotolerant and thus suitable for the SSF process. To assess the thermotolerance characteristics of the MF121 strain, repeated batch fermentations were performed at 42°C in YPD10 medium (Fig. 2b). The ethanol yield of the BY4741 strain was 42% in the first batch samples, but ethanol was not detected in the subsequent batch samples. In contrast, the ethanol yield of the MF-121 strain was approximately 70% in all batch samples. For the BY4741 strain, glucose clearly remained in the culture fluid; for the MF121 strain, in contrast, almost none remained (data not shown). These results indicate that the MF-121 strain is thermotolerant, ethanologic, and also thermostable. Construction of a transformation system for the I. orientalis MF-121 strain In order to express heterologous genes in the MF-121 strain, we needed a transformation system for this strain. As yet, however, a transformation procedure for I. orientalis has not been developed. Yeast transformation often utilizes the URA3 gene (corresponding to orotidine-5′phosphate decarboxylase) as a transformation marker. For yeasts including Candida tropicalis (Haas et al. 1990), Pichia stipitis (Yang et al. 1994), Candida boidinii (Sakai 1.2

100

a

BY4741 MF-121

0.8 0.6 0.4

b

*

BY4741

1

2

MF-121

80

Ethanol yield (%)

1

µ max (h-1)

Fig. 2 Growth and ethanol fermentation of the I. orientalis MF-121 strain at high temperature. a Restriction temperature tests of growth were performed at the indicated temperatures in YPD medium. b Repeated batch fermentation tests were performed microaerobically for 24 h at 42°C in YPD10 medium. All data are given as the mean ± SD of three independent experiments. *P=0.133. Two bars indicate no statistical significant difference

et al. 1991), and Saccharomyces exiguus (Hisatomi et al. 1998), transformation systems have also been constructed using URA3. To screen for uracil auxotrophic mutants of MF-121 strain, positive selection using 5-FOA (Boeke et al. 1984) was performed. We carried out mutagenesis treatment with ultraviolet light, which produced 20 5-FOA-resistant clones. From these, three clones (TTK99, TTK100, and TTK101) were finally isolated as uracil auxotrophic mutants. Because the genomic DNA of I. orientalis has not been sequenced, it was necessary to clone the I. orientalis URA3 (IoURA3) gene to complement the uracil auxotrophic mutants. To isolate the IoURA3 gene, a PCR reaction was carried out using genomic DNA of the MF-121 strain as a template and degenerate primers that were used to isolate Candida bombicola URA3 (Van Bogaert et al. 2007). As a result, a fragment of the IoURA3 gene was cloned into plasmid DNA. This DNA fragment was identical to the URA3 gene of Candida glycerinogenes (CgURA3) previously isolated (DDBJ/EMBL/GenBank accession number AY623794). To isolate the full-length IoURA3 gene, we used genomic DNA of the MF-121 strain and primers designed on the basis of the URA3 sequence data of C. glycerinogenes. As a result, the IoURA3 gene of 1,611 bp containing the predicted 786-bp open reading frame was successfully cloned (DDBJ/EMBL/GenBank accession number AB538280). The amino acid sequence of IoURA3 was 100% identical to that of the CgURA3 gene (DDBJ/ EMBL/GenBank accession number AAT39474), and its DNA sequence had three different base pairs in the promoter, open reading frame, and terminator as compared with CgURA3. C. glycerinogenes was isolated as a glycerol producer (Zhuge et al. 2001). Whereas the rRNA sequence of C. glycerinogenes has not been determined; however, the rRNA sequence of the MF-121 strain has been determined (Hisamatsu et al. 2006), and thus, the MF-121 strain has been classified as I. orientalis. To determine whether or not the isolated mutants could be complemented by the URA3 gene, a PCR DNA fragment of IoURA3 or S. cerevisiae URA3 (ScURA3) was trans-

60 40 20

0.2 0

0 38

40

42

44

46

Temperature (ºC)

48

3

4

Repeated fermentation (times)

Appl Microbiol Biotechnol (2010) 87:1841–1853

1847

Table 2 Transformation of the URA3 gene to uracil auxotrophic mutants with the URA3 gene Strain

TTK99 (I. orientalis) TTK100 (I. orientalis) TTK101 (I. orientalis) BY4741 (S. cerevisiae)

CFUμg−1 DNA IoURA3

ScURA3

No DNA

0 28 0 7

0 0 0 920

0 0 0 0

Transformation of the URA3 DNA fragment into the TTK100 strain and BY4741 strain was performed in a reaction solution comprising 36% w/v PEG3350, 100 mM lithium acetate, 50 mM dithiothreitol, and 0.5 mg/ml of carrier DNA, with heat shock for 60 min. Colonyforming units (CFU) were determined for colonies grown on the medium.

formed into the three uracil auxotrophic mutants and the S. cerevisiae BY4741 strain by means of the lithium acetate method (Table 2). As a result, it was found that the TTK100 strain could be complemented by the DNA fragment of IoURA3. However, the TTK99 strain and TTK101 strain were not complemented by it, and consequently, they may have a mutation in one of the genes of the uracil synthesis pathway other than URA3. In addition, the TTK100 strain was complemented with only the IoURA3 open reading frame (data not shown), suggesting that its mutation is located in the open reading frame of URA3. On the other hand, the BY4741 strain was also complemented by the IoURA3 gene at low transformation efficiency. The transformation efficiency of the TTK100 strain with the IoURA3 gene was approximately 3% of that of BY4741 with the ScURA3 gene. Thus, I. orientalis seemed to exhibit a low transformation efficiency as compared with S. cerevisiae. To enhance the transformation efficiency of the TTK100 strain, the transformation conditions were assessed in terms of optimal heat shock temperature and suitable heat shock duration (Table 3). The transformation efficiency gradually increased with the duration of the heat shock at both 42°C and 45°C. The most efficient transformation conditions for the TTK100 strain were 42°C for 120 min. In contrast, the most efficient conditions for S. cerevisiae were 42°C for 40 min (Gietz and Schiestl 2007). These results indicate that it is necessary to optimize the heat shock conditions to give the best transformation efficiency for a thermotolerant yeast. Isolation of high-expression promoters in I. orientalis Next, we screened for a high-expression promoter to express heterologous proteins in I. orientalis. Total RNA was extracted from MF-121 cells that produced ethanol in YPD10 medium, and the corresponding cDNA was synthesized and used to produce a cDNA library. The 5′ terminus of 571 randomly selected clones was then

sequenced. As a result, predicted IoTDH1 and IoPGK1 genes were found to exist in each of 70 (12.2%) and 16 (2.8%), respectively, of the 571 clones in the cDNA library. We therefore cloned the promoters of IoTDH1 (DDBJ/ EMBL/GenBank accession number AB538278) and IoPGK1 (DDBJ/EMBL/GenBank accession number AB538279) and used them for heterologous protein expression. β-Glucosidase expression in the TTK100 strain In order to determine whether heterologous proteins could be expressed by the TTK100 strain, the A. aculeatus bgl1 gene, which has been previously utilized for lactic acid production by recombinant S. cerevisiae (Tokuhiro et al. 2008), was used. Initially, we tried chromosome integration of bgl1 into the genome of the TTK100 strain, but we did not isolate any transformants (data not shown). Consequently, an episomal plasmid DNA was used to transformation in I. orientalis. A plasmid DNA harboring the bgl1 gene and the 2-μm replication origin of S. cerevisiae was transformed into TTK100 strain, and the resulting transformant could successfully express the bgl1 gene (Fig. 1 and Table 4). Three promoters (IoTDH1, IoPGK1, and ScTDH3) and two terminators (IoTEF1 and ScSAG1) were compared to assess the best conditions for expression of the bgl1 gene. The signal sequence of IoSED1 was fused to the N terminus of the mature β-glucosidase enzyme. Transformants expressing β-glucosidase showed activity in both the supernatant and cell surface fractions. Transformants harboring the control vector (pHU-DEST) exhibited no activity (Table 4f). The β-glucosidase activity of the bgl1 gene with the IoTEF1 terminator was 1.2-fold higher than that of bgl1 with the ScSAG1 terminator in both the supernatant and the cell surface fractions (Table 4a, b). The β-glucosidase activity of the bgl1 gene coupled with the IoTDH1 promoter was 1.7-fold higher than that of bgl1 coupled with the IoPGK1 promoter in both fractions (Table 4b, c). In addition, the IoTDH1 promoter was also functional in S. cerevisiae, but its activity was 1.9-fold lower than that of the ScTDH3 promoter in both fractions (Table 4d, e). Thus, the optimal combination for bgl1 Table 3 Transformation efficiency of the heat reaction of the TTK100 strain Reaction time (min)

42°C 45°C

CFUμg−1 DNA 30

60

90

120

150

180

37 67

57 92

81 6

122 0

10 0

0 0

The transformation procedure was the same that in Table 1, except that heat shock of the transformation solution was performed several times at 42°C or 45°C.

1848

Appl Microbiol Biotechnol (2010) 87:1841–1853

Table 4 β-Glucosidase activity of the TTK100 strain and BY4741 strain harboring bgl1 from A. aculeatus Expression vector

Host strain

Promoter

Specific BGL activityg

Terminator

a b c d e

pHU-IoTDH1p-SSMS-AaBGL(IoTEF1t) pHU-IoTDH1p-SSMS-AaBGL(ScSAG1t) pHU-IoPGK1p-SSMS-AaBGL(ScSAG1t) pHU-IoTDH1p-SSMS-AaBGL(IoTEF1t) pHU-ScTDH3p-SSMS-AaBGL(IoTEF1t)

TTK100 TTK100 TTK100 BY4741 BY4741

IoTDH1pro IoTDH1pro IoPGK1pro IoTDH1pro ScTDH3pro

IoTEF1t ScSAG1t ScSAG1t IoTEF1t IoTEF1t

f

pHU-DEST

TTK100

–

–

Supernatant (mU/ml)

Cell surface (mU/OD600)

343.5±8.9 278.5±18.7 164.6±13.2 177.3±17.7 336.4±61.6

212.7±2.1 173.4±8.2 96.7±8.6 89.5±5.4 175.9±3.3

3.1±3.2

2.9±2.5

Transformants were grown in YCG medium at 30°C for 24 h. Cell surface β-glucosidase activity was determined after the cells were washed twice with deionized water. β-Glucosidase activity was measured at 30°C for 60 min. β-Glucosidase activity is given the mean ± SD of three independent experiments.

Ethanol production with cellobiose as the sole carbon source at high temperature Some yeasts, including S. cerevisiae (Fujita et al. 2002; Rajoka et al. 2003; Fujita et al. 2004; van Zyl et al. 2007; Tokuhiro et al. 2008) and K. marxianus (Hong et al. 2007), have been transformed with the β-glucosidase gene and subsequently shown to convert cellobiose to ethanol. The ethanol fermentation efficiency of the TTK316 strain harboring the IoTDH1 promoter, bgl1, and the IoTEF1 terminator (Table 4a) was assessed with cellobiose or glucose as the sole carbon source at 40°C (Fig. 3a). Ethanol production was saturated at 6–8 h in the case of both glucose and cellobiose, and ethanol yields in both media reached 75–80% of the theoretical yield. The cellobiose consumption rate (8.8 gh−1) was the same as the

Ethanol production at high temperature under acidic conditions with a transformant expressing β-glucosidase To prevent contamination of lignocellulosic biomass, we considered that the SSF process should be performed under acidic conditions. The MF-121 strain can produce ethanol even at pH 2 (Hisamatsu et al. 2006). In addition, the TTK316 strain could convert cellobiose to ethanol at high temperature (Fig. 3a). In order to determine whether or not cellobiose could be converted to ethanol under acidic conditions and at high temperature, ethanol fermentation was performed under a microaerobic condition at 40°C in media with varying initial pH (Fig. 3b and Table 5). The 60

50

25

50

25

40

20

40

20

30

15

30

15

20

10

20

10

10

5

10

5

0

0

a

0 0

2

4

6

8

10

Ethanol (g l-1)

30

60

Suger (g l-1)

glucose consumption rate (8.9 gh−1). Consequently, these results suggest that providing I. orientalis with the ability to express β-glucosidase is enough to produce ethanol at high temperature.

12

Time (h) Fig. 3 Ethanol fermentation with cellobiose at high temperature. Ethanol fermentation was performed microaerobically to give an OD600 value of 10 at 40°C. a Ethanol fermentation with glucose and cellulose was compared in YCC5 medium. Open symbols show the concentrations of glucose (triangles), cellobiose (squares), and ethanol (circles) in YCG5 medium. Filled symbols show the concentrations of glucose (triangles), cellobiose (squares), and ethanol (circles) in YCC5 medium. b Ethanol fermentation was

30

b

Ethanol (g l-1)

expression was a combination of the IoTDH1 promoter, bgl1, and the IoTEF1 terminator (Table 4a).

Cellobiose (g l-1)

g

0 0

2

4

6

8

10

12

Time (h) performed at the indicated pH in YCC5 medium. Open symbols show the concentrations of cellobiose under an initial pH of 5 (circles), 4 (squares), 3 (triangles), and 2 (diamonds) during fermentation. Filled symbols show the concentrations of ethanol under an initial pH of 5 (circles), 4 (squares), 3 (triangles), and 2 (diamonds) during fermentation. All data are the average values of two independent experiments

Appl Microbiol Biotechnol (2010) 87:1841–1853

1849

Table 5 Ethanol fermentation by I. orientalis harboring β-glucosidase under varying initial pH Initial pH for fermentation

Last pH in fermentation

Sugar consumption rate (gh−1)

Ethanol production rate (gh−1)

Ethanol yield (%)

5.5 4.0 3.0 2.0

3.8 3.3 2.7 2.0

8.57 8.28 5.98 0.67

3.09 3.34 1.93 0.35

74.7 75.1 67.6 0

The TTK316 strain expressing β-glucosidase was used for fermentation. All data are the average values of two independent experiments. Fermentation was performed for 12 h at 40°C, and the initial and last pH indicate the pH at 0 and 12 h, respectively. The sugar consumption rate and ethanol production rate were calculated from the cellobiose and ethanol concentrations at 6 h. The ethanol yield shows the values at 12 h.

TTK316 strain converted cellobiose to ethanol for up to 12 h at 40°C at a pH above 3. However, the TTK316 strain could not convert cellobiose to glucose at pH 2 because βglucosidase was inactivated at very low pH. This result indicates that the TTK316 strain can convert cellobiose to ethanol even under acidic conditions and at high temperature. Ethanol fermentation through the SSF process with a transformant expressing β-glucosidase at high temperature Experiments on the SSF process with K. marxianus have been performed by many researchers (Ballesteros et al. 1991; Ballesteros et al. 1993; Bollók et al. 2000; Suryawati et al. 2008; Tomás-Pejó et al. 2009). However, an experiment with K. marxianus expressing cellulase has not been performed. Using the TTK316 strain expressing β-glucosidase, we assessed the ability of I. orientalis to perform SSF in medium containing 100 gl−1 of minicrystalline cellulose (Avicel PH-101) and a commercial cellulase preparation from T. reesei (Fig. 4). The TTK323 strain (TTK100 harboring the pHU-DEST control vector,

Table 4f) produced 4 gl−1 of ethanol at 30°C in 72 h, as compared with 17 gl−1 of ethanol at 40°C. This observation indicates that the SSF process in I. orientalis breaks down crystalline cellulose and produces ethanol more efficiently at high temperature. Because the commercial cellulase preparation does not contain enough β-glucosidase, it is necessary to add a β-glucosidase preparation to hydrolyze cellobiose into glucose. Hence, the β-glucosidase preparation was added to the commercial cellulase preparation. Under these conditions, the TTK323 strain produced 17 g l−1 of ethanol at 30°C in 72 h and 27 gl−1 of ethanol at 40°C (Fig. 4a, b). As a result, ethanol production increased 4-fold at 30°C and 1.5-fold at 40°C as compared within the medium without the β-glucosidase preparation. The ethanol fermentation profile of the TTK316 strain without the βglucosidase preparation was slightly higher than that of the TTK323 strain with the β-glucosidase preparation added to the medium (Fig. 4a, b). In addition, the TTK316 strain with the β-glucosidase preparation produced 30 gl−1 of ethanol at 40°C in 72 h, and its ethanol fermentation profile was slightly higher than that of the TTK316 strain without

30

30

a

b 25

30ºC

20

Ethanol (g l-1)

Ethanol (g l-1)

25

15 10 5

40ºC 20 15 10 5

0

0 0

12

24

36

48

60

72

Time (h) Fig. 4 SSF experiments at high temperature with I. orientalis expressing β-glucosidase. Fermentation tests were performed microaerobically with 100 gl−1 of Avicel and 15 FPU g−1 of Avicel of a commercial cellulase preparation. The fermentation results at a 30°C and b 40°C are shown. Circles show the ethanol concentrations obtained with the TTK316 strain expressing the bgl1 gene. Triangles show the ethanol concentrations obtained with the TTK323 strain and

0

12

24

36

48

60

72

Time (h) a β-glucosidase preparation (Novozyme 188) added at one fifth of the level of the commercial cellulase preparation (resulting in 22.2 FPU g−1 Avicel). Squares show the ethanol concentrations obtained with the TTK323 strain and no addition of the β-glucosidase preparation. All data are the average values of two independent experiments

1850

the β-glucosidase preparation (data not shown). This observation indicates that expression of the bgl1 gene by the TTK316 strain is sufficient for the SSF process without the β-glucosidase preparation. In our experiments, the amount of cellulose remaining was not measured, and Avicel could not be completely hydrolyzed with the commercial cellulase preparation and β-glucosidase preparation. As a result, when the fermentation was conducted using the TTK316 strain coupled with the commercial cellulase from T. reesei, the TTK316 strain converted 100 g l−1 of Avicel to 29 gl−1 of ethanol at 40°C in 72 h even without β-glucosidase addition.

Discussion Although yeast is promising as a microorganism for bioethanol production from lignocellulosic biomass because it is ethanologenic and highly ethanol tolerant, it does not have any saccharification enzymes such as cellulase. In addition, thermotolerant yeasts can grow above 40°C. On the other hand, thermophilic bacteria show poor ethanol productivity and poor ethanol tolerance, and produce various by-products such as acetone and acetic acid. A thermotolerant yeast that can grow healthily at around 60°C has not been found so far. Yeasts and bacteria have varying abilities with regard to CBP but have advantages and disadvantages. The next step for bioethanol production from lignocellulosic biomass is genetic engineering. We previously isolated the I. orientalis MF-121 strain from a river in a hot spring area (Hisamatsu et al. 2006) and confirmed that it can grow at temperatures up to 44–45°C (Fig. 2a). Other isolated strains of I. orientalis have also been shown to be able to grow at 40°C (Hong et al. 2007; Koh and Suh 2009). We found that the MF-121 strain could be used for repeated batch fermentation at 42°C, but the ethanol yield (approximately 70%) was not good in YPD10 medium (Fig. 2b). A culture of the I. orientalis MF-121 strain bubbled owing to unknown materials. It has been reported that the I. orientalis SR4 strain secretes a biosurfactant, oleic acid (Katemai et al. 2008). Thus, the MF-121 strain may secrete oleic acid, similar to the SR4 strain. Another characteristic of I. orientalis MF-121 is that it can produce ethanol at under pH 2–3 and with 5% sodium sulfate (Hisamatsu et al. 2006). Indeed, the TTK316 strain expressing β-glucosidase could convert cellobiose to ethanol at a pH of under and at 40°C (Fig. 3b). Ethanol fermentation under acidic conditions is important to prevent contamination, although it is necessary to use cellulase enzymes that are adapted to low pH. These results suggest that I. orientalis is an efficient strain for producing ethanol under acidic conditions and at high temperature.

Appl Microbiol Biotechnol (2010) 87:1841–1853

A transformation system for I. orientalis is required to transform cellulase genes because the I. orientalis MF-121 strain does not exhibit any cellulase activity. Consequently, we screened for uracil auxotrophic mutants after mutagenesis with ultraviolet light, which resulted in the TTK100 strain, an ura3 auxotrophic mutant. In addition, the IoURA3 gene was also cloned by means of degenerate PCR, and its amino acid sequence was found to be identical to that of the CgURA3 gene previously isolated (Li et al. 2005). The C. glycerinogenes WL2002-5 strain was isolated as a glycerol producer (Zhuge et al. 2001), but its rRNA sequence has not been determined (Wang et al. 1999). The MF-121 strain is not an excellent glycerol producer as compared with the WL2002-5 strain, because it produced only 1–2 gl−1 of glycerol in YPD10 medium (100 gl−1 of glucose; data not shown). However, we did not determine whether or not the I. orientalis MF-121 strain and C. glycerinogenes WL20025 were the same species, although the rRNA sequence of the MF-121 strain has been determined (Hisamatsu et al. 2006). Transformation with the IoURA3 gene was achieved with an efficiency of 28 CFU µg−1 DNA for the TTK100 strain (Table 2). The thermotolerant yeast K. marxianus DMKU3-1042 strain exhibits a transformation efficiency of 4×103 CFU µg−1 DNA (Nonklang et al. 2008). If the transformation efficiency could be increased for I. orientalis, then genetic engineering might be efficiently performed. Indeed, we found that the transformation efficiency could be increased by improving the heat shock duration (Table 3). However, further improvement of the transformation efficiency is required. In S. cerevisiae, deletion of the DNA helicase SGS1 gene causes hyperrecombination (Yamagata et al. 1998). If the SGS1 gene or related genes of I. orientalis could be disrupted, then the transformation efficiency might be increased further. Presumably, homologous recombination of I. orientalis might be weaker than that of S. cerevisiae. Alternatively, DNA uptake by I. orientalis may be seriously inhibited owing to the thickness of its cell wall, because the cell wall of the MF-121 strain is difficult to degrade with yeast cell wall-degrading enzymes such as zymolyase. As another method of transformation, electroporation has been performed with various yeasts including Candida albicans (De Backer et al. 1999), C. glycerinogenes (Chen et al. 2008), Kluyveromyces lactis (Sánchez et al. 1993), H. polymorpha (Faber et al. 1994), Pichia farinosa (Wang et al. 2005), Schizosaccharomyces pombe (Suga and Hatakeyama 2001), and Torulopsis glabrata (Zhou et al. 2009). We have not addressed here whether or not an electroporation method would be useful. Indeed, transformation of the glycerol producer C. glycerinogenes has been achieved by electroporation with a transformation efficiency of approximately 400 CFU µg−1 DNA (Chen et

Appl Microbiol Biotechnol (2010) 87:1841–1853

al. 2008). Our maximum transformation efficiency using lithium acetate method was 122 CFU µg−1 DNA (Table 2). Thus, an electroporation method may improve the efficiency of our transformation system. We succeeded in transforming an episomal plasmid DNA harboring the 2-μm replication origin of S. cerevisiae into the TTK100 strain with a transformation efficiency of 0.4–1.2 CFU µg−1 DNA (data not shown). This low transformation efficiency indicates that it is difficult to maintain the episomal plasmid DNA harboring the replication origin of S. cerevisiae in I. orientalis. Surprisingly, the plasmid maintenance rate of the TTK316 strain, which was measured by auxotrophic assay, was above 95% in nonselective (YPD) medium, in which the TTK316 strain was grown for 3 days, whereas that of the BY4741 strain was only 10% (data not shown). However, we could not recover plasmid DNA from cell extracts of the TTK316 strain, whereas it could be recovered from the BY4741 strain. To assess whether the plasmid DNA was integrated into genomic DNA or maintained as an episomal plasmid, we performed a Southern blot analysis of TTK316 strain expressing the AaBGL gene. A specific band was detected in transformants harboring the AaBGL gene (data not shown). As a result, the plasmid DNA was inserted into the genome. These results assume that the plasmid DNA was inserted into the genome by homologous recombination machinery using the IoURA3 gene in a plasmid DNA or randomly integrated into the genome. If the plasmid DNA was linearized, transformation efficiency may be increased. To express heterologous proteins in I. orientalis, we isolated the IoTDH1 promoter and the IoPGK1 promoter as high-expression promoters that expressed their corresponding genes efficiently during ethanol fermentation. The expression library included the 5′ terminus sequences of IoTDH1 (12.2%) and IoPGK1 (2.8%). In addition, β-glucosidase activity under control of the IoTDH1 promoter was increased approximately 2-fold as compared with the IoPGK1 promoter (Table 4). TDH and PGK are two of the most highly expressed proteins as assessed by global protein tag analysis of S. cerevisiae (Ghaemmaghami et al. 2003). Consequently, these results suggest that the glycolysis pathway is also important for ethanol fermentation with I. orientalis similar to that with S. cerevisiae. Using the TTK316 strain, ethanol fermentation was performed with cellobiose or glucose as a carbon source at 40°C. As a result, we found that the ethanol fermentation and sugar consumption rates with cellobiose were the same as those with glucose (Fig. 3a). This observation means that the β-glucosidase expressed by I. orientalis shows enough activity for ethanol fermentation even at high temperature. Indeed, the culture supernatant of the TTK316 strain

1851

contained 343 mU ml−1 of β-glucosidase, whereas that of K. marxianus expressing β-glucosidase is only 83 mU ml−1 (Hong et al. 2007). This observation suggests that the bgl1 gene from A. aculeatus yields a good β-glucosidase, and the resulting transformants exhibit excellent performance in I. orientalis. The TTK316 strain exhibited an ethanol production rate of 3.25 gh−1 and a cellobiose consumption rate of 8.82 gh−1 with an initial OD600 value of 10 in YCC5 medium containing 50 gl−1 of cellobiose at 40°C (Fig. 3a), whereas K. marxianus exhibits an ethanol production rate of approximately 3.5 gh−1 and a cellobiose consumption rate of approximately 10 gh−1 with an initial OD600 of 15 in SCB-10 medium (6.7 gl−1 YNB without amino acids, 20 g l−1 casamino acids, and 100 gl−1 cellobiose) at 45°C, according to our calculation (Hong et al. 2007). This finding suggests that the ethanol fermentation rate with I. orientalis is as good as that the K. marxianus NBRC1777 strain at high temperature. In our experiments, no carbon source including cellobiose or glucose remained in the medium (Fig. 3a), whereas in the case of K. marxianus, glucose remained in the medium (Hong et al. 2007). With regard to this, I. orientalis is superior to K. marxianus. In addition, I. orientalis also showed acid tolerance and salt tolerance, and produced ethanol under acidic conditions and at high temperature (Fig. 3b and Table 4). Yeasts with these characteristics have not previously been described. Thus, we suggest that I. orientalis is an efficient strain for the SSF process at high temperature. We performed the SSF experiments with I. orientalis using microcrystalline cellulose as the sole carbon source at high temperature (Fig. 4). SSF experiments with K. marxianus have been performed using Solka-Floc or a lignocellulosic biomass containing spruce and switchgrass (Ballesteros et al. 1991, 1993; Bollók et al. 2000; Suryawati et al. 2008). The ethanol fermentation curve of the TTK316 strain without the β-glucosidase preparation was slightly higher than that of the TTK323 strain with the βglucosidase preparation (Novozyme188; Fig. 4). These results indicate that I. orientalis has the ability to express enough β-glucosidase to degrade oligomers from Avicel. If the Avicel source was pretreated by various methods, then a higher ethanol yield might be expected. To our knowledge, we are the first to demonstrate a reduction in the cost of the saccharification enzyme preparation by using a thermotolerant yeast expressing heterologous cellulase in SSF experiments. Further approaches are required to obtain greater a cost reduction for the saccharification enzyme preparation by using heterologous transformation of endoglucanase, cellobiohydrolase, and saccharification-related enzymes. More transformation markers will be required for this purpose. In summary, we have constructed a transformation system for I. orientalis involving an ura3 mutant and the

1852

IoURA3 gene, and succeeded in expressing a heterologous gene. In addition, we have shown that transformants expressing β-glucosidase lead to successful cost reduction for the β-glucosidase preparation in SSF experiments at high temperature. In the future, we hope to achieve CBP through technical improvements with the thermotolerant yeast I. orientalis. Acknowledgments The pYO323 plasmid vector (NBRP ID number BYP563) of YEp type harboring the HIS3 marker was kindly donated by the National Bioresource Project (Yeast) of Japan (http://yeast.lab. nig.ac.jp/nig/index_en.html). We also thank Risa Nagura, Satoshi Katahira, and Takashi Matsuyama for technical assistance.

References Ballesteros I, Ballesteros M, Cabañas A, Carrasco J, Martín C, Negro MJ, Saez F, Saez R (1991) Selection of thermotolerant yeasts for simultaneous saccharification and fermentation (SSF) of cellulose to ethanol. Appl Biochem Biotechnol 28– 29:307–315 Ballesteros I, Oliva JM, Ballesteros M, Carrasco J (1993) Optimization of the simultaneous saccharification and fermentation process using thermotolerant yeasts. Appl Biochem Biotechnol 39–40:201–211 Banat IM, Nigam P, Marchant R (1992) Isolation of thermotolerant, fermentative yeasts growing at 52°C and producing ethanol at 45°C and 50°C. World J Microbiol Biotechnol 8:259–283 Boeke JD, LaCroute F, Fink GR (1984) A positive selection for mutants lacking orotidine-5′-phosphate decarboxylase activity in yeast: 5-fluoro-orotic acid resistance. Mol Gen Genet 197:345– 346 Bollók M, Réczey K, Zacchi G (2000) Simultaneous saccharification and fermentation of steam-pretreated spruce to ethanol. Appl Biochem Biotechnol 84–86:69–80 Chen X, Fang H, Rao Z, Shen W, Zhuge B, Wang Z, Zhuge J (2008) An efficient genetic transformation method for glycerol producer Candida glycerinogenes. Microbiol Res 163:531–537 Daniel HM, Vrancken G, Takrama JF, Camu N, De Vos P, De Vuyst L (2009) Yeast diversity of Ghanaian cocoa bean heap fermentations. FEMS Yeast Res 9:774–783 De Backer MD, Maes D, Vandoninck S, Logghe M, Contreras R, Luyten WH (1999) Transformation of Candida albicans by electroporation. Yeast 15:1609–1618 Dmytruk OV, Dmytruk KV, Abbas CA, Voronovsky AY, Sibirny AA (2008) Engineering of xylose reductase and overexpression of xylitol dehydrogenase and xylulokinase improves xylose alcoholic fermentation in the thermotolerant yeast Hansenula polymorpha. Microb Cell Fact 7:21 Faber KN, Haima P, Harder W, Veenhuis M, AB G (1994) Highlyefficient electrotransformation of the yeast Hansenula polymorpha. Curr Genet 25:305–310 Fujita Y, Takahashi S, Ueda M, Tanaka A, Okada H, Morikawa Y, Kawaguchi T, Arai M, Fukuda H, Kondo A (2002) Direct and efficient production of ethanol from cellulosic material with a yeast strain displaying cellulolytic enzymes. Appl Environ Microbiol 68:5136–5141 Fujita Y, Ito J, Ueda M, Fukuda H, Kondo A (2004) Synergistic saccharification, and direct fermentation to ethanol, of amorphous cellulose by use of an engineered yeast strain codisplaying three types of cellulolytic enzyme. Appl Environ Microbiol 70:1207–1212

Appl Microbiol Biotechnol (2010) 87:1841–1853 Ghaemmaghami S, Huh WK, Bower K, Howson RW, Belle A, Dephoure N, O’Shea EK, Weissman JS (2003) Global analysis of protein expression in yeast. Nature 425:737–741 Gietz RD, Schiestl RH (2007) High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method. Nat Protoc 2:31– 34 Haas LO, Cregg JM, Gleeson MA (1990) Development of an integrative DNA transformation system for the yeast Candida tropicalis. J Bacteriol 172:4571–4577 Hisamatsu M, Furubayashi T, Karita S, Mishima T, Isono N (2006) Isolation and identification of a novel yeast fermenting ethanol under acidic conditions. J Appl Glycosci 53:111–113 Hisatomi T, Kuroyanagi S, Tsuboi M (1998) Identification of the orotidine-5′-phosphate decarboxylase gene and development of a transformation system in the yeast Saccharomyces exiguus Yp74L-3. Biosci Biotechnol Biochem 62:2280–2282 Hong J, Wang Y, Kumagai H, Tamaki H (2007) Construction of thermotolerant yeast expressing thermostable cellulase genes. J Biotechnol 130:114–123 Katemai W, Maneerat S, Kawai F, Kanzaki H, Nitoda T, H-Kittikun A (2008) Purification and characterization of a biosurfactant produced by Issatchenkia orientalis SR4. J Gen Appl Microbiol 54:79–82 Kawaguchi T, Enoki T, Tsurumaki S, Sumitani J, Ueda M, Ooi T, Arai M (1996) Cloning and sequencing of the cDNA encoding β-glucosidase 1 from Aspergillus aculeatus. Gene 173:287– 288 Koh JH, Suh HJ (2009) Biological activities of thermo-tolerant microbes from fermented rice bran as an alternative microbial feed additive. Appl Biochem Biotechnol 157:420–430 Li Y, Shen W, Wang Z, Liu JQ, Rao Z, Tang X, Fang H, Zhuge J (2005) Isolation and sequence analysis of the gene URA3 encoding the orotidine-5′-phosphate decarboxylase from Candida glycerinogenes WL2002-5, an industrial glycerol producer. Yeast 22:423–430 Lynd LR, Weimer PJ, van Zyl WH, Pretorius IS (2002) Microbial cellulose utilization: fundamentals and biotechnology. Microbiol Mol Biol Rev 66:506–577 Mandels M, Andreotti R, Roche C (1976) Measurement of saccharifying cellulase. Biotechnol Bioeng Symp 6:21–33 Nonklang S, Abdel-Banat BM, Cha-aim K, Moonjai N, Hoshida H, Limtong S, Yamada M, Akada R (2008) High-temperature ethanol fermentation and transformation with linear DNA in the thermotolerant yeast Kluyveromyces marxianus DMKU3-1042. Appl Environ Microbiol 74:7514–7521 Nonklang S, Ano A, Abdel-Banat BM, Saito Y, Hoshida H, Akada R (2009) Construction of flocculent Kluyveromyces marxianus strains suitable for high-temperature ethanol fermentation. Biosci Biotechnol Biochem 73:1090–1095 Prillinger H, Molnár O, Eliskases-Lechner F, Lopandic K (1999) Phenotypic and genotypic identification of yeasts from cheese. Antonie Van Leeuwenhoek 75:267–283 Rajoka MI, Shaukat F, Ghauri MT, Shahid R (2003) Kinetics of βglucosidase production by Saccharomyces cerevisiae recombinants harboring heterologous bgl genes. Biotechnol Lett 25:945– 948 Rajoka MI, Ferhan M, Khalid AM (2005) Kinetics and thermodynamics of ethanol production by a thermotolerant mutant of Saccharomyces cerevisiae in a microprocessor-controlled bioreactor. Lett Appl Microbiol 40:316–321 Sakai Y, Kazarimoto T, Tani Y (1991) Transformation system for an asporogenous methylotrophic yeast, Candida boidinii: cloning of the orotidine-5′-phosphate decarboxylase gene (URA3), isolation of uracil auxotrophic mutants, and use of the mutants for integrative transformation. J Bacteriol 173:7458–7463

Appl Microbiol Biotechnol (2010) 87:1841–1853 Sánchez M, Iglesias FJ, Santamaría C, Domínguez A (1993) Transformation of Kluyveromyces lactis by electroporation. Appl Environ Microbiol 59:2087–2092 Schwarz WH, Gräbnitz F, Staudenbauer WL (1986) Properties of a Clostridium thermocellum endoglucanase produced in Escherichia coli. Appl Environ Microbiol 51:1293–1299 Seo SH, Rhee CH, Park HD (2007) Degradation of malic acid by Issatchenkia orientalis KMBL 5774, an acidophilic yeast strain isolated from Korean grape wine pomace. J Microbiol 45:521– 527 Shaw AJ, Podkaminer KK, Desai SG, Bardsley JS, Rogers SR, Thorne PG, Hogsett DA, Lynd LR (2008) Metabolic engineering of a thermophilic bacterium to produce ethanol at high yield. Proc Natl Acad Sci U S A 105:13769–13774 Shimoda C, Itadani A, Sugino A, Furusawa M (2006) Isolation of thermotolerant mutants by using proofreading-deficient DNA polymerase delta as an effective mutator in Saccharomyces cerevisiae. Genes Genet Syst 81:391–397 Sridhar M, Sree NK, Rao LV (2002) Effect of UV radiation on thermotolerance, ethanol tolerance and osmotolerance of Saccharomyces cerevisiae VS1 and VS3 strains. Bioresour Technol 83:199–202 Suga M, Hatakeyama T (2001) High efficiency transformation of Schizosaccharomyces pombe pretreated with thiol compounds by electroporation. Yeast 18:1015–1021 Suryawati L, Wilkins MR, Bellmer DD, Huhnke RL, Maness NO, Banat IM (2008) Simultaneous saccharification and fermentation of Kanlow switchgrass pretreated by hydrothermolysis using Kluyveromyces marxianus IMB4. Biotechnol Bioeng 101:894– 902 Tailliez P, Girard H, Millet J, Beguin P (1989) Enhanced cellulose fermentation by an asporogenous and ethanol-tolerant mutant of Clostridium thermocellum. Appl Environ Microbiol 55:207– 211 Thalagala TATP, Kodama S, Mishima T, Isono N, Furujyo A, Kawasaki Y, Hisamatsu M (2009) Study on ethanol fermentation using D-glucose rich fractions obtained from lignocelluloses by a two-step extraction with sulfuric acid and Issatchenkia orientalis MF 121. J Appl Glycosci 56:7–11 Tokuhiro K, Ishida N, Kondo A, Takahashi H (2008) Lactic fermentation of cellobiose by a yeast strain displaying beta-

1853 glucosidase on the cell surface. Appl Microbiol Biotechnol 79:481–488 Tomás-Pejó E, García-Aparicio M, Negro MJ, Oliva JM, Ballesteros M (2009) Effect of different cellulase dosages on cell viability and ethanol production by Kluyveromyces marxianus in SSF processes. Bioresour Technol 100:890–895 Van Bogaert IN, De Maeseneire SL, De Schamphelaire W, Develter D, Soetaert W, Vandamme EJ (2007) Cloning, characterization and functionality of the orotidine-5'-phosphate decarboxylase gene (URA3) of the glycolipid-producing yeast Candida bombicola. Yeast 24:201–208 van Zyl WH, Lynd LR, den Haan R, McBride JE (2007) Consolidated bioprocessing for bioethanol production using Saccharomyces cerevisiae. Adv Biochem Engin/Biotechnol 108:205–235 Wang Z, Zhuge J, Fang H (1999) A new osmotolerant and glycerolhighly-producing species—Candida glycerolgenesis Zhuge sp. nov. Wei Sheng Wu Xue Bao 39:68–74 Wang X, Li G, Deng Y, Yu X, Chen F (2005) A site-directed integration system for the nonuniversal CUG(Ser) codon usage species Pichia farinosa by electroporation. Arch Microbiol 184:419–424 Watanabe I, Nakamura T, Shima J (2009) Characterization of a spontaneous flocculation mutant derived from Candida glabrata: a useful strain for bioethanol production. J Biosci Bioeng 107:379–382 Yamagata K, Kato J, Shimamoto A, Goto M, Furuichi Y, Ikeda H (1998) Bloom's and Werner's syndrome genes suppress hyperrecombination in yeast sgs1 mutant: implication for genomic instability in human diseases. Proc Natl Acad Sci U S A 95:8733–8738 Yang VW, Marks JA, Davis BP, Jeffries TW (1994) High-efficiency transformation of Pichia stipitis based on its URA3 gene and a homologous autonomous replication sequence, ARS2. Appl Environ Microbiol 60:4245–4254 Zhou J, Dong Z, Liu L, Du G, Chen J (2009) A reusable method for construction of non-marker large fragment deletion yeast auxotroph strains: a practice in Torulopsis glabrata. J Microbiol Methods 76:70–74 Zhuge J, Fang HY, Wang ZX, Chen DZ, Jin HR, Gu HL (2001) Glycerol production by a novel osmotolerant yeast Candida glycerinogenes. Appl Microbiol Biotechnol 55:686–692