Genetically Engineered Wine Yeast Produces a High Concentration of L-Lactic Acid of Extremely High Optical Purity

For mass production of lactic acid, we newly constructed a transgenicwine yeast strain that included six copies of the bovine L-lactatedehydrogenase gene on the genome. On fermentation in inexpensivecane juice-based medium, L-lactate production of this recombinantreached 122 g/liter and the optical purity was 99.9% or higher.

INTRODUCTION

Top

Introduction

Polylactic acid is being developed as a renewable alternativefor conventional petroleum-based plastics. Since it has beenreported that optical purity of L-lactic acid affects physicalcharacteristics such as crystallization, thermostability, biodegradationrate, and performance (4, 14, 25), it is important to establisha processing technology for high purity.

L-Lactic acid is generally produced using lactic acid bacteriasuch as Lactobacillus species. The optical purity of this monomeris not high (approximately 95%), because some lactic acid bacteriahave both L-lactate dehydrogenase (L-LDH) and D-lactate dehydrogenase(D-LDH) genes (9). This purity is not suitable for the highphysical properties of polylactic acid (25). To improve theoptical purity, the separation of optical isomers through crystallizationhas been reported (2, 26), and genetically engineered lacticacid bacteria were also developed (12, 13, 15).

Other hosts, i.e., fungi (17, 24), genetically engineered Escherichiacoli (3, 5, 30), and genetically engineered yeasts (1, 20, 23),have also been investigated as to the production of L-lacticacid. Although several organisms producing lactic acid havebeen discussed in terms of optical purity (29, 30), geneticallyengineered yeasts have not yet been analyzed. In this study,we developed a more recombinant Saccharomyces cerevisiae strainfollowing previous research (10) and examined the optical purityof L-lactic acid. Additionally, fermentation analysis with aninexpensive medium, such as one including an unused resource,would also be significant for producing L-lactic acid of highpurity on an industrial scale. To achieve efficient productionin an inexpensive medium, we examined the lactic acid productivitywith a cane juice-based medium.

Construction of plasmids.

Top

Construction of plasmids.

Maps of the plasmids are shown in Fig. 1, and these vectorswere constructed using pBluescript SKII⁺ (Stratagen, La Jolla,Calif.). The pBBLE-LDHKCB vector (Fig. 1A) consisted of thePDC1 promoter, L-LDH gene, Tn5 bleomycin resistance (Tn5 BLE)gene cassette, PDC5 gene fragment, and SLX4 gene fragment. Thephleomycin resistance gene cassette was Tn5 BLE of bacterialtransposon Tn5 (6), which was fused downstream from the S. cerevisiaecytochrome c gene (CYC1) promoter. The pBG418-LDHKCB vector(Fig. 1B) consisted of the PDC1 promoter, L-LDH gene, APT1 genecassette, PDC6 gene fragment, and cytosolic catalase T1 (CTT1)gene fragment. The kanamycin (G418) resistance gene is the aminoglycosidephosphotransferase gene (7), which confers Geneticin resistance,fused downstream from the S. cerevisiae glyceraldehyde-3-phosphatedehydrogenase 3 (TDH3) promoter. Each DNA fragment, PDC5, SLX4,PDC6, and CTT1, was isolated by PCR using the genomic DNA ofthe S. cerevisiae OC-2T strain as a template.

View larger version (29K):
[in this window]
[in a new window]
FIG. 1. Maps of the plasmid vectors and breeding of transgenic S. cerevisiae. (A) T157 (LDH gene, four copies) strain. The constructed DNA fragment, which was obtained by digesting the pBBLE-LDHKCB vector with SacI and ApaI, was integrated into the PDC5 downstream locus of chromosome XII in the YIBO-7A (LDH gene, two copies) strain, and PDC5 and SLX4 were not disrupted through an integration step. The method used to modify the bovine L-LDH gene and the construction of the YIBO-7A strain were described in detail in the previous report (10). (B) T165 (LDH gene, six copies) strain. The constructed DNA fragment, which was obtained by digesting the pBG418-LDHKCB vector with SacI and KpnI, was integrated into the PDC6 upstream locus of chromosome VII in the T157 (LDH gene, four copies) strain. PDC6 and CTT1 were not disrupted.

Breeding of yeasts.

Top

Breeding of yeasts.

In a previous study, we constructed a recombinant, YIBO-7A strain,including two copies of the bovine LDH gene on the genome (10).For mass production of lactic acid, we newly constructed recombinantswith increased copy numbers of the LDH gene. S. cerevisiae transformationwas performed by the LiAc procedure (11), and each transformantwas selected on YPD medium containing 7.5 µg/ml phleomycin(Sigma, St. Louis, Mo.) or 150 µg/ml G418 (Calbiochem).

We constructed the T157 (LDH gene, four copies) strain by usingthe pBBLE-LDHKCB vector (Fig. 1A). The LDH gene cassette onone side of a chromosome could be duplicated through spore formation,because host strain OC-2T is a diploid and homothallic strain(21). Next, the T165 strain (LDH gene, six copies) was constructedby using the pBG418-LDHKCB vector (Fig. 1B). However, strainT165 could not completely consume the glucose on fermentationanalysis. Following this, we selected the T165 recovered strain(T165R) by ethyl methanesulfonate mutagenesis and this straincompletely consumed the glucose (data not shown).

Effect of increasing copies of the LDH gene.

Top

Effect of increasing copies...

Three recombinants, YIBO-7A (LDH gene, two copies), T157 (LDHgene, four copies), and T165R (LDH gene, six copies), were examinedfor LDH specific activity. LDH specific activity was determinedin freshly prepared extracts as described by Minowa et al. (16).Protein concentrations in cell extracts were determined witha DC protein assay kit (Bio-Rad, Richmond, Calif.). As shownin Fig. 2, the highest activity was observed at 24 h and improvementof the specific activity was observed with an increasing LDHgene copy number on the genome. The T165R (LDH gene, six copies)strain showed high activity (108.2 mU/mg of protein) at 24 h.This is an increase of approximately 2.8 times compared withthe YIBO-7A (LDH gene, two copies) strain.

View larger version (20K):
[in this window]
[in a new window]
FIG. 2. Comparison of LDH specific activity with increasing gene copy number. Shown are the control strain OC-2T with no LDH gene ( ${lozenge}$ ), YIBO-7A strain with two copies of the LDH gene ( ${blacksquare}$ ), T157 strain with four copies of the LDH gene ( ${blacktriangleup}$ ), and T165R strain with six copies of the LDH gene (•). The average and deviation for three independent experiments are presented.

The lactate production of these recombinants was examined inYPD medium (1% yeast extract, 2% Bacto Peptone, 10% D-glucose)containing 3% of sterilized calcium carbonate (wt/vol.). Theculture was performed for 72 h at 30°C, and an inoculumsize was 0.1% PCV (packed cell volume). L-Lactate concentrationswere measured with a Biosensor BF-4 (Oji Keisoku Kiki, Hyogo,Japan). The T165R (LDH gene, six copies) strain was observedto produce 68.0 g/liter of L-lactate (Fig. 3). This productionwas improved by 1.28 times or more compared with that of theYIBO-7A (LDH gene, two copies) strain. Improvement of productionwas observed with increasing LDH gene copy number on the genome,as well as from the LDH specific activity results.

View larger version (17K):
[in this window]
[in a new window]
FIG. 3. Time courses of L-lactate production of four transgenic strains on a flask scale. Shown are the control strain OC-2T with no LDH gene ( ${lozenge}$ ), YIBO-7A strain with two copies of the LDH gene ( ${blacksquare}$ ), T157 strain with four copies of the LDH gene ( ${blacktriangleup}$ ), and T165R strain with six copies of the LDH gene (•). The average and deviation for three independent experiments are presented.

Fermentation analysis in cane juice-based medium.

Top

Fermentation analysis in cane...

To confirm lactate production in inexpensive media, we examinedcane juice-based medium. Cane juice is obtained by squeezingsugar cane and contains glucose, sucrose, and many vitaminsand minerals. This juice was diluted until the sugar level became20%, and then 0.3% yeast extract (wt/vol.) was added. Usinga 1-liter jar-fermenter (Biotto, Tokyo, Japan) with pH control,we examined the lactate production of the 165R strain. As shownin Fig. 4, lactate production reached 122 g/liter, with up to61.0% of the sugar being transformed into lactic acid. Thisproduction was obviously high compared with that of the YIBO-7A(LDH gene, 2 copies) strain (data not shown). Additionally,it was pointed out that intercellular ATP regeneration and theredox balance were important for increasing lactic acid productivity,and oxygen-limited chemostat cultures showed that lactic acid-producingS. cerevisiae strains require oxygen for the generation of ATP(27). In the case of T165R strain fermentation in a jar-fermenter,the microaerobic conditions (aeration at 0.15 liters/min) alsoled to higher proliferation than under anaerobic conditions(data not shown).

View larger version (15K):
[in this window]
[in a new window]
FIG. 4. Fermentation analysis with the S. cerevisiae T165R strain in the cane juice-based medium (sugar concentration, approximately 20%). ${circ}$ , L-lactate; ${blacksquare}$ , ethanol. The jar conditions were kept at 32°C and pH 5.2, with aeration at 0.15 liters/min. The agitation rate was controlled at 60 rpm, and 1 N NaOH was used for neutralization.

Optical purity.

Top

Optical purity.

The optical purity of L-lactic acid produced by the T165R strainwas measured by two independent methods. One involved a BiosensorBF-4 (Oji Keisoku Kiki), and the other involved a Diagnostickit (Roche Diagnostics, Japan; detection limit, 0.01%), andpurity was calculated as follows: optical purity = (L-lacticacid quantity – D-lactic acid quantity)/(L-lactic acidquantity + D-lactic acid quantity). Also, the nonexistence ofL- and D-lactic acid in the media was confirmed in advance.As a result, the optical purity was at least 99.9% in cane juice-basedmedium and this purity was obviously high compared with thatof lactic acid produced by other lactic acid bacteria (Table1).

View this table:
[in this window]
[in a new window]
TABLE 1. Comparison of optical purity of L-lactic acid between the transgenic yeast and lactic acid bacteria

Conclusion.

Top

Conclusion.

We confirmed that a recombinant wine yeast was able to produceL-lactate of high optical purity. With the T165R (LDH gene,six copies) strain, the yield of lactic acid was low comparedwith those of lactic acid bacteria, because ethanol was stillproduced. However, it can be said that there are the followingthree advantages as to the use of this recombinant. First, L-lacticacid of extremely high optical purity can be produced. Second,lactic acid can be produced even if one uses an inexpensivemedium, such as one based on cane juice. Yeast can also be cultivatedat high density as to growth. And last, because yeasts exhibita low pH tolerance, free lactic acid production can be expectedwithout neutralization (10). As to the mass production of L-lacticacid of high optical purity, these results indicate that theuse of this transgenic yeast has several advantages and we expectthis research will lead to further use of transgenic yeasts.

ACKNOWLEDGMENTS

We thank Osamu Saotome, Noriko Yasutani, Tomiko Suzuki, TakashiMatsuyama, and Masana Hirai for valuable discussions. We alsothank Miyoko Imoto, Wakana Takase, and Keiko Uemura for technicalassistance.

FOOTNOTES

* Corresponding author. Mailing address: Biotechnology Laboratory, Toyota Central R&D Labs, Inc., Nagakute-cho, Aichi 480-1192, Japan. Phone: 81-561-63-5375. Fax: 81-561-63-6498. E-mail: e1168{at}mosk.tytlabs.co.jp.

These two authors contributed equally to this work.

REFERENCES

Top