Homofermentative production of D-lactic acid from ... - Springer Link

Biotechnol Lett (2012) 34:2069–2075 DOI 10.1007/s10529-012-1003-7

ORIGINAL RESEARCH PAPER

Homofermentative production of D-lactic acid from sucrose by a metabolically engineered Escherichia coli Yongze Wang • Tian Tian • Jinfang Zhao • Jinhua Wang • Tao Yan • Liyuan Xu • Zao Liu • Erin Garza • Andrew Iverson • Ryan Manow • Chris Finan • Shengde Zhou

Received: 4 May 2012 / Accepted: 28 June 2012 / Published online: 13 July 2012 Ó Springer Science+Business Media B.V. 2012

Abstract Escherichia coli W, a sucrose-positive strain, was engineered for the homofermentative production of D-lactic acid through chromosomal deletion of the competing fermentative pathway genes (adhE, frdABCD, pta, pflB, aldA) and the repressor gene (cscR) of the sucrose operon, and metabolic evolution for improved anaerobic cell growth. The resulting strain, HBUT-D, efficiently fermented 100 g sucrose l-1 into 85 g D-lactic acid l-1 in 72–84 h in mineral salts medium with a volumetric productivity of *1 g l-1 h-1, a product yield of 85 % and D-lactic acid optical purity of 98.3 %, and with a minor by-product of 4 g acetate l-1. HBUT-D thus has

great potential for production of D-lactic acid using an inexpensive substrate, such as sugar cane and/or beet molasses, which are primarily composed of sucrose.

Y. Wang T. Tian J. Zhao J. Wang (&) T. Yan L. Xu Z. Liu Key Laboratory of Fermentation Engineering (Ministry of Education), College of Bioengineering, Hubei University of Technology, Wuhan 430068, People’s Republic of China e-mail: [email protected]

Z. Liu e-mail: [email protected]

Y. Wang e-mail: [email protected] T. Tian e-mail: [email protected] J. Zhao e-mail: [email protected] T. Yan e-mail: [email protected] L. Xu e-mail: [email protected]

Keywords D-Lactic acid E. coli Genetic engineering Polylactic acid Sucrose fermentation

Introduction The applications of polylactic acid (PLA) are expanding and include: industrial packaging, biocompatible/

E. Garza A. Iverson R. Manow C. Finan S. Zhou (&) Department of Biological Sciences, Northern Illinois University, DeKalb, IL 60115, USA e-mail: [email protected] E. Garza e-mail: [email protected] A. Iverson e-mail: [email protected] R. Manow e-mail: [email protected] C. Finan e-mail: [email protected]

123

2070

bio-absorbable medical devices, agriculture-oriented and/or motor vehicle-associated plastic supplies. Successful expansion of these applications, however, relies on the thermal tolerance of PLA (Datta and Henry 2006; Garlotta 2001). Prior studies demonstrated that deliberately blending different ratios of poly D-lactic acid (PDLA) and poly L-lactic acid (PLLA) could create a stereo-complex of poly-lactic acid with a thermal tolerance of 230 °C, which is at least 50 °C higher than those of PDLA or PLLA polymers (Garlotta 2001; Tsuji 2002). This finding in turn requires fermentative production of optically pure D-lactic acid and L-lactic acid. The fermentative production of L-lactic acid has advanced while the technology for D-lactic acid, however, remains immature. Multiple lactic acid bacterial strains have been evaluated for the production of D-lactic acid. These lactic acid bacteria, however, have complex nutritional requirements due to their inability to synthesize some amino acids and B vitamins (Hofvendahl and HahnHagerdal 2000). These expensive complex nutrients further increase the production cost because more downstream processing steps are needed for product recovery. With simple nutritional requirements and often achieving fermentation yields of greater than 90 %, Escherichia coli has been engineered for the production of D-lactic acid (Chang et al. 1999; Mazumdar et al. 2010; Shukla et al. 2004; Zhou et al. 2003; Zhu et al. 2007). However, the starting strains used in these studies, E. coli K12, B, C, W3110 and/or their derivatives, are unable to utilize sucrose (Scr-), the most abundant disaccharide available, as an inexpensive fermentation substrate. Less than 50 % of E. coli strains are Scr?. Of those Scr? strains, many of them contain a plasmid that bears the genes for sucrose utilization (Jahreis et al. 2002). Plasmid stability limits their application of these strains for fermentative production of bulk chemicals from sucrose. Nevertheless, E. coli W, a Scr? strain (Lee et al. 1997), contains the sucrose utilization genes on the chromosome (Jahreis et al. 2002; Sahin-Toth et al. 1999). It would be advantageous to engineer E. coli W for the production of D-lactic acid from an inexpensive sucrose carbon source such as sugarcane and/or beet molasses.

123

Biotechnol Lett (2012) 34:2069–2075

Materials and methods Strains, plasmids, media and growth conditions Escherichia coli W (ATCC 9637) was used to engineer a D-lactic acid-producing strain, E. coli HBUT-D (DfrdBC, DadhE, Dpta, DpflB, DcscR, DaldA), using the following described genetic procedures, and the plasmids (pKD4, pKD46 and pFT-A) described by Datsenko and Wanner (2000) and Posfai et al. (1999). During strain construction, the cultures were grown at 37 °C in modified Luria–Bertani broth (per liter: 10 g Tryptone, 5 g yeast extract, and 5 g NaCl), on LB plates (2 % w/v agar); or in NBS mineral salts medium (Zhou et al. 2005), or on NBS plates (2 % agar) supplemented with 2 % (w/v) glucose or sucrose. Antibiotics were added as needed at the following concentrations: ampicillin, 50 lg ml-1; kanamycin, 50 lg ml-1. Genetic methods Standard methods were used for DNA manipulation. Methods for chromosomal gene deletion, integration, and removing antibiotic resistance cassettes using the k Red recombinase method (Datsenko and Wanner 2000) and the FRT-based flippase (Posfai et al. 1999) technology, have been adopted from previous studies (Zhou et al. 2003, 2010). Briefly, a hybrid primer pair was designed with primer-N (65 bp) containing 45 bp of the N-terminal end of the deleting target gene and a 20 bp P1 sequence of pKD4 (P1-FRT-kan-FRT-P2); primer-C (65 bp) containing 45 bp complementary to the C-terminal end of the deleting target gene and a 20 bp P2 sequence of pKD4 (P1-FRT-kan-FRT-P2) (Zhou et al. 2003). The FRT-kan-FRT cassette was amplified by PCR using this hybrid primer pair and a XmnI digested pKD4 plasmid as the template. After purification (Qiagen PCR purification kit), the amplified DNA was electroporated into pKD46 transformed E. coli W or a derivative by a Micropulser (Bio-Rad) using the vendor’s Ec2 procedure. Kanamycinresistant colonies were selected and verified by PCR. The antibiotic marker (kan) was then removed from the chromosome with flippase by using a temperatureconditional helper plasmid (pFT-A) (Posfai et al. 1999). The chromosomal deletion was verified by analysis of PCR product size and fermentation product profiles.


Fermentations Seed cultures were prepared by inoculating six fresh colonies from NBS plates into 2 l flasks containing 500 ml NBS medium with 2 % (w/v) glucose or sucrose, and incubated at 37 °C, 150 rpm, for 12–15 h to achieve an OD600nm of *1.6 (dry cell wt 1.2 g l-1). Bacteria were harvested by centrifugation (*5,0009g, 5 min, 4 °C). The pellets were resuspended and inoculated (achieving an initial 0.5 OD600nm, dry wt *0.35 g l-1) into a 15 l fermenter (Biotech-15 BS, Biotech, Shanghai, P. R. China) containing 10 l NBS medium with 100 g glucose or sucrose l-1 and 1 mM betaine. The fermentation was carried out at 37 °C, 200 rpm, and pH 7.0. The pH was controlled by automatic addition of a 3.5 M Ca(OH)2 slurry, which was maintained on a magnetic plate to prevent Ca(OH)2 precipitation. All fermentations were carried out at least three times. Analyses Biomass was estimated from the OD600 value; an OD of 1 = 0.75 g dry cell wt l-1. Fermentation samples were taken and acidified with 2 vol 6 % (w/v) H2SO4 to liberate organic acids (lactic acid and other byproducts if any). The acidified samples were centrifuged at *8,0009g for 15 min to remove the cell debris and CaSO4. The supernatant was filtered using a 0.45 lm membrane, and was used for analysis of sugar, organic acids, and optical purity of D-lactic acid. The concentrations of sugars and organic acids were analyzed by HPLC with a refractive index detector and an UV detector using a BioRad HPX 87H column at 45 °C with 4 mM H2SO4 at 0.4 ml min-1 as the mobile phase. Optical isomers of D(-) and L(?)-lactic acids were analyzed by HPLC using a chiral column (Sumitomo OA-5000, Japan).

2071

CO2 and H2), and ethanol (Clark 1989) (Fig. 1b). To convert the mixed acid E. coli W strain into a homofermentative D-lactic acid strain that is able to efficiently ferment sucrose, the competing fermentation pathway genes (pflB, pta, adhE, frdABCD, aldA) were sequentially deleted to block the production of formic acid, acetic acid, ethanol, succinic acid and L-lactic acid, using the k Red recombinase technique and the FRT-based flippase system (Zhou et al. 2010). This resulted in a potential homolactic acid strain HBUT-D1. This strain was further engineered by deleting the CscR regulator gene (cscR) to enhance expression of the sucrose utilization operon (cscA, cscKB), which encodes for sucrose permease (CscB), sucrose-6-phosphate hydrolase (CscA) and fructokinase (CscK) (Fig. 1a) (Jahreis et al. 2002), resulting in strain HBUT-D2. The HBUT-D2 strain was then improved through metabolic evolution to increase anaerobic cell growth. The metabolic evolution was carried out by sequentially transferring 0.5 % of fermentation broth into a new fermentation vessel containing fresh nutrients at 48 h interval for 4 weeks, then at 24 h interval for 4 weeks. A single colony was then isolated from the evolved fermentation broth, and designated as HBUT-D. Compared to the parent HBUT-D2, the new strain, HBUT-D, had improved maximum cell growth by *75 % in anaerobic growth test using screw-cap test tubes (data not shown). This metabolically evolved strain HBUT-D should, in theory, convert one mol sucrose into four mol D-lactic acid with a theoretic yield of 1.05 g g-1 sucrose (sucrose ) [sucrose-6-P hydrolase] ) glucose 1-P ? fructose; glucose 1-P (and/or fructose) ) [glycolysis] ) 2 pyruvic acid ? 2 NADH; pyruvic acid ? NADH ) [lactate dehydrogenase] ) D-lactic acid; summary stoichiometry: sucrose ) 4 D-lactic acid) (Fig. 1b). Fermentation of glucose

Results and discussion Construction of a homolactic acid fermentation pathway Escherichia coli carries out mixed acid fermentation under anaerobic conditions, resulting in the production of redox-neutral lactic acid and acetic acid, oxidized succinic acid and formic acid (further converted to

The engineered strain, E. coli HBUT-D, was initially evaluated for D-lactic acid production using 100 g glucose l-1 as the substrate, with E. coli W as the control. As shown in Fig. 2a, b, E. coli W produced a mixture of lactic acid, succinic acid, acetic acid, ethanol and formic acid, but was unable to complete the fermentation. Although D-lactic acid, produced at 33 g l-1, represented *55 % of the fermentation products, a significant amount of succinic acid

123

2072 Fig. 1 Engineering a homolactic acid pathway via sucrose fermentation under anaerobic conditions. a Sucrose operon in E. coli W; b engineered sucrose-toD-lactic acid pathway. Genes encoding important enzymes are indicated by italics. The relevant gene encoding enzymes are: cscB, sucrose permease; cscK, fructokinase; cscA, sucrose hydrolase (invertase); cscR, sucrose operon repressor. The abbreviated metabolic intermediates are: G3P glyceraldehyde-3phosphate, DHAP dihydroxy acetone phosphate, PEP phosphoenolpyruvate


A Sucrose utilization operon cscB

cacK

cscA

cscR

Operator (CscR binding site)

B Sucrose fermentation in E. coli W and engineered HBUT-D Extracellular sucrose

Glucose-1-P + Fructose

Sucrose

cscK

cscA

Fructose-6-P

Glycolysis

G3P

mgsA

DHAP

Methylglyoxal

NAD

Biosynthesis

aldA

NADH

ppc

Oxaloacetate

PEP

L-lactate

ADP

pps

Malate

lldD

ATP

dld

Pyruvate Fumarate frdABCD

pflB

NAD

NADH

aceEF-lpd

Succinate

D-lactate

ldhA

NAD

H2

Formate

NADH

CO2

Acetyl-CoA NADH pta

adhE NAD

Acetyl-phosphate

Acetaldehyde

ADP

ackA

NADH

adhE NAD

ATP

Acetate

Biosynthesis

Ethanol

Extracellular D-lactate

(*20 %) and acetic acid (*19 %) was also seen. These by-products not only diverted substrate carbon from lactic acid, but also could complicate product recovery due to additional steps required to separate the by-products from D-lactic acid, showing the necessity for genetic modification of E. coli W pathways for D-lactic acid production. As demonstrated in Fig. 2c, d, sequential deletions of the competing pathway genes successfully

123

eliminated production of ethanol, formic acid and succinic acid; and significantly minimized production of acetic acid, resulting in a homolactic acid fermentation strain, HBUT-D, with a balanced redox potential (NADH/NAD?) during anaerobic fermentation. HPLC analysis showed that the fermentation of 100 g glucose l-1 was completed and produced 86 g lactic acid l-1, achieving a yield of *90 % based on sugar metabolized. This lactic acid titer was


C 100

Glucose and lactate (g L -1)

Glucose and lactate (g L -1)

A 75 50 25 0 0

10

20

30

40

50

60

70

100 75 50 25 0

80

0

10

20

30

Time (h)

40

50

60

70

80

60

70

80

B

12

D

12

10

By-products (g L -1)

Time (h)

By-products (g L -1)

Fig. 2 Fermentation of 100 g glucose l-1. a E. coli W, glucose utilization and lactate production; b E. coli W, by-product production. c HBUT-D, glucose utilization and lactate production; d HBUT-D, byproduct production; Symbols: open circle, glucose; filled circle, lactate; open square, ethanol; filled square, acetate; multiple sign, succinate; plus sign, formate. Each data point is the average of three or more replicates with error bar representing SD

2073

10

8 6 4 2 0

8 6 4 2 0

0

10

20

30

40

50

60

70

80

0

Time (h)

Fermentation of sucrose

30

40

50

theoretical yield based on sucrose used. Although the lactic acid titer was about the same (86 vs 85 g l-1) in 100 g l-1 of glucose and sucrose fermentations, the maximum volumetric lactic acid production rate of *1.0 g l-1 h-1 in sucrose was 70 % lower than that (*1.7 g l-1 h-1) of glucose fermentation (Fig. 2c), demonstrating that the sucrose uptake and/or its breakdown into glucose 1-P and fructose remains as the rate limiting step of the fermentation process. Nevertheless, the amount of by-product acetate produced (4 g l-1) in sucrose fermentation was about the same as in glucose fermentation (3.5 g l-1).

100

75

50

25

0 0

Escherichia coli HBUT-D was further evaluated for D-lactic acid production using sucrose as a substrate. As shown in Fig. 3, HBUT-D efficiently fermented 100 g sucrose l-1 to completion in 84–96 h, produced 85 g lactic acid l-1, and achieved an 85 % maximum

20

Time (h)

Sucrose and products (g L-1)

*260 % of that produced by parent E. coli W. Although, lactic acid production was slower than that of the parent strain during the initial 16 h, a volumetric -1 -1 D-lactic acid production rate of 1.7 g l h was obtained after 16 h and was maintained for 44 h (16–60 h) (Fig. 2a, c). As Fig. 2d indicates, *3.5 g acetic acid l-1 was produced as a by-product by HBUT-D, even though the normal acetic acid fermentation pathway was blocked by deletion of the pta gene (encoding for acetyl-phosphate transferase) (Fig. 1b). This additional acetic acid production might be attributed to acs and/or poxB genes which encode for acetate synthase (Acs) and pyruvate oxidase (PoxB), respectively (Clark 1989). Acs converts acetate into pyruvate, enabling E. coli to use acetate as a carbon source for cell growth. In HBUT-D, however, the reverse reaction of Acs might convert pyruvate into acetate during fermentation. In addition, PoxB can oxidize pyruvate into acetate during the stationary phase.

10

8 16 24 32 40 48 56 64 72 80 88 96

Time (h)

Fig. 3 Fermentation of 100 g sucrose l-1. Symbols: open circle, sucrose; filled circle, lactate; filled square, acetate. Each data point is the average of three or more replicates with error bar representing SD

123

2074

About 85 g lactic acid l-1 was produced in both glucose and sucrose fermentations, although the time for this varied. This result may indicate the tolerance limitation of HBUT-D is around 85 g lactic acid l-1. To verify this observation, 120 g sucrose l-1 fermentation was carried out with an extended fermentation time (168 h). No significant increase in lactic acid production was achieved and resulted in incomplete fermentation (data not shown). This confirmed that improving lactic acid tolerance is needed to achieve production of 100 g lactic acid l-1 by HBUT-D, a target titer for economic production of D-lactic acid on a large scale. Optical purity of D-lactic acid Although there are three lactate dehydrogenase genes (dld, lldD, ldhA) in E. coli (Fig. 1b), only ldhA is active, enabling E. coli cells to convert pyruvic acid to D-lactic acid under anaerobic fermentative conditions (Clark 1989). Nevertheless, in our previous study, we noticed that the methylglyoxal bypass could also convert dihydroxyacetone-phosphate into L-lactic acid (and D-lactic acid), resulting in 5 % L-lactic acid contamination of the D-lactic acid produced by the engineered strain (Grabar et al. 2006). Therefore, the aldA gene was deleted during strain construction in order to block the conversion of methylglyoxal to L-lactic acid. The engineered strain HBUT-D should only produce D-lactic acid in theory during anaerobic fermentation. To evaluate the effectiveness of this design, the fermented medium broth was subjected to HPLC analysis to determine the optical purity of D-lactic acid produced. The results showed that the D-lactic acid isomer accounted for 98.3 % of the total lactic acid produced, with a small but detectable amount (1.7 %) of L-lactic acid isomer. We suspect that the L-lactic acid was produced via the ‘‘dld-lldD pathway’’, which encodes enzymes to convert D-lactic acid to pyruvic acid (Dld) and then convert the pyruvic to L-lactic acid (LldD) (Fig. 1b), although both genes are supposed to be repressed under anaerobic fermentative conditions.

Conclusion A D-lactic acid-producing strain was engineered from E. coli W by deleting the competing fermentation

123


pathway genes (pflB, pta, adhE, frdABCD) as well as the cscR and aldA genes. The resulting strain, HBUTD, efficiently fermented sucrose into D-lactic acid with an optical purity of 98 %. Based on our knowledge, this is the first engineered E. coli W strain that can efficiently convert sucrose into optically pure D-lactic acid in mineral salts medium. For large scale D-lactic acid fermentation, however, further improvements to this strain are expected to (1) increase lactic acid tolerance through adaptive evolution; (2) increase optical purity by deleting the dld and lldD genes; (3) eliminate acetate by deleting the poxB and acs genes. Acknowledgments This research was supported by Hubei University of Technology and the Chutian Scholar Program of Hubei province, P. R. China, and the Northern Illinois University, USA.

References Chang DE, Shin S, Rhee JS, Pan JG (1999) Homofermentative production of D- or L-lactate in metabolically engineered Escherichia coli RR1. Appl Environ Microbiol 65:1384– 1389 Clark DP (1989) The fermentative pathways of Escherichia coli. FEMS Microbiol Lett 63(3):223–234 Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci USA 97(12):6640–6645 Datta R, Henry M (2006) Lactic acid: recent advances in products, processes and technologies a review. J Chem Technol Biotechnol 81(7):1119–1129 Garlotta D (2001) A literature review of poly(lactic acid). J Polym Environ 9(2):63–84 Grabar TB, Zhou S, Shanmugam KT, Yomano LP, Ingram LO (2006) Methylglyoxal bypass identified as source of chiral contamination in L(?) and D(-)-lactate fermentations by recombinant Escherichia coli. Biotechnol Lett 28:1527– 1535 Hofvendahl K, Hahn-Hagerdal B (2000) Factors affecting the fermentative lactic acid production from renewable resources. Enzyme Microb Technol 26:87–107 Jahreis K, Bentler L, Bockmann J, Hans S, Meyer A, Siepelmeyer J, Lengeler JW (2002) Adaption of sucrose metabolism in the Escherichia coli wild-type strain EC3132. J Bacteriol 184(19):5307–5316 Lee J, Lee SY, Park S (1997) Fed-batch culture of Escherichia coli W by exponential feeding of sucrose as a carbon source. Biotechnol Tech 11(1):59–62 Mazumdar S, Clomburg JM, Gonzalez R (2010) Escherichia coli strains engineered for homofermentative production of D-lactic acid from glycerol. Appl Environ Microbiol 76(13):4327–4336 Posfai G, Koob MD, Kirkpatrick HA, Blattner FC (1999) Versatile insertion plasmids for targeted genome manipulations in bacteria: isolation, deletion, and rescue of the

Biotechnol Lett (2012) 34:2069–2075 pathogenicity island LEE of the Escherichia coli O157:H7 genome. J Bacteriol 179:4426–4428 Sahin-Toth M, Lengyel Z, Tsunekawa H (1999) Cloning, sequencing, and expression of cscA invertase from Escherichia coli B-62. Can J Microbiol 45:418–422 Shukla VB, Zhou S, Yomano LP, Shanmugam KT, Preston JF, Ingram LO (2004) Production of D(-)-lactate from sucrose and molasses. Biotechnol Lett 26:689–693 Tsuji H (2002) Autocatalytic hydrolysis of amorphous-made polylactides: effects of lactide content, tacticity, and enantiomeric polymer blending. Polymer 43(6):1789–1796 Zhou S, Causey TB, Hasona A, Shanmugam KT, Ingram LO (2003) Production of optically pure D-lactic acid in mineral

2075 salts medium by metabolically engineered Escherichia coli W3110. Appl Environ Microbiol 69(1):399–407 Zhou S, Yomano LP, Shanmugam KT, Ingram LO (2005) Fermentation of 10 % (w/v) sugar to D-lactate by engineered Escherichia coli B. Biotechnol Lett 27:1891–1896 Zhou S, Iverson AG, Grayburn WS (2010) Doubling the catabolic reducing power (NADH) output of Escherichia coli fermentation for production of reduced products. Biotechnol Prog 26(1):45–51 Zhu Y, Eiteman MA, DeWitt K, Altman E (2007) Homolactate fermentation by metabolically engineered Escherichia coli strains. Appl Environ Microbiol 73(2):453–464

123