Modulation of guanosine 5′-diphosphate-d -mannose metabolism in recombinant Escherichia coli for production of guanosine 5′-diphosphate-l -fucose

Bioresource Technology 100 (2009) 6143–6148

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Modulation of guanosine 50 -diphosphate-D-mannose metabolism in recombinant Escherichia coli for production of guanosine 50 -diphosphate-L-fucose Won-Heong Lee a, Nam-Soo Han b, Yong-Cheol Park c,*, Jin-Ho Seo a,* a

Department of Agricultural Biotechnology, Seoul National University, Seoul 151-921, Republic of Korea Department of Food Science and Technology, Chungbuk National University, Cheongju 361-763, Republic of Korea c Department of Advanced Fermentation Fusion Science and Technology, Kookmin University, Seoul 136-702, Republic of Korea b

a r t i c l e

i n f o

Article history: Received 1 April 2009 Received in revised form 10 July 2009 Accepted 16 July 2009 Available online 18 August 2009 Keywords: Recombinant Escherichia coli GDP-L-fucose GDP-D-mannose Fed-batch fermentation

a b s t r a c t Guanosine 5’-diphosphate (GDP)-L-fucose, an activated form of a nucleotide sugar, plays an important role in a wide range of biological functions. In this study, the enhancement of GDP-L-fucose production was attempted by supplementation of mannose, which is a potentially better carbon source to be converted into GDP-L-fucose than glucose, and combinatorial overexpression of the genes involved in the biosynthesis of GDP-D-mannose, a precursor of GDP-L-fucose. Supply of a mannose and glucose led to a 1.3-fold-increase in GDP-L-fucose concentration (52.5 ± 0.8 mg l1) in a fed-batch fermentation of recombinant E. coli BL21star(DE3) overexpressing the gmd and wcaG genes, compared with the case using glucose as a sole carbon source. A maximum GDP-L-fucose concentration of 170.3 ± 2.3 mg l1, corresponding to a 4.4-fold enhancement compared with the control strain overexpressing gmd and wcaG genes only, was achieved in a glucose-limited fed-batch fermentation of a recombinant E. coli BL21star(DE3) strain overexpressing manB, manC, gmd and wcaG genes. Further improvement of GDPL-fucose production was not obtained by additional overexpression of the manA gene. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction Human milk contains large amounts of complex oligosaccharides including lactose, d-galactose, N-acetylglucosamine, L-fucose and sialic acid (Boehm and Stahl, 2007; Kunz et al., 2000) and there has been increasing interest in the biological functions of human milk oligosaccharides, especially fucosyloligosaccharides that are able to defend infants against enteric pathogens (Kunz and Rudloff, 2006; Morrow et al., 2004; Newburg et al., 2004). The availability of large amounts of fucosylated oligosaccharides would make them useful as precursors for therapeutic and protective purposes such as the prevention of pathogen infection, improvement of immune system response and reduction of inflammatory processes (Newburg et al., 2005; Severin and Wenshui, 2005). Fucosylated oligosaccharides such as the Lewis blood group antigen can be chemically synthesized by several complicated procedures consisting of multiple protection and deprotection steps (Kretzschmar and Stahl, 1998). In contrast to the complexity of chemical synthesis, enzymatic methods using glycosyltransferases provide more efficient ways for fucosylated oligosaccharide synthesis. Enzymatic fucosylation of oligosaccharides requires guanosine 5’-diphosphate (GDP)-L-fucose as a donor of L-fucose (Albermann et al., 2000; * Corresponding authors. Tel.: +82 2 880 4855; fax: +82 2 873 5095 (J.H. Seo); tel.: +82 2 880 4889; fax: +82 2 873 5260 (Y.-C. Park). E-mail addresses: [email protected] (Y.-C. Park), [email protected] (J.-H. Seo). 0960-8524/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2009.07.035

Bulter and Elling, 1999). However, the high cost of GDP-L-fucose limits its application for large-scale production of fucosyloligosaccharides. The metabolic pathway of GDP-L-fucose synthesis from glucose is found in bacteria, plants and human (Bulter and Elling, 1999). Fructose-6-phosphate, an intermediate of glycolysis, is converted to GDP-L-fucose by way of GDP-D-mannose in reactions catalyzed by five enzymes. For conversion of fructose-6-phosphate to GDPD-mannose, mannose-6-phosphate isomerase (ManA, E.C. 5.3.1.8), phosphomannomutase (ManB, E.C. 5.4.2.8) and mannose-1-phosphate guanyltransferase (ManC, E.C. 2.7.7.22) are required. Further transformation of GDP-D-mannose to GDP-L-fucose is catalyzed by GDP-D-mannose-4, 6-dehydratase (Gmd, E.C. 4.2.1.47) and GDP-Lfucose synthase (WcaG, E.C. 1.1.1.271). The ManC and WcaG-mediated reactions require guanosine 5’-triphosphate (GTP) and NADPH as cofactors, respectively (Albermann et al., 2000; Sullivan et al., 1998). Biosynthesis of GDP-L-fucose has been developed using recombinant microorganisms. A recombinant Saccharomyces cerevisiae system expressing the gmd and wcaG genes from Escherichia coli K12 produced 0.2 mg l1 GDP-L-fucose from galactose (Mattila et al., 2000). Although S. cerevisiae is known to have a high level of cytosolic GDP-D-mannose useful for protein mannosylation (Hashimoto et al., 1997; Kukuruzinska et al., 1987; Romanos et al., 1992), high concentration of GDP-L-fucose was not obtained. Another microbial system for GDP-L-fucose production was established by combination of Corynebacterium ammoniagenes

6144

W.-H. Lee et al. / Bioresource Technology 100 (2009) 6143–6148

producing GTP and recombinant E. coli overexpressing GDP-Lfucose biosynthetic enzymes (Koizumi et al., 2000). A batch type reaction with the resting cells of the two strains and utilization of GMP and mannose as two starting materials resulted in 18.4 g l1 of GDP-L-fucose concentration in 22 h. However, this coupling method requires complicated procedures including cultivation and separation of two microbes and permeabilization of cell membrane. It also requires two expensive substrates, GMP and mannose. Therefore, it is desirable to develop a recombinant microbial system producing GDP-L-fucose from a relatively cheap carbon source such as glucose. In our previous report, a high titer of GDP-L-fucose was obtained in batch and fed-batch fermentations of a recombinant E. coli B strain overexpressing the gmd and wcaG genes (Byun et al., 2007). In addition, overexpression of glucose-6-phosphate dehydrogenase (G6PDH) improved the volumetric productivity of GDP-L-fucose, which indicated that NADPH supplementation would be a critical factor for GDPL-fucose production. In this study, the enhancement of GDP-L-fucose production was attempted by the supplementation of GDP-D-mannose via both supplying mannose as an alternative carbon source and overexpressing the genes related to the biosynthesis of GDP-D-mannose without addition of external GDP-D-mannose. Fed-batch fermentations of recombinant E. coli overexpressing the gmd and wcaG genes were carried out by using mannose or a mixture of mannose and glucose as carbon sources. Various recombinant E. coli systems were constructed for the combinatorial overexpression of the GDPD-mannose biosynthetic enzymes such as ManA, ManB and ManC. The effects of the overexpression of ManA, ManB, and ManC were assessed in a glucose-limited fed-batch fermentation scheme.

pKJmanA was constructed. The manB gene was amplified by using the manB_F and manB_R primers containing the recognition sites of NcoI and EcoRI enzymes, respectively. By digestion of the manB gene and plasmid pETGW with NcoI and EcoRI, and ligation with each other, plasmid pmBGW was constructed. The manC gene was amplified by using the manC_F and manC_R primers containing the recognition sites of NcoI and SacI enzymes, respectively. By digestion of the manC gene and plasmid pETGW with NcoI and SacI, and ligation with each other, plasmid pmCGW was constructed. The polycistronic manC-manB gene cluster was obtained by PCR using the manC_F and manB_R. The amplified manC-manB gene was combined with plasmid pETDuet-1 after their digestion with NcoI and EcoRI. The resulting plasmid was cut again with KpnI and XhoI, and then connected with the manA gene prepared previously, resulting in the construction of the pmABC plasmid. After the digestion of plasmid pmABC with NcoI and SacI, the released manC-manB gene cluster was inserted into plasmid pETGW, and the resulting vector was called pmBCGW. Two PCR primers of T7 + manA_F and T7 + manA_R were used for the amplification of the 1361 bp DNA fragment containing the T7 promoter and manA gene in plasmid pmABC. After digestion of plasmid pmBCGW and the 1,361 bp DNA fragment with HindIII and NotI, DNA fragments with the estimated sizes were ligated and the resulting plasmid was named pmABCGW. Plasmids and primers used and created in this work are listed in Tables 1 and 2, respectively. Expression of the manA, manB, manC, gmd and wcaG genes were controlled by the IPTG-inducible T7 promoter. All constructed plasmids were subjected to DNA sequencing. PCR reactions, general DNA manipulation and bacterial transformation were carried out as described previously (Byun et al., 2007).

2. Methods 2.1. Genetic manipulation

Table 2 DNA oligomers used in this study.

E. coli TOP10 and BL21star(DE3) were used as bacterial hosts for genetic manipulation and GDP-L-fucose production, respectively. Plasmid pETGW containing the dicistronic gmd and wcaG gene cluster was previously constructed using plasmid pETDuet-1 (Byun et al., 2007). The manA, manB and manC genes were obtained by the polymerase chain reactions (PCR) using genomic DNA of E. coli K12 (ATCC 10798) as template. To amplify the manA gene, two PCR primers, manA_F and manA_R, were designed to contain the 5’- and 3’-end regions of the manA gene, and the recognition sites of the KpnI and XhoI restriction enzymes, respectively. The amplified manA gene digested with KpnI and XhoI was cloned into plasmid pACYCDuet-1 cut with the same enzymes and plasmid

PCR primers

Sequencea

manA_F (KpnI)

50 -AGGAATTCGGTACCATGCAAAAACTCATTAACTCAGTG-3´ 5´-AAGGCTCGAGTTACAGCTTGTTGTAAACACG-3´

manA_R (XhoI) manB_F (NcoI) manB_R (EcoRI) manC_F (NcoI) manC_R (SacI)

50 -ACATGCCATGGATGAAAAAATTAACCTGCTTT-3´ 50 -ACCGGAATTCTTACTCGTTCAGCAACGTCAG-3´ 5´-ACATGCCATGGATGGCGCAGTCGAAACTCTAT-3´ 5´-AGTCCGAGCTCTTACACCCGTCCGTAGCGATC-3´

T7 + manA_F (HindIII)

50 -ATACCCAAGCTTTCGAACAGAAAGTAATCGTATTGT-30

T7 + manA_R (NotI)

50 -ATAAGAATGCGGCCGCTTACAGCTTGTTGTAAACACGC-30

a The underlined sequences indicate the corresponding recognition sites of the restriction enzymes as described in Section 2.

Table 1 Strains and plasmids used in this study. Name E. coli strains TOP10 BL21star(DE3) Plasmids pETDuet-1 pACYCDuet-1 pETGW pKJmanA pmABC pmBGW pmCGW pmBCGW pmABCGW

Description

Source

F-mcrA D(mrr-hsdRMS-mcrBC) u80lacZDM15 DlacX74 recA1 araD139 D(ara-leu) 7697 galU galK rpsL (StrR) endA1 nupG  F-ompT hsdSB(r B mB ) gal dcm rne131(DE3)

Invitrogen

Bi-directional T7 promoter, pBR322 replicon (copy number 40), Ampr Bi-directional T7 promoter, p15A replicon (copy number 10–12), Cmr pETDuet-1 + gmd-wcaG (NdeI/XhoI), Ampr pACYCDuet-1 + manA (KpnI/XhoI), Cmr pETDuet-1 + manA (KpnI/XhoI) + manB-manC (NcoI/EcoRI), Ampr pETGW + manB (NcoI/ EcoRI), Ampr pETGW + manC (NcoI/SacI), Ampr pETGW + manB-manC (NcoI/EcoRI), Ampr pmBCGW + T7 promoter and manA (HindIII/NotI), Ampr

Invitrogen Novagen Novagen Byun and others (2007) This study This study This study This study This study This study

W.-H. Lee et al. / Bioresource Technology 100 (2009) 6143–6148

2.2. Culture conditions Batch fermentation was carried out in a 2.5-l bioreactor (KoBiotech, Incheon, Korea) containing 1.0 l LB medium (0.5% (w/ v) yeast extract, 1% (w/v) tryptone, and 1% (w/v) NaCl) with 50 mg l1 ampicillin or 68 mg l1 chloramphenicol at 25 °C, pH 6.8, 700 rpm and 1 vvm. When cell mass reached 0.35 g l1, 0.1 mM isopropyl-b-D-thiogalactopyranoside was added to the culture medium to a final concentration of 0.1 mM broth. After 9 h of induction, the cells were collected and used for measurement of intracellular GDP-L-fucose concentration. Fed-batch cultivation was performed using the 2.5-l bioreactor containing 1.0 l of a synthetic medium (Park et al., 2005) at 25 °C. After the complete depletion of 2% (w/v) glucose or mannose or a mixture of glucose and mannose (1% (w/v) each) added initially, concentrated solutions of 800 g l1 glucose or mannose, or a mixture of glucose and mannose (400 g l1 each) was fed at a constant feed rate of 7.0 or 3.5 or 6.2 g h1, respectively. All feeding solutions contained 20 g l1 MgSO47H2O. When dry cell mass reached approximately 35 g l1, 0.1 mM IPTG was added for the induction of the T7 promoter-mediated gene expression. Agitation speed was controlled up to 1400 rpm to prevent the limitation of dissolved oxygen. Air supply was maintained at 1 vvm throughout the cultivation. The medium pH was controlled at 6.8 by adding 2 N HCl or 28% (w/v) NH4OH. 2.3. Analytical methods Dry cell mass was obtained by the multiplication of a conversion factor of 0.36 (Park et al., 2005) with optical density which was measured using a spectrophotometer (Ultrospec 2000, Amersham Pharmacia Biotech, Piscataway, NJ, USA) at 600 nm. Concentrations of glucose and acetic acid were determined by the M930 high performance liquid chromatography (HPLC) system (Younglin, Seoul, Korea) equipped with an Aminex HPX-87H cation exchange column (Bio-Rad, Richmond, CA, USA) and a RI detector (Knauer, Germany) as described previously (Park et al., 2005). The column was heated at 60 °C. The mobile phase consisting of 5 mM H2SO4 had a flow rate of 0.6 ml min1. To measure intracellular concentration of GDP-L-fucose, the collected cells were resuspended in extraction buffer containing 50 mM Tris–HCl (pH 7.5), 150 mM NaCl, 10 mM MgCl2, 5 mM bmercaptoethanol and 5 mM EDTA. The resuspended cells were disrupted in a FrenchÒPress (Thermo Spectronic, Rochester, NY, USA). After boiling for 1 min and subsequent centrifugation at 15,000g for 20 min, the supernatant was collected and used for determination of intracellular GDP-L-fucose concentration. The intra- and extracellular concentrations of GDP-L-fucose were determined by the HPLC system equipped with a Synergi hydro-RP column (Phenomenex Co, Torrance, CA, USA). A mobile phase, composed of 40 mM potassium phosphate buffer (pH 6.0), 4% (v/v) acetonitrile and 5 mM tetrabutylammonium hydrogen sulfate, flowed at a rate of 0.8 ml min1. Analysis of the protein fractions using sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS–PAGE, 12% polyacrylamide) followed method of Laemmli (1970). The cells were collected at the end of fermentations and their concentration was adjusted to around 7.2 g l1 by appropriate dilution. The cells, resuspended in 50 mM potassium phosphate buffer (pH 7.0), were disrupted by an ultrasonic processor (Cole-Parmer, Vernon Hills, IL, USA). After centrifugation at 15,000g for 20 min, the supernatant (soluble fraction) and debris (insoluble fraction) were collected separately. Ten microliters of the soluble protein fraction (approximately 0.04 mg) and the same volume of the total and insoluble protein fractions were subjected to SDS–PAGE. Proteins were visu-

6145

alized with Coomassie brilliant blue solution, and their images were analyzed using a densitometer.

3. Results 3.1. Fed-batch production of GDP-L-fucose using mannose as an alternative carbon source In our previous research, batch and fed-batch fermentations of recombinant E. coli BL21(DE3)/pETGW were carried out to obtain a high concentration of GDP-L-fucose from glucose as a sole carbon source (Byun et al., 2007). An alternative carbon source, mannose, is converted into mannose-6-phosphate by the phosphoenolpyruvate:sugar phosphotransferase system (PTS) specific for D-mannose (Postma et al., 1993), which can be introduced into the GDP-L-fucose synthetic pathway. A fed-batch fermentation of recombinant E. coli BL21star(DE3)/pETGW was carried out in a synthetic medium containing 20 g l1 mannose initially (Fig. 1a). After the depletion of initially added mannose, the concentrated mannose solution was fed into the reactor at a constant feed rate of 3.5 g h1. Induction of Gmd and WcaG expression by 0.1 mM IPTG addition at 32 h of culture triggered the production of GDP-L-fucose. The final GDP-L-fucose concentration of 47.8 ± 0.7 mg l1 was obtained after 95 h of cultivation. A mixture of 400 g l1 glucose and 400 g l1 mannose was used as another feeding solution in fed-batch fermentation (Fig. 1b). After the complete consumption of 10 g l1 glucose and 10 g l1 mannose added initially, the glucose and mannose mixture was continuously added into the vessel at 6.2 g h1 of feed rate in the fedbatch mode of operation. GDP-L-fucose was produced rapidly by the addition of IPTG at 23 h of culture, resulting in 52.5 ± 0.8 mg l1 GDP-L-fucose concentration and 3.5 ± 0.07 mg l1 h1 productivity in 47 h of fed-batch fermentation. Residual glucose and mannose, and acetate produced were not detectable in the culture broth throughout the fed-batch operation mode. GDP-Lfucose in the culture broth was not detected in HPLC analysis (data not shown).

3.2. Effects of combinatorial expression of ManA, ManB and ManC A key intermediate for GDP-L-fucose production, GDP-D-mannose is synthesized from fructose-6-phosphate by the ManA, ManB and ManC-mediated reactions. Effects of GDP-D-mannose supply on GDP-L-fucose production were investigated by the combinatorial overexpression of ManA, ManB and ManC in batch fermentations. All recombinant E. coli systems overexpressed Gmd and WcaG except for the pmABC system. After 9 h of 0.1 mM IPTG induction, GDP-L-fucose inside the cells was analyzed by HPLC (Fig. 2). All recombinant E. coli systems showed similar dry cell mass. For GDP-L-fucose production, combinatorial expression of the manA, B, C genes exerted different effects on GDP-L-fucose production. Among the five types of expression systems, BL21star(DE3)/pmBCGW gave the best result with 4.4 ± 0.1 mg l1 of GDP-L-fucose concentration and 0.5 ± 0.02 mg l1 h1 of productivity. These results were 33% greater than those obtained with the E. coli system overexpressing Gmd and WcaG only (pETGW). A considerable improvement of GDP-L-fucose production was not obtained with the other strains (E. coli BL21star(DE3) harboring pmBGW or pmCGW or pmABCGW). GDP-L-fucose was not detected in the pmABC system. Hence, a recombinant E. coli system with the simultaneous expression of manB, manC, gmd and wcaG (the pmBCGW system) was determined as the optimal system for GDP-L-fucose production.

6146

W.-H. Lee et al. / Bioresource Technology 100 (2009) 6143–6148

3.3. Production of GDP-L-fucose in glucose-limited fed-batch fermentation

a 100

-1

Dry cell mass / Mannose / Acetate (g l )

80

-1

GDP-L-fucose (mg l )

80 60 60 40 40 20 20

0 0

20

40

60

0 100

80

Time (h)

100

80

80 -1

60

GDP-L-fucose (mg l )

-1

Dry cell mass / Glucose / Mannose / Acetate (g l )

b

60 40 40 20 20

0 0

10

20

30

0 50

40

Time (h) Fig. 1. Fed-batch fermentations of BL21star(DE3)/pETGW using mannose (a) or a mixture of mannose and glucose (b). After the depletion of carbon sources initially added, 800 g l1 mannose or a mixture of 400 g l1 mannose and 400 g l1 glucose were continuously fed into the reactor. Both solutions contain 20 g l1 MgSO4. The arrow indicates the addition of 0.1 mM IPTG. Symbols denote as follows; d, dry cell mass; N, GDP-L-fucose concentration; s, mannose concentration; h, glucose concentration; j, acetate concentration.

To obtain high concentration of GDP-L-fucose, a fed-batch fermentation of the recombinant E. coli strain containing pmBCGW was carried out in a synthetic medium containing glucose as a sole carbon source (Fig. 3). After the batch-wise growth of the cells, cell growth was maintained by continuous feeding of 800 g l1 glucose at 7.0 g h1 of feed rate. Addition of IPTG at 23 h led to a linear elevation of GDP-L-fucose concentration. Fed-batch fermentation of the pmBCGW system for 49 h resulted in 74.2 ± 0.5 g l1 dry cell mass, 170.3 ± 2.3 mg l1 of GDP-L-fucose concentration and 2.3 ± 0.03 mg g cell1 of specific GDP-L-fucose content. Another fed-batch fermentation of recombinant E. coli BL21star(DE3) harboring both pmBCGW and pKJmanA was conducted in order to investigate the effect of additional overexpression of ManA on GDP-L-fucose production. The fed-batch mode was operated in the same way as the pmBCGW system. As a result, the GDP-L-fucose production did not show any improvement with this strain. Overexpression of the five enzymes yielded 73.3 ± 0.8 g l1 dry cell mass and 170.4 ± 1.0 mg l1 GDP-L-fucose after 49.5 h of culture, which were almost the same as the case without ManA overexpression as shown in Fig. 3. To elucidate the expression patterns of the GDP-L-fucose producing enzymes, total, soluble and insoluble protein fractions of the two recombinant E. coli systems containing plasmid pmBCGW or pmBCGW + pKJmanA were analyzed by SDS–PAGE (Fig. 4). At the end of fermentations, protein samples were collected and separated on the gel. Compared with the protein samples obtained before IPTG induction, all enzymes were expressed highly and stably in the recombinant E. coli cells. Expression levels of ManB, ManC, Gmd and WcaG in the strain containing pmBCGW and pKJmanA (lane 2) were almost the same as those for the strain containing pmBCGW only (lane 1). At the estimated size of 42 kDa, soluble ManA was detected only in the pmBCGW + pKJmanA system (lane 2). 4. Discussion Recently, recombinant DNA technology and metabolic engineering have been used to allow the mass production of GDP-L-fucose from cheap sugars as starting materials. Some reports showed that GDP-L-fucose could be synthesized from GDP-D-mannose via

-1

Specific GDP-L-fucose content (mg g cell ) 1

2

3

4

5

80

200

60

150

40

100

20

50

pmBGW pmCGW pmBCGW pmABCGW pmABC

0 0

1

2

3

4

5

GDP-L-fucose concentration (mg l-1) Fig. 2. Effects of the combinatorial overexpression of the GDP-D-mannose biosynthetic enzymes on GDP-L-fucose production. Maximum GDP-L-fucose concentration (j) and specific GDP-L-fucose content (h) were obtained after 9 h of 0.1 mM IPTG induction in the batch fermentations of recombinant E. coli BL21star(DE3) strains. Analysis of GDP-L-fucose was carried out in triplicate.

0

10

20

30

40

-1

pETGW

GDP-L-fucose (mg l )

-1

Dry cell mass / Glucose / Acetate (g l )

0

0 50

Time (h) Fig. 3. Glucose-limited fed-batch fermentation of recombinant E. coli BL21star(DE3)/pmBCGW. After the depletion of glucose added initially, the concentrated glucose solution was fed constantly at 7 g h1 of feed rate. The arrow indicates the addition of 0.1 mM IPTG. Symbols denote as follows; d, dry cell mass; N, GDP-L-fucose concentration; h, glucose concentration; j, acetate concentration.

6147

W.-H. Lee et al. / Bioresource Technology 100 (2009) 6143–6148

Fig. 4. SDS–PAGE analysis of the cell free extracts of recombinant E. coli BL21star(DE3) strains containing pmBCGW (1) or both pmBCGW and pKJmanA (2). The cells were collected at the end of cultivations. Fractionation of total (T), soluble (S) and insoluble (I) proteins was proceeded as described in Section 2. The arrows beside the protein names indicate the corresponding protein bands with the estimated sizes. Lane M indicated 30, 40, 50 and 60 kDa of protein size markers.

the de novo pathway in a wild type of E. coli (Albermann et al., 2000; Bulter and Elling, 1999; Stevenson et al., 1996). Based in this information, recombinant microorganisms such as S. cerevisiae and C. ammoniagenes have been developed for GDP-L-fucose production (Mattila et al., 2000; Koizumi et al., 2000). In our previous report, metabolically engineered E. coli systems were constructed to overexpress two enzymes involved in the de novo GDP-L-fucose biosynthesis and produced GDP-L-fucose from glucose at high productivity (Byun et al., 2007). In this study, modulation of fermentation processes and additional modification of metabolic pathways were made for the further improvement of GDP-L-fucose production. In the pathway of GDP-L-fucose metabolism catalyzed by Gmd and WcaG, supplementation of GDP-D-mannose, a precursor of GDP-L-fucose, seems to be an important factor for GDP-Lfucose production. For enhanced GDP-L-fucose production, two strategies were designed, supplementation of mannose, which is a potentially better carbon source to be converted into GDP-Dmannose than glucose, and combinatorial overexpression of the GDP-D-mannose producing enzymes. The results of GDP-L-fucose production in fed-batch fermentations using the recombinant E. coli strains are summarized in Table 3. For recombinant E. coli BL21star(DE3) overexpressing Gmd and WcaG only, utilization of mannose gave higher values of dry cell mass and GDP-L-fucose concentration than glucose only. However, the productivity of GDP-L-fucose from mannose decreased by 42% relative to glucose, which could be due to the fact that the mannose specific PTS has a reportedly 0.6-fold lower Vmax/Km value for mannose than the glucose specific PTS for glucose (Garcia-Alles et al., 2002). A feeding solution containing glucose and mannose was designed in order to use glucose for cell growth and mannose for GDP-L-fucose production. The fed-batch fermentation using the glucose and

mannose mixture led to a 1.9- or 4.6-fold increase in GDP-L-fucose productivity compared with the cases using glucose or mannose as a sole carbon source, respectively. As another strategy for the supplementation of GDP-D-mannose, the manA, manB and manC genes from E. coli K12 were overexpressed in recombinant E. coli BL21star(DE3) expressing the gmd and wcaG genes. It was reported that ManA-deficient mutants could not synthesize capsular polysaccharides containing L-fucose so, ManA might be essential for the synthesis of GDP-L-fucose (Markovitz et al., 1967). ManB and ManC are known to be crucial enzymes for the biosynthesis of GDP-L-fucose in enteric bacteria such as E. coli and Salmonella enterica (Stevenson et al., 1996; Stevenson et al., 2000). An increase in the expression levels of ManA, ManB and ManC was expected to enhance the carbon flux from glucose to GDP-D-mannose which proceeds further to GDP-L-fucose. In batch fermentations as shown in Fig. 2, the recombinant E. coli system expressing ManB, ManC, Gmd and WcaG gave the best GDP-Lfucose concentration and specific content. Accompanied with Gmd and WcaG expression, expression of ManB or ManC only, or ManA, ManB and ManC produced similar results as the pETGW system. It was reported that kcat/Km values of ManC and Gmd for mannose-1phosphate and GDP-D-mannose were 324 and 18 s1 mM1, respectively (Elling et al., 1996; Somoza et al., 2000). Also, ManB is known to be inhibited by its product, mannose-1-phosphate (Padgett and Phibbs, 1986). Therefore, reduction of the inhibitory effect of mannose-1-phosphate on ManB and concomitant improvement of GDP-D-mannose supplementation might be possible by coexpression of ManC, which shows a 10-fold higher catalytic efficiency than Gmd, thus resulting in the enhanced production of GDP-L-fucose. To verify the effect of ManA overexpression on GDP-L-fucose production and obtain a large amount of GDP-L-fucose, glucoselimited fed-batch fermentations were carried out using recombinant E. coli systems with plasmid pmBCGW or both pmBCGW and pKJmanA (Table 3). Both experiments gave the same fermentation performances in spite of overexpression of ManA as shown in SDS–PAGE analysis (Fig. 4). Phosphomannose isomerase (ManA) catalyzes the reversible conversion of fructose-6-phosphate to mannose-6-phosphate. Its bi-directional conversion action allows mannose-6-phosphate to be used for both energy metabolism and GDP-L-fucose biosynthesis (Jensen and Reeves, 2001; Stevenson et al., 2000). It was also reported that the equilibrium constant for the ManA-catalyzed-reaction was estimated to be about 1.5 in the presence of partially purified phosphoglucose isomerase (PGI) and ManA when either glucose-, or fructose-, or mannose-6-phosphate was added as a substrate (Slein, 1950). Considering the bidirectional reaction and equilibrium constant of ManA, its overexpression would lead to reduction or maintenance of GDP-L-fucose production as performed in the batch and fed-batch fermentations. As a result, the glucose-limited fed-batch fermentation of BL21star(DE3)/pmBCGW resulted in the highest values for GDP-Lfucose production, which were 4.4- and 3.7-fold increases in

Table 3 Summarized results of fed-batch fermentations of recombinant E. coli BL21star(DE3) strains producing GDP-L-fucose. Plasmid

Carbon source

Dry cell mass (g l1)

GDP-L-fucose concentration (mg l1)

GDP-L-fucose productivitya (mg l1 h1)

Specific GDP-L-fucose content (mg g1 cell)

pETGWb

Glucose Mannose Glucose + mannose Glucose Glucose Glucose

60.2 ± 0.5 75.2 ± 0.4 69.8 ± 0.3 74.2 ± 0.5 73.3 ± 0.8 66.7 ± 0.5

38.9 ± 0.6 47.8 ± 0.7 52.5 ± 0.8 170.3 ± 2.3 170.4 ± 1.0 55.2 ± 0.5

1.8 ± 0.05 0.8 ± 0.04 3.5 ± 0.07 6.7 ± 0.09 6.1 ± 0.04 2.5 ± 0.06

0.6 ± 0.01 0.6 ± 0.01 0.8 ± 0.02 2.3 ± 0.03 2.3 ± 0.01 0.8 ± 0.01

pmBCGW pmBCGW + pKJmanA pETGW + pMWzwfb a

GDP-L-fucose productivity was estimated during the GDP-L-fucose production period after IPTG induction. The results of fed-batch fermentation of recombinant E. coli BL21(DE3) strains using glucose only were cited in Byun and others (2007). Plasmid pMWzwf contains the zwf gene coding for glucose 6-phosphate dehydrogenase from E. coli K-12 catalyzing the NADPH regeneration reaction. b

6148

W.-H. Lee et al. / Bioresource Technology 100 (2009) 6143–6148

GDP-l-fucose concentration and productivity, respectively, compared with the control strain overexpressing Gmd and WcaG only (the pETGW system) (Byun et al., 2007). Considering the mannose effect, further improvement of GDP-L-fucose production is expected in a fed-batch fermentation of the optimal E. coli strain harboring pmBCGW using both mannose and glucose as carbon sources. Moreover, the optimized E. coli system containing plasmid pmBCGW produced 3.1 times more GDP-L-fucose and showed a 2.9-fold higher specific content than E. coli BL21star(DE3)/pETGW + pMWzwf overexpressing the gmd, wcaG and zwf genes (Byun et al., 2007). Glucose-6-phosphate dehydrogenase encoded by the zwf gene catalyzes the first step of the oxidative pentose phosphate pathway and regenerates NADPH, which is a key cofactor for GDPL-fucose biosynthesis (Byun et al., 2007). Thus, more research efforts will be made to engineer recombinant E. coli for the sufficient supply of GTP and NADPH, two major cofactors in the ManC- and WcaG-catalyzing reactions, respectively. 5. Conclusion In this study, optimization of microbial systems for the mass production of GDP-L-fucose was carried out by the supplementation of GDP-D-mannose, a key intermediate for GDP-L-fucose biosynthesis. A considerable improvement of GDP-L-fucose production in fed-batch fermentations of recombinant E. coli strains was obtained by the concerted expression of the key enzymes involved in the GDP-D-mannose and GDP-L-fucose biosynthesis as well as feeding of a mannose and glucose mixture. Acknowledgements This research was supported by the Seoul R&BD Program (No. 11000), WCU (World Class University) program through the Korea Science and Engineering Foundation funded by the Ministry of Education, Science and Technology (Grant No. 08-0238) and the Korean Ministry of Commerce, Industry and Energy (Grant No. 10028380-2006-11). References Albermann, C., Distler, J., Piepersberg, W., 2000. Preparative synthesis of GDP-betaL-fucose by recombinant enzymes from enterobacterial sources. Glycobiology 10, 875–881. Boehm, G., Stahl, B., 2007. Oligosaccharides from milk. J. Nutr. 137, 847S–849S. Bulter, T., Elling, L., 1999. Enzymatic synthesis of nucleotide sugars. Glycoconj. J. 16, 147–159. Byun, S.G., Kim, M.D., Lee, W.H., Lee, K.J., Han, N.S., Seo, J.H., 2007. Production of GDP-L-fucose, L-fucose donor for fucosyloligosaccharide synthesis, in recombinant Escherichia coli. Appl. Microbiol. Biotechnol. 74, 768–775. Elling, L., Ritter, J.E., Verseck, S., 1996. Expression, purification and characterization of recombinant phosphomannomutase and GDP-alpha-D-mannose pyrophosphorylase from Salmonella enterica, group B, for the synthesis of GDP-alpha-Dmannose from D-mannose. Glycobiology 6, 591–597.

Garcia-Alles, L.F., Zahn, A., Erni, B., 2002. Sugar recognition by the glucose and mannose permeases of Escherichia coli. Steady-state kinetics and inhibition studies. Biochemistry 41, 10077–10086. Hashimoto, H., Sakakibara, A., Yamasaki, M., Yoda, K., 1997. Saccharomyces cerevisiae VIG9 encodes GDP-mannose pyrophosphorylase, which is essential for protein glycosylation. J. Biol. Chem. 272, 16308–16314. Jensen, S.O., Reeves, P.R., 2001. Molecular evolution of the GDP-mannose pathway genes (manB and manC) in Salmonella enterica. Microbiology-SGM 147, 599– 610. Koizumi, S., Endo, T., Tabata, K., Nagano, H., Ohnishi, J., Ozaki, A., 2000. Large-scale production of GDP-fucose and Lewis X by bacterial coupling. J. Ind. Microbiol. Biotechnol. 25, 213–217. Kretzschmar, G., Stahl, W., 1998. Large scale synthesis of linker-modified sialyl Lewis(X), Lewis(X) and N-acetyllactosamine. Tetrahedron 54, 6341–6358. Kukuruzinska, M.A., Bergh, M.L.E., Jackson, B.J., 1987. Protein glycosylation in yeast. Ann. Rev. Biochem. 56, 915–944. Kunz, C., Rudloff, S., 2006. Health promoting aspects of milk oligosaccharides. Int. Dairy J. 16, 1341–1346. Kunz, C., Rudloff, S., Baier, W., Klein, N., Strobel, S., 2000. Oligosaccharides in human milk: structural, functional, and metabolic aspects. Ann. Rev. Nutr. 20, 699–722. Laemmli, U.K., 1970. Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 27, 680–685. Markovitz, A., Sydiskis, R.J., Lieberman, M.M., 1967. Genetic and biochemical studies on mannose-negative mutants that are deficient in phosphomannose isomerase in Escherichia coli K-12. J. Bacteriol. 94, 1492–1496. Mattila, P., Rabina, J., Hortling, S., Helin, J., Renkonen, R., 2000. Functional expression of Escherichia coli enzymes synthesizing GDP-L-fucose from inherent GDP-Dmannose in Saccharomyces cerevisiae. Glycobiology 10, 1041–1047. Morrow, A.L., Ruiz-Palacios, G.M., Altaye, M., Jiang, X., Guerrero, M.L., Meinzen-Derr, J.K., Farkas, T., Chaturvedi, P., Pickering, L.K., Newburg, D.S., 2004. Human milk oligosaccharides are associated with protection against diarrhea in breast-fed infants. J. Pediatr. 145, 297–303. Newburg, D.S., Ruiz-Palacios, G.M., Altaye, M., Chaturvedi, P., Meinzen-Derr, J., Guerrero, M.D., Morrow, A.L., 2004. Innate protection conferred by fucosylated oligosaccharides of human milk against diarrhea in breastfed infants. Glycobiology 14, 253–263. Newburg, D.S., Ruiz-Palacios, G.M., Morrow, A.L., 2005. Human milk glycans protect infants against enteric pathogens. Ann. Rev. Nutr. 25, 37–58. Padgett, P.J., Phibbs, P.V., 1986. Phosphomannomutase activity in wild-type and alginate-producing strains of Pseudomonas aeruginosa. Curr. Microbiol. 14, 187– 192. Park, Y.C., Kim, S.J., Choi, J.H., Lee, W.H., Park, K.M., Kawamukai, M., Ryu, Y.W., Seo, J.H., 2005. Batch and fed-batch production of coenzyme Q10 in recombinant Escherichia coli containing the decaprenyl diphosphate synthase gene from Gluconobacter suboxydans. Appl. Microbiol. Biotechnol. 67, 192–196. Postma, P.W., Lengeler, J.W., Jacobson, G.R., 1993. Phosphoenolpyruvate – carbohydrate phosphotransferase systems of bacteria. Microbiol. Rev. 57, 543–594. Romanos, M.A., Scoper, C.A., Clare, J.J., 1992. Foreign gene expression in yeast: a review. Yeast 8, 423–488. Severin, S., Wenshui, X., 2005. Milk biologically active components as nutraceuticals: review. Crit. Rev. Food Sci. Nutr. 45, 645–656. Slein, M.W., 1950. Phosphomannose isomerase. J. Biol. Chem. 186, 753–761. Somoza, J.R., Menon, S., Schmidt, H., Joseph-McCarthy, D., Dessen, A., Stahl, M.L., Somers, W.S., Sullivan, F.X., 2000. Structural and kinetic analysis of Escherichia coli GDP-mannose 4,6 dehydratase provides insights into the enzyme’s catalytic mechanism and regulation by GDP-fucose. Structure 8, 123–135. Stevenson, G., Andrianopoulos, K., Hobbs, M., Reeves, P.R., 1996. Organization of the Escherichia coli K-12 gene cluster responsible for production of the extracellular polysaccharide colanic acid. J. Bacteriol. 178, 4885–4893. Stevenson, G., Lan, R.T., Reeves, P.R., 2000. The colanic acid gene cluster of Salmonella enterica has a complex history. FEMS Microbiol. Lett. 191, 11–16. Sullivan, F.X., Kumar, R., Kriz, R., Stahl, M., Xu, G.Y., Rouse, J., Chang, X.J., Boodhoo, A., Potvin, B., Cumming, D.A., 1998. Molecular cloning of human GDP-mannose 4,6dehydratase and reconstitution of GDP-fucose biosynthesis in vitro. J. Biol. Chem. 273, 8193–8202.