One-pot enzymatic synthesis of deoxy-thymidine-diphosphate (TDP)-2-deoxy-α-d-glucose using phosphomannomutase

Article abstract of DOI:10.1016/j.molcatb.2009.11.008

Production of deoxy-thymidine-diphosphate (TDP)-sugars as substrates of glycosyltransferases, has been one of main hurdles for combinatorial antibiotic biosynthesis, which combines sugar moiety with aglycon of various antibiotics. Here, we report the one-

Full text of DOI:10.1016/j.molcatb.2009.11.008

Journal of Molecular Catalysis B: Enzymatic 62 (2010) 282–287
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
One-pot enzymatic synthesis of deoxy-thymidine-diphosphate
(TDP)-2-deoxy-␣-d-glucose using phosphomannomutase
Yung-Hun Yanga, Young-Bok Kanga, Dae-Hee Kimb, Tek-Hyung Leea, Sung-Hee Parka, Kwangwon Leea,
Dongwon Yooa, Kwang-kyung Liouc, Hee-Chan Leec, Jae-Kyung Sohngc, Byung-Gee Kima,∗
a School of Chemical and Biological Engineering. Seoul National University, Seoul, Republic of Korea
b Genechem INC, Yoosung-ku, Daejeon, Republic of Korea
c Institute of Biomolecule Reconstruction Lab, Sunmoon University, Chunan, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
Production of deoxy-thymidine-diphosphate (TDP)-sugars as substrates of glycosyltransferases, has been
one of main hurdles for combinatorial antibiotic biosynthesis, which combines sugar moiety with aglycon
of various antibiotics. Here, we report the one-pot enzymatic synthesis of TDP-2-deoxy-glucose employ-
ing high efficient TMP kinase (TMK; E.C. 2.7.2.12), acetate kinase (ACK; E.C. 2.7.1.21), and TDP-glucose
synthase (TGS; E.C. 2.7.7.24) with phosphomannomutase (PMM; E.C. 5.4.2.8). In this study, replacing
phosphoglucomutase (PGM; E.C. 5.4.2) by PMM from Escherichia coli gave four times higher specific
activity on 2-deoxy-6-phosphate glucose, suggesting that the activity on 2-deoxy-glucose-6-phosphate
was mainly affected by PMM activity, not PGM activity. Using an in vitro system starting from TMP
and 2-deoxy-glucose-6-phosphate glucose, TDP-2-deoxy-glucose (63% yield) was successfully synthe-
sized. Considering low productivity of NDP-sugars from cheap starting materials, this paper showed how
production of NDP-sugars could be enhanced by controlling mutase activity.
Received 21 February 2009
Received in revised form
21 November 2009
Accepted 21 November 2009
Available online 27 November 2009
Keywords:
TDP-2-deoxy-glucose
Phosphomannomutase
Enzymatic synthesis
TMP
One-pot reaction
© 2009 Published by Elsevier B.V.
1. Introduction
Many macrolide antibiotics are composed of polyketide lactones
substituted with 6-deoxy-hexose sugars [8], and antibiotic activi-
In the status quo, various antibiotic compounds have been
discovered and widely used [1]. As there are many effective antibi-
otics being found, there are also increasing numbers of antibiotic
resistant-bacteria, and the daunting speed of resistance to new
antibiotics seem to exceed the speed of development of new
antibiotics [2,3]. In vivo and in vitro combinatorial biosynthesis,
which combines sugar moiety with a glycon of various antibiotics
and changed the diversity of glycosylated antibiotics, has been
proposed to speed up the development of new antibiotics [4,5].
However, it still has several drawbacks on efficacy, including poor
activity of glycosyltransferase and limited availability of various
activated glycan units [6,7].
ties are typically determined by the number, nature, and relative
distribution of the sugar moieties, which influence the pharma-
cokinetic properties of the drugs [9,10]. Most sugar units can be
attached to aglycon when they are activated by nucleotide diphos-
phates, mainly with deoxy-thymidine-diphosphate (TDP) [11].
To make various TDP-glycans at a large-scale for combi-
natorial biosynthesis, our group and collaborators have been
trying to produce various TDP-sugars enzymatically in an
economic manner, and these attempts were quite successful
with TDP-4-keto-6-deoxy-glucose, TDP-l-rhamonose and TDP-4-
amino-4,6-dideoxy-d-glucose using acetate kinase and thymidine
kinase systems in one pot [12–14].
In this report, TDP-2-deoxy-glucose, the precursor of TDP-
4-keto-2,6-dideoxy-glucose to oliose oleandrose and oliose, was
produced by enzymatic one-pot synthesis from TMP and 2-deoxy-
6-phosphate-glucose, employing acetate kinase (ACK) and TMP
kinase (TMK), phosphomannomutase (PMM), and TDP-glucose
synthase (TGS) (Fig. 1). Introduction of PMM from Escherichia coli
K12 showed four times higher specific activity than previously
reported PGM on 2-deoxy-6-phosphate-glucose with same amount
of substrates [15]. This could increase the conversion from 2-deoxy-
6-phosphate-glucose to 2-deoxy-1-phosphate glucose and make
TDP-2-deoxy-glucose more efficiently.
Abbreviations: TDP, deoxy-thymidine-diphosphate; TMP, deoxy-thymidine-
monophosphate; TMK, deoxy-thymidine-monophosphate kinase; ACK, acetate
kinase; TGS, deoxy-thymidine-diphosphate glucose synthase/glucose-1-phosphate
thymidylytransferase; E. coli, Escherichia coli; PMM, phosphomannomutase; PGM,
phosphoglucomutase; TTP, deoxy-thymidine-triphosphate; NADP, nicotinamide
adenine dinucleotide phosphate; Pi, phosphate; Acetyl Pi, acetyl phosphate; ADP,
adenosine 5ꢀ-diphosphate; ATP, adenosine 5ꢀ-triphosphate.
∗
Corresponding author. Tel.: +82 2 880 6774; fax: +82 2 873 6020.
E-mail address: byungkim@snu.ac.kr (B.-G. Kim).
1381-1177/$ – see front matter © 2009 Published by Elsevier B.V.
doi:10.1016/j.molcatb.2009.11.008

Y.-H. Yang et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 282–287
283
Fig. 1. A diagram of the biosynthesis of TDP-2-deoxy-glucose from 2-deoxy-glucose-6-phosphate and TMP with TMK, ACK, PMM and TGS.
2. Experimental methods
and pelleted. Crude extracts were prepared by the sonication of
the cell suspension. For the enzyme assay, cell extracts were puri-
fied by His-tag agarose beads (Invitrogen, US). TMK, ACK, TGS were
prepared as previously reported [12]. For the multiple enzyme reac-
tion, each gene was expressed under optimal conditions and all
the expressed protein was analyzed by 12% SDS-PAGE (data not
shown).
2.1. Bacterial strains, plasmids and chemical reagents
Chemical reagents used in this study were purchased from
Sigma (MO, USA). E. coli BL21 (DE3) was from Novagen (WI, USA),
and pET24ma was kindly donated by Dr. Hiroshi Sakamoto (Pasteur
Institute, Paris, France). TDP-d-glucose was supplied by GeneChem
(Daejeon, Korea) [12]. PMM was newly amplified using the chro-
mosomal DNA of E. coli K12 as a template and cloned into pET24ma
to give pYH0501. Other bacterial strains and plasmids used in this
work are listed in Table 1.
2.3. Enzyme activity assays
The PMM and TGS activities were identified by determining the
net activity from 2-deoxy-glucose-6-phosphate and TTP. As a con-
trol, TGS from Salmonella typhimurium and phosphoglucomutase
(PGM) from rabbit muscle (Sigma, USA) were used [15]. A reaction
mixture (pH 7.5, volume: 100 ␮l) containing 10 mM TTP, 60 mM
2-deoxy-6-phosphate-glucose, 3 ␮M glucose-1,6-diphosphate,
5 mM MgCl2, 20 mM Tris–HCl buffer, and 0.17 unit of TGS cell
extracts was incubated at 37 ◦C. To compare PGM and PMM activi-
ties, 20 ␮g of PGM and PMM were used with same amount of TGS.
One unit for PGM and PMM was defined as the amount of enzyme
that produces 1 ␮ mol mannose-6-phosphate in 5 min at pH 7.4
and room temperature (15 ◦C), and one unit for TGS was defined
as the amount catalyzing the formation of TDP-2-glucose. Twenty
␮g of PGM and PMM amount to 0.005 and 0.1 units of enzyme,
respectively. The reaction was monitored by HPLC as described
below. Likewise, to examine pH, buffer effect, metal effect and
TMP/2-deoxy-6-phosphate-glucose ratio, the same concentration
and reaction conditions were used with different pH, type of
buffer, metal ion concentration, 2-deoxy-6-phosphate-glucose
concentration, respectively.
2.2. Expression and preparation of enzymes
pYH0501 was transformed into E. coli BL21 (DE3). The cells were
aerobically grown in 50 ml of Luria Bertani (LB) medium. Cells har-
boring each plasmid gene were grown to an OD600 of 0.6 at 30 ◦C,
and then 0.4 mM isopropyl ␤-d-thiogalactopyranoside (IPTG) was
used for the induction. After 12 h, bacterial culture was harvested
Table 1
Plasmids used in this study.
Plasmid
Relevant information
Source or reference
[22]
pET24ma
p15A replication origin, T7 lac promoter,
C-terminal his-tag coding
pYH0504
pYH0503
pYH0502
pYH0505
pET15b carrying PCR product of TMP
kinase (TMK) from E. coli K12
pET24ma carrying PCR product of acetate
kinase (ACK) from E. coli K12
pET24ma carrying PCR product of
TDP-glucose synthase from E. coli K12
pET24ma carrying PCR product of
TDP-glucose synthase from Salmonella
typhimurium
[22]
[22]
[22]
2.4. Kinetic parameters of PMM from E. coli
This work
All experiments were performed as previously published
method [16]. Briefly, for the PMM assay, 1 ␮l of soluble protein
(4.47 × 10−7 M) of purified soluble protein was added to a reaction
buffer consisting of 50 mM Tris–HCl (pH 7.8), 5 ␮l of glucose-
pYH0501
pET24ma carrying PCR product of
Phosphomannomutase (PMM) from E.
coli K12
This work

284
Y.-H. Yang et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 282–287
6-phosphate-dehydrogenase (Roche, Germany), 1 mM ␤-NADP,
0.1 mM D-glucose-1,6-diphosphate, 5 ␮l of phosphomanno iso-
merase (Sigma, USA), 10 ␮l of phosphogluco isomerase (Roche,
Germany), and 100 mM of MgCl2 in a total volume of 100 ␮l.
For the PGM assay, phosphomanno isomerase and phosphogluco
isomerase were left out of the reaction mixture with same volume
of PMM assay. Each assay was measured over a total of 5 min at
25 ◦C with 96-well plate and multiscanner (Thermo Electron Corp.,
Finland).
2.6. Instrumental analysis
Reaction substrates and products were analyzed by HPLC
using an anion exchange analytical column (Carbopac PA1,
4.6 mm × 250 mm, 5 ␮m particle size, Dionex Associates, Nether-
lands) using isocratic 200 mM ammonium dipotassium phosphate
buffer (pH 7.5) with a flow rate of 1 ml/min. The eluted nucleotides
were monitored by absorbance at 254 nm. Retention times at this
condition were as follows: TMP (3.1 min), TDP (12.5 min), TTP
(90.0 min), TDP-2-deoxy-glucose (6.5 min).
To identify the final product, the purified products were dis-
solved in 700 ␮l D2O (99.9 atom % D) and analyzed by NMR. 1H were
recorded with a Varian 300-MHz NMR spectrometer at a probe tem-
perature of 298 K. Chemical shifts were expressed in ppm relative
to an internal HOD signal (4.80).
Mass analysis (ESI-MS) for product was performed with LCQ
Deca XP ion trap mass spectrometer (Thermo Electron Corp, USA)
with standard Thermo ESI source in negative ion mode at the spray
voltage of 5 kV. Sample was infused by syringe pump at a rate of
5 ␮l min−1 and nebulized with dry nitrogen gas. The heated capil-
lary was maintained at a temperature of 270 ◦C. The maximum ion
collection time was set to 50 ms and five microscans were averaged
per scan [17].
2.5. Biosynthesis and purification of TDP-2-deoxy-glucose from
TMP and 2-deoxy-glucose-6-phosphate
Optimization reactions for TDP-2-deoxy-glucose were per-
formed at pH 7.5 using 100 ␮l of 50 mM Tris–HCl buffer
containing TMP, 2-deoxy-glucose-6-phosphate, acetyl phos-
phate, ATP, MgCl2, and cell extracts for four enzymes at
37 ◦C. Ten ␮l of the sample was withdrawn, diluted with
the same buffer (1:10), and the reactions analyzed by HPLC
measurement.
For the preparative synthesis of TDP-2-deoxy-glucose, the reac-
tion was carried out in 50 ml of reaction volume. The initial
concentrations of TMP, ATP, acetyl phosphate, 2-deoxy-glucose-
6-phosphate, glucose-1,6-diphosphate, and MgCl2 were 10 mM,
2 mM, 30 mM, 60 mM, 3 ␮M, and 2 mM, respectively. After reacting
for 3 h, the reaction mixture was boiled for 3 min, and the proteins
were removed from the reaction mixture by ultrafiltration. Then,
the remaining protein was removed by filter (0.45 ␮m, Gelman,
USA). The product was purified using anion exchange (Dowex 1 × 2,
Cl−) and gel filtration (Sephadex G-15) columns (Supplementary
Fig. 1). The lyophilized product was characterized with 1H NMR,
and its purity was analyzed by HPLC [13].
3. Results and discussions
3.1. Preparation of enzymes in E. coli
The molar ratio of glucose-1-phosphate to TMP and activity of
TGS were the most important factors in the production of TDP-4-
keto-6-deoxy-glucose [12] because of the low equilibrium constant
of TGS and the instability of glucose-1-phosphate. To improve
Fig. 2. Optimization of one-pot system. (A) The effects of metal ion concentration, Mg2+ (᭹) and Co2+ (ꢁ) on the production TDP-2-deoxy-glucose. (B) The molar ratio of TMP
and 2-deoxy-6-phosphate-glucose. (C) The effect of pH and D. different types of buffers. Relative activities are based on Tris–HCl buffer, ammonium (ammonium bicarbonate
buffer), BES, HEPES, pi (potassium phosphate buffer), and Tris (Tris–HCl buffer).

Y.-H. Yang et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 282–287
285
the production of TDP-2-deoxy-glucose, most of our attentions
were focused on increasing the rate of producing 2-deoxy-glucose-
1-phosphate from 2-deoxy-glucose-6-phosphate with PGM to
increase the molar ratio of 2-deoxy-glucose-1-phosphate to TTP.
So far, there was only one report on the use of PGM from rab-
bit muscle for the production of 2-deoxy-glucose-1-phosphate [15]
and its kinetic parameters with Kcat/Km = 2.550 min−1 ␮M−1 for
glucose-1-phosphate and, Kcat/Km = 5 min−1 ␮M−1 for mannose-
1-phosphate [18]. This enzyme has high catalytic activity on
glucose-1-phosphate, but less activity on mannose-1-phosphate.
Considering the structural similarities of 2-deoxy-glucose-6-
phosphate to mannose-6-phosphate like glucose-6-phosphate,
phosphomannomutase (PMM), which prefers mannose to PGM,
also could to be another candidate enzyme for phosphate transfer
by mutase. However, there has been no direct report on the kinetic
value of the enzyme reaction of 2-deoxy-glucose-6-phosphate or
2-deoxy-glucose-1-phosphtae by PGM or PMM. For this reason,
PMM from E. coli K12 was searched and examined among easily
accessible strains.
Fig. 3. The reaction profiles of the biosynthesis of TDP-2-deoxy glucose from TMP
and 2-deoxy-6-phosphate-glucose using four enzymes and purification of products.
TMP (᭹), TDP (ꢀ), TTP (ꢁ), TDP-2-deoxy-glucose (ꢁ).
Kinetic
values
of
PMM
were
calculated
as
Kcat/Km = 130 min−1 ␮M−1 for mannose-1-phosphate, but no
activity for glucose-1-phophate. By the introduction of PMM to
the production of TDP-2-deoxy-glucose, it improved the specific
activity on 2-deoxy-glucose-6-phosphate than PGM by almost
four times in a 2 h reaction by monitoring the amount of TDP-
2-deoxy-glucose with HPLC (data not shown), suggesting the
activity on 2-deoxy-6-phosphate-glucose was mainly affected by
phosphomannose activity. It might be possible to add more PGM
for enzymatic synthesis in one pot to increase overall reaction,
however, a small amount of PMM will be preferred.
cloned to compare the activity of TGS from E. coli K12. Under the
conditions described in Experimental methods, the activities of
PMM and PGM with TGSs from different strains with same amount
of substrates were compared, respectively (data not shown).
3.2. Optimization of one-pot reaction
The effects of representative metal ions for overall reaction
were examined to improve productivity. The enzyme activity does
not directly depend on the concentration of enzymes, but the
ratio of the concentrations of the cofactor and the substrate [19].
In another approach to find the genes for TGS, TGS from S.
typhimurium, which has been used in an earlier report [15], was
Fig. 4. 1H NMR results of TDP-2-deoxy-glucose.

286
Y.-H. Yang et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 282–287
Fig. 5. MS/MS spectrum of TDP-2-deoxy-glucose (m/z = 547) in the negative mode.
Fig. 2A shows the optimal metal ion concentration for the over-
all reaction was Mg2+ 0.5 mM. As the main substrate molar ratio
of TMP to 2-deoxy-6-phosphate-glucose increased up to eight, the
amount of synthesized TDP-2-deoxy-glucose increased. Consider-
ing the greater cost of 2-deoxy-6-phosphate-glucose, the optimal
conditions for large-scale production were needed, which had
the maximum value of TDP-2-deoxy-glucose produced/2-deoxy-
6-phosphate supplied (Fig. 2B). The maximum value was gained
when the ratio of 2-deoxy-6-phosphate-glucose/TMP was six. Usu-
ally, the need of a high ratio of TMP/2-deoxy-6-phosphate-glucose
was due to TMP being easily recycled whereas the mutase reaction
is quite slow and rate limiting. In addition, TDP-glucose synthe-
sis reaction is a reversible reaction and the equilibrium constant
appears to be lower than the other two kinases and the instability
of sugar-phosphate in the reaction buffer [12].
2-deoxy-glucose, the purified product was also identified with
300 MHz 1H NMR (Fig. 4) and MS/MS (Fig. 5). 1H NMR (D2O,
300 MHz) 7.57 (s, 1H (B)), 6.16 (app t, 1H, J = 5.7 Hz (C)), 5.52 (app
d, 1H, J = 4.2 Hz (H)), 4.42 (m, 1H (E)), 3.98 (m, 1H (F)), 3.97 (m, 2H
(G)), 3.80 (ddd, 1H, J = 9.1, 8.9 and 4.7 Hz (J)), 3.67 (m, 2H (M)), 3.66
(m, 1H (L)), 3.20 (app t, 1H, J = 8.9 Hz (K)), 2.17 (m, 2H (D)), 2.12 (app
d, 1H, J = 11.8 and 4.7 Hz (I)), 1.73 (s, 3H (A)), 1.51 (ddd, 1H, J = 11.8,
9.1 and 4.2 Hz (I)).
In conclusion, enzymatic synthesis of TMP and 2-deoxy-6-
phosphate-glucose could successfully produce large amount of
TDP-2-deoxy-glucose in one-pot by employing phosphoman-
nomutase, which showed four times higher activity than the
previously reported method [15]. To further improve this sys-
tem, hexose kinase, which can transfer the phosphate group to
2-deoxy-glucose is currently being studied. This enzyme will make
it possible to use a more inexpensive starting material, 2-deoxy-
glucose. Again, by introduction of PMM, the overall production will
be enhanced and cheaper process will be achieved.
For the optimal pH and buffer, many buffers with concentration
of 20 mM such as TES, BES, HEPES, ammonium bicarbonate, phos-
phate, and Tris–HCl, and various pH were examined (Fig. 2C and
D). Among those, Tris–HCl buffer at a pH 8.0 produced the greatest
activity.
Acknowledgment
This work was supported by a grant funded by the Korea gov-
ernment (MOST) (no. R0A-2007-000-10007-0), Republic of Korea.
3.3. One-pot synthesis
Based on all the preliminary experimental results, the one-
pot enzymatic synthesis of TDP-2-deoxy-glucose was performed
(Fig. 3). The product itself was very unstable and easily breakable
to TDP and 2-deoxy-glucose-6-phosphate over time, and as a result,
the reactions were stopped no later than 5 h. By using the one-pot
system with highly active PMM, inhibitions by high concentration
of TDP and TTP [20,21] could be avoided, and separation of TDP and
TDP-2-deoxy-glucose was easier because of high productivity and
low amounts of TDP, which was eluted at a similar time with TDP-
2-deoxy-glucose (Fig. S1a). In addition, the high cost of TTP could
be overcome with one-pot synthesis.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcatb.2009.11.008.
References
[1] A. Paradkar, A. Trefzer, R. Chakraburtty, D. Stassi, Crit. Rev. Biotechnol. 23 (2003)
1–27.
[2] M.J. Rybak, Pharmacotherapy 24 (2004) S203–S215.
[3] S.B. Levy, B. Marshall, Nat. Med. 10 (2004) S122–129.
[4] W.R. Strohl, Biotechnology of Antibiotics, 2nd edition, Marcel-Dekker, Inc.,
1997, pp. 1–20.
[5] C.R. Hutchinson, Curr. Opin. Microbiol. 1 (1998) 319–329.
[6] A. Minami, R. Uchida, T. Eguchi, K. Kakinuma, J. Am. Chem. Soc. 127 (2005)
6148–6149.
By one-pot reaction, 34 mg of products was finally obtained,
which corresponds to a 63% yield based on the initial amount
of TMP. The product was analyzed by HPLC, showing 97% purity,
which has TDP with it as impurity. For the confirmation of TDP-

Y.-H. Yang et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 282–287
287
[7] W.A. Barton, J.B. Biggins, J. Jiang, J.N. Thorson, D.B. Nikolov, Proc. Natl. Acad. Sci.
U.S.A. 99 (2002) 13397–13402.
[15] S. Amann, G. Drager, C. Rupprath, A. Kirschning, L. Elling, Carbohydr. Res. 335
(2001) 23–32.
[8] A.R. Butler, N. Bate, D.E. Kiehl, H.A. Kirst, E. Cundliffe, Nat. Biotechnol. 20 (2002)
713–716.
[9] J.W. Corcoran, M.L. Huber, F.M. Huber, J. Antibiot. (Tokyo) 30 (1977) 1012–1014.
[10] S.A. Fish, E. Cundliffe, J. Antibiot. (Tokyo) 49 (1996) 1044–1048.
[11] M. Perez, F. Lombo, I. Baig, A.F. Brana, J. Rohr, J.A. Salas, C. Mendez, Appl. Environ.
Microbiol. 72 (2006) 6644–6652.
[16] T.R. McCarthy, J.B. Torrelles, A.S. MacFarlane, M. Katawczik, B. Kutzbach, L.E.
Desjardin, S. Clegg, J.B. Goldberg, L.S. Schlesinger, Mol. Microbiol. 58 (2005)
774–790.
[17] Y.H. Yang, Y.B. Kang, K.W. Lee, T.H. Lee, S.S. Park, B.Y. Hwang, B.G. Kim, J. Mol.
Catal. B: Enzym. 37 (2005) 1–8.
[18] R.W. Ye, N.A. Zielinski, A.M. Chakrabarty, J. Bacteriol. 176 (1994) 4851–4857.
[19] S. Fey, L. Elling, U. Kragl, Carbohydr. Res. 305 (1997) 475–481.
[20] L. Elling, Phytochemistry 42 (1996) 955–960.
[12] J. Oh, S.G. Lee, B.G. Kim, J.K. Sohng, K. Liou, H.C. Lee, Biotechnol. Bioeng. 84
(2003) 452–458.
[13] Y.B. Kang, Y.H. Yang, K.W. Lee, S.G. Lee, J.K. Sohng, H.C. Lee, K. Liou, B.G. Kim,
Biotechnol. Bioeng. 93 (2006) 21–27.
[14] Y.S. Chung, D.H. Kim, W.M. Seo, H.C. Lee, K. Liou, T.J. Oh, J.K. Sohng, Carbohydr.
Res. 342 (2007) 1412–1418.
[21] L. Elling, M.R. Kula, Enzym. Microb. Technol. 17 (1995) 929–934.
[22] B.Y. Hwang, H.J. Lee, Y.H. Yang, H.S. Joo, B.G. Kim, Chem. Biol. 11 (2004)
915–925.