A two-stage one-pot enzymatic synthesis of TDP-L-mycarose from thymidine and glucose-1-phosphate

Article abstract of DOI:10.1021/ja0562144

This report describes a procedure combining six enzymes native to Escherichia coli or Salmonella typhi, such as thymidine kinase (TK), thymidylate kinase (TMK), nucleoside diphosphate kinase (NDK), pyruvate kinase (PK; for ATP regeneration), TDP-glucose synthetase (RfbA), and TDP-glucose 4,6-dehydratase (RfbB), with five enzymes from Streptomyces fradiae, such as TylX3, TylC1, TylC3, TylK, and TylC2, that resulted in the biosynthesis of TDP-l-mycarose from glucose-1-phosphate and thymidine. This two-stage one-pot approach can be readily applied to the synthesis of other unusual sugars. Copyright

Full text of DOI:10.1021/ja0562144

Published on Web 01/12/2006
A Two-Stage One-Pot Enzymatic Synthesis of TDP-L-mycarose from
Thymidine and Glucose-1-phosphate
Haruko Takahashi, Yung-nan Liu, and Hung-wen Liu*
DiVision of Medicinal Chemistry, College of Pharmacy and Department of Chemistry and Biochemistry,
UniVersity of Texas at Austin, Austin, Texas 78712
Received September 9, 2005; E-mail: h.w.liu@mail.utexas.edu
Scheme 1
One goal of synthetic chemistry is to prepare a target molecule
by a minimum number of transformations with high stereoselectivity
and efficiency. Despite significant advances made in recent years
in the field of synthetic chemistry, preparation of many organic
compounds still requires multistep chemical reactions. The lengthy
operations can be tedious and time-consuming and can suffer from
low yields. However, as more biosynthetic pathways for natural
products are characterized, the idea of exploiting these pathways
to produce desired products has become increasingly attractive. It
has recently been demonstrated that the biosynthetic machineries
can be manipulated to produce tailor-made new compounds.1
Moreover, due to the high regio- and stereospecificity of enzyme
catalysis, the enzymatic synthesis may be carried out in one pot,2
which is not possible for most chemical reactions. Such an approach
avoids the isolation and/or the accumulation of unstable intermedi-
ates and also minimizes the use of chemicals and production of
waste. The environmentally benign nature of these biosynthetic
approaches is highly attractive.
As a part of our effort to elucidate the biosynthetic pathways of
unusual sugars, we have recently cloned and identified a number
of sugar biosynthetic gene clusters,3 including the entire pathway
for the formation of mycarose (1).4 L-Mycarose is a key component
of the macrolide antibiotic tylosin (2), which is produced by
Streptomyces fradiae. It is also found in a few other clinically useful
antibiotics. To generate new bioactive compounds by derivatization
of structurally diverse macrolide aglycones with L-mycarose using
selected glycosyltransferases, it is necessary to have large quantities
of TDP-L-mycarose (3) readily accessible as it is the sugar substrate
form recognized by the glycosyltransferases. Although L-mycarose
has already been chemically synthesized,5 attempts to prepare TDP-
L-mycarose are hampered by the facile loss of the TDP substituent
at C-1. To circumvent this problem, we explored a biosynthetic
To determine the feasibility of this plan, we purified TylX3,6
TylC1,6 TylC3,7 TylK,4 and TylC24 to near homogeneity, along
with RfbB, a TylA2 homologue in the rhamnose pathway from
Salmonella typhi.9 The reaction was initiated by incubating 4 (7
mM) with RfbB (34 µM) in 50 mM potassium phosphate buffer
(pH 7.5) at 37 °C for 1 h to generate 5.10 This conversion was
nearly quantitative as indicated by HPLC analysis (Figure 1A, peak
a).11 The remaining five mycarose biosynthetic enzymes were then
added to the reaction mixture (30 µM each), together with NADPH
(14 mM) and (S)-adenosyl-L-methionine (SAM, 7 mM). Incubation
was continued at room temperature for 1 h. HPLC analysis11 and
comparison of the isolated material (Figure 1B, peak b) with a
standard (Figure 1C, peak b) showed that TDP-L-mycarose (3) was
the major product. The overall yield of 3 from 4 was approximately
20%.
approach for the preparation of the compound.
As shown in Scheme 1, TDP-L-mycarose (3) is derived from
TDP-D-glucose (4) via six enzymatic reactions.4 The key intermedi-
ate 7 is produced from 4 in three separate steps catalyzed by a
4,6-dehydratase (TylA2, 4 f 5),4 a 2-dehydrase (TylX3, 5 f 6),6
and a 3-reductase (TylC1, 6 f 7).6 Subsequently, 7 is transformed
to 8 by methylation at C-3 by a methyltransferase (TylC3);7 8 is
converted to 9 by epimerization at C-5 by an epimerase (TylK),4
and the reduction of 9 at C-4 by a reductase (TylC2)4 yields TDP-
L-mycarose (3). Two constraints make a one-pot synthesis of 3 a
necessity. First, compound 6 is inherently unstable since incubation
of 5 with TylX3 results in its rapid consumption with the
concomitant formation of TDP and maltol.6,8 Thus, the inclusion
of TylC1 and NADPH in the reaction is required to reduce the
labile 6 in situ to form the more stable product 7. Second, the in
situ reduction catalyzed by TylC2 is necessary to drive the
equilibrium of the epimerization products, 8 and 9, generated in
the TylK reaction to completion, allowing the isolation of TDP-
L-mycarose (3).4
The initial success in obtaining 3 using this method was
encouraging. However, the high cost of TDP-D-glucose (4) limits
its practicality. Hence, we explored the enzymatic generation of 4
for the biosynthesis of 3 in a one-pot preparation. As depicted in
9
1432
J. AM. CHEM. SOC. 2006, 128, 1432-1433
10.1021/ja0562144 CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
1
H2O at 24 mL/h). The identity of 3 was confirmed by H NMR
and high-resolution MS analysis.4
In conclusion, this report describes a procedure combining six
enzymes native to E. coli or S. typhi with five enzymes from S.
fradiae that resulted in the biosynthesis of TDP-L-mycarose from
the readily available precursors, glucose-1-phosphate and thymidine.
This work is significant because (1) it demonstrated that an unstable
TDP-sugar derivative can be successfully prepared by enzymatic
synthesis; (2) while a total of 11 enzymes were used, the preparation
starting from thymidine and glucose-1-phosphate is essentially a
one-pot operation; (3) a convenient enzymatic method for the prep-
aration of TDP-D-glucose, the common precursor for the biosyn-
thesis of many unusual sugars, was developed; and (4) the results
provided further in vitro evidence confirming the TDP-L-mycarose
biosynthetic pathway. In addition, no apparent incompatibility on
the reaction conditions for enzymes used in this multienzyme
synthesis is noted, and no obvious cross-inhibition caused by
substrates/products generated in the incubation is observed. With
the identification of many enzymes for the biosynthesis of a variety
of unusual sugars now complete, the groundwork has been laid for
investigations into the enzymatic preparation of these unusual
sugars, a prerequisite for the in vitro glycodiversification of
secondary metabolites. The two-stage one-pot approach described
here can be readily applied to the synthesis of other unusual sugars.
Figure 1. HPLC analysis of various incubation mixtures. (A) Incubation
of TDP-D-glucose (4) with RfbB to generate 5 (peak a); (B) incubation of
the reaction mixture from (A) and TylX3, TylC1, TylC3, TylK, TylC2, in
the presence of NADPH and SAM to give TDP-L-mycarose (3) (peak b);
(C) TDP-L-mycarose standard (peak b);4 (D) incubation of glucose-1-
phosphate (10) and thymidine (12), PEP, ATP, MgCl2 with TK, TMK, NDK,
PK, and RfbA to produce 4 (peak c); (E) incubation of thymidine (12),
PEP, ATP, MgCl2 with TK, TMK, NDK, and PK to make TTP (11); (F)
incubation of the filtered reaction mixture from (E) with 10 and RfbA to
give 4 (peak c).
Scheme 1, TDP-D-glucose is produced via thymidylylation of 10
by glucose-1-phosphate thymidylyltransferase (TylA1 in the my-
carose pathway of S. fradiae, or RfbA9 in the rhamnose pathway
of S. typhi) using thymidine 5′-triphosphate (TTP, 11) as a cosub-
strate. To further reduce the cost, TTP was generated from thymi-
dine (12) by the action of thymidine kinase (TK), thymidylate kinase
(TMK), and nucleoside diphosphate kinase (NDK), as illustrated
in Scheme 1.12 An ATP regenerating system mediated by pyruvate
kinase (PK) was also included to facilitate the overall conversion.13
To assess the feasibility of making 4 enzymatically in one pot,
a test reaction (12 mL) was carried out in which TK, TMK, NDK,
PK, and RfbA (25 µM each) were incubated with thymidine (12,
20 mM), 10 (80 mM), phosphoenol pyruvate (PEP, 66 mM), ATP
(2 mM), and MgCl2 (33 mM) in 50 mM Tris‚HCl buffer (pH 7.5)
at 37 °C for 6 h.14 However, the yield of 4 (Figure 1D, peak c)15
was low (46% calculated based on 12). This may have resulted
from the hydrolysis of 4. To simplify the situation, the incubation
reaction was conducted in two stages in which TTP was generated
in the first stage and converted to TDP-D-glucose in the second
stage. The first stage enzymes, TK, TMK, NDK, and PK, were re-
moved by ultrafiltration (YM-10 membrane) after thymidine had
been completely consumed (Figure 1E). The filtrate containing TTP
was then incubated with 10 and RfbA at 37 °C for 6 h. The yield
of 4 improved substantially (up to 85%) using this approach (Figure
1F, peak c).
Having solved the TDP-glucose production problem, we pro-
ceeded to synthesize 3 using thymidine and 10 as the starting
materials. The initial reaction mixture contained four enzymes, TK,
TMK, NDK, and PK (75 µM each), 3 mM 12, 12 mM PEP, 0.5
mM ATP, and 10 mM MgCl2 in 100 µL of 50 mM Tris‚HCl buffer
(pH 7.5). After incubation at 37 °C for 10 min, the enzymes were
removed by ultrafiltration (YM-10), and the filtrate was incubated
with 3 mM 10 and 57 µM RfbA at 30 °C for 30 min. This mixture
was subsequently treated with 28 µM RfbB, and the incubation
was continued for 1 h at 37 °C. The mycarose biosynthetic enzymes,
TylX3, TylC1, TylC3, TylK, and TylC2 (30 µM each), together
with 6 mM NADPH and 3 mM SAM, were then introduced into
the above reaction mixture. The final incubation was carried out at
room temperature for 1 h. The yield of 16% for 3 was estimated
based on HPLC analysis. This product was purified by FPLC on a
MonoQ column (0-10 min, H2O; 10-20 min, a linear gradient
from 0 to 280 mM ammonium bicarbonate buffer, pH 7.0) and
desalted using a Sephadex G-10 column (2.5 × 50 cm, eluted with
Acknowledgment. This work was supported by the National
Institutes of Health Grants GM35906 and GM54346.
Supporting Information Available: The construction of the
expression plasmid, the isolation of the expressed enzymes, and
incubation conditions to make 3 are presented. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) He, X.; Liu, H.-w. Annu. ReV. Biochem. 2002, 71, 701-754.
(2) (a) Ozaki, S.-I.; Roessner, C. A.; Stolowich, N. J.; Atshaves, B. P.; Hertle,
R.; Muller, G.; Scott, A. I. J. Am. Chem. Soc. 1993, 115, 7935-7938. (b)
Schoevaart, R.; van Rantwijk, F.; Sheldon, R. A. J. Org. Chem. 2000,
65, 6940-6943. (c) Corbett, K.; Fordham-Skelton, A. P.; Gatehouse, J.
A.; Davis, B. G. FEBS Lett. 2001, 509, 355-360.
(3) (a) Chen, H.; Guo, Z.; Liu, H.-w. J. Am. Chem. Soc. 1998, 120, 9951-
9952. (b) Xue, Y.; Zhao, L.; Liu, H.-w.; Sherman, D. H. Proc. Natl. Acad.
Sci. U.S.A. 1998, 95, 12111-12116. (c) Zhao, Z.; Lin, H.; Liu, H.-w. J.
Am. Chem. Soc. 2005, 127, 7692-7693.
(4) Takahashi, H.; Liu, Y.-n.; Chen, H.; Liu, H.-w. J. Am. Chem. Soc. 2005,
127, 9340-9341.
(5) (a) Toshima, K.; Yoshida, T.; Mukaiyama, S.; Tatsuta, K. Carbohydr.
Res. 1991, 222, 173-188. (b) Lear, M. J.; Hirama, M. Tetrahedron Lett.
1999, 40, 4897-4900.
(6) Chen, H.; Agnihotri, G.; Guo, Z.; Que, N. L. S.; Chen, X. H.; Liu, H.-w.
J. Am. Chem. Soc. 1999, 121, 8124-8125.
(7) Chen, H.; Zhao, Z.; Hallis, T. M.; Guo, Z.; Liu, H.-w. Angew. Chem.,
Int. Ed. 2001, 40, 607-610.
(8) Draeger, G.; Park, S.-H.; Floss, H. G. J. Am. Chem. Soc. 1999, 121, 2611-
2612.
(9) Romana, L. K.; Santiago, F. S.; Reeves, P. R. Biochem. Biophys. Res.
Commun. 1991, 174, 846-852.
(10) An efficient method for the production of TDP-4-keto-6-deoxy-D-glucose
(5) from TMP and 10 has been reported: Oh, J.; Lee, S.-G.; Kim, B.-G.;
Sohng, K.; Liou, K.; Lee, H. C. Biotechnol. Bioeng. 2003, 84, 452-458.
(11) The incubation mixture was loaded on a Dionex PA1 column (5 µm, 4.6
× 250 mm) after ultrafiltration using a YM-10 membrane to remove the
proteins. A linear gradient from 200 to 350 mM ammonium acetate buffer
over 30 min (flow rate 0.6 mL/min) was used to elute the reaction products
(monitored at 267 nm).
(12) See Supporting Information for details.
(13) Hirschbein, B. L.; Mazenod, F. P.; Whitesides, G. M. J. Org. Chem. 1982,
47, 3765-3766.
(14) The workup procedure and the HPLC conditions were the same as those
described in ref 11.
(15) The yield was estimated by dividing the integration of peak c by the
summation of integration of all peaks containing thymidyl substituent
(TMP, TDP, TTP, and 4). TDP-D-glucose (4) could be isolated using a
P2 column (2.5 × 100 cm, elution with H2O at 6 mL/h). Its chromato-
graphic behavior and spectral characteristics are identical to those of a
commercial standard.
JA0562144
9
J. AM. CHEM. SOC. VOL. 128, NO. 5, 2006 1433