Exploiting nucleotidylyltransferases to prepare sugar nucleotides

Article abstract of DOI:10.1021/ol0630853

(Graph Presented) Enzymatic approaches to prepare sugar nucleotides are gaining in importance and offer several advantages over chemical synthesis including high yields and stereospecificity. We report the cloning, expression, and purification of two new wild-type thymidylyltransferases and observed catalysis with a wide variety of substrates. Significant product inhibition was not observed with the enzymes studied over a 24 h period, enabling the efficient preparation of 15 sugar nucleotides, clearly demonstrating the synthetic utility of these biocatalysts.

Full text of DOI:10.1021/ol0630853

ORGANIC
LETTERS
2007
Vol. 9, No. 5
857-860
Exploiting Nucleotidylyltransferases To
Prepare Sugar Nucleotides
Shannon C. Timmons,† Roy H. Mosher,‡,§ Sheryl A. Knowles,‡ and
David L. Jakeman*,†,‡
Department of Chemistry, Dalhousie UniVersity, Halifax, N.S., Canada B3H 4J3,
College of Pharmacy, Dalhousie UniVersity, Halifax, N.S., Canada B3H 3J5, and
Department of Biological Sciences, Wagner College, Staten Island, New York 10301
daVid.jakeman@dal.ca
Received December 20, 2006
ABSTRACT
Enzymatic approaches to prepare sugar nucleotides are gaining in importance and offer several advantages over chemical synthesis including
high yields and stereospecificity. We report the cloning, expression, and purification of two new wild-type thymidylyltransferases and observed
catalysis with a wide variety of substrates. Significant product inhibition was not observed with the enzymes studied over a 24 h period,
enabling the efficient preparation of 15 sugar nucleotides, clearly demonstrating the synthetic utility of these biocatalysts.
Despite advancements in modern drug discovery, the major-
ity of pharmaceuticals continue to be derived from natural
product scaffolds, many of them glycosylated secondary
metabolites.1,2 It is well-established that the carbohydrate
portions of these molecules are essential for bioactivity as
there exists many cases where aglycons show little or no
bioactivity compared to their glycosylated counterparts.3,4
Although the precise role of the sugar residue varies,
carbohydrates have traditionally been implicated in drug
pharmacokinetics and have also been shown to impart
specific interactions with biological targets.4,5 Recently, the
glycosylation of colchicine, a previously nonglycosylated
natural product, has been shown to not only improve
bioactivity but also change the mechanism of action.6
The in vivo glycosylation of secondary metabolites is
mediated by Leloir-type glycosyltransferase enzymes,7 which
rely on sugar nucleotide donors produced by nucleotidylyl-
transferase enzymes (Scheme 1) that are later modified by
sugar processing enzymes. Recent studies have shown that
numerous glycosyltransferases are promiscuous with respect
to their sugar nucleotide donors, allowing these enzymes to
emerge as key glycorandomization tools.8-11 The first
enzyme evolution study of a glycosyltransferase has also
been recently reported, demonstrating the increased viability
of glycorandomization approaches.12 In order to effectively
make use of Nature’s method of glycosylation to expand
(5) Krˇen, V.; Mart´ınkova´, L. Curr. Med. Chem. 2001, 8, 1303.
(6) Ahmed, A.; Peters, N. R.; Fitzgerald, M. K.; Watson, J. A., Jr.;
Hoffmann, F. M.; Thorson, J. S. J. Am. Chem. Soc. 2006, 128, 14224.
(7) Leloir, L. F. Science 1971, 172, 1299.
(8) Jakeman, D. L.; Borissow, C. N.; Graham, C. L.; Timmons, S. C.;
Reid, T. R.; Syvitski, R. T. Chem. Commun. 2006, 35, 3738.
(9) Yang, M.; Proctor, M. R.; Bolam, D. N.; Errey, J. C.; Field, R. A.;
Gilbert, H. J.; Davis, B. G. J. Am. Chem. Soc. 2005, 127, 9336.
(10) Fu, X.; Albermann, C.; Zhang, C.; Thorson, J. S. Org. Lett. 2005,
7, 1513.
(11) Losey, H. C.; Jiang, J.; Biggins, J. B.; Oberthu¨r, M.; Ye, X.-Y.;
Dong, S. D.; Kahne, D.; Thorson, J. S.; Walsh, C. T. Chem. Biol. 2002, 9,
1305.
(12) Aharoni, A.; Thieme, K.; Chiu, C. P. C.; Buchini, S.; Lairson, L.
L.; Chen, H.; Strynadka, N. C. J.; Wakarchuk, W. W.; Withers, S. G. Nat.
Methods 2006, 3, 609.
† Department of Chemistry, Dalhousie University.
‡ College of Pharmacy, Dalhousie University.
§ Department of Biological Sciences, Wagner College.
(1) Butler, M. S. Nat. Prod. Rep. 2005, 22, 162.
(2) Clardy, J.; Walsh, C. Nature 2004, 432, 829.
(3) Thorson, J. S.; Hosted, T. J., Jr.; Jiang, J.; Biggins, J. B.; Ahlert, J.
Curr. Org. Chem. 2001, 5, 139.
(4) Weymouth-Wilson, A. C. Nat. Prod. Rep. 1997, 14, 99.
10.1021/ol0630853 CCC: $37.00
© 2007 American Chemical Society
Published on Web 02/08/2007

Scheme 1. Primary Metabolic Reaction Catalyzed by Cps2L,
RmlA, and RmlA3 R-D-Glucose 1-phosphate
Thymidylyltransferases
libraries of carbohydrate-containing drug candidates, access
to a wide variety of sugar nucleotides is required. Although
advances in chemical approaches towards synthesizing sugar
nucleotides have been reported,13-15 yields are often low and
controlling the stereoselectivity of direct coupling approaches
has proven problematic.16-18 The in vitro preparation of sugar
nucleotides, through the use of nucleotidylyltransferases, is
quickly emerging as an attractive alternative to chemical
synthesis,19-25 although preliminary studies suggested sub-
strate inhibition as a limiting factor in the use of nucleo-
tidylyltransfersaes.26,27
Figure 1. Structures of sugar 1-phosphates and NTPs used in
nucleotidylyltransferase substrate flexibility studies.
The scope and limitations of the nucleotidylyltransferase-
catalyzed formation of sugar nucleotides was evaluated with
a variety of enzymes, sugar 1-phosphates, and nucleoside
triphosphates. We examined the substrate flexibility of three
wild-type bacterial thymidylyltransferases: Cps2L (Strep-
tococcus pneumoniae R6), RmlA (Streptococcus mutans
UA159), and RmlA3 (Aneurinibacillus thermoaerophilus
DSM 10155). Our selection of sugar 1-phosphates was
composed of five commercially available R-D-sugar 1-phos-
phates diverging in structure from R-D-Glc-1-P28 and one
â-L-sugar 1-phosphate prepared by chemical synthesis (Fig-
ure 1). The R-D-sugar 1-phosphates probe the flexibility of
the three enzymes toward changes in configuration at the
C2 and C4 stereocenters of the various R-D-sugar 1-phos-
phates. In addition, synthetic access to â-L-Fuc-1-P29 pro-
vided an opportunity to study the flexibility of the nucle-
otidylyltransferases toward a substrate with a different
1
conformation since â-L-Fuc-1-P adopts a C4 chair confor-
mation whereas the R-D-sugar 1-phosphates adopt 4C1 chair
conformations. It is important to note that â-L-sugar nucleo-
tides are key substrates for many family 1 (GT-B) glyco-
syltransferases30,31 involved in the biosynthesis of bioactive
natural products and thus access to these substrates is of
particular significance for glycorandomization studies.
A series of assays including six pyranosyl-1-phosphates
and five nucleoside triphosphates (Figure 1) were used to
evaluate the substrate flexibility and assess the synthetic
utility of the Cps2L, RmlA, and RmlA3 thymidylyltrans-
ferases. The enzymatic reactions were conducted using
conditions consistent with those described in the literature
including millimolar substrate concentrations and the pres-
(13) Timmons, S. C.; Jakeman, D. L. Submitted for publication.
(14) Marlow, A. L.; Kiessling, L. L. Org. Lett. 2001, 3, 2517.
(15) Wittmann, V.; Wong, C.-H. J. Org. Chem. 1997, 62, 2144.
(16) Ernst, C.; Klaffke, W. Tetrahedron Lett. 2001, 42, 2973.
(17) Uchiyama, T.; Hindsgaul, O. J. Carbohydr. Chem. 1998, 17, 1181.
(18) Arlt, M.; Hindsgaul, O. J. Org. Chem. 1995, 60, 14.
(19) Bae, J.; Kim, K.-H.; Kim, D.; Choi, Y.; Kim, J. S.; Koh, S.; Hong,
S.-I.; Lee, D.-S. ChemBioChem 2005, 6, 1963.
(20) Mizanur, R. M.; Zea, C. J.; Pohl, N. L. J. Am. Chem. Soc. 2004,
126, 15993.
(21) Jiang, J.; Albermann, C.; Thorson, J. S. ChemBioChem 2003, 4,
443.
(22) Barton, W. A.; Biggins, J. B.; Jiang, J.; Thorson, J. S.; Nikolov, D.
B. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 13397.
(23) Barton, W. A.; Lesniak, J.; Biggins, J. B.; Jeffrey, P. D.; Jiang, J.;
Rajashankar, K. R.; Thorson, J. S.; Nikolov, D. B. Nat. Struct. Biol. 2001,
8, 545.
(24) Jiang, J.; Biggins, J. B.; Thorson, J. S. Angew. Chem., Int. Ed. 2001,
40, 1502.
(28) Abbreviations: NTP, nucleoside triphosphate; NDP, nucleoside
diphosphate; NMP, nucleoside monophosphate; ATP, adenosine triphos-
phate; CTP, cytidine triphosphate; GTP, guanosine triphosphate; dTTP,
deoxythymidine triphosphate; UTP, uridine triphosphate; R-D-Glc-1-P, R-D-
glucose 1-phosphate; R-D-GlcNH2-1-P, R-D-glucosamine 1-phosphate; R-D-
GlcNAc-1-P, N-acetyl-R-D-glucosamine 1-phosphate; R-D-Gal-1-P, R-D-
galactose 1-phosphate; R-D-Man-1-P, R-D-mannose 1-phosphate; â-L-Fuc-
1-P, â-;-fucose 1-phosphate.
(29) Synthetic scheme, procedures, and characterization data are available
in the Supporting Information.
(25) Jiang, J.; Biggins, J. B.; Thorson, J. S. J. Am. Chem. Soc. 2000,
122, 6803.
(26) Ma, X.; Sto¨ckigt, J. Carbohydr. Res. 2001, 333, 159.
(27) Ropp, P. A.; Cheng, P.-W. Anal. Biochem. 1990, 187, 104.
(30) Davies, G. J.; Gloster, T. M.; Henrissat, B. Curr. Opin. Struct. Biol.
2005, 15, 637.
(31) Coutinho, P. M.; Deleury, E.; Davies, G. J.; Henrissat, B. J. Mol.
Biol. 2003, 328, 307.
858
Org. Lett., Vol. 9, No. 5, 2007

ence of inorganic pyrophosphatase.21,24,25 Aliquots of each
reaction were quenched with methanol and centrifuged prior
to HPLC and ESI-MS/MS analysis. In the absence of
thymidylyltransferase, NTP, sugar 1-phosphate, or MgCl2,
no product formation was observed.
A summary of the substrate flexibility of these three
enzymes, determined after 30 min, is presented in white in
Figure 2. The incubation of R-D-Glc-1-P with both dTTP
at reasonable rates after 30 min (>50% in all but one case),
demonstrating that changes in configuration at C2 and C4,
and even substitution of an acetyl group at C2, were tolerated
by all three nucleotidylyltransferases. Interestingly, when
both alternative R-D-sugar 1-phosphates and alternative NTPs
were incubated with the nucleotidylyltransferases, yields
decreased significantly in all cases except with R-D-GlcNH2-
1-P and UTP, illustrating that these enzymes readily tolerate
one change in either sugar 1-phosphate or NTP.
It is significant to note that the synthesized â-L-Fuc-1-P
substrate was accepted by all three nucleotidylyltransferases,
albeit at conversions ranging from 3 to 10% after 30 min at
37 °C.
The initial incubation of reactions for 30 min at 37 °C
provided an excellent survey of relative substrate affinity.
In order to more fully assess the synthetic potential of these
biocatalysts, reactions resulting in conversions to product of
less than 50% after 30 min were incubated for 24 h at 37 °C.
A summary of these results, shown in black in Figure 2,
illustrates that significant improvements in conversion were
obtained for all reactions with all enzymes after incubation
for 24 h at 37 °C. All three enzymes were able to synthesize
dTDP-â-L-Fuc in >70% conversion after 24 h. In the case
of RmlA3, conversions of eight substrates after 30 min
ranged from only 3-29%, while conversions after 24 h with
the same group of substrates had increased to 90-99%.
The improved conversions to product observed over a 24 h
time period at 37 °C with all three enzymes suggests that
Cps2L, RmlA, and RmlA3 are stable over a significant
portion of the 24 h time period. It is interesting to note that
RmlA3, a thermophilic enzyme, and Cps2L, a mesophilic
enzyme, generally demonstrated comparable levels of activity
over this period of time at 37 °C. These conversion figures
clearly indicate the potential synthetic utility of these three
nucleotidylyltransferases, particularly as neither substrate nor
product inhibition was observed. Also of particular signifi-
cance is the ability of all three thymidylyltransferases to
synthesize dTDP-â-L-Fuc in high yield (72-98%) after
incubation at 37 °C for 24 h. This is potentially surprising
when you consider that dTDP-â-L-Rha is a competitive
inhibitor of P. aeruginosa RmlA with micromolar affin-
ity,32,33 and the epimerization at C2 and C4 of the hexopy-
ranose ring are the only stereochemical differences between
the products. In addition, only minimal degradation (typically
<5%) of NTPs and NDP-sugars to NMPs and NDPs was
observed over the 24 h incubation period, illustrating the
stability of substrates and products at 37 °C. While the
substrate flexibility of several other thermophilic nucleo-
tidylyltransferases have been reported20,34 these results
provide incentive to study the substrate flexibility of a
psychrophilic enzyme, which would potentially function at
even lower temperatures than those reported herein and
further minimize the degradation of substrates and products
over a given time period.
Figure 2. Substrate flexibility of Cps2L, RmlA, and RmlA3 after
30 min and 24 h incubations at 37 °C.
and UTP consistently produced very high levels of conver-
sion (87-100%) with all three enzymes, indicating that UTP
was readily accepted as an alternative NTP with R-D-Glc-
1-P. Similarly, incubation of all three enzymes with R-D-
GlcNH2-1-P with both dTTP and UTP also produced high
levels of conversion (77-96%) after 30 min at 37 °C. When
the reaction of other R-D-sugar 1-phosphates with dTTP was
explored, it was found that R-D-Man-1-P, R-D-Gal-1-P, and
R-D-GlcNAc-1-P were also accepted by all three enzymes
(32) Blankenfeldt, W.; Asuncion, M.; Lam, J. S.; Naismith, J. H. EMBO
J. 2000, 19, 6652.
(33) Melo, A.; Glaser, L. J. Biol. Chem. 1965, 240, 398.
(34) Zhang, Z.; Tsujimura, M.; Akutsu, J.; Sasaki, M.; Tajima, H.;
Kawarabayasi, Y. J. Biol. Chem. 2005, 280, 9698.
Org. Lett., Vol. 9, No. 5, 2007
859

In comparison with the substrate flexibility of other
characterized nucleotidylyltransferases,20,24,25,34,35 Cps2L, RmlA,
and RmlA3 rank very high in terms of the number of sugar
1-phosphates and NTPs they will accept, and the high
conversions to product obtained demonstrate excellent
synthetic potential. In particular, Ep (Salmonella enterica
LT2)24,25 has demonstrated a broad sugar 1-phosphate
flexibility but produced significantly lower conversions with
UTP than dTTP in several cases. Thermostable UDPG-PPase
(Pyrococcus furiosus DSM 3638)20 again demonstrated broad
sugar 1-phosphate flexibility, but produced much lower
conversions with dTTP than UTP in all but one case. In
addition, UDPG-PPase would not accept ATP, a purine
nucleotide, in contrast to Cps2L and RmlA3, which dem-
onstrated catalysis with ATP and R-D-Glc-1-P in excellent
yield (69% and 94%, respectively). Interestingly, another
thermostable nucleotidylyltransferase, ST0452 (Sulfolobus
tokadaii strain 7)34 has demonstrated stringent substrate
specificity as no catalysis was observed with R-D-GlcNH2-
1-P, R-D-Gal-1-P, or R-D-Man-1-P. In addition, ST0452
would not accept ATP, CTP, or GTP nucleotides with any
sugar 1-phosphate. The uridylyltransferase GalU (Strepto-
coccus pneumoniae)35 has also demonstrated stringent speci-
ficity with respect to nucleotide base as no catalysis was
observed with either ATP or UTP with R-D-Glc-1-P.
Nucleotidylyltransferases involved in secondary metabo-
lism have, in general, been reported to be more substrate
stringent than the primary metabolic enzymes described
above.36,37 In the case of BtrD,36 an aminoglycoside antibiotic
thymidylyltransferase, it was found that, out of six sugar
1-phosphates tested, only two R-D-gluco-hexopyranosyl-1-
phosphates were accepted by the enzyme. In addition, BtrD
was also found to be incapable of accepting ATP, CTP, or
GTP nucleotide bases with any sugar 1-phosphate. The
thymidylyltransferase SgcA137 from the enediyne antibiotic
C-1027 gene cluster has also demonstrated very stringent
specificity accepting only R-D-Glc-1-P of six sugar 1-phos-
phates and only dTTP and UTP (6% conversion) of the five
NTP bases studied.
In summary, Cps2L, RmlA, and RmlA3, three primary
metabolic enzymes, have clearly demonstrated their ability
to use both dTTP and UTP nucleotides along with six sugar
1-phosphates to efficiently produce both dTDP- and UDP-
sugars. In addition, Cps2L and RmlA3 have also demon-
strated their utility in the preparation of ADP-, CDP-, and
GDP-sugars with conversions to product ranging from 69-
99%, illustrating the potential synthetic utility of these
nucleotidylyltransferases over others involved in primary and
secondary metabolism. In addition, this is the first report to
demonstrate the ability of bacterial nucleotidylyltransferases
to prepare dTDP-â-L-Fuc with high conversions ranging from
72 to 98% after incubation at 37 °C for 24 h. The lower
temperature required for catalysis compared to the heat-stable
archaeal UDPG-PPase20 may offer a significant advantage
when utilizing less stable sugar 1-phosphates or NTPs. In
conclusion, these three enzymes provide a convenient route
to the efficient preparation of a library of fifteen sugar
nucleotides with a variety of sugars and nucleotide bases.
Kinetic studies on these thymidylyltransferases will provide
more insight into their substrate flexibility.
Acknowledgment. We gratefully acknowledge Professor
Joseph Lam (Department of Molecular and Cellular Biology,
University of Guelph) for providing us with the pRmlA3
plasmid. We also thank Professor Alison Thompson and
Adeeb Al-Sheikh-Ali (Department of Chemistry, Dalhousie
University) for HPLC assistance and Elden Rowland (At-
lantic Research Centre, Dalhousie University) for ESI-MS/
MS support.
Supporting Information Available: Full experimental
details including Cps2L, RmlA, and RmlA3 cloning, expres-
sion, and purification protocols, synthetic procedures, NMR
and mass spectrometry characterization data, and HPLC
methodology and retention times. This material is available
free of charge via the Internet at http://pubs.acs.org.
(35) Bonofiglio, L.; Garc´ıa, E.; Mollerach, M. Curr. Microbiol. 2005,
51, 217.
(36) Kudo, F.; Kawabe, K.; Kuriki, H.; Eguchi, T.; Kakinuma, K. J.
Am. Chem. Soc. 2005, 127, 1711.
(37) Murrell, J. M.; Liu, W.; Shen, B. J. Nat. Prod. 2004, 67, 206.
OL0630853
860
Org. Lett., Vol. 9, No. 5, 2007