Enzyme-catalyzed synthesis of isosteric phosphono-analogues of sugar nucleotides

Article abstract of DOI:10.1039/b808078j

Efficient enzymatic syntheses of isosteric phosphono analogues of sugar nucleotides have been accomplished using a thymidylyltransferase. The Royal Society of Chemistry.

Full text of DOI:10.1039/b808078j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Enzyme-catalyzed synthesis of isosteric phosphono-analogues of sugar
nucleotidesw
Stephen A. Beaton,a Malcolm P. Huestis,b Ali Sadeghi-Khomami,bc Neil R. Thomasc
and David L. Jakeman*ab
Received (in Cambridge, UK) 13th May 2008, Accepted 20th October 2008
First published as an Advance Article on the web 18th November 2008
DOI: 10.1039/b808078j
Efficient enzymatic syntheses of isosteric phosphono analogues
of sugar nucleotides have been accomplished using a thymidylyl-
transferase.
more sluggish and low yielding than phosphate–phosphate
couplings for the preparation of sugar nucleotides.20 We
envisage the importance of phosphonate analogues of sugar
nucleotides to increase as the field of glycobiology develops and
additional glycosyltransferase targets are identified.22,23 We
hypothesize that nucleotidylyltransferases will be efficient
catalysts for the formation of phosphono analogues of sugar
nucleotides. Our study is the first to examine the interactions of
phosphono analogues of a-D-glucose-1-phosphate (4) and
a-D-galactose-1-phosphate (5) with nucleoside 50-triphosphates
and a recombinant nucleotidylyltransferase to form isosteric
non-hydrolyzable phosphono sugar nucleotides.
Phosphonates are structural mimics of phosphates that have
found utility in probing a diverse array of biological processes
including the development of clinical antiretroviral and osteo-
porosis drugs.1,2 The replacement of a labile phosphate
linkage with a hydrolytically more stable phosphonate3 has
facilitated the study of a variety of enzymes catalyzing phos-
phate bond cleavage4–8 including enzymes where a transient
cleavage of the phosphate occurs during catalysis.9 Phospho-
nates have also been developed as inhibitors of ester hydro-
lysis10,11 or as reagents for proteomic profiling of serine
proteases.12
Glycosyltransferases reversibly13–15 catalyze the transfer of
a donor sugar to an acceptor substrate with the concomitant
release of a pyrophosphoryl moiety, often in the form of a
nucleoside diphosphate,16 although reversibility for mamma-
lian transferases has yet to be demonstrated. The ability to
control glycosylation by developing glycosyltransferase
inhibitors will facilitate entry into new classes of therapeutics
for a wide variety of diseases.17 Replacing the anomeric
oxygen substituent with a methylene unit in a sugar nucleotide
provides an isosteric, non-hydrolyzable, phosphono sugar
nucleotide analogue that acts as a mechanism-based inhibitor
where glycosyl transfer is compromised. Sugar nucleotide
analogues of this type have been shown to be competitive
inhibitors of glycosyltransferases18,19 and surrogate substrates
for sugar nucleotide processing enzymes, particularly if the
corresponding deoxysugar nucleotide is labile.20 Many glycosyl-
transferases are bi-lobal enzymes that undergo significant
conformational changes upon catalysis,21 necessitating access
to non-scissile structural analogues of the enzyme substrates to
probe enzyme function. Synthetic access to appropriate
phosphono sugar nucleotides is limited by the need to perform
phosphonate–phosphate coupling steps, reactions that are even
The synthesis of 4 was accomplished essentially as described
by Nicotra in five facile steps (44%).24 A similar synthetic
route was proposed and executed for the synthesis of 5 in
seven steps (22%, ESIw).
Physiologically, Cps2L catalyzes condensation of
a-Dglucose-1-phosphate (1) and deoxythymidine triphosphate
(2) to yield dTDP-glucose (3, Scheme 1). A series of assays
involving 4 and 5 and five nucleoside triphosphates as sub-
strates were used to evaluate the substrate specificity of Cps2L
thymidylyltransferase. The enzymatic reactions were carried
out using conditions described previously.25 The enzyme
(2 EU) was incubated with phosphonate 4 or 5 (2 mM), MgCl2
(2.2 mM), and dTTP (1 mM) at 37 1C for 24 h.26 HPLC
analysis of quenched aliquots was used to confirm the forma-
tion of product based on conversion of NTP. Control reac-
tions were performed, and in the absence of any one reagent,
no sugar nucleotide product was formed. Product formation
a Department of Chemistry, Dalhousie University, Halifax, Nova
Scotia, Canada B3H 4J3
b College of Pharmacy, Burbidge Building, Dalhousie University, 5968
College St., Halifax, Nova Scotia, Canada B3H 3J5.
E-mail: david.jakeman@dal.ca; Fax: 1 902 494 1396;
Tel: 1 902 494 7159
c School of Chemistry, Centre for Biomolecular Sciences, University of
Nottingham, Nottingham, England NG7 2RD
w Electronic supplementary information (ESI) available: Experimental
details, characterization of synthetic products, HPLC, MS/MS data
and kinetic plot. See DOI: 10.1039/b808078j
Scheme 1 (a) Primary physiological metabolic reaction catalyzed by
Cps2L; (b) substrates and products examined in this study.
ꢀc
This journal is The Royal Society of Chemistry 2009
238 | Chem. Commun., 2009, 238–240

View Article Online
was confirmed by ESI-MS/MS by observing characteristic
fragmentation patterns of product ions. Apparent kinetic
parameters Km and kcat for the sugar-1-phosphate analogues
were determined by varying phosphonate concentrations and
monitoring by discontinuous HPLC using methods described
previously.27
The incubation of 4 with either dTTP or UTP after 30 min
afforded conversions of 95% and 70%, generating products 7
and 8, respectively, indicating that 4 was readily accepted as an
alternative substrate (Fig. 1). Conversely, incubation of Cps2L
with 5 and dTTP or UTP produced only 9 (440%). Signifi-
cant product was also obtained when assays containing ATP,
CTP, or GTP with 4 were conducted although yields remained
below 20%, even after extended incubation. These enzymatic
conversions are significantly greater yielding, more timely and
experimentally more convenient than chemical phosphate–
phosphonate coupling reactions using morpholidate-20 or
our attempts using N-methylimidazole activation strategies
for nucleoside monophosphates.28 A scaled-up enzymatic
reaction containing 4 and dTTP was performed. Initial
purification of the product 7 was challenging due either to
the presence of salt or to the basic pH that caused the break-
down of the product. A similar instability has been observed
by Lowary and co-workers with sugar nucleotides.29 However,
using a weakly acidic ion-pair buffer (tributyl ammonium
bicarbonate, pH 6) as we have reported previously for the
reversed-phase separation of sugar nucleotides,30 followed by
cation exchange and gel filtration, we were able to purify 7
with minimal loss or degradation of product (Fig. 2).
Fig. 2 Reversed-phase ion-pair HPLC traces. HPLC trace for the
dTTP control void of enzyme (top); the enzyme coupled reaction of 4
and dTTP (middle), and the final purified product, 7 (bottom).
substrate binding, yet significantly different deoxysugar
nucleotide yields.31 Relative to the physiological substrate
(1), 4 exhibited a 140-fold decrease in kcat, 5 a 3200-fold
decrease and 6 a 8200-fold decrease. Thus the difference
between conversions for all these enzymatic reactions lies
within the kcat value.
It is known that phosphonate analogues have a wider bond
angle and shorter bond lengths between the C–O–P atoms and
the C–C–P atoms, respectively.3 By titration, we determined
that the pKa2 of phosphonate (4) was half a unit higher than
the physiological substrate (1). These geometric and ionization
changes presumably account for the decrease in turnover
efficiency for 4 versus 1, but this modification has less effect
on the levels of conversion than the change in stereochemistry
at C4 (1 versus 6). This suggests that the stereochemistry at C4
is of greater importance in terms of the overall catalytic
efficiency than the ionization state of the phosphoryl centre;
thus the configuration of the sugar affects the reactivity more
than the electron density on the phosphoryl centre. This
is corroborated by the fact that an even lower kcat was
observed for a furanosyl sugar phosphate, and highlights the
importance of sugar stereochemistry.27
Apparent kinetic parameters were determined for Cps2L
with respect to 4–6 in the forward direction upon coupling
with dTTP (Table 1). The Km values were similar for 4–6, and
are within the same range as previously reported values of Km
for a-D-Glcp-1-P and even b-L-Araf-1-P.27 This indicates
that the formation of an activated Michaelis complex is
accomplished approximately equally. Similar trends in
Michaelis parameters have been shown with a series of
Two contrasting observations result from the comparison
between the phosphono sugar-1-phosphate analogues relative
to the parent sugar-1-phosphates. First, the phosphonate 4 has
a lower turnover efficiency in comparison to 1, and second, the
phosphonate 5 has a higher turnover efficiency, relative to 6. It
is interesting that a stereochemical change at C4, distant from
the reactive site, can partially be countered by modification of
the phosphate moiety (5). Potentially, the change in stereo-
chemistry at C4 may perturb the trajectory of the phosphate,
and replacing the anomeric oxygen with a methylene unit may
partially realign the nucleophile trajectory as a result of the
differences in geometry of the phosphonate. Increases in pKa2
values were also observed for a series of 3-O-alkyl glucose-
1-phosphate Cps2L substrates, indicating that the enzyme is
capable of accepting substrates with modified pKa2 values.32
Our results clearly demonstrate that the anomeric oxygen
substituent is not critical for nucleotidylyltransferase activity
and further elaboration at this site will potentially result
in structurally and functionally important sugar nucleotide
analogues.
deoxyglucose-1-phosphates and a related nucleotidylyl-
transferase; Pohl observed limited differences between
A study on the conversion of carbaglucose-1-phosphate
using
revealed that the carbocyclic analogue of the physiological
a
catalytically analogous nucleotidylyltransferase
Fig. 1 Substrate conversions by Cps2L after 30 min ( ) and 24 h ( )
incubations at 37 1C. (*ref. 27).
substrate displayed a 725-fold decrease in kcat/Km with respect
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 238–240 | 239

View Article Online
Table 1 Apparent kinetic parameters for Cps2L substrates with dTTP
Parameter
127
4
5
6
b-L-Araf-1-P27
Vmax (mM minꢁ1
Km (mM)
)
2.818
0.4321
124
0.198
0.00166
0.493
176
0.00849
0.0000483
0.1022
107
0.00335
0.0000313
0.2801
90
0.00218
0.0000242
139
27.6
0.199
kcat (minꢁ1
)
kcat/Km (mMꢁ1 minꢁ1
)
to the physiological substrate.33 In contrast, the phosphonate
analogue 4, exhibited a 120-fold decrease in kcat/Km. The
presence of the methylene functionality as a non-scissile
isosteric analogue of the glycosidic linkage in these sugar
nucleotide analogues (7–9), and the facile access to these
compounds by enzymatic coupling, will enable their use as
substrates for subsequent sugar nucleotide processing enzymes
to generate specific isosteric glycosyltransferase inhibitors.
In conclusion, nucleotidylyltransferases have been shown to
solve the issues of low yields and sluggish reaction times for
the formation of phosphono analogues of sugar nucleotides.
Cps2L is the first nucleotidylyltransferase to couple dTTP and
UTP nucleotides with a phosphonate analogue of glucose-
1-phosphate to produce phosphono analogues of dTDP- and
UDP-sugars. It is also capable of producing a phosphono
analogue of dTDP-Gal. The enzymatic production of these
products indicate that increases in the second ionization
constant (pKa2) does not significantly hinder the catalytic
activity. In the case of the galacto-configured substrates, the
phosphonate is a better substrate than the phosphate. Further-
more, the ready conversion of the phosphono analogue
of glucose-1-phosphate to phosphono sugar nucleotide
analogues indicates the potential of efficient enzymatic
transformations to generate novel glycosyltransferase probes.
We thank the Canadian Institutes of Health Research, the
Natural Sciences and Engineering Research Council, The
Royal Society (UK) and the Mizutani Glycoscience Founda-
tion of Japan for funding. We thank Jessica Pearson for
purification of the enzyme.
10 L. Han, J. Hiratake, A. Kamiyama and K. Sakata, Biochemistry,
2007, 46, 1432–1447.
11 G. W. Zhou, J. C. Guo, W. Huang, R. J. Fletterick and
T. S. Scanlan, Science, 1994, 265, 1059–1064.
12 D. Kidd, Y. S. Liu and B. F. Cravatt, Biochemistry, 2001, 40,
4005–4015.
13 C. Zhang, B. R. Griffith, Q. Fu, C. Albermann, X. Fu, I. Lee, L. Li
and J. S. Thorson, Science, 2006, 313, 1291–1294.
14 A. Minami, R. Uchida, T. Eguchi and K. Kakinuma, J. Am. Chem.
Soc., 2005, 127, 6148–6149.
15 K. D. Miller, V. Guyon, J. N. Evans, W. A. Shuttleworth and
L. P. Taylor, J. Biol. Chem., 1999, 274, 34011–34019.
16 C. Breton, L. Snajdrova, C. Jeanneau, J. Koca and A. Imberty,
Glycobiology, 2006, 16, 29R–37R.
17 A. L. Lovering, L. H. de Castro, D. Lim and N. C. Strynadka,
Science, 2007, 315, 1402–1405.
18 R. D. Gordon, P. Sivarajah, M. Satkunarajah, D. Ma,
C. A. Tarling, D. Vizitiu, S. G. Withers and J. M. Rini, J. Mol.
Biol., 2006, 360, 67–79.
19 R. N. Armstrong, J. C. Andre and J. G. M. Bessems, Colloq.
INSERM, Cell. Mol. Aspects Glucuronidation, 1988, 173,
51–58.
20 Z. B. Zhao, L. Hong and H. W. Liu, J. Am. Chem. Soc., 2005, 127,
7692–7693.
21 D. N. Bolam, S. Roberts, M. R. Proctor, J. P. Turkenburg,
E. J. Dodson, C. Martinez-Fleites, M. Yang, B. G. Davis,
G. J. Davies and H. J. Gilbert, Proc. Natl. Acad. Sci. U. S. A.,
2007, 104, 5336–5341.
22 T. Kajimoto and M. Node, J. Synth. Org. Chem. Jpn., 2006, 64,
894–912.
23 A. Dove, Nat. Biotech., 2002, 19, 913–917.
24 F. Nicotra, F. Ronchetti and G. Russo, J. Org. Chem., 1982, 47,
4459–4462.
25 S. C. Timmons, R. H. Mosher, S. A. Knowles and D. L. Jakeman,
Org. Lett., 2007, 9, 857–860.
26 Abbreviations: dTTP, deoxythymidine triphosphate; UTP, uridine
triphosphate; ATP, adenosine triphosphate; CTP, cytidine triphos-
phate; GTP, guanosine triphosphate; dTDP, deoxythymidine
diphosphate; a-D-Glc-1-P, a-D-glucose-1-phosphate; a-D-Gal-1-P,
a-D-galactose-1-phosphate; b-L-Araf-1-P, b-L-arabinofuranose-
1-phosphate; EU, Enzyme Unit.
27 S. C. Timmons, J. P. Hui, J. L. Pearson, P. Peltier, R. Daniellou,
C. Nugier-Chauvin, E. C. Soo, R. T. Syvitski, V. Ferrieres and
D. L. Jakeman, Org. Lett., 2008, 10, 161–163.
Notes and references
1 E. De Clercq, Biochem. Pharmacol., 2007, 73, 911–922.
2 R. G. G. Russell and M. J. Rogers, Bone, 1999, 25, 97–106.
3 R. Engel, Chem. Rev., 1977, 77, 349–367.
4 A. K. Bera, L. S. Polovnikova, J. Roestamadji, T. S. Widlanski,
G. L. Kenyon, M. J. McLeish and M. S. Hasson, J. Am. Chem.
Soc., 2007, 129, 4120–4121.
28 S. C. Timmons and D. L. Jakeman, Carbohydr. Res., 2008,
865–874.
5 D. L. Jakeman, A. J. Ivory, G. M. Blackburn and
M. P. Williamson, J. Biol. Chem., 2003, 278, 10957–10962.
6 I. P. Street and S. G. Withers, Biochem. J., 1995, 308, 1017–1023.
7 H. M. Seidel and J. R. Knowles, Biochemistry, 1994, 33,
5641–5646.
29 N. L. Rose, R. B. Zheng, J. Pearcey, R. Zhou, G. C. Completo and
T. L. Lowary, Carbohydr. Res., 2008, 340, 2130–2139.
30 S. C. Timmons and D. L. Jakeman, Org. Lett., 2007, 9, 1227–1230.
31 K. S. Ko, C. J. Zea and N. L. Pohl, J. Org. Chem., 2005, 70,
1919–1921.
8 M. M. Vaghefi, R. J. Bernacki, W. J. Hennen and R. K. Robins,
J. Med. Chem., 1987, 30, 1391–1399.
32 M. P. Huestis, G. A. Aish, J. P. Hui, E. C. Soo and D. L. Jakeman,
Org. Biomol. Chem., 2008, 6, 477–484.
9 A. Caravano, S. P. Vincent and P. Sinay, Chem. Commun., 2004,
1216–1217.
33 K. S. Ko, C. J. Zea and N. L. Pohl, J. Am. Chem. Soc., 2004, 126,
13188–13189.
ꢀc
This journal is The Royal Society of Chemistry 2009
240 | Chem. Commun., 2009, 238–240