One-step synthesis of labeled sugar nucleotides for protein O-GlcNAc modification studies by chemical function analysis of an archaeal protein

Article abstract of DOI:10.1021/ja044117p

Herein we present the chemical function analysis of a recombinant sugar nucleotidyltransferase from the hyperthermophile Pyrococcus furiosus and its use in the one-pot synthesis of chloroacetyl- and alkyne-tagged analogues of uridinediphospho-N-acetylglucosamine (UDP-GlcNAc). The gene was originally annotated as a glucose-1-phosphate deoxythymidylyltransferase; however, kinetic analysis of a panel of sugar-1-phosphates with the protein shows that it is better described as a bifunctional protein that synthesizes UDP-GlcNAc from glucosamine-1-phosphate and acetyl coenzyme A (CoA). A new mass-spectrometry-based assay for the rapid analysis of the acyltransferase activity demonstrates that the enzyme can also accept cheaper truncated N-acetylcysteamine thioester substrates in place of the natural acetyl CoA. The enzyme can tolerate alkyne or chloride substitutions in the acyl moiety, thereby allowing the facile synthesis of tagged sugar nucleotides for future use in protein O-GlcNAc modification studies. Copyright

Full text of DOI:10.1021/ja044117p

Published on Web 12/29/2004
One-Step Synthesis of Labeled Sugar Nucleotides for Protein O-GlcNAc
Modification Studies by Chemical Function Analysis of an Archaeal Protein
Rahman M. Mizanur, Firoz A. Jaipuri, and Nicola L. Pohl*
Department of Chemistry and the Plant Sciences Institute, Gilman Hall, Iowa State UniVersity,
Ames, Iowa 50011-3111
Received September 27, 2004; E-mail: npohl@iastate.edu
Cells use uridinediphospho-N-acetylglucosamine (UDP-GlcNAc)
to transfer N-acetylglucosamine to an incredible range of proteins,
including enzymes and transcription factors, as a transient post-
translational modification implicated in signaling pathways.1 How-
ever, the low concentration of this modification makes identification
of glycosylated sites on proteins challenging.1,2 Facile access to
Figure 1. Synthesis of sugar nucleotides by an enzyme from P. furiosus.
labeled sugar nucleotides would allow the modification to be tagged
for easy isolation and mass spectrometry-based sequencing. Herein
we present the chemical function analysis of a recombinant sugar
nucleotidyltransferase from the hyperthermophile Pyrococcus fu-
riosus and its use in the one-pot synthesis of chloroacetyl- and
alkyne-tagged analogues of UDP-GlcNAc.
Table 1. Kinetic Values for Each of the Functions of the P.
furiosus Protein Assayed Independently with Varying
Concentrations of One Substrate (Fixed Substrate Concentration
in Parentheses)
KM
M)
kcat
kcat/KM
M-1 s-
)
1
1
substrate
(
µ
(s-
)
(µ
Sugar nucleotides for natural products and glycobiology studies
have attracted much recent interest, but chemical strategies for their
synthesis remain cumbersome.1a,b,3 Synthetic strategies incorporating
biocatalysts are often limited by the stability of the enzymes and
their limited tolerance of nonnatural substrates.3 The majority of
these studies have focused on bacterial and eukaryotic proteins.
The third branch of life, archaea,4 provides a largely untapped
genetic source of potential synthetic enzymes, many of which are
predicted to be unusually thermostable.5 The thermal vent archaea
P. furiosus is of particular interest as a large amount of information
is expected from a structural proteomics effort and will require
correlation to verified chemical function data.6 To explore the value
of these proteins as glycobiology reagents, we copied a 1260 base
pair P. furiosus gene annotated as a glucose-1-phosphate thymidy-
lyltransferase using the polymerase chain reaction and ligated it
into a vector that provides a polyhistidine tag at the C-terminus of
the expressed protein for ease of purification by affinity chroma-
tography.
To evaluate the function and substrate specificity of this new
gene product, the enzyme was first tested with glucose-1-phosphate
(1) and deoxythymidinetriphosphate (dTTP) using our recently
developed mass-spectrometry (MS)-based assay.7 Indeed, deox-
ythymidinephosphoglucose was formed as expected from the
genome annotation (Figure 1). The optimal activity was at 80 °C
and a pH of 7.5, reaction parameters which would inactivate any
E. coli proteins that might have coeluted with the P. furiosus protein.
A divalent cation was necessary for activity with Mg2+ serving
best. Nonlinear regression analysis of reactions run in triplicate
Nucleotidyltransferase
glucose-1-phosphate (UTP)
glucose-1-phosphate (dTTP)
mannose-1-phosphate (UTP)
mannose-1-phosphate (dTTP)
N-acetylglucosamine-1-phosphate (UTP)
N-acetylglucosamine-1-phosphate (dTTP)
9 ( 1
0.5
2.9
1.2
1.4
0.05
0.35
0.03
0.03
0.45
0.42
8 ( 1
40 ( 4
44 ( 5
100 ( 10 44
90 ( 10 40
Acetyltransferase
glucosamine-1-phosphate (7)
7 (glucosamine-1-phosphate)
63 ( 9
74 ( 7
44 ( 4
9.5
6.5
1.5
0.15
0.08
0.03
8 (glucosamine-1-phosphate)
delineate the most likely in vivo chemical function of this gene
product (Table 1). The kcat/KM values calculated from three
independent experiments indicated that N-acetylglucosamine-1-
phosphate was actually a better substrate than glucose or mannose
in the synthesis of both UDP- and dTDP-activated sugars.
The high kinetic competence of this enzyme with N-acetylglu-
cosamine-1-phosphate hinted that this enzyme might actually be a
bifunctional enzyme.8 These enzymes catalyze both the acylation
of glucosamine-1-phosphate (Figure 2) and the subsequent nucleo-
tide transfer. The standard assay for this acyltransferase activity is
the indirect detection of released thiols using Ellman’s reagent.9
Because this assay measures the decomposition of thioesters by a
variety of pathways that do not lead to product, we reasoned that
an assay that directly measured product formation would be
preferable, especially at elevated temperatures. Reaction analysis
by ESI-MS rapidly confirmed our hypothesis. In the presence of
acetyl coenzyme A but in the absence of UTP, peaks corresponding
to glucosamine-1-phosphate (m/z ) 258) disappeared over time as
a peak corresponding to N-acetylglucosamine-1-phosphate (m/z )
300) appeared with the enzyme present. Acylation precedes
nucleotide addition. No change was observed in the absence of the
enzyme. With the addition of UTP, too, glucosamine-1-phosphate
was converted to UDP-GlcNAc. Peak quantification with internal
standards and calibration curves provides the first MS-based
acyltransferase assay. To probe the steric limitations of the acyl
moiety, propyl CoA and butyryl CoA were tested. As was found
resulted in a KM ) 8.1 ( 0.7 µM and kcat ) 2.9 (s-1) (kcat/KM
)
0.35 (µM-1 s-1) using glucose-1-phosphate as a substrate (Table
1). The enzyme also showed substantial activity with uridinetri-
phosphate (UTP) as the nucleotide triphosphate.
The utility of this novel nucleotidyltransferase in the synthesis
of sugar nucleotides was then tested by evaluating the transfer
efficiency of six different sugar-1-phosphates. Surprisingly, the
enzyme turned over a variety of substrates, including mannose-,
glucosamine- (3), and N-acetylglucosamine-1-phosphate (2). Clearly,
only an analysis of the kinetic competence of these substrates would
9
836
J. AM. CHEM. SOC. 2005, 127, 836-837
10.1021/ja044117p CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Archaeal genomes clearly are a rich source of proteins; however,
genome annotation based on primary sequence homologies some-
times does not even hint at the true synthetic potential of the
corresponding proteins. Chemical function analysis is required,
ideally, with a range of substrates to ascertain the most likely in
vivo function and to uncover unexpected properties. The archaeal
gene denoted originally as a glucose-1-phosphate deoxythymidy-
lyltransferase is actually a bifunctional acetyltransferase/N-acetyl-
glucosamine-1-phosphate uridylyltransferase. The use of thermo-
stable enzymes in synthetic schemes also allows chemical steps
that are sluggish at lower temperatures to be incorporated into the
same step with an enzyme. Current studies are underway to use
these tagged sugar nucleotides to probe protein modifications by
N-acetylglucosamine moieties.
Acknowledgment. We are grateful for an NSF-CAREER award
and a Shimadzu University Research Grant, and we thank the
Herman Frasch Foundation, administered by the American Chemi-
cal Society, for support of this research. N.L.P. is a Cottrell Scholar
of Research Corporation.
Figure 2. Acyltransferase activity of the bifunctional enzyme with natural
and truncated acetylCoA derivatives. R ) CH3, CH2Cl, CCH.
in the analogous bacterial enzyme,10 the enzyme tolerated a propyl
group reasonably well with only an 8-fold decrease in kcat/KM, but
a two-carbon extension proved detrimental to efficient product
formation.
Supporting Information Available: Experimental details for
cloning, expression, and purification of the hyperthermophilic enzyme,
details of the mass spectrometry assays including calibration curves,
and synthetic details for thioesters 9 and 10. This material is available
free of charge via the Internet at http://pubs.acs.org.
Reexamination of the protein sequence indeed shows alignment
of the first 240 amino acids with NTP-transferases. However, the
C-terminal portion contains a hexapeptide repeat motif often
associated with acyltransferase activity.11 Analysis of a structure
of the analogous bifunctional enzyme from Escherichia coli with
bound acetyl CoA (7) shows several contacts between CoA and
the protein. However, no contacts are made with the acyl side chain,
and the diphosphate is solvent exposed.12 Fatty acid synthases
(FASs) and polyketide synthases (PKSs) are known to accept
truncated CoA thioesters as substrates, but these enzymes form
covalent acylated intermediates before transfer to a bound sub-
strate.13 The structure of the bifunctional E. coli enzyme does not
support a covalent enzyme intermediate for the acyltransfer reaction
to glucosamine.
References
(1) (a) Tai, H.-C.; Khidekel, N.; Ficarro, S. B.; Peters, E. C.; Hsieh-Wilson,
L. C. J. Am. Chem. Soc. 2004, 126, 10500-105001. (b) Vocadlo, D. J.;
Hang, H. C.; Kim, E.-J.; Hanover, J. A.; Bertozzi, C. R. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 9116-9121. (c) Zachara, N. E.; Hart, G. W. Chem.
ReV. 2002, 102, 431-438.
(2) (a) Khidekel, N.; Arndt, S.; Lamarre-Vincent, N.; Lippert, A.; Poulin-
Kerstein, K. G.; Ramakrisnan, B.; Qasba, P. K.; Hsieh-Wilson, L. C. J.
Am. Chem. Soc. 2003, 125, 16162-16163. (b) Haynes, P. A.; Aebersold,
R. Anal. Chem. 2002, 72, 5402-5410. (c) Wells, L.; Vosseller, K.; Cole,
R. N.; Cronshaw, J. M.; Matunis, M. J.; Hart, G. W. Mol. Cell. Proteomics
2002, 1, 791-804.
(3) Thorson, J. S.; Barton, W. A.; Hoffmeister, D.; Albermann, C.; Nikolov,
D. B. ChemBioChem 2004, 5, 16-25. (b) Koeller, K. M.; Wong, C.-H.
Nat. Biotechnol. 2000, 18, 835-841. (c) Elhalabi, K.; Rice, K. G. Curr.
Med. Chem. 1999, 6, 93-116.
The ability of FASs and PKSs to accept significantly cheaper
and less synthetically challenging analogues such as the N-
acetylcysteamine thioester 8 has greatly aided experiments to
understand and exploit these enzymes; therefore, we decided to
synthesize and test 8 with the P. furiosus enzyme. In fact, this
truncated analogue was readily accepted and turned over with only
a 2-3-fold reduction in kcat/KM compared to the full length substrate
(Table 1). Emboldened by this discovery, we synthesized analogues
9 and 10 (Figure 2) containing sterically conservative modifications
that are nevertheless amenable to selective tagging reactions via
thiol substitution14 or Sonogashira reactions15 and alkyne/azide
coupling,16 respectively. The enzyme not only was able to acylate
the amine but also could convert the resulting modified glucos-
amine-1-phosphate into the activated UDP-sugar with 80% conver-
sion in one pot. The acylation reactions did not occur at 37 °C
with enzyme or at 80 °C without enzyme with one exception.
Significant uncatalyzed acylation took place with the alkyne
analogue at 80 °C. Therefore, even if the acyltransfer domain did
not readily accept this modification, the thermostability of the
enzyme allows the coupling of the chemical acylation with the
enzymatic uridyl transfer in one pot. For comparison, the chemical
synthesis of an azide-labeled UDP-GlcNAc required 11 chemical
steps with less than a 15% overall yield.1b
(4) Rivera, M. C.; Lake, J. A. Nature 2004, 431, 152-155.
(5) Sterner, R.; Liebl, W. Crit. ReV. Biochem. Mol. Biol. 2001, 36, 39-106.
(6) Adams, M. W. W.; Dailey, H. A.; DeLucas, L. J.; Luo, M.; Prestegaard,
J. H.; Rose, J. P.; Wang, B.-C. Acc. Chem. Res. 2003, 36, 191-198.
(7) Zea, C. J.; Pohl, N. L. Anal. Biochem. 2004, 328, 196-202.
(8) Mengin-Lecreulx, D.; Heijenoort, J. V. J. Bacteriol. 1994, 18, 5788-
5795.
(9) Riddles, P. W.; Blakeley, R. L.; Zenner, B. Methods Enzymol. 1983, 91,
49-60.
(10) Gehring, A. M.; Lees, W. J.; Mindiola, D. J.; Walsh, C. T.; Brown, E. D.
Biochemistry 1996, 35, 579-585.
(11) (a) Bairoch, A. Nucleic Acids Res. 1993, 21, 3097-3103. (b) Dicker, I.
B.; Seetharam, S. Mol. Microbiol. 1992, 6, 817-823. (c) Vourio, R.;
Harkonen, T.; Tolvanen, M.; Vaara, M. FEBS Lett. 1994, 337, 289-292.
(12) Olsen, L. R.; Roderick, S. L. Biochemistry 2001, 40, 1913-1921.
(13) See, for example: (a) Mishra, P. K.; Drueckhammer, D. G. Chem. ReV.
2000, 100, 3283-3309. (b) Pohl, N. L.; Gokhale, R. S.; Cane, D. E.;
Khosla, C. J. Am. Chem. Soc. 1998, 120, 11206-11207.
(14) Svensson, R.; Greno¨, C.; Johansson, A.-S.; Mannervik, B.; Morgenstern,
R. Anal. Biochem. 2002, 311, 171-178.
(15) (a) Bong, D. T.; Ghadiri, M. R. Org. Lett. 2000, 3, 2509-2511. (b)
Dibowski, H.; Schmidtchen, F. P. Angew. Chem., Int. Ed. 1998, 37, 476-
478.
(16) Speers, A. E.; Cravatt, B. F. Chem. Biol. 2004, 11, 535-546. (b) Lewis,
W. G.; Magallon, F. G.; Fokin, V. V.; Finn, M. G. J. Am. Chem. Soc.
2004, 126, 9152-9153. (c) Link, J. A.; Vink, M. K. S.; Tirrell, D. A. J.
Am. Chem. Soc. 2004, 126, 10598-10602. (d) Speers, A. E.; Adam, G.
C.; Cravatt, B. F. J. Am. Chem. Soc. 2003, 125, 4686-4687. (e)
Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596-2599. (f) Huisgen, R. Pure Appl. Chem.
1989, 61, 613-628.
JA044117P
9
J. AM. CHEM. SOC. VOL. 127, NO. 3, 2005 837