A Chemoenzymatic Approach toward the Rapid and Sensitive Detection of O-GlcNAc Posttranslational Modifications

Article abstract of DOI:10.1021/ja038545r

We report a new chemoenzymatic strategy for the rapid and sensitive detection of O-GlcNAc posttranslational modifications. The approach exploits the ability of an engineered mutant of β-1,4-galactosyltransferase to selectively transfer an unnatural ketone

Full text of DOI:10.1021/ja038545r

Published on Web 12/06/2003
A Chemoenzymatic Approach toward the Rapid and Sensitive Detection of
O-GlcNAc Posttranslational Modifications
†
†
†
†
Nelly Khidekel, Sabine Arndt, Nathan Lamarre-Vincent, Alexander Lippert,
†
‡,§
‡
Katherine G. Poulin-Kerstien, Boopathy Ramakrishnan, Pradman K. Qasba, and
,
†
Linda C. Hsieh-Wilson*
DiVision of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125,
and Structural Glycobiology Section, Laboratory of Experimental and Computational Biology, CCR, NCI,
SAIC-Frederic, Inc., Frederick, Maryland 21702-1201
Received September 16, 2003; E-mail: lhw@caltech.edu
Protein glycosylation is one of the most abundant posttransla-
tional modifications and plays a fundamental role in the control of
biological systems. For example, carbohydrate modifications are
important for host-pathogen interactions, inflammation, develop-
1
ment, and malignancy. As part of a broader program to understand
the role of protein glycosylation in neuronal communication, we
are investigating O-GlcNAc glycosylation, which is the covalent
modification of serine and threonine residues by â-N-acetylgluco-
2
samine. The O-GlcNAc modification is found in all higher eu-
karyotic organisms from C. elegans to man and has been shown to
be ubiquitous, inducible, and highly dynamic, suggesting a regula-
tory role analogous to phosphorylation. However, the regulatory
nature of the modification (i.e., dynamic, low cellular abundance)
also represents a central challenge in its detection and study.
A common method to observe O-GlcNAc involves labeling
proteins with â-1,4-galactosyltransferase (GalT), an enzyme that
Figure 1. Strategy for the detection of O-GlcNAc glycosylated proteins.
Scheme 1 a
3
3
catalyzes the transfer of [ H]galactose from UDP-[ H]galactose to
3
terminal GlcNAc groups. Unfortunately, this approach is expensive,
involves handling of radioactive material, and requires exposure
times of days to months. Antibodies4 and lectins offer alternative
means of detection, but they can suffer from weak binding affinity
and limited specificity.
,5
3
a
Conditions: (a) Me2NH, THF (53%); (b) (BnO)2PNiPr2, then mCPBA
(54%); (c) Pd/C, H2, tri-n-octylamine; (d) UMP-morpholidate, 1H-tetrazole,
pyr; (e) TEA, H2O/MeOH (45%, three steps).
In this Communication, we report a new strategy for the rapid
and sensitive detection of O-GlcNAc glycosylated proteins. Our
approach capitalizes on the substrate tolerance of GalT, which
allows for chemoselective installation of an unnatural ketone
functionality to O-GlcNAc modified proteins (Figure 1). The ketone
moiety has been well-characterized in cellular systems as a neutral,
tation enlarges the binding pocket and enhances the catalytic activity
toward GalNAc substrates without compromising specificity.
Analogue 1 was synthesized from the previously reported ketone
9
210
as shown in Scheme 1. Selective anomeric deacetylation
11
followed by treatment with (BnO)
2
PNiPr
2
afforded the phosphite,
12
which was directly oxidized with mCPBA to produce dibenzyl
phosphate 3. Hydrogenolytic debenzylation yielded the unprotected
phosphate as the trioctylammonium salt, which was coupled with
UMP-morpholidate in pyridine13 to provide the target molecule 1
upon deacetylation with TEA.
6
yet versatile, chemical handle. Here, it serves as a unique marker
to “tag” O-GlcNAc glycosylated proteins with biotin. Once
biotinylated, the glycoconjugates can be readily detected by
chemiluminescence using streptavidin conjugated to horseradish
peroxidase (HRP).
With analogue 1 in hand, we examined the ability of GalT to
label the peptide TAPTS(O-GlcNAc)TIAPG, which encompasses
We designed UDP analogue 1 on the basis of previous biochemi-
cal and structural studies of GalT (Figure 1). We chose to append
the ketone functionality at the C-2 position of the galactose ring
because GalT has been shown to tolerate unnatural substrates
containing minor substitutions at the C-2 position, includ-
1
4
an O-GlcNAc modification site within the protein CREB. Using
wild-type GalT, only partial transfer of the keto-sugar was observed
by LC-MS (∼1.5%). As anticipated, however, the Y289L mutant
showed greater activity and afforded complete conversion after 6
h at 4 °C (Figure 2). Subsequent reaction of the ketone-labeled
peptide with the aminooxy biotin derivative, N-(aminooxyacetyl)-
N′-(D-biotinoyl) hydrazine, under mild conditions (pH 6.7 buffer,
7
ing 2-deoxy, 2-amino, and 2-N-acetyl substituents. Moreover, 2-
deoxy-Gal was transferred at rates comparable to Gal, whereas 3-,
,8
4
-, and 6-deoxy-Gal were transferred at reduced rates.7 Analysis
of the crystal structures of GalT complexed with UDP-GalNAc re-
8
h, 25 °C) gave complete formation of the corresponding O-alkyl
oxime.
Having demonstrated the labeling of a peptide, we applied our
vealed that the C-2 N-acetyl moiety is accommodated in a shallow
9
pocket within the active site. Importantly, the single Y289L mu-
strategy to the O-GlcNAc glycosylated protein CREB. Recombinant
CREB from Sf9 cells was incubated with 1 and Y289L GalT for
12 h at 4 °C. Following reaction with aminooxy biotin, the mixture
†
California Institute of Technology.
14
‡
SAIC-Frederic, Inc.
§
CCR, NCI, Frederick, MD.
16162
9
J. AM. CHEM. SOC. 2003, 125, 16162-16163
10.1021/ja038545r CCC: $25.00 © 2003 American Chemical Society

C O MMU N I C A T I O N S
proach and is distinct in several key respects. First, the use of an
engineered GalT and 1 enables near stoichiometric labeling,
resulting in higher sensitivity. Enhanced sensitivity is crucial in
studying O-GlcNAc as the regulatory nature of the modification
means that it is often present only in low cellular abundance.
Second, the use of an engineered GalT rather than the native
O-GlcNAc glycosyltransferase allows one to capture the glycosy-
lated species directly and avoid perturbation of metabolic pathways.
Thus, our approach should permit the observation of O-GlcNAc
signaling pathways after cellular stimulation, an important frontier
in the field.
In conclusion, we have developed a novel chemoenzymatic
strategy that detects O-GlcNAc modifications with an efficiency
and sensitivity that is unrivaled by existing methods. Given the
chemical versatility of the ketone handle, we can envision a variety
of applications, including direct fluorescence detection, affinity
enrichment, and isotopic labeling for comparative proteomics.
Moreover, the study of other enzymes (e.g., farnesyltransferases
and other glycosyltransferases) may also benefit from this approach.
Current efforts in our laboratory are focused on the extension of
the approach to novel glycosylated proteins and to the dynamic
regulation of the modification in cells.
Figure 2. Labeling of the peptide TAPTS(O-GlcNAc)TIAPG. LC-MS
traces monitoring the reaction progress at (A) time 0, (B) 6 h after the
addition of 1 and Y289L GalT, and (C) 8 h after biotin addition. A and B
represent base peak chromatograms, and C is the extracted ion chromato-
gram within 1319.0-1321.0 and 1633.0-1635.5 m/z. The peak at 8 min in
C is a biotin impurity. See Supporting Information for details.
Acknowledgment. We thank Dr. M. Shahgholi, Dr. P. Snow,
H.-C. Tai, and S. Tully for helpful discussions and assistance. This
research was supported by an NSF CAREER Award (CHE-
0239861) and an Alfred P. Sloan Fellowship.
Figure 3. Selective labeling of glycosylated CREB. CREB from Sf9 cells
Supporting Information Available: Synthetic procedures, MS
analysis, and detailed labeling procedures (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
(lanes 1-3) or E. coli (lane 4) was tested.
References
(
1) (a) Varki, A. Glycobiology 1993, 3, 97-130. (b) Lasky, L. A. Annu. ReV.
Biochem. 1995, 64, 113-139. (c) Capila, I.; Linhardt, R. J. Angew. Chem.,
Int. Ed. 2002, 41, 391-412. (d) Rudd, P. M.; Elliott, T.; Cresswell, P.;
Wilson, I. A.; Dwek, R. A. Science 2001, 291, 2370-2376.
Figure 4. Labeling of R-crystallin, and comparison with several existing
detection methods. For the ketone and tritium labeling studies, 0.75 µg of
protein was used; for the lectin and antibodies, 5 µg of protein was used.
See Supporting Information for conditions and full experimental controls.
(2) (a) Wells, L.; Vosseller, K.; Hart, G. W. Science 2001, 291, 2376-2378.
b) Zachara, N. E.; Hart, G. W. Chem. ReV. 2002, 102, 431-438.
3) Roquemore, E. P.; Chou, T. Y.; Hart, G. W. Methods Enzymol. 1994,
30, 443-460. As GalT labels terminal GlcNAc sugars, experiments using
PNGaseF and â-elimination serve as valuable controls.
(4) Snow, C. M.; Senior, A.; Gerace, L. J. Cell Biol. 1987, 104, 1143-1156.
(
(
2
was resolved by SDS-PAGE, transferred to nitrocellulose, and
probed with streptavidin-HRP. Strong labeling of CREB was
observed by chemiluminescence within seconds of exposure to film
(
5) Comer, F. I.; Vosseller, K.; Wells, L.; Accavitti, M. A.; Hart, G. W. Anal.
Biochem. 2001, 293, 169-177.
(
6) (a) Cornish, V. W.; Hahn, K. M.; Schultz, P. G. J. Am. Chem. Soc. 1996,
118, 8150-8151. (b) Mahal, L. K.; Yarema, K. J.; Bertozzi, C. R. Science
(Figure 3). In contrast, no signal was observed over the same time
1
997, 276, 1125-1128. (c) Datta, D.; Wang, P.; Carrico, I. S.; Mayo, S.
period for unglycosylated CREB (from E. coli) or when reactions
were performed in the absence of either 1 or enzyme, demonstrating
the selectivity of the transfer.
L.; Tirrell, D. A. J. Am. Chem. Soc. 2002, 124, 5652-5653.
(
7) (a) Qian, X.; Sujino, K.; Palcic, M. M.; Ratcliffe, R. M. In Glycochem-
istry: Principles, Synthesis, and Application; Wang, P. G., Bertozzi, C.
R., Eds.; Marcel Dekker: New York, 2001; pp 535-565. (b) Wong, C.-
H.; Halcomb, R. L.; Ichikawa, Y.; Kajimoto, T. Angew. Chem., Int. Ed.
Engl. 1995, 34, 521-546.
We next explored the sensitivity of the approach using the
challenging target, R-crystallin. Detection of the O-GlcNAc moiety
on R-crystallin has been reported to be particularly difficult due to
its low stoichiometry of glycosylation (∼10%) and the presence
of only one major modification site.15 Indeed, we found that existing
(
8) Notably, a biotin moiety appended to the 6-position of UDP-Gal can serve
as a substrate for human (but not bovine) GalT: B u¨ tler, T.; Schumacher,
T.; Namdjou, D.-J.; Gallego, R. G.; Clausen, H.; Elling, L. ChemBioChem
2001, 2, 884-894.
3
(9) Ramakrishnan, B.; Qasba, P. K. J. Biol. Chem. 2002, 277, 20833-20839.
10) Hang, H. C.; Bertozzi, C. R. J. Am. Chem. Soc. 2001, 123, 1242-1243.
methods such as wheat-germ agglutinin (WGA) lectin and the
(
(
4
5
O-GlcNAc-specific antibodies RL-2 and CTD110.6 failed to detect
any O-GlcNAc modification on R-crystallin, even when 10 µg of
R-crystallin was used (Figure 4 and Supporting Information). In
contrast, our approach enabled detection of the O-GlcNAc modi-
fication within minutes using 0.75 µg of R-crystallin. For com-
parison, tritium labeling with wild-type GalT required 8 days of
exposure to film for a weaker signal. Thus, our approach represents
at least a 380-fold enhancement in signal over traditional methods.
Recently, Vocadlo et al. have reported the extension of their
Staudinger methodology to O-GlcNAc glycosylated proteins.16 The
strategy described here complements the metabolic labeling ap-
11) Sim, M. M.; Kondo, H.; Wong, C.-H. J. Am. Chem. Soc. 1993, 115, 2260-
2267.
(12) Ha, S.; Chang, E.; Lo, M.-C.; Men, H.; Park, P.; Ge, M.; Walker, S. J.
Am. Chem. Soc. 1999, 121, 8415-8426.
13) Wittmann, V.; Wong, C.-H. J. Org. Chem. 1997, 62, 2144-2147.
(
(
14) Lamarre-Vincent, N.; Hsieh-Wilson, L. C. J. Am. Chem. Soc. 2003, 125,
6612-6613.
(
15) (a) Chalkley, R. J.; Burlingame, A. L. J. Am. Soc. Mass Spectrom. 2001,
1
2, 1106-1113. (b) Haynes, P. A.; Aebersold, R. Anal. Chem. 2000, 72,
5
402-5410.
(16) 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.
JA038545R
J. AM. CHEM. SOC.
9
VOL. 125, NO. 52, 2003 16163