Sortase-catalyzed peptide-glycosylphosphatidylinositol analogue ligation

Article abstract of DOI:10.1021/ja903231v

(Figure Presented) It is demonstrated that sortase A (SrtA) can catalyze efficient coupling of peptides to GPI analogues with a glycine residue attached to the phosphoethanolamine moiety at the nonreducing end to form GPI-linked peptides. This represents

Full text of DOI:10.1021/ja903231v

Published on Web 07/07/2009
Sortase-Catalyzed Peptide-Glycosylphosphatidylinositol Analogue Ligation
Xueqing Guo, Qianli Wang, Benjamin M. Swarts, and Zhongwu Guo*
Department of Chemistry, Wayne State UniVersity, 5101 Cass AVenue, Detroit, Michigan 48202
Received April 22, 2009; E-mail: zwguo@chem.wayne.edu
Scheme 1
Glycosylphosphatidylinositol (GPI) anchoring of cell surface
proteins and glycoproteins onto cell membranes is ubiquitous in
eukaryotic species,1 and GPI-anchored proteins and glycoproteins
play a vital role in various biological processes.2 It has been well-
established that GPI-anchored proteins and glycoproteins cannot
function properly without the GPI anchor. On the other hand, linking
naturally non-GPI-anchored proteins to GPI anchors may assist in
the targeting, trafficking, and shedding of the proteins and help
tune their biological function. For instance, Cathepsin D is a non-
GPI-anchored protein that has no affinity for the pore-forming toxin
aerolysin, but when Cathepsin D was expressed in GPI-anchored
form via bioengineering, it was converted to an aerolysin-binding
variant.3 Consequently, methods to access GPI-anchored proteins
and glycoproteins and related structures are valuable for studying
GPI anchoring and developing novel biotechnologies.4
Scheme 2
Because of the problem of microheterogeneity, it is particularly
difficult to obtain homogeneous GPI-linked proteins/glycoproteins
from biological resources. Meanwhile, the chemical synthesis of
GPI-anchored proteins currently remains a significant challenge.
Though GPI conjugates containing short peptides have been
prepared chemically by total synthesis,5-7 the strategy is not well-
suited for GPI conjugates containing full-sized proteins. The only
described synthesis of any well-defined GPI-anchored protein was
based on native ligation,8 but this synthetic strategy requires Cys-
modified GPI anchors and C-terminus-modified proteins, both of
which are difficult to acquire.
To develop a practical and widely applicable method for the
synthesis of GPI-anchored proteins (1), we became interested in a
biomimetic approach. GPI addition to proteins is a post-translational
process mediated by GPI transamidase.9 All proteins are attached
to GPIs at the nonreducing-end mannose 6-O position of the GPI
glycan core through the peptidic C-terminus and a phosphoetha-
nolamine bridge (dashed box in Scheme 1). Therefore, we
envisioned a strategy for ligating synthetic GPIs to synthetic or
recombinant peptides/proteins by enzymatic methods (Scheme 1).
In this regard, sortases, a group of bacterial transpeptidases that
catalyze surface protein anchoring to bacterial cell walls by
mechanisms similar to that of GPI transamidase,10,11 appeared to
be extremely attractive.
Sortase A (SrtA) from Staphylococcus aureus recognizes a
pentapeptide LPXTG sequence near the C-terminus of a protein,12
breaks the peptide bond between T and G of LPXTG, and links
the new C-terminus of the target protein to the amino group of the
N-terminal glycine of a peptidoglycan.10 SrtA is rather substrate
promiscuous, and therefore, it has been utilized for site-specific
attachment of peptides and proteins to synthetic structures that
contain one or multiple glycine residues13-18 and to carbohydrate
derivatives directly.19 Accordingly, we expected that SrtA might
be able to link peptides and proteins to GPI anchors and thus be
used to prepare GPI-anchored peptides and proteins.
of SrtA. Compound 2 has the same anomeric configuration as the
mannose in natural GPIs, while the methyl group is a useful
diagnostic signal for NMR analysis. SrtA used in the research was
a truncated, water-soluble form of the natural enzyme containing
the extracellular but not the transmembrane sequence, and it was
prepared in our laboratory by a standard recombinant technique.
Peptide 3 had LPETG linked to the hexahistidine tag commonly
used in recombinant techniques to facilitate protein purification.
The reaction was performed at 37 °C in 0.3 M Tris-HCl buffer
(pH 7.5) containing 150 mM NaCl, 5 mM CaCl2, 2 mM 2-mer-
captoethanol, and ∼6.8 µM SrtA, and its progress was monitored
by MALDI-TOF mass spectrometry (MS) and reversed-phased
(RP) HPLC. The reaction did slowly yield a new compound that
was not the desired product but proved to be the hydrolyzed peptide
4 (MS m/z: calcd, 1514.7; found, 1515.8 [M + H]+) (Scheme 2).
This result indicated that 3 was indeed activated by SrtA but that
2 is not a sufficiently active nucleophile to react with the reactive
enzyme-peptide complex; consequently, the activated peptide
slowly hydrolyzed to afford 4. We believed that the poor nucleo-
philicity of 2 was not likely due to steric hindrance, since much
more sterically demanding sugar derivatives are easily accom-
To explore the above hypothesis, we synthesized 2 as a simple
GPI analogue and studied its reaction with peptide 3 in the presence
9
9878 J. AM. CHEM. SOC. 2009, 131, 9878–9879
10.1021/ja903231v CCC: $40.75  2009 American Chemical Society

C O M M U N I C A T I O N S
sufficiently reactive to suppress the hydrolysis and reverse reactions.
Therefore, in theory, complete conversion of 5 and 7 into the desired
conjugates should be possible.
To explore the optimal conditions for these transformations, we
next examined the reactions between 3 (0.5 mM) and GPI analogues
5 and 7 (2.5 mM) in the presence of an increased concentration
(30 µM) of SrtA in the same buffer as described above. It was
observed that under these conditions, the rates and efficiencies of
the two reactions were significantly improved. HPLC analyses
(Figure 1) revealed that after only 2.5 h of incubation at 37 °C, the
reactions afforded the desired GPI conjugates 6 and 8 in yields
greater than 95%, and no significant hydrolysis was observed.
In conclusion, this work has demonstrated that introducing a
glycine residue onto the phosphoethanolamine moiety of the
nonreducing-end glycan of GPI analogues can transform them into
SrtA-acceptable substrates and that SrtA can be utilized to ef-
fectively ligate small peptides and modified GPI analogues. This
represents the first chemoenzymatic synthesis of any GPI-peptide
conjugate and is also a proof-of-concept for the application of SrtA
to the synthesis of GPI-anchored proteins and glycoproteins.
Currently, we are examining the SrtA-catalyzed attachment of
peptides and full-sized proteins to complex GPI analogues and intact
GPIs. On the basis of the present results and literature reports that
SrtA accepts large proteins and various sterically demanding
nucleophiles,13-19 we are confident that SrtA can be a powerful
tool for coupling proteins and glycoproteins to GPIs for chemoen-
zymatic syntheses of natively linked GPI-anchored proteins and
glycoproteins.
Acknowledgment. This work was supported by Wayne State
University and in part by the NSF (CHE-0715275).
Figure 1. HPLC traces for (a) peptide 3, (b) the reaction of 3 and 5, and
(c) the reaction of 3 and 7 (C18 column, gradient eluent using CH3CN/
H2O).
Supporting Information Available: Preparation of SrtA, 2, 3, 5,
and 7; enzymatic reaction conditions and procedures; HPLC results
for the reactions; and MS and NMR spectra of SrtA and 2-8. This
material is available free of charge via the Internet at http://pubs.acs.org.
modated by SrtA.19 Most probably, the nearby phosphate func-
tionality had a negative impact on the nucleophilicity of the amino
group of 2.
To establish the GPI forms acceptable for SrtA, we introduced
a glycine amino acid to the phosphoethanolamine group to obtain
5 and investigated its reaction with 3 in the presence of SrtA. To
our satisfaction, SrtA did catalyze an efficient coupling between 3
and 5 to afford 6 under the reaction conditions described above.
After ∼36 h of incubation at 37 °C, the reaction was quenched,
and purification by RP HPLC (see the Supporting Information)
afforded a 45% isolated yield of the product 6, which was
characterized by both NMR and MS analyses (MS m/z: calcd,
1870.8; found, 1871.8 [M + H]+, 1893.9 [M + Na]+).
To inspect whether elongating the peptide chain linked to the
GPI analogue would have a further impact on the enzymatic
reaction, we introduced another glycine residue onto 5 to obtain 7.
The reaction between 3 and 7 in the presence of SrtA afforded 8
(MS m/z: calcd, 1927.8; found, 1928.8 [M + H]+), and its rate and
efficiency were similar to those of the reaction between 3 and 5,
suggesting that incorporation of a single glycine residue onto the
phosphoethanolamine moiety at the nonreducing end of the GPI is
probably sufficient for SrtA-catalyzed ligations between GPIs and
peptides or proteins.
References
(1) Ferguson, M. A. J. J. Cell Sci. 1999, 112, 2799.
(2) Menon, A. K.; Baumann, N. A.; Rancour, D. M. In Carbohydrates in
Chemistry and Biology; Hart, G. W., Hart, G., Sinay¨, P., Ernst, B., Sinaikin,
P., Eds.; Wiley-VCH: New York, 2000; Vol. 4, pp 757-769.
(3) Diep, D. B.; Nelson, K. L.; Raja, S. M.; Pleshak, E. N.; Buckley, J. T.
J. Biol. Chem. 1998, 273, 2355.
(4) Mayor, S.; Riezman, H. Nat. ReV. Mol. Cell Biol. 2004, 5, 110.
(5) Xue, J.; Shao, N.; Guo, Z. J. Org. Chem. 2003, 68, 4020.
(6) Shao, N.; Xue, J.; Guo, Z. Angew. Chem., Int. Ed. 2004, 43, 1569.
(7) Tanaka, Y.; Nakahara, Y.; Hojo, H.; Nakahara, Y. Tetrahedron 2003, 59,
4059.
(8) Becker, C. F. W.; Liu, X.; Olschewski, D.; Castelli, R.; Seidel, R.;
Seeberger, P. H. Angew. Chem., Int. Ed. 2008, 47, 8215.
(9) Udenfriend, S.; Kodukula, K. Annu. ReV. Biochem. 1995, 64, 563.
(10) Marraffini, L. A.; DeDent, A. C.; Schneewind, O. Microbiol. Mol. Biol.
ReV. 2006, 70, 192.
(11) Perry, A. M.; Ton-That, H.; Mazmanian, S. K.; Schneewind, O. J. Biol.
Chem. 2002, 277, 16241.
(12) Mazmanian, S. K.; Liu, G.; Ton-That, H.; Schneewind, O. Science 1999,
285, 760.
(13) Pritz, S.; Kraetke, O.; Klose, A.; Klose, J.; Rothemund, S.; Fechner, K.;
Bienert, M.; Beyermann, M. Angew. Chem., Int. Ed. 2008, 47, 3642.
(14) Tanaka, T.; Yamamoto, T.; Tsukiji, S.; Nagamune, T. ChemBioChem. 2008,
9, 802.
(15) Parthasarathy, R.; Subramanian, S.; Boder, E. T. Bioconjugate Chem. 2007,
18, 469.
(16) Pritz, S.; Wolf, Y.; Kraetke, O.; Klose, J.; Bienert, M.; Beyermann, M. J.
Org. Chem. 2007, 72, 3909.
In the above experiments, we found that the reactions reached
the maximal yield of desired conjugate products after ∼24 h of
incubation. Thereafter, the remaining substrate peptide 3 could not
be transformed into GPI conjugate, but it was not hydrolyzed either.
This result suggests that SrtA may lose its activity over a long
incubation time. Moreover, because these reactions did not give
rise to significant hydrolysis products, both 5 and 7 seemed to be
(17) Mao, H.; Hart, S. A.; Schink, A.; Pollok, B. A. J. Am. Chem. Soc. 2004,
126, 2670.
(18) Antos, J. M.; Miller, G. M.; Grotenbreg, G. M.; Ploegh, H. L. J. Am. Chem.
Soc. 2008, 130, 16338.
(19) Samantaray, S.; Marathe, U.; Dasgupta, S.; Nandicoori, V. K.; Roy, R. P.
J. Am. Chem. Soc. 2008, 130, 2132.
JA903231V
9
J. AM. CHEM. SOC. VOL. 131, NO. 29, 2009 9879