Synthesis of truncated analogues for studying the process of glycosyl phosphatidylinositol modification

Article abstract of DOI:10.1021/ol100575q

Figure presented Many eukaryotic proteins are modified with a glycosylphosphatidylinositol (GPI) anchor at their C-termini. This post-translational modification causes these proteins to be noncovalently tethered to the plasma membrane. The synthesis of tr

Full text of DOI:10.1021/ol100575q

ORGANIC
LETTERS
2010
Vol. 12, No. 9
2080-2083
Synthesis of Truncated Analogues for
Studying the Process of Glycosyl
Phosphatidylinositol Modification
Franklin John and Tamara L. Hendrickson*
Department of Chemistry, 5101 Cass AVenue, Wayne State UniVersity,
Detroit, Michigan 48202
tamara.hendrickson@chem.wayne.edu
Received March 9, 2010
ABSTRACT
Many eukaryotic proteins are modified with a glycosylphosphatidylinositol (GPI) anchor at their C-termini. This post-translational modification
causes these proteins to be noncovalently tethered to the plasma membrane. The synthesis of truncated GPI anchor analogues is reported;
these compounds were designed for use as soluble substrates for GPI transamidase (GPI-T), the enzyme that appends the GPI anchor onto
proteins.
Glycosylphosphatidylinositols (GPIs) represent a class of
naturally occurring glycophospholipids that anchor the
C-termini of proteins and glycoproteins to the membrane
of eukaryotic cells. GPI-anchored proteins mediate many
different cellular functions such as signal transduction,
cell-cell recognition, and ion transport.1 The core of the
GPI anchor, 6-Man-R-(1f2)-ManR(1f6)-ManR(1f4)-
GlcNR(1f6)-myo-inosityl-1-phospholipid, is conserved
throughout eukaryotes.1d-f A phosphoethanolamine attached
to C-6 of the ultimate mannose residue forms a bridge
between the modified protein and the glycolipid. The GPI
anchor from S. cereVisiae includes an additional mannose
(1, Figure 1).2
Methods to obtain quantities of pure GPIs from natural
sources are inefficient and hindered by complexity and
heterogeneity in the oligosaccharide. Chemical synthesis
can provide access to both native and novel GPI-anchored
protein structures. Most of the synthetic attempts reported
to date have focused on the full-length GPI anchor or
(1) (a) Varma, Y.; Hendrickson, T. L. ChemBioChem 2010, 11, 623.
(b) Lakhan, S. E.; Sabharanjak, S.; De, A. J. Biomed. Sci. 2009, 16, 93. (c)
Priola, S. A.; McNally, K. L. Prions 2009, 3, 134. (d) Paulick, M. G.;
Bertozzi, C. R. Biochemistry 2008, 47, 6991. (e) Nosjean, O.; Briolay, A;
Roux, B. Biochim. Biophys. Acta 1997, 1331, 153. (f) McConville, M. J.;
Ferguson, M. A. Biochem. J. 1993, 294, 305. (g) Homans, S. W.; Ferguson,
M. A.; Dwek, R. A.; Rademacher, T. W.; Anand, R.; Williams, A. F. Nature
1988, 333, 269.
(2) Fankhauser, C.; Homans, S. W.; Thomas-Oates, J. E.; McConville,
M. J.; Desponds, C.; Conzelmann, A.; Ferguson, M. A. J. Biol. Chem. 1993,
268, 26365.
(3) (a) Bosson, R.; Guillas, I.; Vionnet, C.; Roubaty, C.; Conzelmann,
A. Eukaryot. Cell 2009, 8, 306. (b) Imhof, I.; Flury, I.; Vionnet, C.; Roubaty,
C.; Egger, D.; Conzelmann, A. J. Biol. Chem. 2004, 279, 19614.
10.1021/ol100575q  2010 American Chemical Society
Published on Web 04/09/2010

Figure 2. Structures of GPI anchor analogues 2-4.
the 6-position of Man3. Building from this position, each
analogue contains one or more mannose residues, based on
the structure of the S. cereVisiae GPI (1). Consequently, they
have been truncated to remove the terminal lipid domain,
the myo-inositol, and two or more sugar units. These
analogues were designed to be used as soluble substrates
for GPI-T.
The synthetic strategy adopted here involves the assembly
of common building blocks and synthetic routes. Orthogonal
protecting groups for the glycosyl donor-acceptor pairs were
optimized for yield and to obtain the desired diastereo-
selectivity in the glycosylation reactions. In order to introduce
the phosphoethanolamine moiety onto each glycan, phos-
phoramidite reagent 5 was synthesized in two steps following
published procedure.7
Figure 1. Structure of the yeast GPI anchor 1. R groups denote
sites of known species-specific modifications.1d,3
analogues that contain both carbohydrate and lipid func-
tionalities.4 These lipid-linked compounds are useful for
studying the functions of the GPI anchor and anchored
proteins, but their amphipathic nature hinders their utility
for soluble experiments.1a,d
Synthesis of GPI anchor analogue 2 started from com-
mercially available 1,2,3,4-tetra-O-benzyl-R-D-mannopyra-
noside 6 (Scheme 1). A standard phosphoramidite coupling
reaction employing 4,5-dicyanoimidazole (DCI) as the
coupling agent and previously synthesized phosphoramidate
5 resulted in an intermediate phosphite, which was subse-
quently oxidized in a one-pot reaction to generate 7 in good
yield (73%) over two steps.4a,8 A methanolic solution of
formic acid with Pd/C and hydrogen9 provided the best
conditions for cleavage of the benzyl and benzyloxy carbonyl
(Cbz) groups in 7 to furnish 2 in quantitative yield (Scheme
1).
Synthesis of GPI analogues 3 and 4 required glycosyl
donor-acceptor pairs. The synthesis of mannose building
blocks 11 and 13 started from known ortho ester 810 (Scheme
2), obtained from D-mannose. Acid-catalyzed ortho ester
deprotection of 8, followed by acetylation of the intermediate,
afforded 9. Selective deacetylation using hydrazinium acetate
afforded 10r,ꢀ in excellent yield (83%). Successive treat-
ment with trichloroacetonitrile in the presence of 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU)4e almost exclusively
afforded R-trichloroacetimidate 11 in 71% yield. Neighboring
group participation of the 2-O-acetyl group in 11 enables
R-selectivity in glycosylation reactions.4e Reaction of com-
In vivo, GPI anchors are attached to proteins via the
action of GPI transamidase (GPI-T); this enzyme catalyzes
the replacement of a C-terminal peptide signal sequence
with the full-length GPI anchor.5 Remarkably, the entire
GPI anchor substrate can be replaced by potent small
nucleophiles like hydrazine and hydroxylamine, if they
are presented to GPI-T in sufficiently high concentration.6
Here we describe the synthesis of three GPI anchor
analogues (2-4, Figure 2), which we predict will be more
effective soluble substrates for GPI-T than either hydrazine
or hydroxylamine because they more closely resemble the
full-length GPI anchor. Peptides modified with a C-
terminal disaccharide similar to 3 have been reported
previously.4c
Compounds 2-4 each begin with the site of protein
attachment, namely the phosphoethanolamine side chain at
(4) (a) Paulick, M. G.; Wise, A. R.; Forstner, M. B.; Groves, J. T.;
Bertozzi, C. R. J. Am. Chem. Soc. 2007, 129, 11543. (b) Wu, X.; Guo, Z.
Org. Lett. 2007, 9, 4311. (c) Tanaka, Y.; Nakahara, Y.; Hojo, H; Nakahara,
Y. Tetrahedron 2003, 59, 4059. (d) Xue, J.; Shao, N.; Guo, Z. J. Org.
Chem. 2003, 68, 4020. (e) Mayer, T. G.; Schmidt, R. R. Eur. J. Org. Chem.
1999, 5, 959. (f) Udodong, U. E.; Madsen, R.; Roberts, C.; Roberts, S. C.;
Fraser-Reid, B. J. Am. Chem. Soc. 1993, 115, 7886.
(5) (a) Hong, Y.; Ohishi, K.; Kang, J. Y.; Tanaka, S.; Inoue, N.;
Nishimura, J.; Maeda, Y.; Kinoshita, T. Mol. Biol. Cell 2003, 13, 1780. (b)
Chen, R.; Anderson, V.; Hiroi, Y.; Medof, M. E. J. Cell Biochem. 2003,
88, 1025. (c) Fraering, P.; Imhof, I.; Meyer, U.; Strub, J. M.; van Dorsselaer,
A.; Vionnet, C.; Conzelmann, A. Mol. Biol. Cell 2001, 12, 3295. (d)
Amathauer, R.; Kodukula, K.; Gerber, L.; Udenfriend, S. Proc. Natl. Acad.
Sci. U.S.A. 1993, 90, 3973.
(7) Campbell, A. S.; Fraser-Reid, B. Bioorg. Med. Chem. 1994, 2, 1209.
(8) Gu, Q.-M.; Prestwich, G. D. J. Org. Chem. 1996, 61, 8642
.
(9) (a) Becker, C. F. W.; Liu, X.; Olschewski, D.; Castelli, R.; Seidel,
R.; Seebeger, P. H. Angew. Chem., Int. Ed. 2008, 47, 8215. (b) Yagi, H.;
Thakker, D. R.; Lehr, R. E.; Jerina, D. M. J. Org. Chem. 2004, 69, 6341.
(10) Beignet, J.; Tiernan, J.; Woo, C. H.; Kariuki, B. M.; Cox, L. R. J.
Org. Chem. 2004, 69, 6341.
(6) (a) Ramalingam, S.; Maxwell, S. E.; Medof, M. E.; Chen, R.; Gerber,
L. D.; Udenfriend, S. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7528. (b)
Maxwell, S. E.; Ramalingam, S.; Gerber, L. D.; Brink, L.; Udenfriend, S.
J. Biol. Chem. 1995, 270, 19576.
Org. Lett., Vol. 12, No. 9, 2010
2081

Scheme 1. Synthesis of Monosaccharide 2
Scheme 2. Synthesis of Building Blocks 11 and 13
Scheme 3. Synthesis of Building Blocks 17 and 19
4). Treatment of 20 with BF3·OEt2 in CH2Cl213 proved to be
most effective for 6-O-TIPS deprotection, which afforded
21 in 58% yield. Phosphoramidate 5 was coupled to 21 and
oxidized as described above to give 22 in good yield (57%).
Cleavage of the 2-O-acetyl group11 afforded 23 in high yield
(80%). Selective removal of the allyl group was achieved
by acetic acid and PdCl24e to produce disaccharide 24 (70%
yield). Final deprotection was carried out in one step under
reductive conditions as described above to quantitatively
yield GPI analogue 3.
Synthesis of analogue 4 followed a similar strategy
(Scheme 5) beginning with reaction of the donor-acceptor
pairs 11 and 19 to produce the R(1-6)-linked disaccharide
25 in good yield (62%). TIPS deprotection13 afforded 26 in
79% yield, which was coupled to phosphoramidite 5 and
oxidized to afford 27 in 60% yield over two steps. Subse-
quent deprotections furnished 4.11
pound 11 with allyl alcohol in the presence of trimethylsilyl
trifluoromethanesulfonate (TMSOTf) as catalyst afforded
allyl R-mannoside 12 in 60% yield. 2-O-Deacetylation11
resulted in the formation of compound 13. Compounds 11
and 13 were obtained in multigram scale in eight and nine
steps, respectively.
The synthesis of glycosyl donor-acceptor pairs 17 and
19 began with the acid-catalyzed ortho ester cleavage of
known compound 14 (Scheme 3).12 Acetylation resulted in
the formation of 1,2-di-O-acetyl derivative 15r,ꢀ in 60%
yield. Selective 1-O-deacetylation with hydrazinium acetate
afforded 16r,ꢀ in excellent yield (95%). Trichloroacetimidate
17 (R,ꢀ ≈ 50:1) was prepared from 16 as described above.
Compound 16 was also used in the generation of 1-O-allyl
mannoside 18 via reaction with allyl bromide in the presence
of sodium hydride;4e ensuing 2-O-deacetylation afforded
19.11
The synthetic strategy described herein will facilitate
the synthesis of other truncated GPI analogues with
variations within the different mannose positions. Readily
accessible synthetic analogues of GPI anchors opens
various possibilties for understanding the catalytic activity
of the enzyme GPI transamidase (GPI-T). These analogues
Reaction of mannosyl donor 17 with acceptor 13 in the
presence of catalytic TMSOTf resulted in formation of the
R(1-6)-linked disaccharide 20 in good yield (52%) (Scheme
(11) Zemplen, G. Ber. Dtsch. Chem. Ges. 1927, 60, 1554.
(12) (a) Ravida, A.; Liu, X.; Kovacs, L.; Seeberger, P. H. Org. Lett.
2006, 8, 1815. (b) Ho¨lemann, A.; Stocker, B. L.; Seeberger, P. J. Org.
Chem. 2006, 71, 8071.
(13) Mabic, S.; Lepoittevin, J. -P. Synlett 1994, 10, 851.
Org. Lett., Vol. 12, No. 9, 2010
2082

Scheme 4. Synthesis of Disaccharide 3
Scheme 5. Synthesis of Disaccharide 4
lack the lipid moiety, in comparison to other GPI ana-
logues reported so far; their solubility in aqueous media
is enhanced. Evaluation of these compounds as substrates
for GPI-T is underway.
(University of Zurich) and Prof. Zhongwu Guo (Wayne State
University) for helpful discussions.
Supporting Information Available: Full experimental
1
procedures; H and/or 13C NMR spectra and HRMS of
compounds 2-4, 7, and 9-29. This material is available
free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We thank Wayne State University and
The American Cancer Society (RSG-07-217-01-TBE) for
financial support. We also thank Prof. Nathaniel Finney
OL100575Q
Org. Lett., Vol. 12, No. 9, 2010
2083