Asparagine surrogates for the assembly of N-linked glycopeptide mimetics by chemoselective ligation

Article abstract of DOI:10.1016/S0040-4039(01)00104-6

Alanine-β-hydroxylamine (Aβx) and alanine-β-hydrazide (Aβz) have been developed as asparagine surrogates for the assembly of N-linked glycopeptide mimetics by chemoselective ligation. The utility of these residues is illustrated with the synthesis of the oligosaccharyl transferase substrate mimetics 1 and 2, and conjugation thereof with GlcNAc to obtain the N-glycopeptide mimetics 3 and 4.

Full text of DOI:10.1016/S0040-4039(01)00104-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 2085–2087
Asparagine surrogates for the assembly of N-linked glycopeptide
mimetics by chemoselective ligation
Ste´phane Peluso and Barbara Imperiali*
Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA
Received 20 December 2000; accepted 9 January 2001
Abstract—Alanine-b-hydroxylamine (Abx) and alanine-b-hydrazide (Abz) have been developed as asparagine surrogates for the
assembly of N-linked glycopeptide mimetics by chemoselective ligation. The utility of these residues is illustrated with the synthesis
of the oligosaccharyl transferase substrate mimetics 1 and 2, and conjugation thereof with GlcNAc to obtain the N-glycopeptide
mimetics 3 and 4. © 2001 Elsevier Science Ltd. All rights reserved.
The synthesis of glycopeptide mimetics has recently
received significant attention with the development of
C-glycoside and S-glycoside analogs as well as
chemoselective ligation techniques.1 The latter is based
on the use of saccharides and peptides bearing mutually
and selectively reactive functionalities. This strategy is
particularly attractive as it allows the convergent syn-
thesis of neoglycopeptides without the need for protect-
ing groups or activating agents. Recent examples
include reaction of reducing sugars with peptides fea-
Asn
Aβx
Aβz
NH2
H2N
H2N
HN
OH
NH2
O
H
N
OH
NH2
O
OH
NH2
O
O
O
O
O
H
N
O
O
H
N
H
N
H
N
N
H
N
H
N
H
N
H
O
O
O
O
O
O
1
2
oligosaccharyl
transferase
chemoselective ligation
OH
O
HO
HO
HO
HO
HO
HO
OH
R= Man9Glc3GlcNAc
RO
HO
NH
Ac
OH
OH
OH
Ac
Ac
O
O
OH
Ac
N
H
N
H
O
N
H
N
NH
NH
HN
OH
NH2
O
O
OH
NH2
OH
NH2
O
H
N
O
H
N
O
O
O
H
N
O
N
H
N
H
N
H
N
H
N
H
N
H
O
O
O
O
O
O
3
4
N-linked glycopeptide
N-linked glycopeptide mimetics
Figure 1. Alanine-b-hydroxylamine (Abx) and alanine-b-hydrazide (Abz) as asparagine surrogates for the assembly of N-linked
glycopeptide mimetics by chemoselective ligation.
Keywords: glycopeptide mimetic; solid-phase synthesis; chemoselective ligation.
* Corresponding author. Tel.: (617) 253-1838; fax: (617) 452-2419; e-mail: imper@mit.edu
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00104-6

2086
S. Peluso, B. Imperiali / Tetrahedron Letters 42 (2001) 2085–2087
turing an aminooxy group2 and hydrazide- and
thiosemicarbazide-functionalized saccharides with pep-
tides featuring a carbonyl group.3
in quantitative yield by using FmocCl and collidine in
water/dioxane (1:1) to afford 9. The weak base collidine
was used to prevent elimination of the susceptible a-
amino ester. Finally, the allyl ester was removed by
using a catalytic amount of Pd(PPh3)4 and PhSiH3, as
scavenger, in DCM. The crude product was purified by
flash chromatography followed by size exclusion chro-
matography to remove all traces of palladium impuri-
ties and acetic acid to obtain 5.11
As part of an ongoing program focused on understand-
ing enzyme catalyzed N-linked protein glycosylation,4
we have developed alanine-b-hydroxylamine and ala-
nine-b-hydrazide as asparagine surrogates for the
assembly of N-glycopeptide mimetics by chemoselective
ligation. This approach is illustrated with the synthesis
of the oligosaccharyl transferase substrate mimetics 1
and 2, and the N-linked glycopeptide mimetics 3 and 4
(Fig. 1).
Use of 5 in SPPS was illustrated in the synthesis of 1
from H2N-Leu-Thr(tBu)-XAL-PS. Coupling of 5 was
realized using PyAOP and collidine as a weak base in
order to prevent elimination of the hydroxyphthalimide
group. Fmoc deprotection was performed using pipe-
ridine/DMF (1:4) buffered with dinitrophenol followed
by the N-terminal capping with benzoic anhydride
and pyridine as base. At this stage, release from the
solid support using TFA/DCM/H2O/triisopropylsilane
(90:5:2.5:2.5) afforded the phthalimide-protected pep-
tide. Removal of the protecting group proceeded
smoothly and quantitatively (HPLC analysis) using
methylhydrazine in acetonitrile/water (4:1) to afford
1.12 Alternatively, the phthalimide protecting group can
be removed on solid support using hydrazine:allyl alco-
hol:trifluoroethanol (1:3:46). Trifluoroethanol was
essential; attempts with DMF, DCM or MeCN/H2O as
solvent gave inferior yields.
Alanine-b-hydroxylamine is related to the natural
product canaline and to the antibiotics
D- and L-cyclo-
serine. The syntheses of canaline5 and homocanaline6
have been reported. Specifically, hydrolysis of L-cyclo-
serine affords alanine-b-hydroxylamine7 which can then
be sequentially protected to access the FmocAbx-
(Boc)OH8 building block. However, this semi synthetic
approach suffers from the price of commercially avail-
able L-cycloserine (Aldrich, $ 20/25 mg).
In order to generate more abundant quantities of a
versatile building block for solid phase peptide synthe-
sis (SPPS) we have chosen to target FmocAbx(Pht)OH.
The Fmoc protecting group is compatible with the
Fmoc/tBu SPPS strategy. The phthalimide protecting
group was selected over carbamate protecting groups
because carbamate protected hydroxylamines have been
shown to be vulnerable to alkylation or acylation at the
remaining amide proton.9 FmocAbx(Pht)OH 5 was
synthesized in five steps from serine, in an overall yield
of 25%. The key step of the synthetic strategy is a
Mitsunobu reaction of N-hydroxyphthalimide with an
N-a-trityl protected serine (Scheme 1). Serine was pro-
tected as the N-trityl and allyl ester in a single pot
procedure to give 6 in 52% overall yield. The allyl ester
was selected over a methyl ester since all attempts to
remove the latter in Abx(Pht)OMe led excusively to
elimination side products. The use of the N-trityl pro-
tecting group has been shown to prevent elimination in
Mitsunobu reactions of serine,10 and 7 was obtained
from 6 in 68% yield. Removal of the N-a-trityl protect-
ing group was achieved by treating 7 with HCl/Et2O.
After trituration and washing with diethyl ether, pure 8
was isolated by centrifugation in 87% yield. Introduc-
tion of the N-a-Fmoc protecting group was performed
In contrast to the building block approach of Abx,
Abz-containing peptide 2 was prepared directly on solid
support by modification of the aspartic acid side chain
of Bz-Asp(OAll)-Leu-Thr(tBu)-XAL-PS. Upon palla-
dium-catalyzed removal of the allylic ester using cata-
lytic Pd(PPh3)4 and PhSiH3, as scavenger, in DCM, the
hydrazide functionality was obtained by coupling of
t-butylcarbazate in the presence of PyAOP and col-
lidine. Release from the solid support and deprotection
were achieved using TFA/DCM/H2O/triisopropylsilane
(90:5:2.5:2.5) to afford 2.13
Conjugation of 1 and 2 with GlcNAc was realized
according to chemoselective ligation protocols (Fig.
1).14 For example, a solution of GlcNAc 1 M in buffer
(AcONa 0.1 M pH 5.6; 60 ml, 5 equiv.) was added to a
solution of 1 (12 mmol) in DMSO (50 ml). After 24 h at
room temperature, the solution was concentrated in
vacuo. HPLC analysis showed the total conversion of 1
to afford neoglycoconjugate 3.15 Neoglycoconjugate 415
was obtained from 2 following the same procedure. As
OH
OPht
1) HOAll, pTsOH /Toluene
PPh3, DEAD, HONPht
O
O
L-serine
TrtHN
TrtHN
2) TrtCl, TEA
52%
68%
O
O
6
7
HCl / Et2O
87%
FmocCl
collidine
OPht
O
Pd(PPh3)4,
OPht
OPht
PhSiH3
OH
O
HCl.H2N
99%
FmocHN
84%
FmocHN
O
O
8
O
9
5
Scheme 1. Synthesis of FmocAbx(Pht)OH.

S. Peluso, B. Imperiali / Tetrahedron Letters 42 (2001) 2085–2087
2087
1
AcOH, 90:10:1); 13C NMR (CDCl3) l (ppm): 47.21,
expected, H NMR analysis of 3 in DMSO showed two
sets of peaks arising from the E- and Z-acyclic glycosyl
53.38, 67.90, 77.88, 120.15, 124.13, 125.56, 127.34, 127.92,
oxime isomers.2a,16 Likewise, H NMR analysis of 4 in
1
1
128.83, 135.11, 141.45, 143.95, 156.52, 163.78, 172.45; H
DMSO showed two sets of peaks arising from the cis
NMR (CDCl3) l (ppm): 7.86 (2H, dd, J=3.0 Hz, J=5.5
Hz), 7.77 (4H, m), 7.67 (2H, m), 7.41 (2H, t, J=7.3 Hz),
7.33 (2H, t, J=7.3 Hz), 6.48 (1H, d, J=8.2 Hz), 4.86
(1H, dd, J=3.0 Hz, J=11.0 Hz), 4.68 (1H, m), 4.44–4.35
(3H, m), 4.28 (1H, m).
and trans b-glycosyl hydrazide isomers.17
In conclusion, we have developed two asparagine surro-
gates, alanine-b-hydroxylamine and alanine-b-hydra-
zide, for the assembly of N-glycopeptides mimetics by
chemoselective ligation. These surrogates are currently
being used for biochemical studies of oligosaccharyl
transferase. The results of these studies will be reported
elsewhere.
12. Characterization of 1: C20H31N5O6 (Mw=437.52),
HPLC: tR=16.52 min (C18 7–100% B in 28 min), ESMS:
[M+H]+=438.2; 1H NMR (DMSO-d6) l (ppm): 8.70 (1H,
d, J=7.6 Hz), 8.31 (1H, d, J=7.9 Hz), 7.87 (2H, m), 7.62
(1H, d, J=8.8 Hz), 7.55 (1H, m), 7.49 (2H, m), 7.14 (1H,
s), 7.13 (1H, s), 4.82 (1H, td, J=7.6 Hz, J=5.2 Hz), 4.36
(1H, dd, J=7.9 Hz, J=14.9 Hz), 4.16 (br, 1H), 4.10 (2H,
m), 4.09 (1H, m), 1.62 (1H, m), 1.52 (2H, m), 1.00 (3H, d,
J=6.4 Hz), 0.87 (3H, d, J=6.7 Hz), 0.82 (3H, d, J=6.7
Hz).
13. Characterization of 2: C21H32N6O6 (Mw=464.54);
HPLC: tR=16.70 min (7–100% B in 28 min); ESMS:
[M+H]+=465.21; 1H NMR (DMSO-d6) l (ppm): 10.29
(1H, br s), 8.71 (1H, d, J=7.9 Hz), 8.17 (1H, d, J=8.2
Hz), 7.85 (2H, m), 7.62 (1H, d, J=8.8 Hz), 7.55 (1H, m),
7.48 (2H, m), 7.13 (1H, s), 7.11 (1H, s), 4.87 (1H, td,
J=5.8 Hz, J=8.1 Hz), 4.32 (1H, td, J=5.9 Hz, J=8.4
Hz), 4.08 (1H, dd, J=3.7 Hz, J=8.8 Hz), 4.01 (1H, m),
2.80 (1H, dd, J=5.8 Hz, J=15.2 Hz), 2.64 (1H, dd,
J=8.5 Hz, J=15.2 Hz), 1.60 (1H, m), 1.50 (2H, m), 1.01
(3H, d, J=6.1 Hz), 0.85 (3H, d, J=6.5 Hz), 0.81 (3H, d,
J=6.5 Hz).
Acknowledgements
This work was supported by NIH (GM39334).
References
1. Marcaurelle, L. A.; Bertozzi, C. R. Chem. Eur. J. 1999, 5,
1384–1390.
2. (a) Cervigni, S. E.; Dumy, P.; Mutter, M. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 1230–1232; (b) Zhao, Y.; Kent, S.
B. H.; Chait, B. T. Proc. Natl. Acad. Sci. USA 1997, 94,
1629–1633; (c) Peri, F.; Dumy, P.; Mutter, M. Tetra-
hedron 1998, 54, 12269–12278.
3. Rodriguez, E. C.; Marcaurelle, L. A.; Bertozzi, C. R. J.
14. Chemoselective ligation through oxime and hydrazone
bond formation is commonly used in peptide and protein
synthesis and has been found to be highly selective in the
presence of competing amino acid side chain functional-
ity. For an example see: Shao, J.; Tam, J. P. J. Am.
Chem. Soc. 1995, 117, 3893–3899.
Org. Chem. 1998, 63, 7134–7135.
4. Imperiali, B. Acc. Chem. Res. 1997, 30, 452–459.
5. (a) Barlos, K.; Papaioannou, D.; Sanida, C. Liebigs Ann.
Chem. 1986, 287–291; (b) Ozinskas, A. J.; Rosenthal, G.
A. J. Org. Chem. 1986, 51, 5047–5050.
6. Lang, I.; Donze, N.; Garrouste, P.; Dumy, P.; Mutter,
15. Characterization of 3: C28H44N6O11 (Mw=640.72);
HPLC: tR=16.22 min (7–100% B in 28 min); ESMS:
[M+H]+=641.26. Characterization of 4: C29H45N7O11
(Mw=667.54); HPLC: tR=16.40 min (7–100% B in 28
min); ESMS: [M+H]+=668.30.
16. Vasella, A.; Witzig, C.; Waldraff, C.; Ulhmann, P.;
Briner, K.; Bernet, B.; Panza, L.; Husi, R. Helv. Chim.
Acta 1993, 76, 2847–2875.
M. J. Pept. Sci. 1998, 4, 72–80.
7. Korpela, T.; Makeal, M. Synthesis 1980, 997–998.
8. Spetzler, J. C.; Hoeg-Jensen, T. J. Pept. Sci. 1999, 5,
582–592.
9. Ngu, K.; Patel, D. V. J. Org. Chem. 1997, 62, 7088–7089.
10. Cheney, R. J.; Wang, L. J. Org. Chem. 1996, 61, 2544–
2546.
11. Characterization of FmocAbx(Pht)OH 5: [h]2D5=+37.0
(c=1, CHCl3); mp=88°C; Rf=0.62 (DCM:MeOH:
17. Bendiak, B. Carbohydr. Res. 1997, 304, 85–90.
.