A general strategy toward S-linked glycopeptides

Article abstract of DOI:10.1002/anie.200500090

(Chemical Equation Presented) A high-yield one-pot synthesis of S-linked glycosyl amino acids 1 has been developed (see scheme; DMF = dimethyl formamide) and used in the solid-phase synthesis of S-linked glycopeptides. This approach has been shown to be efficient for the synthesis of various S-linked glycosyl amino acid building blocks.

Full text of DOI:10.1002/anie.200500090

Communications
Glycopeptide Synthesis
A General Strategy toward S-Linked
Glycopeptides**
Desiree A. Thayer, Henry N. Yu, M. Carmen Galan, and
Chi-Huey Wong*
Glycosylation, the most complex posttranslational modifica-
tion of proteins, is involved in a number of biological
processes. Such highly complex glycan structures have
challenged present methods used in the functional study of
glycoproteins.[1–4] Therefore, the development of new meth-
ods for the syntheses of natural and unnatural homogeneously
glycosylated proteins[5–7] is needed for a systematic under-
standing of glycan function. In this context, the syntheses of
natural and modified glycopeptides represent a means for
additional glycoprotein studies and therapeutic develop-
ments.[8–11]
Naturally occurring glycopeptides most commonly incor-
porate an O- or N-glycosidic linkage between the saccharide
moiety and the side chain of an appropriate amino acid
residue. Replacement of the anomeric oxygen or nitrogen
atom by sulfur results in the corresponding S-linked glyco-
peptide, which is known to be more stable chemically,[12] more
resistant to the action of glycosidases, and tolerated by most
biological systems. Furthermore, members of the closely
related S-linked oligosaccharides have been used as enzyme
inhibitors[13,14] and are suggested to be better immunogens
than their natural O-linked analogues.[15–18] An alternate
modification is the C-glycosyl linkage, for which several
synthetic methods exist.[19]
The common synthetic approach for S-linked glycosyl
amino acids uses an anomeric thiolate nucleophile in reaction
with an alanine derivative equipped with a leaving group
(Figure 1). The major side reaction is b-elimination, and the
subsequent Michael addition results in a diastereomeric
mixture at the a-carbon of the amino acid product. Several
procedures have since emerged to overcome this prob-
lem.[20–24] The most advanced among these is a method
reported by Schmidt and co-workers,[24–27] in which the
reaction of the thiosugar 1 with b-bromoalanine was per-
formed in an ethyl acetate/water two-phase system containing
tetra-n-butylammonium hydrogen sulfate (TBAHS) and
NaHCO3 (Figure 2). We have used this protocol for the
[*] D. A. Thayer,+ Dr. H. N. Yu,+ Dr. M. C. Galan, Prof. C.-H. Wong
Department of Chemistry and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax: (+1)858-784-2409
E-mail: wong@scripps.edu
[+] Contributed equally to this work.
[**] This work was supported by the National Institutes of Health (C.-
H.W.) and National Science Foundation Predoctoral and Norton B.
Gilula Graduate Student fellowships (D.A.T.).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or fromthe author.
4596
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200500090
Angew. Chem. Int. Ed. 2005, 44, 4596 –4599

Angewandte
Chemie
Table 1: General applicability of the synthesis of S-linked glycosyl amino
acids.
R1SAc
R1
R2 =Fmoc
R2 =Boc
(Yield [%])
(Yield [%])
4
12 (79)
19 (75)
Figure 1. The common synthetic approach used for the synthesis of
S-linked glycosyl amino acids.
20
20a (74)
21
22
23
24
21a (77)
22a (81)
23a (75)
24a (71)
Figure 2. Reported procedure for the two-step synthesis of S-linked
glycosyl amino acids and an alternative one-pot approach.
25
26
27
28
25a (66)
26a (68)
27a (71)
28a (78)
25b (75)
26b (70)
27b (78)
preparation of several different S-linked Boc- and Fmoc-
modified glycosyl amino acids (Boc = tert-butoxycarbonyl,
Fmoc = 9-fluorenylmethyloxycarbonyl). The Fmoc-protected
compounds, however, can only be obtained indirectly from
the S-linked Boc-modified glycosyl amino acids owing to the
basicity of the reaction. The step for the generation of 1-
thiosugars requires a special precaution, as it may give low
yields due to the facile formation of disulfides during
deacetylation of the sulfur atom and subsequent silica gel
chromatography.
We therefore explored alternate reaction conditions in
which thioacetate 2 and the alanine derivative 3 can be
subjected to a one-pot reaction to generate the desired thio-
linked glycosyl amino acid product (Figure 2). Such reaction
conditions should meet several important criteria: first,
deacetylation of the thioacetate group should be selective to
generate the desired thiolate anion in situ, and to avoid b-
elimination of the alanine derivative; second, dimethyl
formamide (DMF) should be the solvent of choice, as
solubility of large peptides in other solvents can be problem-
atic;[24] furthermore, a set of optimal amino acid protecting
groups should be employed to facilitate large-scale solid-
phase peptide synthesis (SPPS).
Scheme 1. A two-step one-pot synthesis of a Boc-protected glycosyl
amino acid (Bn=benzyl).
With these goals in mind, we first screened numerous one-
galactosyl amino acid 7. This reaction is simple yet very
pot reaction conditions (Table 1, Supporting Information). A effective. It selectively deacetylates the sulfur atom, preserves
simple “two-step one-pot” reaction emerged as the method of
choice (Scheme 1). Therein, with 2,3,4,6-tetra-O-acetyl-1-S-
acetyl-1-thio-b-d-galactopyranose (4)[28] as a starting material,
the anomeric thioacetate was selectively deprotected in situ
with NaOH in anhydrous methanol (pH ꢀ 7.5). After
removing the methanol solvent, the resulting thiolate
sodium salt 5 was subjected to nucleophilic substitution with
b-bromoalanine 6 in dry DMF to give the desired S-linked
the anomeric configuration,[29] and carries out the SN2
displacement reaction effectively without disrupting the
integrity of the a-carbon center. It is necessary to remove
the methanol before the nucleophilic substitution reaction, as
electrophile 6 would otherwise undergo b-elimination with
sodium methoxide generated in situ.
As the Fmoc protecting group is labile in basic conditions,
most of the procedures reported in the literature use Boc as
Angew. Chem. Int. Ed. 2005, 44, 4596 –4599
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4597

Communications
the temporary protecting group to sustain the alkaline
conditions in the nucleophilic substitution reaction.[20,24]
However, the Boc group must then be replaced with the
Fmoc protecting group for the preferred Fmoc SPPS of
glycopeptides.[30,31] After the reaction shown in Scheme 1 was
carried out following this traditional method, we hypothe-
sized that the basicity of intermediate 5 in DMF would be
sufficiently mild to permit the SN2 reaction with the Fmoc-
protected b-bromoalanine species without affecting the Fmoc
group itself. Thus, compound 8a was prepared as reported in
the literature (Scheme 2).[32] The allyl ester protecting group
Scheme 3. Solid-phase peptide synthesis of glycopeptide 14 with
Fmoc-modified glycosyl amino acid 12. The yield from 12 to 14 is 60%
(DIC=1,3-diisopropylcarbodiimide, HOBt=1-hydroxybenzotriazole,
Tr =trityl).
conformation in solution.[35–38] A recent report has shown
that several cyclic glycopeptide analogues of tyrocidine A
retain biological activity as antibiotics.[39] To further test the
method reported herein for the synthesis of S-linked glyco-
peptides, glycosyl amino acid 15 was synthesized and con-
verted into 16. After the removal of palladium by scavenger
resin, the tyrocidine analogue precursor 17 was produced
(Scheme 4). Conversion of 17 to cyclic glycopeptide 18 (an
analogue of tyrocidine A in which the wild-type Tyr group at
position 7 is replaced by a d-Ser(b-S-GlcNH2) residue) was
carried out in four steps on solid support followed by cleavage
and amino acid side chain deprotection (29% overall yield
from 16 to 18). The biological activity of this glycosylated
tyrocidine analogue was evaluated with the minimal inhib-
itory concentration (MIC) against B. subtilis as a measure of
Scheme 2. A direct two-step one-pot synthesis of the S-linked Fmoc-modified
glycosyl amino acid 10 with suppression of side product 11.
was favored over the benzyl ester group, as it is easier to
remove in high yield. Bromination of 8a went smoothly with
the solid-support Ph3P resin and CBr4 in dichloromethane to
give 9a (100%, TLC),[33] which was pure enough to carry out
the next step without purification.[34] Upon performing the
SN2 reaction at ambient temperature, the desired amino acid
10 was obtained (54%), along with side-product 11 (27%).
Thiolate 5 presumably attacks Fmoc-CH2 to give 11. T o the antibiotic activity and the minimal hemolytic concentra-
suppress the formation of this side-product, the reaction was
performed at À788C to room temperature. We were gratified
to observe that only compound 10 (79%) was formed without
any trace of side product 11.
tion (MHC) against human erythrocytes as an evaluation of
eukaryotic membrane toxicity. The S-linked glycopeptide
analogue 18 was shown to have a better antibiotic profile than
tyrocidine A, with MIC values of 12.5 mm for both and MHC
values of 100 and 50 mm, respectively.
Although Fmoc SPPS is preferred for glycopeptides, the
synthetic approach reported herein should also be applicable
for the synthesis of Boc-modified amino acid building blocks.
Therefore, Boc-protected S-linked glycosyl amino acid build-
ing block 19 (Table 1) was synthesized from 4 and 9b
(Scheme 2) in a similar fashion.
To demonstrate that our approach is also applicable to
other thiosugars, we synthesized several different peracety-
lated 1-thiosugars (Supporting Information). These represent
all four possible classes of products: 1,2-trans-a (21 and 23[15]);
1,2-trans-b (4,[28] 24,[40] 26,[41] 27,[42] and 28[43]); 1,2-cis-a (20 and
25[23]); and 1,2-cis-b (22[29]). They are conveniently grouped
based on the stereochemistry at C1 and its relationship to the
functional group at C2 of the sugar ring. All compounds were
subjected to the “two-step one-pot” conditions mentioned
above to give the corresponding Fmoc- or Boc-protected S-
linked glycosyl amino acid building blocks, in yields ranging
from 66 to 81%.
S-linked glycosyl amino acid 10 was treated with tetra-
kis(triphenylphosphine)palladium(0) catalyst and morpholine
in dichloromethane to give the Fmoc-modified building block
12 swiftly (100%, TLC), which was used without purification
to afford the protected S-linked glycodecapeptide 13
(Scheme 3). Deacetylation of the sugar moiety was carried
out on-resin by transesterification with hydrazine hydrate in
methanol. Removal of the acid-labile amino acid protecting
groups concomitant with cleavage from the 2-chlorotrityl
resin by Reagent K afforded free S-linked glycopeptide 14 in
pure form after HPLC. The whole process, starting from the
readily available peracetylated 1-thio galactose to the final
glycopeptide 14, involved only two purifications (silica gel
chromatography for 10 and HPLC for 14). In fact, it is also
possible to start with thiolate salt 5, which can be stored under
argon for up to a month without degradation.
Glycopeptide 14 is a linear analogue of tyrocidine A,
which is a cyclic cationic decapeptide antibiotic produced in
Bacillus brevis with an antiparallel amphipathic b-sheet
4598
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 4596 –4599

Angewandte
Chemie
[5] H. T. Liu, L. Wang, A. Brock, C. H. Wong, P. G.
Schultz, J. Am. Chem. Soc. 2003, 125, 1702.
[6] Z. W. Zhang, J. Gildersleeve, Y. Y. Yang, R. Xu,
J. A. Loo, S. Uryu, C. H. Wong, P. G. Schultz,
Science 2004, 303, 371.
[7] J. D. Warren, J. S. Miller, S. J. Keding, S. J. Dani-
shefsky, J. Am. Chem. Soc. 2004, 126, 6576.
[8] O. Sietz, ChemBioChem 2000, 1, 214.
[9] H. Herzner, T. Reipen, M. Schultz, H. Kunz,Chem.
Rev. 2000, 100, 4495.
[10] P. M. St. Hilaire, M. Meldal, Angew. Chem. 2000,
112, 1210; Angew. Chem. Int. Ed. 2000, 39, 1162.
[11] B. G. Davis, Chem. Rev. 2002, 102, 579.
[12] B. Capon, Chem. Rev. 1969, 69, 407.
[13] H. Driguez, Top. Curr. Chem. 1997, 187, 85.
[14] Z. J. Witczak, Curr. Med. Chem. 1999, 6, 165.
[15] H. N. Yu, PhD Thesis, University of Alberta, 2002.
[16] E. Bousquet, A. Spadaro, M. S. Pappalardo, R.
Bernardini, R. Romeo, L. Panza, G. Ronsisvalle, J.
Carbohydr. Chem. 2000, 19, 527.
[17] J. R. Rich, D. R. Bundle, Org. Lett. 2004, 6, 897.
[18] B. Kuberan, S. A. Sikkander, H. Tomiyama, R. J.
Linhardt, Angew. Chem. 2003, 115, 2119; Angew.
Chem. Int. Ed. 2003, 42, 2073.
[19] A. Dondoni, A. Marra, Chem. Rev. 2000, 100, 4395.
[20] Y. Ohnishi, M. Ichikawa, Y. Ichikawa, Bioorg.
Med. Chem. Lett. 2000, 10, 1289.
[21] L. Jobron, G. Hummel, Org. Lett. 2000, 2, 2265.
[22] S. B. Cohen, R. L. Halcomb, Org. Lett. 2001, 3, 405.
[23] S. Knapp, D. S. Myers, J. Org. Chem. 2002, 67, 2995.
[24] X. M. Zhu, R. R. Schmidt, Chem. Eur. J. 2004, 10,
875.
[25] X. M. Zhu, R. R. Schmidt, Tetrahedron Lett. 2003,
44, 6063.
[26] X. M. Zhu, K. Pachamuthu, R. R. Schmidt, J. Org.
Chem. 2003, 68, 5641.
Scheme 4. Solid-phase glycopeptide synthesis of the tyrocidine analogue 18.
Reagents and conditions: a) [Pd(PPh3)4]/4-methylmorpholine, CHCl3; b) piperidine/
DMF (20%); c) PyBOP/HOBt, DMF; d) H2NNH2·H2O, MeOH; e) TFA/CH2Cl2/triiso-
propylsilane (80:15:5); 29% overall yield from 16. PyBOP=1-benzotriazolyloxy-tris-
(pyrollidino)phosphonium; TFA =trifluoroacetate.
[27] X. M. Zhu, K. Pachamuthu, R. R. Schmidt, Org.
Lett. 2004, 6, 1083.
[28] K. P. R. Kartha, P. Cura, M. Aloui, S. K. Readman, T. J.
Rutherford, R. A. Field, Tetrahedron: Asymmetry 2000, 11, 581.
[29] H. N. Yu, C. C. Ling, D. R. Bundle, J. Chem. Soc. Perkin Trans. 1
2001, 832.
[30] E. Atherton, R. C. Sheppard, Solid Phase Peptide Synthesis: A
Practical Approach, IRL Press at Oxford University Press,
Oxford, 1989.
[31] G. B. Fields, R. L. Noble, Int. J. Pept. Protein Res. 1990, 35, 161.
[32] S. Friedrich-Bochnitschek, H. Waldmann, H. Kunz, J. Org.
Chem. 1989, 54, 751.
[33] D. H. Drewry, D. M. Coe, S. Poon, Med. Res. Rev. 1999, 19, 97.
[34] Compound 9a is not stable in silica gel.
[35] R. D. Hotchkiss, R. J. Dubos, J. Biol. Chem. 1940, 132, 791.
[36] R. D. Hotchkiss, R. J. Dubos, J. Biol. Chem. 1940, 132, 793.
[37] R. D. Hotchkiss, R. J. Dubos, J. Biol. Chem. 1940, 136, 803.
[38] M.-C. Kuo, W. A. Gibbons, Biophys. J. 1980, 32, 807.
[39] H. Lin, D. A. Thayer, C. H. Wong, C. T. Walsh, Chem. Biol. 2004,
11, 1635.
In summary, we have developed a general strategy for S-
linked glycopeptide synthesis. With a simple “two-step one-
pot” reaction, peracetylated thiosugars are quickly converted
into masked S-linked amino acid building blocks. Under such
reaction conditions, deacetylation at the sulfur atom is
selective, anomeric integrity is retained, and SN2 substitution
is stereoselective. Unmasking the building blocks by deal-
lylation affords Fmoc- or Boc-protected amino acids, which
are ready for SPPS. This strategy is applicable for large-scale
S-linked glycopeptide synthesis and could also be used for
other S-linked glycoconjugates.
Received: January 11, 2005
Revised: April 21, 2005
Published online: June 30, 2005
[40] Obtained by the same procedure as used for compound 22.
[41] Obtained by the same procedure as used for compound 4, but
with DMF as the solvent.
[42] Obtained by the same procedure as used for compound 4, but
with acetone as the solvent.
Keywords: glycopeptides · S-linked glycopeptides ·
solid-phase synthesis · sulfur
.
[43] V. Moreau, J.-L. Viladot, E. Samain, A. Planas, H. Driguez,
Bioorg. Med. Chem. 1996, 4, 1849.
[1] R. A. Dwek, Chem. Rev. 1996, 96, 683.
[2] H. Lis, N. Sharon, Eur. J. Biochem. 1993, 218, 1.
[3] A. Varki, Glycobiology 1993, 3, 97.
[4] R. S. Rush, P. L. Derby, D. M. Smith, C. Merry, G. Rogers, M. F.
Rohde, Anal. Chem. 1995, 67, 1442.
Angew. Chem. Int. Ed. 2005, 44, 4596 –4599
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4599