Solid-phase synthesis of thioether-linked glycopeptide mimics for application to glycoprotein semisynthesis.

Article abstract of DOI:10.1021/ol025627w

[reaction: see text]. Glycoproteins are particularly suited to protein semisynthesis since homogeneous samples for biological analyses are not readily available using traditional recombinant techniques. Here we apply glycosyl iodoacetamides, normally used for the modification of bacterially derived proteins, to solid-phase glycopeptide synthesis. This provides access to glycopeptide alpha-thioesters, which may lend themselves to the semisynthesis of homogeneous N-linked glycoprotein mimics and novel glycopeptide libraries.

Full text of DOI:10.1021/ol025627w

ORGANIC
LETTERS
2002
Vol. 4, No. 9
1467-1470
Solid-Phase Synthesis of
Thioether-Linked Glycopeptide Mimics
for Application to Glycoprotein
Semisynthesis
Derek Macmillan,* Alison M. Daines, Monika Bayrhuber, and Sabine L. Flitsch*
Department of Chemistry, UniVersity of Edinburgh, Kings Buildings, West Mains Road,
Edinburgh EH9 3JJ, Scotland
derek.macmillan@ed.ac.uk
Received January 26, 2002
ABSTRACT
Glycoproteins are particularly suited to protein semisynthesis since homogeneous samples for biological analyses are not readily available
using traditional recombinant techniques. Here we apply glycosyl iodoacetamides, normally used for the modification of bacterially derived
proteins, to solid-phase glycopeptide synthesis. This provides access to glycopeptide r-thioesters, which may lend themselves to the
semisynthesis of homogeneous N-linked glycoprotein mimics and novel glycopeptide libraries.
It is well established that protein glycosylation can be a vital
co- and post-translational modification for the normal growth
and development of organisms. This is compounded by the
knowledge that the glycan moiety of glycoproteins is
implicated in an increasing number of important biological
processes from correct protein folding and secretion to cell-
cell recognition.1 Understanding how and why these highly
specific interactions come about has always been hindered
by the difficulties associated with the “microheterogeneity”
of glycoproteins.2
novel methods for the synthesis of simpler structural
analogues (glycopeptide mimics) that have been employed
successfully in various glycobiological studies.3 The main
aim of such an approach is to simplify the synthesis relative
to the natural products. However, some additional advanta-
geous properties such as resistance to hydrolysis may also
be conferred while remaining functionally silent.
The first report on the use of glycosyl iodoacetamides such
as 1 (Figure 1) for the synthesis of N-linked glycoprotein
mimics4 was soon followed by a report discussing modifica-
tion of short unprotected synthetic peptides sequences
containing a single cysteine residue in solution.5 More
As a result of the problems associated with the synthesis
of glycopeptides and glycoproteins, chemists have sought
* Email for S.L.F.: s.flitsch@ed.ac.uk.
(3) Reviewed in: Marcaurelle, L. A.; Bertozzi, C. R. Chem. Eur. J. 1999,
5, 1384. Davis, B. G.; Jones, J. B. Synlett 1999, 1495. See also: Allen, J.
R.; Harris, C. R.; Danishefsky, S. J. J. Am. Chem. Soc. 2001, 123, 1890.
Marcaurelle, L. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2001, 123, 1587.
Schafer, A.; Klich, G.; Schreiber, M.; Paulsen, H.; Thiem, J. Carbohydr.
Res. 1998, 313, 107. Wang, L. X.; Tang, M.; Suzuki, T.; Kitajima, K.;
Inoue, Y.; Inoue, S.; Fan, J. Q.; Lee, Y. C. J. Am. Chem. Soc. 1997, 119,
11137.
(1) Parodi, A. J. Annu. ReV. Biochem. 2000, 69, 69. Varki, A. Glyco-
biology 1993, 3, 97. Rudd, P. M.; Wormald, M. R.; Stanfield, R. L.; Huang,
M. D.; Mattsson, N.; Speir, J. A.; DiGennaro, J. A.; Fetrow, J. S.; Dwek,
R. A.; Wilson, I. A. J. Mol. Biol. 1999, 293, 351. Helenius, A.; Aebi, M.
Science 2001, 291, 2364. Frigerio, L.; Lord, J. M. Curr. Biol. 2000, 10,
R674. Dwek, R. A. Chem. ReV. 1996, 96, 683.
(2) Rush, R. S.; Derby, P. L.; Smith, D. M.; Merry, C.; Rogers, G.;
Rohde, M. F.; Katta, V. Anal. Chem. 1995, 67, 1442.
(4) Davis, N. J.; Flitsch, S. L. Tetrahedron Lett. 1991, 32, 6793.
10.1021/ol025627w CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/09/2002

Table 1. Examples of Glycopeptide Mimics Formed Using
Solid-Phase Chemistry Depicted in Figure 1 and “thio”-CD52 as
the Peptide Scaffolda
a Enzymatic transformations can also be carried out on the glycopeptide
products.
Figure 1. Model CD52 study. (a) 10% DTT sat. NH4HCO3/DMF,
24 h, quant. (b) Glycosyl iodoacetamide (3 equiv), DMF, 2.5%
v/v pyridine, 3 h, 95-100%. (c) 95% TFA, 3 h.
the synthesis of peptides containing post-translational modi-
fications such as phosphorylation9 and glycosylation10 that
are not under direct genetic control. We therefore aimed to
widen the scope of the original iodoacetamide method to
include protein semisynthesis. Through chemical synthesis,
it should be trivial to differentiate between cysteines involved
in disulfide bonds and those to be modified by glycosylation
using orthogonally protected cysteine residues. We also
aimed to conduct all reactions and glycopeptide synthesis
on the solid phase using 4-sulfamylbutyryl-AM resin, al-
lowing glycopeptide mimetic R-thioesters to be prepared
using standard Fmoc chemistry.11 This facilitated application
recently this and closely related methodologies have provided
the first examples of homogeneously glycosylated, bacterially
derived proteins. In such cases, the site of modification has
been specifically targeted using a mutagenesis/modification
approach. In this way specific protein glycoforms in the 20-
25 KDa molecular weight range have been prepared.6 The
iodoacetamide methodology may, however, be limited by
obstacles, including incomplete modification of free thiols
on recombinant proteins, which then require further separa-
tion. Furthermore, the introduction of extra cysteine residues
into proteins may not be wholly tolerated, for example, where
the introduction of more than one cysteine residue into a
recombinantly derived protein is desirable, yet the protein
already contains many other cysteine residues normally
involved in disulfide bond formation. These potential limita-
tions have driven the search for a means of expanding the
present methodology.
(8) Muir, T. W.; Sondhi, D.; Cole, P. A. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 6705. Noren, C. J.; Wang, J. M.; Perler, F. B. Angew. Chem.,
Int. Ed. 2000, 39, 451.
(9) Huse, M.; Holford, M. N.; Kuriyan, J.; Muir, T. W. J. Am. Chem.
Soc. 2000, 122, 8337. Flavell, R. R.; Huse, M.; Goger, M.; Trester-Zedlitz,
M.; Kuriyan, J.; Muir, T. W. Org. Lett. 2002, 4, 165.
(10) Shin, Y.; Winans, K. A.; Backes, B. J.; Kent, S. B. H.; Ellman, J.
A.; Bertozzi, C. R. J. Am. Chem. Soc. 1999, 121, 11684. Winans, K. A.;
King, D. S.; Rao, V. R.; Bertozzi, C. R. Biochemistry 1999, 38, 11700.
Marcaurelle, L. A.; Mizoue, L. S.; Wilken, J.; Oldham, L.; Kent, S. B. H.;
Handel, T. M.; Bertozzi, C. R. Chem. Eur. J. 2001, 7, 1129. Tolbert, T. J.;
Wong, C.-H. J. Am. Chem. Soc. 2000, 122, 5421. Macmillan, D.; Bertozzi,
C. R. Tetrahedron 2000, 56, 9515.
Methods for protein semisynthesis, particularly native
chemical ligation7 and expressed protein ligation8 are par-
ticularly suited to glycoprotein synthesis since they enable
(11) Backes, B. J.; Ellman, J. A. J. Org. Chem. 1999, 64, 2322. Shin,
Y.; Winans, K. A.; Backes, B. J.; Kent, S. B. H.; Ellman, J. A.; Bertozzi,
C. R. J. Am. Chem. Soc. 1999, 121, 11684.
(12) Herzner, H.; Reipen, T.; Schultz, M.; Kunz, H. Chem. ReV. 2000,
100, 4495.
(13) Diekman, A. B.; Norton, E. J.; Klotz, K. L.; Westbrook, V. A.;
Herr, J. C. Immunol. ReV. 1999, 171, 203.
(14) For peptide synthesis, see: Macmillan, D.; Bertozzi, C. R. Tetra-
hedron 2000, 56, 9515.
(15) St Hilaire, P. M.; Meldal, M. Angew. Chem., Int. Ed. 2000, 39,
1163.
(5) Wong, S. Y. C.; Guile, G. R.; Dwek, R. A.; Arsequell, G. Biochem.
J. 1994, 300, 843.
(6) Bill, R. M.; Winter, P. C.; McHale, C. M.; Hodges, V. M.; Elder, G.
E.; Caley, J.; Flitsch, S. L.; Bicknell, R.; Lappin, T. R. J. Biochim. Biophys.
Acta-Gene Struct. Express. 1995, 1261, 35. Davis, B. G. J. Chem. Soc.,
Perkin Trans. 1 1999, 3215. Davis, B. G.; Lloyd, R. C.; Jones, J. B. Bioorg.
Med. Chem. 2000, 8, 1527. Davis, B. G. Chem. Commun. 2001, 351.
Macmillan, D.; Bill, R. M.; Sage, K. A.; Fern, D.; Flitsch, S. L. Chem.
Biol. 2001, 8, 133.
(7) Dawson, P. E.; Kent, S. B. H. Annu. ReV. Biochem. 2000, 69, 923.
1468
Org. Lett., Vol. 4, No. 9, 2002

merization of glycosylamines (which results in an R/â
mixture of glycopeptide products).12
We began with a model system on which to optimize the
chemistry. CD52, a cell surface glycopeptide, was chosen
as it contains N-linked glycosylation and is modified at its
C-terminus with a GPI anchor.13 The peptide moiety was
first synthesized on acid-labile NovaSyn-TGT resin, allowing
facile monitoring of each process. The peptide was prepared
using standard Fmoc chemistry, replacing the native aspar-
agine residue with the commercially available S-tButhio
protected cysteine. Following peptide synthesis,14 the tert-
butylthio protecting group was removed by subjecting the
resin to 10% DTT in DMF (stirred previously with solid
(NH4)2CO3). The free thiol was then derivatized using
synthetic saccharide bromoacetamides or iodoacetamides
(Figure 1).
Generally, quantitative sulfhydryl modification was achieved
within 1.5 h using 3 equiv of glycosyl haloacetamide. The
resin-bound products were then cleaved with 95% TFA,
purified by semipreparative HPLC, and characterized by
mass spectrometry (Table 1).
Figure 2. Extension to sulfonamide resin affords the “thio”-CD52
glycopeptide mimic R-benzylthioester after cleavage. Calculated
mass ) 1563.9 Da, observed mass ) 1563.9 Da.
The ability for different saccharides to be efficiently ligated
to a peptide scaffold may also facilitate glycopeptide library
synthesis,15 potentially useful for the display of carbohydrate-
dependent T-cell antigens, where the peptide moiety is
present only to enable binding within the major histo-
compatibility complex (MHC) peptide binding cleft.16
to native chemical ligation or further C-terminal modification.
This also overcomes various problems associated with
forming native linkages between the carbohydrate and
peptide moieties such as aspartamide formation and ano-
Next, the potential for this chemistry to be applied to native
chemical ligation and expressed protein ligation was ex-
Figure 3. (a) Fmoc-Gly-OH, PyBOP, DIPEA, DMF, rt. (b) Fmoc SPPS. (c) 10% DTT, 2.5% v/v DIPEA, DMF, 24 h. (d) 2.5% v/v
pyridine, DMF, 3 h (e) (1) ICH2CN, DIPEA, DMF, 16 h; (2) BnSH, NaSPh, THF, 16 h. (f) 85% TFA, 5% H2O 5% EDT, 5% thioanisole.
Org. Lett., Vol. 4, No. 9, 2002
1469

plored. The same CD52 glycopeptide mimic was thus
prepared on 4-sulfamylbutyryl-AM (safety-catch) resin and
cleaved as described.17 The glycopeptide was released from
an aliquot (50 mg) of resin with benzyl mercaptan in the
presence of catalytic NaSPh after activation with iodoaceto-
nitrile. The backbone protecting groups were then cleaved
(85% TFA, 5% ethanedithiol, 5% thioanisole, 5% water, 3
h), and the pure glycopeptide R-thioester product (2.5 mg,
32%) was obtained after ether precipitation and semiprepara-
tive HPLC (Figure 2).
To test the methodology further and to test specifically
for the ability to differentiate between cysteine residues
within the same peptide sequence, we applied this method
to the synthesis of the first 28 amino acids of the glycoprotein
hormone erythropoietin (EPO). Within this fragment is
contained the first glycosylation site (Asn24) and a cysteine
residue (Cys-6) that is involved in disulfide bond formation.
We aimed to differentiate between the cysteine residues using
the commercially available tert-butylthio-protected cysteine
as described above (Figure 1) to release the free thiol for
modification (Asn24Cys) and S-trityl modification for Cys-
6. The peptide was prepared using standard Fmoc solid-phase
synthesis on 4-sufamylbutyryl resin. This would ultimately
afford a C-terminal thioester for native chemical ligation.
Figure 3 demonstrates how this peptide was prepared.
Again, the glycopeptide product was characterized by
electrospray mass spectrometry after HPLC purification
(Figure 4) and will be used for the synthesis of semisynthetic
erythropoietin in future studies.
Figure 4. Electrospray mass spectrum of the glycosylated EPO
fragment (residues 1-28), prepared as a C-terminal thioester.
proteins has been expanded to encompass solid-phase glyco-
peptide synthesis, providing access to homogeneously gly-
cosylated N-linked glycopeptide mimic R-thioesters. The
products can be applied to native chemical ligation and
expressed protein ligation. Furthermore, we can achieve
quantitative modification of specific, introduced cysteine
residues using as little as 3 equiv of glycosyl haloacetamides.
In doing so, some potential problems associated with the
original approach such as incomplete sulfhydryl modification,
lengthy purification, and ambiguity in the site of modification
can be abolished. It should be noted that it is the stability of
the thioether-linked saccharides that allows this chemistry
to be applied to native chemical ligation. Other common
approaches that rely on linking oligosaccharides through a
disulfide bond to cysteine would likely be unsuitable for such
methods.
In summary, the use of glycosyl haloacetamides for the
incorporation of oligosaccharides into bacterially derived
(16) Komba, S.; Meldal, M.; Werdelin, O.; Jensen, T.; Bock, K. J. Chem.
Soc., Perkin Trans. 1 1999, 415. Gagne, R.; Werdelin, O.; Gad, M.; Komba,
S.; Meldal, M. Glycobiology 2000, 10, 202. Keil, S.; Claus, C.; Dippold,
W.; Kunz, H. Angew. Chem., Int. Ed. 2001, 40, 366. Hilaire, P. M. S.;
Cipolla, L.; Franco, A.; Tedebark, U.; Tilly, D. A.; Meldal, M. J. Chem.
Soc., Perkin Trans. 1 1999, 3559.
(17) The resin-bound, protected glycopeptide mimetic (50-100 mg) was
taken up in DMF (2.0 mL), treated with DIPEA (100 µL) and then
iodoacetonitrile (filtered through basic alumina, 100 µL), and stirred at room
temperature with the exclusion of light. After 16 h the resin was washed
(DMF, then DCM, then THF) and taken up in anhydrous THF (2.0 mL).
Benzyl mercaptan (100 µL) and solid NaSPh (2 mg) were added, and stirring
was continued. After 16 h the solvent was removed by evaporation, and
the crude peptide was obtained after precipitation with ether. See also: Shin,
Y.; Winans, K. A.; Backes, B. J.; Kent, S. B. H.; Ellman, J. A.; Bertozzi,
C. R. J. Am. Chem. Soc. 1999, 121, 11684. Ingenito, R.; Bianchi, E.; Fattori,
D.; Pessi, A. J. Am. Chem. Soc. 1999, 121, 11369.
Acknowledgment. We acknowledge financial support
from the BBSRC.
Supporting Information Available: Full experimental
1
procedures, MS and H and 13C NMR data for glycosyl
haloacetamide, and mass spectra for selected glycopeptides
(6-8, 10, 11). This material is available free of charge via
the Internet at http://pubs.acs.org.
OL025627W
1470
Org. Lett., Vol. 4, No. 9, 2002