Native N-glycopeptide thioester synthesis through N→S acyl transfer

Article abstract of DOI:10.1016/j.bmcl.2011.05.059

Peptide thioesters are important tools for the total synthesis of proteins using native chemical ligation (NCL). Preparation of glycopeptide thioesters, that enable the assembly of homogeneously glycosylated proteins, is complicated by the perceived fragile nature of the sugar moiety. Herein, we demonstrate the compatibility of thioester formation via N→S acyl transfer with native N-glycopeptides and report observations that will aid in their preparation.

Full text of DOI:10.1016/j.bmcl.2011.05.059

Bioorganic & Medicinal Chemistry Letters 21 (2011) 4973–4975
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Native N-glycopeptide thioester synthesis through N?S acyl transfer
⇑
Bhavesh Premdjee, Anna L. Adams, Derek Macmillan
Department of Chemistry, University College London, 20 Gordon Street, London WC1H 0AJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Peptide thioesters are important tools for the total synthesis of proteins using native chemical ligation
(NCL). Preparation of glycopeptide thioesters, that enable the assembly of homogeneously glycosylated
proteins, is complicated by the perceived fragile nature of the sugar moiety. Herein, we demonstrate
the compatibility of thioester formation via N?S acyl transfer with native N-glycopeptides and report
observations that will aid in their preparation.
Received 20 March 2011
Revised 17 May 2011
Accepted 18 May 2011
Available online 25 May 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Glycopeptides
Native chemical ligation
Peptide thioesters
Glycoproteins are increasingly becoming viewed as tractable
synthetic targets for development and exploitation as therapeu-
tics.1 Consequently, efficient methods for the preparation of glyco-
peptide thioesters, the key building blocks for protein synthesis
using Native Chemical Ligation (NCL), are highly desirable.2
Previously, we showed that peptide thioesters (1) can be prepared
via N?S acyl transfer from readily available precursors such as 2,
equipped simply with a C-terminal cysteine residue (Scheme 1).3
While demonstrating a preference for thioester formation at
certain motifs over others,3 several additional factors appear to
influence this transformation. In particular, variables such as
peptide concentration, solubility and complexity appear impor-
tant but have not been dissected in detail. Peptide thioesters
adorned with chemically fragile post-translational modifications
such as glycosylation are likely to benefit most from an opti-
mized reaction protocol. Therefore model experiments were con-
ducted with a view to accessing native N-glycopeptide thioesters
that could be used for glycoprotein assembly.
A model N-glycopeptide (Scheme 2) corresponding to erythro-
poietin (EPO) residues 22–29 was prepared (see Supplementary
data for experimental details). Acetate esters were retained on
the sugar moiety to aid analysis and purification and prevent the
fully deprotected glycopeptide from eluting too rapidly from the
reverse-phase column. Glycopeptide 3 was subjected to thioester
formation at 1 mg mlÀ1 (approx. 0.9 mM) peptide concentration
in 0.1 M Na phosphate buffer; pH 5.8, containing 10% w/v sodium
2-mercaptoethanesulfonate (MESNa) and 0.5% w/v tris-carboxy-
ethylphosphine (TCEP) at 55 °C for 48 h. The reaction was moni-
tored by HPLC, and LC–MS, and showed smooth transition to
thioester 4
within 48 h (Fig. 1a). Furthermore, 1H NMR
spectroscopy of the purified product (Fig. 1b) confirmed thioester
formation. Conversion to the desired thioester was additionally
supported through NCL between 4 and an EPO fragment corre-
sponding to residues 29–166 (Fig. 1c). The large EPO fragment
was expressed in Escherichia coli and isolated as previously
described.4 The ligation reaction was conducted in 0.3 M sodium
phosphate buffer; pH 7.0 containing 6 M guanidineÁHCl, 0.1 M
4-mercaptophenylacetic acid, and 40 mM TCEP. After 3 h LC–MS
indicated that the initial recombinant fragment had been con-
sumed. Treatment of the ligation reaction mixture with hydrazine
hydrate in the presence of dithiothreitol for 1 h afforded the full
length glycopolypeptide (Fig. 1d).
While the model experiments demonstrated that N-glycopoly-
peptides can be assembled using simpler N-glycopeptide thioes-
ters formed through N?S acyl transfer, the isolated yield of 4
was rather low (20–40%). We considered thioester formation,
as depicted in Scheme 1, to be essentially unidirectional, since
the relatively low pH (pH 2–6) and high concentration of thiol
additive (usually 10% w/v MESNa) employed are known to inhi-
bit NCL.5 Consequently ligation between the thioester products
and the released C-terminal cysteine should be negligible under
the reaction conditions. However, the observation that thioester
synthesis did not proceed to completion meant we could not
rule out the possibility that competing NCL compromised
H3N
CONH2
MESNa
SH
O
O
H3O+
O
SO3H
S
N
H
CONH2
S
2
1
⇑
Corresponding author.
E-mail address: d.macmillan@ucl.ac.uk (D. Macmillan).
Scheme 1. Peptide thioester synthesis employing an N?S acyl shift.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.05.059

4974
B. Premdjee et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4973–4975
OAc
O
SH
O
SH
O
O
D-Cys
MESNa
CONH2
L-Cys
H
N
AcO
AcO
SO3H
N
H
S
N
H
CO2H
MESNa
NHAc
O
O
OH
R
R
R
O
O
H
N
H
N
H
H
N
5
7
m/z = 1138.5
m/z= 1137.5
6
m/z =1159.5
N
CO2H
H2N
N
H
N
H
N
H
H2O
O
O
O
O
O
OH
HS
CO2H
3
8 m/z= 1035.5
OH
10% w/v MESNa
R
0.1 M Na phosphate buffer; pH 5.8
0.5% w/v TCEP, 55 oC, 48h
OAc
O
Scheme 3. Thioester formation and competing NCL were investigated, under
H
N
AcO
AcO
identical reaction conditions, by conducting the reaction in the presence of
cysteine.
D-
NHAc
O
O
OH
O
O
H
H
H
N
N
N
S
H2N
N
N
N
H
SO3H
H
H
simpler detection of 7 by virtue of the C-terminal carboxyl, rather
than carboxamide group present in 5, which additionally allowed 5
and 7 to be unambiguously distinguished by mass.
O
O
O
O
OH
4
CO2H
Scheme 2. N-Glycopeptide thioester formation at pH 5.8.
In the absence of D-cysteine, thioester 6 was observed to accu-
mulate to approximately 60% conversion after 48 h with hydrolysis
to 8 appearing as the only significant side reaction (Fig. 2).
Surprisingly, 7 could be observed in the reaction mixture upon
reaction efficiency. If NCL was to influence the reaction outcome
then we postulated that lowering the initial peptide concentra-
tion and reaction pH (from pH 5.8 to pH 2.0) should benefit thi-
oester formation.
To test this hypothesis, a model peptide 5 (Sequence: H-MEE-
LYKSHC-NH2) derived from the C-terminal residues of green fluo-
rescent protein was first subjected to thioester formation in the
addition of less than 1 equiv of
come a significant constituent of the reaction mixture at less than
10 equiv. As the concentration of -cysteine increased (>350 equiv)
in the reaction mixture thioester formation was dramatically re-
duced to only 30%. At high -Cys concentrations an additional
D-Cys and could accumulate to be-
D
D
product, 9, potentially corresponding to a disulfide-bonded dimer
of 7 was also observed. Inter- or intramolecular disulfide bond for-
mation is also a competing reaction that can inhibit N?S acyl
transfer, although no disulfide-bonded adduct between 5 and
presence of increasing concentrations of
D
-cysteineÁHCl. We envis-
aged that overall epimerization of the C-terminal cysteine resulting
from the cysteine exchange reaction (Scheme 3) should, in itself,
produce products that were sufficiently resolvable so as to assess
D-Cys could be observed. Thioester formation in the presence of
D
-Cys was additionally investigated at lower pH (pH ꢀ2) employ-
the occurrence of NCL by HPLC. Furthermore, addition of
D-Cys-
ing 10% v/v AcOH as solvent (see Supplementary data). In this case,
OH, rather than -Cys-NH2 to the reaction mixture facilitated
D
4
a
c
d
Asn24
Cys29
*
3
EPO (22-28)
*
*
48 h
24 h
6 h
EPO (29-166)
4
i) 0.1 M MPAA, pH 7.0,3 h
ii) 5% v/v hydrazine hydrate, DTT
0 h
Asn24
Cys29
EPO (29-166)
Calculated
m/z= 16182Da
EPO (22-28)
b
2 x MESNa CH2
Asn CH2- β
Glu CH2
16181 Da
Cys CH2- β
Asn CH2- β
Glu CH2
ppm
mass
Figure 1. (a) HPLC analysis of thioester formation. Peaks marked with asterisks correspond to derivatives of 3 or 4 with a cleaved acetate ester. (b) 1H NMR analysis of the
purified thioester 4 (upper trace) and 3 (lower trace, see Supplementary data for full spectrum). (c) NCL between 4 and recombinant EPO fragment comprising residues 29–
166. (d) ESI mass spectrum of the crude ligation product following carbohydrate deacetylation.

B. Premdjee et al. / Bioorg. Med. Chem. Lett. 21 (2011) 4973–4975
4975
thioester formation. However, this should not be significant under
normal reaction conditions. The results also emphasize the
remarkable dynamic nature of amide bonds susceptible to an
NCL/retro-NCL process, even in the absence of synthetic ‘devices’7
or inteins.8
Acknowledgments
The authors acknowledge Professor Philip E. Dawson (Scripps
Research Institute) for helpful discussions, and Jaskiranjit Kang
for performing preliminary experiments. We also acknowledge
financial support from The Royal Society, Dextra Laboratories, Uni-
versity College London, and The Wellcome Trust.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.05.059.
References and notes
Figure 2. Relative composition (%) of the reaction mixture as a function of time and
added
D
-CysÁHCl. The reactions were all conducted at 1 mg mlÀ1 (approx. 0.9 mM)
1. (a) Hojo, H.; Nakahara, Y. Peptide Sci. 2007, 88, 308; (b) Boltje, T. J.; Buskas, T.;
Boons, G.-J. Nat. Chem. 2009, 1, 611; (c) Gamblin, D. P.; Scanlan, E. M.; Davis, B. G.
Chem. Rev. 2009, 109, 131; (d) Seeberger, P. H. Nat. Chem. Biol. 2009, 5, 368; (e)
Wang, P.; Danishefsky, S. J. J. Am. Chem. Soc. 2010, 132, 17045; (f) Kan, C.;
Danishefsky, S. J. Tetrahedron 2009, 65, 9047; (g) Hirano, K.; Macmillan, D.;
Tezuka, K.; Tsuji, T.; Kajihara, Y. Angew. Chem., Int. Ed. 2009, 48, 9557; (h)
Yamamoto, N.; Tanabe, Y.; Okamoto, R.; Dawson, P. E.; Kajihara, Y. J. Am. Chem.
Soc. 2008, 130, 501; (i) Piontek, C.; Silva, D. V.; Heinlein, C.; Pöhner, C.; Mezzato,
S.; Ring, P.; Martin, A.; Schmid, Franz X.; Unverzagt, C. Angew. Chem., Int. Ed.
2009, 48, 1941; (j) Piontek, C.; Ring, P.; Harjes, O.; Heinlein, C.; Mezzato, S.;
Lombana, N.; Pöhner, C.; Püttner, M.; Silva, D. V.; Martin, A.; Schmid, Franz X.;
Unverzagt, C. Angew. Chem., Int. Ed. 2009, 48, 1936; (k) Macmillan, D.; Bertozzi,
C. R. Angew. Chem., Int. Ed. 2004, 43, 1355.
2. (a) 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; (b) Mezzato, S.; Schaffrath, M.; Unverzagt, C.
Angew. Chem., Int. Ed. 2005, 44, 1650; (c) Ozawa, C.; Katayama, H.; Hojo, H.;
Nakahara, Y.; Nakahara, Y. Org. Lett. 2008, 10, 3531; (d) Hirano, K.; Kajihara, Y. J.
Carbohydr. Chem. 2010, 29, 84.
3. (a) Kang, J.; Richardson, J. P.; Macmillan, D. Chem. Commun. 2009, 407; (b) Kang,
J.; Reynolds, N. L.; Tyrrell, C.; Dorin, J. R.; Macmillan, D. Org. Biomol. Chem. 2009,
7, 4918.
peptide concentration in 0.1 M Na phosphate buffer; pH 5.8, 10% w/v MESNa, 0.5%
w/v TCEPÁHCl, 60 °C, 48 h.
the reaction proceeded essentially as before although thioester
hydrolysis appeared slightly reduced.
The results suggest that such reversibility in amide bond forma-
tion, enabled by an NCL/retro-NCL sequence, could be extended to
dynamic combinatorial processes. However, it is unlikely that, un-
der usual reaction conditions, the single equivalent of cysteine that
is released will significantly compromise thioester production
through NCL.
In light of these findings it was subsequently unsurprising that a
fivefold reduction in the concentration of our glycopeptide thioes-
ter precursor 3 did not dramatically influence the production of 4,
as determined by HPLC. Notably, 4 appeared to accumulate more
rapidly when the reaction was conducted at lower concentration.
Furthermore, lowering the reaction pH, to inhibit NCL, did not give
rise to an appreciable amount of 4 after 48 h. Instead a complex
mixture of products resulting from extensive deacetylation of both
3 and 4 was obtained.
In summary, our results show that thioester formation via N?S
acyl shift is compatible with native N-glycopeptides at pH 5.8.
Deacetylated sugars should be employed in reactions conducted
at lower pH otherwise partial sugar deacetylation can complicate
analysis. These simply glycosylated thioesters can, through NCL,
be assembled into glycoproteins.6 Furthermore, we investigated
the potential for NCL to occur during thioester formation using a
cysteine exchange reaction and found that NCL can clearly reverse
4. Macmillan, D.; Arham, L. J. Am. Chem. Soc. 2004, 126, 9530.
5. Burlina, F.; Dixson, D. D.; Doyle, R. P.; Chassaing, G.; Boddy, C. N.; Dawson, P.;
Offer, J. Chem. Commun. 2008, 2785.
6. (a) Schwarz, F.; Huang, W.; Li, C.; Schulz, B. L.; Lizak, C.; Palumbo, A.; Numao, S.;
Neri, D.; Aebi, M.; Wang, L.-X. Nat. Chem. Biol. 2010, 6, 264; (b) Wang, L.-X.;
Huang, W. Curr. Opin. Chem. Biol. 2009, 13, 592.
7. (a) Kang, J.; Macmillan, D. Org. Biomol. Chem. 2010, 8, 1993; (b) Kawakami, T.;
Shimizu, S.; Aimoto, S. Bull. Chem. Soc. Jpn. 2010, 83, 570; (c) Nakamura, K. i.;
Sumida, M.; Kawakami, T.; Vorherr, T.; Aimoto, S. Bull. Chem. Soc. Jpn. 2006, 79,
1773; (d) Hojo, H.; Onuma, Y.; Akimoto, Y.; Nakahara, Y.; Nakahara, Y.
Tetrahedron Lett. 2007, 48, 25; (e) Tsuda, S.; Shigenaga, A.; Bando, K.; Otaka, A.
Org. Lett. 2009, 11, 823; (f) Ohta, Y.; Itoh, S.; Shigenaga, A.; Shintaku, S.; Fujii, N.;
Otaka, A. Org. Lett. 2006, 8, 467.
8. (a) Saleh, L.; Perler, F. B. Chem. Rec. 2006, 6, 183; (b) Pellois, J.-P.; Muir, T. W. Curr.
Opin. Chem. Biol. 2006, 10, 487; (c) Mootz, H. D. Chembiochem 2009, 10, 2579.