Convergent solid-phase synthesis of N-glycopeptides facilitated by pseudoprolines at consensus-sequence Ser/Thr residues

Article abstract of DOI:10.1002/anie.201204272

Remote control: The formation of aspartimides is greatly decreased during peptide elongation and during convergent sugar couplings of Asp-X-Ser/Thr peptides that contain a pseudoproline (red; see scheme). The robust approach efficiently joins complex peptides and N-glycans on the solid phase thus facilitating the availability of glycopeptides and glycoproteins. Copyright

Full text of DOI:10.1002/anie.201204272

Angewandte
Chemie
DOI: 10.1002/anie.201204272
Peptide Synthesis
Convergent Solid-Phase Synthesis of N-Glycopeptides Facilitated by
Pseudoprolines at Consensus-Sequence Ser/Thr Residues**
Vera Ullmann, Marisa Rꢀdisch, Irene Boos, Jutta Freund, Claudia Pçhner, Stefan Schwarzinger,
and Carlo Unverzagt*
Dedicated to Professor Hans Paulsen on the occasion of his 90th birthday
N-Glycosylation is an
important posttransla-
tional modification of
proteins.
A carbohy-
drate is transferred to
an asparagine within an
Asn-X-Ser/Thr consen-
sus sequence.[1] The
study of the biological
aspects of N-glycosyla-
tion often requires the
synthesis of N-glyco-
peptides,[2] which are
accessible by two main
approaches. In the
sequential mode glyco-
sylamino acid cassettes
are used for peptide
Scheme 1. Lansbury aspartylation leads to glycopeptide 3, Asn peptide 3Asn and aspartimide side product 4ai.
PG=protecting group.
elongation. After incorporation of larger oligosaccharides
the solubility and reactivity of the peptide is affected and side
reactions, for example, involving free OH groups[3] complicate
further elongation. In the convergent mode (Lansbury
aspartylation)[4] the sugar is connected to an aspartate after
complete assembly of the peptide (Scheme 1). The main
drawback of the convergent mode is the formation of cyclic
aspartimides during peptide elongation and sugar coupling;[5]
this formation depends on the peptide sequence,[6] and the
syntheses of longer glycopeptide thioesters are particularly
difficult as they require, for example, additional ligation
steps[9f,10] or segment couplings.[3b] For longer N-glycopeptides
the convergent approach provides advantages over the
sequential approach,[11] especially as the Lansbury aspartyla-
tion[4] has been shown to be efficient also on the solid phase.[12]
The undesired aspartimide formation can be avoided by
peptide backbone (NH) protection,[13] however, this approach
is tedious for residues other than glycine, and causes
racemization when coupling backbone-protected dipepti-
des.[12b] Aspartimide formation during peptide elongation
can be reduced by using bulky groups to protect the Asp side
chain,[14] for example, the 2-phenylisopropylester (PhiPr);[12a]
however, trityl anchors are also cleaved under the reaction
conditions for PhiPr removal. Dmab-protected[15] Asp resi-
dues are compatible with trityl anchors, but the backbone
protection of the neighboring amino acid is required.[13a,16]
To provide complex N-glycopeptide thioesters by solid-
phase Lansbury aspartylation we compared three Asp-side-
chain protecting groups for the Fmoc-SPPS of an interleukin-
6 (IL-6) 43–48 hexapeptide. The allyl-protected peptide[17] 6a
showed the highest percentage of aspartimide 7ai formation
(15%), followed by the Dmab peptide 6b (8%), and the
PhiPr[12a] peptide 6c (< 1%; Scheme 2). As well as the
protecting group many factors are known to contribute to
aspartimide formation[5] for example, the steric bulk of the
neighboring amino acid, the basicity of the reaction media,
and also the overall conformation[18] of the peptide C-terminal
relative to the Asp moiety. We reasoned that constraining the
coupling conditions.[7] We found that
a pseudoproline
(Ypro)[8] at the consensus-sequence Ser/Thr residue (Asn-
X-Ser/Thr(Ypro)) efficiently suppresses the formation of
aspartimides in the convergent synthesis of N-glycopeptides
on the solid phase.
The lack of pure N-glycoproteins for biological studies has
stimulated research into their synthesis; these syntheses were
carried out mainly by ligation techniques.[9] The required
glycopeptides and their thioesters are difficult to obtain as the
sugar component interferes with the peptide synthesis. The
[*] V. Ullmann, M. Rꢀdisch, I. Boos, J. Freund, Dr. C. Pçhner,
Dr. S. Schwarzinger, Prof. C. Unverzagt
Bioorganische Chemie, Gebꢀude NW1, Universitꢀt Bayreuth
95440 Bayreuth (Germany)
E-mail: carlo.unverzagt@uni-bayreuth.de
[**] This work was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201204272.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Scheme 2. a) Fmoc-SPPS of IL-6 43-48Nle 6a–c with different Asp protection groups, b) IL-6 43-48Nle with Ser(Ypro); c, d) synthesis of ADGSTG
peptides with Ser(OtBu) (12) or Ser(Ypro) (15). All=allyl, Boc=tert-butyloxycarbonyl, 2-Cl-Trt-PS=2-chlorotrityl polystyrene, Dmab=4-(N-[1-(4,4-
dimethyl-2,6-dioxocyclohexylidene)-3-methyl-butyl]-amino)benzyl, Fmoc=9-fluorenylmethoxycarbonyl, HFIP=hexafluoroisopropanol, Nle=norleu-
cine, SPPS=solid phase peptide synthesis, Trt=trityl.
flexibility of this peptide part might influence aspartimide
formation. This hypothesis was investigated by converting the
trans conformation of the Lys-Ser amide bond into a stable cis
conformation by using a Ser(YMe,Mepro) pseudoproline di-
peptide.[8] Pleasingly, in the synthesis of peptide 9 the
presence of the pseudoproline almost completely eliminated
the succinimide formation (product 10ai) despite the pres-
ence of the susceptible allyl ester (Scheme 2b).[17] This
hitherto unnoticed pseudoproline-based conformational
effect was tested by investigating the synthesis of peptides
12 and 15, in which a glycine, which is the residue that
promotes aspartimide formation the most,[6,19] is next to the
Asp allylester. In the case of Ser(OtBu) peptide 12 asparti-
mide formation was extensive (68% 13ai; Scheme 2c),
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
whereas the Ser(Ypro) peptide 15 gave rise to only 5% of
cyclized side product 16ai (Scheme 2d). Aspartimide forma-
tion mainly occurred during Fmoc removal. The purified Asp-
Gly-containing peptides 12 and 15 were stable under simu-
lated coupling conditions (DIPEA, 3 equiv), whereas simu-
lated cleavage conditions (20% piperidine) gave asparti-
mides. Thus, mild conditions including a hexamethylene-
imine/N-methylpyrrolidine/Cl-HOBt mixture[20] were tested
for the deprotection of elongated Asp-Gly-containing pep-
tides, and only 7% of aspartimides 35ai and 38ai were found
(Figure S3).
The allyl protecting groups of Asp-Gly-containing peptide
resins 11 and 14 were selectively cleaved[21] in the presence of
DIPEA, which prevented acidic cleavage of the peptide from
the linker. Solid-phase Lansbury aspartylation was carried out
by activating the peptides with PyBOP in the presence of
GlcNAc-NH2 2 (Scheme 3). The ratio of glycopeptide 17 to
aspartimide 13ai was 23:77, thus indicating the formation of
additional aspartimide during aspartylation (Scheme 3 a). For
the pseudoproline peptide 14 efficient conversion into
glycopeptide 18 (84%) took place accompanied by formation
of aspartimide 16ai (12%) and only 4% of the Asn peptide
18Asn (Scheme 3 b). Glycopeptide 18 was obtained in 68%
yield after HPLC. The allyl protecting group was removed
from Asp-Lys-containing peptide 8 and then it was coupled to
2 to give glycopeptide 19 in 62% yield upon isolation
(Scheme 3 c).
The preparative scope of the pseudoproline-assisted
convergent N-glycopeptide formation was probed with the
synthesis of a ribonuclease (RNase) fragment. The RNase 27–
39 peptide containing a Thr(Ypro) was synthesized on 2-Cl-
Trt-PS resin, the allyl group was cleaved at Asp 34, and then it
was coupled with 2. Glycopeptide 20 and the Asn side product
20Asn were obtained in a 83:17 ratio with no detectable
aspartimide (Scheme 3 d). Cleavage from the resin and
HPLC purification gave 20 (59%). The GlcNAc-b-Asn
linkage was confirmed by NMR spectroscopy for glycopep-
tides 18, 19, and 20 (J1,2 ꢀ 9 Hz; see the Supporting Informa-
tion Scheme S6).
The advantages of pseudoprolines in the convergent
assembly of peptides and glycopeptides were combined in
the case of difficult RNase 1–39 glycopeptides (Scheme 4 a).
The RNase 17–39 peptide 21 was synthesized on a 2-Cl-trityl-
CM resin and coupled with the RNase 1–16 fragment 22.
Racemization during segment condensation on the solid
support is avoided owing to the presence of the C-terminal
Scheme 3. Solid phase coupling of GlcNAc-NH2 2 to peptides. DIPEA=diisopropylethylamine, DMF=N,N-dimethylformamide, GlcNAc=N-
acetylglucosamine, PyBOP=(benzotriazol-1-yloxy)tripyrrolidinophosphonium hexafluorophosphate.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Scheme 4. Convergent synthesis of glycoprotein segments: a) RNase 1–39, b) EPO 1–28). DIC=diisopropylcarbodiimide, DMSO=dimethylsulf-
oxide, Cl-HOBt=6-chloro-1-hydroxybenzotriazole, 2-Cl-Trt-CM=2-chlorotrityl ChemMatrix, HATU=2-(7-Aza-1H-benzotriazole-1-yl)-1,1,3,3-tetra-
methyluronium hexafluorophosphate, HOAt=1-hydroxy-7-azabenzotriazole.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
pseudoproline of 22.[3b] Removal of the allyl groups of the
RNase 1–39 peptide gave 23, which was coupled with 2 and
converted into thioester 24 by the in situ thioesterification
method.[22] After HPLC, thioester 24 was obtained in 20%
yield, which is the same yield as for a previous approach using
GlcNAc-Asn.[3b] Coupling of the complex nonasaccharide
amine 25 gave RNase glycopeptide 26 in 14% yield despite
considerable formation of the asparagine side product. The
nonasaccharide amine 25[23] was obtained from a sialoglyco-
peptide[23b] by enzymatic cleavage followed by anomeric
azidation,[24] and reduction (Supporting Information
Scheme S7).
[3] a) N. Yamamoto, Y. Tanabe, R. Okamoto, P. E. Dawson, Y.
Kajihara, J. Am. Chem. Soc. 2008, 130, 501; b) C. Heinlein, D.
Varon Silva, A. Trçster, J. Schmidt, A. Gross, C. Unverzagt,
Angew. Chem. 2011, 123, 6530; Angew. Chem. Int. Ed. 2011, 50,
6406.
[4] S. T. Cohen-Anisfeld, P. T. Lansbury, J. Am. Chem. Soc. 1993,
115, 10531.
[5] R. Subirꢀs-Funosas, A. El-Faham, F. Albericio, Tetrahedron
2011, 67, 8595.
[6] M. Bodanszky, J. Z. Kwei, Int. J. Pept. Protein. Res. 1978, 12, 69.
[7] J. Martinez, M. Bodanszky, Int. J. Pept. Protein. Res. 1978, 12,
277.
[8] P. Dumy, M. Keller, D. E. Ryan, B. Rohwedder, T. Woehr, M.
Mutter, J. Am. Chem. Soc. 1997, 119, 918.
The pseudoproline-assisted Lansbury aspartylation was
then applied to the erythropoietin (EPO) 1–28 sequence
(Scheme 4 b). Resin-bound EPO 1–28 27 was synthesized in
satisfactory purity by linear Fmoc-SPPS on the polar 2-Cl-Trt-
CM resin, and no accompanying cyclization of the Asp 24
allylester ocurred despite the numerous coupling and depro-
tection steps. Using PyBOP 27 was coupled to 2, then the
glycopeptide was cleaved, thioesterification was carried out
followed by deprotection. HPLC analysis showed high
conversion into the glycopeptide thioester (78%) accompa-
nied by some asparagine formation (22%) and no detectable
aspartimide. Purification by HPLC gave the EPO 1–28
GlcNAc thioester 28 in a yield of 37%. When the EPO 1–
28 peptide 27 was reacted with nonasaccharide amine 25,
there was extensive asparagine formation. Thus, a series of
coupling reagents was screened for 2 and the best reaction
conditions were applied to 25 (Supporting Information,
Table 2). In situ thioesterification of the resulting nonasac-
charide glycopeptide was unexpectedly difficult because the
glycopeptide acid was retained on the resin after mild acidic
cleavage. An optimized thioesterification procedure, how-
ever, gave the desired glycopeptide thioester 29 in 24% yield
after HPLC.
[9] a) R. J. Payne, C. H. Wong, Chem. Commun. 2010, 46, 21;
b) D. P. Gamblin, E. M. Scanlan, B. G. Davis, Chem. Rev. 2009,
109, 131; c) M. Murakami, R. Okamoto, M. Izumi, Y. Kajihara,
Angew. Chem. 2012, 124, 3627; Angew. Chem. Int. Ed. 2012, 51,
3567; d) C. Piontek, D. Varꢀn Silva, C. Heinlein, C. Pçhner, S.
Mezzato, P. Ring, A. Martin, F. X. Schmid, C. Unverzagt, Angew.
Chem. 2009, 121, 1974; Angew. Chem. Int. Ed. 2009, 48, 1941;
e) P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. Kent, Science
1994, 266, 776; f) P. Nagorny, N. Sane, B. Fasching, B. Aussedat,
S. J. Danishefsky, Angew. Chem. 2012, 124, 999; Angew. Chem.
Int. Ed. 2012, 51, 975; g) C. P. Hackenberger, D. Schwarzer,
Angew. Chem. 2008, 120, 10182; Angew. Chem. Int. Ed. 2008, 47,
10030.
[10] a) Y. Yuan, J. Chen, Q. Wan, Z. Tan, G. Chen, C. Kan, S. J.
Danishefsky, J. Am. Chem. Soc. 2009, 131, 5432; b) R. Okamoto,
K. Morooka, Y. Kajihara, Angew. Chem. 2012, 124, 195; Angew.
Chem. Int. Ed. 2012, 51, 191.
[11] P. Wang, X. Li, J. Zhu, J. Chen, Y. Yuan, X. Wu, S. J. Danishefsky,
J. Am. Chem. Soc. 2011, 133, 1597.
[12] a) R. Chen, T. J. Tolbert, J. Am. Chem. Soc. 2010, 132, 3211; b) T.
Conroy, K. A. Jolliffe, R. J. Payne, Org. Biomol. Chem. 2010, 8,
3723; c) D. Vetter, D. Tumelty, S. K. Singh, M. A. Gallop, Angew.
Chem. 1995, 107, 94; Angew. Chem. Int. Ed. Engl. 1995, 34, 60.
´
[13] a) J. Ruczynski, B. Lewandowska, P. Mucha, P. Rekowski, J.
Pept. Sci. 2008, 14, 335; b) M. Quibell, D. Owen, L. C. Peckman,
T. Johnson, J. Chem. Soc. Chem. Commun. 1994, 2343.
[14] A. Karlstrçm, A. Unden, Tetrahedron Lett. 1996, 37, 4243.
[15] W. C. Chan, B. W. Bycroft, D. J. Evans, P. D. White, J. Chem. Soc.
Chem. Commun. 1995, 2209.
[16] Z. Tan, S. Shang, T. Halkina, Y. Yuan, S. J. Danishefsky, J. Am.
Chem. Soc. 2009, 131, 5424.
[17] H. Kunz, H. Waldmann, C. Unverzagt, Int. J. Pept. Protein Res.
1985, 26, 493.
We have found a general method for improved conver-
gent N-glycopeptide synthesis on the solid phase. By con-
verting the consensus-sequence Ser/Thr moieties into a pseu-
doproline the formation of aspartimides is highly reduced,
both in peptide elongation and in the subsequent aspartyla-
tion. The robust approach uses standard building blocks
compatible with automated peptide synthesis, thus facilitating
the availability of glycopeptide and glycoprotein libraries.
[18] R. Doelling, M. Beyermann, J. Haenel, F. Kernchen, E. Krause,
P. Franke, M. Brudel, M. Bienert, J. Chem. Soc. Chem. Commun.
1994, 853.
[19] E. Nicolꢁs, E. Pedroso, E. Giralt, Tetrahedron Lett. 1989, 30, 497.
[20] X. Li, T. Kawakami, S. Aimoto, Tetrahedron Lett. 1998, 39, 8669.
[21] M. Dessolin, M.-G. Guillerez, N. Thieriet, F. Guibe, A. Loffet,
Tetrahedron Lett. 1995, 36, 5741.
Received: June 1, 2012
Published online: && &&, &&&&
Keywords: aspartylation · glycopeptides · protecting groups ·
[22] S. Flemer, Jr., J. Pept. Sci. 2009, 15, 693.
.
pseudoproline · solid-phase synthesis
[23] a) N. Yamamoto, T. Sakakibara, Y. Kajihara, Tetrahedron Lett.
2004, 45, 3287; b) A. Seko, M. Koketsu, M. Nishizono, Y. Enoki,
H. R. Ibrahim, L. R. Juneja, M. Kim, T. Yamamoto, Biochim.
Biophys. Acta Gen. Subj. 1997, 1335, 23 – 32.
[24] T. Tanaka, H. Nagai, M. Noguchi, A. Kobayashi, S. Shoda, Chem.
Commun. 2009, 3378.
[1] R. Kornfeld, S. Kornfeld, Annu. Rev. Biochem. 1985, 54, 631.
[2] a) H. Herzner, T. Reipen, M. Schultz, H. Kunz, Chem. Rev. 2000,
100, 4495; b) O. Seitz, ChemBioChem 2000, 1, 214; c) T. Buskas,
S. Ingale, G. J. Boons, Glycobiology 2006, 16, 113R; d) M.
Meldal, K. Bock, Glycoconjugate J. 1994, 11, 59.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Peptide Synthesis
V. Ullmann, M. Rꢁdisch, I. Boos,
J. Freund, C. Pçhner, S. Schwarzinger,
C. Unverzagt*
&&&& — &&&&
Convergent Solid-Phase Synthesis of N-
Glycopeptides Facilitated by
Pseudoprolines at Consensus-Sequence
Ser/Thr Residues
Remote control: The formation of aspar-
timides is highly reduced during peptide
elongation and convergent sugar cou-
plings of Asp-X-Ser/Thr peptides that
contain a pseudoproline (red; see
scheme). The robust approach efficiently
joins complex peptides and N-glycans on
the solid phase thus facilitating the
availability of glycopeptides and glyco-
proteins.
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!