Rapid formation of N-glycopeptides via Cu(II)-promoted glycosylative ligation

Article abstract of DOI:10.1021/ol302961s

Herein is described the chemoselective Cu(II)-HOBt promoted chemical ligation of glycosylamines and peptide thioacids to give N-glycosylated peptides. The method is distinguished from other chemical approaches to peptide N-glycosylation in that (1) it can be employed in the presence of unprotected N-terminal and Lys side chain amines; (2) it is remarkably fast, going to completion in under 30 min; and (3) it produces glycopeptides without attendant aspartimide formation.

Full text of DOI:10.1021/ol302961s

ORGANIC
LETTERS
2013
Vol. 15, No. 4
732–735
Rapid Formation of N‑Glycopeptides
via Cu(II)-Promoted Glycosylative Ligation
Ryan Joseph, Frank Brock Dyer, and Philip Garner*
Department of Chemistry, Washington State University, Pullman,
Washington 99164-4630, United States
ppg@wsu.edu
Received October 26, 2012
ABSTRACT
Herein is described the chemoselective Cu(II)-HOBt promoted chemical ligation of glycosylamines and peptide thioacids to give N-glycosylated
peptides. The method is distinguished from other chemical approaches to peptide N-glycosylation in that (1) it can be employed in the presence of
unprotected N-terminal and Lys side chain amines; (2) it is remarkably fast, going to completion in under 30 min; and (3) it produces glycopeptides
without attendant aspartimide formation.
Protein N-glycosylation, wherein a glycan is attached
to an asparagine residue, is an important co- and post-
translational modification that influences a diverse array
of biological processes ranging from protein quality con-
trol to cell recognition.1,2 It has been hypothesized that the
epigenetically controlled biosynthesis of glycoproteins has
evolved in higher organisms as a relatively rapid means
to modulate protein function in response to environmental
changes.3 Glycoproteins have also been implicated in
disease states. For example, cancer cells undergo adaptive
regulation of their cell surface glycosylation patterns in
order to acquire a survival advantage.4 Progress toward
elucidating the glycoproteome has been hindered by the
difficulty in obtaining homogeneous glycoproteins for
characterization and biological studies. The challenges of
glycopeptide synthesis using recombinant techniques make
the case for developing chemical synthetic methods to
access homogeneous glycoproteins.5
constructs have generally utilized the native chemical
ligation6 of glycopeptide segments, which are prepared
by incorporating the glycosylated amino acids during a
solid phase peptide synthesis.5,7 Syntheses of N-glycopep-
tides arising from the coupling of a glycosylamine such
as 1 with an aspartic acid residue embedded in a peptide 2
to give the N-glycosylated peptide 3 (Scheme 1) have also
been reported.8À11
Inherent limitations of these protocols necessitate the
masking of free amino groups (N-terminus, Lys side chains)
in the peptide, restricting their application to the middle
and early stages of a projected glycoprotein synthesis.12
(7) Piontek, C.; Ring, P.; Harjes, O.; Heinlein, C.; Mezzato, S.;
€
€
Lombana, N.; Pohner, C.; Puttner, M.; Silva, D. V.; Martin, A.;
Schmidt, F. X.; Unverzagt, C. Angew. Chem., Int. Ed. 2009, 48, 1936.
€
Piontek, C.; Silva, D. V.; Heinlein, C.; Pohner, C.; Mezzato, S.; Ring, P.;
Martin, A.; Schmidt, F. X.; Unverzagt, C. Angew. Chem., Int. Ed. 2009,
48, 1941. (b) Aussedat, B.; Fasching, B.; Johnston, E.; Sane, N.;
Nagorny, P.; Danishefsky, S. J. J. Am. Chem. Soc. 2012, 134, 3532.
(c) Sakamoto, I.; Tezuka, K.; Fukae, K.; Ishii, K.; Tadurur, K.; Maeda,
M.; Ouchi, M.; Yoshida, K.; Nambu, Y.; Igarashi, J.; Hayashi, N.; Tsuji,
T.; Kajihara, Y. J. Am. Chem. Soc. 2012, 134, 5428.
(8) (a) Anisfeld, S. T.; Lansbury, P. T., Jr. J. Org. Chem. 1990, 55,
5560. (b) Cohen-Anisfeld, S. T.; Lansbury, P. T., Jr. J. Am. Chem. Soc.
1993, 115, 10531.
The covalent union of the carbohydrate and peptide
domains of glycoproteins remains a formidable challenge.
Recent synthetic approaches to N-linked glycoprotein
(1) Dwek, R. A. Chem. Rev. 1996, 96, 683.
(2) Gamblin, D. P.; Scanlan, E. M.; Davis, B. G. Chem. Rev. 2009,
109, 131.
(9) Wang, P.; Li, X.; Zhu, J.; Chen, J.; Yuan, Y.; Wu, X.; Danishefsky,
S. J. J. Am. Chem. Soc. 2011, 133, 1597.
(3) (a) Lauc, G.; Zoldos, V. Med. Hypotheses 2009, 73, 510. (b) Lauc,
G.; Zoldos, V. Molecular Biosystems 2010, 6, 2373.
(4) Kannagi, R.; Yin, J.; Miyazaki, K.; Izawa, M. Biochim. Biophys.
Acta 2008, 1780, 525.
(5) Rich, J. R.; Withers, S. G. Nat. Chem. Biol. 2009, 5, 206.
(6) Kent, S. B. H. Chem. Soc. Rev. 2009, 38, 338.
(10) Kaneshiro, C. M.; Michael, K. Angew. Chem., Int. Ed. 2006, 45,
1077.
€
€
(11) (a) Ullmann, V.; Radisch, M.; Boos, I.; Freund, J.; Pohner, C.;
Schwarzinger, S.; Unverzagt, C. Angew. Chem., Int. Ed. 2012, 51, 11566.
(b) Wang, P.; Aussedat, B.; Vohra, Y.; Danishefsky, S. J. Angew. Chem.,
Int. Ed. 2012, 51, 11571.
r
10.1021/ol302961s
Published on Web 01/24/2013
2013 American Chemical Society

An aspartylation procedure that circumvents this che-
moselectivity issue would permit the introduction of the
glycan moiety at a later stage in the synthesis, thus making
it more convergent. Finally, several protecting group
schemes have been developed as potential solutions to
the ubiquitous problem of aspartimide formation.11 While
these represent significant advances, the development of
an aspartylation protocol which bypassed aspartimide
formation altogether could reduce protecting group mani-
pulations and simplify the chemical synthesis of glycopep-
tides. We now report a method that enables the rapid
chemical ligation of a glycosylamine and a specific aspartic
acid residue in an unprotected peptide to give a native
N-linked glycopeptide fragment.
the readilyisolatedproduct 6(Table1), confirmingthatthe
aziridine-mediated ligation conditions did, indeed, result
in the desired amide bond formation (entry 1).17 Reaction
optimization studies (entries 2À6) indicated that the max-
imum yield of 6 was obtained when the thioacid was added
to the Cu(II)ÀHOBt complex followed by a 4-fold excess
of glycosylamine (entry 5). These conditions minimized
the detrimental effect resulting from the known propen-
sity of glycosylamines toward hydrolysis.18 Only margin-
al improvement in yield accompanied a further increase
in the amount of 6 used. This reaction was complete
after 30 min at room temperature. Control experiments
demonstrated the necessity for both Cu(II) and HOBt in
the reaction (entries 7 and 8).19 A known oxidative
protocol9 did not produce 6 after 30 min (entry 9). The
desired reaction could be “rescued” by the inclusion of
Cu(OAc)2, but DMSO was deemed to be inferior to DMF
(entry 10).
Scheme 1
Table 1. Reaction Optimization and Control Experiments
Our approach is derived from our recently reported
Cu(II)-promoted ligation of unprotected peptide thioacids
and aziridine-2-carbonyl peptides.13 This facile coupling
reaction proceeds chemoselectively in the presence of
unprotected primary amines, which we tentatively attri-
bute to the unique properties of the aziridine lone pair
(reduced basicity and reduced steric hindrance) working
in concert with the cupric ion activated thioacid. We
reasoned that a glycosylamine, with an estimated ammo-
nium ion pKa of approximately 6,14 might exhibit similar
behavior with unprotected peptide thioacids, specifically
one derived from an aspartic acid side chain. If this logic
held, these properties would enable the analogous Cu(II)-
promoted N-glycosylative ligation to proceed chemoselec-
tively in the presence of (protonated) free amines.
equiv
equiv
reaction
yield
entry
1
of 4
of 5
conditions
of 6
1.0
1.0
2.0
3.0
4.0
5.0
1.0
1.0
1.0
1.0
1.0
1.0
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMFa
64%c
68%d
75%d
81%d
90%c
92%d
2
3
4
5
6
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMFb
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMFb
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMFb
To test this idea, we initially examined the coupling of
Ac3GlcNAcβ-NH2 (4)15 and Cbz-Gly-SH (5)16 to provide
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMFb
(12) (a) Davis and co-workers reported the coupling of GlcNAcβ-N3
and Fmoc-Ser-Asp(OBt)-Leu-Thr-NH2 using the Staudinger reaction:
Doores, K. J.; Mimura, Y.; Dwek, R. A.; Rudd, P. M.; Elliott, T.; Davis,
B. G. Chem. Commun. 2006, 1401. (b) Damkaci, F.; DeShong, P. J. Am.
Chem. Soc. 2003, 125, 4408.
(13) Dyer, F. B.; Park, C.-M.; Joseph, R.; Garner, P. J. Am. Chem.
Soc. 2011, 133, 20033.
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMFb
7
1.0
1.0
4.0
4.0
1.1
1.1
1.0
1.0
HOBt (2.2),
DMFb
0%e
0%e
(14) Legler, G. Biochim. Biophys. Acta 1978, 524, 94.
8
Cu(OAc)2 (1.0),
aq. DMFb
(15) Glycosylamine 4 was prepared from commercially available
glucosamine hydrochloride (1a. NaOMe/MeOH; 1b. Ac2O 0°C to rt,
92% yield; 2. AcCl, rt, 65% yield; 3. NaN3, DMF, 70°C, 96% yield; 4. H2
Pd/C, EtOAc, rt, 90% yield) using a known method: Cunha, A.; Pereira,
L. c.; de Souza, R.; de Souza, M. C. l.; Ferreira, V. Nucleosides,
Nucleotides, Nucleic Acids 2001, 20, 1555.
(16) Thioacid 5 was prepared from commercially available Cbz-
Gly-OH (1. TrtSH, EDC, DMAP, CH2Cl2, rt, 90% yield; 2. TFA,
Et3SiH, CH2Cl2, rt, 85% yield) using a known method: Crich, D.;
Sharma, I. Angew. Chem., Int. Ed. 2009, 48, 7591.
(17) While our work was already in progress, Gopi and coworkers
reported the facile, Cu(II)-promoted coupling of thioacids and amines in
MeOH. The method was applied to the synthesis of fully protected
peptides. They also showed that this amidation was catalyzed by in situ
generated copper sulfide. Mali, S. M.; Jadhav, S. V.; Gopi, H. N. Chem.
Commun. 2012, 48, 7085–7087.
9
HOBt (2.0),
DMSO
0%d
83%d
10
Cu(OAc)2 (1.0),
HOBt (2.0),
aq. DMSOb
a Order of addition: aq Cu(OAc)2•H2O, HOBt, amine, thioacid.
b Order of addition: aq Cu(OAc)2•H2O, thioacid, amine. c Isolated yield.
d Quantitative HPLC analysis. e TLC analysis.
(18) Isbell, H. S.; Frush, H. J. J. Org. Chem. 1958, 23, 1309.
Org. Lett., Vol. 15, No. 4, 2013
733

Table 2. Chemoselective N-Glycosylation Reactionsa
a The optimized conditions from Table 1, entry 5 were used for these reactions. b 19% of hydrolyzed thioacid was isolated. c 5% of hydrolyzed
thioacid was isolated. d 32% of hydrolyzed thioacid was isolated. e 21% of hydrolyzed thioacid was isolated. f 30% of hydrolyzed thioacid was isolated.
With an optimized glycosylation protocol in hand, we
next examined the chemoselectivity issue employing our
optimized reaction conditions and substrates that did not
incorporate protecting groups (Table 2). The coupling of
GlcNAcβ-NH2 (7)20 and thioacid 5 gave glycosylated
amino acid8 in94% isolatedyield (entry 1), demonstrating
the feasibility of using unprotected glycosylamines in the
reaction. The coupling of peracetylated glycosylamine 4
and the unprotected tripeptide thioacid H-Leu-Asn-Phe-
SH (9)21 provided our first test of peptide amine chemos-
electivity, cleanly producing 10 in 72% isolated yield
(entry 2). Careful HPLC-MS analysis of the reaction mix-
ture showed no evidence of interference by the N-terminal
amine, confirming the chemoselectivity.22 This result was
in line with our aziridine-mediated peptide ligation
studies.13 Building on these two experiments, the coupling
of unprotected glycosylamine 7 and unprotected thioacid 9
afforded the product 11 in 81% isolated yield (entry 3).
Moving on to examples that are directly related to
glycobiology, we next examined the glycosylative ligation
in the context of aspartylation using the heptapeptide
thioacid 12,23 a compound that contains unprotected
amine and alcohol moieties. The coupling of 4 and 12
proceeded smoothly in the presence of both the N-terminal
amine and unprotected Lys side chain to afford the
glycopeptide 13 in 55% isolated yield (entry 4). As ex-
pected, the ligation of 7 and 12 gave the glycopeptide 14
in 65% isolated yield (entry 5). Finally, the reaction of
chitobiosylamine (15)24 and thioacid 12 produced the
glycopeptide 16 in 53% isolated yield (entry 6). This
example demonstrated the compatibility of glycosylative
ligation with an intersaccharide acetal linkage. None of
the aspartimide that often accompanies N-glycosylation
procedures was observed in these ligation reactions.25
The point of glycan attachment to the peptide in 16
(ω-aspartylation) was confirmed by means of a ROESY
NMR experiment. In each case, the mass balance was
accounted for by isolation of the thioacid hydrolysis
byproduct. Products 14 and 16 correspond to truncated
segments of the common R-subunit found in human
(19) The benefit of adding Cu(II) salts together with HOBt or
Cu(OBt)2 to standard peptide coupling reactions in order to suppress
racemization has been known for some time: (a) Miyazawa, T.;
Otomatsu, T.; Fukui, Y.; Yamada, T.; Kuwata, S. Int. J. Peptide Protein
Res. 1992, 39, 308. (b) Gibson, F. S.; Rapoport, H. J. Org. Chem. 1995,
60, 2615. (c) Ryadnov, M. G.; Klimenko, L. V.; Mitin, Y. V. J. Pep-
tide Res. 1999, 53, 322. (d) Van den Nest, W.; Yuval, S.; Albericio, F.
J. Peptide Sci. 2001, 7, 115.
(20) Kiyozumi, M.; Kato, K.; Komori, T.; Yamamoto, A.; Kawasaki,
T.; Tsukamoto, H. Carbohyd. Res. 1970, 14, 355.
(21) Known compound Boc-Phe-SFm was elongated to Fmoc-Leu-
Asn-Phe-SFm with HATU/Boc chemistry and then globally depro-
tected with DBU and purified by HPLC. See: Wu, W.; Zhang, Z.;
Liebeskind, L. S. J. Am. Chem. Soc. 2011, 133, 14256.
(23) Synthesis of peptide thioacid 12: Boc-Val-Gln(Tr)-Lys(Boc)-
Asp(OAll)-Val-Thr(tBu)-Ser(tBu) on 2-chlorotrityl resin was synthe-
sized using standard Fmoc-based solid phase synthesis protocols. Con-
version to 12 was accomplished in solution (1. AcOH/TFE/DCM to
obtain Boc-Val-Gln(Tr)-Lys(Boc)-Asp(OAll)-Val-Thr(tBu)-Ser(tBu)-OH;
2. CH2N2; 3. Pd(PPh3)4/pTolSO2Na; 4. DIC, cat. DMAP, HS-Tmb; 5.
TFA global deprotection).
(24) Hackenberger, C. P. R.; O’Reilly, M. K.; Imperiali, B. J. Org.
Chem. 2005, 70, 3574.
(25) This conclusion was confirmed by comparing with a genuine
(22) Lack of selectivity would lead to peptide coupling products of
the type H-Leu-Asn-Phe-Leu-Asn-Phe-X (X = SH, OH, Ac3GlcNAcβ-
NH), which were not detected in the crude reaction mixture.
sample of the expected aspartimide.
734
Org. Lett., Vol. 15, No. 4, 2013

glycoprotein hormones (e.g., human chorionic gonado-
tropin, hCG).26
shows promise for the convergent synthesis of complex
N-glycopeptides. Extension of this reaction to larger
systems will be the next challenge to face.
The N-glycosylative ligation provides a means for the
chemoselective attachment of a glycosylamine to a specific
aspartic acid residue in an unprotected peptide. Further-
more, this Cu(II)-mediated ligation reaction appears to
proceed significantly faster than aspartylation protocols
that have previously been reported. Finally, the Cu(II)-
promoted glycosylative ligation formed the expected
N-glycopeptides with no detectable amount of aspartimide
side product. Although the mechanism of this reaction
remains to be determined, the critical role played by the
cuprate ion is established. While the N-glycosylative liga-
tion is still at the earliest stages of development, this reaction
Acknowledgment. This research was supported by a
grant from the National Science Foundation (CHE-
1149327). We thank Dr. Greg Helms (WSU Center for
NMR Spectroscopy) for his technical assistance with the
2D NMR experiments on glycopeptide 16.
Supporting Information Available. Experimental pro-
cedures and characterization data for all new compounds
are provided. This material is available free of charge via
the Internet at http://pubs.acs.org.
(26) Pierce, J. G.; Parsons, T. F. Annu. Rev. Biochem. 1981, 50, 465.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 4, 2013
735