Solid-Phase O-Glycosylation with a Glucosamine Derivative for the Synthesis of a Glycopeptide

Article abstract of DOI:10.1071/CH17201

An efficient synthesis of the O-linked glycosylamino acid Fmoc-l-Ser((Ac)₃-β-d-GlcNAc)-OH building block is described. The utility of the method was demonstrated with direct solid-phase O-glycosylation of the hydroxyl group on the amino acid (Ser) side chain of a human A crystallin-derived peptide (AIPVSREEK) in nearly quantitative glycosylation yield.

Full text of DOI:10.1071/CH17201

CSIRO PUBLISHING
Aust. J. Chem. 2017, 70, 1151–1157
Full Paper
https://doi.org/10.1071/CH17201
Solid-Phase O-Glycosylation with a Glucosamine
Derivative for the Synthesis of a Glycopeptide
A B C
, ,
D
D
Philip Ryan,
Andy Hsien Wei Koh, Anna Elizabeth Lohning,
A B C E F
, , , ,
and Santosh Rudrawar
A
Menzies Health Institute Queensland, Griffith University, Gold Coast,
Qld 4222, Australia.
B
School of Pharmacy and Pharmacology, Griffith University, Gold Coast,
Qld 4222, Australia.
C
School of Chemistry, The University of Sydney, Sydney, NSW 2006, Australia.
D
Faculty of Health Sciences and Medicine, Bond University, Robina,
Qld 4229, Australia.
E
Quality Use of Medicines Network, Gold Coast, Qld 4222, Australia.
Corresponding author. Email: s.rudrawar@griffith.edu.au
F
An efficient synthesis of the O-linked glycosylamino acid Fmoc–L-Ser((Ac)3–b-D-GlcNAc)-OH building block is
described. The utility of the method was demonstrated with direct solid-phase O-glycosylation of the hydroxyl group
on the amino acid (Ser) side chain of a human a-A crystallin-derived peptide (AIPVSREEK) in nearly quantitative
glycosylation yield.
Manuscript received: 10 April 2017.
Manuscript accepted: 16 June 2017.
Published online: 20 July 2017.
Introduction
linked (‘mucin-type’ a-linked GalNAc residue) oligosaccharides
and cell surface and secreted proteins.[14]
O-GlcNAcylation, the attachment of O-linked b-N-acetyl-
glucosamine (O-GlcNAc) to serines and threonines of cyto-
plasmic, nuclear, and mitochondrial proteins is an ubiquitous
and essential post-translational modification of multicellular
organisms (Fig. 1).[1] The O-GlcNAc modification is governed
by the essential enzyme O-linked b-N-acetylglucosamine
transferase (O-GlcNAc transferase, OGT).[2] To date, more than
4000 proteins[3] have been identified as OGT substrates,
including numerous transcription factors,[4] tumour suppressors,
kinases,[5] phosphatases,[6] and histone-modifying proteins.[7]
OGT-catalysed O-GlcNAcylation has been found to play key
roles in many fundamental biological processes including epi-
genetic regulation, transcription, stress response, and homeo-
stasis.[8] Therefore, it is not surprising that dysregulation of
protein O-GlcNAcylation underlies the pathological progres-
sion of several chronic metabolic diseases including diabe-
tes,[9,10] cancer,[11] and neurodegenerative disorders.[12]
O-GlcNAcylation exerts diverse actions in a site-specific
manner, with glycosylation of many proteins conferring signifi-
cant functions including alterations in protein stability and
enzymatic activity, translocation, and regulation of gene expres-
sion.[3] O-GlcNAcylation can occur on multiple Ser and Thr
residues of proteins and modification on different sites often has
distinct functional consequences. Therefore, the determination
of these locations is a prerequisite to study the structural and
functional significance of the glycan. To date, only a handful of
O-GlcNAc modification sites have been mapped; hence, chem-
ical synthesis of O-linked GlcNAc glycopeptides offers an
important tool to decipher the roles of this modification in
chronic disease conditions such as diabetes and cancer.
Results and Discussion
O-GlcNAcylation (addition of b-linked N-acetylglucosamine)
is in many ways distinct from ‘classical’ protein glycosylation.[6]
O-GlcNAcylation is: (i) reversible, highly dynamic, and mediated
exclusively bya singleglycosyltransferase– OGT;(ii) an addition
of the GlcNAc monosaccharide residue and not any elongated
complex sugar structure, and (iii) interplayed with phosphoryla-
tion as most of the known serine (Ser) and threonine (Thr) sites for
O-GlcNAcylation are also sites for phosphorylation or adjacent to
phosphorylation sites.[13] Notably, the classical glycosylation
mainly occurs between Asn-linked (N-linked) or Ser- or Thr-
Traditionally, glycopeptides are synthesised by incorporating
preformed glycosylated amino acid building blocks (b-GlcNAc-
Ser/Thr and a-GlcNAc-Ser/Thr) into standard solid-phase pep-
tide synthesis (SPPS) techniques.[15] Two factors generally limit
this approach. First, the synthesis of glycosyl-amino acids
involves protection and deprotection of the carboxyl groups of
amino acids and tedious purification processes.[16] Second, most
building blocks are extremely expensive to purchase in the large
quantities required for SPPS. A large excess of glycosyl-amino
acids is required to increase the rate and drive the reaction to
Journal compilation Ó CSIRO 2017
www.publish.csiro.au/journals/ajc

1152
P. Ryan et al.
OH
β
O
HO
HO
H (CH3)
O
HO
(CH )
H
O
3
AcHN
O
O
H
N
H
N
UDP
UDP-GlcNAc
N
H
N
H
OGT
O
Ser or Thr
Ser or Thr
Naked protein
Glycoprotein
Fig. 1. OGT-catalysed O-glycosylation.
completion (Fmoc–L-Ser((Ac)3–b-D-GlcNAc)–OH costs US
$1185 per 100 mg, a quantity sufficient for the incorporation of
just one O-GlcNAcylated residue in a single peptide synthesised
at small scale). Additionally, glycosylated amino acids are prone
to epimerisation during SPPS.[17] For rapid synthesis of a library
of glycopeptides with variations in amino acid residues in the
peptide or synthesis of peptides with multiple glycosylated resi-
dues, it is more efficient to stereoselectively couple a sugar res-
idue directly to the free hydroxyl groups of the amino acid side
chains on the resin-bound peptides. Convergent syntheses of
O-linked glycopeptides by attachment of glycan fragments to
peptides are constrained by the poor solubility of peptides under
conditions used for glycoside synthesis, and by poor stereo-
selectivity of glycosylation reactions. Direct solid-phase glyco-
sylation of the hydroxyl groups of Ser, Thr or tyrosine (Tyr)
residues of peptides on solid support with a range of glycosyl
donors (monosaccharide: Gal, Glu, Man, Fuc, Lac, Mal; disac-
charide: Galb1, 3GalN3b) have been reported; however, glyco-
sylation with glucosamine is still challenging.[18–21]
The common approach in the installation of the b-glycosidic
linkage employs a glycosyl donor with a leaving group at the
anomeric centre, activated with a promoter to yield an oxocarbe-
nium ion intermediate, which in turn is susceptible to nucleophilic
attackby a glycosyl acceptor.[22,23] The O-acetylglycosyl donor of
N-acetylglucosaminepreviouslyusedextensivelyinglycosylation
reactions required stoichiometric amounts of acidic activators or
promoters,[24] or a catalytic promoter under microwave condi-
tion.[21,25,26] Seibel et al. reported the microwave-assisted FeCl3-
promoted glycosylation of Fmoc–Ser–OBn (OBn ¼ O-benzyl)
with peracetylatedmono- and disaccharides. Theper-O-acetylated
galactose, glucose, maltose and lactose were converted to corre-
sponding b-glycosides in moderate yields (,61 %). However,
glycosylation reaction conditions failed to form b-O-GlcNAc–
Ser–OBn from 2-acetamido-2-deoxy-1,3,4,6-tetra-O-acetyl-b-D-
glucopyranose, and only the oxazoline intermediate was isolated
asthe major product(see SupplementaryMaterial, Table S1, Entry
2).[21] Arsequell et al. prepared b-O-GlcNAc–Ser using an excess
(3 equiv.) of boron trifluoride etherate as a Lewis acid promoter.
The reaction was thought to proceed through the oxazoline
intermediate (see Supplementary Material, Table S1, Entry
3).[25] Carvalho et al. previously reported mercuric bromide-
promoted traditional Koenigs–Knorr glycosylation of Fmoc–
Ser–OBn with 2-acetamido-2-deoxy-3,4,6-tri-O-acetyl-a-D-
glucopyranosyl chloride, offering the corresponding b-glycoside
in 64 % yield (see Supplementary Material, Table S1, Entry 4).[27]
Recently, Brimble et al. achieved HgBr2-mediated glycosylation
of protected serine (Fmoc–Ser–O-tBu) in 35 % yield (see Supple-
mentaryMaterial,TableS1, Entry5).[28] Takasuetal. reportedthat
silver triflate (AgOTf) promoted glycosylation of Cbz–Ser–OBn
(Cbz ¼ carboxybenzyl) with 2-acetamido-2-deoxy-3,4,6-tri-O-
acetyl-a-D-glucopyranosyl chloride, furnishing the b-glycoside
in 37 % yield (see Supplementary Material, Table S1, Entry 6).[29]
In most cases, however, these reactions did not exclusively afford
the b-anomer. Glycoside bond formation with donors derived
from 2-acetamido-2-deoxy sugars (O-acetylglycosyl donor or
glycosyl chloride donor) normally proceeds via an oxazoline
intermediate to give a- and b-linked O-glycans. An alternative
strategyfor theinductionofcompleteb-selectivityinthe chemical
synthesis of the b-glycoside of N-acetyl-D-glucosamine employs a
glycosyl donor boasting a participating group on the C-2 nitrogen.
N-Phthaloyl (N-Phth)-protected donors usually give the desired
b-glycoside in high yield[23] and stereoselectivity; however, it is
difficult to maintain this high yield owing to partial product
decomposition under subsequent basic cleavage conditions. The
alternative 2,2,2-trichloroethoxycarbonylamino (N-Troc) group-
containing donor N-Troc-glucosamine[22] gave b-glycoside in
higher yields than the N-Phth-protected counterparts.[23] Addi-
tionally, the Troc group can be removed in high yield under mild
conditions.[21] Salvador et al. reported treatment of an N-Troc-
protected O-acetylglycosyl donor with Fmoc-protected serine
under the influence of an excess (3 equiv.) of boron trifluoride
etherate gave amino acid b-glycoside in 70 % yield (see Supple-
mentary Material, Table S1, Entry 7).[22,23,30] Recently, Norgren
et al. described difficulty in reproducing this reaction and isolated
46 % b-glycoside (see Supplementary Material, Table S1, Entry
7).[31] Mitchell et al. reported glycosylation of a Schiff-base
benzhydryl ester serine derivative withN-Troc-protected glycosyl
bromide (2-deoxy-2-trichloroethoxycarbonylamide-a-bromo-D-
Glc(OAc)3). The glycosylation reaction was promoted by AgOTf
and the desired glycoside was isolated in 72 % yield with a 15 :1
b : a ratio (see Supplementary Material, Table S1, Entry 8).[32]
Reipen et al. reported glycosylation of Fmoc–Ser(OH)–O-tBu
with b-thioglycoside. The reaction required activation with
N-iodosuccinimide (NIS) and trifluoromethanesulfonic acid
(TfOH) to furnish the desired b-glycoside in 80 % yield (see
Supplementary Material, Table S1, Entry 9).[33] Ro¨hrig et al.
reported glycosylation of a trichloroacetimidate-activated glyco-
syl donor with Cbz–Ser–OBn with trimethylsilyl trifluorometha-
nesulfonate (TMSOTf) catalysis gave the glycosylamino acid in
90 % yield (see Supplementary Material, Table S1, Entry 10).[34]
Alternatively, the N-allyloxycarbonyl (Alloc)-glucosamine
derivative has stereoselectively glycosylated Fmoc–Ser–OPfp
(OPfp ¼ pentafluorophenyl). Vargas-Berenguel et al. investigated
anAgOTf-mediatedglycosylation reactionwithglycosyl bromide
and trichloroacetimidate donors and isolated b-glycoside in 37
and 63 % yield respectively (see Supplementary Material, Table
S1, Entry 11).[35] Garg et al. reported glycosylation of Cbz–Ser–
OMe with a glycosyl bromide containing a non-participating
group at C-2 and were able to isolate the b-glycoside in 30 %
yield (see Supplementary Material, Table S1, Entry 12).[36]
We were interested in the efficient b-glycosylation of Fmoc–
Ser(OH)–O-tBu, which could then be applied in the direct

Glycopeptide Synthesis on the Solid Support
1153
OH
O
OAc
O
OAc
O
OAc
O
HO
HO
AcO
AcO
AcO
AcO
d
AcO
AcO
a, b
c
OH
OH
OAc
NH2
TrocHN
TrocHN
TrocHN
O
NH
CCl3
HCl
3
4
2
1
HO
OtBu
e
FmocHN
O
5
OAc
O
OAc
O
OAc
O
AcO
AcO
AcO
AcO
AcO
AcO
O
O
f
O
g
AcHN
TrocHN
AcHN
OtBu
OtBu
OH
FmocHN
FmocHN
FmocHN
O
O
6
7
O
8
Scheme 1. (a) TrocCl, K2CO3, H2O; (b) Ac2O, pyridine, 61 % over 2 steps; (c) ethylenediamine, AcOH, THF, 79 %; (d) CCl3CN, DBU, DCM,
60 %; (e) Fmoc–Ser–O-tBu, TMSOTf, DCM, 97 %; (f) Zn, Ac2O, 78 %; (g) TFA, DCM (1 : 1).
solid-phase glycosylation of peptides for synthesis of glycopep-
tides. Following consideration of these various glycosyl donors,
weaimed to establish a new reliableprocedure for O-glycosylation
and, as such, were led to revisit the stereoselective formation
of Fmoc–L-Ser((Ac)3–b-D-GlcNAc)–O-tBu using the readily
(10–15 %). Increasing the concentration of TMSOTf resulted in
more acidic conditions and, as a result, a polar side product was
also isolated, which was assigned as glycosylated serine, with a
loss of its tert-butyl ester, 9 (,30–35 %).[23] Previously, Halkes
et al. observed that a higher concentration (0.1 equivalents) of
TMSOTf was incompatible with the use of the tert-butyl
protecting group in the peptide template.[40]
available
3,4,6-tri-O-acetyl-2-deoxy-2-(2,2,2-trichloroethoxy-
carbonylamino)-a-D-glucopyranosyltrichloroacetimidate.Herein,
we report improvement in the b-glycosylation of trichloro-
acetimidate-activated N-Troc-protected glucosamine with an
appropriately protected serine building block for application in
solid-phase O-glycosylation. Synthesis of the known N-Troc-
protected glycosyl donor was carried out from D-glucosamine
hydrochloride (1) according to a reported method[22] (Scheme 1).
Briefly, D-glucosamine hydrochloride (1) was treated with
2,2,2-trichloroethoxycarbonyl chloride (TrocCl) in aqueous
K2CO3 to yield the N-Troc-protected intermediate, and subse-
quent acetylation gave N-Troc-1,3,4,6-tetra-O-acetyl glucos-
amine (2) in 61 % yield after two steps. Selective removal of
the anomeric acetyl group was performed by using ethylenedia-
mine/acetic acid buffer, affording compound 3 in 79 % yield.
Treatment of a,b-1-O-unprotected glucosamine derivative 3
with trichloroacetonitrile in the presence of 1,8-diazabicyclo
[5.4.0]undec-7-ene (DBU) as a base afforded the desired glyco-
syl trichloroacetamidate 4 in 60 % yield. The glycosyl acceptor,
Fmoc–Ser(OH)–O-tBu (5), was synthesised from commercially
available Fmoc–Ser–OH in one step.[28,37,38] With glycosyl
donor 4 and the glycosyl acceptor 5 in hand, we then focussed
on the glycosylation reaction.
The glycosylation of Fmoc–Ser(OH)–O-tBu (5) with 4 was
attempted under various conditions derived from changes to the
donor : acceptor ratio, reaction temperature, time, and promoter
TMSOTf quantity. Initial attempts following a literature proce-
dure[39] with 0.9 equivalents of serine derivative 5 and
0.02 equivalents of TMSOTf resulted in poor yields (35 %) of
the desired b-glycoside 6. Unreacted glycosyl donor 4 (10 %)
and a side product with loss of trichloroacetimidate 3 were also
isolated (,30 %) (Table 1, Entry 1). A slight improvement in
yield was obtained (41 %) when the reaction was carried out
using 0.1 equivalents of promoter TMSOTf (Table 1, Entry 2).
The reaction yield (45 %) was not significantly improved even
when conducting the reaction using 1 equivalent of glycosyl
acceptor 5 (Table 1, Entry 3). Although the glycosyl donor was
mostly consumed, hydrolysis of the trichloroacetimidate had
also occurred as evidenced by the presence of side product 3
In order to avoid hydrolysis of trichloroacetimidate and tert-
butyl ester, the reaction was carried out in the presence of
˚
activated 4-A molecular sieves at low temperature (ꢀ788C to
08C; 2 h). This change in reaction conditions resulted in only a
slight improvement in yield, with the desired glycoside 6
isolated in moderate yield (50 %). Though the reaction pro-
ceeded without hydrolysis of trichloroacetimidate, ,35 % gly-
cosylated acid derivative 9 was again isolated owing to
hydrolysis of the tert-butyl ester (Table 1, Entry 4).[23] When
the reaction was repeated using 0.02 equivalents of TMSOTf
in the presence of activated 4-A molecular sieves,[39] there was
˚
an improvement in product yield, with the b-glycoside being
isolated in 70 % yield with only a trace amount (,5 %) of acid
derivative observed (Table 1, Entry 5). Gratifyingly, glycosyla-
tion was successfully achieved (b-glycoside, 98 % yield) on
reacting glycosyl donor with acceptor Fmoc–Ser(OH)–O-tBu
(5) (1.5 equiv), and promoter TMSOTf (0.02 equiv.) in the
˚
presence of activated 4-A molecular sieves at low temperature
(ꢀ788C to 08C; 2h) (Table 1, Entry 6).
Herein, we report the first use of a glycosyl trichloroaceta-
midate, 4, in the stereoselective glycosylation of Fmoc–Ser
(OH)–O-tBu (5). The tert-butyl ester protecting group was
sufficiently stable under the glycosylation conditions. It is
important to note that the tert-butyl ester protecting group is
commonly employed for glutamic acid (Glu) side chain protec-
tion and tert-butyl ether groups are routinely used for Ser or Thr
side chain protection in SPPS. The b-configuration of the
glycosidic bond was established from 1H and 13C NMR spectra
(serine glycoside 6: H-1 doublet at d 4.66 ppm, J1,2 7.5 Hz and
C-1 at d 100.2 ppm). Reproducibility of the reaction was
confirmed by conducting the reaction on a 5-mmol scale, where
the b-glycoside was isolated in 97 % yield. For further use in
solid phase glycopeptide synthesis, the a-amino acid carboxyl
group of 7 was released by treatment with 50 % trifluoroacetic
acid (TFA) in dichloromethane to give glycosylated amino acid
[Fmoc–L-Ser((Ac)3–b-D-GlcNAc)–OH] building block
(Scheme 1).
8

1154
P. Ryan et al.
Table 1. Variation of the glycosylation conditions
OAc
O
OAc
O
OAc
O
AcO
AcO
AcO
AcO
OAc
O
AcO
AcO
O
O
acceptor 5
TMSOTf
DCM
AcO
AcO
TrocHN
TrocHN
OH
TrocHN
OtBu
OH
O
NH
CCl
FmocHN
FmocHN
TrocHN
O
O
3
9
Condition
6
3
4
Entry
Conditions
YieldA [%]
3
4
6
9
1
2
3
4
5
6
5 (0.9 equiv.), TMSOTf (0.02 equiv.), rt, 2 h
5 (0.9 equiv.), TMSOTf (0.1 equiv.), rt, 2 h
5 (1 equiv.), TMSOTf (0.1 equiv.), rt, 2 h
30
15
10
–
10
–
35B
41C
45D
50E
70F
98G
–
35
30
35
,5
–
–
˚
5 (1 equiv.), TMSOTf (0.1 equiv.), molecular sieves 4 A, ꢀ788C to 08C, 2 h
–
˚
5 (1 equiv.), TMSOTf (0.02 equiv.), molecular sieves 4 A, ꢀ788C to 08C, 2 h
–
–
˚
5 (1.5 equiv.), TMSOTf (0.02 equiv.), molecular sieves 4 A, ꢀ788C to 08C, 2 h
–
–
AIsolated yields.
BReaction conditions: glycosyl donor 4 (0.2 mmol), glycosyl acceptor 5 (0.18 mmol), TMSOTf (0.004 mmol), DCM (2 mL).
CReaction conditions: glycosyl donor 4 (0.2 mmol), glycosyl acceptor 5 (0.18 mmol), TMSOTf (0.02 mmol), DCM (2 mL).
DReaction conditions: glycosyl donor 4 (0.2 mmol), glycosyl acceptor 5 (0.2 mmol), TMSOTf (0.02 mmol), DCM (2 mL).
E
˚
Reaction conditions: glycosyl donor 4 (0.2 mmol), glycosyl acceptor 5 (0.2 mmol), TMSOTf (0.02 mmol), DCM (2 mL), MS 4 A (0.3 g).
F
˚
Reaction conditions: glycosyl donor 4 (0.2 mmol), glycosyl acceptor 5 (0.2 mmol), TMSOTf (0.004 mmol), DCM (2 mL), MS 4 A (0.3 g).
G
˚
Reaction conditions: glycosyl donor 4 (0.2mmol), glycosyl acceptor 5 (0.30 mmol), TMSOTf (0.004mmol), DCM (2mL), MS 4 A (0.3g).
OGT-catalysed O-GlcNAcylation of specific serines or
threonines on more than 1000 nucleocytoplasmic proteins is
an abundant and essential post-translational modification.[41]
Analysis of experimentally determined O-GlcNAc sites has
suggested the modification predominantly takes place on serine
residues. Moreover, substitution of serine with threonine on an
OGT substrate reference peptide caused a 50% reduction in OGT
activity.[42] Therefore, we chose the serine-containing nonapep-
tide from human a-A crystallin (NH2–AIPVSREEK–COOH;
group on the acceptor peptide was glycosylated with trichlor-
oacetimidate-activated N-Troc-protected glucose building
block 4 (3 equiv.) in the presence of promoter TMSOTf (0.02
˚
equiv.) in dry dichloromethane over a few pellets of 4-A
molecular sieves at room temperature for 2 h. Side chain
deprotection and cleavage from the resin using an acidic cocktail
gave the desired glycopeptide 12 in high yield (quantitative,
84 %) as determined by liquid chromatography–mass spectrom-
etry (LC-MS) analysis. Isolation of the glycopeptide by prepar-
ative HPLC provided the glycopeptide in 68 % yield.
Glycopeptide formation was confirmed by HRMS and the
b-configuration of the glycosidic bond was established from
1H (1D and correlation spectroscopy (COSY)) NMR spectra
(serine glycoside 12: H-1 doublet at d 5.62 ppm, J1,2 8.8 Hz). We
observed a downfield shift of the anomeric proton as compared
with glycosylated serine 6, possibly due to long-range correla-
tions through space with the peptide backbone. It is important to
note that only 0.02 equiv. of TMSOTf was employed, so as to
prevent acidolysis of the tert-butyl ester or cleavage from the
resin. In the previous studies, the glycosylation of a simple
pentapeptide Ile-Ser-Gly-Ile-Gly (ISGIG) on polyethylene
glycol-based resin solid support with trichloroacetimidate-
activated N-Troc-protected glycosyl donor 4 under TMSOTf
(0.1 equiv) promotion was reported in 53 % yield.[45]
AIPVSREEK ¼ Ala-Ile-Pro-Val-Ser-Arg-Glu-Glu-Lys) as
a
model system to test this methodology for direct solid-phase
glycosylation. Crystallins are structural proteins responsible for
the optical properties of the lens and the a-A crystallin subunit isa
bona fide physiological OGT substrate.[43,44] The a-A crystallin-
derived peptide NH2–AIPVSREEK–COOH has been shown to
exhibit OGT-catalysed glycosylation with UDP-GlcNAz (UDP
¼ uridine diphosphate).[44] It is a nonapeptide comprising one
serine amino acid, and so is an ideal substrate for testing solid-
phase O-glycosylation methodology.
Synthetic efforts commenced with the preparation of a model
resin-bound peptide (AIPVSREEK) bearing an unprotected side
chain –OH serine residue 11 (Scheme 2). The resin-bound
peptide was synthesised on Wang resin following the Fmoc–
SPPS strategy, to initially afford resin-bound peptide 11 (see
Supplementary Material). Fmoc–Ser(OH)–OH was coupled for
the synthesis of peptide 11. Yields of amino acid coupling steps
were ascertained to be quantitative in all cases, as determined by
measuring the absorbance of the piperidine–fulvene adduct at
l ¼ 301 nm. Nonapeptide formation was confirmed by 1H NMR
and high-resolution mass spectrometry (HRMS) (see Supple-
mentary Material for details). It is interesting to note that the
coupling of four amino acid (Ala–Ile–Pro–Val) on unprotected
H–Ser(OH)–OH yielded a clean product peptide 11 (crude
peptide LC-MS; see Supplementary Material, p. S25).
The effect of TMSOTf on Wang resin-bound peptide was
studied by conducting the glycosylation reaction using an excess
of TMSOTf (0.3 equiv.). LC-MS analysis of the reaction
washings indicated the absence of even trace amounts of peptide
or glycopeptide (see the Supplementary Material for experimen-
tal details).
Conclusion
We have shown the efficient synthesis of the glycosylated amino
acid building block Fmoc–L-Ser((Ac)3–b-D-GlcNAc)–OH.
b-Anomeric selectivity provides the potential for solid-phase
O-glycosylation, which is demonstrated through direct solid-
phase glycosylation of the peptide template AIPVSREEK on
With the solid-support-bound peptide 11 carrying a free side
chain hydroxyl group in hand, the glycosylation reaction was
planned to place the glycan on the peptide. The free hydroxyl

Glycopeptide Synthesis on the Solid Support
1155
HO
Wang resin
1. Fmoc-Lys(Boc)-OH (10 equiv.), DIC (5 equiv.)
DMAP (0.1 equiv.)m DCM/DMF
2. Ac2O/pyridine (1:9 v/v)
NHBoc
O
FmocHN
O
10
1. Fmoc removal: 10 vol-% piperidine in DMF
2. amino acid coupling: 0.2 M Fmoc-AA or Fmoc–Ser–OH in DMF,
PyBOP (4 equiv.), NMM (8 equiv.)
SPPS
NHBoc
tBuO
O
OH
O
O
O
O
O
H
H
H
N
H
N
H
N
O
N
N
N
H
N
N
H
FmocHN
H
O
O
O
O
NH
O
tBuO
HN
NHPbf
11
1. Glycosylation: Glycosyl donor 4, TMSOTf, DCM
2. Cleavage: TFA/DCM (1:1)
3. Preparative HPLC
OAc
AcO
AcO
O
NH2
OH
O
O
HO
NHTroc
O
O
O
O
H
N
H
N
H
N
H
H
N
N
N
N
N
H
FmocHN
H
H
O
O
O
O
O
NH
HO
O
HN
NH2
68 %
12
Scheme 2. Solid-phase synthesis of O-linked glycopeptide.
Wang resin. Moreover, this mild glycosylation protocol may be
also applicable to solid-phase glycopeptide syntheses contain-
ing vulnerable peptide sequences.
million (ppm) downfield shift using the solvent peak as internal
reference. The data are reported as chemical shift (d), multiplicity
(s, br s, d, t, dd, ddd, m, etc.), relative integral, coupling constant
(J in Hz) and assignment where possible. Low-resolution mass
spectra were recorded on a Finnigan LCQ Deca ion trap mass
spectrometer (electrospray ionisation (ESI)). High-resolution
mass spectra were recorded on a Bruker 7T Fourier-Transform
Ion Cyclotron Resonance (FTICR) mass spectrometer.
Analytical reverse-phase HPLC was performed on a Waters
System 2695 separations module with an Alliance Series col-
umn heater at 308C and 2996 photodiode array detector. A
Waters Sunfire 5-mm, 2.1 ꢁ 150-mm column was used at a flow
rate of 0.2 mL minꢀ1 using a mobile phase of 0.1 % TFA in water
(Solvent A) and 0.1 % TFA in acetonitrile (Solvent B) and a
linear gradient of 2–50 % B over 30 min. The results were
analysed with Waters Empower software. Preparative reverse-
phase HPLC was performed using a Waters 600 Multisolvent
Experimental
General
Reactions were monitored by analytical thin-layer chromatog-
raphy (TLC). TLC was performed on commercially prepared
silica plates (Merck Kieselgel 60, 0.25 mm F254). Spots were
detected under UV (254 nm) and or by staining with 7 % (v/v)
ethanolic sulfuric acid. Flash column chromatography was
performed using 230–400 mesh Kieselgel 60 silica, eluting with
analytical-grade solvents.
1H NMR spectra were recorded using a Bruker Avance DPX
400 spectrometer at a frequency of 400.2 MHz. 13C NMR spectra
were recorded on a Bruker Avance DPX 400 spectrometer at a
frequency of100 MHz. The spectra peaks are reportedin parts per

1156
P. Ryan et al.
Delivery System and Waters 500 pump with a 2996 photodiode
array detector or Waters 490E programmable wavelength detec-
tor operating at 254 and 280 nm. A Waters Sunfire 5-mm,
19 ꢁ 150-mm column was used at a flow rate of 7 mL minꢀ1
using a mobile phase of 0.1 % TFA in water (Solvent A) and
0.1 % TFA in acetonitrile (Solvent B) using a linear gradient of
0–100 % B over 60 min.
LC-MS was performed on a Thermo Separation Products
Spectra System consisting of a p400 pump and a UV6000LP
photodiode array detector on a Phenomenex Jupiter 5-mm,
2.1 ꢁ 150-mm column at a flow rate of 0.2 mL minꢀ1 coupled
to a Thermoquest Finnigan LCQ Deca mass spectrometer (ESI)
operating in positive mode. Separations involved a mobile phase
of 0.1 % methanoic acid in water (Solvent A) and 0.1 %
methanoic acid in acetonitrile (Solvent B) using a linear gradient
of 0–100 % B over 15 min.
trichloroethoxycarbonylamino)-a-D-glucopyranosyl trichloroa-
cetimidate (4) (62.5 mg, 0.15mmol, 3 equiv.) were dried over-
˚
night under high vacuum over few pellets of 4-A molecular sieves
in the syringe. DCM (4mL) was injected into the syringe to allow
complete swelling of the resin. After 30min, dilute TMSOTf
(0.18 mL, 0.001 mmol, 0.02 equiv.) in DCM (18 mL of 10mL
mLꢀ1 v/v) was injected and then the syringe was agitated at room
temperature (rt) for 2 h. Then, the resin was washed with DMF
(5ꢁ 5 mL), DCM (5ꢁ 5 mL), and DMF (5 ꢁ 5 mL). The resin
was subsequently washed with DCM (10 ꢁ 3 mL). A mixture of
TFA/DCM (1: 1 v/v, 2 mL) was added to the resin and it was
shaken for 2 h at rt. The resin was then washed with TFA/DCM
(1: 1 v/v, 2 ꢁ 3 mL), and the combined cleavage and washing
solutions were concentrated under vacuum. The crude glycopep-
tide was suspended in cold diethyl ether (,10mL). The suspen-
sion was transferred to a 50-mL Eppendorf tube and centrifuged
at 2415 g for 5 min. The supernatant ether was carefully decanted
and the precipitate was dried under high vacuum (quantitative
crude yield83 mg). LC-MS analysisof the crude product indicated
84 % desired glycopeptide along with 16 % unreacted peptide.
The crude glycopeptide was purified by preparative HPLC and
lyophilised to afford the desired glycopeptide 12 as a fluffy white
solid white solid (57.4 mg, 67 % yield).
TMSOTf was purchased from Aldrich. All other commercial
materials were used as received unless otherwise noted. Amino
acids, coupling reagents and resins were obtained from GL
Biochem and Novabiochem. Dichloromethane and methanol
were distilled over calcium hydride. DMF was obtained as
peptide synthesis grade from Auspep or Labscan.
Experimental Procedures
Preparative HPLC: Rt 37.84 min (0–100 % B over 60 min,
Na-Fluoren-9-ylmethoxycarbonyl-O-3,4,6-tri-O-Acetyl-
2-deoxy-2-(2,2,2-trichloroethoxycarbonylamino)-b-D-
Glucopyranosyl-L-Serine-tert-Butyl Ester (6)
l 254 nm).
LC-MS: Rt 10.94 min (0–100 % B over 15 min, l 254 nm).
m/z (ESI) 1758.4 ([M þ 2Na – H]þ, 100 %). m/z (HRMS)
1755.61017; calc. for C74H104Cl3N14O26Na2: 1755.62103
[M þ 2Na – H]þ.
A solution of glucopyranoside (3.1g, 5.0 mmol) and Fmoc–
Ser–OtBu (2.87 g, 7.5 mmol) in dry dichloromethane (DCM,
˚
50.0mL) was prepared and mixed with freshly activated 4-A
Supplementary Material
molecular sieves (7.5g) at ꢀ788C. To it was added dilute
TMSOTf (18 mL, 0.10 mmol) in DCM (0.90 mL of 20 mL mLꢀ1
v/v) with vigorous stirring. The mixture was allowed to slowly
reach room temperature over 1 h before being neutralised with
triethylamine (13 mL, 0.10 mmol), diluted with toluene, and
concentrated under reduced pressure before purification by flash
chromatography (3: 1 ethyl acetate/toluene) to give the title
compound 6 (4.09 g, 4.83 mmol, 97 %) as a colourless, amor-
phous foam. dH (400MHz CDCl3) 7.77 (d, 2H, J 7.48, Ar), 7.64
(t, 2H, J 7.7, Ar), 7.41 (t, 2H, J 7.4, Ar), 7.33 (tdd, 2H, J 1.1, 4.8,
7.4, Ar), 5.78 (d, 1H, J 6.48, NHFmoc), 5.21–5.29 (m, 2H, 3-CH,
NHTroc), 5.05 (t, 1H, J 9.6, 4-CH), 4.78 (d, 1H, J 11.4, CHH
Troc), 4.66 (d, 1H, J 7.5, 1-CH), 4.52–4.41 (m, 3H, CHH Fmoc,
CHH Troc, a-CH), 4.32–4.22 (m, 3H, CHH Fmoc, 6-CHa, CH
Fmoc), 4.18 (dd, 1H, J 4.1, 10.4, b-CHa), 4.13 (dd, 1H, J 2.1, 12.3,
6-CHb), 3.88 (dd, 1H, J 2.2, 10.3, b-CHb), 3.68 (dq, 1H, J 2.3,
9.8 Hz, 5-CH), 3.59 (q, 1H, 9.2 Hz, 2-CH), 2.07 (s, 3H, COCH3),
2.03(s, 3H, COCH3), 2.02 (s, 3H, COCH3), 1.48 (s, 9H, OtBu). dC
(100 MHz CDCl3) 170.6, 170.6, 169.5, 168.5 (COCH3), 156.2
(COCH2), 154.2 (CO Fmoc), 144.0, 143.6, 141.3, 141.3 (Cquat),
127.8, 127.2, 125.3, 125.1, 120.1, 120.0 (CH Ph), 100.2 (C1),
95.4 (CCl3), 82.9 (CO2C(CH3), 74.4 (CH2 Troc), 71.9 (C3, C5),
69.4 (b-CH2), 68.6 (C4), 67.3 (CH2 Fmoc), 62.0 (C6), 56.3 (C2),
54.4(a-CH), 47.1(CHFmoc), 27.9(CO2C(CH3)), 20.7(COCH3),
20.6 (COCH3). m/z (ESI) 869.3 ([M þ Na]þ, 100 %).
Full experimental details, characterisation of all compounds,
copies of mass spectra, 1H and 13C spectra are available on the
Journal’s website.
Conflicts of Interest
The authors declare no conflicts of interest
Acknowledgements
Financial support to S. R. from an Australian Research Council – Discovery
Early Career Research Award (DE140101632) is gratefully acknowledged.
The authors express their sincere gratitude to Professor Richard J. Payne and
Professor Kate Jolliffe (The University of Sydney) for their support. We
thank Professor Andrew Davey for infrastructure support and Dr Matt Zunk
for his comments on the manuscript.
References
[1] G. W. Hart, J. Biol. Chem. 2014, 289, 34422. doi:10.1074/JBC.R114.
609776
[2] M. B. Lazarus, Y. Nam, J. Jiang, P. Sliz, S. Walker, Nature 2011, 469,
564. doi:10.1038/NATURE09638
[3] J. Ma, G. W. Hart, Clin. Proteomics 2014, 11, 8. doi:10.1186/1559-
0275-11-8
[4] X. Yang, F. Zhang, J. E. Kudlow, Cell 2002, 110, 69. doi:10.1016/
S0092-8674(02)00810-3
[5] W. B. Dias, W. D. Cheung, G. W. Hart, Biochem. Biophys. Res.
Commun. 2012, 422, 224. doi:10.1016/J.BBRC.2012.04.124
[6] G. W. Hart, M. P. Housley, C. Slawson, Nature 2007, 446, 1017.
doi:10.1038/NATURE05815
[7] R. Fujiki, T. Chikanishi, W. Hashiba, H. Ito, I. Takada, R. G. Roeder,
H. Kitagawa, S. Kato, Nature 2009, 459, 455. doi:10.1038/
NATURE07954
Glycosylation Protocol for Synthesis of Glycopeptide 12
The glycosylation was performed with dry dichloromethane
under an argon atmosphere in a Torviq polypropylene syringe.
The reaction was carried out on Wang resin using the model
peptide Fmoc–AIPVSR(Pbf)E(OtBu)E(OtBu)K(Boc) (11). The
resin and the glycosyl donor 3,4,6-tri-O-acetyl-2-deoxy-2-(2,2,2-

Glycopeptide Synthesis on the Solid Support
1157
[8] M. R. Bond, J. A. Hanover, J. Cell Biol. 2015, 208, 869. doi:10.1083/
JCB.201501101
[28] M. A. Brimble, P. J. Edwards, P. W. R. Harris, G. E. Norris, M. L.
Patchett, T. H. Wright, S.-H. Yang, S. E. Carley, Chem. – Eur. J. 2015,
21, 3556. doi:10.1002/CHEM.201405692
[9] X. Yang, P. P. Ongusaha, P. D. Miles, J. C. Havstad, F. Zhang,
W. Venus So, J. E. Kudlow, R. H. Michell, J. M. Olefsky, S. J. Field,
R. M. Evans, Nature 2008, 451, 964. doi:10.1038/NATURE06668
[10] M. Brownlee, Nature 2001, 414, 813. doi:10.1038/414813A
[11] C. M. Ferrer, T. P. Lynch, V. L. Sodi, J. N. Falcone, L. P. Schwab, D. L.
Peacock, D. J. Vocadlo, T. N. Seagroves, M. J. Reginato, Mol. Cell
2014, 54, 820. doi:10.1016/J.MOLCEL.2014.04.026
[12] S. A. Yuzwa, D. J. Vocadlo, Chem. Soc. Rev. 2014, 43, 6839.
doi:10.1039/C4CS00038B
[29] A. Takasu, T. Houjyou, Y. Inai, T. Hirabayashi, Biomacromolecules
2002, 3, 775. doi:10.1021/BM020014S
[30] L. A. Salvador, M. Elofsson, J. Kihlberg, Tetrahedron 1995, 51, 5643.
doi:10.1016/0040-4020(95)00224-V
[31] A. S. Norgren, T. Norberg, P. I. Arvidsson, J. Pept. Sci. 2007, 13, 717.
doi:10.1002/PSC.832
[32] S. A. Mitchell, M. R. Pratt, V. J. Hruby, R. Polt, J. Org. Chem. 2001,
66, 2327. doi:10.1021/JO005712M
[13] J. W. Bullen, J. L. Balsbaugh, D. Chanda, J. Shabanowitz, D. F. Hunt,
D. Neumann, G. W. Hart, J. Biol. Chem. 2014, 289, 10592.
doi:10.1074/JBC.M113.523068
[14] R. G. Spiro, Glycobiology 2002, 12, 43R. doi:10.1093/GLYCOB/12.4.
43R
[15] P. H. Seeberger, W.-C. Haase, Chem. Rev. 2000, 100, 4349.
doi:10.1021/CR9903104
[16] V. L. Campo, A. D. L. Borges, I. Carvalho, J. Braz. Chem. Soc. 2006,
[33] T. Reipen, H. Kunz, Synthesis 2003, 2487.
[34] C. H. Ro¨hrig, O. A. Retz, L. Hareng, T. Hartung, R. R. Schmidt,
ChemBioChem 2005, 6, 1805. doi:10.1002/CBIC.200500079
[35] A. Vargas-Berenguel, M. Meldal, H. Paulsen, K. Bock, J. Chem. Soc.,
Perkin Trans. 1 1994, 2615. doi:10.1039/P19940002615
[36] H. G. Garg, R. W. Jeanloz, Carb. Res. 1976, 52, 246. doi:10.1016/
S0008-6215(00)85969-8
[37] H. Kunz, in Preparative Carbohydrate Chemistry (Ed. S. Hanessian)
17, 648. doi:10.1590/S0103-50532006000400004
1997, pp. 265–281 (Marcel Decker: New York, NY).
[17] Y. Zhang, S. M. Muthana, D. Farnsworth, O. Ludek, K. Adams, J. J.
Barchi, Jr, J. C. Gildersleeve, J. Am. Chem. Soc. 2012, 134, 6316.
doi:10.1021/JA212188R
[38] G. Braum, Synthese von Glycopeptiden mit photoaktivierbaren Grup-
pen im Saccharidteil 1991, Ph.D. thesis, University of Mainz,
Germany.
[18] N.-H. Yao, W.-Y. He, K. S. Lam, G. Liu, J. Comb. Chem. 2004, 6, 214.
doi:10.1021/CC034005E
[39] S. Talat, M. Thiruvikraman, S. Kumari, K. J. Kaur, Glycoconj. J. 2011,
28, 537. doi:10.1007/S10719-011-9353-2
[19] K. M. Halkes, C. H. Gotfredsen, M. Grotli, L. P. Miranda, J. O. Duus,
M. Meldal, Chem. – Eur. J. 2001, 7, 3584. doi:10.1002/1521-3765
(20010817)7:16,3584::AID-CHEM3584.3.0.CO;2-Z
[20] M. Hurevich, P. H. Seeberger, Chem. Comm. 2014, 1851. doi:10.1039/
C3CC48761J
[40] K. M. Halkes, C. H. Gotfredsen, M. Grøtli, L. P. Miranda, J. O. Duus,
M. Meldal, Chem. – Eur. J. 2001, 7, 3584. doi:10.1002/1521-3765
(20010817)7:16,3584::AID-CHEM3584.3.0.CO;2-Z
[41] M. Schimpl, X. Zheng, V. S. Borodkin, D. E. Blair, A. T. Ferenbach,
A. W. Schu¨ttelkopf, I. Navratilova, T. Aristotelous, O. Albarbarawi,
D. A. Robinson, M. A. Macnaughtan, D. M. F. van Alten, Nat. Chem.
Biol. 2012, 8, 969. doi:10.1038/NCHEMBIO.1108
[21] J. Seibel, L. Hillringhaus, R. Moraru, Carb. Res. 2005, 340, 507.
doi:10.1016/J.CARRES.2004.12.014
[22] W. Dullenkopf, J. C. Castro-Palomino, L. Manzoni, R. R. Schmidt,
Carb. Res. 1996, 296, 135. doi:10.1016/S0008-6215(96)00237-6
[23] U. Ellervik, G. Magnusson, Carb. Res. 1996, 280, 251. doi:10.1016/
0008-6215(95)00318-5
[24] K. Sasaki, S. Matsumura, K. Toshima, Tetrahedron Lett. 2006, 47,
9039. doi:10.1016/J.TETLET.2006.10.095
[42] S. Pathak, J. Alonso, M. Schimpl, K. Rafie, D. E. Blair, V. S. Borodkin,
A. W. Schu¨ttelkopf, O. Albarbarawi, D. M. F. van Alten, Nat. Struct.
Mol. Biol. 2015, 22, 744. doi:10.1038/NSMB.3063
[43] E. P. Roquemore, A. Dell, H. R. Morris, M. Panico, A. J. Reason, L.-A.
Savoy, G. J. Wistow, J. S. Zigler, Jr, B. J. Earles, G. W. Hart, J. Biol.
Chem. 1992, 267, 555.
[25] G. Arsequell, L. Krippner, R. A. Dwek, S. Y. C. Wong, J. Chem. Soc.
Chem. Commun. 1994, 2383. doi:10.1039/C39940002383
[26] A. Ste´venin, F.-D. Boyer, J.-M. Beau, Eur. J. Org. Chem. 2012, 1699.
doi:10.1002/EJOC.201200062
[44] T. M. Leavy, C. R. Bertozzi, Bioorg. Med. Chem. Lett. 2007, 17, 3851.
doi:10.1016/J.BMCL.2007.05.008
[45] A. Schleyer, M. Meldal, R. Manat, H. Paulsen, K. Bock, Angew. Chem.
Int. Ed. Engl. 1997, 36, 1976. doi:10.1002/ANIE.199719761
[27] I. Carvalho, S. L. Scheuerl, K. P. R. Kartha, R. A. Field, Carb. Res.
2003, 338, 1039. doi:10.1016/S0008-6215(03)00071-5