Internally quenched fluorogenic, α-helical dimeric peptides and glycopeptides for the evaluation of the effect of glycosylation on the conformation of peptides

Article abstract of DOI:10.1039/a607624f

A panel of α-helical, dimeric coiled-coil peptides has been designed and synthesized for the evaluation of the effect of glycosylation on the conformation of these coiled-coil peptides. Two glycosylated building blocks, N ^α-(fluoren-9-ylmethoxycarbonyl)-O-(2,3,4-tri-O-acetyl-6-azido- 6-deoxy-β-D-glucopyranosyl)-L-threonine pentafluorophenyl ester 8 and N ^α-(fluoren-9-ylmethoxycarbonyl)-O-{2,3,4-tri-O-acetyl-6- [2′-(tert-butoxycarbonylamino)benzoylamino]-6-deoxy-β-D- glucopyranosyl}-L-threonine pentafluorophenyl ester 9 containing the fluorogenic 2-aminobenzamide (Abz) group, have been synthesized. These compounds have been obtained by the glycosylation of N^α-Fmoc-Thr-OPfp with the corresponding glycosyl trichloroacetimidate donors and have been incorporated into the solid-phase synthesis of the peptides 1-3 and 7 and glycopeptides 4-6. Compounds 1 and 4-7 have been synthesized as internally quenched fluorogenic compounds where the Abz group has been employed as the fluorogenic probe and 3-nitrotyrosine Tyr(NO₂) as the quenching chromophore. Steady-state fluorescence studies have provided evidence to support the dimerization of the α-helical peptides. Denaturation studies, by fluorescence as well as CD spectroscopy, indicate that the introduction of a carbohydrate moiety into the coiled-coil peptides has a significant destabilizing effect on the α-helicity.

Full text of DOI:10.1039/a607624f

View Article Online / Journal Homepage / Table of Contents for this issue
Internally quenched fluorogenic, á-helical dimeric peptides and
glycopeptides for the evaluation of the effect of glycosylation on the
conformation of peptides
Seema Mehta, Morten Meldal, Vito Ferro, Jens Ø. Duus and Klaus Bock
Department of Chemistry, Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-2500 Valby,
Copenhagen, Denmark
A panel of á-helical, dimeric coiled-coil peptides has been designed and synthesized for the evaluation of
the effect of glycosylation on the conformation of these coiled-coil peptides. Two glycosylated building
blocks, N á-(fluoren-9-ylmethoxycarbonyl)-O-(2,3,4-tri-O-acetyl-6-azido-6-deoxy-â-D -glucopyranosyl)-L-
threonine pentafluorophenyl ester 8 and N á-(fluoren-9-ylmethoxycarbonyl)-O-{2,3,4-tri-O-acetyl-6-[2Ј-
(tert-butoxycarbonylamino)benzoylamino]-6-deoxy-â-D -glucopyranosyl}-L-threonine pentafluorophenyl
ester 9 containing the fluorogenic 2-aminobenzamide (Abz) group, have been synthesized. These
compounds have been obtained by the glycosylation of N á-F moc-Thr-OPfp with the corresponding
glycosyl trichloroacetimidate donors and have been incorporated into the solid-phase synthesis of the
peptides 1–3 and 7 and glycopeptides 4–6. Compounds 1 and 4–7 have been synthesized as internally
quenched fluorogenic compounds where the Abz group has been employed as the fluorogenic probe and
3-nitrotyrosine Tyr(N O2) as the quenching chromophore. Steady-state fluorescence studies have provided
evidence to support the dimerization of the á-helical peptides. D enaturation studies, by fluorescence as
well as CD spectroscopy, indicate that the introduction of a carbohydrate moiety into the coiled-coil
peptides has a significant destabilizing effect on the á-helicity.
mational averaging over millisecond time-scales and hence are
Introduction
somewhat limited. Fluorescence energy transfer (FET) has fre-
quently been used to study the structure and dynamics of vari-
ous biomolecules. More recently, this technique has been
adopted for the examination of glycopeptide conformation.9
The technique complements the established methods by
operating on a more rapid nanosecond time-scale, and by
providing long-range, intramolecular distances between the
fluorogenic probes. This enables the detection of even subtle
conformational changes. Moreover, FET is a highly sensitive
technique which requires substrates in only micromolar
concentrations.
During the last decade research has emphasized the important
role played by carbohydrates in both the structure and bio-
logical functions of proteins.1 Glycosylation has been shown to
be important for cell surface expression,1 modulation of inter-
molecular interactions,1 efficient secretion of proteins,1,2 the
thermal stability of proteins,1,2 and protein solubility1,3 to name
but a few. Glycosylation is essential for the proper folding and
transport of certain proteins.1,4 In order for us to have a better
understanding of the mechanisms involved in these processes it
is important to have an insight into the conformational flex-
ibilities of the glycosylated peptide chain. One of the points
that needs to be addressed concerns the role of carbohydrates in
imparting particular conformations to the glycopeptides. We
are interested in studying the effect of protein glycosylation on
the secondary structure of peptides.
The present paper describes the synthesis of fluorescent pep-
tides and glycopeptides and the initial investigations of the
effect of glycosylation on the secondary structure of α-helical
coiled-coils.
The coiled-coil structural motif is frequently found in pro-
teins. It has been known for its structural role in fibrous pro-
teins such as myosin, keratin and fibrinogen.5 It is encountered
in a number of DNA-binding proteins and is a convenient
model system with which to study the interaction of peptide
chains during protein folding.
Results and discussion
Design of á-helical dimeric peptides and glycopeptides
Previous work 10 has provided a design for double-stranded,
coiled-coil peptides. These peptides form parallel α-helical
homodimers. The design consists of a heptapeptide repeating
unit denoted as ‘abcdefg’ (Fig. 1 and structures 1–7). The
choice and position of the amino acid residues in each heptad,
as well as the length of the peptide chain, are important for the
stability of the coiled-coil. Leucine residues are placed at the ‘a’
and ‘d’ positions of the heptapeptide. They form the internal
core of the coiled-coil and provide stability through van der
Waalls and hydrophobic interactions [Fig. 1(a)]. Positioning of
oppositely charged amino acids, specifically, glutamic acid at
the ‘e’ position and lysine at the ‘g’ position of the heptapeptide
provide additional stability to the dimer by interhelical electro-
static interaction. Due to the high intrinsic α-helical propensity
of alanine and its small steric requirements,6,10 it has been
incorporated into the peptide at a number of positions. The
heptapeptide is repeated four times to provide a peptide with 28
residues, which is the minimum length required to form stable
The study of the influence of glycosylation on the conform-
ation of peptides which form well defined parallel α-helical
dimers is of considerable interest. In aqueous solution, such
peptides exist in an equilibrium between the α-helical dimeric
state and the random coil monomeric state.6 The α-helical
dimeric state may be visualized as a pair of α-helices wrapped
around each other to form a superhelix. In order to assess the
effect of glycosylation on the conformation of such peptides a
shift in the monomer–dimer equilibrium upon glycosylation of
the parent peptide could be measured. Earlier studies aimed at
investigating the conformational effects of glycosylation in
glycopeptides have employed a number of techniques, such as
nuclear magnetic resonance (NMR) spectroscopy7 and circular
dichroism (CD) spectroscopy.8 However, solution conform-
ations derived from NMR spectroscopy are results of confor-
J. Chem. Soc., Perkin Trans. 1, 1997
1365

View Article Online
abcdefg
1 Abz-ELEALEKKLAALETKLQALEKKLEALEY(NO2)G-NH2
2 Ac-Y(NO2)ELEALEKKLAALETKLQALEKKLEALEYG-NH2
3 Abz-ELEALEKKLAALETKLQALEKKLEALEYG-NH2
N3
O
HO
HO
O
OH
4 Abz-ELEALEKKLAALETKLQALEKKLEALEY(NO2)G-NH2
N3
O
HO
HO
O
OH
5 Abz-ELEALEKKLAATETKLQALEKKLEALEY(NO2)G-NH2
H2N
O
NH
O
HO
HO
O
OH
6 Ac-ELEALEKKLAALETKLQALEKKLEALEY(NO2)G-NH2
7 Abz-ELEALEKKLAALETKLQY(NO2)LEKKLEALEYG-NH2
replaced by tyrosine and Tyr(NO2), respectively. Peptide 3 was
designed as a fluorogenic peptide, devoid of the quenching
chromophore. The Abz moiety of peptide 1 was retained and
the quencher Tyr(NO2) was replaced by a non-quenching tyro-
sine (Tyr) residue. We contemplated that by mixing dilute solu-
tions of peptides 2 and 3, heterodimers would result and the
fluorescence of the peptide 3 would be strongly quenched by the
quenching chromophore of peptide 2. A study of the rate of
decay of the fluorescence by time-resolved fluorescence spec-
troscopy indeed confirmed that the dimerization occurred in a
parallel manner.15
In order to investigate the influence of glycosylation on the
conformation of peptide 1, the glycopeptides 4 and 5 were
designed. For the synthesis of the O-linked glycopeptide sub-
strates 4–6 the method of choice was to synthesize glycosylated
threonine building blocks and subsequently to incorporate
them into the solid-phase peptide synthesis. In the glycopeptide
4, a glycosylated threonine residue was introduced at position
15. The helical wheel representation of this glycopeptide [Fig.
1(b)] illustrates the position of the sugar residue with respect
to the coiled-coil. In this position, the sugar is pointing out-
wards, away from the α-helical dimer, so that any effect on
the tendency to form α-helical dimers can be assigned to the
tendency to form the secondary structure and is not due to
any influence on the dimerization. In the peptide 5 the site of
glycosylation was shifted to position 13. The Leu13 in pep-
tide 1 was replaced by a glycosylated threonine residue. This
corresponds to the position ‘d’ of the second heptad. In this
arrangement the sugar residue is positioned on the inner
hydrophobic face of the α-helical dimer, directly between the
coiled-coil [Fig. 1(c)].
Fig. 1
coiled-coils.11 The terminal carboxy group is protected as the
carboxamide and the terminal amino group as its acetyl deriv-
ative, to avoid helix-destabilizing interactions with the helix
dipole. We maintained this basic design of O’Neill and
DeGrado 6 and performed certain modifications. The peptides
and glycopeptides described in this paper were synthesized as
internally quenched fluorescent compounds in which a fluores-
cent group was placed at one end of the substrate and a quench-
ing group at the other end. In order to incorporate fluorescent
probes into this design, the modifications presented in Fig. 1
and structures 1–7 were performed.
FET between the fluorogenic chromophore (energy donor)
and the quenching chromophore (energy acceptor) occurs via
resonance (or long-range) energy transfer.12–14 To ensure effi-
cient FET it is important to have a high quantum yield of the
fluorescence donor and a good spectral overlap between the
emission band of the fluorogenic chromophore and the absorp-
tion band of the quenching chromophore.13 One donor–
acceptor pair that meets these criteria very well is Abz and
Tyr(NO2).14 The Abz group exhibits an emission maximum at
420 nm upon excitation at 320 nm. This coincides with the
absorption maxima of Tyr(NO2). Moreover, both these groups
are easily incorporated by solid-phase synthesis and they have
minimal influence on the conformational and amphipathic
properties of the peptide products. Thus, the fluorescent probe
Abz was introduced at the amino terminus of the peptides and
glycopeptides and the quenching chromophore, Tyr(NO2), was
positioned towards the carboxy terminus (structures 1–7). A
threonine residue was introduced at position 15 in the peptides,
which corresponded to the ‘f’ position of the second heptad.
This facilitated the introduction of the carbohydrate moiety, in
the case of glycopeptides.
FET varies inversely with the sixth power of the distance
between the fluorescence donor and the fluorescence acceptor.13
In order to vary the interprobe distance and to provide another
handle for conformational studies, the glycopeptide 6 and the
peptide 7 were synthesized. In the glycopeptide 6, the fluor-
escent Abz group was positioned on the sugar moiety rather
than the peptide backbone. For the peptide 7, the quenching
chromophore was moved from the fourth heptad to the third
heptad. Leu19 was replaced by a Tyr(NO2).
It was expected that peptides of this design would dimerize in
a parallel fashion.6,10 In order to provide further support for
this by fluorescence the peptides 2 and 3 were designed as con-
trol compounds. Peptide 2 was a nonfluorogenic peptide that
contained only the quenching chromophore at the amino ter-
minus. Thus, the Tyr(NO2) and Abz groups in peptide 1 were
Design and synthesis of the building blocks
Two glycosylated building blocks, the non-fluorogenic com-
1366
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
pound 8 and the fluorogenic compound 9, were designed for the
incorporation into solid-phase peptide synthesis. The synthesis
of compounds 8 and 9 is presented below.
obtained from compound 13 by the selective removal of the
anomeric acetate with hydrazine acetate20 in DMF to provide
the hemiacetals 14 in 82% yield, followed by their reaction with
trichloroacetonitrile and potassium carbonate in dichloro-
methane.19 The desired α-trichloroacetimidate 15 was obtained
in a yield of 81%. The corresponding β-compound was also
isolated, in 6% yield. Nα-Fmoc--threonine-OPfp 16 was
obtained from its commercially available tert-butyl derivative,
by treatment with trifluoroacetic acid for 1.5 h followed by crys-
tallization from diethyl ether–hexane. Glycosylation of Nα-
Fmoc--threonine-OPfp 16 with the glycosyl trichloroacetimi-
date 15 in the presence of 0.1 mol equiv. of trimethylsilyl tri-
fluoromethanesulfonate as the promoter afforded the desired
building block 8 in 85% yield.
O
H
N
O
N3
O
O
AcO
AcO
O
CH3
OPfp
NH
O
OAc
FmocNH
AcO
AcO
O
CH3
OPfp
O
OAc
FmocNH
8
9
O
The synthetic approach to compound 9 involved the intro-
duction of the Abz-functionality at the 6-position, followed by
activation of the anomeric centre and subsequent coupling with
the protected amino acid. Initial efforts to reduce the 6-azido
group of compound 13 and to introduce the Abz group with
F
F
O
F
F
O
Fmoc =
Pfp =
F
3,4-dihydro-4-oxo-1,2,3-benzotriazol-3-yl
2-tert-butyloxy-
carbonylamino benzoate (Boc-Abz-ODhbt),14 were not success-
ful. The reduction of the azido group was performed in the
presence of Pd/C in methanol and acetic acid. Attempts to iso-
late the amine 17, and to treat it quickly with Boc-Abz-ODhbt
did not afford the desired compound 18. The major product 19,
was a result of the migration of the acetyl group from the 4-
to the 6-position. To overcome this difficulty, we attempted to
do the two-step reaction in a single step and to perform the
reduction in the presence of Boc-Abz-ODhbt. This however
was not effective. The use of zinc/acetic acid instead of hydro-
gen and Pd/C did not offer any advantage. However, improved
reaction conditions were realized when the reduction of com-
pound 13 was performed in the presence of Pd/C and the
acylating agent, with only tetrahydrofuran (THF) as the solv-
ent. The reduction was conducted in the presence of 3 mole
equivalents of Boc-Abz-ODhbt to afford the desired com-
pound 18 in 88%.
The phenyl thioglycosyl donor 20 was synthesized. Attempts
to effect glycosylations of amino acids with this building block
were unsuccessful due to the unreactive nature of the com-
pound. In order to investigate other activating procedures, as
well as other glycosyl donors, the reactive glycosyl acceptor
21,21 which contained a primary hydroxy group at the 6 pos-
ition, was synthesized as a model acceptor. Glycosylation of the
methyl glycoside 21 with the thioglycoside 20 was attempted
under nitrosyl tetrafluoroborate promotion.22 However, the
major product was compound 22, formed as a result of trans-
esterification. A more reactive glycosyl donor was required and
the glycosyl trichloroacetimidate approach was contemplated.
A mixture of the α- and β-hemiacetal 23 was obtained by the
treatment of compounds 18 with hydrazine acetate,20 in 82%
yield. Reaction of the hemiacetals 23 with trichloroacetonitrile
and potassium carbonate afforded a mixture of the α- and β-
glycosyl trichloroacetimidates 24 (87%) in the ratio of 2:1.
These were used without separation. Glycosylation of the
glycosyl acceptor 21 with glycosyl donor 24 was performed in
the presence of a catalytic amount of trimethylsilyl trifluoro-
methanesulfonate. The desired disaccharide 25 was obtained
stereoselectively in an unoptimized yield of 51%. The low yield
was a result of transesterification side products and disacchar-
ides with less than the expected number of acetyl protecting
groups were also isolated. The next step was the glycosylation
of the protected amino acid Nα-Fmoc--threonine-OPfp 16
with the glycosyl trichloroacetimidate 24. This was also success-
ful and afforded, stereoselectively, the building block 9 in 60%
yield.
The selective tritylation of the 6-position of glucose with
triphenylmethyl (trityl) chloride in pyridine followed by acetyl-
ation of the remaining hydroxy groups with acetic anhydride
and pyridine afforded compound 10.16 Removal of the trityl
group was effected with 80% aq. acetic acid at 50 ЊC in 2 h.
The resulting compound 11, upon treatment with trifluoro-
methanesulfonic anhydride and 2,6-di-tert-butyl-4-methyl-
pyridine provided compound 1217 in 85% yield. The trifluoro-
methanesulfonate at the 6-position was displaced with sodium
azide in N,N-dimethylformamide (DMF) to afford compound
1318 (91%).
O
H
N
O
O
R2
O
AcO
AcO
NH
O
OR1
AcO
AcO
AcO
R
OAc
R1
R2
R
Ac
Ac
Ac
Ac
H
OTr
OH
OTf
N3
N3
N3
10
11
12
13
14
15
17
OAc
β-SPh
OH
18
20
23
24
27
OC( NH)CCl3
β-SOPh
α-C( NH)CCl3
Ac
NH2
HO
FmocNH
CH3
NHAc
O
OR
O
OPfp
HO
AcO
BzlO
OAc BzlO
OAc
BzlO
O
OMe
16
19
21 R = H
22 R = Ac
O
H
N
O
O
N3
O
NH
O
AcO
AcO
SPh
AcO
AcO
OAc
O
26
OAc
BzlO
O
BzlO
BzlO
OMe
25
Synthesis of á-helical dimeric peptides and glycopeptides
The substrates 1–7 were synthesized by solid-phase peptide
synthesis on the PEGA resin 23 in DMF. Two peptides–
The glycosyl donor of choice for the synthesis of the building
block was the glycosyl trichloroacetimidate 15.19 This was
J. Chem. Soc., Perkin Trans. 1, 1997
1367

View Article Online
glycopeptides were synthesized simultaneously in 2 parallel
columns using an automated peptide synthesizer.24 The p-(α-
amino-2,4-dimethoxybenzyl)phenoxyacetic acid 25 (Rink) was
employed as the linker. The linker was coupled to the resin
by activation with O-(1H-benzotriazol-1-yl)-N,N,NЈ,NЈ-tetra-
methyluronium tetrafluoroborate (TBTU)26 and 4-ethylmor-
pholine (NEM). Amino acids were added as their Nα-Fmoc-
protected Pfp ester derivatives. The use of the Fmoc group
enables deprotection of the α-amino group under mild con-
ditions of 20% piperidine in DMF. The Pfp esters27 served to
protect the carboxy group during glycosylations as well as to
activate it during acylation.28 3,4-Dihydro-3-hydroxy-4-oxo-
1,2,3-benzotriazine (DhbtOH) was added as indicator and aux-
iliary nucleophile to enhance the reactivity of the Pfp-esters.
Acylation can be followed by observing the disappearance of
the yellow colour of DhbtOϪ, either visually or with a solid-
phase photometer.24 The control peptide 2 was synthesized first.
In this peptide the yellow Tyr(NO2) residue was the final one to
be incorporated into the chain and the synthesis of this peptide
enabled the reaction times of each amino acid coupling to be
noted and subsequently applied for the synthesis of analogue
peptides 3–7.
Fig. 2 Urea denaturation profile of a mixture of compounds 2 and 3
studied by fluorescence spectroscopy at substrate concentration of 5.0
µM in 50 mM TRIS chloride buffer, pH 8.5
ation studies were performed on a mixture of the control pep-
tides 2 and 3. A mixture of the quenching peptide 2 and the
fluorogenic peptide 3 is capable of forming the homodimers 2–2
and 3–3, as well as the heterodimer 2–3 (Scheme 1). This
Cleavage of the Nα-Fmoc protecting group was accomplished
by treatment with 20% piperidine in DMF for 15 min. The
cleavage was monitored by UV spectroscopy at 320 nm.
Tyr(NO2) was coupled as its Nα-Fmoc derivative in the presence
of TBTU and NEM. The coupling of subsequent amino acids
was performed with their Nα-Fmoc-protected Pfp ester deriv-
atives (3 mol equiv.). Dhbt-OH (1 mol equiv.) was added for
each coupling as the auxiliary nucleophile to catalyse the acyl-
ation. Couplings were allowed to proceed for 1–3 h. Longer
coupling times were required for the coupling of the threonine
and the glutamine residues and these were allowed to proceed
overnight. The glycosylated threonine building blocks 8 or 9
(2.5–3 mol equiv.) were allowed to react for 6 h. The final Abz
group was introduced with Boc-Abz-ODhbt 14 and the coupling
was allowed to proceed for 7 h. Following the final coupling, the
resin was washed with piperidine (30 min). Substrates 2 and 6
were further treated with 10% acetic anhydride in DMF to
acylate the free terminal amine. Cleavage of the substrates from
the resin was carried out by treatment with 95% aq. TFA with
concurrent removal of the side-chain-protecting groups on the
amino acids. Removal of the O-acetyl groups on the carbo-
hydrates provided the glycopeptides 4–6 by treatment with a
catalytic amount of sodium methoxide in methanol at a con-
trolled pH of 10 for 4–5 h. All compounds were purified by
reversed-phase HPLC and analysed by amino acid analysis and
electrospray mass spectroscopy (ES-MS). The pure peptides
and glycopeptides 1–7 were obtained in yields of 60–80%.
homodimer 2–2
homodimer 3–3
2 ϩ 2 random-coil monomer
3 ϩ 3 random-coil monomer
ϩ
2–3 heterodimer
Scheme 1
experiment furthermore probes the assumption that the pep-
tides align in parallel fashion and not in an anti-parallel one, as
only the parallel arrangement will position the fluorescence
donor and quencher in close proximity to give a stronger
quenching than that observed in the homodimers.
Peptides 2 and 3 were mixed in equimolar concentrations, the
mixture was subjected to varying concentrations of urea (1.0–
7.0 M) and the fluorescence intensity of the solution was moni-
tored. As the urea concentration was increased, an increase in
the fluorescence was observed. This corresponded to a shift
in the equilibrium from the α-helical heterodimeric state to
the random-coil monomeric state. When the fluorescence was
plotted as a function of urea concentration, a typical sigmoidal
urea denaturation curve was obtained (Fig. 2).
To obtain the urea denaturation curve of the internally
quenched peptide 1, the peptide was subjected to varying con-
centrations of urea and the changes in fluorescence were moni-
tored. However, well defined denaturation curves were not
obtained. At this point it was decided to study the urea
denaturation of the heterodimer 1–2, as opposed to the
homodimer 1–1, in a series of experiments similar to those per-
formed for the control heterodimer 2–3. In this case, the
changes in fluorescence would be more pronounced and thus
more easily observed. Indeed, the denaturation study of the
heterodimer 1–2 afforded well defined denaturation curves.
Similar experiments were repeated with the heterodimers 2–4
and 2–5. The results are presented in Fig. 3.
A comparison of the urea denaturation profiles allowed us to
comment on the relative stabilities of the heterodimers formed
by the peptide 1 and the glycopeptides 4 and 5 with the control
peptide 2. The midpoints of the denaturation curve of peptide 1
(3.8 M) and the corresponding glycopeptide 4 (3.3 M) were
separated by about 0.5 M urea, where the peptide 1 was more
resistant to denaturation than the glycopeptide 4. This result
suggested that the sugar residue, on the external face of the
coiled-coil (Fig. 1), had an adverse effect on the ability of the
peptide to dimerize.
Circular dichroism characterization of compounds 1–7
The CD spectra of the peptides 1–3 and 7 and the glycopeptides
4–6 were measured at 5 µM concentrations and the spectra
exhibit minima at 220 and 207 nm as well as a maximum at 192
nm indicative of an α-helical conformation. The glycopeptide 5
showed less ellipticity at 207 and 220 nm, indicating less α-
helical structure. In general the CD spectra cannot be used for
quantification of the fraction of α-helical structure, as the CD
signals from both the fluorescent donor acceptor pair and the
sugar moiety are not known. This can indirectly be seen in the
fact that using the normal signal intensity for a purely α-helical
peptide for the 220 nm signal some of the peptides would have
an estimated fraction of α-helical structure slightly more than
100%.
Stability of the á-helical homodimers and the heterodimers.
In order to evaluate the influence of the carbohydrate moiety
on the stability of α-helical peptides, it was decided to assess
and compare the stability of the substrates 1, 4 and 5 by moni-
toring their fluorescence intensity as a function of urea concen-
tration at a fixed peptide concentration. Initial urea denatur-
Interestingly, the fluorescence of the mixture of the glyco-
1368
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Fig. 5 Formation of ‘dimers’: 2 ϩ 1, 2 ϩ 3, 2 ϩ 4, 2 ϩ 5, studied by
fluorescence spectroscopy at substrate concentrations of 5.0 µM in 50
mM TRIS chloride buffer, pH 8.5
amino-2-(hydroxymethyl)propane-1,3-diol (TRIS) buffer or 2
M urea solutions. The intensity of fluorescence at the begin-
ning of the experiment was lower for the mixtures 2 ϩ 1,
2 ϩ 4, 2 ϩ 5 as compared to 2 ϩ 3. Since compounds 1, 4, 5
possess a fluorogenic group as well as a quenching chromo-
phore, their initial fluorescence was lower due to the internal
quenching factor. For peptide 3, the only mode of quenching
is external, hence the fluorescence intensity was comparatively
elevated.
Fig. 3 Urea denaturation profiles of a mixture of compounds 2 ϩ 1
(—᭝—), 2 ϩ 4 (— —), 2 ϩ 5 (—ᮀ—), studied by fluorescence spec-
*
troscopy at substrate concentrations of 5.0 µM in 50 mM TRIS chloride
buffer, pH 8.5
The fluorescence of the solutions was monitored over time
(Fig. 5). For the mixtures 2 ϩ 1, 2 ϩ 3 and 2 ϩ 4, in pure buffer
as well as in 2 M urea (data not shown), the fluorescence
decreased with time. This can be attributed to the formation of
hetero-aggregates, where the fluorescence of compounds 1, 3
and 4 was quenched by the quenching peptide 2. Interestingly,
the fluorescence of the glycopeptide 5 remained constant in
spite of the glycopeptide being mixed with the quenching pep-
tide 2. This result supports the conclusion that the presence of
the sugar residue at site 13 on glycopeptide 5 hinders the form-
ation of dimers or higher aggregates.
A comparison of the curves obtained in buffer alone (Fig. 5)
with those obtained in 2 M urea indicated that the rate of for-
mation of hetero-aggregates from homo-aggregates is higher in
the latter situation. In 2 M urea, the equilibrium is shifted
slightly towards the random-coil monomeric state (Scheme 1)
and thereby enables faster heterodimerization.
In conclusion a set of analogues of 4 heptad repeat coiled-
coil double-stranded peptides and glycopeptides were success-
fully synthesized in order to study peptide alignment and
dynamics of the formation of the coiled-coil dimer. The peptide
analogue 6 containing a fluorescent label in the glycan linked to
a central amino acid residue was prepared through several
routes, and the route using azido protection during the glyco-
sylation and reduction of azide in the presence of acylating
agents seems more general. An alternative route involving
reduction of the azido group after incorporation on the solid
phase has recently been described.29
The fluorescence and CD data obtained clearly demonstrate
that the introduction of a single O-linked carbohydrate unit at
the threonine of the solvent-exposed face of a coiled-coil pep-
tide reduces the tendency to form α-helical dimers. In analogy
to the study by O’Neill and DeGrado 6 this demonstrates that
the inherent tendency to form α-helical structure is lowered by
the glycosylation. The introduction of a carbohydrate moiety at
the internal face of the coiled-coil peptide effectively destroys
the dimerization. A further investigation using time-resolved
fluorescence measurements in order to obtain more detailed
information on the structure and dynamics of the dimeric pep-
tides is currently being performed.
Fig. 4 Urea denaturation profiles of a solution of compounds 1
(—᭝—), 4 (—×—), 5 (—ᮀ—), studied by CD spectroscopy at substrate
concentrations of 120 µM in 50 mM TRIS chloride buffer, pH 8.5
peptide 5 with 2 is substantially higher at 0 M urea than for the
mixtures 2 ϩ 3 and 2 ϩ 4. Moreover, the fluorescence remained
constant in spite of the increased urea concentration. These
results indicate the reduced ability of the glycopeptide 5 to form
homodimers as well as heterodimers and that the dimerization
of compounds 2 and 5 was indeed abolished when the sugar
residue is positioned along the internal face of the coiled-coil
(Fig. 1).
In addition to the fluorescence study, the stability of the het-
erodimers was also corroborated by CD spectroscopy (Fig. 4).
The urea denaturation profile of compounds 1 and 3–5 revealed
that the glycopeptide 4 was more easily denatured as compared
to the parent peptide 1; the midpoints of their denaturation
curves were separated by ~1 M urea (2.8 and 3.8 M, respect-
ively, at 5 µM peptide concentrations). These results are in
agreement with the fluorescence study. The denaturation
experiments investigated by monitoring changes in fluorescence
involved hetero-aggregates. Hence the interaction of one sugar
residue, between a pair of peptides in a coiled-coil, was exam-
ined. In the case of the CD experiments, the species under
investigation were homo-aggregates and the effect of the sugar
residue was doubled, demonstrating the destabilizing effect of
the glycosylation to be additive.
For the glycopeptide 5, well defined urea denaturation curves
were not obtained. The ellipticity remained more or less con-
stant as the concentration of urea was increased from 0 M to
8 M. This result, in conjunction with the fluorescence results,
imply the inability of the glycopeptide 5 to aggregate, possibly
due to the interaction of the sugar residues.
In order to detect the interaction between the peptide 2
and the compounds 1 and 3–5 upon mixing, and to investi-
gate the kinetics of dimerization, the following experiments
were performed. Peptide 2 was mixed, in turn, with an equi-
molar concentration of compounds 1 and 3–5, in either 2-
Experimental
General
Analytical TLC was performed on Merck Silica Gel 60 F254
aluminium sheets with detection by UV light and by charring
with sulfuric acid. Vacuum liquid chromatography (VLC) was
J. Chem. Soc., Perkin Trans. 1, 1997
1369

View Article Online
performed on an open glass column with a sintered filter
packed dry with suction using Merck silica gel H60 and equi-
librated with a mixture of EtOAc and hexanes (1:1). Com-
pounds were purified by medium-pressure chromatography
on Kieselgel 60 (230–400 mesh). For chromatography under
anhydrous conditions, Kieselgel 60 (230–400 mesh) was dried at
120 ЊC for >24 h prior to use. Solvents were purchased from
Labscan Ltd. (Dublin, Ireland). Light petroleum was the frac-
tion boiling at 60–80 ЊC. Dichloromethane was distilled over
P2O5. DMF was freshly distilled by fractional distillation at
reduced pressure and kept over 4 Å molecular sieves. Con-
centrations were performed under reduced pressure at temper-
atures <40 ЊC (bath). 4-(Hydroxymethyl)benzoic acid (HMBA)
and suitably protected Nα-Fmoc amino acids were purchased
from NovaBiochem (Switzerland), DhbtOH, NEM and TBTU
from Fluka (Switzerland). Nα-Fmoc-Tyr(NO2)-OH and Boc-
Abz-ODhbt were prepared as previously described.14 The pep-
tides and the glycopeptides were hydrolysed with 6 M HCl at
110 ЊC for 24 h and the amino acid composition was deter-
mined on a Pharmacia LKB Alpha Plus amino acid analyser,
Asn and Gln were determined as Asp and Glu, respectively.
ES-MS spectra were recorded in the positive mode on a VG
Quattro Mass Spectrometer from Fisons, with 50% aq.
acetonitrile as the liquid phase. Matrix-assisted laser-
desorption time-of-flight mass spectrometry (MALDI-TOF-
MS) was performed on a Finnigan Mat 2000 mass spectrometer
using a matrix of α-cyano-4-hydroxycinnamic acid. NMR
spectra were recorded on a Bruker AM-500 or a Bruker AMX-
600 MHz spectrometer. For all compounds, the assignment of
1H NMR spectra was based on 2D proton–proton shift-
correlation spectra. The assignment of 13C NMR spectra was
based on carbon–proton shift correlation. NMR spectra were
recorded in CDCl3 (reference: CDCl3 at δH 7.3 and δC 77.0).
J-Values are given in Hz. Preparative HPLC was performed on
a Waters system with a 600 controller, a 991 photodiode array
detector, equipped with a preparative flow cell, and a model 600
pump with modified 80 cm3 minϪ1 pump heads. The system was
mmol).26 Cleavage of the Nα-Fmoc protecting group was
accomplished by treatment with 20% piperidine in DMF for 15
min. The cleavage was monitored by UV spectroscopy at 320
nm. This was followed by washing with DMF (30 cm3 for 18
min). Tyr(NO2) was coupled as its Nα-Fmoc derivative (52 mg,
0.12 mmol) in the presence of TBTU (35 mg, 0.11 mmol), and
NEM (30 mm3, 0.24 mmol). A coupling time of 3 h was
allowed. Coupling of subsequent amino acids was performed
with their Nα-Fmoc-protected Pfp ester derivatives (3 mol
equiv.). The side chains on the glutamic acid were protected
with tert-butyl groups. Glutamine was used as its trityl deriv-
ative, lysine as its tert-butoxycarbonyl (Boc) derivative, and
threonine as its tert-butyl derivative. DhbtOH (1 mol equiv.)
was added for each Pfp ester coupling as the auxiliary nucleo-
phile to catalyse the acylation. Couplings were allowed to pro-
ceed for 1–3 h. Longer coupling times were required for the
coupling of the threonine and the glutamine residues and these
were allowed to proceed overnight. Building blocks 8 and 9
(2.5–3 mol equiv.) were employed and the coupling was allowed
to proceed for 6 h. Finally the Abz group was introduced with
Boc-Abz-ODhbt 14 and the coupling was allowed to proceed for
7 h. Following the final coupling, the resin was washed with
piperidine (30 min). Resins with protected peptides 2 and 6 were
further treated with 10% acetic anhydride in DMF (30 min) to
acylate the free terminal amine. The resin was transferred into a
syringe and washed with DMF (×10) and CH2Cl2 (×10), dried
for 2 h and lyophilized overnight. Cleavage of the peptides from
the resin was carried out by treatment with 95% aq. TFA (5–6
cm3) for 3 h. This effected concomitant removal of the side-
chain-protecting groups on the amino acids. The resin was
transferred into a sintered funnel, washed with 95% TFA fol-
lowed by 95% aq. acetic acid (×4). The filtrate was concentrated
and lyophilized. All compounds were purified by reversed-
phase HPLC and were analysed by amino acid analysis and
ES-MS. For the deprotection of glycopeptides 4–6 they were
dissolved in anhydrous methanol (1 cm3 per 2 mg) and a freshly
prepared solution of sodium methoxide in methanol (0.2 M)
was added dropwise (to pH 10–11). The reaction mixtures were
stirred for 4–5 h and neutralized with acetic acid. The solutions
were concentrated, lyophilized, purified by reversed-phase
HPLC and analysed by amino acid analysis and ES-MS.
fitted with a switchable Delta Pak (25 × 200 mm; 10 cm3 minϪ1
)
and a preparative radial pack column (50 × 300 mm; 20 cm3
minϪ1), both packed with reversed-phase C18 or a Shodex DS-
2013 column (20 × 300 mm; 4 cm3 minϪ1). Analytical HPLC
was performed using a Waters RCM 8 × 10 module with a
Waters 8 NV C18 (4 µm) column (1 cm3 minϪ1). The solvent
system was buffer A: 0.1% aq. TFA and buffer B: 0.1% TFA in
90% acetonitrile–10% water and detection was at 215 or 320
nm. Mps were measured on a Buchi melting point apparatus
and are uncorrected. Optical rotations were recorded on a
Perkin-Elmer 241 polarimeter; [α]D -values are given in units of
10Ϫ1 deg cm2 gϪ1. Microanalysis was provided by Leo Pharm-
aceutical Products (Ballerup, Denmark) or Department of
Chemistry, Copenhagen University (Denmark).
Abz-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-Leu-Glu-
Thr-Lys-Leu-Gln-Ala-Leu-Glu-Lys-Lys-Leu-Glu-Ala-Leu-Glu-
Tyr(NO2)-Gly-NH2 1
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 10% B, then a linear gradient of 10–
90% B during 70 min; retention time 52.0 min) to give pure
compound 1 (120.0 mg, 87%). Amino acid analysis is presented
in Table 1 {Found (ES-MS): m/z, 1146.6 [M ϩ 3 H]3ϩ. 860.4
[M ϩ 4 H]4ϩ; (MALDI) 3436.6 (M ϩ H)ϩ. C155H256N38O49
requires M, 3435.9}.
Fluorescence measurements were performed on a Perkin-
Elmer luminescence spectrometer LS50. The substrates were
excited at 320nm and their fluorescencewasmonitored at 420nm,
both with 5 nm slits. All measurements were performed at 25 ЊC.
The CD measurements were carried out as previously des-
cribed on a Jobin Yvon Model IV dischrograph at NOVO
Nordisk, Bagsværd, Denmark.30
Ac-Tyr(NO2)-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-
Leu-Glu-Thr-Lys-Leu-Gln-Ala-Leu-Glu-Lys-Lys-Leu-Glu-Ala-
Leu-Glu-Tyr-Gly-NH2 2
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 10% B, then a linear gradient of 10–
100% B during 70 min; retention time 53.0 min) to give pure
compound 2 (99.0 mg, 72 %). Amino acid analysis is presented
in Table 1 {Found: (MALDI) 3562.5 (M ϩ K)ϩ. C159H262N38O51
requires M, 3522.1}.
Solid-phase peptide synthesis. General procedure
Synthesis of the peptides/glycopeptides 1–7 was performed on
the PEGA resin 23 in DMF. Two peptides/glycopeptides were
synthesized simultaneously in two parallel columns using an
automated peptide synthesizer.24 PEGA resin (200 mg, 0.04
mmol) was transferred into each column. The resin was allowed
to swell in DMF for 2 h and was then treated with 20% piperi-
dine in DMF for 10 min to remove any chloride ions and to
expose the free amino groups on the resin. The resin was deriv-
atized with the Rink linker 25 (54 mg, 0.1 mmol) in the pres-
ence of TBTU (32 mg, 0.09 mmol) and NEM (25 mm3, 0.2
Abz-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-Leu-Glu-
Thr-Lys-Leu-Gln-Ala-Leu-Glu-Lys-Lys-Leu-Glu-Ala-Leu-Glu-
Tyr-Gly-NH2 3
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 20% B, then a linear gradient of 20–
90% B during 80 min; retention time 62.0 min) to give pure
compound 3 (132.0 mg, 78%). Amino acid analysis is presented
1370
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
Table 1 Amino acid analysis of glycopeptides and peptides 1–7. Theoretical values are given in parentheses
Compound
Thr
Glu
Gly
Ala
Leu
Tyr
Tyr(NO2)
0.97 (1)
Lys
1
2
3
4
5
6
7
0.93 (1) 7.89 (8) 1.22 (1) 4.70 (5) 8.50 (8)
0.87 (1) 8.04 (8) 1.11 (1) 5.05 (5) 8.16 (8) 0.87 (1) 0.96 (1)
0.91 (1) 7.91 (8) 1.18 (1) 4.70 (5) 8.60 (8)
0.88 (1) 8.01 (8) 1.11 (1) 5.10 (5) 8.50 (8)
2.00 (2) 8.50 (8) 1.20 (1) 4.80 (5) 6.61 (7)
0.79 (1) 7.94 (8) 1.13 (1) 5.26 (5) 8.34 (8)
4.85 (5)
4.91 (5)
4.92 (5)
4.72 (5)
4.80 (5)
4.72 (5)
4.85 (5)
0.91 (1)
0.88 (1)
0.90 (1)
0.83 (1)
0.86 (1) 8.14 (8) 1.10 (1) 4.40 (4) 8.60 (8) 0.87 (1) 0.93 (1)
in Table 1 {Found (ES-MS): m/z, 1131.6 [M ϩ 3 H]3ϩ; 858.7
[M ϩ Na ϩ 3 Li]4ϩ; 849.06 [M ϩ 4 H]4ϩ; (MALDI) 3392.1
(M ϩ H)ϩ. C155H257N37O47 requires M, 3391.0}.
neutralized with acetic acid and concentrated. The crude prod-
uct was purified by semipreparative HPLC (Delta Pak column;
10 min 20% B, then a linear gradient of 20–80% B during 70
min; retention time 52.5 min) to give pure compound 6 (21
mg, 83%). Amino acid analysis is presented in Table 1 {Found
(ES-MS): m/z, 1228.1 [M ϩ Na ϩ 3 Li]3ϩ; (MALDI) 3639.5
(M ϩ H)ϩ. C163H269N39O54 requires M, 3639.2}.
Abz-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-Leu-Glu-
Thr(6-azido-6-deoxy-â-D-glucopyrannosyl)-Lys-Leu-Gln-Ala-
Leu-Glu-Lys-Lys-Leu-Glu-Ala-Leu-Glu-Tyr(NO2)-Gly-NH2 4
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 15% B, then a linear gradient of 15–
85% B during 60 min; retention time 52.0 min) to give pure
protected glycopeptide (105.0 mg, 70%). The protected sub-
strate (25 mg) was dissolved in anhydrous methanol (15 cm3)
and a freshly prepared solution of sodium methoxide in metha-
nol (0.2 M) was added dropwise (to pH 10–11), the reaction
mixtures were stirred for 3.5 h, at which point analytical HPLC
indicated complete deacetylation. The reaction mixture was
neutralized with acetic acid and concentrated. The crude prod-
uct was purified by semipreparative HPLC (Delta Pak column;
10 min 15% B, then a linear gradient of 15–85% B during 70
min; retention time 56.0 min) to give pure compound 4 (15 mg,
63%). Amino acid analysis is presented in Table 1 {Found (ES-
Abz-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-Leu-Glu-
Thr-Lys-Leu-Gln-Tyr(NO2)-Leu-Glu-Lys-Lys-Leu-Glu-Ala-
Leu-Glu-Tyr-Gly-NH2 7
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 15% B, then a linear gradient of 15–
85% B during 70 min; retention time 52.0 min) to give pure
compound 7 (120.0 mg, 85%). Amino acid analysis is presented
in Table 1 {Found (ES-MS): m/z, 1765.5 [M ϩ 2 H]2ϩ; 1177.1
[M ϩ 3 H]3ϩ. C161H260N38O50 requires M, 3528.1}.
1,2,3,4-Tetra-O-acetyl-6-O-trifluoromethylsulfonyl-â-D-gluco-
pyranose 12
1,2,3,4-Tetra-O-acetyl-β--glucopyranose 1117,31,32 (1.20 g, 3.45
mmol) was added portionwise to a solution of trifluoro-
methanesulfonic anhydride (1.36 g, 0.81 cm3, 4.82 mmol) and
2,6-di-tert-butyl-4-methylpyridine (1.00 g, 4.88 mmol) in
dichloromethane (60 cm3). The mixture was stirred at 25 ЊC for
2 h and then was poured into cold, saturated aq. NaHCO3. The
organic phase was separated and the aqueous phase was
extracted with dichloromethane. The combined organic phases
were washed successively with water and brine, dried (MgSO4)
and concentrated. Recrystallization of the residue from
dichloromethane–hexanes then gave the trifluoromethane-
sulfonate 12 (1.36 g) as needles, mp 95–96 ЊC (decomp.) [lit.,17
85 ЊC (decomp.)] [α]D ϩ5.3 (c 0.8, CHCl3). The mother liquors
were subjected to vacuum liquid chromatography (VLC)
[EtOAc–hexanes (1:1)] to give a further 0.20 g of product.
Total yield: 1.56 g (94%); (CDCl3) 20.4, 20.5, 20.7 and 20.8
(4 × COCH3), 72.7 (C-6), 67.8 (C-4), 69.9 (C-2), 72.4 (C-3), 71.9
(C-5), 91.4 (C-1), 118.5 (CF3Ј, JCF 319.8) and 168.7, 169.1, 169.3
and 170.0 (4 × COCH3); δH (CDCl3) 2.06, 2.08, 2.11 and 2.16
(12 H, 4 s, 4 × Ac), 4.55 (1 H, dd, J5,6a 4.4, J6a,6b 11.3, Ha-6), 4.58
(1 H, dd, J5,6b 3.1, Hb-6), 4.00 (1 H, ddd, J4,5 9.4, H-5), 5.12 (1 H,
t, J3,4 9.2, H-4), 5.17 (1 H, dd, J1,2 8.2, J2,3 9.2, H-2), 5.31 (1 H, t,
H-3) and 5.78 (1 H, d, H-1).
MS): m/z, 1812.62 [M ϩ 2 H]2ϩ; 1222.2 [M ϩ 2 Li ϩ Na]3ϩ
;
1215.9 [M ϩ Na ϩ 2 H]3ϩ; 1208.8 [M ϩ 3 H]3ϩ; (MALDI)
3623.6 (M ϩ H)ϩ. C161H265N41O53 requires M, 3623.1}.
Abz-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-
Thr(6-azido-6-deoxy-â-D-glucopyrannosyl)-Glu-Thr-Lys-Leu-
Gln-Ala-Leu-Glu-Lys-Lys-Leu-Glu-Ala-Leu-Glu-Tyr(NO2)-
Gly-NH2 5
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 15% B, then a linear gradient of 15–
85% B during 70 min; retention time 68.0 min) to give pure
protected glycopeptide (130.0 mg, 87%). The protected sub-
strate (80 mg) was dissolved in anhydrous methanol (35 cm3)
and a freshly prepared solution of sodium methoxide in
methanol (0.2 M) was added dropwise (to pH 9–10); the reac-
tion mixture was stirred for 4 h, at which point analytical HPLC
indicated complete deacetylation. The reaction mixture was
neutralized with acetic acid and concentrated. The crude prod-
uct was purified by semipreparative HPLC (Delta Pak column;
10 min 15% B, then a linear gradient of 15–85% B during 60
min; retention time 49.0 min) to give pure compound 5 (70 mg,
90%). Amino acid analysis is presented in Table 1 {Found (ES-
MS): m/z, 1204.5 [M ϩ 3 H]3ϩ; 903.9 [M ϩ 4 H]4ϩ; 723.4
[M ϩ 5 H]5ϩ. C159H261N41O54 requires M, 3611.1}.
1,2,3,4-Tetra-O-acetyl-6-azido-6-deoxy-â-D-glucopyranose 13
To a mixture of the trifluoromethanesulfonate 12 (3.04 g, 6.33
mmol) and sodium azide (2.0 g, 30.8 mmol) was added DMF
(40 cm3) and the mixture then was stirred at 25 ЊC for 1 h before
being poured into water and extracted with ethyl acetate. The
extracts were washed successively with water and brine, dried
(MgSO4) and concentrated. Recrystallization of the residue
from diethyl ether–hexanes then gave the azide 13 (1.81 g) as
needles, mp 86 ЊC; [α]D ϩ9.5 (c 2.3, CHCl3); δC(CDCl3) 20.4,
20.6, 20.6 and 20.8 (4 × COCH3), 50.6 (C-6), 68.9 (C-4), 70.1
(C-2), 72.6 (C-3), 73.8 (C-5), 91.5 (C-1) and 168.7, 169.4, 169.6
and 170.2 (4 × COCH3); δH (CDCl3) 2.04, 2.07, 2.09 and 2.23
(12 H, 4 s, 4 × Ac), 3.37 (1 H, dd, J5,6a 5.3, J6a,6b 13.5, Ha-6), 3.40
(1 H, dd, J5,6b 3.3, Hb-6), 3.84 (1 H, ddd, J4,5 9.4, H-5), 5.10 (1 H,
t, J3,4 9.4, H-4), 5.16 (1 H, dd, J1,2 10.1, J2,3 9.4, H-2), 5.27 (1 H,
Abz-Glu-Leu-Glu-Ala-Leu-Glu-Lys-Lys-Leu-Ala-Ala-Leu-Glu-
Thr[6-(2Ј-aminobenzoylamino)-6-deoxy-â-D-glucopyrannosyl]-
Lys-Leu-Gln-Ala-Leu-Glu-Lys-Lys-Leu-Glu-Ala-Leu-Glu-
Tyr(NO2)-Gly-NH2 6
The crude product was purified by semipreparative HPLC
(Delta Pak column; 10 min 15% B, then a linear gradient of 15–
85% B during 70 min; retention time 52.0 min) to give pure
protected glycopeptide (90.0 mg, 60%). The protected substrate
(20 mg) was dissolved in anhydrous methanol (8 cm3) and a
freshly prepared solution of sodium methoxide in methanol
(0.2 M) was added dropwise (to pH 10–11). The reaction
mixture was stirred for 5 h, at which point analytical HPLC
indicated complete deacetylation. The reaction mixture was
J. Chem. Soc., Perkin Trans. 1, 1997
1371

View Article Online
t, H-3) and 5.75 (1 H, d, H-1) (Found: C, 45.1; H, 5.0.
C14H19N3O9 requires C, 45.0; H, 5.1%). The mother liquors
were subjected to VLC [EtOAc–hexanes (2:3)] to give a further
crop (0.44 g) of product. Total yield: 2.25 g (95%).
(1:1)] of the residue then gave the sulfoxide 27 (118 mg, 94%) as
a solid 5.5:1 mixture of diastereomers, used without further
purification; m/z 633.3 (M ϩ H)ϩ and 655.3 (M ϩ Na)ϩ
(Found: C, 56.9; H, 5.9. C30H36N2O11S requires C, 57.0; H,
5.7%).
Phenyl 2,3,4-tri-O-acetyl-6-azido-6-deoxy-1-thio-â-D-gluco-
pyranoside 26
A small portion was recrystallized from EtOAc–Et2O to give
the major diastereomer as needles, mp 142–143 ЊC; δC(CDCl3)
20.6 (3 × COCH3), 28.3 [C(CH3)3], 39.8 (C-6), 67.6 (C-2), 68.9
(C-4), 73.5 (C-3), 77.2 (C-5), 80.2 [C(CH3)3], 90.5 (C-1), 119.8–
140.3 (Ar), 152.9 [C(O)C(CH3)3] and 168.7, 168.9, 169.6 and
A mixture of compound 13 (373 mg, 1.0 mmol), thiophenol
(220 mg, 0.204 cm3, 2.0 mmol), boron trifluoride–diethyl ether
(710 mg, 0.628 cm3, 5.0 mmol) and 3 Å molecular sieves in
dichloromethane (2 cm3) was stirred at 25 ЊC for 21 h. The
mixture was then diluted with dichloromethane and poured into
saturated aq. NaHCO3. The organic phase was separated,
washed successively with water and brine, dried (MgSO4) and
concentrated. VLC [EtOAc–hexanes (2:3)] of the residue then
gave the thioglycoside 26 (359 mg, 85%) as needles, mp 120–
121 ЊC (from CH2Cl2–hexanes); [α]D Ϫ11 (c 0.6, CHCl3);
δC(CDCl3) 20.6 and 20.7 (3 × COCH3), 51.3 (C-6), 69.2
(C-4), 69.9 (C-2), 73.8 (C-3), 77.0 (C-5), 85.7 (C-1), 128.7–
133.8 (arom. C) and 169.2, 169.4 and 170.2 (3 × COCH3);
δH (CDCl3) 2.03, 2.06 and 2.14 (9 H, 3 s, 3 × Ac), 3.34 (1 H, dd,
J5,6a 6.5, J6a,6b 13.4, Ha-6), 3.40 (1 H, dd, J5,6b 2.9, Hb-6), 3.70
(1 H, ddd, J4,5 9.7, H-5), 4.76 (1 H, d, J1,2 10.1, H-1), 4.99 (1 H,
dd, J2,3 9.7, H-2), 5.00 (1 H, t, J3,4 9.3, H-4) and 5.25 (1 H, t,
H-3) (Found: C, 51.0; H, 5.0. C18H21N3O7S requires C, 51.0;
H, 5.0%).
170.3 (C᎐O); δ (CDCl ) 1.55 (9 H, s, But), 2.07, 2.11 and 2.13
᎐
H
3
(9 H, 3 s, 3 × Ac), 3.26 (1 H, ddd, J5,6a 7.6, Ha-6), 3.64 (1 H, ddd,
J4,5 9.6, H-5), 3.89 (1 H, m, J5,6b 2.5, J6a,6b 14.4, Hb-6), 4.30 (1 H,
d, J1,2 9.9, H-1), 5.01 (1 H, t, J3,4 9.2, H-4), 5.37 (1 H, t, H-3),
5.44 (1 H, dd, J2,3 9.5, H-2), 6.22 (1 H, br t, NHAbz), 7.02–8.42 (9
H, ArH) and 9.95 (1 H, s, NHBoc).
2,3,4-Tri-O-acetyl-6-azido-6-deoxy-á-D-glucopyranosyl
trichloroacetimidate 15
A mixture of α- and β-peracetylated 6-azido-6-deoxygluco-
pyranoside 13 (0.6 g, 1.6 mmol) was dissolved in DMF (16
cm3). Hydrazinium acetate (0.2 g, 2.4 mmol) was added and the
reaction mixture was stirred under argon for 3 h. The reaction
was quenched with ethyl acetate (20 cm3), diluted with dichlo-
romethane (20 cm3), stirred for 5 min, and washed with aq.
sodium chloride (5%; 20 cm3). The organic extracts were dried
over magnesium sulfate, filtered and concentrated. The syrup
was purified by column chromatography with hexane–ethyl
acetate (1:1) as eluent (Rf 0.4). A mixture of 2,3,4-tri-O-acetyl-
6-azido-6-deoxy-α/β--glucopyranose 14 was obtained as a
foam (0.43 g, 81%).
Compounds 14 (0.3 g, 0.9 mmol) were dissolved in CH2Cl2
(4.5 cm3). Trichloroacetonitrile (0.9 cm3, 9.0 mmol) and potas-
sium carbonate (1.0 g, 7.2 mmol) were added and the reaction
mixture was stirred under argon for 12 h before being filtered
through Celite and concentrated. The residue was chromato-
graphed with hexane–ethyl acetate (1.75:1) as eluent [Rf α-
isomer 0.5; β-isomer 0.35]. The title compound was obtained as
a foam (α-isomer: 0.33 g, 75%, β-isomer: 0.26 g, 6%). α-Isomer:
[α]2D0 91.7 (c 9.3, CH2Cl2); δC(CDCl3) 20.4, 20.6 and 20.7
(3 × COCH3), 50.6 (C-6), 68.9 (C-4), 69.7 (C-3, -2), 71.1 (C-5),
Phenyl 2,3,4-tri-O-acetyl-6-[2Ј-(tert-butoxycarbonylamino)-
benzoylamino]-6-deoxy-1-thio-â-D-glucopyranoside 20
The thioglycoside 26 (1.27 g, 3.0 mmol) was suspended in dry
methanol (25 cm3). Methanolic sodium methoxide (1 cm3; 1
mol dmϪ3) was then added and the mixture was stirred at room
temperature until dissolution was complete (10 min). The mix-
ture was then stirred for a further 1 h, neutralized (Amberlite
IRC-50, Hϩ), filtered and concentrated. The residue was redis-
solved in methanol (25 cm3), triethylamine (1.21 g, 1.67 cm3,
12.0 mmol) and dithiothreitol (1.85 g, 12.0 mmol) were added,
and the mixture then was stirred at room temperature for 16 h
before being concentrated, the residue was dissolved in THF
(25 cm3) and Boc-Abz-ODhbt (1.26 g, 3.3 mmol) was added.
The mixture was stirred for 1.5 h and then acetic anhydride (5
cm3) and pyridine (5 cm3) were added. The mixture was stirred
overnight and then was poured into water and extracted with
ethyl acetate. The extracts were washed successively with dil.
HCl, water, saturated aq. NaHCO3 and brine, dried (MgSO4)
and concentrated. VLC [EtOAc–hexanes (1:4–3:7)] of the
residue then gave the title compound 20 (1.56 g, 84% from 26) as
plates, mp 145–146 ЊC (from EtOAc–hexanes); [α]D ϩ4 (c 0.5,
CHCl3); δC(CDCl3) 20.5, 20.6 and 20.7 (3 × COCH3), 28.3
[C(CH3)3], 39.7 (C-6), 69.2 (C-4), 70.1 (C-2), 73.6 (C-3), 76.4
(C-5), 80.2 [C(CH3)3], 85.3 (C-1), 119.0–140.5 (Ar), 153.0
92.7 (C-1), 160.7 (C᎐N) and 169.5, 169.8 and 170.0
᎐
(3 × COCH3); δH (CDCl3) 2.06, 2.08 and 2.10 (9 H, 3 s, 3 × Ac),
3.36 (1 H, dd, J5,6a 5.4, J6a,6b 13.5, Ha-6), 3.45 (1 H, dd, J5,6b 2.9,
Hb-6), 4.22 (1 H, ddd, J4,5 10.2, H-5), 5.16 (1 H, dd, J1,2 3.7, J2,3
10.2, H-2), 5.18 (1 H, t, J3,4ϩ4,5 19.4, H-4), 5.5 (1 H, t, J2,3ϩ3,4
19.6, H-3), 6.63 (1 H, d, J1,2 3.7, H-1) and 8.76 (1 H, s, NHCCl3)
(Found: C, 35.4; H, 3.7; N, 11.6. Calc. for C14H17N3O8Cl3:
C, 35.4; H, 3.6; N, 11.8%).
Ná-(Fluoren-9-ylmethoxycarbonyl)-O-(2,3,4-tri-O-acetyl-6-
azido-6-deoxy-á-D-glucopyranosyl)-L-threonine pentafluoro-
phenyl ester 8
[C(O)C(CH ) ] and 168.7, 169.3, 169.8 and 170.1 (C᎐O);
᎐
3
3
δH (CDCl3) 1.54 (9 H, s, But), 2.03, 2.12 and 2.13 (9 H, 3 s,
3 × Ac), 3.41 (1 H, ddd, J5,6a 6.6, Ha-6), 3.96 (1 H, m, J5,6b 2.7,
J6a,6b 14.3, Hb-6), 3.74 (1 H, ddd, J4,5 9.8, H-5), 4.80 (1 H, d, J1,2
10.1, H-1), 4.96 (1 H, t, J3,4 9.4, H-4), 5.00 (1 H, dd, J2,3 9.7,
H-2), 5.29 (1 H, t, H-3), 6.53 (1 H, br t, NHAbz), 6.98–8.44 (9 H,
ArH) and 10.14 (1 H, s, NHBoc) (Found: C, 58.5; H, 5.9.
C30H36N2O10S requires C, 58.4; H, 5.9%).
A
mixture of 2,3,4-tri-O-acetyl-6-azido-6-deoxy-α--gluco-
pyranosyl trichloroacetimidate 15 (0.10 g, 0.20 mmol), Nα-
(fluoren-9-ylmethoxycarbonyl)--threonine pentafluorophenyl
ester 16 (0.13 g, 0.25 mmol), freshly activated 4 Å molecular
sieves in anhydrous CH2Cl2 (1 cm3) was stirred under argon for
1 h. The reaction mixture was cooled to Ϫ78 ЊC and trimethyl-
silyl trifluoromethanesulfonate (3 mm3, 0.015 mmol) was
added. After 0.5 h the reaction mixture was warmed to room
temperature during 10 min to ensure completion of the reac-
tion. The mixture was recooled to Ϫ78 ЊC and quenched with
1 drop of triethylamine. The reaction mixture was filtered, con-
centrated and subjected to column chromatography on a short
column of pre-dried silica gel with hexane–ethyl acetate (1.5:1)
as eluent (Rf 0.4). The title compound was obtained as a powder
(0.15 g, 85%), [α]D20 4.6 (c 9.9, CH2Cl2); δC(CDCl3) 14.2 (Thrg),
20.6 (COCH3), 47.1 (Fmocβ), 50.9 (C-6), 58.5 (Thrα), 67.4
(Fmocα), 69.4 (C-4), 71.4 (C-2), 72.2 (C-3), 73.3 (Thrβ), 73.5
2,3,4-Tri-O-acetyl-6-[2Ј-(tert-butoxycarbonylamino)benzoyl-
amino]-6-deoxy-1-phenylsulfinyl-â-D-glucopyranose 27
To a solution of the sulfide 20 (123 mg, 0.20 mmol) at Ϫ78 ЊC
was added 3-chloroperbenzoic acid (MCPBA) (49 mg of 85%
from Fluka, 0.24 mmol). The mixture was allowed to warm to
Ϫ20 ЊC over a period of 1.5 h and then was poured into satur-
ated aq. NaHCO3. The organic phase was separated and the
aqueous phase was extracted with dichloromethane. The com-
bined organic phases were washed successively with water and
brine, dried (MgSO4) and concentrated. VLC [EtOAc–hexanes
1372
J. Chem. Soc., Perkin Trans. 1, 1997

View Article Online
(C-5), 97.6 (C-1), 119.9–143.3 (Ar), 156.5 (CONH), 166.3
19.6, H-4), 5.48 (1 H, d, H-1), 5.60 (1 H, t, J2,3ϩ3,4 19.5, H-3),
[C(O)OPfp] and 169.3, 169.4 and 170.2 (COCH3); δH (CDCl3)
6.79 (1 H, t, NHAbz), 7.0–8.4 (4 H, ArH) and 10.00 (1 H, s,
NHBoc).
1.35 (3 H, d, JThr ,Thr 6.3, Thrg), 2.07, 2.08 and 2.09 (9 H, 3 s,
4 × COCH3), 3.28 (1 H, dd, J5,6a 2.2 J6a,6b 13.0, Ha-6), 3.35 (1 H,
dd, J5,6b 6.7, J6a,6b 13.0, Hb-6), 3.70 (1 H, m, J4,5 9.0, H-5), 4.30
β
g
β-Isomer: δH (CDCl3) 1.55 (9 H, s, But), 2.038, 2.10 and 2.13
(9 H, 3 s, 3 × Ac), 3.50 (1 H, dt, J6a,6b 14.5, J6a,6bϩ6a,NH 11.9, Ha-
6), 3.74 (1 H, m, H-5), 3.88 (1 H, m, Hb-6), 4.80 (1 H, d, J1,2 8.0,
H-1), 4.90 (1 H, dd, J1,2 8.0, J2,3 9.5 H-2), 4.99 (1 H, t, J3,4ϩ4,5
18.8, H-4), 5.29 (1 H, t, J2,3ϩ3,4 19.0, H-3), 6.79 (1 H, m, NHAbz),
7.0–8.4 (4 H, ArH) and 9.96 (1 H, s, NHBoc).
(1 H, t, JF moc ,F moc 7.0, Fmocβ), 4.48 (1 H, t, JF moc ,F moc 10.0,
α
β
α
α
α
α
α
β
JF moc ,F moc 6.8, Fmoc ), 4.54 (1 H, t, Fmoc ), 4.64–4.70 (2 H, m,
Thrβ and H-1), 4,79 (1 H, dd, JN H ,Thr 9.1, JThr ,Thr 2.0, Thrα),
5.0 (1 H, t, J1,2ϩ2,3 18.0, H-2), 5.04 (1 H, t, J3,4ϩ4,5 18.8, H-4), 5.28
α
α
β
α
(1 H, t, J2,3ϩ3,4 19.0, H-3), 5.63 [1 H, d, JN H ,Thr 9.2, C(O)NH]
A mixture of α- and β-2,3,4-tri-O-acetyl-6-[2Ј-(tert-butoxy-
and 7.20–7.85 (8 H, ArH); ES-MS 827.0 [M ϩ Li]ϩ.
C37H33F5N4O12 requires M, 820.2 (Found: C, 53.91; H, 4.10;
N, 6.86. Calc. for C37H33N4O12F5: C, 54.15; H, 4.05; N,
6.83%).
carbonylamino)benzoylamino]-6-deoxy--glucopyranose
23
(0.6 g, 1.0 mmol) was dissolved in CH2Cl2 (5.5 cm3). Trichloro-
acetonitrile (0.72 cm3, 7.3 mmol) and potassium carbonate (0.5
g, 3.5 mmol) were added and the reaction was stirred under
argon for 5 h. The reaction mixture was filtered through Celite
and concentrated. The residue was chromatographed with
hexane–ethyl acetate (1.2:1) as eluent (Rf 0.44). The title com-
pounds 24 were obtained as a foam (0.6 g, 87%; α/β = 2).
1,2,3,4-Tetra-O-acetyl-6-[2Ј-(tert-butoxycarbonylamino)-
benzoylamino]-6-deoxy-á/â-D-glucopyranose 18
A
mixture of 1,2,3,4-tetra-O-acetyl-6-azido-6-deoxy-β--
glucopyranose 13 (0.6 g, 1.6 mmol) and Boc-Abz-ODhbt (2.0 g,
5.2 mmol) was dissolved in anhydrous THF (36 cm3) and Pd/C
(0.12 g) was added. The reaction mixture was stirred under a
positive pressure of hydrogen for 18 h, filtered through Celite
and concentrated. The residue was chromatographed with
hexane–ethyl acetate (1:1.5) as eluent (Rf 0.3). The inseparable
impurity was crystallized out with hexane–ethyl acetate (1:2)
and the concentrated filtrate was chromatographed with
hexane–ethyl acetate (1:1.5) to afford the title compounds 18
(0.8 g, 88%).
α-Isomer: δC(CDCl3) 20.4, 20.47 and 20.5 (3 × COCH3), 28.3
[C(CH3)3], 39.1 (C-6), 68.9 (C-4), 69.7 (C-3), 69.9 (C-2), 70.6
(C-5), 80.2 [C(CH3)3], 92.7 (C-1), 119.1–140.4 (Ar), 152.9
[C(O)C(CH ) ], 160.9 (C᎐N) and 168.8–169.9 (C᎐O); δ -
᎐
᎐
3
3
H
(CDCl3) 1.56 (9 H, s, But), 2.06, 2.07 and 2.17 (9 H, 3 s, 3 × Ac),
3.53 (1 H, dt, J6a,6b 14.5, J5,6aϩ6a,N H 12.0, Ha-6), 3.87 (1 H, ddd,
J5,6b 2.6, J6b,N H 6.7, Hb-6), 4.24 (1 H, ddd, J4,5 10.3, J5,6a 6.1,
H-5), 5.06 (1 H, t, J3,4ϩ4,5 19.9, H-4), 5.13 (1 H, t, J1,2 3.8 J2,3
10.0, H-2), 5.62 (1 H, t, J2,3ϩ3,4 19.8, H-3), 6.56 (1 H, d, H-1),
6.57 (1 H, t, JN H ,6 6.3, NHAbz), 7.0–8.4 (4 H, ArH), 8.69 (1 H, s,
NHCCl3) and 10.05 (1 H, s, NHBoc).
α-Isomer: δC(CDCl3) 20.4–20.9 (3 × COCH3), 29.0 [C(CH3)3],
39.09 (C-6), 68.9 (C-4), 69.4 (C-2), 69.7 (C-3), 70.6 (C-5), 80.2
[C(CH3)3], 89.1 (C-1), 119.3–140.5 (Ar), 153.1 [C(O)C(CH3)3]
β-Isomer: δC(CDCl3) 95.8 (C-1); δH (CDCl3) 1.55 (9 H, s,
But), 2.069, 2.07 and 2.15 (9 H, 3 s, 3 × Ac), 3.90–3.97 (3 H, m,
H-5 and H2-6), 5.30–5.37 (3 H, m, H-2,-3 and -4), 5.86 (1 H, d,
J1,2 7.5, H-1), 6.61 (1 H, NHAbz), 7.0–8.4 (4 H, ArH), 8.75 (1 H,
s, NHCCl3) and 10.03 (1 H, s, NHBoc) (Found: C, 46.55; H,
4.99; N, 6.07. C26H32N3O11Cl3 requires C, 46.69; H, 4.82; N,
6.28%).
and 168.9–170.2 (C᎐O); δ (CDCl ) 1.55 (9 H, s, But), 2.05,
᎐
H
3
2.06, 2.14 and 2.2 (12 H, 4 s, 4 × Ac), 3.54 (1 H, m, Ha-6),
3.85 (1 H, m, Hb-6), 4.14 (1 H, m, H-5), 5.05 (1 H, t, J3,4ϩ4,5 19.0,
H-4), 5.10 (1 H, dd, J1,2 3.8 J2,3 10.3, H-2), 5.53 (1 H, t, J2,3ϩ3,4
19.9, H-3), 6.35 (1 H, d, H-1), 6.54 (1 H, br t, NHAbz), 7.0–8.4
(4 H, ArH) and 10.01 (1 H, s, NHBoc).
β-Isomer: δC(CDCl3) 20.4–20.9 (3 × COCH3), 29.0
[C(CH3)3], 39.3 (C-6), 68.8 (C-4), 70.4 (C-2), 72.7 (C-3), 73.5
(C-5), 80.2 [C(CH3)3], 89.1 (C-1), 119.3–140.5 (Ar), 153.1
Methyl 2,3,4-tri-O-benzyl-6-O-{2Ј,3Ј,4-tri-O-acetyl-6Ј-[2Ј-(tert-
butoxycarbonylamino)benzoylamino]-6-deoxy-â-D-gluco-
pyranosyl}-á-D-glucopyranoside 25
[C(O)C(CH ) ] and 168.9–170.2 (C᎐O); δ (CDCl ) 1.55 (9 H,
A
mixture of 2,3,4-tri-O-acetyl-6-[2Ј-tert-butoxycarbonyl-
᎐
3
3
H
3
s, But), 2.057, 2.08, 2.159 and 2.16 (12 H, 4 s, 4 × Ac), 3.57 (1 H,
m, Ha-6), 3.85 (2 H, m, H-5 and Hb-6), 5.03 (1 H, t, J3,4ϩ4,5 20.0,
H-4), 5.16 (1 H, dd, J1,2 8.2, J2,3 9.3, H-2), 5.31 (1 H, t, J2,3ϩ3,4
18.8, H-3), 5.73 (1 H, d, H-1), 6.57 (1 H, br t, NHAbz), 7.0–8.4 (4
H, ArH) and 10.06 (1 H, s, NHBoc); ES-MS (Found: 589.4
[M ϩ Na]ϩ. C26H34N2O12 requires M, 566.3).
amino)benzoylamino]-6-deoxy-α/β--glucopyranosyl trichloro-
acetimidate 24 (0.1 g, 0.15 mmol), methyl 2,3,4-tri-O-benzyl-α-
-glucopyranoside 21 (0.1 g, 0.22 mmol) and freshly activated
4 Å molecular sieves in anhydrous CH2Cl2 (2 cm3) was stirred
under argon for 1 h. The reaction mixture was cooled to Ϫ78 ЊC
and trimethylsilyl trifluoromethanesulfonate (3 mm3, 0.015
mmol) was added. After 10 min the reaction mixture was
warmed to room temperature and was stirred for 5 h. The mix-
ture was recooled to Ϫ30 ЊC and quenched with 1 drop of tri-
ethylamine before being filtered, concentrated and subjected to
column chromatography with hexane–ethyl acetate (1:1) as
eluent (Rf 0.35). The title compound 25 was obtained as a syrup
(0.07 g, 51%); δC(CDCl3) 20.7 (COCH3), 28.3 [C(CH3)3], 39.3
(C-6Ј), 55.2 (OCH3), 68.5 (C-6), 69.0 (C-4Ј), 69.6 (C-5Ј), 71.3
(C-2Ј), 72.4 (C-5), 72.8 (C-3Ј), 73.4, 74.9 and 75.7 (CH2C6H5),
77.5 (C-4), 79.7 (C-2), 80.2 [C(CH3)3], 81.9 (C-3), 98.2 (C-1Ј),
100.9 (C-1), 121.4–141.0 (Ar), 153.0 [COC(CH3)3] and 169.0,
2,3,4-Tri-O-acetyl-6-[2Ј-(tert-butoxycarbonylamino)benzoyl-
amino]-6-deoxy-á/â-D-glucopyranosyl trichloroacetimidate 24
A mixture of 1,2,3,4-tetra-O-acetyl-6-[2Ј-(tert-butoxycarbonyl-
amino)benzoylamino]-6-deoxy-α/β--glucopyranose 18 (0.75 g,
1.3 mmol) was dissolved in DMF (13 cm3). Hydrazinium acet-
ate (0.18, 2.0 mmol) was added and the reaction mixture was
stirred under argon for 2 h. The reaction was quenched with
ethyl acetate (20 cm3), diluted with dichloromethane (20 cm3),
stirred for 5 min, and washed with aq. sodium chloride (5%; 20
cm3). The organic extracts were dried over magnesium sulfate,
filtered and concentrated. The syrup was purified by column
chromatography with hexane–ethyl acetate (1:1) as eluent
(Rf 0.45). A 2.5:1 mixture of α- and β-2,3,4-tri-O-acetyl-6-[2Ј-
(tert-butoxycarbonylamino)benzoylamino]-6-deoxy--gluco-
pyranose 23 (0.57 g, 82%) was obtained. α-Isomer: δC(CDCl3)
20.5, 20.7 and 21.0 (3 × COCH3), 26.2 [C(CH3)3], 39.2 (C-6),
67.6 (C-5), 69.5 (C-3), 69.7 (C-4), 71.3 (C-2), 80.4 [C(CH3)3],
119.5–140.0 (Ar), 153.1 [C(O)C(CH3)3], 169.9, 170.1, 170.2 and
169.6 and 170.2 (C᎐O); δ (CDCl ) 1.55 (9 H, s, But), 1.98, 2.03
᎐
H
3
and 2.12 (9 H, 3 s, 3 × Ac), 3.37 (3 H, s, OCH3), 3.47 (1 H, t,
J3,4ϩ4,5 18.7, H-4), 3.53 (1 H, dd, J1,2 3.5, J2,3 9.8, H-2), 3.63–3.80
(5 H, m, H-5,-5Ј, H2-6Ј and Ha-6), 4.0 (1 H, t, J2,3ϩ3,4 18.5, H-3),
4.07 (1 H, m, Hb-6), 4.55 (1 H, d, J 11.0, CHHC6H5), 4.56 (1 H,
d, J1Ј,2Ј 8.0, H-1Ј), 4.59 (1 H, d, H-1), 4.67 (1 H, d, J 12.0,
CHHC6H5), 4.81 (1 H, d, J 12.0, CHHC6H5), 4.82 (1 H, d, J
10.9, CHHC6H5), 4.89 (1 H, d, J 11.0, CHHC6H5), 4.99 (1 H, d,
J 10.9, CHHC6H5), 5.00 (1 H, t, J3Ј,4Јϩ4Ј,5Ј 18.9, H-4Ј), 5.08 (1 H,
t, J2Ј,3Ј 9.8, H-2Ј), 5.23 (1 H, t, J2Ј,3Јϩ3Ј,4Ј 19.0, H-3Ј), 6.55 (1 H,
t, JN H ,6 5.9, NHAbz), 7.0–8.4 (19 H, ArH) and 10.0 (1 H, s,
NHBoc); ES-MS 977.6 [M ϩ Li]ϩ (Found: C, 64.01; H, 6.50; N,
170.4 (C᎐O); δ (CDCl ) 1.55 (9 H, s, But), 2.04, 2.09 and 2.14 (9
᎐
H
3
H, 3 s, 4 × Ac), 3.36 (1 H, dt, J5,6a 5.5, J6a,6b 14.3, J6a,N H 6.5, Ha-6),
3.84 (1 H, ddd, J5,6b 2.8, J6b,N H 6.5, Hb-6), 4.28 (1 H, ddd, J4,5 10.0,
H-5), 4.87 (1 H, dd, J1,2 3.7, J2,3 10.1, H-2), 4.95 (1 H, t, J3,4ϩ4,5
J. Chem. Soc., Perkin Trans. 1, 1997
1373

View Article Online
2.92%. C52H62N2O16 requires M, 970.4; C, 64.36; H, 6.44; N,
2.89%).
ing the CD data was provided by Niels Kaarsholm at NOVO
Nordisk A/S.
N á-(Fluoren-9-ylmethoxycarbonyl)-O-{2,3,4-tri-O-acetyl-6-
[2Ј-(tert-butoxycarbonylamino)benzoylamino]-6-deoxy-â-D-
glucopyranoxy}-L-threonine pentafluorophenyl ester 9
References
1 H. Lis and N. Sharon, Eur. J. Biochem, 1993, 218, 1; H. C. Joao and
R. A. Dwek, Eur. J. Biochem. 1993, 218, 239; J. R. Rasmussen, Curr.
Opin. Struct. Biol., 1992, 2, 682; R. B. Parekh, Curr. Opin. Struct.
Biol., 1991, 1, 750; R. A. Dwek, Chem. Rev., 1996, 96, 683.
2 F. C. Grochee, M. J. Gramer, D. C. Andersen, J. B. Bahr and
J. R. Rasmussen, Frontiers in Bioprocessing II, ed. C. P. Todd,
S. K. Sikdar and M. Bier, ACS-series, Washington, DC, 1992,
p. 199.
3 T. Arakawa and S. N. Timasheff, Methods Enzymol., 1985, 114,
49.
4 K. Olden, J. B. Parent and S. L. White, Biochim. Biophys. Acta, 1982,
650, 209; A. D. Elbein, Methods Enzymol., 1987, 138, 661.
5 M. A. Titus, Curr. Opin. Cell Biol., 1993, 5, 77; A. Columbus, Curr.
Opin. Cell Biol., 1993, 5, 17.
6 K. T. O’Neil and W. F. DeGrado, Science, 1990, 250, 646.
7 L. Otvos, J. Thurin, E. Kollat, L. Urge, H. M. Mantsch and
M. Hollosi, Int. J. Pept. Protein Res., 1991, 38, 476; J. P. Aubert,
N. Helbecque and M. H. Loueheux-Lefebvre, Arch. Biochem.
Biophys., 1981, 208, 20.
8 H. C. Joao, I. G. Scragg and R. A. Dwek, FEBS Lett., 1992, 307,
343; J. T. Davis, S. Hirani, C, Bartlett and B. R Reid, J. Biol. Chem.,
1994, 269, 3331.
9 K. G. Rice, P. Wu, L. Brand and Y. C. Lee, Biochemistry, 1993, 32,
7264; 1991, 30, 6646; B. Imperiali and K. W. Rickert, Proc. Natl.
Acad. Sci. USA, 1995, 92, 97.
10 P. Y. Chou and G. D. Fasman, Annu. Rev. Biochem., 1978, 47, 251;
N. E. Zhou, C. M. Kay and R. S. Hodges, J. Mol. Biol., 1994, 237,
500.
A
mixture of 2,3,4-tri-O-acetyl-6-[2Ј-(tert-butoxycarbonyl-
amino)benzoylamino]-6-deoxy-α/β--glucopyranosyl trichloro-
acetimidate 24 (0.10 g, 0.20 mmol), Nα-(fluoren-9-ylmethoxy-
carbonyl)--threonine pentafluorophenyl ester 16 (0.15 g, 0.30
mmol) and freshly activated 4 Å molecular sieves in anhydrous
CH2Cl2 (2 cm3) was stirred under argon for 1 h. The reaction
mixture was cooled to Ϫ78 ЊC and trimethylsilyl trifluoro-
methanesulfonate (3 mm3, 0.015 mmol) was added. After 0.5 h
the reaction mixture was warmed to room temperature and was
stirred for 1 h before being recooled to Ϫ78 ЊC, quenched with
1 drop of triethylamine, filtered, concentrated and subjected to
column chromatography on a short column of pre-dried silica
gel with hexane–ethyl acetate (1:1) as eluent (Rf 0.4). The title
compound 9 was obtained as a foam (0.09 g, 60%), [α]2D0 Ϫ27.3
(c 3.3, CH2Cl2); δC(CDCl3) 17.1 (Thrg), 20.6 and 20.7 (COCH3),
28.3 [C(CH3)3], 39.9 (C-6), 47.1 (Fmocβ), 58.7 (Thrα), 67.2
(Fmocα), 69.4 (C-2), 71.35 (C-4), 72.1 (C-3), 72.9 (C-5), 75.1
(Thrβ), 80.2 [C(CH3)3], 99.3 (C-1), 119.5–144.0 (Ar), 152.9
[COC(CH3)3], 156.2 (CONH), 167.0 [C(O)OPfp] and 168.8,
β
g
169.2, 169.8 and 170.1 (C᎐O); δ (CDCl ) 1.33 (3 H, d, J
᎐
H
3
Thr ,Thr
6.4, Thrg), 1.56 (9 H, s, But), 2.07, 2.09 and 2.16 (9 H, 3 s,
3 × Ac), 3.07 (1 H, m, Ha-6), 3.67 (1 H, m, H-5), 3.92 (1 H, m,
Hb-6), 4.25 (1 H, t, JF moc ,Fmoc 6.0, Fmocβ), 4.47 (1 H, m,
Thrβ), 4.52 (2 H, m, Fmocα), 4.60 (1 H, d, J1,2 8.0, H-1), 4.75 (1
α
β
11 S. Y. M. Lau, A. K. Taneja and R. S. Hodges, J. Biol. Chem., 1984,
259, 13 253.
12 V. T. Förster, Ann. Phys., 1948, 6, 55.
13 P. Well and L. Brand, Anal. Biochem., 1994, 218, 1.
14 M. Meldal and K. Breddam, Anal. Biochem., 1991, 195, 141.
15 J. Ø. Duus, J. Winkler and M. Meldal, manuscript in preparation.
16 R. L. Whistler, L. W. Doner and M. Kosik, Methods Carbohydr.
Chem., 1972, 6, 411.
H, dd, JThr ,Thr 3.0, JThr ,N H 8.4, Thrα), 4.95 (1 H, t, J1,2ϩ2,3 19.5,
H-2), 4.99 (1 H, t, J3,4ϩ4,5 17.3, H-4), 5.27 (1 H, t, J2,3ϩ3,4 19.1, H-
α
β
α
3), 5.82 [1 H, d, JN H Thr 8.4, NHThr], 6.67 (1 H, br t, JN H ,6 5.8,
α
NHAbz), 6.8–8.4 (12 H, ArH) and 10.15 (1 H, s, NHBoc); ES-MS
1020.3 [M ϩ Li]ϩ (C39H48F5N3O15 requires M, 1013.9).
17 M. G. Ambrose and R. W. Binkley, J. Org. Chem., 1983, 48, 674.
18 Z. Gyorgydeak and L. Szilagyi, Leibigs Ann. Chem., 1987, 235.
19 R. R. Schmidt, Angew. Chem., Int. Ed. Engl., 1986, 25, 212.
20 G. Excoffier, D. Gagnaire and J.-P. Utille, Carbohydr. Res., 1975, 39,
368.
21 P. J. Garegg, T. Iverson and S. Oscarson, Carbohydr. Res., 1976, 50,
c12.
22 V. Pozgay and H. Jennings, J. Org. Chem., 1987, 52, 4635; 1988, 53,
4042.
23 M. Meldal, Tetrahedron Lett., 1992, 33, 3077.
24 A. Dryland and R. C. Sheppard, Tetrahedron, 1988, 44, 859;
M. Meldal, A. Holm and O. Buchardt, PCT Int.Appl. WO 90 07,975
(Chem. Abstr., 1991, 114, 82556v); L. R. Cameron, J. L. Holder,
M. Meldal and R. C. Sheppard, J. Chem. Soc., Perkin Trans. 1, 1988,
2895.
25 H. Rink, Tetrahedron Lett., 1987, 28, 3787.
26 R. Knorr, A. Trzeciak, W. Bannwarth and D. Gillessen, Tetrahedron
Lett., 1989, 30, 1927.
27 E. Atherton, L. R. Cammeron and R. C. Sheppard, Tetrahedron,
1988, 44, 843.
28 M. Meldal and K. J. Jensen, J. Chem. Soc., Chem. Commun., 1990,
438.
29 E. Meinjohanns, M. Meldal, T. Jensen, O. Werdelin, L. Galli-
Stampino, S. Mouritsen and K. Bock, J. Chem. Soc., Perkin Trans. 1,
1997, 871.
30 I. Christiansen-Brams, M. Meldal and K. Bock, J. Chem. Soc.,
Perkin Trans. 1, 1993, 1461.
31 K. Dax, W. Wolflehner and H. Weidmann, Carbohydr. Res., 1978,
65, 132.
32 T. Utamura, K. Kuromatsu, K. Suwa, K. Koizumi and T. Shingu,
Chem. Pharm. Bull., 1986, 34, 2341.
Fluorescence experiments
Stock solutions of compounds 1–5 were prepared in 50 mmol
TRIS chloride buffer at pH 8.5, to give substrate concentrations
of ~500 µM. For urea-denaturation studies, a series of ali-
quots of ~10 mm3 of the stock solution were transferred into
vials containing solutions of urea in TRIS buffer (0, 1.0, 2.0,
2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 6.0 and 7.0 M) to make the final
volume equal to 1 cm3, and the final concentration of the
peptide equal to 5.0 µM. The solutions were allowed to
equilibrate for 2 h at 25ЊC before fluorescence measurements
were commenced.
For the comparison of the rate of heterodimer formation of
peptide 2 with the peptides and glycopeptides 1 and 3–5 in
TRIS and 2 M urea, ~10 mm3 of the stock solution of com-
pound 2 (500 µM) and ~10 mm3 of the stock solutions of one of
the compounds 1, 3–5 (500 µM) was added to an appropriate
volume of either TRIS or 2 M urea in a cuvette to make the
final volume 1 cm3 and a final concentration of the com-
pounds equal to 5.0 µM. Fluorescence measurements were
commenced 30 s after mixing of any two substrates.
CD experiments
For the CD spectra, solutions of the peptide 1–7 were prepared
in degassed water (~0.5 mg cmϪ3) and concentrations were
determined accurately by UV spectroscopy. For the denatur-
ation studies stock solutions of the substrates 1, 4 and 5 were
prepared in 50 mM TRIS perchlorate buffer at pH 8.5 at con-
centrations of 120 µM.
Acknowledgements
Paper 6/07624F
Received 8th November 1996
Accepted 20th January 1997
Amino acid analysis was performed by Dr Ib Svendsen, and
Dr Anita M. Jansson performed the ES-MS. Support in obtain-
1374
J. Chem. Soc., Perkin Trans. 1, 1997