Synthesis of photopolymerizable glycolipids and their application as scaffolds to immobilize proteins with a micron-sized pattern

Article abstract of DOI:10.1071/CH02182

Photopolymerizable glycolipids incorporating ceramide- or amido-type linkers and able to form stable monolayers were efficiently synthesized by chemical and enzymatic methods. Glycolipid polymer films served as platforms for the immobilization of proteins

Full text of DOI:10.1071/CH02182

Full Papers
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2003, 56, 567–576
Synthesis of Photopolymerizable Glycolipids and their Application as
Scaffolds to Immobilize Proteins with a Micron-Sized Pattern
Noriko Nagahori,A Kenichi Niikura,B Reiko Sadamoto,B Kenji MondeA and
Shin-Ichiro NishimuraA,C
A Laboratory for Bio-Macromolecular Chemistry, Division of Biological Sciences,
Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan.
B Sapporo Laboratory for Glycocluster Project, Japan Bioindustry Association, Sapporo 060-0810, Japan.
C Author to whom correspondence should be addressed (e-mail: shin@glyco.sci.hokudai.ac.jp).
Photopolymerizable glycolipids incorporating ceramide- or amido-type linkers and able to form stable monolayers
were efficiently synthesized by chemical and enzymatic methods. Glycolipid polymer films served as platforms
for the immobilization of proteins through specific carbohydrate–protein interactions at the air–water interface.
Carbohydrate-binding proteins deposited on the glycolipid film were observed by atomic force microscopy, which
showed varying submicron-sized protein patterns such as dendrites, dots, and networks, depending on the lipid
structure, membrane preparation process, and sugar density of the membrane. Surface plasmon resonance mea-
surement confirmed that the subunit-type lectins immobilized on the glycolipid membranes exhibited the ability to
interact specifically with carbohydrate ligands by using unoccupied binding sites.
Manuscript received: 4 September 2002.
Final version: 28 January 2003.
Introduction
characterization of protein–carbohydrate interactions.[26–28]
The development of carbohydrate biosensors and arrays
would enable high-throughput analysis of carbohydrate–
protein and carbohydrate–carbohydrate recognition, which is
useful for biomedical research and clinical diagnostics.
In this study, our aim was to assemble proteins on the
surface of glycolipid films for future applications to the con-
struction of ordered protein microarrays such as cell-adhesive
substrates and sensors. We utilized glycolipid monolayers at
the air–water interface and used them as scaffolds to display
proteins in a variety of ordered arrangements.
Langmuir–Blodgett (LB) membranes of functional lipids
on a solid support are usually unstable in air, due to ‘flip-
flopping’ of the lipid molecules inducing disruption of
the ordered membrane structure. Therefore, a photopoly-
merizable diacetylene group was introduced in the lipid
moiety of the glycolipids to prevent this undesired lipid
movement[29–31] (Scheme 1).
For further practical applications of these protein arrays,
the immobilized proteins must be oriented so that their active
sites are available. Protein-immobilization methods using
chemicals to crosslink them with supporting materials[32]
are not site-specific and align the proteins in a variety
of orientations, in which functional sites may be hidden.
In our method, the proteins are attached to the mem-
brane via specific carbohydrate–protein interactions, thus
the proteins are expected to share the same orientation
Enhanced binding of carbohydrates and proteins is often
achieved through the introduction of multivalency to gly-
coconjugates, and this phenomenon has been known as
the ‘glycoside cluster effect’.[1–3] We have reported that
multivalent sugar side-chains displayed on water-soluble
polymers exhibit high affinity to guest proteins.[4–6] In addi-
tion, synthetic glycopolymers are also excellent acceptors of
glycosyltransferases,[7,8] which have been used in our previ-
ously developed enzyme-assisted glycosynthesizer.[9,10]
Multivalent recognition is also employed for amplifying
individual weak carbohydrate–protein interactions to attain
strong, specific recognitions in biological systems.[11–13]
Especially on the cell surface, receptors aggregate to form
ordered oligomeric structures,[14,15] and glycolipids usually
cluster and localize to form microdomains through which
various recognition events occur.[16,17] In artificial lipid
membranes such as monolayers, glycolipids often produce
domains with highly ordered and patterned morphologies on
a sub-micron scale due to phase separation.[18–20] Recently,
besides conventional lithographic technology, it was found
that synthetic polymers can be micropatterned on a solid
support by self-assembly or self-organization.[21,22] How-
ever, the techniques for patterning biomolecules such as
proteins and carbohydrates on the micro- or nano-scale have
not yet been fully explored, except for the recent studies
on protein arrays[23–25] and on carbohydrate arrays for the
© CSIRO 2003
10.1071/CH02182
0004-9425/03/060567

568
S.-I. Nishimura et al.
OH
O
R
HN
Natural glycosphingolipid
O
i
SCDase
OH
O
R
Lyso-glycosphingolipid
NH2
SCDase
PDA (1)
ii
OH
O
O
R
Scheme 1. Topochemical polymerization of diacetylene-containing
glycolipids at the interface by UV irradiation.
HN
HO
and to exhibit satisfactory binding activity for further
functionalization.
O
(1)
OH
R=
HO
HO3SO
O
Results and Discussion
OH
HO
HO
Synthesis of Photopolymerizable Glycolipids
OH
(2)
OH
O
HO
O
O
We employed enzymatic modification of natural glycolipids
as a rapid entry to prepare a series of polymerizable glyco-
lipids, because the chemical assembly of the carbohydrate
is often a time-consuming process. Firstly, we designed and
synthesized diacetylene-containing ceramide-type glyco-
lipids using an enzyme-assisted technique. Next, we prepared
the glycosphingolipid mimetics containing two amido groups
in the lipid moiety by chemical synthesis.
O
OH
NHAc
COOH
HO
OH
HO
O
HO
O
OH
O
O
O
HO
AcHN
OH
OH
HO
(3)
HO
HO
O
OH
O
OH
O
HOOC
OH
O
O
OH
O
OH
O
AcHN
OH
HO
HO
O
O
Enzymatic Synthesis of Diacetylene-Containing
Glycolipids
HOOC
O
HO
OH
OH
O
HO
HO
OH
(4)
AcHN
Scheme 2 shows the synthetic route to ceramide-type glycol-
ipids. Natural glycolipids, such as sulfatide (2), ganglioside
GM1 (3), and GD1a (4), were converted to the corre-
sponding diacetylene-containing glycolipids using sphin-
golipid ceramide N-deacylase (SCDase).[32] The enzyme
catalyzes hydrolysis and re-condensation of the amide moi-
ety according to the pH and surfactant concentration in the
reaction mixtures. The hydrolysis reactions were performed
in an aqueous buffer containing 0.8%Triton X-100 at pH 6.0.
Then, to the reaction mixture, n-decane was added to remove
the released fatty acid. The resultant lyso form was coupled
with 10,12-pentacosadiynoic acid (1) in a phosphate buffer
(pH 7.0) containing 0.1% surfactant at 37◦C. The reaction
mixture was purified with a reverse phase column (SM-2
beads) to remove detergent, followed by silica gel (IATRO
beads) chromatography. The average yields of the hydrolysis
(step i) and condensation (step ii) reactions were 50% and
70%, respectively.
HO
AcHN
OH
Scheme 2. Synthesis of ceramide-type glycolipids.
Chemical Synthesis of Diacetylene-Containing
Glycolipids
Ceramide-mimetic glycolipids of galactose (5) and N-
acetyl-glucosamine (6) were synthesized (Scheme 3). The
two amide residues in the compounds were expected to
enhance self-assembly by their intermolecular hydrogen
bonds. Although we tried to synthesize ceramide-type glyco-
lipids carrying maltotriose, the intermediate showed poor
solubility in organic solvents. Thus, we synthesized single-
alkyl-chain glycolipids carrying maltotriose, (12) and (13),
Scheme 4. Synthetic details are described in the Experimental
section.

Photopolymerizable Glycolipids as Scaffolds to Immobilize Proteins
569
O
OH
O
N
O
R
OH
HO
HO
H
HN
O
HO
OH
O
HO
O
O
O
HO
O
R
3
HO
HO
OH
OH
HO
HO
O
R=
HO
O
HO
HO
R3=
O
OH
NHAc
(5)
(6)
N
H
(12)
(13)
COOH
Cbz
ϩ
H2N
HN
EEDQ
OAc
O
EtOH-benzene
O
OAc
AcO
AcO
HO
N
H
O
AcO
OAc
O
O
AcO
(7)
HN
S
AcO
O
Cbz
(14)
AcO
OAc
AcO
AcO
AcO
(7)
O
O
CCl3
NH
BF3·OEt2 / CH2Cl2
OAc
NHCbz
HO
NIS, TfOH / CH2Cl2
OAc
O
AcO
AcO
O
OAc
N
O
O
H
OAc
AcO
AcO
HN
OAc
Cbz
O
i) NaOMe / MeOH
ii) Pd-C / MeOH
AcO
OAc
O
(8)
AcO
O
NHCbz
O
AcO
O
AcO
(15)
OH
O
HO
HO
O
AcO
N
O
H
OH
OH
NH2
(9)
O
PDA (1)
HO
HO
(5)
OH
O
HO
EEDQ / CHCl3-MeOH
O
HO
NaOMe / MeOH
Pd-C / MeOH
OH
OAc
O
HO
O
O
NH2
O
(7)
HO
S
AcO
AcO
HO
NIS, TfOH / CH2Cl2
PDA
NHTroc
OAc
(12)
WSC / MeOH
O
O
N
H
O
AcO
AcO
HN
NHTroc
(14)
+
HO
Cbz
Zn powder
(10)
/ THF-Ac2O-AcOH
NIS, TfOH / CH2Cl2
OAc
OAc
O
O
O
AcO
AcO
OAc
O
N
O
AcO
AcO
AcO
O
AcO
H
HN
NHAc
OAc
O
NaOMe / MeOH
Pd-C / MeOH
Cbz
AcO
(11)
O
O
AcO
(16)
OH
O
AcO
O
HO
HO
N
H
O
NaOMe / MeOH
(13)
Scheme 4. Synthesis of maltotriose-type glycolipids.
NHAc
PDA (1)
(6)
NH2
EDC / MeOH
Scheme 3. Synthesis of ceramide-mimetic glycolipids.

570
S.-I. Nishimura et al.
Characterization of the Glycolipid Monolayers
Surface pressure–area isotherms (π–A) of monolayers con-
sisting of varying ratios of the glycolipid (5) and the matrix
lipid (1) are shown in Figure 1. The limiting area of com-
pound (5) is estimated to be about 0.7 nm2 per molecule,
which is more than twice that (about 0.23 nm2) of the matrix
lipid (1).
The monolayer surface was irradiated with ultraviolet light
(254 nm, 8W) from a distance of 12 cm.The surface pressure
was kept at 30 mN m−1 during the UV polymerization. The
monolayer was transferred onto a hydrophobic glass plate
with the carbohydrate moiety displayed on the surface. Poly-
merization was confirmed by the increase in absorbance at
400–700 nm. Typical absorbance spectra of the glycolipid
LB film before and after photopolymerization are shown in
Figure 2. Before photopolymerization, the LB film of glyco-
lipid (5) and matrix lipid (1) did not show absorbance, while
the polymerized LB films exhibited optimal absorbance at
around 530 nm (the LB film of (5), red phase), and 650 and
540 nm (the LB film of (1), blue phase).
Fig. 2. Absorption spectra of the LB film before and after polymeriza-
tion. Red solid line, polymerized LB film of glycolipid (5); Red dashed
line, LB film of (5) before polymerization; Blue solid line, polymerized
LB film of matrix lipid (1); Blue dashed line, LB film of (1) before
polymerization.
A red-phase polydiacetylene film is reported to have an
expanded limiting area and larger tilt angle than that of blue-
phase film.[33] Our π–A data and absorbance spectra also
suggest that the glycolipid (5) has a larger tilt angle than that
of the lipid (1).
(a) Procedure 1
(i)
(ii)
(iii)
Adsorption of PNA Lectin onto the Surface of
Galactose-Displayed Membranes
(b)
(c)
Figure 3a illustrates the preparation by Procedure 1
(see Experimental section) of the microarray of lectin, a
(d)
(e)
GalCer
PDA (1)
(matrix)
GalCer (%)
(PDA)
(f )
Fig. 3. Adsorption of PNA on the glycolipid LB film. (a) (i) Photo-
polymerization of the mixed glycolipid monolayer; (ii) injection of PNA
into the subphase for binding to the monolayer; (iii) transfer of the
monolayer bound with PNA onto a hydrophobic glass plate. Light brown
ball represents the matrix lipid; green hexagon the glycolipid; orange
ball the PNA. AFM images of PNA-adsorbed glycolipid LB film are at
concentrations of (5)/(1) of (b) 0, (c) 10, (d) 50, and (e) 100% mol/mol.
The scale is 10 µm × 10 µm. (f )The height of the patterned PNA on the
LB film of (c) by cross-section analysis. The scale is 2.5 µm × 2.5 µm.
Area (nm2)
Fig. 1. Surface pressure–area isotherms of mixtures of glycolipid (5)
and matrix lipid (1). The density of glycolipid (5) in the monolayer was
0, 20, 80, or 100% (mol/mol), as shown.

Photopolymerizable Glycolipids as Scaffolds to Immobilize Proteins
571
carbohydrate-recognizing protein, on the film. The mixed
glycolipid monolayers (5)/(1) were photopolymerized on the
subphase of 50 mM phosphate buffer (pH 7.4). A solution
of peanut lectin (PNA), which has four binding sites recog-
nizing a terminal galactose residue,[34] was injected into the
subphase.The protein-bound monolayer was transferred onto
a hydrophobic glass plate.
(a)
Figures 3b–3e shows AFM images of the PNA on the gly-
colipid LB film prepared following Procedure 1 using various
ratios of the lipids. When the PNA was added to the mem-
brane in the absence of the glycolipid (100% matrix lipid
(1), Fig. 3b), the amount of protein binding was smaller than
those on the sugar-displaying membranes. Interestingly, PNA
was arrayed in a dendritic pattern on the glycolipid film con-
taining 10% GalCer (5) (Fig. 3c), whereas the non-specific
protein binding to the matrix film did not show any specific
micropattern.
(b)
When the ratio of glycolipid in the LB film was
50 or 100%, no obvious dendritic pattern was observed
(Figs. 3d, 3e). The formation of the protein pattern depended
on the density of the glycolipid in the membrane and might
be mainly due to the segregation of the glycolipid and matrix
lipid. In Figure 3e, the protein did not show an homogeneous
pattern, although the glass surface is covered with 100% gly-
colipid. In our previous reports,[4,5] we showed the amount
of glycopolymer binding to lectins in solution is also reduced
when the carbohydrate ligands on the polymer backbone are
too crowded. We considered this to be due to steric hindrance
of the sugar moiety.
Time (s)
Fig. 4. The binding profiles of PNA to the glycolipid LB film mea-
sured by SPR. (a) A cartoon of the process. (b) Sensor response upon
injection of PNA to films of different glycolipid densities. The solution
of PNA (1 µM in the buffer) was injected at point A and changed to a
dissociation phase at point B.
Figure 3f shows AFM height information for the pat-
tern on the glycolipid LB film of 10% GalCer (Fig. 3c,
2.5 µm × 2.5µm). The height (about 10 nm) corresponds to
the diameter of a single molecule of PNA, which was revealed
from the crystal structure.[35] The width of the protein assem-
bly was 50–100 nm in average, indicating that the proteins
formed clusters of 5–10 molecules.
Among the AFM images of the adsorbed proteins, some
(Figs. 3b, 3c, 6b, 6e, and 7c) showed one-way alignment.
In this experiment, the LB films were deposited on the plate
vertically, and the alignment of protein patterns corresponded
to the dipping direction of the LB film. Thus, we considered
that the alignment was due to the fluidity of the polymerized
monolayer.
The amount of PNA binding onto the glycolipid LB films
was determined quantitatively by surface plasmon resonance
(SPR). The polymerized films prepared from different ratios
of (5) and (1) were transferred onto the surface of the
SPR sensor chips carrying alkanethiol (n-octadecyl mercap-
tan) monolayers. This modified sensor chip was set in the
SPR instrument, then a buffer solution containing PNA was
injected into the flow cell. As shown in Figure 4, the amount
of adsorption of PNA depended on the density of (5) in the
film. Only a small amount of protein was adsorbed to the
matrix lipid (1) in the film not containing glycolipid (GalCer
0%).The amount of PNA binding (represented as a resonance
unit (RU) value) increased with increases in the ratio of (5),
indicating that PNA was immobilized on the glycolipid LB
films in a carbohydrate-dependent manner.
We estimated the contribution of nonspecific binding to
the films using bovine serum albumin (BSA), which is a
protein having no affinity for the sugar. A BSA solution
(1 µM in a 10 mM Tris buffer) was added to the SPR sen-
sor chip coated with matrix lipid or GalCer lipid monolayer.
The maximum intensity of signals after injecting BSA onto
the surface of the matrix lipid was as small as approximately
100 RU (data not shown), and about 80 RU for the GalCer
lipid monolayer. Our data indicates that the galactose moi-
ety on the surface reduced the non-specific absorption of
proteins driven by hydrophobic interactions. Therefore, the
galactose concentration-dependent increase in the amount of
PNA binding (Fig. 4b) indicates that the force for binding the
PNA to the membrane was the specific interaction with the
sugar.
Protein Immobilization on the Glycolipid Assembly
Protein patterns formed by several combinations of sug-
ars and proteins were examined using the same procedure
(Fig. 5). The surface of the glycolipid LB film before pro-
tein binding did not show any specific structure (Fig. 5a).
When l-selectin–immunoglobulin (l-selectin–IgG) complex
was bound on the film of 1 : 9 sulfatide (2)[36,37]/matrix lipid
(1), the morphology of the protein pattern was a long-range
ordered dendrite (Fig. 5b). Cholera Toxin-B subunit (CTX-
B) on the LB film of 1 : 9 ganglioside GM1 (3)/phospholipid
(20), Diagram 1, gave a mottled pattern (Fig. 5c). In both

572
S.-I. Nishimura et al.
(a)
(a)
Procedure 2
(b)
(c)
(b)
(c)
Fig. 6. Protein microarrays prepared following Procedure 2. Light
brown ball, matrix lipid; green hexagon, glycolipid; orange ball, PNA.
(a) (i) Polymerizable glycolipid and protein were incubated, then the
complex and lipid (1) were compressed to form a monolayer; (ii)
photopolymerization of the monolayer; (iii) transfer of the complex
monolayer with protein to a hydrophobic glass plate. AFM images of
(b) l-selectin–IgG on (2)/(1), 1 : 9 and (c) CTX-B on (3)/(20), 1 : 9.
(d)
(e)
procedure for immobilizing proteins on the surface of a
glycolipid membrane. As in Procedure 1, the mixed-lipid
solution was first spread onto an aqueous subphase. How-
ever, in this procedure, a concentrated protein solution was
injected into the subphase before compression of the mono-
layer. After compression of the protein–glycolipid complex,
the monolayer was photopolymerized by UV irradiation.
l-Selectin–IgG and CTX-B were employed for the immo-
bilization experiments to allow comparison with the data
shown in Figures 5b and 5c. Interestingly, the pattern of l-
selectin adsorbed on the LB film exhibited a long-range net
(Fig. 6b) and CTX-B formed a small dot pattern (Fig. 6c).
The morphology of these protein patterns produced by Proce-
dure 2 markedly differed from those produced by Procedure 1
(Figs. 5b, 5c). In the case of Procedure 2, proteins injected
into the subphase would aggregate on the flexible mono-
layer surface after binding with the glycolipids during the
monolayer compression process. This step seems to account
for the significant difference in the formation of the protein
patterns. The result suggests that the present method might
aid understanding of the mechanism of domain formation on
biomembranes.
Fig. 5. AFM images of the proteins displayed on glycolipid LB films
prepared following Procedure 1. (a) Glycolipid LB film of (2)/(1), 1 : 9.
(b) l-Selectin–IgG on a film of (2)/(1), 1 : 9. (c) CTX-B on a film of
(3)/(20), 1 : 9. (d) MBP on a film of (12)/(1), 1 : 9. (e) MBP on a film of
(13)/(1), 1 : 9. Scale bars 1 µm.
O
O
O
Nϩ
O
P
OϪ
O
O
O
Diagram 1.
cases, when the glycolipid density was increased to 50%, no
obvious patterns were observed.
Theeffectoftheglycolipidstructureonproteinpatternwas
examinedusingmaltotriose-carryinglipids(12)and(13).The
pattern of maltose-binding protein (MBP) bound onto the LB
film of lipids 1 : 9 (12)/(1) was found to be dots (Fig. 5d), and
MBP on the LB film of lipids 1 : 9 (13)/(1) was ramiform,
terminating in small dots (Fig. 5e). The difference between
lipids (12) and (13) is an amide bond in the linker moiety of
the glycolipid. Thus, it was suggested that the morphology
of the protein patterns bound on the glycolipid domains was
determined by the chemical structure of the lipids.
Molecular Recognition by Lectin Microarray
The micrometre-scale dendritic patterns of lectins revealed
by AFM prompted us to apply these lectin microarrays to the
development of patterned cell-adhesive substrates or biosen-
sor chips. For this purpose, it is essential that the patterned
lectins on the LB film retain their sugar-binding activity,
which is dependent on the specific orientation of the bind-
ing sites. The carbohydrate-binding activity of the dendrite
PNA on the glycolipid LB film (10% (5) in (1), Fig. 3c)
was examined by SPR measurements (Fig. 7). The lectin was
immobilized on the SPR sensor chip according to Procedure 1
When the protein was bound to glycolipid prior to the
lipid segregation in the membrane (Procedure 2), the pro-
tein patterns were different. Figure 6a outlines the alternative

Photopolymerizable Glycolipids as Scaffolds to Immobilize Proteins
573
of the polymer was effectively inhibited by excess lactose
(black line). The result means that the interaction is based
on the specific carbohydrate–protein interaction, and the
polymer did not remove PNA from the LB surface under
these experimental conditions. The binding constant between
the polymer (19) and immobilized PNA was calculated to
be 5 × 106 M−1 by varying the concentration as previously
reported.[38,39] This value is approximately 100-fold higher
than the interaction of monovalent carbohydrate and a single
lectin.
(a)
Conclusion
(b)
Several types of glycolipids having a photopolymerizable
diacetylene moiety were designed and synthesized by chem-
ical and enzymatic methods. Photopolymerization of the
mixed glycolipid membrane provided a platform on which
carbohydrate-binding proteins were bound via specific inter-
actions. The proteins bound on the glycolipid LB films
clustered to form a protein monolayer with a variety of
ordered patterns. The carbohydrate-binding activity of the
immobilized protein was confirmed by SPR. It was sug-
gested that the procedure provided a practical method for
protein immobilization on a flat solid support while main-
taining activity, which could in time lead to the development
of a novel class of biosensors and cell-adhesive substrates.
Time (s)
Fig. 7. Binding of lactose-appending polymer to the PNA-
immobilized LB film. (a) Illustration of the experimental outline using
SPR. (b) SPR sensorgrams upon addition of glycopolymer (19) (1 µM)
alone (red line) or in the presence of 5 mM lactose (black line). Samples
were injected at point A and the sensor chip was washed with buffer at
point B.
Experimental
General Procedures
1H NMR spectra were recorded at 600 MHz on an AVANCE 600
(Bruker) in CDCl3, D2O, CD3OD, or [D6]DMSO. Tetramethylsilane,
methanol (3.3 ppm for 1H and 49.0 ppm for 13C), and chloroform
(77.0 ppm for 13C) were used as internal standards. Molecular weights
were determined by using JMS-HX110 (JEOL) for FAB-MS and Reflex
(Bruker) for TOF-MS. Elemental analyses were performed with a Sar-
torius 4503. Reactions were monitored by thin-layer chromatography
(TLC) on precoated plates of silica gel 60 F254 (layer thickness 0.25
mm, Merck). For detection of compounds,TLC sheets were sprayed with
(a) a solution of 85 : 10 : 5 v/v/v methanol/sulfuric acid/p-anisaldehyde
followed by a few minutes heating (for carbohydrates), (b) a solution
of 5 wt% ninhydrin in ethanol followed by a few minutes heating (for
OH
OH
O
O
OH
HO
O
O
N
H
HO
O
O
OH
HN
OH
N
H
1
O
(19)
H2N
O
amino groups), or (c) an aqueous solution of 5 wt% KMnO4 (for C
C
15
bonds). Column chromatography was performed on silica gel (silica
gel 60N or IATRO 6RS-8060 beads, purchased from Kanto Chemical
or Iatron Laboratories, respectively). Lipid monolayers were prepared
and π–A isotherms were recorded on an LB trough (U.S.I., Kyushu,
Japan). A UVG-54 lamp (UVP) was used for the photopolymeriza-
tion of the monolayer. AFM measurements were carried out at 23◦C
on a Nanoscope III (Digital Instruments) with tapping mode in air.
The imaging was performed using silicon cantilevers (NanoDevices)
with a resonance frequency of 300–400 Hz and a spring constant of 20–
70 N m−1.Absorbance of the polydiacetylene films was measured with a
U-3010 spectrophotometer (Hitachi). SPR measurement was performed
on a Biacore-X (Biacore AB.)
Diagram 2.
by using a coated (n-octadecyl mercaptan) SPR sensor chip
instead of a hydrophobic glass plate. In this system, it was
assumed that one or two binding sites of the four sites of PNA
were occupied by the glycolipids on the LB film and that the
othersiteswereunoccupied. Figure 7b showstheSPRsensor-
gram upon the addition of a lactose-appending glycopolymer
((19), Diagram 2) or the polymer with 5 mM lactose (red line
and black line, respectively). When the polymer (19) (1µM
in 50 mM phosphate buffer, pH 7.4) was added (point A), the
SPRresponseincreasedandtheboundpolymerwasnotdisso-
ciated from the sensor chip by washing with buffer (point B).
On the other hand, when the polymer (19) was injected in the
presence of 5 mM of lactose, the baseline returned to the orig-
inal level after washing with the buffer.Therefore, the binding
Materials
1,2-Bis(10,12-tricosadiynoyl)-sn-glycero-3-phosphocholine (20) was
purchasedfromAvantiPolarLipids. 10,12-Pentacosadiynoicacid(PDA)
(1) was purchased from TCI. Bio-Beads SM-2 Adsorbent (20–50 mesh)
was purchased from Bio-Rad Laboratories. Peanut lectin (PNA) and
AR grade chloroform were purchased from Wako. Cholera Toxin B
subunit (CTX-B) were purchased from Sigma-Aldrich. l-Selectin–IgG
chimera protein was kindly supplied by Dr. H. Kondo (Kanebo, Ltd).

574
S.-I. Nishimura et al.
Maltose-binding protein (MBP) was purchased from New England
Biolabs.
Enzymatic Synthesis of Glycolipids (2)–(4) by SCDase
SCDase (50 µU) was added to a suspension of the natural glycosphin-
golipid in a sodium acetate buffer (pH 6.0) containing 0.8% Triton X-
100. n-Decane was placed on the reaction buffer to extract released fatty
acids. The reaction mixture was incubated at 37◦C for about 24 h under
static conditions. n-Decane was removed, and the residual buffer was
subjected to a column of SM-2 beads, to remove the surfactant, eluted
with 95 : 5 v/v MeOH/H2O. The eluate was purified on a column of
IATRO beads and eluted with CHCl3/MeOH to obtain the lyso form of
the glycosphingolipid. The lyso-glycosphingolipid and PDA were sus-
pended in 50 mM sodium acetate buffer (pH 7.0) containing 0.1%Triton
X-100. SCDase (50 µU) and the mixture were incubated at 37◦C and the
reaction was monitored by TLC. The mixture was dried, suspended in
a small amount of CHCl3, and applied to the IATRO bead column. The
column was eluted with CHCl3/MeOH to obtain crude product fractions
contaminated withTriton X-100. The surfactant was removed by a short
column of SM-2 beads eluted with 95 : 5 v/v MeOH/H2O.
Lipid Monolayers
The surface area-pressure isotherms of the amphiphiles were mea-
sured on a Teflon-coated trough (10 cm × 13 cm) with a Teflon barrier.
Each lipid sample was dissolved in chloroform at a concentration of
0.1–1.0 mM (about 1 mg mL−1) and spread on the subphase at 23◦C.
After 10 min for the evaporation of the organic solvent, the lipids were
compressed at a rate of 0.1 nm2 molec−1 min−1
.
Preparation of Photopolymerized LB Membranes (Procedure 1)
The mixed lipids were dissolved in chloroform at a concentration of
10−4–10−3 M (e.g. 0.53 mM for the mixed lipid of (1) and (5)) and
were spread on the aqueous subphase.The lipid monolayer was prepared
as described and compressed to 30 mN m−1. After the monolayer was
equilibrated for 10 min under the same surface pressure, the membrane
was irradiated with UV light (254 nm, 8W) for 10 min for polymer-
ization. The distance between the lamp and the monolayer surface was
12 cm.A concentrated solution of protein was injected into the subphase
from outside of the barrier to avoid disturbing the monolayer surface.
(2): δH 5.61 (1 H, CH CHCH2), 5.36 (1 H, CH(OH)CH CHCH2),
4.39 (1 H, m, H-4), 4.24 (1 H, d, J 8.0, H-1), 4.18 (1 H, m,
H-3), 4.05 (1 H, t,
CH(NH)CH(OH)), 3.76−3.64 (3 H, m, H-2 and OCH2CH(NH)),
CCH2 and COCH2), 1.92 (2 H, t,
J 8.2, CH(NH)CH(OH)), 3.88 (1 H, m,
2.17−2.07 (6 H, t, CH2C
C C
The final concentration of the protein was adjusted to 10−7 mol L−1
.
CH CHCH2), 0.80 (6 H, t, CH3). δC (CD3OD) 175.4 (NHCO), 134.7
(CH(OH)CH CH), 130.1 (CH(OH)CH = CH), 104.1 (C-1), 81.1 (C-
3), 75.6 (CH(NH)CH(OH)CH CH), 72.3, 70.3, 69.4, 68.0, 65.9
After allowing 30 min for the system to reach equilibrium, a hydropho-
bic glass plate (or an n-octadecyl mercaptan coated SPR sensor chip
for SPR measurement of lactose-appending glycopolymer, described
below) was vertically immersed into the subphase at a dipping rate of
10 mm min−1. The surface pressure was kept at 30 mN m−1 during the
transfer, and the transfer ratio was between 0.9 and 1.0. After removal
of the extra lipids from the air–water interface with an aspirator, the
plate was lifted from the solution. For AFM observation, the surface of
the plate on which the protein was immobilized was rinsed with buffer
solution, pure water (Milli-Q), and dried in air.
(C
C
C
C), 54.0 (CH(NH)CH(OH)), 36.9, 33.0, 32.5, 30.3−28.9
(CH2), 26.5, 23.2 (CH2CH3), 19.5 (CH2C
C C CCH2), 14.3
(CH3). FAB-MS m/z 898.5852 [M + H]− (Calc. for C49H87NO11S2:
898.6084).
(3): δH 5.69−5.41 (2 H, m, CH CH), 4.36 (1 H, d, J 7.81), 4.31
(1 H, d, J 7.67), 4.26 (1 H, d, J 7.81), 4.18 (1 H, dd, J 4.46, 10.2), 4.12
(1 H, s), 3.95 (1 H, m), 3.42 (2 H, dd, J 3.34, 9.85), 2.65 (1 H, dd, J 4.62,
13.2, H-3eq), 2.22 (4 H, t, J 6.6, CH2C
C C CCH2), 2.15 (2 H, t, J
8.4, COCH2), 2.0 and 1.94 (6 H, s, 2 ×AcN), 0.85 (6 H, t, J 6.6, CH3).
TOF-MS m/z 1681.4 [M + H]+ (Calc. for C83H144N3O31: 1680.3).
(4): δH 5.39−5.68 (2 H, m, CH CH), 4.83 (1 H, d, J 8.55, H-1),
4.35 (1 H, d, J 7.8), 4.22 (1 H, d, J 7.8), 4.04 (1 H, t, J 7.8), 2.84
(1 H, d, J 10.8, H-3eq), 2.71 (1 H, d, J 11.2, H-3eq), 2.19 (4 H, t, J
Preparation of Photopolymerized LB Membranes (Procedure 2)
The monolayer was formed after the protein had bound to the glyco-
lipids. A protein solution was added into the subphase after the mixed
lipids were spread on the surface of the subphase.The binding of the pro-
teinswasequilibratedfor30 min.Themixedglycolipidsbearingproteins
at the interface were then compressed to 25 mN m−1 to form a mono-
layer. Next, the complex monolayer was photopolymerized under the
same conditions as those for Procedure 1. The polymerized monolayer
was transferred onto a hydrophobic glass plate for AFM measurements.
7.0, CH2C
C C CCH2), 2.12 (2 H, t, J 7.9, COCH2), 1.98, 1.95,
and 1.90 (9 H, s, 3 × AcN), 0.83 (6 H, t, CH3). TOF-MS m/z 1972.41
[M + Na]+ (Calc. for C93H156N4O39:1953.03).
Chemical Synthesis of Glycolipids
N-Benzyloxycarbonyl-L-serineoctadecylamide (7)
2-Ethyl-1-ethoxycarbonyl-1,2-dihydroquinoline
(EEDQ,
6.82 g,
SPR Measurement of PNA Binding to the LB Films
27.6 mmol) and octadecylamine (7.1 g, 26.3 mmol) were added to
a solution of N-benzyloxycarbonyl-l-serine (6 g, 25 mmol) in 1 : 1
ethanol/benzene (50 mL). The reaction mixture was stirred for 12 h at
room temperature. The solvent was evaporated under reduced pressure
and the residuewas recrystalized from chloroform/ethanol to afford pure
(7) (9.8 g, 85%): δH (CDCl3) 7.32 (5 H, m,ArH), 6.58 (1 H, br s, NHSer),
5.87 (1 H, d, NH), 5.11 (2 H, s, OCH2Ph), 4.16 (1 H, ddd, HαSer), 4.09
(1 H, d, HβSer), 3.64 (1 H, dd, HβSer), 3.21 (2 H, dd, NHCH2), 1.45−1.20
(34 H, 17 × CH2), 0.86 (3 H, t, CH3). TOF-MS m/z 491.343 [M + H]+
(Calc. for C29H50N2O4: 491.3843).
An SIA Kit Au sensor surface (Biacore) was immersed into an ethanol
solution of 1 mM n-octadecyl mercaptan for 12 h. After removal from
the solution, the surface was washed with ethanol and benzene and
dried under a nitrogen stream. The plate was then vertically immersed
in the LB trough through the photopolymerized monolayer formed from
a mixture of lipids (1) and (5) (0 : 10, 2 : 8, 5 : 5, 8 : 2, or 10 : 0 (1)/(5))
while the surface pressure was kept at 30 mN m−1. The transfer ratio of
the film onto the glass plate was 1.0. The extra lipid was removed from
the air–water surface, and the plate with the monolayer was taken from
the subphase, and washed with Milli-Q water. The plate was set on an
SPR sensor chip support for measurements. After the SPR stabilized at
25◦C with an elution buffer at a flow rate of 10 µL min−1, an aqueous
solution of peanut lectin (1 µM) was injected.
N-(Benzyloxycarbonyl)-O-(2,3,4,6-tetra-O-acetyl-
β-D-galactopyranosyl)-L-serineoctadecylamide (8)
Boron trifluoride etherate (17 µL, 0.06 mmol) was added
to
a suspension of 2,3,4,6-tetra-O-acetyl-β-d-galactopyranosyl-
SPR Measurement of Lactose-Appending Glycopolymer
trichloroacetimidate (100 mg, 0.2 mmol), compound (7) (119 mg,
0.24 mmol), and powdered 4 Å molecular sieves in dichloroethane at
0◦C under a nitrogen atmosphere. The reaction mixture was warmed to
room temperature and stirred for 8 h. The mixture was neutralized with
triethylamine and filtrated through celite. The filtrate was washed with
saturated sodium hydrogen carbonate and brine, and dried over magne-
sium sulfate. The mixture was filtrated and concentrated. The residue
was then separated on silica gel with 3 : 1 toluene/ethylacetate as an
PNA microarrays were constructed on an SPR sensor plate coated with
a mixed lipid (1 : 9 (5)/(1)) according to Procedure 1 and kept wet until
the SPR measurement. After the SPR signal stabilized at 25◦C with
an elution buffer (10 mM Tris buffer, pH 7.6, 50 mM NaCl) at flow
rate of 10 µL min−1, a solution of 1 µM lactose-appending acrylamide
polymer[8,9] in the elution buffer was injected with or without 5 mM
lactose.

Photopolymerizable Glycolipids as Scaffolds to Immobilize Proteins
575
eluent to give (8) (67 mg, 40%): δH (CDCl3) 7.28 (5 H, d, ArH), 6.27
(1 H, br s, NHSer), 5.61 (1 H, br s, NH), 5.31 (1 H, d, J 7.04, H-4), 5.13
(1 H, d, J 8.8, 10.4, H-2), 5.04 (2 H, m, OCH2Ph), 4.97 (1 H, dd, J
6.08, 10.5, H-3), 4. 48 (1 H, d, J 6.72, H-1), 4.27 (1 H, d, HαSer), 4.06
(2 H, m, H-5 and H-6a), 3.98 (1 H, dd, HβSer), 3.88 (1 H, s, H-6b), 3.68
(1 H, dd, J 8.24, 10.4, HβSer), 3.16 (2 H, m, CONHCH2), 2.01−1.91
(12 H, 4 × s, 4 × OAc), 1.18 (36 H, m, 18 × CH2), 0.81 (3 H, t, CH3).
δC (CDCl3) 170.3 (NHCO), 127.9−128.5 (Ph), 101.8 (C-1), 70.9, 70.5,
69.8, 68.5, 67.0, 66.9, 61.1, 39.7 (CONHCH2), 31.8, 29.6−29.1 (CH2),
22.5 (CH2CH3), 20.6 (Ac), 20.5 (Ac), 20.4 (Ac), 14.0, (CH3). Anal:
Calc. for C43H68N2O13 H2O: C, 61.56; H, 8.41; N, 3.34. Found: C,
61.33; H, 8.19; N, 3.34.
3 ×Ac), 1.45−1.20 (34 H, m, 17 × CH2), 0.86 (3 H, t, CH3). FAB-MS
m/z 952 [M + H]+ (Calc. for C44H69Cl3O13N3:953).
N-(Benzyloxycarbonyl)-O-(3,4,6-tri-O-acetyl-(2-acetoamide-
2-deoxy)-β-D-glucopyranosyl)-L-serineoctadecylamide (11)
Compound (10) (500 mg) was dissolved in a mixture of 4 : 2 : 1 v/v/v
THF/acetic anhydride/acetic acid (10 mL). Zinc powder activated with
1 M aqueous CuSO4 was added to the solution and the mixture was
stirred for 5 h at room temperature.The solution was filtered and allowed
to evaporate. The residue was purified by silica gel chromatography
eluted with 95 : 5 v/v CHCl3/MeOH to afford solid (11) in a quantitative
yield. δH (CDCl3) 7.28−7.36 (1 H, m, ArH), 6.43 (1 H, bs, NHSer), 5.75
(1 H, d, J 4.73), 5.58 (1 H, d, J 8.82), 5.08 (5 H, m), 4.57 (1 H, d, J
7.57), 4.34 (1 H, bs), 4.25−4.10 (2 H, m), 4.05 (1 H, m), 3.95 (1 H,
m), 3.25−3.15 (5 H, m), 2.03, 2.01, and 1.88 (12 H, 3 × s, 3 ×AcO
and AcN), 1.28−1.22 (34 H, m), 0.86 (3 H, t, CH3). FAB-MS m/z 820
[M + H]+ (Calc. for C43H68N2O13: 820.5).
β-D-Galactopyranosyl-L-serineoctadecylamide (9)
Sodium methoxide (17 mg, 0.31 mmol) was added to a solution of com-
pound (8) (220 mg, 0.27 mmol) in dry methanol (10 mL) and the mixture
was stirred for 5 h at room temperature. The solution was neutralized
with Dowex 50-X8 (H+) resin, filtered, and evaparated in vacuo to afford
an amorphous powder. Subsequently, 20% palladium hydroxide on car-
bon (49 mg) was added to the residue in methanol (15 mL) and stirred for
18 h at room temperature under a hydrogen atmosphere.The mixture was
filtrated through celite and evaporated in vacuo, then subjected to silica
gel chromatography with 65 : 25 : 4 v/v/v chloroform/methanol/water as
an eluent to afford pure (9) (120 mg, 86%): FAB-MS m/z 519.4003 [M
+ H]+ (Calc. for C27H55O7N2: 519.4009).
N-(10,12-pentacosadinoyl)-O-(2-acetoamide-2-deoxy)-
β-D-glucopyranosyl-L-serineoctadecylamide (6)
PDA (100 mg, 0.27 mmol) and 1-ethyl-3-(3ꢀ-dimethylaminopropyl) car-
bodiimide hydrochloride (EDC, 77 mg, 0.4 mmol) were added to a
solution of compound (11) (150 mg, 0.27 mmol) in dry methanol. The
mixture was stirred for 14 h at room temperature, and purified by
silica gel chromatography, eluent chloroform/methanol, to afford (6)
(183 mg, 75%). δH (CD3Cl3/CD3OD, 1 : 1) 4.40 (1 H, t, J 5.4, HαSer),
4.25 (1 H, d, J 8.4), 3.78 (1 H, dd, J 4.8, 10.2), 3.74 (1 H, dd, J 2.4,
12.0), 3.58 (1 H, dd, J 6.6, 10.8), 3.52 (1 H, dd, J 6.0, 12.0), 3.47
(1 H, dd, J 8.4, 10.2), 3.26 (1 H, dd, J 8.4, 10.2), 3.15 (2 H, m), 3.04
N-(10,12-pentacosadiynoyl)-O-β-D-galactopyranosyl-
L-serineoctadecylamide (5)
EEDQ (15.5 mg, 0.063 mmol) and 10,12-pentacosadiynoic acid
(23.5 mg, 0.063 mmol) were added to a solution of (9) (36 mg,
0.052 mmol) in dry methanol (3 mL). The mixture was stirred for 18 h
at room temperature. During the stirring, chloroform (3 mL) was added
to dissolve formed precipitation. Additional EEDQ (15 mg, 0.06 mmol)
was added to the reaction mixture and the mixture was stirred for 20 h.
The mixture was evaporated and chromatographed on silica gel, eluted
with 9 : 1 v/v chloroform/methanol to afford pure (5) (21.5 mg, 47%).
δH ([D6]DMSO) 4.41 (1 H, t, J 5.4 HαSer), 4.10 (1 H, d, J 7.2, H-1),
3.91 (1 H, m, HβSer), 3.67 (1 H, d, J 2.0, H-6a), 3.61 (1 H, dd, J 5.8,
10.6, HβSer), 3.56 (1 H, dd, J 1.4, 6.2, H-4), 3.37 (1 H, dd, J 5.6, 6.8,
H-3), 3.33 (1 H, t, J 7.0, H-2), 3.30 (1 H, dd, J 3.0, 9.4, H-6b), 3.06
(2 H, t, J 7.2, CONHCH2), 2.09 (4 H, t, J 6.6, CH2C
C C CH2),
1.88 (3 H, s, Ac), 0.73 (3 H, t, J 6.6, CH2CH3). δC (CD3Cl3/CD3OD,
1 : 1) 174.3 and 172.9 (NHCO), 170.0 (CH3COHN), 100.7 (C-1), 77.3
(C
C C C), 76.1 (C-5), 75.1 (C-3 and C-4), 52.5 (CαSer), 39.6
(C-2), 36.1 (CONHCH2), 31.8 (NHCOCH2)„ 29.6−28.2, 26.2, 25.5,
22.6 (CH2CH3), 19.0 (CH3CONH), 19.0 (CH2C
C C CCH2), 13.9
(CH3). TOF-MS m/z 916.7 [M + H]+ (Calc. for C54H97N3O8: 916.74.)
Bezyloxycarbonylaminopropyldeca-O-acetyl-
β-D-maltotrioside (15)
N-iodosuccinimide (421 mg, 1.87 mmol) was added to a mixture of
(14) (1.58 g, 1.55 mmol), benzyloxycarbonylamino propanol (390 mg,
1.87 mmol), and powdered 4 Å molecular sieves (220 mg) in anhydrous
CH2Cl2under a nitrogen atmosphere. The mixture was cooled to 0◦C
and then TfOH (50 µL) was added. The reaction mixture was kept for
30 min at 0◦C and then warmed to room temperature. The reaction was
quenched by adding a few drop of TEA, and then the reaction mixture
was diluted with CH2Cl2 and filtered. The filtrate was washed succes-
sively with saturated aqueous solution of Na2SO3, NaHCO3, and brine.
The residue was purified by silica gel chromatography eluted with 5 : 4
to 1 : 1 v/v toluene/EtOAc to give compound (15) (225 mg). δH 7.15
(4 H, m, ArH), 5.39 (1 H, d, J 4.2), 5.38−5.32 (2 H, m), 5.25−5.20
(2 H, m), 5.09−5.03 (3 H, m), 4.83 (1 H, dd, J 4.2, 10.2), 4.74 (1 H, dd,
J 8.4), 4.71 (1 H, dd, J 3.6, 10.2), 4.49 (1 H, m), 4.45 (1 H, t, J 10.8),
4.22 (1 H, dd, J 3.6, 12.6), 4.15 (1 H, dd, J 3.0), 4.02 (1 H, dd, J 1.8),
3.95−3.86 (5 H, m), 2.15−1.96 (30 H, 10 × s). TOF-MS m/z 1138.2 [M
+ Na]+ (Calc. for C49H65NO28: 1138.36).
(2 H, t, J 6.8, NHCH2), 2.21 (4 H, t, J 6.8, CH2C
C C CCH2),
2.15 (2 H, dd, COCH2), 1.46−1.09 (64 H, m, 32 × CH2), 0.83 (6 H, t,
CH3). δC ([D6]DMSO) 174.2 and 170.0 (NHCO), 103.7 (C-1), 77.9
(C
65.6 (C
C
C
C
C), 75.7 (C-3), 73.6, 71.0 (C-2), 69.7 (CβSer), 68.6 (C-6),
C), 61.0 (C-4), 35.6 (CONHCH2), 31.6 (NHCOCH2),
CCH2), 13.8
C
29.4−28.1, 25.4, 22.3 (CH2CH3), 18.6 (CH2C
C C
(CH3). Anal: Calc. for C52H94O8N2 H2O: C, 69.91; H, 10.83; N, 3.14.
Found: C, 70.48; H, 10.70; N, 3.08. FAB-MS m/z 875.7075 [M + H]+
(Calc. for C52H94O8N2: 875.7083).
N-(Benzyloxycarbonyl)-O-(3,4,6-tri-O-acetyl-2-deoxy-2-
(2,2,2-trichloroethoxylcarbonylamino)-β-D-glucopyranosyl)-
L-serineoctadecylamide (10)
TfOH (20 µL) was added to
a mixture of p-methylphenyl
3,4,6- tri-O- acetyl-2-deoxy-2-(2,2,2-trichloroethoxylcarbonylamino)-
1- thioglucopyranoside (600 mg, 1.02 mmol), compound (7) (400 mg,
0.82 mmol), N-iodosuccinimide (400 mg, 1.78 mmol), and molecular
sieves in anhydrous CH2Cl2 at −20◦C. The mixture was stirred for
20 min and warmed to room temperature.The reaction was quenched by
adding a few drops of triethylamine (TEA) and the reaction mixture was
then diluted with CH2Cl2. The mixture was filtered and washed succes-
sively with saturated aqueous solution of Na2SO3, NaHCO3, and brine,
andthenthesolventswereallowedtoevaporate.Theresiduewaspurified
by silica gel chromatography eluted with 9.5 : 0. 5 v/v CHCl3/MeOH to
afford solid (10) (603 mg, 90%). δH (CDCl3) 7.5−7.3 (5 H, m, ArH),
6.39 (1 H, bs, NHSer), 5.70 (1 H, bs, NH), 5.3−5.0 (4 H, m, OCH2Ph),
4.76 (1 H, d), 4.69 (1 H, bs), 4.51 (1 H, bs), 4.35 (1 H, bs), 4.2 (2 H, dd),
4.05 (1 H, bs), 3.9−3.6 (3 H, m, HSerβ), 2.05, 2.03, and 1.99 (9 H, 3 × s,
Pentacosa-10,12-diynoic acid {3-[β-D-maltotriosyl]-
hydroxy-propyl}amide (12)
Sodium methoxide (10 mg, 0.18 mmol) was added to a solution of com-
pound (15) (225 mg, 0.2 mmol) in dry methanol, and the mixture was
stirred for 6 h at room temperature. The solution was neutralized with
Dowex 50-X8 (H+) resin, filtered, and evaparated in vacuo. The residue
was dissolved in methanol (10 mL) and Pd-C (130 mg) was added to
the mixture. The mixture was then stirred for 48 h under a hygrogen
atmosphere. The mixture was filtered to remove the catalyst, and the
solvent allowed to evaporate. To complete the reaction, the residue
was dissolved in 4 : 1 methanol/H2O (25 mL) and Pd/C (220 mg) was

576
S.-I. Nishimura et al.
added. The reaction mixture was kept for 17 h under a hydrogen atmo-
sphere at a pressure of 0.32 MPa. The mixture was filtered and the
filtrate was evaporated to afford crude 3-aminopropyl β-maltotrioside.
EDC (163 mg, 0.85 mmol) and DIEA (50 µL, 0.28 mmol) were added
to the suspension of crude 3-aminopropyl β-maltotrioside and PDA
(100 mg, 0.27 mmol) in methanol. The mixture was stirred for 23 h at
room temperature, then purified with silica gel chromatography, elu-
ent CHCl3/MeOH/H2O (65 : 15 : 1). Pure product (12) was obtained by
purification on a Sephadex G-10 column eluted with water (150 mg,
80%). δH 5.14 (1 H, d, J 4.2), 5.12 (1 H, d, J 3.6), 4.26 (1 H, d, H-1),
3.42 (1 H, dd, J 3.6, 9.6), 3.34 (1 H, m, H-5), 3.22 (1 H, dd, J 9.0, H-2),
[8] S.-I. Nishimura, K. Yamada, J. Am. Chem. Soc. 1997, 119,
10 555.
[9] S. Nishiguchi, K. Yamada, Y. Fuji, S. Shibatani, A. Toda,
S.-I. Nishimura, Chem. Comm. 2001, 1944.
[10] S.-I. Nishimura, Curr. Opin. Chem. Biol. 2001, 5, 325.
[11] M. Mammen, S. Choi, G. M. Whitesides, Angew. Chem. Int.
Ed. 1998, 37, 2754.
[12] J. C. Sacchettini, L. G. Baum, C. F. Brewer, Biochemistry 2001,
40, 3009.
[13] P. R. Crocker, T. Feizi, Curr. Opin. Struct. Biol. 1996, 6, 679.
[14] Y. C. Lee, R. R. Townsend, M. R. Hardy, J. Lonngren, J. Arnarp,
M. Haraldsson, H. Lonn, J. Biol. Chem. 1983, 258, 199.
[15] L. L. Kiessling, J. E. Gestwicki, L. E. Strong, Curr. Opin.
Chem. Biol. 2000, 4, 696.
2.22 (4 H, t, J 7.2, CH2C
0.88 (3 H, t, CH3). δC 176.4 (CONH), 104.3 (C-1), 102.9 and 102.8
(Cꢀ-1 and Cꢀꢀ-1), 81.4, 81.3, 77.8 (C
C), 76.6, 75.1, 74.9, 74.8,
C C CCH2), 2.15 (2 H, J 7.2, t, COCH2),
C
C
74.7, 74.3, 73.8, 73.4, 71.5, 68.4, 66.4, 62.8, 62.2, 37.6 (CH2NHCO),
37.2, 33.1 (NHCOCH2), 30.7, 30.6, 30.5, 30.4, 30.3, 30.1, 30.0, 29.9,
[16] K. Simons, E. Ikonen, Nature 1997, 387, 569.
[17] S. Hakomori, Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 225.
[18] Y. F. Défrene, W. R. Barger, J.-B. D. Green, G. U. Lee, Langmuir
1997, 13, 4779.
[19] C. Yuan, L. J. Johnston, Biophys. J. 2001, 81, 1059.
[20] H. A. Rinia, M. M. Snel, J. P. van der Eerden, B. de Kruijff,
FEBS Lett. 2001, 501, 92.
29.6, 27.0, 23.7 (CH2CH3), 19.7 (CH2C
C C CCH2), 14.4, (CH3).
TOF-MS m/z 940.9 [M + Na]+ (Calc. for C46H79NO17: 940.52).
10,12-Pentacosadinyl-β-D-maltotrioside (13)
N-iodosuccinimide (400 mg, 1.78 mmol) and molecular sieves were
added to a mixture of thioglycoside (14) (500 mg, 0.49 mmol) and 10,12
pentacosadiyn-1-ol(200 mg, 0.55 mmol)indryCH2Cl2 at−20◦C.After
stirring for 20 min, TfOH (20 µL) was added. The mixture was stirred
for 10 min at −20◦C and then warmed to room temperature.The reaction
was quenched by adding a few drop of TEA, and the reaction mixture
was then diluted with CH2Cl2 and filtered. The filtrate was washed
with saturated Na2SO3, saturated NaHCO3, and brine. The residue was
purified by silica gel chromatography, eluent 8 : 2 v/v CH2Cl2/EtOAc,
to afford (16) (120 mg, 90%). Sodium methoxide (17 mg, 0.31 mmol)
was added to a solution of compound (16) (100 mg, 0.28 mmol) in dry
methanol (10 mL) and the mixture was stirred for 5 h at room tem-
perature. The solution was neutralized with Dowex 50-X8 (H+) resin,
filtered and evaparated in vacuo to afford (13) as an amorphous powder.
δH 5. 14 (2 H, dd, J 3.6, 9.6, Hꢀ-1 and Hꢀꢀ-1), 4.25 (1 H, d, J 7.8, H-1),
3.42 (1 H, dd, J 3.6, 9.6), 3.41 (1 H, m), 3.20 (1 H, dd, J 7.8, 9.0), 2.22
[21] S. Förster, T. Plantenberg, Angew. Chem. Int. Ed. 2002, 41, 688.
[22] O. Karthaus, N. Maruyama, X. Cieren, M. Shimomura,
H. Hasegawa, T. Hashimoto, Langmuir 2000, 16, 6071.
[23] G. MacBeath, S. L. Schreiber, Science 2000, 289, 1760.
[24] H. Zhu, M. Bilgin, R. Bangham, D. Hall, A. Casamayor,
P. Bertone, N. Lan, R. Jansen, S. Bidlingmaier, T. Houfek,
T. Mitchell, P. Miller, R. A. Dean, M. Gerstein, M. Snyder,
Science 2001, 293, 2101.
[25] T. Kodadek, Chem. Biol. 2001, 8, 105.
[26] B. T. Houseman, M. Mrksich, Chem. Biol. 2002, 9, 443.
[27] D. Wang, S. Liu, B. J. Trummer, C. Deng, A. Wang, Nat.
Biotechnol. 2002, 20, 275.
[28] M. C. Bryan, O. Plettenburg, P. Sears, D. Rabuka, S. Wacowich-
Sgarbi, C. H. Wong, Chem. Biol. 2002, 9, 713.
[29] D. H. Charych, J. O. Nagy, W. Spevak, M. D. Bednarski, Science
1993, 261, 585.
[30] D. Charych, Q. Cheng, A. Reichert, G. Kuziemko, M. Stroh,
J. O. Nagy, W. Spevak, R. C. Stevens, Chem. Biol. 1996, 3,
113.
[31] D. Day, H. Ringsdorf, J. Polym. Sci. 1978, 16, 205.
[32] M. Ito, T. Kurita, K. Kita, J. Biol. Chem. 1995, 270, 24 370.
[33] H. Tachibana, Y. Yamanaka, H. Sakai, M. Abe, M. Matsumoto,
Macromolecules 1999, 32, 8306.
[34] R. Lotan, E. Skutelsky, D. Danon, N. Sharon, J. Biol. Chem.
1975, 250, 8518.
[35] R. Banerjee, S. C. Mande, V. Ganesh, K. Das, V. Dhanaraj,
S. K. Mahanta, K. Suguna, A. Surolia, M. Vijayan, Proc. Natl.
Acad. Sci. U. S. A. 1994, 91, 227.
[36] J. Kajihara, Y. Guoji, K. Kato, Y. Suzuki, Biosci. Biotechnol.
Biochem. 1995, 59, 155.
[37] A. Konno, K. Nunogami, T. Wada, A. Yachie, Y. Suzuki,
N. Takahashi, T. Suzuki, D. Miyamoto, M. Kiso, A. Hasegawa,
T. Miyawaki, Int. Immunol. 1996, 8, 1905.
(4 H, t, J 7.2, CH2C
δC 102.9 (C-1), 101.7 (Cꢀ-1 and Cꢀꢀ-1), 85.5, 80.4, 77.0 (C
76.3, 75.1, 74.9, 73.7, 73.6, 73.2, 72.8, 72.3, 72.0, 70.6, 70.2, 70.1, 61.7
(C
C), 61.5, 60.8 (Cꢀ-6 and Cꢀꢀ-6), 31.8, 29.5, 29.4, 29.3, 29.2,
C
C
CCH2), 0.88 (3 H, t, J 7.2, CH2CH3).
C
C
C),
C
C
28.9, 28.7, 28.2, 25.8, 22.5 (CH2CH3), 18.8 (CH2C
C C CCH2),
13.6 (CH3). FAB-MS m/z 847 [M + H]+ (Calc. for C43H74O16: 847.5).
References
[1] R. T. Lee, Y. C. Lee, Glycoconj. J. 2000, 17, 543.
[2] S.-I. Nishimura, Y. C. Lee, in Polysaccharides: Structural
Diversity and Functional Versatility (Ed. S. Dumitriu) 1998,
Ch. 15, pp. 523–537 (Marcel Dekker: New York, NY).
[3] J. J. Lundquist, E. J. Toone, Chem. Rev. 2002, 102, 555.
[4] S.-I. Nishimura, T. Furuike, K. Matsuoka, K. Maruyama,
K. Nagata, K. Kurita, N. Nishi, S. Tokura, Macromolecules
1994, 27, 4876.
[5] T. Furuike, N. Nishi, S. Tokura, S.-I. Nishimura, Macromolecules
1995, 28, 7241.
[6] K. Matsuoka, S.-I. Nishimura, Macromolecules 1995, 28, 2961.
[7] K. Yamada, E. Fujita, S.-I. Nishimura, Carbohydr. Res. 1997,
305, 443.
[38] S. Hirmo, E. Artursson, G. Puu, T. Wadstrom, B. Nilsson, Anal.
Biochem. 1998, 257, 63.
[39] M. Matsuda, S. Nishimura, F. Nakajima, T. Nishimura, J. Med.
Chem. 2001, 44, 715.