High-throughput evaluation system based on fluorescence intensity distribution analysis-polarization to investigate carbohydratecarbohydrate interactions

Article abstract of DOI:10.1246/bcsj.20150369

We synthesized rhodamine-labeled β-lactoside and assessed its rotatory diffusion rates in aqueous media containing various oligosaccharides through fluorescence intensity distribution analysis-polarization measurements to find that its rotatory diffusion rates decrease in the presence of the coexisting oligosaccharides. Especially, the coexisting lactose lowered the rotatory diffusion rate more effectively than the other disaccharides did. This restrained rotatory diffusion arises from intermolecular carbohydratecarbohydrate interactions between the rhodaminelabeled β-lactoside and the coexisting lactose. We also detected intermolecular bindings between the rhodamine-labeled β-lactoside and monosialyl-disaccharides (3′- and 6′-sialyllactoses). These data clearly show that the fluorescence intensity distribution analysis-polarization-based system is quite advantageous to probe carbohydratecarbohydrate interactions.

Full text of DOI:10.1246/bcsj.20150369

High-Throughput Evaluation System based on Fluorescence
Intensity Distribution Analysis-Polarization to Investigate
Carbohydrate­Carbohydrate Interactions
1
2
1
1
1
Maho Iwamura, Ryoichi Koyama, Yuki Nonaka, Fumiko Dai, Ryoji Matsuoka,
1
2
2
2,3
Masaki Nakamura, Haruo Iwabuchi, Takahiro Okada, and Teruaki Hasegawa*
1Graduate School of Life Sciences, Toyo University, 1-1-1 Izumino, Itakura-machi, Ora-gun, Gunma 374-0193
2Department of Life Sciences, Toyo University, 1-1-1 Izumino, Itakura-machi, Ora-gun, Gunma 374-0193
3Bio-Nano Electronics Research Centre, Toyo University, 2100 Kujirai, Kawagoe, Saitama 350-8585
E-mail: t-hasegawa@toyo.jp
Received: October 23, 2015; Accepted: February 15, 2016; Web Released: February 26, 2016
Teruaki Hasegawa
Teruaki Hasegawa received his Ph.D. in Engineering from Nagoya University in 2001. After postdoctoral
work as a JSPS research fellow at University of Washington and Kyushu University, he was appointed as a
research associate at Kyushu University in 2004. He then moved to Toyo University in 2006. He is currently a
professor at Department of Life Sciences, Toyo University. In addition to the FIDAPo-based approach for
investigating CCIs, his ongoing research includes click chemistry-based modifications of native
polysaccharides to develop various eco, chiral, and biomaterials.
Abstract
increasing research interest has been, however, placed on
carbohydrate­carbohydrate interactions (CCIs), in which car-
bohydrates recognize carbohydrates in specific and, in some
We synthesized rhodamine-labeled β-lactoside and assessed
its rotatory diffusion rates in aqueous media containing various
oligosaccharides through fluorescence intensity distribution
analysis-polarization measurements to find that its rotatory dif-
fusion rates decrease in the presence of the coexisting oligosac-
charides. Especially, the coexisting lactose lowered the rotatory
diffusion rate more effectively than the other disaccharides did.
This restrained rotatory diffusion arises from intermolecular car-
bohydrate­carbohydrate interactions between the rhodamine-
labeled β-lactoside and the coexisting lactose. We also de-
tected intermolecular bindings between the rhodamine-labeled
β-lactoside and monosialyl-disaccharides (3¤- and 6¤-sialyl-
lactoses). These data clearly show that the fluorescence intensity
distribution analysis-polarization-based system is quite advan-
tageous to probe carbohydrate­carbohydrate interactions.
2+
5,6
cases, Ca -dependent manners. In these newly-recognized
binding events, the GSLs laterally aggregate within the mem-
branes to form microdomains presenting densely packed car-
bohydrates (glycoclusters) on their surfaces. Interfacial CCIs
between such glycoclusters on two adjoining cells are the very
first steps to initiate their cell­cell adhesions; those would
be then reinforced by strong carbohydrate­protein/protein­
X
X
protein interactions. For example, Le ­Le interactions be-
tween blastmeres are essential for their compactions and G 3­
M
7
,8
Gg3 interactions act as adhesive forces to induce metastases.
Investigation of CCIs is of significant importance, since it
should supply useful information not only to understand
mechanisms of the cell­cell adhesions but also to design new
drugs to prevent various diseases triggered by unfavorable cell­
cell adhesions (cancer metastases, inflammations, etc.). In spite
of such importance, not only heterogeneic/fluidic nature of
the cell membranes but also fluctuation of GSLs-levels on
living cell surfaces make it quite difficult to investigate CCIs in
detailed and quantitative manners. Simple and well-designed
model systems are, therefore, highly required to obtain molec-
ular-level information on CCIs.
Introduction
Considerable research efforts have been focused on func-
tional analyses of glycosphingolipids (GSLs) on cell surfaces
to reveal that these GSLs play substantial roles in various
biological events including fertilizations, differentiations, and
cell­cell adhesions.1 In these biological events, hydrophilic
carbohydrate units of the GSLs act as ligands for carbohydrate
binding proteins (lectins or lectin-like proteins) and such car-
bohydrate­protein interactions have been recognized as main
­3
Many research groups have reported varieties of model
9
­27
systems to access molecular-level information for CCIs.
For
example, Penadés et al. developed gold nanoparticles (GNPs)
9
,10
coated with densely packed glycoclusters to probe CCIs.
In
4
driving forces to induce these biological events. Recently,
their work, they investigated effects of cations to induce CCIs-
Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan | 617

Figure 1. Schematic illustration of the rotatory diffusions of the Rho-labeled oligosaccharides with/without CCIs.
Chart 1. Structures and abbreviations of the oligosaccharides used in this work.
X
mediated aggregations of the GNPs carrying Le -trisaccharide
rotatory diffusion rates of the former and the resultant increased
degree of polarization (P) of its fluorescence would be readily
monitored through the FIDAPo measurements. This FIDAPo-
based system has great advantages as the HTE systems as
follows. The instrument adopts 386-well glass-bottom plates
and it takes only one minute or less to complete single FIDAPo
measurement, making it possible to conduct analyses of many
sample solutions in a short time. In addition, varieties of com-
mercially available free oligosaccharides can be directly used
in our system as the coexisting oligosaccharides. Time/labor-
consuming chemical modification processes are, therefore,
avoidable in this system. Furthermore, only limited volume
(30­80 ¯L) of sample solution is required for each well so that
amounts of the coexisting oligosaccharides are highly reduced,
ascertaining low-cost measurements and wide applicability of
this system to evaluate CCIs with high-priced bioactive oligo-
saccharides. We herein report some results in our study using
this FIDAPo-based system, in which we investigated the CCIs
between Rho-labeled β-lactoside and the coexisting oligo-
saccharides including lactose (Lac), cellobiose (Cel), maltose
(Mal), 3¤-sialyllactose (3¤SLac), and 6¤-sialyllactose (6¤SLac)
(Chart 1).
X
(
Le : Galβ1,4(Fucα1,3)GlcNAcβ) by TEM imagings to find
that Ca induces aggregations of such glyco-GNPs more
effectively than Na does. Russel et al. also carried out
similar TEM imagings and they found that Ca2 also medi-
3
2
+
+
+
ates aggregations of GNPs immobilized with β-lactosides
(
(
Lac: Galβ1,4Glcβ).11 Glycoclusters having synthetic scaffolds
glycopolymers) have been also developed to probe CCIs. For
example, interactions between Langmuir monolayers com-
posed of G 3 (NeuAcα2,3Galβ1,4Glcβ-ceramide) and linear/
dendritic polymers presenting multiple Lacs at air/water inter-
faces were investigated by using ³­A isotherms.1
In spite of these excellent works, progress of the inves-
tigations on CCIs is still limited. In order to make a break-
through in this research area, we are now directing our atten-
tion to fluorescence intensity distribution analysis-polarization
M
2,13
(
FIDAPo) as a potential high-throughput evaluation (HTE)
2
8,29
system.
In this FIDAPo-based system, small numbers
ca. 1­5) of fluorescent molecules within a small confocal area
ca. 1 femto liter) are excited to monitor their rotatory diffusion
rates. Especially, in this work, we used an MF20 (Olympus) as
a highly automated apparatus for the FIDAPo measurements
and monitored the rotatory diffusion rates of rhodamine (Rho)-
labeled oligosaccharides with/without coexisting oligosaccha-
rides (Figure 1). In this system, if adhesive CCIs work between
the Rho-labeled oligosaccharides and the coexisting (non-
labeled/free) oligosaccharides, such CCIs should lower the
(
(
Results and Discussion
Synthesis. The Rho-labeled β-lactoside (RhoLac, Chart 2)
was synthesized through a coupling between per-O-acetylated
β-lactoside 4 and commercially available Rho functionalized
618 | Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan

with N-hydroxysuccinimide ester (RhoNHS, Chart S1)
Scheme 1). Briefly, in the synthesis of 4, p-aminophenyl-β-
areas (Figure 2). Since commercially available RhoNHS is a
mixture of two isomers having NHS groups at different (o- or
p-) positions, these two predominant peaks can be assigned
to two isomeric RhoLac. In fact, these peaks both showed
identical electrospray ionization (ESI) mass spectra. Our trial to
clarify which peak is assignable to o- or p-isomeric RhoLac
turned out to be a failure, since these two peaks appeared
almost identical in ESI mass spectrometry/mass spectrometry
(MS/MS). These identical MS/MS spectra also clearly indicate
that these two HPLC peaks are assignable to the two isomeric
RhoLac. We also synthesized another Rho derivative having no
Lac unit (RhoH, 11, Scheme 2) as a negative reference through
a coupling between RhoNHS and aniline derivative. After
purification on HP20, RhoH also showed two predominant
peaks (TR = 23.1 and 23.5 min) in its HPLC chart (Figure S1).
ESI-MS analyses of these two peaks also showed identical MS
and MS/MS spectra, again indicating that these two HPLC
peaks can be assigned to the two isomeric RhoH.
(
2
6
lactoside 1 was amidated with chloroacetyl chloride and the
subsequent S 2 reaction using NaN afforded β-lactoside 3
N
3
having methylene-tethered azide terminal. The subsequent
reduction on Pd/C afforded 4 that was then treated with
RhoNHS in N,N-dimethylformamide (DMF) containing NaH-
CO . After deacetylation in aqueous ammonia, the deacetylated
3
product (RhoLac, 6) was purified through column chroma-
tography packed with hydrophobic resin (Diaion HP20,
Mitsubishi chemical) to remove impurities including hydro-
lysates (RhoCOOH) originated from unreacted RhoNHS.
HPLC analyses of the purified RhoLac showed two predom-
inant peaks (T = 12.6 and 13.0 min) with comparable peak-
R
Molecular Dynamics Calculation. Before the FIDAPo
measurements, we carried out molecular dynamics (MD)
calculations on RhoLac to confirm that RhoLac takes suitable
conformations in bulk water to allow intermolecular CCIs with
the coexisting oligosaccharides. The most stable conformation
of RhoLac in its MD calculations showed that the spacer unit
spatially keeps the Lac unit apart from the bulky Rho to assure
good accessibility of the coexisting oligosaccharides toward the
Lac unit (Figure 3).
FIDAPo Measurements. We then carried out FIDAPo
measurements on RhoLac and RhoH. Note that, in these
FIDAPo measurements, rotatory diffusion rates of the fluo-
rescence-labeled molecules were shown in mP values (mP =
1
000 © P, eq 1). We found that mP values of RhoLac and
RhoH in pure water (mP ) are 33.72 and 51.41, respectively.
0
These small mP0 values clearly indicate their fast rotatory
diffusions in these media. It should be noted that RhoH showed
Chart 2. Structures of RhoLac and RhoH, in which gray
spots indicate the positions onto which the lactosylphenyl
a larger mP value than that obtained for RhoLac, indicating
0
slower rotatory diffusion rate of RhoH than that of RhoLac
although the smaller molecules usually have higher rotatory
diffusion rates and lower mP values. We assume that some
(RhoLac) and phenyl (RhoH) units are attached through
the spacers.
Scheme 1. Synthetic route to access RhoLac (6).
Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan | 619

Figure 2. HPLC chart of RhoLac after purification through column chromatography packed with HP20, Conditions: Inertsil ODS-4,
¹1
5
¯m, 4.6 © 250 mm, flow rate: 1.0 mL min , ­ = 560 nm, aqueous acetonitrile containing TFA (0.05%), [acetonitrile] = 5­95%
v/v (0­20 min), and LC-MS/MS spectra by ESI (m/z 903.3261, collision energy: 40 V, insets) of the two peaks.
Figure 3. The most stable conformation of RhoLac in bulk
water during the MD calculation (the Rho/Lac and spacer
units are depicted as CPK and ball-and-stick models,
respectively, and the water molecules surrounding RhoLac
were removed after the MD calculation for clear
presentation).
CCIs between Lac and Disaccharides. We carried out
Scheme 2. Synthetic route to access RhoH (11).
similar FIDAPo measurements on the Rho derivatives in
aqueous media containing the oligosaccharides and/or cations
RhoH molecules, in our study, might aggregate in the media
through hydrophobic RhoH­RhoH interactions to lower its
rotatory diffusion rates.
(Table 1). Interestingly, all samples showed mP values larger
than the corresponding mP values, indicating that the rotatory
0
diffusion rates of RhoLac and RhoH were lowered in the media
620 | Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan

Table 1. Molecular polarization (mP) of Rho-derivatives with/without the coexisting oligosaccharides or cationsa)
RhoLac
mP
RhoH
Oligosaccharides/Cations
¦mPr
mPr-RhoLac
mP
mPr-RhoH
¹
/¹b)
33.72 « 0.04
1.000 « 0.002
50.3 « 0.2
1.000 « 0.005
0.000 « 0.006
+
Na
Ca
35.5 « 0.6
37.2 « 0.4
1.05 « 0.02
1.10 « 0.01
50 « 1
51.8 « 0.6
0.99 « 0.02
1.03 « 0.01
0.06 « 0.03
0.07 « 0.02
2
+
Lac
Lac/Na
Lac/Ca2
41.3 « 0.2
42.5 « 0.3
46.1 « 0.1
1.226 « 0.006
1.260 « 0.008
1.367 « 0.004
62 « 4
59.9 « 0.2
61.6 « 0.2
1.22 « 0.08
1.191 « 0.006
1.225 « 0.006
0.00 « 0.06
0.058 « 0.009
0.116 « 0.007
+
+
Mal
Mal/Na
Mal/Ca2
41.2 « 0.2
42.3 « 0.1
45.9 « 0.3
1.223 « 0.006
1.255 « 0.004
1.36 « 0.01
59.8 « 0.4
58.6 « 0.9
62.1 « 0.7
1.188 « 0.009
1.17 « 0.02
1.23 « 0.02
0.030 « 0.009
0.08 « 0.02
0.10 « 0.02
+
+
Cel
Cel/Na
Cel/Ca2
40.9 « 0.3
42.3 « 0.6
45.3 « 0.4
1.21 « 0.01
1.26 « 0.02
1.34 « 0.01
55.7 « 0.2
58.9 « 0.5
62 « 1
1.106 « 0.005
1.17 « 0.01
1.23 « 0.02
0.09 « 0.01
0.07 « 0.02
0.10 « 0.02
+
+
3
3
3
¤SLac
¤SLac/Na
¤SLac/Ca
44.5 « 0.2
46.3 « 0.4
50.1 « 0.4
1.320 « 0.006
1.37 « 0.01
1.49 « 0.01
62.2 « 0.2
65.1 « 0.2
67.8 « 0.1
1.237 « 0.007
1.294 « 0.006
1.348 « 0.006
0.067 « 0.008
0.06 « 0.01
0.102 « 0.009
+
2
+
6
6
6
¤SLac
¤SLac/Na
¤SLac/Ca
44.6 « 0.3
46.0 « 0.3
51.7 « 0.4
1.32 « 0.01
1.364 « 0.008
1.53 « 0.01
62.7 « 0.2
66.1 « 0.7
67.3 « 0.4
1.246 « 0.007
1.31 « 0.01
1.338 « 0.009
0.06 « 0.01
0.04 « 0.01
0.15 « 0.01
+
2
+
a) Conditions: [RhoLac] or [RhoH] = ca. 5 nM, [oligosaccharides] = 0 or 167 mM, [cations] = 0 or 200 mM, in
water, 25 °C, n = 10 (see Experimental section for detail), b) No oligosaccharide or cation.
containing these oligosaccharides and/or cations. For example,
mP value of RhoLac greatly increased to 46.1 in the presence
2
+
2+
of both Lac and Ca (Lac/Ca ). It should be noted, however,
that RhoH as the negative reference also showed increased mP
value (61.6) in the presence of Lac/Ca2 . The additions of the
free oligosaccharides and/or cations would increase viscosity
of the sample solutions and such increased viscosities should
be responsible for the restrained molecular rotation of RhoH.
As shown in eq 1, mP value of Rho derivatives is thought to be
a function of temperature (T), fluorescence lifetime (¸), viscos-
ity of the media (©), and volume of fluorescent molecules (V).
In the case of small molecules, 5kBT¸/6©V in eq 1 is greatly
larger than 1 and therefore, eq 1 can be roughly simplified into
eq 2, implying that increased viscosities of the media increase
mP values of the Rho derivatives in almost linear fashions. We,
therefore, calculated relative mP (mP = mP/mP ) values of
+
Figure 4. ¦mPr values of RhoLac in the presence of the
oligosaccharides and/or the cations in which the error bar
indicates the standard deviation (n = 10, see Experimental
section for detail).
r
0
the Rho derivatives to evaluate the effects of the increased
viscosities of the media on their rotatory diffusion rates. For
example, mP values of RhoLac and RhoH can be estimated
r
2
+
as 1.367 and 1.255, respectively, in the presence of Lac/Ca
.
indicative of higher Lac­Lac affinity than Lac­Mal and Lac­
Cel.
Note that the mP value of RhoLac is significantly higher than
r
that of RhoH. This fact indicates that not only the increased
viscosity of the media but also the intermolecular CCIs with the
5
00
ð5kBT¸=6©VÞ þ 1
00©V
mP ¼
ð1Þ
ð2Þ
coexisting Lacs is responsible for the increased mP value of
r
6
RhoLac. We then evaluated contributions of the intermolecular
;
kBT¸
CCIs (¦mP ) in the restrained rotatory diffusions of RhoLac by
r
using eq 3. For example, ¦mP of RhoLac in the presence of
where k : Boltzmann constant, T: temperature (K), ¸: fluo-
rescence lifetime, ©: viscosity of the media, V: volume of the
fluorescent molecule.
r
B
2
+
Lac/Ca was estimated to be 0.116 through this calculation.
On the other hand, the coexistence of the other disaccharides,
2
+
2+
such as Mal/Ca and Cel/Ca , induced less ¦mP values
r
mPr-RhoLac ꢀ mPr-RhoH
3
0
2+
ꢀ
mPr ¼
ð3Þ
than Lac/Ca did (Figure 4). The larger ¦mP value observed
for Lac/Ca than those for Mal/Ca and Cel/Ca is clearly
r
mPr-RhoH
2
+
2+
2+
Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan | 621

As mentioned above, our FIDAPo-based system clearly
showed that the coexisting Lacs interact with the Lac unit of
RhoLac to lower its rotatory diffusion rates. It should be,
however, noted that the specificity of the Lac­Lac interaction
obtained through our FIDAPo measurements is relatively low
dependent.12,13 We also probed this Ca2+-mediated Lac-3¤SLac
interaction by using our FIDAPo-based system; that is, ¦mP
r
2+
value of RhoLac on the coexistence of 3¤SLac/Ca was larger
+
than those of 3¤SLac/Na and 3¤SLac. These data indicate that
2+
Ca also plays substantial roles to induce the CCIs between
Lac and 3¤SLac. It is, however, to our great surprise that the
and there are only small differences in the ¦mP values induced
r
2+
by Lac­Lac, Lac­Mal, and Lac­Cel interactions. This low
specificity might arise from the mechanism to increase ¦mPr
coexisting 6¤SLac/Ca induced much higher ¦mPr value
2+
(0.15) than that induced by 3¤SLac/Ca , indicating that Lac
has higher affinity toward 6¤SLac over 3¤SLac in the presence
values in our FIDAPo-based system; that is, the ¦mP values
r
2+
2+
originate from monomeric CCI between the Lac units of
RhoLac and the coexisting oligosaccharides. On the other
hand, the aggregations of glyco-GNPs as well as the adhesions
between the Langmuir monolayers and glycopolymers both
arise from multivalent CCIs between the glycoclusters. In the
latter two systems, the carbohydrate specificity might be
enhanced by the well-known “carbohydrate cluster effects”.
It might be readily assumed that multivalent CCIs between
glycoclusters are also adopted in nature to overcome such low
specificity of each CCI.
of Ca . Note that, with the best of our knowledge, such Ca -
mediated Lac-6¤SLac interaction has been never reported so far.
In this respect, our FIDAPo-based system is quite useful not
only as HTE systems for detailed investigation of the mecha-
2+
nisms of the “already-known” CCIs (e.g. Ca -mediated Lac­
Lac interactions), but also as a high-throughput screening
(HTS) system for discovering new CCIs. On the other hand,
most of the conventional model systems are not suitable for the
HTS, suffering not only from labor-consuming procedures to
construct the model systems with different carbohydrate units
but also time-consuming processes to complete the measure-
ments on each sample.
Cation-Dependency of the CCIs. In spite of the excellent
work so far conducted by many research groups, the data from
these studies sometimes conflict with each other and the discus-
sion concerning cations to induce CCIs has not come to a con-
clusion yet. For example, Russel et al. reported in their paper
that Lac-immobilized GNPs (Lac-GNPs) form large aggregates
Conclusion
We newly established the FIDAPo-based HTE/HTS system
to probe CCIs between the Rho-labeled oligosaccharides and
the coexisting (nonlabeled/free) oligosaccharides in aqueous
media. We monitored, by using this system, the degree of fluo-
rescence polarization of RhoLac to find that Lac, 3¤SLac, and
6¤SLac interact with RhoLac to lower its rotatory diffusion
in the aqueous media on additions of Ca2 while Na did not
+
+
induce such aggregations, showing that the Lac­Lac inter-
2
+
11
actions are mediated by Ca in a specific manner. Similar
2
+
Ca -induced aggregation was also observed for polyfluorene-
14
+
2+
based glycoclusters. Again, in this case, Na did not induce
any aggregations of the multi-lactosylated polyfluorene. In
rates in Ca -dependent fashions. These data clearly show our
FIDAPo-based system has great advantages as a THE/THS
system for probing CCIs. Our FIDAPo-based system, however,
still has some problems mainly arising from weak/monomeric
CCI between RhoLac and the coexisting oligosaccharides, and
2
+
contrast to these clear Ca -specificities, we reported in our
paper that not only Ca2 but also Na induced the Lac­Lac
+
+
24
interactions in an intramolecular fashion. In the latter work,
not only Ca but also Na induced a twisted conformation of
a ferrocene derivative having a Lac-appendage at each cyclo-
2+
+
the subsequent small ¦mP values obtained after the data
r
processing. We, however, still believe that our FIDAPo-based
system has great advantages for quick and comprehensive
investigations to understand the molecular-level mechanism of
CCIs.
pentadienyl ring, possibly through Ca2 /Na -induced Lac­Lac
interaction within the molecule. This difference in the cation-
dependency observed in these model systems should again
arise from the structural difference in their carbohydrate
clusters; that is, the former systems detected the multivalent
CCIs between the glycoclusters immobilized onto GNPs/
polyfluorene and the latter one did monovalent CCI between
two carbohydrate units attached onto single ferrocene mole-
+
+
Experimental
Materials.
Silica gel 60 N (spherical, neutral, particle
size 63­210 ¯m) for column chromatography was purchased
from KANTO CHEMICAL. Thin layer chromatographic
(TLC) analyses were carried out on Merck TLC glass plates
(1.05715.0001) precoated with silica gel 60 with fluorescent
indicator (­ = 254 nm). DIAION HP20 was purchased from
Mitsubishi Chemical. Chloroacetyl chloride, sodium hydrogen
carbonate, sodium chloride, and anhydrous tetrahydrofuran
(THF) were purchased from Wako Pure Chemicals. Sodium
azide, N,N-dimethylformamide (DMF), and methanol were
purchased from Kishida Chemicals. Ethyl acetate and tolu-
ene were purchased from GODO. RhoNHS (PCC46406) was
purchased from Funakoshi. Aqueous ammonia, triethylamine,
acetonitrile (HPLC grade), and acetic acid were purchased from
KANTO KAGAKU. Palladium on carbon (Pd/C, 5%) was
purchased from KOJIMA CHEMICALS. Anhydrous magne-
sium sulfate was purchased from SHOWA CHEMICAL. These
all chemicals were used without further purification. The start-
cule. The ¦mP values herein reported, however, can afford
r
decisive information on the cation dependency of CCIs. For
example, the coexisting Lac/Ca2 induced much higher ¦mPr
+
2
+
value of RhoLac than those induced by sole additions of Ca
+
or Lac. On the other hand, the additions of Lac/Na showed
similar ¦mP value to that induced by Na only, indicating that
Na negligibly induces the Lac­Lac interactions. These data
clearly show that Ca selectively mediates the Lac­Lac inter-
+
r
+
2+
2
+
+
actions (Ca º Na ¿ no ion).
CCIs between Lac and Monosialyl-disaccharides. It has
been already reported that lactosylceramide interacts with GM3
on the cell surfaces through CCIs between Lac units of the
former and 3¤-sialyllactoside (3¤SLac: NeuAcα2,3Galβ1,4Glc)
unit of the latter. Detailed investigations using artificial model
2+
systems revealed that this Lac-3¤SLac interaction is Ca
622 | Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan

ing material (pAPLac, 1) for the synthesis of RhoLac was
synthesized from lactose monohydrate through the 4-steps
chemical modification as reported in the literature (ref. 26).
(200 mL), sodium azide (1.1 g) was added and the resultant
mixture was stirred at 70 °C for 15 h. The mixture was then
diluted with ethyl acetate and washed with aqueous NaCl
solution. The organic layer was dried over anhydrous MgSO4,
filtered, concentrated to dryness and purified through chroma-
tography using a glass column packed with silica gel (40 cm
long; 3 cm i.d.; toluene­ethyl acetate = 2:1­1:1 in v/v) to give
1
13
General.
H and C NMR spectra were acquired on an
AL300 (JEOL RESONANCE) in CDCl . The chemical shifts
3
are reported in ppm (¤) relative to Me Si. IR spectra were
4
recorded on a FT/IR-4100 Fourier transform infrared spec-
trometer (JASCO). Matrix-assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectra were recorded on an
AXIMA-CFR+ (SHIMADZU). High-resolution LC-MS and
LC-MS/MS spectra by ESI were recorded on a SYNAPT G2
HDMS (Waters, Co.). UV­vis spectra were acquired on a V-
the desired product as marmalade powders (87%, 2 steps).
1
H NMR (300 MHz, CDCl , r.t.): ¤ 7.96 (s, 1H), 7.46 (d, J =
3
8.7 Hz, 2H), 6.96 (d, J = 9.3 Hz, 2H), 5.36 (d, J = 3.3 Hz, 1H),
5.28 (t, J = 7.7 Hz, 1H), 5.16 (t, J = 8.3 Hz, 1H), 5.13 (dd, J =
7.5 and 11.4 Hz, 1H), 5.01 (d, J = 7.8 Hz, 1H), 4.97 (dd, J =
3.3 and 10.2 Hz, 1H), 4.51 (d, J = 7.5 Hz, 1H), 4.51­4.48 (m,
1H), 4.18­4.05 (m, 5H), 3.92­3.86 (m, 2H), 3.79­3.75 (m, 1H),
2.16 (s, 3H), 2.10 (s, 3H), 2.08 (s, 3H), 2.07 (s, 3H), 2.07 (s,
6
1
30 (JASCO). Lyophilizations were carried out using a FDU-
200 (TOKYO RIKAKIKAI). HPLC analyses were carried out
on our integrated HPLC system (JASCO, pump: PU-2089,
detector: UV-2095, MD-2018, and FP-2020) equipped with
Inertsil ODS-4 (5 ¯m, 4.6 © 250 mm). The FIDAPo measure-
ments were conducted on an MF20 (Olympus). Molecular
dynamics calculation was carried out by using Discovery
Studio 4.0 program (Accelrys, CHARMm, 10 ps equilibrium,
1
3
3H), 2.06 (s, 3H), 1.97 (s, 3H); C NMR (75 MHz, CDCl3,
r.t.): ¤ 170.255, 170.223, 170.042, 169.943, 169.680, 169.499,
169.047, 164.764, 153.609, 132.251, 121.499, 117.380,
100.881, 98.842, 75.981, 72.610, 72.528, 71.196, 70.744,
70.497, 68.894, 66.478, 61.816, 60.682, 52.609, 20.623,
¹
1
1
000 ps dynamics, 300 K).
20.507, 20.458, 20.351; IR (KBr, cm ): 2113, 1753, 1695,
FIDAPo Measurements. We prepared three sample solu-
1608, 1534, 1225; MALDI-TOF-MS (positive mode): m/z
833.11 [M + Na] .
p-2¤¤-Aminoacetamidophenyl-2,3,6,2¤,3¤,4¤,6¤-hepta-O-
+
tions for each condition (oligosaccharide/cation), put them into
three separate wells, and carried out the FIDAPo measurements
on each well ten times (n = 30). Some of the data, however,
largely deviated from the normal distributions, possibly suf-
fering from light scattering arising from small dust particles
and/or bubbles attached onto the glass bottoms. We therefore
removed some data within the highest and the lowest one-third
from the whole data sets to obtain reliable mP values. The mP
values in Table 1 are, therefore, the mean mP values obtained
from the data sets within the middle one-third (n = 10).
p-2¤¤-Chloroacetamidophenyl-2,3,6,2¤,3¤,4¤,6¤-hepta-O-
acetyl-β-lactoside (4).
To p-2¤¤-azidoacetamidophenyl-
2,3,6,2¤,3¤,4¤,6¤-hepta-O-acetyl-β-lactoside (3, 0.41 g) in THF
(40 mL), Pd/C (5%, 0.13 g), acetic acid (0.3 mL), and methanol
(1.0 mL) were added carefully. The resultant mixture was
vigorously stirred under hydrogen atmosphere at the ambient
temperature for overnight until complete consumption of the
substrate was confirmed through TLC analyses (toluene­ethyl
acetate = 1:1 v/v). We then removed Pd/C through filtration
and the resultant filtrate was evaporated, dried in vacuo, re-
dissolved into water, and lyophilized to give the desired
product in its acetate salt form as pale yellow powders without
any purification processes (94%). The acetic acid was essential
to eliminate side reactions (e.g. migrations of the acetyl groups
from the hydroxy groups to the highly reactive primary amino
acetyl-β-lactoside (2).
To p-aminophenyl-2,3,6,2¤,3¤,4¤,6¤-
hepta-O-acetyl-β-lactoside (1, 0.75 g) in anhydrous THF (70
mL), chloroacetyl chloride (1.5 mL) and triethylamine (2.8 mL)
were added, and the resultant mixture was stirred at room
temperature for 30 min. The mixture was then poured into iced
water, diluted with ethyl acetate, and washed repeatedly with
1
group) in this reaction. H NMR (300 MHz, CDCl3 + CD3OD,
NaHCO saturated aqueous solution. The organic layer was
r.t.): ¤ 7.54 (br s, 2H), 6.96 (br s, 2H), 5.36 (d, J = 2.7 Hz, 1H),
5.28 (t, J = 9.6 Hz, 1H), 5.16 (t, J = 8.4 Hz, 1H), 5.12 (dd,
J = 7.8 and 10.2 Hz, 1H), 5.03 (d, J = 7.5 Hz, 1H), 4.97 (dd,
J = 3.0 and 10.5 Hz, 1H), 4.55­4.48 (m, 2H), 4.19­4.07 (m,
3H), 3.96­3.37 (m, 2H), 3.83­3.76 (m, 1H), 3.43 (s, 2H), 2.17
(s, 3H), 2.11 (s, 3H), 2.08 (s, 9H), 2.07 (s, 3H), 1.98 (s, 3H);
3
dried over anhydrous MgSO , filtered, and concentrated to
4
dryness to give the desired product as pale yellow powders.
1
H NMR (300 MHz, CDCl , r.t.): ¤ 8.19 (s, 1H), 7.46 (d, J =
3
9.0 Hz, 2H), 6.98 (d, J = 9.0 Hz, 2H), 5.36 (d, J = 2.7 Hz, 1H),
5.28 (t, J = 9.0 Hz, 1H), 5.16 (dd, J = 7.5 and 9.0 Hz, 1H),
5.13 (dd, J = 7.5 and 10.5 Hz, 1H), 4.97 (dd, J = 3.3 and
10.2 Hz, 1H), 4.52­4.48 (m, 2H), 4.19 (s, 2H), 4.17­4.14 (m,
1H), 4.13­4.08 (m, 2H), 3.92­3.86 (m, 2H), 3.80­3.75 (m, 1H),
2.16 (s, 3H), 2.10 (s, 3H), 2.08 (s, 3H), 2.07 (s, 3H), 2.07 (s,
3
H), 2.06 (s, 3H), 1.97 (s, 3H); 13C NMR (75 MHz, CDCl3,
1
3
C NMR (75 MHz, CDCl , r.t.): ¤ 170.338, 170.124, 170.042,
3
169.738, 169.581, 169.088, 151.422, 135.679, 125.445,
117.528, 101.021, 98.999, 76.104, 72.676, 71.352, 70.859,
70.621, 68.985, 66.535, 61.915, 60.748, 30.224, 21.108,
¹1
20.738, 20.581, 20.458; IR (KBr, cm ): 2959, 1752, 1509,
+
r.t.): ¤ 170.798, 170.741, 170.576, 170.494, 170.190, 170.034,
1227; ESI-MS (positive mode): obs. m/z 785.2593 [M + H]
1
1
7
2
69.540, 164.238, 154.480, 132.465, 122.189, 118.062,
01.522, 99.451, 76.573, 73.218, 73.144, 71.829, 71.328,
1.130, 69.462, 66.987, 62.384, 61.208, 43.205, 21.206,
(calc. 785.2617 for C H N O ).
3
4
45
2
19
RhoLac (6).
To p-2¤¤-aminoacetamidephenyl-
2,3,6,2¤,3¤,4¤,6¤-hepta-O-acetyl-β-lactoside (4, 15 mg) in DMF
(1 mL), NaHCO3 (15 mg) and RhoNHS (4.8 mg) were added,
and the resultant mixture was incubated at the ambient temper-
ature for overnight. After the successful coupling to access 5
was confirmed through TLC (chloroform/methanol = 9/1 v/v)
¹
1
1.058, 20.927; IR (KBr, cm ): 1753, 1684, 1540, 1227;
+
MALDI-TOF-MS (positive mode): m/z 826.10 [M + Na] .
p-2¤¤-Azidoacetamidophenyl-2,3,6,2¤,3¤,4¤,6¤-hepta-O-acet-
yl-β-lactoside (3).
To p-2¤¤-chloroacetamidophenyl-
+
2
,3,6,2¤,3¤,4¤,6¤-hepta-O-acetyl-β-lactoside (2, 0.75 g) in DMF
and MALDI-TOF-MS ([M + H] = 1197.03, calc. 1197.40)
Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan | 623

analyses, the resultant mixture was then evaporated and the
residue was re-dissolved in methanol (10 mL) to which we
added aqueous ammonia (10 mL) dropwise over 3 h. After
evaporation and drying in vacuo, the residue was re-dissolved
into water and subjected to purification using a glass column
packed with HP20 (aqueous MeOH = 0, 10, 20, 30, 40, 50, 60,
dissolved in DMF (1 mL), NaHCO (15 mg) and Rho-NHS
3
(5.3 mg) were added and the resultant mixture was incubated
at the ambient temperature for overnight. After the successful
coupling was confirmed through MALDI-TOF-MS and HPLC
analyses, the resultant mixture was then evaporated, redis-
solved into water and purified through glass column packed
with HP20 (water/methanol = 10/0, 9/1, 8/2, 7/3, 6/4,
5/5, 4/6, 3/7, 2/8, 1/9, and 0/10, stepwise gradient). The
fraction (ca. 20 mL) containing the desired product with the
highest purity (Figure S1) was concentrated, dried in vacuo,
re-dissolved in water, and lyophilized to afford the desired
product in 84% yield as brown powders; ESI-MS (positive
7
0, 80, 90, 100% v/v, stepwise gradient). The fraction (ca. 20
mL) containing the desired product with the highest purity
Figure S1) was concentrated, dried in vacuo, re-dissolved
(
into water, and lyophilized to afford the desired product as
brown powders. ESI-MS (positive mode): obs. m/z 903.3261
+
[
M + H] (calc. 903.3300 for C H N O ).
45
51
4
16
+
Chloroacetamidobenzene (8). To aniline (7, 1 mL) in THF
70 mL), chloroacetyl chloride (1.6 mL) and triethylamine (3.1
mode): obs. m/z 563.2310 [M + H] (calc. 563.2295 for
(
C H N O ).
3
3
31
4
5
mL) were added and the mixture was stirred at room temper-
ature for 30 min. The mixture was then poured into iced water
and diluted with ethyl acetate. The resultant organic layer was
This research was partially supported by Special Research
Fund of Toyo University. We are also so grateful to Dr. Noriko
Kato (Olympus) for her kind support provided not only for
conducting FIDAPo measurements but also for revising this
manuscript.
washed with NaHCO saturated aqueous solution repeatedly
3
and dried over anhydrous MgSO . After removal of MgSO by
4
4
filtration, the filtrate was evaporated and dried in vacuo to give
1
the desired product in pale yellow powders. H NMR (300
Supporting Information
MHz, CDCl , r.t.): ¤ 8.26 (s, 1H), 7.55 (d, J = 7.8 Hz, 2H),
3
1
13
7
.36 (t, J = 7.8 Hz, 2H), 7.18 (t, J = 7.5 Hz, 1H), 4.19 (s, 2H);
H/ C NMR spectra, HPLC chart, and MS/MS spectra.
13
C NMR (75 MHz, CDCl , r.t.): ¤ 163.852, 136.575, 129.127,
This material is available on http://dx.doi.org/10.1246/bcsj.
20150369.
3
¹
1
1
25.272, 120.118, 42.843; IR (KBr, cm ): 1673, 1603, 1559;
+
ESI-MS (positive mode): obs. m/z 170.0399 [M + H] (calc.
70.0373 for C H ClNO).
1
8
9
References
Azidoacetamidobenzene (9).
8, 0.21 g) was dissolved in DMF (100 ml) containing sodium
azide (1.4 g) and the resultant mixture was stirred at 70 °C for
7 h. The resultant mixture was then diluted with ethyl acetate
Chloroacetamidobenzene
1
N. Hirohashi, N. Kamei, H. Kubo, H. Sawada, M.
Matsumoto, M. Hoshi, Dev. Growth Differ. 2008, 50, S221.
J. L. Dupree, A. D. Pomicter, Prostaglandins Other Lipid
Mediators 2010, 91, 118.
(
2
1
and washed with aqueous NaCl. The organic layer was dried
3
4
F. M. Platt, Nature 2014, 510, 68.
T. K. Dam, C. F. Brewer, Adv. Carbohydr. Chem. Biochem.
over anhydrous MgSO , filtered, concentrated to dryness, and
4
purified on a silica gel column (40 cm long; 3 cm i.d.; hexane/
2
010, 63, 139.
Y. Song, D. A. Withers, S. Hakomori, J. Biol. Chem. 1998,
273, 2517.
ethyl acetate = 1/0­1/1 in v/v) to give the desired product as
5
1
white powders (62%, 2 steps). H NMR (300 MHz, CDCl , r.t.):
3
¤
2
8.01 (s, 1H), 7.55 (d, J = 9.0 Hz, 2H), 7.35 (t, J = 9.0 Hz,
H), 7.16 (t, J = 7.5 Hz, 1H), 4.15 (s, 2H); 13C NMR (75 MHz,
6
S. Hakomori, K. Handa, K. Iwabuchi, S. Yamamura, A.
Prinetti, Glycobiology 1998, 8, xi.
CDCl3, r.t.): ¤ 164.452, 136.682, 129.119, 125.042, 120.019,
7
K. Iwabuchi, S. Yamamura, A. Prinetti, K. Handa, S.
Hakomori, J. Biol. Chem. 1998, 273, 9130.
A. R. Todeschini, S. Hakomori, Biochim. Biophys. Acta,
Gen. Subj. 2008, 1780, 421.
¹1
5
2.971; IR (KBr, cm ): 3261, 2103, 1673, 1556; MALDI-
+
TOF-MS (positive mode): m/z 177.10 [M + H] .
8
Aminoacetamidobenzene (10). To azidoacetamidobenzene
9
J. M. de la Fuente, S. Penadés, Tetrahedron: Asymmetry
(
9, 0.14 g) in THF (40 mL), we added Pd/C (5%, 50 mg) and
2
002, 13, 1879.
methanol (1.0 mL), and the resultant mixture was vigorously
stirred under hydrogen atmosphere at the ambient temperature
overnight until complete consumption of the substrate was con-
firmed through TLC analyses (CHCl3/methanol = 9/1 v/v).
We then removed Pd/C through filtration, and the resultant
filtrate was evaporated and dried in vacuo. The residue was
then purified on a silica gel column (40 cm long; 3 cm i.d.;
1
0
J. M. de la Fuente, P. Eaton, A. G. Barrientos, M.
Menéndez, S. Penadés, J. Am. Chem. Soc. 2005, 127, 6192.
A. J. Reynolds, A. H. Haines, D. A. Russell, Langmuir
006, 22, 1156.
11
2
1
2
K. Matsuura, K. Kobayashi, Glycoconjugate J. 2004, 21,
1
39.
1
1
3
4
N. Seah, P. V. Santacrose, A. Basu, Org. Lett. 2009, 11, 559.
Q. Chen, Y. Cui, T.-L. Zhang, J. Cao, B.-H. Han,
CHCl /methanol = 9/1 v/v) to give the desired product as
3
1
white powders in 90% yield. H NMR (300 MHz, CDCl +
3
Biomacromolecules 2010, 11, 13.
J. Zhao, Y. Liu, H.-J. Park, J. M. Boggs, A. Basu,
Bioconjugate Chem. 2012, 23, 1166.
G. L. Simpson, A. H. Gordon, D. M. Lindsay, N.
CD OD, r.t.): ¤ 9.39 (s, 1H), 7.61 (d, J = 9.0 Hz, 2H), 7.34
3
1
5
(
(
1
t, J = 7.5 Hz, 2H), 7.11 (t, J = 7.5 Hz, 1H), 3.49 (s, 2H), 1.60
1
3
s, 2H); C NMR (75 MHz, CDCl3 + CD3OD, r.t.): ¤ 171.357,
1
6
¹1
37.356, 128.757, 124.145, 119.542, 44.374; IR (KBr, cm ):
Promsawan, M. P. Crump, K. Mulholland, B. R. Hayter, T.
Gallagher, J. Am. Chem. Soc. 2006, 128, 10638.
17 T. M. Altamore, C. Fernández-García, A. H. Gordon, T.
Hübscher, N. Promsawan, M. G. Ryadnov, A. J. Doig, D. N.
1
621, 1557; ESI-MS (positive mode): obs. m/z 151.0896
+
[M + H] (calc. 151.0871 for C H N O).
8
11
2
RhoH (11).
To aminoacetamidobenzene (10, 3.1 mg)
624 | Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan

Woolfson, T. Gallagher, Angew. Chem., Int. Ed. 2011, 50, 11167.
C. Gege, A. Geyer, R. R. Schmidt, Eur. J. Org. Chem.
002, 2475.
B. Lorenz, L. Á. de Cienfuegos, M. Oelkers, E. Kriemen,
C. Brand, M. Stephan, E. Sunnick, D. Yüksel, V. Kalsani, K.
Kumar, D. B. Werz, A. Janshoff, J. Am. Chem. Soc. 2012, 134,
25 T. Hasegawa, Trends Glycosci. Glycotechnol. 2009, 21,
345.
1
8
2
26 M. Nakamura, M. Tsutsumi, Y. Ishikawa, H. Umemiya, T.
Hasegawa, K. Izawa, H. Abe, Y. Togashi, T. Kinone, S. Sekiguchi,
M. Igumi, K. Ide, T. Hasegawa, Tetrahedron 2013, 69, 3019.
27 Y. Nishida, T. Tsurumi, K. Sasaki, K. Watanabe, H. Dohi,
K. Kobayashi, Org. Lett. 2003, 5, 3775.
1
9
3
326.
2
0
N. Jayaraman, K. Maiti, K. Naresh, Chem. Soc. Rev. 2013,
2, 4640.
Y. Luo, F. Barbault, C. Gourmala, Y. Zhang, F. Maurel, Y.
Hu, B. T. Fan, J. Mol. Model. 2008, 14, 901.
28 A. Murata, S. Arai, S.-I. Yoon, M. Takabayashi, M. Ozaki,
S. Takeoka, Bioorg. Med. Chem. Lett. 2010, 20, 6905.
29 T. Abe, K. Goda, K. Futami, Y. Furuichi, Nucleic Acids
Res. 2009, 37, e56.
4
2
1
2
2
2
3
T. Hasegawa, T. Sasaki, Chem. Commun. 2003, 978.
T. Hasegawa, M. Numata, M. Asai, M. Takeuchi, S.
30 mPr-Rhos (Rhos = RhoLac or RhoH): relative mP values of
Rhos (mPRhos/mP0-Rhos) in which mPRhos and mP0-Rhos indicate mP
values of Rhos in the media containing the oligosaccharides and/
or the cations and that in pure water, respectively.
Shinkai, Tetrahedron 2005, 61, 7783.
A. Otsuka, K. Sakurai, T. Hasegawa, Chem. Commun.
009, 5442.
2
4
2
Bull. Chem. Soc. Jpn. 2016, 89, 617–625 | doi:10.1246/bcsj.20150369
© 2016 The Chemical Society of Japan | 625