Coupling a Natural Receptor Protein with an Artificial Receptor to Afford a Semisynthetic Fluorescent Biosensor

Article abstract of DOI:10.1021/ja035631i

An artificial receptor and a signal transducer have been engineered on a lectin (saccharide-binding protein) surface by a post-photoaffinity labeling modification method. Saccharide binding can be directly and selectively read out by the fluorescence chan

Full text of DOI:10.1021/ja035631i

Published on Web 12/20/2003
Coupling a Natural Receptor Protein with an Artificial
Receptor to Afford a Semisynthetic Fluorescent Biosensor
Eiji Nakata,† Tsuyoshi Nagase,† Seiji Shinkai,† and Itaru Hamachi*,†,‡,§
Contribution from PRESTO (Organization and Function, JST), Institute for Materials Chemistry
and Engineering (IMCE), and Department of Chemistry and Biochemistry, Graduate School of
Engineering, Kyushu UniVersity, Fukuoka 812-8581 Japan
Received April 14, 2003; E-mail: itarutcm@mbox.nc.kyushu-u.ac.jp
Abstract: An artificial receptor and a signal transducer have been engineered on a lectin (saccharide-
binding protein) surface by a post-photoaffinity labeling modification method. Saccharide binding can be
directly and selectively read out by the fluorescence changes of the fluorophore via photoinduced electron
transfer (PET) mode. Fluorescence titration with various saccharides reveals that molecular recognition by
the artificial receptor is successfully coupled to the native binding site of the lectin, producing a novel
fluorescent saccharide biosensor showing modulated specificity and enhanced affinity. Designed cooper-
ativity between artificial and native molecular recognition modules was quantitatively demonstrated by the
comparison of the binding affinities, and it represents a new strategy in molecular recognition. By using
appropriate artificial receptors and various native lectins, this approach may provide many new semisynthetic
biosensors for saccharide derivatives such as glycolipids and glycopeptides/proteins. An extended library
of lectin-based biosensors is envisioned to be useful for glycome research, a newly emerging field of the
post-genomic era.
Introduction
biosensors. In this report, we show that the introduction of an
artificial sugar-binding site into a native lectin greatly enhances
The monitoring of chemical substances of biological impor-
tance is required in many fields of biological science and in
medical diagnosis.1-5 In glycome research, which seeks to
elucidate the complicated biological functions of saccharides
in the post-genomic era, it is essential to sense many saccharides
in a high-throughput manner.6-8 Although native lectins, a
family of saccharide-binding proteins, have been used to
qualitatively analyze carbohydrates, their low selectivity prevents
the accurate high-throughput quantitation and assay of saccha-
rides. Synthetic chemosensors may overcome these limitations,
but they have not yet been successfully applied to the detection
of complicated oligosaccharides. Advanced mass spectroscopic
methods may be useful for the structural analysis of complex
oligosaccharides,9 but these are not so suitable to rapid and
convenient sensing of saccharides in a high-throughput manner.
We propose that the rational hybridization of naturally occurring
receptor proteins with artificial receptor molecules represents a
practical, new approach for the construction of sophisticated
the binding affinity of the lectin toward specific saccharides
having interaction moieties to both the lectin and the artificial
receptor. As a result, one can engineer a fluorescent semisyn-
thetic lectin with modulated saccharide selectivity.
Results and Discussion
Molecular Design of Phenylboronic Acid-Appended Co-
nA.10 We recently developed a new method (post-photoaffinity
labeling modification: P-PALM) to incorporate a unique
benzylthiol group proximal to the sugar-binding pocket of a
natural lectin11 (Concanavalin A (ConA), an R-mannoside or
R-glucoside binding protein, was used in this study12). The
benzylthiol was utilized as a reactive tag for fluorescent probes
(Scheme 1). Using environmentally sensitive fluorophores (such
as the Dansyl derivative, IAEDANS) as a signal transducer,
we could sense R-mannoside or R-glucoside derivatives by a
(10) (a) Hamachi, I.; Tajiri, Y.; Shinkai, S. J. Am. Chem. Soc. 1994, 116, 7437.
(b) Hamachi, I.; Tajiri, Y.; Nagase, T.; Shinkai, S. Chem.-Eur. J. 1997,
3, 1025.
† Department of Chemistry and Biochemistry.
‡ PRESTO (Organization and Function, JST).
(11) (a) Hamachi, I.; Nagase, T.; Shinkai, S. J. Am. Chem. Soc. 2000, 122, 12065.
(b) Nagase, T.; Shinkai, S.; Hamachi, I. Chem. Commun. 2001, 229. (c)
The conventional peptide digestion, HPLC separation, and tandem mass
spectroscopy indicate that the photolabeling site can be assigned to Y100,
a proximal amino acid in the saccharide-binding pocket. It is clear that the
photoaffinity labeling process of the present P-PALM method is remarkably
site-selective on a protein surface. Nagase, T.; Nakata, E.; Shinkai, S.;
Hamachi, I. Chem.-Eur. J. 2003, 9, 3660.
(12) The crystal data of ConA complexed with Man-tri showed the extended
interaction between Man-tri and the ConA surface using all the OH groups
of Man-tri and several amino acid residues of the ConA surface. The
lessened flexibility of the branched Man units of Man-tri due to such an
extended interaction may be a factor for no sensing to Man-tri. Lis, H.;
Sharon, N. Chem. ReV. 1998, 98, 637.
§ Institute for Materials Chemistry and Engineering.
(1) Lavigne, J. J.; Anslyn, E. V. Angew. Chem., Int. Ed. 2001, 40, 3118.
(2) Gu, L.; Braha, O.; Conlan, S.; Cheley, S.; Bayley, H. Nature 1999, 398,
686.
(3) Bayley, H.; Cremer, P. S. Nature 2001, 413, 226.
(4) Benson, D. E.; Conrad, D. W.; de Lorimier, R. M.; Trammell, S. A.;
Hellinga, H. W. Science 2001, 293, 1641.
(5) Giuliano, K. A.; Taylor, D. L. Trends Biotechnol. 1998, 16, 135.
(6) Bertozzi, C. R.; Kiessling, L. L. Science 2001, 291, 2357.
(7) Roseman, S. J. Biol. Chem. 2001, 276, 41527.
(8) Dwek, R. A. Chem. ReV. 1996, 96, 683.
(9) Harvey, J. D. Mass Spectrom. ReV. 1999, 18, 349.
9
490
J. AM. CHEM. SOC. 2004, 126, 490-495
10.1021/ja035631i CCC: $27.50 © 2004 American Chemical Society

Coupling a Natural Receptor with an Artificial Receptor
A R T I C L E S
Scheme 1. Molecular Structure of a P-PALM Reagent, IAEDANS, and APET (a) and P-PALM Scheme for Semisynthesis of the
Fluorescent Biosensor Discussed in This Study (b)
decrease in fluorescence intensity. This was ascribed to the
DANS moiety being ejected upon saccharide binding, causing
the fluorophore to move from the hydrophobic binding pocket
to the bulk water. The slight red-shift of the fluorescence and
decrease in the fluorescence anisotropy upon saccharide com-
plexation reasonably supported such a mechanism. Importantly,
the sugar selectivity and binding affinity of the IAEDANS-ConA
evaluated by the fluorescence titration are almost identical to
those of native ConA determined by isothermal calorimetry.13
In addition to the simple fluorescent probe, an artificial
receptor can be newly introduced to the proximity of the sugar-
binding pocket using the benzylthiol tag attached by P-PALM.
Phenylboronic acid (PBA) is capable of binding to 1,2- or 1,3-
diol derivatives. According to previous reports by us and
others,14 we equipped PBA with a fluorescent transducer motif,
in which a photoinduced electron transfer (PET)-type of readout
mode can operate upon complexation with a corresponding diol
(Figure 1a). As shown in Scheme 2, bromoacetyl group was
introduced to the receptor molecule (APET-Br) as a reactive
site to ConA. 9,10-Bis-N-methyaminomethyl-anthracene was
monoalkylated with o-bromomethyl-PBA, and the remaining
methylamino site was acylated with bromoacetyl bromide to
yield APET-Br. The resultant APET-Br was incubated with SH-
ConA to afford APET-ConA via a nucleophilic substitution
reaction on the ConA surface (Scheme 1b). APET-ConA was
purified through gel chromatography (TOYO PEARL), and the
purity was confirmed by analytical reversed-phase HPLC, in
which a single ConA band having both UV absorbance and
fluorescent emission was detected. In UV-visible spectroscopy
of the purified APET-ConA, a strong absorbance at 257 nm
and three peaks at 357, 375, and 397 nm are characteristic of
the anthracene moiety (Figure 1b). The emission at 407, 432,
and 454 nm and excitation spectra at 357, 375, and 397 nm of
APET-ConA indicate a clear mirror image (Figure 1c), and the
excitation spectrum agrees well with the absorption spectrum.
These results reveal the successful attachment of the APET
moiety to ConA.
Fluorescence Sensing and Saccharide Selectivity of APET-
ConA. Prior to saccharide sensing, we conducted the fluores-
cence pH titration of APET-ConA with and without a saccharide
(Figure 2a). The emission due to the anthracene moiety lessened
in the basic pH region. This is ascribed to the emission
quenching by the deprotonated tertiary amine group of APET.
It is clear that addition of saccharide induces a basic pKa shift
of the tertiary amine, which results in retaining the fluorescence
intensity at neutral pH. Consequently, the fluorescence of the
anthracene moiety in APET-ConA increases following saccha-
ride addition (Figure 2b). This is in sharp contrast to the
decreased fluorescence of IAEDANS-ConA induced by sac-
charide binding (Figure 2c). Fluorescence intensification by
sugars is direct evidence that the PBA site binds to a diol unit
of an appropriate sugar and that the PET mechanism works well
on a ConA surface. Table 1 summarizes the association
constants of engineered ConA for various saccharides compared
to native ConA. Figure 3 shows among monosaccharides, it was
(13) Dam, T. K.; Brewer, C. F. Chem. ReV. 2002, 102, 387.
(14) (a) James, T. D.; Sadanayake, K. R. A. S.; Iguchi, R.; Shinkai, S. J. Am.
Chem. Soc. 1995, 117, 8982. (b) James, T. D.; Sadanayake, K. R. A. S.;
Shinkai, S. Nature 1995, 375, 345. (c) James, T. D.; Sadanayake, K. R. A.
S.; Shinkai, S. Angew. Chem., Int. Ed. Engl. 1996, 35, 1920. (d) Bielecki,
M.; Eggert, H.; Norrild, J. C. J. Chem. Soc., Perkin Trans. 2 1999, 449.
(e) Norrild, J. C.; Eggert, H. J. Am. Chem. Soc. 1995, 117, 1479.
9
J. AM. CHEM. SOC. VOL. 126, NO. 2, 2004 491

A R T I C L E S
Nakata et al.
Figure 1. (a) A calculated (Insight II discover) 3D structure of APET-ConA and illustration of PET mechanism of saccharide sensing by APET-ConA. (b)
UV-visible spectroscopy of the APET-ConA. (c) Excitation and emission spectra of APET-ConA. [APET-ConA] ) 0.7 µM, 50 mM HEPES buffer (pH
7.5), 5 mM CaCl2, 0.1 M NaCl, T ) 15 ( 1 °C, λex ) 375 nm, λem ) 425 nm.
Scheme 2. Synthetic Route of APET-Br
Table 1. Comparison of the Association Constants of
Semisynthetic ConA with Those of Native ConA and PBA
d
log K
e
APET-ConA
log KAPET-ConA
PBA
log KPBA
IAEDANS-ConA
log K
IAEDANS-ConA
c
saccharide
native ConA
Me-R-Man
Me-R-Glc
Fru
Glucitol
R-1,3-Man-bi
Isomal
a
a
a
a
2.28
3.68
a
a
a
a
3.89
3.20
a
4.04
3.48
a
2.85
3.52
a
4.08
a
a
b
3.76
3.11
2.76
a
4.48
3.23
3.20
a
Mal
Cel
a
Pal
5.18
6.84
a
5.26
5.81
2.76
4.51
a
4.56
a
3.20
3.00
5.40
3.00
5.45
b
b
5.40
b
5.30
Maltitol
Man-tri
Maltotriitol
Man-tetra
a
Precise values cannot be determined because of low affinity (log K <
found that D-fructose (Fru) and D-glucitol (Glucitol) are
moderately bound to APET-ConA, showing good agreement
with the binding affinity of PBA. Me-R-Man and Me-R-Glc,
good ligands to native ConA, cannot be sensed by APET-ConA
because of their poor affinity to the PBA site, while they can
be most efficiently sensed by IAEDANS-ConA. These data
indicate that any fluorescence signal from APET-ConA must
be produced by a saccharide-binding event at the PBA site, not
at the native ConA pocket, because the signal transducer is
directly coupled to PBA via the PET mechanism (Figure 1a).
Figure 3 shows the saccharide structures used in this study.
Most significantly, APET-ConA shows higher affinity and
selectivity for specific oligosaccharides over monosaccharides.
b
c
2). Not reported previously. Reported values determined by isothermal
titration calorimetry.13 The averaged value of at least three independent
d
titration experiments. Our previous work.11c
e
Maltitol, a disaccharide consisting of D-Glc and D-glucitol
connected by an R-1,4-linkage, is most sensitively detected in
the sub-micromolar range by APET-ConA. Fluorescence titra-
tion determined an association constant (log K ) 6.84) (Figure
2c), which is approximately 3 orders of magnitude tighter
binding than that by native ConA. Similarly, palatinose, a
disaccharide (Pal: D-Glc-R-1,5-D-Fru), which is a poor ligand
for native ConA, is effectively sensed (at the micromolar level)
by APET-ConA. We found that the association constants for
both disaccharides are in good agreement with the value of the
9
492 J. AM. CHEM. SOC. VOL. 126, NO. 2, 2004

Coupling a Natural Receptor with an Artificial Receptor
A R T I C L E S
Figure 2. (a) Fluorescence pH titration profiles of APET-ConA in the absence (O) and presence (b) of 2 µΜ Maltitol. [APET-ConA] ) 0.5 µΜ, 5 mM
CaCl2, 0.1 M NaCl, T ) 15 ( 1 °C, λex ) 375 nm, λem ) 425 nm. (b) Fluorescence spectral change of APET-ConA upon addition of Maltitol (0-10 µM).
[APET-ConA] ) 0.7 µM, 50 mM HEPES buffer (pH 7.5), 5 mM CaCl2, 0.1 M NaCl, T ) 15 ( 1 °C, λex ) 375 nm, λem ) 425 nm. (c) Fluorescence
titration plots of the relative intensity (I/I0) versus saccharide concentration (log[saccharide]) for APET-ConA and IAEDANS-ConA: APET-ConA (Maltitol
(b), Pal ([), Isomal (9)) and IAEDANS-ConA (Maltitol (O), Pal (]), Isomal (0)). (d) Fluorescence titration of APET-ConA by addition of Maltitol in the
absence (b) and presence (O) of 10 mM of Me-R-Man. [APET-ConA] ) 0.7 µΜ, 50 mM HEPES buffer (pH 7.5), 5 mM CaCl2, 0.1 M NaCl, T ) 15 (
1 °C, λex ) 375 nm, λem ) 425 nm.
Figure 3. Saccharide structures used in this study. The abbreviation is depicted below each structure.
sum for each monosaccharide fragment, that is, log KAPET-ConA
(Maltitol) ) log KIAEDANS-ConA (Me-R-Glc) + log KPBA (Glu-
citol) and log KAPET-ConA (Pal) ) log KIAEDANS-ConA (Me-R-
Glc) + log KPBA (D-Fru) (see Table 1). This implies that two
distinct binding sites composed of an artificial framework and
a native protein scaffold successfully cooperate in APET-ConA,
so as to enhance the affinity for the specific disaccharide by
102-104-fold relative to each monosaccharide fragment. The
cooperative effect of Maltitol binding was confirmed by a
competition assay (Figure 2d). In the presence of excess Me-
9
J. AM. CHEM. SOC. VOL. 126, NO. 2, 2004 493

A R T I C L E S
Nakata et al.
R-Man (10 mM), the apparent titration curve flattened (an
apparent association constant (log K ) 4.76)), whereas Me-â-
Glc did not show such an inhibition effect. The lessened affinity
can be reasonably attributed to the masking of the natural
binding pocket of ConA by Me-R-Man and not by Me-â-Glc.
It is clear that the hybridization of an artificial receptor to ConA
successfully allows us to modulate the saccharide selectivity
of a natural lectin by enhancing the affinity of the semisynthetic
receptor toward the specific saccharides.
On the other hand, cellobiose (Cel: D-Glc-â-1,4-D-Glc), a
â-linked disaccharide, is not fluorescently sensed, probably
because of the preference of the R-configuration over â found
in the native ConA pocket. In addition, regioisomers among
the disaccharides composed of two D-Glc units (R-configuration)
can be discriminated by APET-ConA, that is, isomaltose
(Isomal) with a 1,6-connection is selectively detected but not
maltose (Mal) with a 1,4-connection. In Isomal bearing a 1,6-
linkage, the second glucose unit can isomerize to a furanose
ring, 1,2-diol or 1,3-diol of which is favorable to binding by
PBA,14d,e whereas such a pyranose/furanose ring isomerization
cannot occur in Mal bearing 1,4-linkage. Additionally, it seems
that the more flexible motion of the second glucose unit through
the 1,6-connection is readily bound to the PBA site of APET-
ConA. Among the more complicated oligosaccharides, 1,3- and
1,6-Mannotriose (Man-tri), which is the strongest ligand for
native ConA and thus is detected by IAEDANS-ConA, is not
sensed by APET-ConA. This is presumably due to all the OH
groups of Man-tri interacting with the ConA surface as well as
the ConA pocket12 and being unavailable to interact with the
PBA unit. Consistent with this explanation, we found that Man-
tetra, which bears an additional mannose unit to Man-tri, can
be sensed by APET-ConA. Although PBA itself shows the poor
affinity toward Isomal and Man-tetra (log K is less than 2),
APET-ConA gains a moderate affinity to these saccharides. This
may be explained by the proximity effect of PBA to the native
binding pocket of ConA, that is, the PBA site close to the ConA
pocket effectively acts as a subsite for saccharide binding, so
as to cause the PET fluorescence sensing, even though the net
binding affinity is not remarkably enhanced.
Scope and Limitation. In conclusion, the rational coupling
of an artificial receptor with a native receptor protein could
evolve toward a novel semisynthetic receptor and sensor. As
the seminal work by Norrild previously pointed out, PBA
preferentially forms a tight complex with cis-1,2-furanose type
of diol involving the C1 hydroxyl of sugars,14d and the
hybridization of PBA to lectin has limited potential in the
practical application. However, it is important that the present
proof-of-principle experiment clearly demonstrates the coopera-
tive action between an artificial and a native receptor. By using
alternative approaches, host-guest chemistry has produced a
number of artificial receptors such as crown ether, cyclodextrin,
cyclophane, and others based on hydrogen-bonding force,
etc.,15,16 and biochemical studies have already unveiled various
natural receptor proteins for the last several decades.17 Therefore,
it is envisioned that the designed hybrid of these two families
will facilitate the generation of semisynthetic receptors with
Figure 4. Summary of the binding constants evaluated by fluorescence
change of various bio- or chemosensors in this study.
enhanced affinity and selectivity, not only toward a variety of
simple carbohydrates, but also for other biological substances
such as glycolipids, glycopeptides/proteins, and bioactive
hormones, etc.6-8
In cases where targets are structurally complicated and
diverse, such as saccharide derivatives, it is generally difficult
to determine the target structure by a typical one-to-one
recognition. Instead, the pattern recognition is advantageous to
such a case as Anslyn pointed out.1 However, to carry out the
pattern recognition, a set of receptors bearing various selectivity
is required. In the present case, for example, IAEDANS-ConA
cannot distinguish Man-tri from Man-tetra and APET-ConA can
scarcely distinguish Man-tetra from Maltitol. However, compar-
ing the sensing pattern of IAEDANS-ConA and APET-ConA
enables one to distinguish the three saccharides (see the sensing
pattern shown in Figure 4). Although this is still primitive, we
can expect a more precise pattern recognition when a rich variety
of fluorescent biosensors will be prepared according to our
approach. We are currently investigating along this line.
Experimental Section
General Methods. Matrix-assisted laser desorption/ionization time-
of-flight (MALDI-TOF) mass spectrometry (MS) was recorded on PE
Voyager DE-RP, where sinapinic acid (SA) was used as a matrix. UV-
visible spectra were obtained on a Hitachi U-3000 spectrometer.
Fluorescence spectra were recorded on a Hitachi F-4500 spectrometer.
ConA was purchased from Funakoshi and used without further
purification. IAEDANS was purchased from Molecular Probes. The
other chemical reagents were purchased from Aldrich or Tokyo
Chemical Industry.
Photoaffinity Label of ConA with P-PALM Reagent. This process
was conducted according to our previous method.11c The second fraction
was collected and mainly used in the following study.18
Preparation of APET-ConA. To a solution of the labeled ConA
(a heterodimer consisted of the labeled monomoer and the unlabeled
(native) monomer, that is, the second fraction in the affinity column)
(20 µM, 10 mL) in 50 mM phosphate buffer (pH 7.5) was added Tris-
carboxyethylphosphine (TCEP) (a final concentration of TCEP is 100
(18) Since the fluorescence signal from APET-ConA is produced by a saccharide-
binding event at the PBA site, not at the native ConA pocket (Figure 1a),
the binding constants are evaluated on the basis of the saccharide binding
at the PBA-modified ConA site. Although the other unmodified binding
site (i.e., without PBA) may bind saccharides with a weak affinity, it does
not affect the determination of binding constants under conditions of the
excess amount of saccharide (for Benesi-Hildebrandt analysis). In case
of the strong binding affinity observed at the almost stoichiometric amount
of the saccharide, the affinity may be slightly underestimated because the
native binding pocket binds the corresponding saccharide without fluores-
cence sensing so as to decrease the free concentration of the saccharide.
(15) Lehn, J. M. Chem.-Eur. J. 1999, 5, 2455.
(16) Rowan, S. J.; Cantrill, S. J.; Cousins, G. R. L.; Sanders, J. K. M.; Stoddart,
J. F. Angew. Chem., Int. Ed. 2002, 41, 898.
(17) Subrahmanyam, S.; Piletsky, S. A.; Turner, A. P. F. Anal. Chem. 2002,
74, 3942.
9
494 J. AM. CHEM. SOC. VOL. 126, NO. 2, 2004

Coupling a Natural Receptor with an Artificial Receptor
A R T I C L E S
µM: 10 equiv for labeled ConA), and the solution was incubated for
5 h at 4 °C under nitrogen atmosphere. The reaction solution was
purified by gel filtration chromatography (Bio-Gel P-30, 2 cm × 20
cm, eluent: 50 mM phosphate buffer (pH 7.5)) to give SH-ConA (12
µM, 16 mL) in quantitative yield. The thiol content in SH-ConA was
determined by the Ellman method, showing that one ConA dimer has
one SH group. MALDI-TOF MS (SA) demonstrated the cleavage of
the thiomannose unit: m/z, calcd 25 787, obsd 25 796. The SH-ConA
was quickly subjected to the next modification. To a solution of SH-
ConA (20 µM, 5 mL) in 50 mM phosphate buffer (pH 7.5), 100 mM
D-Glc, APET-Br (20 equiv to one SH group) in dimethyl formamide
(DMF) (500 µL) was added slowly at 0 °C and incubated for 12 h at
r.t. in the dark. After incubation, the solution was loaded on TOYO
PEARL (1 cm × 15 cm), eluted with the buffer (C) (50 mM HEPES
buffer (pH 7.5), 5 mM CaCl2, 0.1 M NaCl). The protein fraction was
collected, concentrated by ultrafiltration, and dialyzed against the buffer
(C) to afford APET-ConA (12 µM, 8 mL) in quantitative yield. APET-
ConA displays the absorbance at 280 nm due to the aromatic ring of
ConA and 375 nm due to the anthracene unit. The emission maximum
of APET-ConA at 407, 432, and 454 nm and the excitation maximum
at 357, 375 and 397 nm are both characteristic of the anthracene
chromophore. The labeling efficiency was estimated from the ratio of
the corresponding absorbance to be 0.75-0.90.
Fluorescence Titration. Saccharide solution was added dropwise
to a 0.7 µM APET-ConA solution [50 mM HEPES buffer (pH 7.5)
containing 5 mM CaCl2 and 0.1 M NaCl at 15 ( 1 °C] or to a 0.2 µM
IAEDANS-ConA solution [50 mM HEPES buffer (pH 7.0) containing
1 mM CaCl2, 1 mM MnCl2 and 0.1 M NaCl at 15 ( 1 °C], and
fluorescence spectra were measured. The slit widths for the excitation
and emission were set to 10 nm and 10 nm, respectively. The excitation
and emission wavelengths used were λex ) 340 nm and λem ) 484 nm
(IAEDANS-ConA) or λex ) 375 nm and λem ) 425 nm (APET-ConA),
respectively. Fluorescence titration curves were analyzed with the
nonlinear least-squares curve-fitting method or the Benesi-Hildebrandt
plot to give the association constants (log K) for various saccharides.18
pH Titration Experiments in the Absence and Presence of
Saccharides. The fluorescence of APET-ConA (0.5 µM) dissolved in
aqueous solution with various pH was measured. We adjusted the pH
by addition of aqueous HCl or aqueous NaOH to the unbuffered APET-
ConA solution. The optimal pH for the saccharide titration was
determined at pH 7.5.
Synthesis of APET-Br. The following solvents were distilled under
nitrogen atmosphere prior to usage: MeOH was dried over Mg and I2.
Unless otherwise stated, all of the air- or moisture-sensitive reactions
were performed under nitrogen atmosphere.
PBA and 9,10-bis[(methylamino)methyl]anthracene were prepared
according to literature procedure.14a
9-((N-Methyl-N-(o-boronobenzyl)amino)methyl)-10-((methylami-
no)methyl)anthracene (2). 2,2-Dimethylpropane-1,3-diyl(o-(bromom-
ethyl)phenyl)boronate (760 mg, 3.0 mmol), 9,10-bis-((methylamino)-
methyl)anthracene (880 mg, 3.33 mmol, 1.1 equiv), and potassium
carbonate (1.24 g, 3.0 mmol) in anhydrous THF (40 mL) were refluxed
for 3 h. The mixture was filtered, the solvent was removed in vacuo,
and the residue was subjected to column chromatography (CHCl3/
MeOH (1% TEA): a linear gradient from 50/1 to 6/1 (v/v)) to give
1
compound 2 as a yellow powder in 18% yield (210 mg). H NMR
(CDCl3,TMS,250 MHz): δ/ppm; 8.34 (d, J ) 8.5 Hz, 2H, Ar-H), 8.13
(d, J ) 8.5 Hz, 2H, Ar-H), 8.11 (d, J ) 8.5 Hz, 1H, Ar-H), 7.6-7.4
(m, 7H, Ar-H), 4.68 (s, 2H, -CH2-), 4.50 (s, 2H, -CH2-), 3.92 (s,
2H, -CH2-), 2.65 (s, 3H, -CH3), 2.21 (s, 3H, -CH3). ESI-TOF MS:
+
calcd for C25H28BN2O2 ([M + H]+), 399.22; found, 399.09.
9-((N-Methyl-N-(o-boronobenzyl)amino)methyl)-10-((N-methyl-
N-(bromoacetyl) amino) methyl)anthracene (APET-Br). The com-
pound 2 (30 mg, 75.3 mmol), DIEA (97 mg, 753 mmol), and
bromoacetylbromide (30 mg, 150 mmol, 2 equiv) in anhydrous
dichloromethane (10 mL) were stirred at 0 °C for 30 min. The solvent
was removed in vacuo, and the residue was subjected to column
chromatography (CHCl3/MeOH: a linear gradient from 20/1 to 10/1
(v/v)) to give APET-Br as a yellow powder in 82% yield (32 mg). 1H
NMR (CDCl3, TMS, 250 MHz): δ/ppm; 8.29 (d, J ) 8.5 Hz, 2H,
Ar-H), 8.16 (d, J ) 8.5 Hz, 2H, Ar-H), 8.05 (d, 1H, Ar-H), 7.60-
7.35 (m, 7H, Ar-H), 5.68 (s, 2H, -CH2-), 4.53 (s, 2H, -CH2-),
4.53 (s, 2H, -CH2-), 4.06 (s, 2H, Br-CH2-), 3.94 (s, 2H, -CH2-),
2.52 (s, 3H, N-CH3), 2.24 (s, 3H, N-CH3). SIMS-MS (matrix: NBA
and glycerol): calcd for C30H32BBrN2O4+ ([M + glycerol - 2H2O]+),
574.576; obsd, 574.576.
JA035631I
9
J. AM. CHEM. SOC. VOL. 126, NO. 2, 2004 495