NMR Studies of Conformations of N-Dansyl-L-leucine-Appended and N-Dansyl-D-leucine-Appended β-Cyclodextrin as Fluorescent Indicators for Molecular Recognition

Article abstract of DOI:10.1021/jo960425x

The structures of N-dansyl-L-leucine-appended β-cyclodextrin (1) and N-dansyl-D-leucine-appended β-cyclodextrin (2) were estimated by the combined use of 1D and 2D NMR techniques (1D and 2D TOCSY, ROESY, PFG MQF-COSY, PFG HSQC, and NOE difference spectra). The dansyl moiety of 2 is included in its own cavity more deeply than that of 1. The difference in the properties between 1 and 2 was interpreted by the difference in the inclusion depth of the dansyl moieties. Their conformational changes upon addition of 1-adamantanol were also studied by NMR. The chemical shifts and patterns of ¹H resonances for the protons of the dansyl, leucine, and cyclodextrin parts were changed upon addition of the guest. These changes indicated the exclusion of the dansyl moiety from the cyclodextrin cavity to bulk water upon addition of the guest.

Full text of DOI:10.1021/jo960425x

J . Org. Chem. 1997, 62, 1411-1418
1411
NMR Stu d ies of Con for m a tion s of N-Da n syl-L-leu cin e-Ap p en d ed
a n d N-Da n syl-D-leu cin e-Ap p en d ed â-Cyclod extr in a s F lu or escen t
In d ica tor s for Molecu la r Recogn ition
Hiroshi Ikeda,* Michiei Nakamura, Nobuyuki Ise, Fujio Toda,† and Akihiko Ueno*
Department of Bioengineering, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology,
4259 Nagatsuta-cho, Midori-ku, Yokohama 226, J apan
Received March 1, 1996 (Revised Manuscript Received J anuary 2, 1997X)
The structures of N-dansyl-L-leucine-appended â-cyclodextrin (1) and N-dansyl-D-leucine-appended
â-cyclodextrin (2) were estimated by the combined use of 1D and 2D NMR techniques (1D and 2D
TOCSY, ROESY, PFG MQF-COSY, PFG HSQC, and NOE difference spectra). The dansyl moiety
of 2 is included in its own cavity more deeply than that of 1. The difference in the properties
between 1 and 2 was interpreted by the difference in the inclusion depth of the dansyl moieties.
Their conformational changes upon addition of 1-adamantanol were also studied by NMR. The
chemical shifts and patterns of 1H resonances for the protons of the dansyl, leucine, and cyclodextrin
parts were changed upon addition of the guest. These changes indicated the exclusion of the dansyl
moiety from the cyclodextrin cavity to bulk water upon addition of the guest.
In tr od u ction
of the pendant groups of some modified CDs by the
combined use of 1D and 2D NMR techniques.14-17
We have been studying fluorophore-appended CDs as
fluorescent indicators for molecular recognition.5,7,17 The
fluorescence intensities of the fluorophore-appended CDs
are decreased or increased upon addition of a guest
molecule. The degree of this guest-induced variation of
the fluorescence intensity depends on the polarity, shape,
and size of the guest molecule, and spectroscopically inert
molecules can also be detected by these indicators by
changes in the fluorescence intensity. We synthesized
N-dansyl-L-leucine-appended â-CD (1) and N-dansyl-D-
leucine-appended â-CD (2) as new fluorescent indicators
for molecular recognition (dansyl: 5-(dimethylamino)-1-
naphthalenesulfonyl). Their properties were significantly
different from each other, although the structural dif-
ference between them was only the enantiomeric config-
uration of their amino acid residues.18 For example, the
binding ability of 1 was over twice that of 2. The degree
of guest-induced variation in the fluorescence intensity
of 1 was greater than that of 2, whereas the former
fluorescence intensity in the absence of the guest was
smaller than the latter. These differences would be
derived from the difference in the orientation of the
dansyl moiety. Therefore, it is necessary to elucidate the
orientation of the pendant groups in order to understand
the difference in their properties. In this paper, we
describe the conformational analyses of N-dansyl-L-
leucine-appended and N-dansyl-D-leucine-appended â-CD
using NMR techniques.
In supramolecular chemistry, there has been a great
deal of interest in molecular recognition by cyclodextrins
(CDs) and their derivatives.1-9 Detailed understanding
of the structure-function relationships of modified CDs
will be one of the effective ways to know how to design
more excellent modified CDs; however, only a few at-
tempts have been made to determine the detailed struc-
tures of the modified CDs including the orientation of
the pendant groups. The current NMR technique is a
powerful tool for the structural and conformational
analysis of the modified CDs.10-17 On this basis, we have
been studying the structural and conformational analyses
of the modified CDs and have determined the orientation
* To whom correspondence should be addressed.
†Present address: Tokyo Polytechnic College, 2-32-1, Ogawanisi,
Kodaira 187, J apan.
X Abstract published in Advance ACS Abstracts, February 1, 1997.
(1) Bender, M. L.; Komiyama, M. Cyclodextrin Chemistry;
Springer-Verlag: Berlin, 1978.
(2) Szejtli, S. Cyclodextrin Technology; Kluwer Academic Publish-
ers: Dordrecht, 1988.
(3) Ducheˆne, D., Ed. New Trends in Cyclodextrin and Derivatives;
de Sante: Paris, 1991.
(4) Breslow, R.; Halfon, S.; Zhang, B. Tetrahedron 1995, 51, 377-
388.
(5) Ueno, A. In Fluorescent Chemosensors for Ion and Molecule
Recognition; Czarnik, A. W., Ed.; ACS Symposium Series 538; Ameri-
can Chemical Society: Washington, DC, 1993; pp 74-84.
(6) Kuwabara, T.; Nakamura, A.; Ueno, A.; Toda, F. J . Phys. Chem.
1994, 98, 6297-6303.
(7) Wang, Y.; Ikeda, T.; Ikeda, H.; Ueno, A.; Toda, F. Bull. Chem.
Soc. J pn. 1994, 67, 1598-1607.
(8) Ueno, A.; Chen, Q.; Suzuki, I.; Osa, T. Anal. Chem. 1992, 64,
1650-1655.
Exp er im en ta l Section
(9) Corradini, R.; Dossena, A.; Impellizzeri, G.; Maccarrone, G.;
Marchelli, R.; Rizzarelli, E.; Sartor, G.; Vecchio, G. J . Am. Chem. Soc.
1994, 116, 10267-10274.
Ma ter ia ls. â-Cyclodextrin was kindly gifted by Nihon
Shokuhin Kako Co., Ltd. All chemicals were reagent grade
and were used without further purification unless otherwise
noted. Distilled water and dimethyl sulfoxide as solvents for
spectroscopy were special grade for fluorometry (Uvasol) from
Kanto Chemicals. Deuterium oxide, with an isotopic purity
of 99.95%, was purchased from the Merck Co.
(10) Inoue, Y. Annu. Rep. NMR Spectrosc. 1993, 27, 59-101.
(11) Yamamoto, Y.; Inoue, Y. J . Carbohydr. Chem. 1989, 8, 29-46.
(12) Deschenaux, R.; Harding, M. M.; Ruch, T. J . Chem. Soc., Perkin
Trans. 2 1993, 1251-1258.
(13) Djeda¨ıni-Pilard, F.; Azaroual-Bellanger, N.; Gosnat, M.; Vernet,
D.; Perly, B. J . Chem. Soc., Perkin Trans. 2 1995, 723-730.
(14) Ikeda, H.; Nagano, Y.; Du Y-q.; Ikeda, T.; Toda, F. Tetrahedron
Lett. 1990, 31, 5045-5048.
Gen er a l Meth od s. Thin-layer chromatography (TLC) was
carried out with silica gel 60 F254 (Merck Co.). Absorbance
(15) Ikeda, H.; Du, Y-q.; Nakamura, A.; Toda, F. Chem. Lett. 1991,
1495-1498.
(16) Ikeda, H.; Moon, H-t.; Du Y-q.; Toda, F. Supramol. Chem. 1993,
1, 337-342.
(18) Ikeda, H.; Nakamura, M.; Ise, N.; Oguma, N.; Nakamura, A.;
Ikeda, T.; Toda, F.; Ueno, A. J . Am. Chem. Soc. 1996, 118, 10980-
10988.
(17) Hamasaki, K.; Ikeda, H.; Nakamura, A.; Ueno, A.; Toda, F.;
Suzuki, I.; Osa, T. J . Am. Chem. Soc. 1993, 115, 5035-5040.
S0022-3263(96)00425-2 CCC: $14.00 © 1997 American Chemical Society

1412 J . Org. Chem., Vol. 62, No. 5, 1997
Ikeda et al.
H-7′); 8.25 (dd, J ) 1.1 and 7.4 Hz, 1H, H-2′); 8.37 (d, J ) 8.5
Hz, 1H, H-8′); 8.43 (d, J ) 8.4 Hz, 1H, H-4′).
Ch a r t 1
Syn th esis of Mon o-6-(N-dan syl-L-leu cylam in o)-6-deoxy-
â-cyclod extr in (1). To a solution of 6-amino-6-deoxy-â-
cyclodextrin17 (2.1 g, 1.9 mmol) in 10 mL of DMF was added
N-dansyl-L-leucine (0.73 g, 2.0 mmol). After the solution was
cooled below 0 °C, N,N-dicyclohexylcarbodiimide (0.45 g, 2.2
mmol) and 1-hydroxybenzotriazole (0.30 g, 2.2 mmol) were
added. The resulting mixture was stirred at room temperature
for 2 days. After the insoluble materials were removed with
filtration, the filtrate was poured into acetone, and the
precipitate was collected and dried in vacuo. The crude
product was purified by column chromatography on highly
porous polystyrene gel, DIAION HP-20 (eluted with water/
methanol ) 100/0 to 50/50). The eluent was concentrated to
give the desired product as a pale yellow solid (50% yield). TLC
Rf ) 0.60 (n-BuOH:EtOH:H2O ) 5:4:3). Anal. Calcd for
C60H93O37N3S‚3H2O: C, 47.0; H, 6.50; N, 2.74; S, 2.09.
Found: C, 47.3; H, 6.52; N, 2.76; S, 2.05. MS (ESI) m/e 1478
(calcd for (M - H)-, 1478).
Syn th esis of Mon o-6-(N-Dan syl-D-leu cylam in o)-6-deoxy-
â-cyclod extr in (2). This compound was prepared by the
condensation of 6-amino-6-deoxy-â-cyclodextrin and N-dansyl-
D-leucine by the same method as for 1 (40% yield). TLC
Rf ) 0.61 (n-BuOH:EtOH:H2O ) 5:4:3). Anal. Calcd for
C60H93O37N3S‚4H2O: C, 46.4; H, 6.56; N, 2.71; S, 2.07.
Found: C, 46.5; H, 6.52; N, 2.84; S, 2.47. MS (ESI) m/e 1478
(calcd for (M - H)-, 1478).
NMR Sp ectr oscop y. 1D and 2D NMR spectra were
recorded in D2O at 25 °C on Varian VXR-500S and UNITY
plus-400 spectrometers operating at 499.843 and 399.973 MHz,
respectively, for 1H. The probe of a UNITY plus-400 spec-
trometer was equipped with an Actively-Shielded Pulsed Field
Gradient coil. All the NMR spectra were measured using pulse
sequences and standard procedures offered by Varian. HDO
(δ ) 4.70) was used as an internal standard, and 3-(trimeth-
ylsilyl)propionic acid-d4 sodium salt (TSP, δ ) 0) was used as
an external standard. A 1-3 mM sample in D2O was used
for NMR spectroscopy unless otherwise noted. One-dimen-
sional total correlation spectroscopy (1D TOCSY): 1D TOCSY
spectra were obtained by the selective irradiation of the H-1
protons of the glucose unit with the shaped pulse, hermite
180,20 followed by various mixing times (MLEV-17 spin lock)
on a UNITY plus-400 spectrometer; 256 scans, 1.2 s repetition
delay, 27392 data point, and 3.8 s acquisition time were used.
Two-dimensional total correlation spectroscopy (2D TOCSY):
2D TOCSY spectra were obtained on a Varian VXR-500S; 100
ms MLEV-17 spin lock, 32 scans per t1 increment, 512 × 2 t1
increment, 2048 × 2048 data points, 0.227 s acquisition time,
and 2.0 s repetition delay were used. HDO was suppressed
by selective irradiation during the repetition delays. PFG
multiple-quantum filtered correlated spectroscopy (PFG
MQF-COSY): PFG MQF-COSY spectra were obtained on a
UNITY plus-400 spectrometer; 32 scans per t1 increment, 512
× 2 t1 increment, 2048 × 2048 data points, 0.205 s acquisition
time, and 1.5 s repetition delay were used. PFG 1H-13C 1H
detected heteronuclear single-quantum coherence spectroscopy
(PFG 1H-13C HSQC): PFG 1H-13C HSQC spectra were
obtained on a UNITY plus-400 spectrometer; 128 scans per t1
increment, 128 × 2 t1 increment, 2048 × 2048 data points,
0.285 s acquisition time, and 2.0 s repetition delay were used.
Rotating frame nuclear Overhauser effect spectroscopy
(ROESY): ROESY spectra were obtained on a Varian VXR-
500S; 300 ms MLEV-17 spin lock, 32 scans per t1 increment,
512 × 2 t1 increment, 2048 × 2048 data points, 0.227 s
acquisition time, and 2.0 s repetition delay were used. HDO
was suppressed by selective irradiation during the repetition
delays. The NOE difference spectrum of dansylamide (4) was
obtained by subtracting the FID without NOE caused by
irradiation at a no-resonance area from that with NOE caused
by selective irradiation of the methyl resonance (δ ) 2.88 ppm)
on a UNITY plus-400 spectrometer.
and fluorescence spectra were recorded on a Shimadzu UV-
3100 spectrometer and a Hitachi 850 fluorescence spectrom-
eter, respectively. Fluorescence decay was measured by a
time-correlated single-photon counting method on a Horiba
NAES-550 system. A self-oscillating flash lamp filled with H2
was used as a light source. The excitation beam was passed
through the filters, UV34, U340, and U350 (Hoya), and the
emission beam was passed through the filters, L42 (Hoya) and
Y46 (Toshiba) and an aqueous solution of NiSO4‚6H2O (500
g/L) in a 1-cm pathlength cell. The accumulated data in the
memory of the system were transferred to and analyzed on a
desktop computer (Hewlett-Packard HP9000 Series Model
330). Lifetime was obtained by deconvolution with a nonlinear
least-square fitting procedure. Mass spectrometry was per-
formed on a Hitachi M-1200H mass spectrometer. Optical
rotation was measured with a J ASCO DIP-1000 digital pola-
rimeter at room temperature. Melting point was measured
by a Yanaco MP-J 3 micromelting point apparatus without any
correction. Elemental analyses were performed by the Ana-
lytical Division in Research Laboratory of Resources Utiliza-
tion of Tokyo Institute of Technology.
Syn th esis of N-Da n syl-D-leu cin e. To a solution of D-
leucine (0.75 g, 5.7 mmol) in 10 mL of 1 N NaOH aqueous
solution was added a solution of dansyl chloride (2.7 g, 10.0
mmol) in 10 mL of acetone. The resulting mixture was stirred
at room temperature for 4 h. After evaporation of acetone,
the insoluble materials were removed with filtration. The
filtrate was made acidic to Congo red with 1 N HCl aqueous
solution to obtain a crude product as a solid. Column chro-
matography on silica gel (eluted with methanol/chloroform )
1/9) afforded the desired product as a pale yellow solid (30%
yield). TLC Rf ) 0.34 (CHCl3: MeOH ) 4:1). Anal. Calcd
for C18H24N2 O4S: C, 59.32; H, 6.64; N, 7.69; S, 8.78. Found:
C, 59.18; H, 6.57; N, 7.64; S, 8.55. [R]D ) 37.4° (c 0.1,
methanol) ([R]D ) -37.7° for N-dansyl-L-leucine). Mp 80-81
°C. 1H NMR (D2O) δ 0.01 (d, J ) 6.6, 3H, H-δ): 0.51 (d, J )
6.6 Hz, 3H, H-δ′); 0.98 (m, 1H, H-γ); 1.16 (ddd, J ) 4.4, 9.5,
and 14.1 Hz, 1H, H-â); 1.28 (ddd, J ) 4.4, 10.0 and 14.1 Hz,
1H, H-â′); 2.92 (s, 6H, NMe2); 3.33 (dd, J ) 4.4 and 10.0 Hz,
1H, H-R); 7.47 (d, J ) 7.8 Hz, 1H, H-6′); 7.67 (m, 2H, H-3′ and
(19) Dunbar, R. A.; Bright, F. V. Supramol. Chem. 1994, 3, 93-99.
(20) Warren, W. S.; Silver, M. Adv. Magn. Reson. 1988, 12, 248-
384.

NMR Studies of N-Dansylleucine-Appended â-Cyclodextrins
J . Org. Chem., Vol. 62, No. 5, 1997 1413
Ta ble 1. F lu or escen ce Lifetim es of 1-3 in Aqu eou s
use of various of 1D and 2D NMR techniques. In addition
to the method already reported,14-17 the 1D total correla-
tion spectroscopy (TOCSY)21,22 and the nuclear Over-
hauser effect (NOE) difference spectra27 were applied and
the pulsed field gradient (PFG)23-25 was used for the
selection of a particular coherence-transfer pathway and
the suppression of the HDO resonance in 2D NMR
experiments. A shaped pulse was also used for increas-
ing the selectivity of the irradiation.
Solu tion a
τ1/ns
A1
τ2/ns
A2
ø2
1
2
3
5.7
6.9
3.5
0.33
0.23
-
17.7
17.8
-
0.67
0.77
-
1.32
1.38
1.28
[1-3] ) 2 × 10-5 M. Decay curves were fitted to the following
a
equation: I(t) ) A1 exp(-t/τ1) + A2 exp(-t/τ2).
1
At first, the overlapped signals of the H NMR spectra
of 1 and 2 were separated into each set of resonances for
the protons belonging to the same D-glucose unit using
the 1D and 2D TOCSY spectra. Using the 1D and 2D
TOCSY spectra and the PFG multiple-quantum filtered
correlated spectroscopy (MQF-COSY) spectra,24 all the
resonances of 1 and 2 were then assigned to the protons
of each glucose unit, except the sequence of the glucose
units. Because the 1D TOCSY spectrum with variation
of the duration of the propagation period can give not
only the extracts of each set of resonances for the protons
belonging to the same D-glucose unit but also the assign-
ments of their resonances as shown in Figure 2, the
combined use of 1D and 2D TOCSY and COSY spectra
provides more exact assignment. It is difficult to assign
the resonances of the H-6 protons of the glucose unit A
F igu r e 1. Conformational equilibrium in aqueous solution
and guest-induced conformational change in 1 and 2.
Resu lts a n d Discu ssion
F lu or escen ce Lifetim e. In aqueous solution, the
pendant groups of modified CDs are flexible and not fixed
in many cases.7,17 In the case of the modified CD that
has a fluorophore as a pendant, the analysis of the
fluorescence decay of the fluorophore pendant provides
useful information with respect to the conformational
features, because the fluorescence lifetimes of usual
fluorophores are in the range of a measurable time scale
(nanoseconds) for the conformations. This is particularly
true for the fluorescent CDs with a fluorophore pendant,
which is sensitive to the environment and has different
lifetimes when located inside the cavity and in bulk water
solution.7,19
All the fluorescence decays of the dansyl moiety of
N-dansyl-L-leucine-appended â-cyclodextrin (1) and N-
dansyl-D-leucine-appended â-cyclodextrin (2) were ana-
lyzed by a simple double-exponential function, and the
results are summarized in Table 1. These results indi-
cate that there are two kinds of observable conforma-
tional isomers. The longer and shorter lifetime species
are with the dansyl moiety inside and outside the cavity,
respectively.7,19 These two species are in equilibrium as
shown in parenthesis in Figure 1.
in the
1H NMR spectrum because of their dispersed and
shifted resonances due to the anisotropic shielding effect
of the carbonyl group of the leucine residue, but the C-6
carbon resonance of the glucose unit A was found at 43.5
ppm in the 13C NMR spectrum. The resonances of the
H-6 protons of the glucose unit A were assigned by the
1
1
PFG H-13C H detected heteronuclear single-quantum
coherence spectroscopy (HSQC) spectrum25 as shown in
Figure 3. Finally, the sequence of the glucose units was
determined by the rotating frame nuclear Overhauser
effect spectroscopy (ROESY) spectrum.26 The distance
between H-1 of a glucose unit and H-4 of the adjacent
glucose unit is short enough to give an NOE, which can
be observed by ROESY. The glucose unit of which the
H-4 resonance has a negative crosspeak with the H-1
resonance of the glucose unit A in the ROESY spectrum
was identified as the glucose unit B. In a similar way,
the sequences of all the glucose units were successively
determined. The results of the assignments are shown
in Figures 4 and 5.
The 1H resonances for H-A5 (the proton at the C5
position in the glucose unit A), H-E3, and H-E5 of 1 and
H-A3, H-A5, H-D3, H-D5, H-E3, H-E5, and H-G5 of 2,
these protons being located at the inside of the cavities,
are shifted to the upper field as shown in Figures 4 and
5. These shifts are attributed to the anisotropic ring
current effects from the dansyl moieties, and this indi-
cates that the dansyl moieties are included in the CD
cavities.
Assign m en ts of th e CD Region s of th e 1H Reso-
n a n ces of 1 a n d 2. The observed NMR spectra of 1 and
2 are the averages of the spectra of the species in various
conformational states weighted by the fractions of the
conformers. This situation is due to the fact that the
interconversion rates are fast on the NMR time scale.
The analysis of the fluorescence decay of the dansyl
moiety indicates that there are two kinds of observable
conformational isomers, which are in equilibrium, and
that the dansyl moieties of 1 and 2 are located in the
cavity in the major state. The concentration dependence
of various spectra of 1 and 2 (10-0.1 mM) substantiates
that this inclusion complex is not the intermolecular
complex but the intramolecular complex (self-inclusion
complex). Therefore, the NMR spectra of 1 and 2 are
expected to provide information about their structures
in the self-inclusion state.
Assign m en ts of th e Da n syl Region s of th e 1H
Reson a n ces of 1 a n d 2. Figure 6 shows the naphthyl
1
region of the H NMR spectra of 1, 2, and N-dansyl-L-
leucine (3). The resonance patterns of 1 and 2 are
(21) Davis, D. G.; Bax, A. J . Am. Chem. Soc. 1985, 107, 7197-7198.
(22) Inagaki, F.; Kodama, C.; Suzuki, M.; Suzuki, A. FEBS Lett.
1987, 219, 45-50.
(23) Hurd, R. E. J . Magn. Reson. 1990, 87, 422-428.
(24) Davis, A. L.; Laue, E. D.; Keeler, J . J . Magn. Reson. 1991, 94,
637-644.
1
The full assignments of the H NMR spectra of 1 and
(25) Davis, A. L.; Keeler, J .; Laue, E. D.; Moskau, D. J . Magn. Reson.
1992, 98, 207-216.
2 are required to elucidate their structures. Although
1
the H NMR spectra of 1 and 2 are complex, it is possible
(26) Bothner-By, A. A.; Stephens, R. L.; Lee, J -m.; Warren, C. D.;
J eanloz, R. W. J . Am. Chem. Soc. 1984, 106, 811-813.
to assign almost all of their resonances by the combined

1414 J . Org. Chem., Vol. 62, No. 5, 1997
Ikeda et al.
1
F igu r e 3. 400 MHz PFG H-13C HSQC spectrum of 2 in D2O
at 25 °C.
F igu r e 2. 400 MHz 1D-TOCSY spectra of 2; each spectrum
was obtained with a different mixing time after the selective
irradiation of the anomeric proton in the glucose unit A with
a shaped pulse, hermite 180.
different from each other and are also different from that
of 3. These spectra suggest that the dansyl moieties of
1 and 2 are included in the cavities, and the positions of
their dansyl moieties in the cavities are different from
1
each other. The H resonances for the naphthyl region
of the dansyl moiety were assigned by the use of the NOE
difference spectra and the COSY spectra. Because the
dimethylamino group of the dansyl moiety prefers to
adopt conformation a rather than b in Figure 7,28 the
1
F igu r e 4. A part of the 500 MHz H NMR spectrum of 1 in
D2O at 25 °C, showing only the CD region with assignments.
irradiation of the methyl resonance would produce the
largest enhancement on the H-6′ proton and a smaller
enhancement on the H-4′ proton. The NOE difference
spectrum of dansylamide (4) was obtained by subtracting
the FID without NOE caused by irradiation at a no-
resonance area from that with NOE caused by selective
(27) Neuhaus, D.; Williamson, M. The Nuclear Overhauser Effect
in Structural and Conformational Analysis; VCH Publishers: New
York, 1989.
(28) Freeman, H. S.; Hsu, W. N.; Esancy, J . F.; Esancy, M. K. Dyes
Pigm. 1988, 9, 67-82.

NMR Studies of N-Dansylleucine-Appended â-Cyclodextrins
J . Org. Chem., Vol. 62, No. 5, 1997 1415
F igu r e 8. Aromatic regions of (a) 400 MHz NOE difference
spectrum of dansyl amide (4) on preirradiation of N(CH3)2
signal (δ ) 2.88) and (b) its normal 1H NMR spectrum in
methanol-d4 at 25 °C: 1 Hz line broadening was applied in
both spectra.
spectrum. The resonances of the dansyl protons of 1, 2,
and 3 were also assigned by the same method.
1
F igu r e 5. A part of the 500 MHz H NMR spectrum of 2 in
D2O at 25 °C, showing only the CD region with assignments.
Elu cid a tion of th e Or ien ta tion s of th e Da n syl-
leu cin e Moieties. The NOE correlation between a
proton of the dansyl moiety and a proton of the CD moiety
will be observed by the ROESY spectrum, if they are
closer than 4 Å through space.27 Therefore, if the dansyl
moiety is included in the CD cavity, the NOE correlations
between the protons of the dansyl moiety and the protons
(H-3, H-5, or H-6) of the CD moiety will be observed, and
it is possible to estimate the orientation of the dansyl
moiety in the CD cavity using the assigned NOE cor-
relations. Figures 9 and 10 show the ROESY spectra
with assignments. The NOE correlations were observed
between the protons of CD and the protons of the
naphthyl and methyl parts of the dansyl moiety. A weak
NOE correlation was observed between the protons of CD
and the δ protons of the leucine moiety. The flexible
movement of the δ protons makes the NOE weaker.
Although the NOE correlation between the protons of the
dansyl moiety and the protons of CD moiety can be also
interpreted by the head-to-head bimolecular complex in
which each dansyl moiety is included in the cavity of each
other (intermolecular complex), the NOE correlation
between the δ protons of the leucine moiety and the CD
protons cannot be interpreted by the head-to-head bimo-
lecular complex, because some additional NOE correla-
tion between the δ protons and the CD protons should
be observed in the case of the head-to-head bimolecular
complex. Therefore, the head-to-head bimolecular CD
complex is completely ruled out. The head-to-tail bimo-
lecular complex is also ruled out by the data of the NOE
correlation, and the NMR data can be explained only by
the self-inclusion complex (intramolecular complex). Fig-
ure 11 shows the structures of 1 and 2 estimated from
the NOE data and the degree of the anisotropic ring
current effect from the dansyl moiety in the 1H NMR
spectra of the CD parts. The dansyl moiety of 2 is
included in the cavity more deeply than that of 1. These
structures can elucidate the difference in the resonance
of the δ protons of the leucine between 1 and 2. The
resonances of the δ protons of the leucine moiety of 1 are
more dispersed and are shifted more upfield than those
of 2. This difference is caused by the difference in the
position of the dansyl moiety, where the δ protons of 1
are more influenced by the anisotropic ring current
F igu r e 6. The naphthyl regions of the 500 MHz 1H NMR
spectra of (a) 1, (b) 2, and (c) 3 in D2O at 25 °C.
F igu r e 7. Conformational equilibrium of the dansyl moiety.
irradiation of the methyl resonance at 2.88 ppm (Figure
8). From this spectrum, the resonances at 7.27 and 8.53
ppm were identified as the H-6′ resonance and the H-4′
resonance, respectively. The other 1H resonances of 4
were assigned by tracing the coupling relationships
starting from H-6′ or H-4′ in the PFG MQF-COSY

1416 J . Org. Chem., Vol. 62, No. 5, 1997
Ikeda et al.
F igu r e 9. Portions of the 500 MHz ROESY spectrum of 1 in D2O at 25 °C with mixing time of 300 ms, indicating the cross peaks
between the protons of the dansyl-L-leucine moiety and the protons of the â-CD moiety.
effects of the dansyl moiety than those of 2. The
difference in the fluorescence intensity of 1 and 2 is also
interpreted by the difference in the inclusion depth of the
dansyl moiety, because the fluorescent intensity is sensi-
tive to the microenvironment, and the fluorescence
intensity is stronger in a hydrophobic environment than
in a hydrophilic environment.29-32
The degree of the stability in the self-inclusion state
can be compared between 1 and 2 by the use of the
fluorescence lifetime data, because the equilibrium con-
stant between the species A and the species B in Figure
1 can be roughly estimated from the A2/A1 ratio. The
equilibrium constants for 1 and 2 were calculated to be
2.0 and 3.4, respectively. Therefore, the self-inclusion
state of 2 is 1.7-fold more stable than that of 1. The
difference in the stability in the self-inclusion state
produces the difference in the binding ability, because
the guest binding is the competitive action with the
dansyl moiety binding. In fact, the difference in the
stability in the self-inclusion complex between 1 and 2
(i.e. 1.7-fold) is almost consistent with the difference in
the binding ability (i.e. about 2-fold or more).
The difference in the self-inclusion depth is attributed
only to the difference in the enantiomeric configuration
of the leucine residue. The CPK model experiment
suggested that the self-inclusion complex formation of 1
required larger angle bending in the spacer part than
(29) Chen, R. F. Arch. Biochem. Biophys. 1967, 120, 609-620.
(30) Li, Y.-H.; Chan, L.-M.; Tyer, L.; Moody, R. T.; Himel, C. M.;
Hercules, D. M. J . Am. Chem. Soc. 1975, 97, 3118-3126.
(31) Bridge, J . B.; J ohnson, P. Eur. Polym. J . 1973, 9, 1327-1346.
(32) Hoenes, G.; Hauser, M.; Pfleiderer, G. Photochem. Photobiol.
1986, 43, 133-137.

NMR Studies of N-Dansylleucine-Appended â-Cyclodextrins
J . Org. Chem., Vol. 62, No. 5, 1997 1417
F igu r e 10. Portions of the 500 MHz ROESY spectrum of 2 in D2O at 25 °C with mixing time of 300 ms, indicating the cross
peaks between the protons of the dansyl-D-leucine moiety and the protons of the â-CD moiety.
that of 2. This energy defect for the angle bending is an
important factor for the determination of the self-inclu-
sion depth, and the deeper inclusion of the dansyl moiety
into its own cavity results in the stronger van der Waals
interaction and the more stable self-inclusion complex.
Hydroxyl groups of CD moiety would form hydrogen
bonding with the dimethylamino group or the sulfona-
mide group of the dansyl moiety, and this hydrogen
bonding is another important factor for the stabilization
of the self-inclusion complex. More detailed structural
studies are needed for discussion about the stabilization
factors and are now under way.
Con for m a tion a l Ch a n ge u p on Ad d ition of 1-Ad a -
m a n ta n ol. Upon addition of 1-adamantanol, the fluo-
rescence intensities of 1 and 2 were dramatically de-
creased. This guest-induced fluorescence variation
suggests that the hosts exclude the dansyl moiety from
their hydrophobic cavities to the bulk water environment
associated with the guest accommodation (Figure 1).
These conformational changes are also supported from
1
the changes in their H NMR spectra as shown in Figures
1
12, 13, and 14. The H resonances of the dansyl moieties
of 1 and 2 in the presence of 1-adamantanol are different
from those of the hosts alone, but are similar to those of
3 or 4. The resonances of the δ protons of the leucine
moieties of 1 and 2 were also changed in a similar way.
These changes suggest that the dansyl moiety is excluded
from the cavity upon addition of the guest molecule. The
dansyl moiety excluded from the cavity causes a larger
anisotropic ring current effect on the δ protons of the
leucine moiety, and the resonances of the δ protons of
the leucine moiety are more dispersed (Figure 13).
The resonances of the CD protons were also changed
upon addition of 1-adamantanol. The upfield shifted

1418 J . Org. Chem., Vol. 62, No. 5, 1997
Ikeda et al.
F igu r e 11. The estimated structures of 1 and 2. The energy was not minimized.
F igu r e 14. Portions of the 500 MHz 1H NMR spectra of 1
(a-1) alone, (a-2) in the presence of 1-adamantanol, 2 (b-1)
alone, and (b-2) in the presence of 1-adamantanol in D2O at
25 °C, showing the region of the anomeric protons of CD.
F igu r e 12. The aromatic regions of the 500 MHz 1H NMR
spectra of 1 (a-1) alone, (a-2) in the presence of 1-adamantanol,
2 (b-1) alone, (b-2) in the presence of 1-adamantanol, and (c)
3 in D2O at 25 °C.
nances were almost overlapped around 5.0 ppm (Figure
14). This change also indicates the exclusion of the
dansyl moiety. In general, the resonances of the ano-
meric protons are dispersed, if self-inclusion of a pendant
group occurs.15-17 Even if the pendant group is not
included in its own cavity but is fixed at a position near
the entrance of the cavity, the resonances are dispersed
to some extent. These spectral dispersions of the ano-
meric protons represent the reduction of the sevenfold
symmetry of the modified â-CD. However, if the globular
guest is included and the pendant group is moved away
from the cavity, this spectral dispersion is lost.15-17 The
spectral changes in the CD protons of 1 and 2 also
suggest the exclusion of the dansyl moieties from the
cavities and the reduction of the distortion of the CD
macrocyclic rings.
In conclusion, the structures of N-dansyl-L-leucine-
appended â-CD (1) and N-dansyl-D-leucine-appended
â-CD (2) were estimated by the combined use of 1D and
2D NMR techniques. The dansyl moiety of 2 was
included in its own cavity more deeply than that of 1.
Only the difference in the enantiomeric configuration in
the spacer unit caused the difference in the inclusion
depth, and it is due to the difference in the angle bending
to make the self-inclusion complex. The difference in
their properties was interpreted by the difference in the
inclusion depth of the dansyl moieties. The exclusion of
the dansyl moiety upon addition of the guest was also
monitored by NMR.
F igu r e 13. Portions of the 500 MHz 1H NMR spectra of 1
(a-1) alone, (a-2) in the presence of 1-adamantanol, 2 (b-1)
alone, (b-2) in the presence of 1-adamantanol, and (c) 3 in D2O
at 25 °C, showing the region of δ protons of leucine.
peaks due to the ring current effect of the dansyl moiety
were not observed in the spectra. The resonances for the
anomeric protons of 1 and 2 were also changed upon
addition of 1-adamantanol. The resonances for the
anomeric protons of 1 and 2 in the absence of the guest
are dispersed in the range from 4.7 to 5.2 ppm, and all
of them are completely separated as shown in Figure 14.
Upon addition of the globular guest, 1-adamantanol,
these spectral dispersions were collapsed, and the reso-
J O960425X