Molecular cleft possessing a cholic acid moiety as a podant and its conformation

Article abstract of DOI:10.1039/a808755e

A cleft type host molecule 4 possessing a cholic acid moiety as a podant was synthesized from the condensation of naphthalene-1,4,5,8-tetracarboxylic dianhydride with a 3α-aminocholanoate derivative. It was found that 4 could be associated with naphthalen

Full text of DOI:10.1039/a808755e

View Article Online / Journal Homepage / Table of Contents for this issue
Molecular cleft possessing a cholic acid moiety as a podant and its
conformation
Shigeo Kohmoto,* Katsuya Sakayori, Keiki Kishikawa and Makoto Yamamoto
Department of Materials Science, Faculty of Engineering, Chiba University, 1-33, Yayoi-cho,
Inage-ku, Chiba 263-8522, Japan
Received (in Cambridge) 9th November 1998, Accepted 1st February 1999
A cleft type host molecule 4 possessing a cholic acid moiety as a podant was synthesized from the condensation of
naphthalene-1,4,5,8-tetracarboxylic dianhydride with a 3α-aminocholanoate derivative. It was found that 4 could be
associated with naphthalene-2,6-dimethanol 5 with the binding constant of 71–92 MϪ1 with a host–guest ratio of 1:1
from 1H NMR spectrometric titration. Conformational analysis of 4 in the absence or the presence of 5 was carried
out by variable temperature 1H NMR spectroscopy. Without a guest molecule, 4 adopted two stable conformations at
213 K, however, a single different conformation was found in the presence of 5. The cleft type conformation of 4 was
induced by the inclusion of the guest molecule.
Considerable effort has been devoted to developing efficient
host molecules for molecular recognition.1 A large number of
host molecules are designed to be cyclic in order to give rigid
conformations for binding of guest molecules. Some devices are
required in acyclic hosts to fix their conformations. The fixation
of their conformation in a C-shape creates a cleft-like structure,
which provides a concave area for binding of guest molecules.
In order to create a molecular cleft, it is highly desirable to have
a rigid podant and a spacer unit and also to restrict the rotation
of a podant along the spacer–podant single bond. Owing to its
rigidity and unique amphiphilicity, having a hydrophobic outer
surface and a hydrophilic inner surface, cholic acid has been
used in building molecules for molecular recognition.2 The con-
cept of linking two cholic acids to form an acyclic cleft was
Fig. 1 Job plot and the simulated curve fit for titration of 4 with 5.
first reported by McKenna et al.,3 and other examples have
appeared since.4 As a continuation of our study, the application
1
was added to 4 and the H NMR spectrum of the mixture
was measured at ambient temperature. Out of the compounds
examined, the addition of naphthalene-2,6-dimethanol 5, 1,6-
dihydroxyhexa-2,4-diyne, p-xylylenediol, p-nitrophenol, and
N-cbz-L-tryptophan caused the chemical shift change of the
aromatic protons of 4, suggesting an interaction between 4 and
these guest molecules. Except for 1,6-dihydroxyhexa-2,4-diyne,
upfield shifts of the aromatic protons were observed. Upfield
shifts of 0.4, 0.05, 0.05, and 0.10 ppm were observed when
one equivalent amount of 5, p-xylylenediol, p-nitrophenol, and
N-cbz-L-tryptophan were added, respectively. The results indi-
cated that the aromatic protons of 4 were shielded magnetically
by the aromatic rings of the guest molecules. In the case of 1,6-
dihydroxyhexa-2,4-diyne, downfield shifts of the protons (0.01
ppm with one equivalent amount) were observed due to
the deshielding effect of its triple bonds. From the observed
shielding and deshielding effects, it can be deduced that guest
molecules are associated in parallel to the diimide moiety of 4.
The binding constants for all the guest molecules except 5 could
not be determined due to the large errors originated in the small
changes in chemical shifts of both 4 and guest molecules. The
host–guest ratio between 4 and 5 was determined to be 1:1
of cholic acid (3α,7α,12α-trihydroxy-5β-cholan-24-oic acid) as
a potent building block for host molecules,5 we have been inter-
ested in the conformational regulation of an acyclic host
possessing this building block. In this paper, we report the syn-
thesis of a cholic acid based molecular cleft with a rigid spacer
unit, naphthalenetetracarboximide, and its conformational
analysis. An advantage of using the large aromatic imide
moiety was demonstrated by Rebek et al.6 in which the
restricted rotation along its C–N single bond was crucial.7
Results and discussion
The cholic acid based molecular cleft 4 was synthesized as fol-
lows. Methyl 3α-azido-7α,12α-bis(formyloxy)cholanoate 18 was
reduced with zinc powder to give the 3α-amino derivative 2.
Crude 2 was heated with naphthalene-1,4,5,8-tetracarboxylic
dianhydride at 120 ЊC in DMF for 2 h to give tetrakisformyloxy
derivative 3 in 37% yield. Deprotection of the formyloxy
groups was achieved by adding dropwise a saturated meth-
anolic lithium hydroxide solution to a THF solution of 3 to
1
afford 4 in 50% yield. The structure of 4 was confirmed by H
NMR (500 MHz) and HRMS (FAB). Since cholic acid has the
L-type shape and the 3-imido group has the α-configuration,
the molecule can take a concave structure. In this structure, the
hydroxy groups at the positions 7α and 12α can be positioned as
binding sites, orthogonal to the concave surface.
from the Job plot9 (Fig. 1) using H NMR spectroscopy since
1
the maximum value was observed at the [4]/([4] ϩ [5]) ratio of
0.5. The binding constant between 4 and 5 was determined by
1H NMR spectrometric titration by adding the CDCl3 solution
of 4 to the CDCl3 solution of 5. Fig. 2 shows the induced
chemical shift changes of the hydroxy protons of 5. The bind-
In order to survey the recognition of guest molecules with di-
functional groups, an equivalent amount of guest molecule
J. Chem. Soc., Perkin Trans. 2, 1999, 833–836
833

View Article Online
ing constant of 91 9 MϪ1 was obtained using non-linear least
squares regression.10 A similar value (72 9 MϪ1) was obtained
from the induced chemical shift changes of aromatic protons of
6. The binding ability of 4 is almost equivalent to the Rebek’s
cleft.6 The conformational flexibility of 4 may be responsible
for the relatively small binding constant.
In order to examine the conformational flexibility of 4 in the
absence or the presence of 5, the variable temperature (VT) 1H
NMR study of 4 was carried out. Fig. 3 shows the VT 1H NMR
study of 4 (the aromatic protons) in the absence of 5. The sharp
peak at 273 K was gradually broadened and split into two peaks
(243 K). According to the MM2 calculation of 4, this molecule
can adopt two stable conformations, conformations A and B, as
shown in Scheme 1. The conformation B is 1.7 kJ molϪ1 more
stable than the conformation A according to the calculation.
These conformations seem to be appropriate since imide com-
pounds which showed a restricted rotation about the C–N
single bond, were known to adopt two stable conformations in
which the imide ring would be positioned perpendicular to the
N-substituents to avoid the steric repulsion between them. In
the conformation A, there are two kinds of aromatic protons.
Two protons inside the cavity are magnetically equivalent and
two other protons outside the cavity are also magnetically
equivalent. Therefore, the 1H NMR spectrum corresponding to
the conformation A should give two singlet peaks due to the
above mentioned two kinds of protons. On other hand, the
conformation B has two different types of aromatic protons.
Due to the C2 symmetric nature of this conformation, two pro-
tons across the naphthalenediimide skeleton are magnetically
equivalent. In this situation, the aromatic protons correspond-
ing to the conformation B should give two doublet peaks.
1
Therefore, it is reasonable to consider that the observed H
NMR spectrum of 4 at 213 K is the mixture of two conform-
ations A and B. From the coalescence temperature at 250 K, the
rotational barrier11 is determined to be 57.5 kJ molϪ1. The sing-
let peaks corresponding to the aromatic protons of conformer
A might be overlapped with the doublet peaks of the aromatic
protons of conformer B. When the spectral changes of 4 with
varying amounts of 5 were examined at 213 K (Fig. 4), two
doublet peaks appeared with J = 7.8 Hz (4:5, 5:1) which might
correspond to the aromatic protons of conformer B together
with a broad singlet located between the two doublets. The con-
former A might be associated preferentially with 5, which
resulted in the formation of the above-mentioned broad singlet.
At the end of the titration, all the peaks joined together to give
a single singlet peak corresponding to the aromatic protons
of the complex. A similar phenomenon was observed for the
18β-methyl protons too. The peak corresponding to 4 gradually
changed to the peak corresponding to the complex and con-
verted completely at the host–guest ratio of 1:8.
Fig. 2 1H NMR (500 MHz) titration data and the simulated curve fit
for 4 binding 5 in CDCl3, monitoring the hydroxy proton of 5.
1
Similarly, the VT H NMR study of 4 in the presence of
excess (8 equiv.) 5 was carried out. At ambient temperature, the
signal corresponding to the aromatic protons of 4 became very
broad and it was hard to identify its chemical shift (Fig. 5),
which suggested a rapid equilibrium between the host and the
guest. However, it gradually became sharp with the lowering of
the measured temperature and finally a sharp singlet peak was
observed at 213 K. The results showed a nice contrast to those
in the absence of the guest. The observation of a single singlet
peak suggested that all four aromatic protons of 4 were in a
magnetically equivalent environment. This indicated that a
1
Fig. 3 VT H NMR (400 MHz) spectra of the aromatic protons of
4 in CDCl3.
CO2CH3
CO2CH3
OH
OH
OH
OH
O
O
N
N
O
O
N
N
O
O
O
O
OH
OH
OH
OH
H3CO2C
CO2CH3
conformation A
conformation B
Scheme 1
834
J. Chem. Soc., Perkin Trans. 2, 1999, 833–836

View Article Online
CO2CH3
CO2CH3
Zn
OCHO
OCHO
AcOH
OCHO
OCHO
H2N
2
1
N3
O
O
O
DMF/120 o
C
O
O
O
CO2CH3
CO2CH3
H3CO2C
H3CO2C
LiOH/MeOH/THF
OCHO
OH
OH
OHCO
OHCO
HO
HO
OCHO
O
O
O
O
N
O
N
O
N
O
N
O
3
4
O
N
O
N
OH
HO
O
O
5
6
Scheme 2
1
Fig. 5 VT H NMR (500 MHz) spectra of the aromatic protons of
4 with 8 equiv. amount of 5 in CDCl3.
perature (228 K). In contrast to 4, 6 showed no chemical shift
change upon addition of 5. No change was also observed in the
chemical shift of 5, which indicated that 6 did not have the
ability to recognize 4. Thus the cholic acid moiety is required
for association with 5. The importance of the hydroxy groups in
the cholic acid moiety for the recognition of 5 was confirmed by
the titration experiment of the formyl derivative 3 with 5. No
meaningful binding (binding constant; 5 13 MϪ1) of 5 with 3
was observed. The large error was associated with the small
shift change. These results suggested that the hydrogen bonding
between the hydroxy groups of 4 and 5 would be the driving
force for the formation of cleft type conformation.
Fig. 4 1H NMR (500 MHz) spectra of the aromatic protons of 4 with
varying amounts of 5 in CDCl3 at 213 K.
unique fixed conformation of the associated complex between 4
and 5 existed with the host–guest ratio of 1:1. Such a conform-
ation should be the cleft type conformation depicted as the
structure of 4 in Scheme 2. This conformation is 33.6 kJ molϪ1
less stable than the conformation B by MM2 calculation.
In order to demonstrate the significance of cholic acid
1
moieties as the binding sites, a similar H NMR study of the
anhydride 6 possessing two cyclohexyl groups substituted at
As we demonstrated, the conformation of 4 was flexible in
the absence of the guest molecule. Its podant part, the cholic
acid moiety, was freely rotated at ambient temperature and it
showed two stable conformations at 213 K, though these
nitrogen atoms and the formyl derivative 3 was carried out with
1
5. From the VT H NMR study of 6 in CDCl3, its rotational
barrier was found to be 51.2 kJ molϪ1 at the coalescence tem-
J. Chem. Soc., Perkin Trans. 2, 1999, 833–836
835

View Article Online
two conformations were not suitable as a cleft. However, the
addition of 5 at 213 K caused the fixation of the conformation
of 4 as a cleft type. This kind of guest responsive conform-
ational change of a host molecule is interesting to study
dynamic process of self-assembling system.
Preparation of dipodant 6
To a DMF solution (20 cm3) of naphthalene-1,4,5,8-tetra-
carboxylic dianhydride (262.9 mg, 1.00 mmol) was added
cyclohexylamine (0.229 cm3, 2.00 mmol) quickly at 120 ЊC. The
resulting mixture was continued to be heated for 15 h. After
evaporation of solvent, chloroform (100 cm3) was added to the
residue. The chloroform layer was washed with aqueous 1 M
HCl solution (50 cm3) and then dried over anhydrous Na2SO4.
After evaporation of solvent, the residue was chromatographed
on silica gel with chloroform as eluent to give 6 (441 mg, 0.96
mmol, 96%) as pale yellow blocks, mp > 300 ЊC; νmax(KBr)/
cmϪ1 2940, 1710, 1670, 1585, 1455, 1330, 1255, 1180 and 770;
δH(400 MHz; CDCl3) 1.33 (2 H, qt, J 13.3 and 3.4 ), 1.44 (2 H,
qt, J 13.3 and 3.4), 1.65–1.80 (8 H, br m), 1.89 (4 H, br d,
J 13.3), 2.31 (4 H, qd, J 12.0 and 3.6), 4.99 (2 H, tt, J 12.0
and 3.6) and 8.68 (4 H, s); MS (EI); 431 (M ϩ 1), 430 (Mϩ, 22),
349 (95), 267 (31), 248 (21), 129 (14), 97 (27) and 57 (100);
HRMS (FAB) Calcd for C26H26N2O4: (MHϩ), 430.1893. Found:
430.1894.
Experimental
General
The melting points were determined on a Yanako MP-S3
melting-point apparatus or a Mac Science DSC 3100S and are
uncorrected. IR spectra were obtained in a HITACHI I-2000
spectrometer. NMR spectra were recorded on JEOL LA-400 or
LA-500 spectrometers in CDCl3 with Me4Si as an internal
standard; J values are given in Hz. Mass spectra were obtained
on a JEOL JMS-HX110. Reaction mixtures were concentrated
on a rotary evaporator at 15–20 mmHg (1 mmHg = 133.322
Pa). Chromatographic separations were accomplished by flash
column chromatography on silica gel (Fuji gel BW 200).
1H NMR study
Preparation of tetrakisformyl dipodant 3
An acetic acid solution (20 cm3) of 3α-azidocholanoate 1 (944.7
mg, 1.876 mmol) and zinc powder (4.000 g) was stirred for 24 h
at room temperature. The progress of the reaction was moni-
tored by tlc (CHCl3 :MeOH = 10:1). The reaction mixture was
filtered by suction and the residue was washed with acetic acid
(10 cm3 × 2). A combined filtrate was evaporated and the resi-
due was dissolved in chloroform (100 cm3) and washed with
saturated aqueous NaHCO3 solution (100 cm3). The organic
layer was separated and washed with distilled water (100 cm3).
After evaporation of solvent, the crude 3α-aminocholanoate 2
thus obtained, was used for the preparation of 3 without
further purification. To a dimethylformamide (DMF) solution
of naphthalene-1,4,5,8-tetracarboxylic dianhydride was added
dropwise a solution of crude 2 (927 mg, prepared as above
mentioned) in 10 cm3 of DMF during 0.5 h at 120 ЊC and
stirred for an additional 2 h at this temperature. After evapor-
ation of solvent, the residue was chromatographed on silica gel
with CHCl3–ethyl acetate (20:1) to give 3 (408 mg, 37%) as
brown powder, mp 271.7 ЊC; νmax(KBr)/cmϪ1 2940, 1710, 1670,
1585, 1455, 1330, 1255, and 1180; δH(400 MHz; CDCl3) 0.80
(6 H, s), 0.87 (6 H, d, J 6.3), 1.01 (6 H, s), 1.17–2.60 (46 H, m),
3.21 (2 H, q, J 12.9), 3.68 (6 H, s), 4.95 (2 H, br m), 5.11 (2 H,
br s), 5.32 (2 H, br s), 8.09 (2 H, s), 8.34 (2 H, s) and 8.70 (4 H,
s); HRMS (FAB) Calcd for C68H87N2O16: (MHϩ), 1187.6056.
Found: 1187.6044.
1H NMR spectroscopic titration for the determination of bind-
ing constant between 4 and 5 was carried out as follows. A
measured amount of CDCl3 solution of 4 (25 mmol dmϪ3) was
added to the CDCl3 solution of 5 (2.0 mmol dmϪ3, 0.50 cm3),
1
and after each addition the H NMR spectrum of the mixture
was recorded. The induced chemical shift value was plotted vs.
the ratio of 4 to 5. The binding constant was calculated using
non-linear least square regression minimizing the value of
Σ(δobs Ϫ δcalc). For the variable temperature (VT) 1H NMR
study, the CDCl3 solution (5.0 mml dmϪ3, 0.50 cm3) of 4 was
used.
References
1 C. Seel and F. Vögtle, Angew. Chem., Int. Ed. Engl., 1992, 31, 528;
R. M. Izatt, J. S. Bradshaw, K. Pawlak, R. L. Bruening and B. J.
Tarbet, Chem. Rev., 1992, 92, 1261; A. P. Davis, Chem. Soc. Rev.,
1993, 900; S. Anderson, H. L. Anderson and J. K. M. Sanders, Acc.
Chem. Res., 1993, 26, 469; F. C. J. M. van Veggel, W. Verboom and
D. N. Reinhoudt, Chem. Rev., 1994, 94, 279; J.-M. Lehn,
Supramolecular Chemistry, VCH, Weinheim, 1995; A. D. Hamilton,
Ed., Supramolecular Control of Structure and Reactivity, John Wiley
& Sons, Chichester, 1996.
2 A. P. Davis, Chem. Soc. Rev., 1993, 243; A. P. Davis, R. P.
Bonar-Law and J. K. M. Sanders, Comprehensive Supramolecular
Chemistry, Vol. 4, Y. Murakami, Ed., Ch. 7, pp. 257, Pergamon
Press, Oxford, 1996.
3 J. McKenna, J. M. McKenna and D. W. Thornthwaite, J. Chem.
Soc., Chem. Commun., 1977, 809.
Preparation of tetraol dipodant 4
4 C. J. Burrows and R. A. Sauter, J. Inclusion Phenom., Mol. Recognit.
Chem., 1987, 5, 117; J. F. Kenneary, T. M. Roy, J. S. Albert,
H. Yoon, T. R. Wagler, L. Shen and C. J. Burrows, J. Inclusion
Phenom., Mol. Recognit. Chem., 1989, 7, 155; J. Kikuchi,
C. Matsushima, K. Suehiro, R. Oda and Y. Murakami, Chem. Lett.,
1991, 1807.
5 S. Kohmoto, D. Fukui, T. Nagashima, K. Kishikawa, M.
Yamamoto and K. Yamada, Chem. Commun., 1996, 1869.
6 K. D. Shimizu, T. M. Dewey and J. Rebek, Jr., J. Am. Chem. Soc.,
1994, 116, 5145.
7 S. M. Verma and N. B. Singh, Aust. J. Chem., 1976, 29, 295.
8 A. P. Davis, J. J. Perry and R. P. Williams, J. Am. Chem. Soc., 1997,
119, 1793.
9 K. A. Connors, Binding Constants, Wiley, New York, 1987.
10 Simulated curve fits were carried out using GRAFIT Version 3. 0;
R. J. Leatherbarrow, Erithacus Software Ltd, Staines, 1992.
11 A. Mannschreck, G. Rissmann, F. Vögtle and D. Wild, Chem. Ber.,
1967, 100, 335; I. C. Calder and P. J. Garratt, J. Chem. Soc. B., 1967,
660.
Deprotection of 3 was carried out by hydrolysis with LiOH. To
a THF solution (15 cm3) of 3 was added dropwise a saturated
methanolic solution of LiOH under stirring. The progress of
the reaction was monitored by tlc (CHCl3–MeOH = 40:1).
When the spot corresponding to 3 (Rf = 0.95) was completely
(after the addition of 50 drops) converted to that of 4 (Rf = 0.2),
distilled water (50 cm3), brine (70 cm3) and THF (100 cm3) were
added to the reaction mixture and an organic layer was separ-
ated. After evaporation of solvent, the residue was chromato-
graphed on silica gel with CHCl3–MeOH (20:1) to afford 4
(180 mg, 50%), mp 421.5 ЊC; νmax(KBr)/cmϪ1 3540, 2940, 1710,
1665, 1585, 1455, 1335, 1250, 1180 and 1095; δH(500 MHz;
CDCl3) 0.72 (6 H, s), 0.96 (6 H, s), 1.01 (6 H, d, J 6.1), 1.16–1.92
(38 H, m), 2.02 (4 H, m), 2.25 (2 H, m), 2.38 (2 H, m), 2.47 (2 H,
m), 2.69 (2 H, m), 2.71 (2 H, q, J 12.0), 3.42 (2 H, q, J 13.0), 3.66
(6 H, s), 3.87 (2 H, br s), 4.02 (2 H, br s), 4.93 (2 H, br m) and
8.66 (4 H, s); HRMS (FAB) Calcd for C64H87N2O12: (MHϩ),
1075.6259. Found: 1075.6204.
Paper 8/08755E
836
J. Chem. Soc., Perkin Trans. 2, 1999, 833–836