Cholic acid dimers as invertible amphiphilic pockets: Synthesis, molecular modeling, and inclusion studies

Article abstract of DOI:10.1139/cjc-2016-0621

Two dimers of cholic acid were synthesized through simple covalent linkers. The dimers form invertible molecular pockets in media of different polarity; hydrophobic pockets are formed in water and hydrophilic pockets are formed in organic media. Fluorescence studies show that pockets formed by these dimers can serve as invertible hosts for the hydrophobic guest pyrene and the hydrophilic guest coumarin 343. The molecular pocket also enhances dissolution of the weakly soluble cresol red sodium salt in organic media. Molecular modeling was performed to better understand the host-guest complexation process of the invertible amphiphilic pockets. The calculated free energy changes indicate that the two dimers form the most stable complexes with coumarin 343 at a host to guest ratio of 2:2, whereas the host to guest ratio differs in the formation of complexes with pyrene for the two dimers. The dimer with the shorter, less flexible linker seems to form host-guest complexes that are more stable in both water and organic solvents.

Full text of DOI:10.1139/cjc-2016-0621

792
ARTICLE
Cholic acid dimers as invertible amphiphilic pockets:
synthesis, molecular modeling, and inclusion studies
Meng Zhang, Nicolas Levaray, Josée R. Daniel, Karen C. Waldron, and X.X. Zhu
Abstract: Two dimers of cholic acid were synthesized through simple covalent linkers. The dimers form invertible molecular pockets
in media of different polarity; hydrophobic pockets are formed in water and hydrophilic pockets are formed in organic media.
Fluorescence studies show that pockets formed by these dimers can serve as invertible hosts for the hydrophobic guest pyrene and the
hydrophilic guest coumarin 343. The molecular pocket also enhances dissolution of the weakly soluble cresol red sodium salt in
organic media. Molecular modeling was performed to better understand the host–guest complexation process of the invertible
amphiphilic pockets. The calculated free energy changes indicate that the two dimers form the most stable complexes with coumarin 343
at a host to guest ratio of 2:2, whereas the host to guest ratio differs in the formation of complexes with pyrene for the two dimers. The
dimer with the shorter, less flexible linker seems to form host–guest complexes that are more stable in both water and organic solvents.
Key words: cholic acid dimers, molecular pockets, host–guest, molecular modeling.
Résumé : Nous avons synthétisé deux dimères de l’acide cholique a` l’aide de coupleurs covalents simples. Les dimères forment
des pochettes moléculaires réversibles dans des milieux de différents degrés de polarité; les pochettes hydrophobes se forment
dans l’eau, et les pochettes hydrophiles, en milieu organique. Des études de fluorescence montrent que les pochettes formées par
ces dimères peuvent servir d’hôtes réversibles au pyrène comme invité hydrophobe, et a` la coumarine 343 comme invité
hydrophile. La pochette moléculaire permet aussi de dissoudre plus facilement le sel de sodium du rouge de crésol, un composé peu
soluble en milieu organique. Nous avons réalisé une étude de modélisation moléculaire afin de mieux comprendre le processus de
complexation hôte–invité des pochettes amphiphiles réversibles. Les calculs des variations de l’énergie libre indiquent que les
deux dimères forment les complexes les plus stables avec la coumarine 343, selon un rapport hôte–invité de 2:2. Cependant, dans
le cas de la formation des complexes avec le pyrène, les calculs indiquent un rapport hôte-invité différent pour les deux dimères.
Le dimère possédant le coupleur le plus court et le moins flexible semble former des complexes hôte–invité qui sont plus stables
tant dans l’eau que dans les solvants organiques. [Traduit par la Rédaction]
Mots-clés : dimères de l’acide cholique, alvéoles moléculaires, hôte–invité, modélisation moléculaire.
prising a tertiary amine and a butylene spacer) and a longer linker
(dimer 2, with a spacer of 15 atoms comprising two amide groups
Introduction
Bile acids are natural amphiphilic compounds that exist in the
and a secondary amine).
digestive tract of humans and animals.1,2 They possess a convex
Molecular modeling, often used in pharmacokinetics and in the
hydrophobic face bearing three methyl groups and a concave hy-
study of biological receptor–ligand geometry, has been shown to
drophilic face with hydroxyl and carboxylic acid groups.3 Bile acids
be useful for investigating the dynamics and geometry of host–
have been used to prepare molecular devices for use as drug deliv-
ery agents,4–6 sensors for metal ions,7,8 and ion transporters.9–11
Bile acids also serve as attractive building blocks in the construc-
tion of star-shaped derivatives known as molecular pockets or
umbrellas.12–17 Due to their amphiphilicity, bile acid based mate-
rials are capable of responding to changes in solvent polarity by
exposing either their hydrophobic or hydrophilic faces.18,19
Our group has previously synthesized dimers, trimers, and te-
tramers based on cholic acid,7,20–23 the most abundant bile acid.
These oligomers, possessing different linkages between cholic
acid units, can form invertible pockets for binding and encapsu-
lating metal ions and hydrophobic or hydrophilic species. The
effect of the linker length on the stability of such complexes
remains to be investigated. To this end, we have designed and
synthesized two new dimers based on cholic acid through a
shorter linker (dimer 1, with an alkyl spacer of nine atoms com-
guest complexes24 such as cyclodextrin25 and calixarene,26 and
was thus applied to the cholic acid dimers. The combination of
experimental and computational studies has been recognized as a
powerful tool27 to elucidate the effect of solvation energy,28 ionic
interactions29 and insertion dynamics.30,31 The two dimers used
in this work were designed to have uncomplicated linear linkers
to simplify the molecular modeling and to obtain a better under-
standing of the host–guest interaction in relation to the inclusion
properties of the molecular pockets and solvation effects. For the
guest molecules, we have selected polarity-sensitive probes such
as pyrene, coumarin 343, and cresol red sodium salt.
Experimental
Materials and methods
All chemicals were purchased from Sigma-Aldrich and used
without further purification, unless otherwise specified.
Received 25 November 2016. Accepted 4 May 2017.
M. Zhang,* N. Levaray,* J.R. Daniel, K.C. Waldron, and X.X. Zhu. Department of Chemistry, Université de Montréal, C.P. 6128, Succursale Centre-
ville, Montréal, QC H3C 3J7, Canada.
Corresponding author: Julian Zhu (email: julian.zhu@umontreal.ca).
*These authors contributed equally to this work.
Copyright remains with the author(s) or their institution(s). Permission for reuse (free in most cases) can be obtained from RightsLink.
Can. J. Chem. 95: 792–798 (2017) dx.doi.org/10.1139/cjc-2016-0621
Published at www.nrcresearchpress.com/cjc on 17 May 2017.

Zhang et al.
793
Scheme 1. Synthesis of dimers 1 and 2.
Dimer 2
To a solution of diethylenetriamine (440 ␮L, 4 mmol) in tetrahy-
drofuran (THF, 200 mL), cholic acid succinimide ester (4.04 g,
8 mmol) in THF (40 mL) was added dropwise. The reaction mixture
was stirred at room temperature for 10 h and then filtered. The
crude product was obtained by evaporation of the filtrate under
reduced pressure and purified by flash chromatography on neu-
tral aluminium oxide (80/20 chloroform–methanol). Dimer 2 was
obtained as a white solid (2.58 g, 73%). 1H NMR (400 MHz, DMSO-d6,
␦): 4.32 (d, J = 4.18 Hz, 2H, OH), 4.10 (d, J = 3.33, 2H, OH), 4.01 (d, J =
3.35, 2H, OH), 3.78 (m, 2H, CHOH), 3.62 (m, 2H, CHOH), 3.17 (m, 2H,
CHOH), 3.10 (q, J1 = J2 = 6.24, J3 = 6.16, 4H, CONHCH2), 2.58 (t, J1 =
J2 = 6.44, 4H, CH2NH), 0.92 (d, J = 6.36, 6H, CH3), 0.81 (s, 6H, CH3),
0.58 (s, 6H, CH3) ppm. 13C NMR (100 MHz, DMSO-d6, ␦),: 172.81,
70.98, 70.40, 66.21, 48.12, 46.10, 45.69, 41.47, 41.32, 38.14, 35.26,
35.13, 34.84, 34.33, 32.48, 31.62, 30.34, 28.51, 27.26, 26.17, 22.77,
22.57, 17.07, 12.29 ppm. MS m/z, 884.673 [M + H]+, 906.656 [M + Na]+.
Fluorescence spectroscopy
A stock solution of pyrene was prepared in methanol (2 mmol/L) then
diluted in Milli-Q water (30 ␮L in 200 mL, 0.3 ␮mol/L). A stock solution of
coumarin 343 was prepared in anhydrous THF (2 mmol/L) and then
diluted with THF (30 ␮L in 200 mL, 0.3 ␮mol/L). Stock solutions of dimers
1 and 2 (25 ␮mol/L) were made separately in the solutions of either
probe (0.3 ␮mol/L) and diluted to obtain desired concentrations
for the experiments, in which the two probes were studied sepa-
rately or as a mixture in water. The water solutions were purged
with N2 for 60 min to remove the trace of methanol used initially
to dissolve pyrene. The fluorescence spectra of pyrene and cou-
marin 343 were recorded at room temperature on a Varian fluores-
cence spectrophotometer equipped with a Xe-900 lamp using
excitation wavelengths of 336 nm and 410 nm for pyrene and cou-
marin 343, respectively. The bandwidths for excitation and emission
were both 2.5 nm for pyrene, whereas the excitation and emission
bandwidths for coumarin 343 were 5 and 4 nm, respectively.
Absorbance spectroscopy
Cholic acid methyl ester,32 the alcohol derivative from the re-
duction of cholic acid, mesyl cholate,33 and cholic acid succinim-
ide ester34 were synthesized according to the literature with
minor modifications for dimer 2.35 The synthetic pathway for
dimer 1 started with esterification, followed by reduction and me-
sylation of cholic acid. After these three steps, two equivalents of
mesylated cholic acid were reacted with butylamine to form dimer 1
with a global yield of 70%. The synthetic pathway for dimer 2was first
published by Salunke et al.35 and slightly modified in this work.
Cholic acid was first activated by N-hydroxysuccinimide34 and then
reacted with diethylenetriamine to afford dimer 2 with a global yield
of 67% (Scheme 1).
A stock solution of dimer 1 (2 mmol/L) in chloroform and a
solution of dimer 2 in chloroform–methanol (49/1, v/v, where the
small quantity of methanol was added to help dissolve the dimer)
were prepared and diluted to obtain a series of concentrations
ranging from 0 to 1.0 mmol/L for dimer 1 and from 0 to 2.0 mmol/L
for dimer 2. A fixed amount (10 mg) of cresol red sodium salt was
added to each of the solutions, which were then shaken at room
temperature for 24 h to reach equilibrium and filtered through
a 0.2 ␮m PTFE filter. The filtrates were diluted with methanol,
and the absorbance spectra were recorded on a Cary Series
UV–vis – NIR spectrophotometer (Agilent Technologies). The
dissolved concentration of cresol red probe was calculated based
on the maximum absorption (426 nm) for each solution with
Dimer 1
␧
= 17 500 (mol/L)−1 cm−1 36
.
max
To a solution of mesyl cholate (2.5 g, 5.28 mmol) in anhydrous
dimethylformamide (DMF, 25 mL), butylamine (250 ␮L, 2.52 mmol)
was added. The reaction mixture was stirred at 60 °C for 12 h. After
cooling, water (100 mL) and ethyl acetate (100 mL) were added and
the mixture was separated. The aqueous layer was extracted twice
with ethyl acetate (2 × 50 mL). The organic layer was washed with
brine, dried over MgSO4, filtered, and evaporated under reduced
pressure. Flash chromatography on silica gel (80/20 hexane – ethyl
acetate) afforded dimer 1 as a solid (1.85 g, 88%). 1H NMR (400 MHz,
CDCl3, ␦): 4.06–4.00 (m, 2H, CHOH), 3.96–3.91 (m, 2H, CHOH),
3.90–3.86 (m, 2H, CHOH), 3.32–3.21 (m, 4H, CH2N), 2.61–2.52 (m,
2H, CH2N), 2.18–2.11 (m, 2H), 2.08–1.10 (m, 71H), 1.02 (d, J = 6.60 Hz,
6H, CH3), 0.96 (s, 6H, CH3), 0.73 (s, 6H, CH3) ppm. 13C NMR
(100 MHz, CDCl3, ␦): 77.20, 72.90, 68.16, 58.43, 51.56, 47.18, 46.30,
41.83, 39.43, 36.53, 35.13, 34.96, 33.86, 32.96, 32.80, 32.71, 30.92,
30.43, 29.53, 28.50, 27.42, 26.30, 25.52, 24.53, 23.08, 22.92, 22.59,
17.64 ppm. MS m/z, 826.689 [M + H]+, 848.561 [M + Na]+.
Molecular modeling
A 150 ps molecular dynamics study was performed using Hyper-
Chem (Hypercube Inc.) for the formation of the inclusion com-
plexes; dynamics were conducted at 273 K. A docking approach
for molecular mechanics optimization was used so that the free
energy of binding could be estimated. Calculations were per-
formed using the AMBER Force Field method and using dielectric
constants (THF, 7.6; water, 80) to mimic the solvents. Initial geo-
metric optimization was performed with the optimization algo-
rithm Polak–Ribière with a convergence of 0.01 for the energy
minimization.
Results and discussion
The formation of invertible molecular pockets was confirmed
using the guest compounds pyrene and coumarin 343. To demon-
strate the existence of the hydrophobic pocket in aqueous media,
Published by NRC Research Press

794
Can. J. Chem. Vol. 95, 2017
Fig. 1. Fluorescence spectra of each probe (0.3 ␮mol/L) in the absence and presence of the dimers added at various concentrations (solid lines,
0 ␮mol/L; dashed lines, 5 ␮mol/L; dotted lines, 25 ␮mol/L). (A) Dimer 1 with pyrene in water; (B) dimer 1 with coumarin 343 in THF; (C) dimer 2
with pyrene in water; (D) dimer 2 with coumarin 343 in THF. [Colour online.]
Fig. 2. The concentration of a hydrophilic probe, cresol red sodium
salt, in pure chloroform and a mixture of chloroform and methanol
(49/1, v/v) in the presence of either dimer 1 or 2 at varying
concentrations. [Colour online.]
Fig. 3. Fluorescence spectra of (A) pyrene and (B) coumarin 343,
mixed together in water (both concentrations at 0.3 ␮mol/L) for
increasing concentrations of dimer 1 (solid lines, 0 ␮mol/L; dashed
lines, 5 ␮mol/L; dotted lines, 25 ␮mol/L). [Colour online.]
dimer 1, which bears a butyl group, showed a greater reduction of
intensity in the fluorescence spectra. It was reported previously
that the amine groups on the dimers did not quench the fluores-
cence intensity of pyrene.20,23 The decreased intensity of pyrene
fluorescence spectra in the presence of either of the dimers may
be caused by the phase separation of the dimer–pyrene complex
in the aqueous solutions, which will be discussed later.
To demonstrate the existence of the hydrophilic pocket in a
relatively nonpolar medium, coumarin 343 was added as it is a
typical probe used to identify polar environments.39,40 Both
dimers and the coumarin 343 are quite soluble in THF, and there
was no precipitate observed throughout the fluorescence study in
pyrene was used as the probe. The intensity ratio of the third
emission peak (I3 at 383 nm) to the first emission peak (I1 at
372 nm) in the pyrene fluorescence spectrum is known to increase
when pyrene moves from a hydrophilic environment to a more
hydrophobic one.37,38 A slight increase in the ratio of I3 to I1 was
indeed observed for pyrene in the presence of both dimers
(Figs. 1A and 1C). The same figures also show an overall decrease of
fluorescence intensity when dimer 1 or 2 was added, which may
be caused by the precipitation that was observed during mixing.
This suggests that the host–guest complex of pyrene with each
dimer has very low solubility in the aqueous media. The less polar
Published by NRC Research Press

Zhang et al.
795
Table 1. Molecular dynamics of host–guest approach for dimer 1 and dimer 2 (host) with pyrene (guest) in water at
different stages of complexation.
Time (ps)
0
Dimer 1
Dimer 2
0.5
20
THF, in contrast to the fluorescence tests using pyrene. A red shift,
accompanied by a significant increase in fluorescence intensity, is
expected when the environment surrounding coumarin 343 be-
comes more polar or hydrophilic. The fluorescence spectrum of
coumarin 343 in THF showed weak fluorescence intensity, with a
maximum around 470 nm in the absence of the dimers (solid line
in Figs. 1B and 1D). The addition of cholic acid dimer caused an
increase in the fluorescence intensity of coumarin 343, and the
maxima of the spectra shifted to 485 and 474 nm when in the
presence of dimers 1 and 2, respectively. These results indicate
that coumarin 343 enters the hydrophilic pocket of the dimer to
form a host–guest complex. Dimer 1 may form a more polar
pocket in THF for the accommodation of coumarin 343 than
dimer 2, as evidenced by the larger maximum emission shift on
the fluorescence spectra, as shown in Figs. 1B and 1D.
Cresol red sodium salt, another hydrophilic probe investigated,
presents limited solubility in organic solvents such as chloroform,
and thus its apparent dissolution in such nonpolar media can
indicate the presence of a localized hydrophilic environment.41,42
Therefore, the formation of host–guest complexes with the cholic
acid dimers providing a hydrophilic cavity that helps dissolve a
polar probe in organic solvents was investigated. The absorbance
of homogeneous solutions filtered from a mixture containing
10 mg of the solid probe cresol red sodium salt and increasing
concentrations of dimer were measured at 426 nm, then the ab-
solute absorbance was converted to the concentration of solubi-
Published by NRC Research Press

796
Can. J. Chem. Vol. 95, 2017
Table 2. Computed free energy changes (in kJ/mol) at 0 K for the
complexation of guests (pyrene or coumarin 343) with host dimers 1
and 2 in vacuum and in a solvent.
lized probe. Figure 2 shows that the apparent solubility of the
cresol red probe in either pure chloroform or a mixture of chlo-
roform and methanol (49/1, v/v) increased significantly with in-
creasing concentration of each dimer. The linear relationship
between the concentration of dimer and that of the dissolved
probe confirms the formation of a hydrophilic pocket by both
dimers in this relatively nonpolar solvent.
Guest
Pyrene
Coumarin 343
Host
Host:guest In vacuum In water In vacuum In THF
We also investigated the behavior of a mixture of both pyrene
and coumarin 343 in water as a function of dimer 1 concentration.
This experiment was designed to test whether dimer 1 interacts
preferentially with one of the probes, with both probes simulta-
neously, or if the dissolved probe will interact with the pocket.
Different excitation wavelengths were used to follow the fluores-
cence spectra of the two probes separately in the same solution.
The fluorescence spectra of pyrene (Fig. 3A) remained almost iden-
tical to those of pyrene alone in aqueous solution (Fig. 1A) except
for a red-shifted fluorescence emission band of pyrene at about
470 nm that was not detected for pyrene alone. The existence of
such a red-shifted band in the presence of coumarin 343 may
indicate the formation of a pyrene–coumarin exciplex via ␲–␲
stacking, which has been previously reported to appear at a sim-
ilar shift of about 100 nm.43 Precipitation occurred, as with pyrene
probe alone, with increasing concentration of dimer 1. The fluo-
rescence spectrum of coumarin 343 (Fig. 3B) showed no signifi-
cant change with increasing concentrations of dimer 1, indicating
that there was no observable interaction between coumarin 343
and dimer 1 in water.
The binding stoichiometry with a specific guest probe is an
important characteristic for the evaluation of host molecular
pocket size. If a molecular pocket can encapsulate the probe mol-
ecule at a binding ratio of 1:n (host:guest), the value n can be
calculated by fitting the fluorescence signal of the guest versus the
concentration of the host according to the Benesi–Hildebrand
equation.20 However, the fluorescence studies in this work were
not able to calculate valid values of host:guest binding ratios with
either probe for different reasons. Firstly, due to the low solubility
in water of both dimers and pyrene, the dimer:pyrene complexes
were observed to form suspensions in aqueous solutions, result-
ing in only a decrease of overall fluorescence intensities, instead
of an obvious change in the vibronic band intensity ratio. Sec-
ondly, although both dimers and coumarin 343 formed homoge-
neous solutions in THF, no effective binding ratio could be
obtained from the Benesi–Hildebrand equation by fitting either
the fluorescence intensities or the emission shift. This indicates
that both dimers may interact with coumarin 343 to form com-
plexes other than a simple 1:n (host:guest) binding. As an alternative
method for the evaluation of binding ratios, we used molecular mod-
eling to speculate the structure of host–guest complexes of these two
dimers with the probes pyrene and coumarin 343.
Molecular modeling of the host–guest complexes with a 150 ps
molecular dynamics was carried out. Table 1 shows the simulation
results of the molecular dynamics of dimers 1 and 2 with pyrene
in water up to 20 ps equilibration time, beyond which no signifi-
cant changes were observed. The guest molecule was placed along
a vector orthogonal to the cavity’s opening. As the simulation
progresses, the guest molecule moves into the cholic acid dimer
pocket, mainly due to hydrophobic forces.44 In the last stage of
the approach, pyrene seems to penetrate further inside the cavity
of dimer 1 than that of dimer 2.
To estimate the free energy of binding, a docking approach for
molecular mechanics optimization was used. Molecular docking
is typically used to predict the structure of an intermolecular
complex of two molecules. The preferred orientation of one mol-
ecule is calculated in the presence of the other molecule when
they interact to form a stable complex. The complexation energies
of the hydrophobic guest (pyrene) and hydrophilic guest (cou-
marin 343) with dimers 1 and 2 were calculated by energy mini-
mization in vacuum and in water. The experiments were conducted
Dimer 1 1:1
−91.5
−85.2
−112.9
−193.7
−91.2
−111.6
−104.7
−103.5
−82.4
−66.7
−112.0
−174.7
−81.1
−60.8
−119.4
−108.0
−98.1
−109.5
−160.4
−192.0
−68.1
−97.0
−119.8
−161.7
−70.5
−65.6
−117.7
−162.2
−61.9
−65.6
−108.7
−141.0
1:2
2:1
2:2
Dimer 2 1:1
1:2
2:1
2:2
Note: The bold numbers illustrate the most stable complexes in each case.
in 1:1, 1:2, 2:1, and 2:2 molar ratios of dimer to guest probe to
compare stabilities (Table 2). For example, the 1:1 ratio represents
one host molecule (dimer) encapsulating one guest molecule,
whereas the 2:2 ratio represents a situation where two dimers are
needed to have inclusion of two stacked guest molecules. Table 2
shows that for dimer 1, complexation with more than one pyrene
molecule is favored (ratio of 2:2). A different behavior was pre-
dicted for dimer 2, for which the 2:1 ratio yields a more stable
complex (–119 kJ/mol). Such binding ratios may favor the forma-
tion of dimer–pyrene complexes of a large size, which may be
phase separated from aqueous solutions due to the low solubility
of both dimers in water. The pyrene encapsulated in dimers can-
not be detected by fluorescence spectroscopy and only a fraction
of unencapsulated pyrene remains in the solution to account for
the reduced intensity across the whole fluorescence spectrum.
This is further confirmed by the more reduced fluorescence inten-
sity for the more hydrophobic dimer 1 versus dimer 2. For the
interaction of the inverted pockets with coumarin 343 in THF, a
2:2 host–guest ratio seems to provide the most stable inclusion
complex for both dimers (–162 and –141 kJ/mol, respectively). This
explains why no effective binding ratio could be obtained for 1:n
complexes using the Benesi–Hildebrand equation. A comparison
of the computed free energy changes for the complexes with the
same guests listed in Table 2 seems to indicate that dimer 1, with
a shorter and less flexible linker, forms more stable complexes
with pyrene in water (–175 vs –119 kJ/mol) and with coumarin 343
in THF (–162 vs –141 kJ/mol) than dimer 2.
Figure 4 shows pyrene and coumarin 343 as guests inside the
pockets of dimers 1 and 2 for the most stable conformations in
solvent based on the computed ⌬G changes (Table 1). The 2:2 mix-
ture of the complex of dimer 1 with pyrene, which showed better
stability (–174.7 kJ/mol), reveals the two molecules of pyrene are
facing each other surrounded by four cholic acid moieties (Fig. 4A).
In the dimer 2 : pyrene complex, two molecules of dimer 2 are
needed to encapsulate one hydrophobic probe (Fig. 4B). The ␲–␲
interaction in the stacked pyrene molecules in the 2:2 complexes
is likely responsible for the better stability of the dimer 1 complex
compared with a 1:1 ratio. Simulation of the host:guest complexes
with coumarin 343 show two guest molecules facing each other in
pockets formed by two molecules of either dimer 1 or dimer 2
(Figs. 4C and 4D). It is interesting to note that the improvement in
stability (calculated ⌬G values, Table 2) of the dimer:coumarin 343
complexes was identical (i.e., 2.4 fold) for the two dimers when
going from a ratio of 1:1 to 2:2. This is in contrast to the dimer:
pyrene complexes for which the stability (⌬G values) improved
over 2 fold for dimer 1 and only 1.3 fold for dimer 2 for the same
change from a ratio of 1:1 to 2:2. In addition, the stabilities differ
by only 10% between the 2:1 and 2:2 complexes of dimer 2 with
pyrene in water and by only 1% in vacuum (Table 2) compared with
differences of 56% and 71% in water and vacuum, respectively, for
Published by NRC Research Press

Zhang et al.
797
Fig. 4. Molecular modeling of the host–guest systems. (A) dimer 1 : pyrene = 2:2, (B) dimer 2 : pyrene = 2:1, (C) dimer 1 : coumarin 343 = 2:2,
(D) dimer 2 : coumarin 343 = 2:2. [Colour online.]
dimer 1. The longer linker in dimer 2 may lead to some uncer-
tainty during energy minimization over the 150 ps dynamics such
that the probability of forming a 2:2 complex is similar to that of
the 2:1 complex for dimer 2 : pyrene.
A. Schmitzer for useful discussion regarding molecular modeling
and W. Baille for helpful suggestions.
References
(1) Small, D. M., In The Bile Acids: Chemistry, Physiology, and Metabolism; Nair, P. P.,
Kritchevsky, D., Eds.; Plenum: New York, 1971; Vol. 2.
(2) Sobotka, H. Chem. Rev. 1934, 15, 311. doi:10.1021/cr60052a002.
(3) Stein, T. M.; Gellman, S. H. J. Am. Chem. Soc. 1992, 114, 3943. doi:10.1021/
ja00036a048.
Conclusion
Two dimers of cholic acid with simple linkers were synthesized
and used in the study of host–guest inclusion complexation with
both hydrophobic and hydrophilic guest molecules in water and
organic media. The pockets may be induced to invert themselves
by switching the polarity of the media from water to organic
solvents. This was confirmed by the behavior of three guest mol-
ecules of different polarity. Molecular modeling helped to gain
insight into the behavior of the dimers with the hydrophobic
(pyrene) and hydrophilic (coumarin 343) guests. Dimer 1 formed
the most energetically stable complex with pyrene with a host–
guest ratio of 2:2, whereas dimer 2 formed a less stable 2:2 com-
plex with pyrene, as indicated by a 38% difference in calculated
free energy. The modeling showed that dimer 2 favored the for-
mation of a 2:1 complex with pyrene but only by a slight margin
compared with the 2:2 complex. With coumarin 343, both dimers
favored the formation of host–guest complexes with a ratio of 2:2,
with dimer 1 showing 15% better stability. The results suggest that
the host–guest complex stability may depend on the length and
the functionality of the linker between the two cholic acid moi-
eties in the dimer. Therefore, changes in the linker may allow
fine-tuning of the host–guest complexation behavior for specific
applications of these biomaterials such as selective extraction and
release of environmentally or biologically relevant molecules.
(4) Zhu, X. X.; Nichifor, M. Acc. Chem. Res. 2002, 35, 539. doi:10.1021/ar0101180.
(5) Zhang, J.; Zhu, X. Sci. China, Ser. B: Chem. 2009, 52, 849. doi:10.1007/s11426-
009-0124-x.
(6) Virtanen, E.; Kolehmainen, E. Eur. J. Org. Chem. 2004, 2004, 3385. doi:10.1002/
ejoc.200300699.
(7) Zhang, J.; Luo, J.; Zhu, X. X.; Junk, M. J.; Hinderberger, D. Langmuir 2010, 26,
2958. doi:10.1021/la9028996.
(8) Chakrabarty, A.; Maitra, U. J. Phys. Chem. B 2013, 117, 8039. doi:10.1021/
jp4029497.
(9) Janout, V.; Regen, S. L. Bioconjugate Chem. 2009, 20, 183. doi:10.1021/
bc800296g.
(10) Chen, W. H.; Janout, V.; Kondo, M.; Mosoian, A.; Mosoyan, G.; Petrov, R. R.;
Klotman, M. E.; Regen, S. L. Bioconjugate Chem. 2009, 20, 1711. doi:10.1021/
bc900246u.
(11) Albalak, A.; Zeidel, M. L.; Zucker, S. D.; Jackson, A. A.; Donovan, J. M.
Biochemistry 1996, 35, 7936. doi:10.1021/bi960497i.
(12) Janout, V.; Cline, L. L.; Feuston, B. P.; Klein, L.; O’Brien, A.; Tucker, T.;
Yuan, Y.; O’Neill-Davis, L. A.; Peiffer, R. L.; Nerurkar, S. S.; Jadhav, V.;
Tellers, D. M.; Regen, S. L. Bioconjugate Chem. 2014, 25, 197. doi:10.1021/
bc400506m.
(13) Maitra, U.; Mukhopadhyay, S.; Sarkar, A.; Rao, P.; Indi, S. S. Angew. Chem. Int. Ed.
2001, 40, 2281. doi:10.1002/1521-3773(20010618)40:12<2281::AID-ANIE2281>
3.0.CO;2-L.
(14) Mukhopadhyay, S.; Krishnamoorthy, G.; Maitra, U. J. Phys. Chem. B 2003, 107,
2189. doi:10.1021/jp027079+.
(15) Janout, V.; Lanier, M.; Regen, S. L. J. Am. Chem. Soc. 1996, 118, 1573. doi:10.
1021/ja953261v.
(16) Cline, L. L.; Janout, V.; Fisher, M.; Juliano, R. L.; Regen, S. L. Bioconjugate Chem.
2011, 22, 2210. doi:10.1021/bc200347x.
(17) Chen, W. H.; Wennersten, C.; Moellering, R. C., Jr.; Regen, S. L. Chem.
Biodiversity 2013, 10, 385. doi:10.1002/cbdv.201100431.
Acknowledgements
Financial support from NSERC, FQRNT, and the Canada Re-
search Chair program is gratefully acknowledged. We thank
Published by NRC Research Press

798
Can. J. Chem. Vol. 95, 2017
(18) Zou, T.; Cheng, S. X.; Zhang, X. Z.; Zhuo, R. X. J. Biomed. Mater. Res., Part B
2007, 82B, 400. doi:10.1002/jbm.b.30745.
(19) Wan, T.; Liu, Y.; Yu, J. Q.; Chen, S.; Li, F.; Zhang, X. Z.; Cheng, S. X.;
Zhuo, R. X. J. Polym. Sci., Part A: Polym. Chem. 2006, 44, 6688. doi:10.1002/pola.
21793.
(20) Luo, J.; Chen, Y.; Zhu, X. X. Langmuir 2009, 25, 10913. doi:10.1021/la9013703.
(21) Zhang, J.; Junk, M. J.; Luo, J.; Hinderberger, D.; Zhu, X. X. Langmuir 2010, 26,
13415. doi:10.1021/la102158a.
(32) Gouin, S.; Zhu, X. X. Steroids 1996, 61, 664. doi:10.1016/S0039-128X(96)00187-0.
(33) Aher, N. G.; Pore, V. S. Synlett 2005, 2005, 2155. doi:10.1055/s-2005-872223.
(34) Okahata, Y.; Ando, R.; Kunitake, T. Bull. Chem. Soc. Jpn. 1979, 52, 3647. doi:
10.1246/bcsj.52.3647.
(35) Salunke, D. B.; Hazra, B. G.; Pore, V. S.; Bhat, M. K.; Nahar, P. B.;
Deshpande, M. V. J. Med. Chem. 2004, 47, 1591. doi:10.1021/jm030376y.
(36) Satyanarayana, T. B.; Maitra, U. Chem. Asian J. 2012, 7, 321. doi:10.1002/asia.
201100597.
(22) Luo, J.; Chen, Y.; Zhu, X. X. Synlett 2007, 2007, 2201. doi:10.1055/s-2007-
984909.
(23) Chen, Y. L.; Luo, J. T.; Zhu, X. X. J. Phys. Chem. B 2008, 112, 3402. doi:10.1021/
jp711447r.
(37) Almgren, M.; Grieser, F.; Thomas, J. K. J. Am. Chem. Soc. 1979, 101, 279. doi:
10.1021/ja00496a001.
(38) Dong, D. C.; Winnik, M. A. Photochem. Photobiol. 1982, 35, 17. doi:10.1111/j.1751-
1097.1982.tb03805.x.
(24) Kurtán, T.; Nesnas, N.; Koehn, F. E.; Li, Y. Q.; Nakanishi, K.; Berova, N. J. Am.
Chem. Soc. 2001, 123, 5974. doi:10.1021/ja010250v.
(25) Chelli, S.; Majdoub, M.; Jouini, M.; Aeiyach, S.; Maurel, F.; Chane-Ching, K. I.;
Lacaze, P. C. J. Phys. Org. Chem. 2007, 20, 30. doi:10.1002/poc.1122.
(26) Ghoufi, A.; Morel, J. P.; Morel-Desrosiers, N.; Malfreyt, P. J. Phys. Chem. B
2005, 109, 23579. doi:10.1021/jp054925i.
(39) Grant, C. D.; DeRitter, M. R.; Steege, K. E.; Fadeeva, T. A.; Castner, E. W., Jr.
Langmuir 2005, 21, 1745. doi:10.1021/la047560m.
(40) Jones, G.; Rahman, M. A. J. Phys. Chem. 1994, 98, 13028. doi:10.1021/
j100100a035.
(41) Borucki, B.; Devanathan, S.; Otto, H.; Cusanovich, M. A.; Tollin, G.;
Heyn, M. P. Biochemistry 2002, 41, 10026. doi:10.1021/bi0256227.
(42) Fierke, C. A.; Calderone, T. L.; Krebs, J. F. Biochemistry 1991, 30, 11054. doi:10.
1021/bi00110a007.
(27) Fanelli, F.; De Benedetti, P. G. Chem. Rev. 2005, 105, 3297. doi:10.1021/
cr000095n.
(28) Suzuki, T. J. Chem. Inf. Comput. Sci. 2001, 41, 1266. doi:10.1021/ci010295f.
(29) Tkatchenko, A.; Alfè, D.; Kim, K. S. J. Chem. Theory Comput. 2012, 8, 4317.
doi:10.1021/ct300711r.
(30) Fois, E.; Tabacchi, G.; Calzaferri, G. J. Phys. Chem. C 2012, 116, 16784. doi:10.
1021/jp304962w.
(43) Lee, Y. O.; Lee, J. Y.; Quang, D. T.; Lee, M. H.; Kim, J. S. Bull. Korean Chem. Soc.
2006, 27, 1469. doi:10.5012/bkcs.2006.27.9.1469.
(44) Rao, V. P.; Turro, N. J. Tetrahedron Lett. 1989, 30, 4641. doi:10.1016/S0040-
4039(01)80763-2.
(31) Leahy-Hoppa, M. R.; Cunningham, P. D.; French, J. A.; Hayden, L. M. J. Phys.
Chem. A 2006, 110, 5792. doi:10.1021/jp0565397.
Published by NRC Research Press