Artificial receptors that provides a preorganized hydrophobic environment: A biomimetic approach to dopamine recognition in water

Article abstract of DOI:10.1021/jo051630s

The recognition of dopamine in water has been achieved with tripodal oxazoline-based artificial receptors, capable of providing a preorganized hydrophobic environment by rational design, which mimics a hydrophobic pocket predicted for a human D2 receptor.

Full text of DOI:10.1021/jo051630s

Artificial Receptors That Provides a Preorganized Hydrophobic
Environment: A Biomimetic Approach to Dopamine Recognition in
Water
Jeongryul Kim, Balamurali Raman, and Kyo Han Ahn*
Department of Chemistry and Centre for Integrated Molecular Systems, Pohang UniVersity of Science and
Technology, San 31 Hyoja Dong, Pohang 790-784, Republic of Korea
ahn@postech.ac.kr
ReceiVed August 4, 2005
The recognition of dopamine in water has been achieved with tripodal oxazoline-based artificial receptors,
capable of providing a preorganized hydrophobic environment by rational design, which mimics a hy-
drophobic pocket predicted for a human D2 receptor. The receptors show an amphiphilic nature owing
to the presence of hydrophilic sulfonate groups at the periphery of the tripodal oxazoline ligands, which
seems to contribute in forming the preorganized hydrophobic environment. The artificial receptors
recognized dopamine hydrochloride in water with reasonable selectivity among various organoammonium
guests examined. The observed binding behavior of the receptors was explained by evoking guest inclusion
in the preorganized hydrophobic pocket-like environment and not by simple ion-pairing interactions.
The rationally predicted 1:1 inclusion binding mode was supported by binding studies such as with a ref-
erence receptor that cannot provide a similar binding pocket, Job and VT-NMR experiments, electrospray
ionization mass analysis, and guest selectivity data. This study implies that an effective hydrophobic en-
vironment can be generated even from an acyclic, small molecular artificial receptor. Such a preorganized
hydrophobic environment, as being utilized in biological systems, can be effectively used as a comple-
mentary binding force for the recognition of organoammonium guests such as dopamine hydrochloride
in water.
Introduction
features of most of these artificial DRs, however, are far from
those in biosystems. Most of these artificial receptors provide
a single binding site either for the ammonium ion or the catechol
hydroxyl groups of dopamine, and in most cases their binding
ability is limited to organic solvents only.3 Thus, the recognition
of dopamine in water or at biological pH still remains a
challenging task.4 A recently predicted binding domain of a D2
Dopamine receptors (DRs), a member of the super family of
G-Protein coupled receptors, are known to play an important
role in cellular signaling processes in the nervous system.1 These
DRs are also ideal targets for treating schizophrenia and
Parkinson’s disease. Despite the biological importance, their
exact binding mechanisms in the biological systems are yet to
be understood completely. Meanwhile, considerable efforts have
been focused on the development of artificial DRs in order to
unravel dopamine binding mechanisms at the molecular level
in biological systems.2 Consequently, a sizable number of
artificial receptors have been developed to date. The structural
(2) (a) Schmidtchen, F. P. Z. Naturforsch., C: J. Biosci. 1987, 42, 476-
485. (b) Kimura, E.; Fujioka, H.; Kodama, M. J. Chem. Soc., Chem.
Commun. 1986, 1158-1159. (c) Hayakawa, K.; Kido, K.; Kanematsu, K.
J. Chem. Soc., Perkin Trans. 1 1988, 511-519. (d) Park, C. E.; Jung, Y.
G.; Hong, J. I. Tetrahedron Lett. 1998, 39, 2353-2356. (e) Paugam, M.-
F.; Valencia, L. S.; Bogess, B.; Smith, B. D. J. Am. Chem. Soc. 1994, 116,
11203-11204. (f) Paugam, M.-F.; Biens, J. T.; Smith, B. D.; Christoffels,
A. J.; de Jong, F.; Reinhoudt, D. N. J. Am. Chem. Soc. 1996, 118, 9820-
9825.
(1) (a) Jackson, D. M.; Westlind-Danielsson, A. Pharmacol. Ther. 1994,
64, 291-369. (b) Strange, P. G. AdV. Drug Res. 1996, 28, 313-351.
10.1021/jo051630s CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/24/2005
38
J. Org. Chem. 2006, 71, 38-45

Dopamine Recognition in Water
FIGURE 1. (a) Simplified schematic representation of dopamine binding interactions predicted in a biosystem. (b) PhBTOs and designed biomimetic
receptors 1. (c) Proposed interactions of designed hosts 1 with dopamine‚HCl in water.
human dopamine receptor5 (Figure 1a) indicates that dopamine
is buried in a hydrophobic pocket provided by three aromatic
nuclei of Phe-189, Phe-390, and Trp-386. Additionally, hydro-
gen bonding interactions between the dopamine ammonium ion
and carboxylate group of Asp-114 and between the dopamine
catechol hydroxyl groups and serine-197 and serine-193 are
found to exist as major forces of binding interactions. In general,
it is believed that a balanced combination of electrostatic
interactions, hydrogen bonding, a hydrophobic environment,
π-π stackingm and van der Waals forces are crucial for the
design of artificial receptors in water.6 Biomimetic and shape-
selective artificial adrenergic receptors that can provide ditopic
binding sites, i.e., two different binding sites for ammonium
and catechol, were employed with reasonable success to achieve
shape-selective recognition of noradrenaline in aqueous medium
by Schrader and workers.7 Recently, the same group reported
excellent adrenergic artificial receptors that can provide strong
hydrophobic interactions along with other binding interactions.7c,d
In these receptors the hydrophobic environment is two-
dimensional and arises from either a rigid or cyclic structure of
the host. We reasoned that artificial receptors that can provide
a preorganized, three-dimensional hydrophobic environment
similar to the recently predicted D2 human dopamine receptors
might also be useful for binding catechol amines, particularly
dopamine.5 As far as we know, a preorganized hydrophobic
environment,8 particularly built from an acyclic host system,
has been rarely realized for the molecular recognition in aqueous
medium. Herein, we wish to present a novel class of artificial
dopamine receptors, partial biomimics of a recently predicted
D2 human dopamine receptor, which can provide a preorganized
hydrophobic environment for dopamine binding in sole water.
On the basis of our experience in the field of molecular
recognition and sensing with benzene-based tripodal receptors,
particularly, the phenylglycinol-derived tripodal oxazolines
(PhBTOs),9 we reasoned that a tripodal receptor unit should be
an ideal candidate for the recognition of dopamine as its
ammonium salt, because C3 symmetric tripodal oxazoline
receptors can provide both a complementary binding geometry
for the dopamine ammonium ion and a preorganized hydro-
phobic environment, albeit very flexible, for the dopamine
phenylethyl moiety. To this end, we were particularly intrigued
with the preorganized hydrophobic environment, whether it can
be generated from acyclic precursors as those in biomolecules
and, once generated, if it can be effective or not in water for
the recognition of organoammonium salts such as dopamine and
related amines. The structures of the designed receptors 1 are
shown in Figure 1b. The butanesulfonate groups at the periphery
of the receptors were introduced for dual purposes: the one
obvious reason is to confer water solubility to the receptor and
the other is to provide enhanced preorganization as reasoned in
the following. We reasoned that there could be an additional
driving force for the preorganization of the designed host mole-
cules, that is, an amphiphilic nature of the receptor structure,
in addition to the 2,4,6-trialkyl substituents on the core benzene
ring in 1 (1a, R ) Me; 1b, R ) Et), which drive the three
oxazoline ligands to organize on one side of the core benzene
ring.10 The amphiphilic nature of the host could result from the
self-association of all the sulfonate groups in water. Through
the sulfonate self-association, the oxazolinyl phenyl substituents
of the receptors may provide a highly preorganized and
reasonably “rigid” hydrophobic environment, suitable for bind-
ing the hydrophobic moiety of guest organoammonium ions.
Thus, from Figure 1, it can be seen that the receptor design is
(3) (a) Behr, J.-P.; Lehn, J.-M.; Vierling, P. HelV. Chim. Acta 1982, 65,
1853-1867. (b) Kimura, E.; Watanabe, A.; Kodama, M. J. Am. Chem. Soc.
1983, 105, 2063-2066. (c) Campayo, L.; Bueno, J. M.; Ochoa, C.; Navarro,
P.; Jimenez- Barbero, J.; Pepe, G.; Samat, A. J. Org. Chem. 1997, 62, 2684-
2693.
(4) For dopamine recognition in aqueous media, see: (a) Inoue, M. B.;
Velazquez, E. F.; Inoue, M.; Fernando, Q. J. Chem. Soc., Perkin Trans. 2
1997, 2113-2118. (b) Lamarque, L.; Navarro, P.; Miranda, C.; Ara´n, V.
J.; Ochoa, C.; Escart´ı, F.; Espan˜a, E. G.; LaTorre, J.; Luis, S. V.; Miravet,
J. F. J. Am. Chem. Soc. 2001, 123, 10560-10570. (c) Coskun, A.; Akkaya,
E. U. Org. Lett. 2004, 6, 3107-3109. (d) Secor, K. E.; Glass, T. E. Org.
Lett. 2004, 6, 3727-3730. (e) Mannironi, C.; Di Nardo, A.; Fruscoloni, P.;
Tocchini-Valentini, G. P. Biochemistry 1997, 36, 9726-9734.
(5) (a) Kalani, M. Y. S.; Nagarajan, V.; Spencer, E. H.; Rene, J. T.;
Peter, L. F.; Kalani, M. A. K.; Wely, B. F.; Victor, W. T. K.; Goddard, W.
A., III. Proc. Natl. Acad. Sci. 2004, 101, 3815-3820. For the role of serine
in dopamine binding, see: (b) Liapakis, G.; Ballesteros, J. A.; Papachristou,
S.; Chan, W. C.; Chen, X.; Jonathan, A. J. J. Biol. Chem. 2000, 275, 37779-
37788.
(6) Fersht, A. R.; Serrano, L. Curr. Opin. Struct. Biol. 1993, 3, 75-83.
(7) (a) Herm, M.; Molt, O.; Schrader, T. Angew. Chem., Int. Ed. 2001,
40, 3148-3151. (b) Herm, M.; Molt, O.; Schrader, T. Chem. Eur. J. 2002,
8, 1485-1498. (c) Molt, O.; Rubeling, D.; Schrader, Y. J. Am. Chem. Soc.
2003, 125, 12086-12087. (d) Molt, O.; Rubeling, D.; Schafer, G.; Schrader,
T. Chem. Eur. J. 2004, 10, 4225-4232.
(8) For preorganization effects, see: (a) Lehn, J. M. Supramolecular
Chemistry, Concepts and PerspectiVe; VCH: Weinheim, 1995. (b) Still,
W. C. Acc. Chem. Res. 1996, 29, 155 and references therein.
(9) (a) Kim S.-G.; Ahn, K. H. Chem. Eur. J. 2000, 6, 3399-3403. (b)
Kim, S.-G.; Kim, K.-H.; Jung, J.; Shin, S. K.; Ahn, K. H. J. Am. Chem.
Soc. 2002, 124, 591-596. (c) Kim, S.-G.; Kim, K.-H.; Kim, Y. K.; Shin,
S. K.; Ahn, K. H. J. Am. Chem. Soc. 2003, 125, 13819-13824.
J. Org. Chem, Vol. 71, No. 1, 2006 39

Kim et al.
SCHEME 1. Synthesis of Receptors 1 and Reference
Compound 2a
FIGURE 2. 1H NMR spectral changes of 1a (4.0 mM) upon addition
of dopamine hydrochloride (0, 0.8, 1.6, 2.5, 3.3, 4.1, 4.9, 6.6, 9.8, 16.5,
50 equiv) in water. Only an enlarged region (6-8 ppm) is shown, in
which a phenyl proton Hf of the receptor is indicated with arrows (guest
equivalents from 0.0 to 16.5).
a Reaction conditions: (a) (COCl)2, Et3N, CH2Cl2; (b) MsCl, Et3N,
DMAP, CH2Cl2, 45% (5a), 29% (5b); (c) 1 N NaOH, MeOH, rt, 6 h, 99%;
(d) NaH, DMF, 1,4-butanesultone, rt, 36 h, 63% (1a), 65% (1b).
intended to provide a preorganized hydrophobic pocket-like
environment to the dopamine’s lipophilic moiety. The receptor
is expected to bind dopamine in a 1:1 fashion via tripodal
hydrogen bonding between the oxazolines of the receptor and
the ammonium ion of the dopamine.9 The ether linkage allows
a synthetically simple manipulation for the introduction of the
butanesulfonate groups to our PhBTOs. Also, these ethereal or
the sulfonate oxygens may provide hydrogen bonding sites for
the catechol hydroxyl groups of dopamine. In such a case, the
designed receptors 1 would show higher selectivity toward
dopamine compared to simple 2-phenethylamine. Furthermore,
a higher selectivity for dopamine over the other adrenergic
amines with R- or â-branches could be expected for the simple
reason that PhBTOs accommodate linear ammonium ions better
than R- or â-branched ammonium ions owing to their shape
complemetarity.9a,b Also, cation-π interactions between the
ammonium and the core benzene ring of the host are expected
to have a stabilizing influence on the host-guest complexation
process.
tanesulfonate group yielded receptors 1 in high yields. The bis-
(oxazoline) analogue 2 as a reference compound was synthesized
similarly, with which valuable information on the binding mode
was obtained.
Our initial experimentation to evaluate the potential of our
receptors 1 as DRs and to validate our reasoning began with
the 1H NMR titration of receptor 1a with increasing concentra-
tions of dopamine hydrochloride in pure D2O (Figure 2).13 The
protons of the oxazoline phenyl rings (peaks at δ 6.78, 6.89,
6.92, 7.26 ppm) and the oxazoline ring protons (peaks at δ 4.11,
4.67, 5.12 ppm) showed significant complexation-induced
chemical shifts (CICS). However, the CICS of the oxazoline
ring protons could not be followed during NMR titration because
they merged with the solvent peak; therefore, the phenyl ring
protons were followed. We observed upfield CICS for the
oxazoline phenyl protons (for example, peaks at δ 6.78, 6.89
ppm), which is ascribed to the aromatic π-π stacking between
the aromatic ring of the dopamine and that of the receptor 1a.
Also, the CH2 protons of the dopamine (peaks at δ 2.82, 3.12
ppm) displayed significant upfield shifts upon complexation (∆δ
) 0.45, 0.75 ppm, respectively, at the saturation point), which
can be explained by assuming that the dopamine binds inside
the hydrophobic pocket composed of the three phenyl rings as
predicted. Such upfield shifts of the R- and â-methylene units
of organoammonium ions were ascribed to diamagnetic shield-
ing by the surrounding phenyl rings, which was previously
identified by an X-ray crystal structure of an inclusion complex
of PhBTOs.9a A Job plot14 was obtained by the continuous
variations method (Figure 3), which also suggested the proposed
1:1 binding fashion. An association constant of Kassoc ) 161
M-1 was obtained by nonlinear least-squares fitting for the 1H
Results and Discussion
The synthesis of the receptors 1 is depicted in Scheme 1.
The tricarboxylic acid 4a and 4b9b were converted to the
corresponding oxazolines by treating with amino alcohol 3.11
Deprotection of the tert-butyldimethylsilyl (TBS) group with
aqueous hydroxide12 and subsequent introduction of the bu-
(10) For selected examples for the use of other benzene-based tripodal
ligands in molecular recognition and self-assembly, see: (a) Metzger, A.;
Lynch, V. M.; Anslyn, E. V. Angew. Chem., Int. Ed. Engl. 1997, 36, 862-
865. (b) Niikura, K.; Metzger, A.; Anslyn, E. V. J. Am. Chem. Soc. 1998,
120, 8533-8534. (c) Szabo, T.; O’Leary, B. M.; Rebek, J., Jr. Angew.
Chem., Int. Ed. 1998, 37, 3410-3413 (d) Sato, K.; Arai, S.; Yamagishi, T.
Tetrahedron Lett. 1999, 40, 5219-5222. (e) Chin, J.; Walsdorff, C.; Stranix,
B.; Oh, J.; Chung, H. J.; Park, S.-M.; Kim, K. Angew. Chem., Int Ed. 1999,
38, 2756-2759. (f) Lavigne, J. L.; Anslyn, E. V. Angew. Chem., Int. Ed.
1999, 38, 3666-3669. (g) Rekharsky, M.; Inoue, Y.; Tobey, S.; Metzger,
A.; Anslyn, E. J. Am. Chem. Soc. 2002, 124, 14959-14967.
(12) Attempts to deprotect the silyl group under various other known
conditions that use a fluoride or hydrofluoric acid source failed.
(13) See Supporting Information for complete spectral changes.
(14) (a) Job, P. Compt. Rend. 1925, 180, 928. (b) Blanda, M. T.; Horner,
J. H.; Newcomb, M. J. Org. Chem. 1989, 54, 4626-4636.
(11) Ahn, K. H. and co-workers, manuscript in preparation.
40 J. Org. Chem., Vol. 71, No. 1, 2006

Dopamine Recognition in Water
complex, [R + 2G + H+ - Cl-]2+ at m/z ) 713.73, was also
observed as minor peak. The observation of such a 1R:2G
complex peak may be explained by further existence of ionic
interactions between the 1R:1G inclusion complex and the guest.
To get an insight on the host-guest complexation process,
we evaluated the thermodynamic parameters for the binding
process using the variable temperature NMR technique (VT
NMR). Thus, the binding process was followed by 1H NMR at
temperatures in the range of 5-65 °C, and the van’t Hoff
equation was employed to extract the thermodynamic data. The
thermodynamic data so obtained for the binding process between
receptor 1a and dopamine hydrochloride showed that the host-
guest complexation involves negative enthalpy and positive
entropy changes (∆H° ) -3.8 kJ‚mol-1 and T∆S° ) 8.9
kJ‚mol-1 (T ) 303 K); from which Kassoc ) 148 M-1 can be
extracted, similar to that obtained by the NMR titration). Thus,
the host-guest complexation process is entropy-driven rather
than enthalpy-driven. The entropy gain would have resulted from
the release of a number of water molecules during the binding
process of the ionic guest in going from the solvated state into
the preorganized hydrophobic pocket. This entropy-driven
binding process implies that simple ion-pairing interactions
between the guest ammonium ions with the host sulfonate
groups are not the main interactions responsible for the observed
association constants, since there is no reason to expect a
positive entropy change for such a recombination process of
ion pairs between the sodium butanesulfonate and the am-
monium chloride in water. If such a recombination process of
ion pairs were the dominant source of the binding affinity
observed, we should have obtained a different stoichiometry,
not the 1:1 binding mode, from the Job plot and ESI MS
analysis, because 3 molar equiv of ionic guests were required
to match the three sulfonate groups of the hosts. Also, a large
difference in the basicity of oxazoline nitrogen (pKa ≈ 5)9a and
the dopamine amine (pKa ) 10.6) excludes a possible proton-
transfer complex formation between these two groups.17 To
exclude a possibility of such simple ionic interactions as the
major binding force, we also performed an NMR titration of
1a with dopamine hydrochloride in a phosphate buffer solution
(pH 7.0). The significant binding affinity (Kassoc ) 122 ( 18
M-1) so obtained, which is almost comparable with those in
pure water, rules out such an ion-pairing process for the observed
binding affinity. However, the ion-pairing process should not
be excluded but rather be incorporated with “overall” binding
processes, of which only the inclusion binding mode as shown
in Figure 1c manifests itself in the binding behavior observed
by NMR experiments.
FIGURE 3. (a) A Job plot for the same process. (b) 1H NMR titration
curve for complexation of receptor 1a with dopamine hydrochloride
in D2O: complexation-inducedchemical shift of the Hf vs guest
equivalent.
NMR titration curve using WinEQNMR program.15 This Kassoc
value is better than or comparable to literature values determined
in aqueous medium.16 In water, the desired hydrogen bonding
and hydrophobic interactions are likely to compete with
“nonspecific” interactions such as ion-pairing, which may
account for the observed weak association constants.
Further support for the predicted 1:1 host-guest complexation
process was obtained by electrospray ionization (ESI) mass
spectrometry.
The ESI MS spectrum of an equimolar solution of the receptor
1a and dopamine hydrochloride (Figure 4) exhibited peaks
corresponding to the 1:1 receptor-guest complex: [R + G +
H+ - Cl-]2+, m/z ) 637.18; [R + G + Na+ - Cl-]2+, 648.19;
[R + G - Cl-]+, 1273.34; where R represents the receptor 1a
and G represents dopamine‚HCl. In addition to these, a 1R:2G
The ability of our receptors 1 to provide the predicted, highly
preorganized hydrophobic pocket-like environment for the
dopamine binding could be further supported by comparing its
dopamine binding ability with the reference receptor 2, which,
as can be easily reasoned from its open structure (Scheme 1),
cannot provide a preorganized hydrophobic pocket-like environ-
ment for the guest binding. When the complexation of the
reference receptor with dopamine hydrochloride was followed
1
by H NMR, otherwise under identical conditions as that of
receptor 1a, we could not observe any CICS. This observation
supports that the binding affinity observed for our receptor 1a
is originated mainly from its 3D hydrophobic environment and
(15) (a) Schneider, H. J.; Kramer, R.; Simova, S.; Schneider, U. J. Am.
Chem. Soc. 1988, 110, 6442-6448. (b) Hynes, M. J. J. Chem. Soc., Dalton
Trans. 1993, 311-312.
(16) Association constants reported for dopamine in aqueous media. In
ref 7a, Kassoc ) 246 ( 38% M-1 in MeOH/D2O (1: 1). In ref 7b, Kassoc
)
142 ( 14% M-1 in D2O/MeOH (1:1). In ref 4a log Kassoc ) 1.2 M-1 in
D2O by 1H NMR. In ref 4b, maximum of Kassoc ) 63.4 ( 2 M-1 in aqueous
NaCl by pH-metric titration.
(17) Sa´nchez-Revera, A. E.; Corona-Avendan˜o, S.; Alarco´n-Angeles, G.;
Rojas-Herna´ndez, A.; Ram´ırez-Silva, M. T.; Romero-Romo, M. A. Spec-
trochim. Acta, Part A 2003, 59, 3193-3203.
J. Org. Chem, Vol. 71, No. 1, 2006 41

Kim et al.
FIGURE 4. ESI mass spectrum of a 1:1 mixture of receptor 1a and dopamine‚HCl.
as the function of host concentration yielded a curve, and
analysis of the curve yielded the corresponding self-association
constants for the monomer-dimer equilibrium. Although the
self-association constants (Ksa) obtained were quite low, being
Ksa ) 40 ( 3 and 73 ( 9 [M-1] for the receptors 1a and 1b,
respectively, this result shows an amphiphilic nature of our
receptors in water.19
The slightly higher self-association constant for the receptor
1b compared to 1a could have resulted from its higher degree
of preorganization, as the 2,4,6-triethyl substituents are known
to show higher preorganizing ability than the 2,4,6-trimethyl
substituents. Moreover, this difference in the self-association
constants indicates that the receptors’ self-association results
from the preorganized shuttlecock structures in which the
sulfonate groups are on one side of the core benzene ring. In
other words, the butanesulfonate groups associate themselves
in water, and this process assists the preorganization of the
receptors as we intended in their design. We observed frothing
upon shaking the test solutions containing either free receptors
or a mixture of receptor and guest. This observation also
suggests the existence of the amphiphilic average nature of our
receptors. Thus, our reasoning that the introduction of alkyl
sulfonate groups at appropriate position to our PhBTO system
would lead to water-soluble receptors that can provide highly
preorganized structures is validated.
FIGURE 5. Model structure and its CPK view of an inclusion complex
between 1b and dopamine ammonium ion.
not from the simple ion-pairing; were the latter situation
operating, there is no reason to expect such a difference in the
binding affinity between the two receptors 1a and 2. Our efforts
to get additional evidence for the guest binding in the hydro-
phobic pocket by NOESY experiments did not yield fruitful
results, possibly owing to a highly flexible nature of the
inclusion complex.
A molecular modeling study18 indicated that the distance
between the aromatic protons of the hosts and the guest protons
were above the NOE scale (Figure 5).
To validate our reasoning that the butanesulfonate arms may
associate themselves, which in turn assist the preorganization
process, we have studied the self-associating ability of our
receptors 1. The receptors 1 may undergo self-association
through a shuttlecock shape, which can be obtained as the
sulfonate groups associate themselves. Thus, the 1H NMR
spectra of the receptors were recorded at various concentrations
(0.1-20 mM in D2O) and the chemical shifts of the phenyl
ring protons were followed (Figure 6). A plot of chemical shift
A close look at Tables 1 and 2 reveals an interesting and
nonnegligible observation that the association constants of the
receptors 1 toward a particular ammonium guest studied do not
show a considerable increase upon changing from the 2,4,6-
trimethyl based receptor 1a to the 2,4,6-triethyl based 1b. We
were intrigued by this little change because it is well-known
(19) The amphiphilic nature of our receptors open up their possible
application to the molecular recognition at the air/water interface by
embedding them in a monolayer of stearic acid, as beautifully demonstrated
by Schrader and co-workers; see: Molt, O.; Schrader, T. Angew. Chem.,
Int. Ed. 2003, 42, 5509-5513.
(18) Molecular mechanics computation was performed using Spartan ‘04
Windows from Wavefunction, Inc.
42 J. Org. Chem., Vol. 71, No. 1, 2006

Dopamine Recognition in Water
FIGURE 6. Stack plots of 1H NMR spectra of receptor 1a showing its self-aggregation behavior upon increasing concentration (A region around
7 ppm is enlarged and shown at left).
TABLE 1. Association Constants Obtained for the Complexation
of Receptor 1a toward Various Organoammonium Ions in D2O by
1H NMR Titrations
was attributed to the tendency of the 2,4,6-triethyl groups to
force a higher degree of preorganization than that of the 2,4,6-
trimethyl groups. The absence of such substantial increase in
the association constants in the cases of receptors 1a and 1b
may be understood as follows. Even with the 2,4,6-trimethyl
substituents, the receptor 1a might have gained a fair degree of
preorganization due to the amphiphilic nature of the receptors
in water by the presence of the ionic sulfonate groups; hence,
the effect of the 2,4,6-triethyl groups in the preorganization is
less felt. Thus, the above observation further augments the
predicted participation of the sulfonate groups in the preorga-
nization of the receptors.
Kassoc(1:1)
-∆G
∆δsat
stoichio-
metryd
b
guesta
[M-1
]
[kJ mol-1
]
[ppm]c
2-phenethylamine
tyramine
dopamine
82 ( 12%
101 ( 7%
161 ( 16%
<1
10.99
11.51
12.67
0.29
0.32
0.32
1:1
1:1
1:1
catechol
ethanolamine
acetylcholine
adrenaline
noradrenaline
DL-tyrosine methyl ester 65 ( 11%
GABA
<1
<1
<1
67 ( 6%
10.48
10.41
0.13
0.17
1:1
1:1
<1
The complex forming ability of receptors 1 was examined
for other biogenic organoammonium ions (Figure 7), which are
structurally related to dopamine, and the obtained association
constants (Kassoc) are collected in Tables 1 and 2.
a As the hydrochloride salt. b Errors are calculated as standard deviations
from the nonlinear regression. c Largest shifts from selected CH protons.
d Determined by Job plots and curve fitting of the titration curves.
The trend of the binding constants observed reveals again
that the aromatic π-π interactions and hydrophobic nature of
the guest are essential for the binding process; i.e., ethanolamine,
acetylcholine, and GABA which do not have an aromatic
nucleus did not show any binding. This result clearly indicates
that a simple ion-pairing process between the host and guest,
apparent in these cases if any, does not lead to any change in
the NMR titration spectra; hence, the observed association
constants in other cases are owing to the predicted hydrogen
bonding and hydrophobic interactions. If only the hydrophobic
interactions alone were responsible for binding, then catechol
should have shown association. However, absence of such
association indicates that the hydrogen bonding of ammonium
ions to the tripodal oxazoline nitrogens, which can be only
provided in the hydrophobic pocket-like environment. Thus, how
well the ammonium guests fit in the binding pocket of receptors
1 should affect the binding process. Consequently, the un-
branched 2-phenylethylammonium analogues (2-phenylethyl-
amine, tyramine, and dopamine salts) that fit better in the
binding pocket showed stronger binding affinities compared to
the branched ones (noradrenaline and tyrosine methyl ester
salts).
TABLE 2. Association Constants Obtained for Complexes of
1
Receptor 1b and Various Organoammonium Ions in D2O by H
NMR Titrations
Kassoc(1:1)
-∆G
∆δsat
stoichio-
metryd
b
guesta
[M-1
]
[kJ mol-1
]
[ppm]c
2-phenethylamine
tyramine
dopamine
86 ( 14%
92 ( 16%
178 ( 15%
<1
11.11
11.27
12.92
0.32
0.33
0.36
1:1
1:1
1:1
catechol
ethanolamine
acetylcholine
adrenaline
noradrenaline
DL-tyrosine methyl ester 72 ( 17%
GABA
<1
<1
<1
74 ( 14%
10.73
10.66
0.21
0.22
1:1
1:1
<1
a As the hydrochloride salt. b Errors are calculated as standard deviations
from the nonlinear regression. c Largest shifts from selected CH protons.
d Determined by Job plots and curve fitting of the titration curves.
that a benzene based tripodal system with 2,4,6-triethyl sub-
stituents has better tendency toward preorganization of the 1,3,5-
tripodal ligands than does the 2,4,6-triemethyl analogue.10 We
have observed, in the case of a simple PhBTO receptor, an
increase up to 2 orders of magnitude in the association constants
toward linear organoammonium ions, in going from the 2,4,6-
trimethyl based receptor to the 2,4,6-triethyl based ones in
organic medium.9a This increase in the association constants
As predicted, the receptors 1 showed higher selectivity toward
the linear 2-phenethylammonium analogues such as dopamine
over the branched ones such as noradrenaline salt. As mentioned
previously, the tripodal oxazolines offer only three complement-
J. Org. Chem, Vol. 71, No. 1, 2006 43

Kim et al.
FIGURE 7. Structures of the guest ammonium ions studied.
ing hydrogen bonding sites that can accommodate linear
ammonium ions much better than R- or â-branched ammonium
ions.9 In the hydrophobic pocket generated by the tripodal
oxazoline ligands, we can expect that guests such as catechol
that has no ammonium group, acetylcholine that is quaternary
ammonium salt, and adrenaline that is a secondary amine salt
cannot or weakly bind, as demonstrated. This relationship
between the substrate structures and binding affinity is correlated
very well with the previously observed results with PhBTOs.9
This structure-affinity correlation is the most obvious evidence
for the proposed inclusion binding mode.
A notable feature is that the receptors 1 show higher affinity
toward dopamine compared to other structurally related am-
monium ions, 2-phenylethylamine and tyramine. This selectivity
suggests the participation of the catechol hydroxyl groups in
the host-guest interactions. The binding of the dopamine in
the hydrophobic pocket results in such a fashion that the catechol
hydroxyl groups extend out of the hydrophobic pocket (Figure
1b),20 where they may enter into hydrogen bonding with the
sulfonate or the phenyl ether oxygens of the receptor. The CPK
model in Figure 5 suggests that the catechol hydroxyl groups
can have hydrogen bonding with sulfonate groups. The trend
of the Kassoc values in the Table 1 and Table 2, dopamine (two
OH groups) > tyramine (one OH group) > 2-phenethylamine
(no OH group), is in accordance with this explanation. Our
attempts to get supporting evidence for the supposed hydrogen
bonding participation by the sulfonate or phenyl ether oxygens
in aqueous media, which is an obviously challenging task,
through IR and NMR studies were not successful.
recognized dopamine hydrochloride with reasonable selectivity
among various organoammonium salts evaluated, which could
be explained by complementary molecular interactions involving
a hydrophobic pocket-like environment and not by simple ion-
pairing interactions. Binding studies such as with a reference
compound that lacks a similar hydrophobic environment as well
as by Job and VT-NMR experiments, electrospray ionization
mass analysis, and guest selectivity data gave supporting results
for the inclusion binding mode suggested. An amphiphilic nature
of the receptors, evidenced by their self-association constants
and their froth-forming tendency in water, is supposed to
contribute to the receptors’ preorganization, thereby enabling
the acyclic receptors with an effective pocket-like hydrophobic
environment. We took an advantage of this preorganization to
recognize dopamine hydrochloride and related ions in water.
An acyclic receptor system providing such a highly preorganized
hydrophobic pocket for guest binding in water is indeed unique.
This idea seems to be general and could be extended to design
a range of receptors for recognition and sensing of biogenic
molecules in water. A synergistic effect of stronger binding
interactions such as ionic salt bridges and preorganized hydro-
phobic interactions should result in new receptors with more
improved affinity, which is a subject of our next study.
Experimental Section
All commercial reagents are of ACS reagent grade and used as
supplied. All solvents were dried over 4 Å molecular sieves when
necessary. Column chromatography was carried out on silica gel
having 230-400 mesh. Melting points were obtained with an
electrothermal capillary apparatus and are uncorrected. Optical
rotations were measured using a sodium lamp (D line, 589 nm)
and are reported in degrees with concentration in unit of 10 mg
Conclusion
The molecular recognition of dopamine hydrochloride in sole
water has been achieved by a rationally designed, new class of
artificial receptors that provides a highly preorganized hydro-
phobic environment, mimicking a hydrophobic pocket predicted
for a human D2 receptor. The new receptors were derived from
our benzene-based tripodal oxazoline system by introducing
butanesulfonate groups at the periphery of the oxazoline ligands,
which endowed the receptors with not only water solubility but
also enhanced preorganization proposed. Our artificial receptors
1
mL-1. H and 13C spectra were recorded at ambient temperature
and all chemical shifts are reported as δ in parts per million (ppm)
downfield from tetramethylsilane (δ ) 0.0) using the residual
solvent signal as an internal standard. Mass spectral data are
reported in the unit of mass to charge (m/z).
(S,S,S)-2-{[3,5-Bis({4-[3-(tert-butyldimethylsilanyloxy)phenyl]-
4,5-dihydrooxazol-2-yl}methyl)-2,4,6-trimethyl]phenyl}methyl-
4-[3-(tert-butyldimethylsilanyloxy)phenyl]-4,5-dihydrooxazole (5a).
To a suspension of triacid 4 (1.30 g, 4.42 mmol) in dichloromethane
(63 mL) were added oxalyl chloride (1.93 mL, 22.1 mmol) and
N,N-dimethylformamide (0.15 mL, 2.21 mmol). After being stirred
for 24 h at room temperature, solvent and excess oxalyl chloride
were evaporated under reduced pressure. The crude acyl chloride
was immediately used for the next reaction without further
(20) See ref 9b for the X-ray crystal structure of the inclusion complex
of PhBTO with 2-phenethylammonium ion. This reveals that the phenyl
ring of a 2-phenyethylammonium ion is not completely buried in the
hydrophobic pocket provided by the tripodal host.
44 J. Org. Chem., Vol. 71, No. 1, 2006

Dopamine Recognition in Water
purification. To a solution of amino alcohol 3 (3.31 mg, 12.4 mmol)
and triethylamine (3.1 mL, 22.1 mmol) in dichloromethane at 0
°C was added the above acyl chloride in dichloromethane (43 mL)
dropwise via cannula. After being stirred for 12 h at room
temperature, methanesulfonyl chloride (1.13 mL, 14.6 mmol),
triethylamine (4.9 mL, 35.4 mmol), and 4-(dimethylamino)pyridine
(162 mg, 1.33 mmol) were added to the reaction mixture. After
being stirred for additional 24 h at room temperature, the mixture
was poured into an Erlenmeyer flask containing a mixture of water/
dichloromethane. The combined organic layer was extracted,
washed with brine, dried over anhydrous MgSO4, and concentrated
to dryness. The crude residue was purified by column chromatog-
raphy (hexane/EtOAc, 6:4) to afford the desired tris(oxazoline) 5a
(1.84 g, 45%) as a white solid: mp 93-94 °C; [R]18D ) -40.2 (c
,and filtered to give a pale brown powder. The crude product was
dissolved in a minimum amount of methanol and precipitated out
with EtOAc. The precipitate was filtered and washed with EtOAc
to give the desired product 1a (220 mg, 63%): mp >270 °C,
1
decomposed; [R]23 ) -39.2 (c ) 0.5, CH3OH); H NMR (300
D
MHz, D2O) δ 7.25 (t, J ) 7.9 Hz, 3H), 6.91 (dd, J ) 1.9, 8.2 Hz,
3H), 6.82 (d, J ) 7.7 Hz, 3H), 6.78 (s, 3H), 5.12 (t, J ) 8.1 Hz,
3H), 4.66 (t, J ) 9.2 Hz, 3H), 4.08 (t, J ) 8.2 Hz, 3H), 4.01 (t, J
) 5.5 Hz, 6H), 3.89 (s, 6H), 2.96 (t, J ) 7.3 Hz, 6H), 2.37 (s, 9H),
1.86 (m, 12H); 13C NMR (75 MHz, D2O) δ 168.1, 157.9, 143.2,
135.2, 129.9, 129.6, 118.4, 113.3, 112.2, 74.3, 67.2, 67.2, 50.2,
28.6, 26.9, 20.4, 16.0; MS (FAB) m/z (rel intensity) 1120 (M + 1,
90), 1098 (100), 1076 (49), 1053 (53); HRMS (FAB) calcd for
C51H61N3O15S3Na3 1120.2958, found 1120.2946.
1
) 0.5, CHCl3); H NMR (300 MHz, CHCl3) δ 7.13 (dd, J ) 8.4,
Evaluation of the Complex Stoichiometry. The stoichiometry
of the complexes was determined according to the Job’s method
of continuous variations. Equimolar amounts of host and guest were
dissolved in D2O. These solutions were distributed among 10 NMR
tubes in such a way that the molar fraction XH (XH ) [H]0/([H]0 +
[G]0)) in the resulting solutions decreased from 0.0 to 1.0. The
complexation-induced chemical shifts (CICS) were multiplied by
XH and plotted against XH itself (Job plot).
ESI Mass Analysis. A sample solution (20 µL) of an 1:1 mixture
of host 1a and dopamine hydrochloride (each 1.4 µL in distilled
water) was introduced at flow rates of 5 µL min-1 and ion spray
potential of 4.0 kV (positive ESI). About 60 scans were averaged
to improve the signal-to-noise ratio. The region of m/z 1260-1290
is magnified and shown separately after m/z 800 (Figure 4).
Evaluation of Association Constant Kassoc. The host compound
was dissolved in 6.6 mL of D2O, and the resulting solution was
evenly distributed among 11 NMR tubes. The first NMR tube was
sealed without any guest. The guest (100 equiv corresponding to
the host) was also dissolved in 1.22 mL of D2O and added in
increasing amounts to the NMR tubes, so that finally solutions with
the following relative amounts (equiv) of the guest versus host
compound were obtained: 0, 0.8, 1.6, 2.4, 3.2, 4.1, 4.9, 6.5, 9.8,
16.4, 50.0. All ∆δ values refer to the standard of the pure host
compound. Volume and concentrations changes were taken into
account during analysis. The association constants were calculated
by nonlinear regression methods.
7.8 Hz, 3H), 6.78 (d, J ) 7.8 Hz, 3H), 6.70 (d, J ) 8.4 Hz, 3H),
6.69 (s, 3H), 5.07 (dd, J ) 10.2, 9.0 Hz, 3H), 4.53 (dd, J ) 10.2,
8.4 Hz, 3H), 3.40 (dd, J ) 9.0, 8.4 Hz, 3H), 3.83 (s, 6H), 2.48 (s,
9H), 0.96 (s, 27H), 0.16 (s, 18H); 13C NMR (75 MHz, CHCl3) δ
167.5, 156.2, 144.6, 136.3, 131.1, 130.0, 119.9, 119.6, 118.8, 75.2,
69.8, 30.5, 26.1, 18.6, 17.7, -4.0; MS (FAB) m/z (rel inten-
sity) 988 (M + 1, 100), 738 (12), 250 (31). Anal. Calcd for
C57H81N3O6Si3‚1/2H2O: C, 68.63; H, 8.29; N, 4.21. Found: C,
68.60; H, 8.44; N 4.32.
(S,S,S)-2-[(3,5-Bis{[4-(3-hydroxyphenyl)-4,5-dihyrooxazol-2-
yl]methyl}-2,4,6-trimethyl)phenyl]methyl-4-(3-hydroxyphenyl)-
4,5-dihydrooxazole (6a). Compound 5a (50 mg, 0.05 mmol) was
dissolved in methanol (1.7 mL) and 1 N NaOH solution (0.5 mL,
0.5 mmol) was added dropwise during 30 min under stirring at
room temperature. After being stirred for 6 h, the crude product
was cooled to 0 °C and neutralized with 0.5 N HCl, during which
the product was precipitated. The precipitate was filtered and
washed with distilled water to afford a desired product 6a (33 mg,
99%) as white solids: mp > 195 °C, decomposed; [R]19D ) -42.0
1
(c ) 1.20, DMSO); H NMR (300 MHz, DMSO-d6) δ 9.63 (s,
3H), 7.04 (dd, J ) 8.4, 7.8 Hz, 3H), 6.60-6.54 (m, 9H), 4.98 (dd,
J ) 9.6, 8.1 Hz, 3H), 4.52 (dd, J ) 9.6, 8.7 Hz, 3H), 3.84 (dd, J
) 8.7, 8.1 Hz, 3H), 3.71 (s, 6H), 2.34 (s, 9H); 13C NMR (75 MHz,
DMSO-d6) δ 166.7, 158.3, 145.3, 135.9, 131.4, 130.3, 117.8, 114.9,
114.2, 75.0, 69.4, 30.3, 17.6; MS (FAB) m/z (rel intensity) 646 (M
+ 1, 5), 460 (12), 307 (64), 154 (100); HRMS (FAB) calcd for
C39H39N3O6 646.2917, found 646.2916.
Acknowledgment. We thank Prof. Hynes, M. J. at National
University of Ireland for providing us with the Win EQNMR
program. This work was financially supported by the Center
for Integrated Molecular Systems, POSTECH.
(S,S,S)-4-{3-[2-(2,4,6-Trimethyl-3,5-bis-{4-[3-(4-sulfobutoxy)-
phenyl]-4,5-dihydrooxazol-2-ylmethyl}benzyl)-4,5-dihydrooxazol-
4-yl]phenoxy}butane-1-sulfonic Acid Trisodium Salt (1a). Solid
NaH (30 mg, 1.24 mmol) was added in small portions into a solution
of phenol-BTO 6a (200 mg, 0.31 mmol) in dry DMF (5 mL). After
the generation of hydrogen gas subsided, the reaction mixture was
stirred at room temperature for 30 min, and then 1,4-butanesultone
(0.13 mL, 1.24 mmol) was introduced into the reaction through
the septum via a syringe. The resulting mixture was subsequently
stirred at room temperature for 36 h, diluted with CH2Cl2 (5 mL)
Supporting Information Available: Synthesis and character-
ization of compounds 5b, 6b, 1b, and 2; NMR titration data, and
1H/13C NMR spectra of compounds 1, 2, 5, and 6. This material is
available free of charge via the Internet at http://pubs.acs.org.
JO051630S
J. Org. Chem, Vol. 71, No. 1, 2006 45