Enantioselective binding of amino acids and amino alcohols by self-assembled chiral basket-shaped receptors

Article abstract of DOI:10.1016/j.tet.2003.11.019

Amino acid appended diphenylglycoluril-based chiral molecular receptors 2 and 3 have been prepared and their aggregation has been studied in water at various pH's and in chloroform. The binding of several biologically relevant guests with aromatic moietie

Full text of DOI:10.1016/j.tet.2003.11.019

Tetrahedron 60 (2004) 291–300
Enantioselective binding of amino acids and amino alcohols by
self-assembled chiral basket-shaped receptors
*,
†
Beatriu Escuder, Alan E. Rowan, Martinus C. Feiters and Roeland J. M. Nolte
*
Department of Organic Chemistry, Katholieke Universiteit Nijmegen, Toernooiveld 1, NL-6525 ED Nijmegen, The Netherlands
Received 22 October 2002; revised 25 July 2003; accepted 7 November 2003
Abstract—Amino acid appended diphenylglycoluril-based chiral molecular receptors 2 and 3 have been prepared and their aggregation has
been studied in water at various pH’s and in chloroform. The binding of several biologically relevant guests with aromatic moieties to these
aggregates has been studied with UV–Vis spectroscopy in competition experiments with 4-(4-nitrophenylazo)resorcinol (Magneson) and
2-(4-hydroxyphenylazo)benzoic acid (HABA) as probes. Aggregates of chiral host 2b showed binding of catecholamines and aromatic
amino acids in an aqueous environment, as well as discrimination between amino acid enantiomers, and can be considered a mimic for
adrenergic receptors.
q 2003 Elsevier Ltd. All rights reserved.
1
. Introduction
some cases transport, adrenaline, ephedrine, L-dopa and
dopamine in water although with moderately low binding
1
3–20
The design of synthetic molecular receptors that mimic the
natural binding sites for hormones, neurotransmitters, and
other essential messengers is a topic of intense research with
a special emphasis on their enantioselective recognition.
Although there is significant progress toward the under-
standing of the natural systems, synthesis of model
compounds can contribute to this understanding and in
constants.
pyrazole-containing cryptand that binds dopamine in water
More recently, a copper complex of a
2
1,22
with a high binding constant has been reported.
Diphenylglycoluril based clip molecules have been pre-
pared and extensively studied in the Nolte group over the
past 10 years. These hosts possess a well defined cavity that
allows the binding of different phenolic guests via a
combination of several non covalent interactions (H-bond-
1
,2
addition find applications in drug delivery, catalysis, etc.
L-Adrenaline has some important biological functions. It
belongs to the family of adrenal medulla hormones that have
a large influence on the storage and mobilisation of
glycogen and fatty acids and the corresponding metabolic
pathways. In addition, it is a neurotransmitter of the
adrenergic nervous system and has an effect on a and b
receptors. The biosynthetic precursors of adrenaline, the
catecholamines, have also very interesting biological
properties and are of great therapeutical value. In the recent
years a great effort has been made in the X-ray crystal-
lographic characterization and modeling of membrane-
bound proteins as well as the design of synthetic model
receptors for their binding sites.
phosphonate containing cyclophanes that bind catechol-
amines and amino acids in organic solvents such as DMSO
or methanol and in water. Other authors have reported
several crown ether containing receptors that bind, and in
2
3
ing, p–p stacking, and the cavity effect). The introduction
of crown ether chains and alkyl tails to these receptors leads
to a new generation of the so-called amphiphilic basket
receptors, which are able to bind also alkaline metal ions
and ammonium salts, and aggregate in water into well
2
4,25
defined nanostructures.
Here we describe a new series
of amino acid appended diphenylglycoluril based receptors
bearing L-lysine and L-2,3-diamino propionic acid residues
that bind aromatic amino acids and catecholamines in water
at different pH values and, in some cases, recognize them
2
6
enantioselectively. As these receptor molecules have
pronounced polar and apolar sides, they are amphiphilic
and can be expected to aggregate in both aqueous and
3
–5
Schrader described
2
7
organic media.
6
–12
2
. Results and discussion
Keywords: Supramolecular chemistry; Receptors; Enantioselectivity;
Amino acids.
2.1. Synthesis
*
Corresponding authors. Tel.: þ34-964-728235; fax: þ34-964-728214;
e-mail address: escuder@qio.uji.es
Present address: Dep. Qu ´ı mica Inorg a` nica i Org a` nica, Universitat Jaume
I, E-12080 Castell o´ , Spain.
†
The synthesis of molecular receptors 2a,b and 3a,b was
carried out as described in Scheme 1. Compounds 2a,b were
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.11.019

292
B. Escuder et al. / Tetrahedron 60 (2004) 291–300
Figure 1. Plot of the conductivity vs concentration of solutions of
compounds 2a,b (A) and (B) 3a,b in water at 20 8C. (The critical
aggregation concentration for each compound is pointed by an arrow. All
the experiments were carried out in duplo).
tration (CAC) of these compounds at 20 8C was determined
by measuring the conductivity at different concentrations.
As can be seen in Figure 1, a change in the slope of the plot
of the conductivity against the concentration is clearly
2
5
26
produced at 3.5£10
and 1.5£10 M for 2a and 2b
2
6
26
respectively, and at 3£10 and 4£10 M for 3a and 3b.
The observed differences are the result of an increase in the
mobility of the ions as well as in the number of independent
charge carriers upon going from an aggregated towards a
non-aggregated state. The difference of one order of
magnitude observed between the CAC of compounds 2a
and 2b is significant enough to be noteworthy. Both
compounds exist as zwitterionic species in water but in
the case of compound 2a a six-membered cyclic intra-
molecular ionic pair could be formed between the
carboxylate group and the protonated tertiary nitrogen
atom of the aza-crown moiety.
Transmission electron microscopy reveals that receptor 2b
forms vesicles in water with diameters ranging between 50
and 100 nm (Fig. 2B). The vesicular structure of the
aggregates was confirmed by encapsulation of ethydium
bromide. A dispersion of compound 2b was prepared in the
presence of 102 M ethidium bromide and subsequently
filtered through a Sephadex G25 column monitoring the
absorbance of 2b and the fluorescence of the entrapped dye.
Fractions containing both compounds were found at elution
volumes of ca. 25–100 mL, whereas the free dye was
4
Scheme 1.
prepared in a good yield from the tetrachloride 1 and the
a
correspondingN -BocprotectedaminoacidusingFinkelstein
conditions. The carboxylic acid functions were coupled with
two equivalents of hexadecyl amine to give the N-protected
precursors of amphiphiles 3a,b. The t-butoxy-carbonyl
protecting group was removed with 4 M HCl in ethyl acetate
to give the corresponding hydrochloride as a white precipitate.
After slightly basic work-up the free amines were fully
characterised by NMR, MS and elemental analysis.
2
.2. Aggregation studies
Figure 2. (A) Transmission electron micrograph of a 1% wt dispersion of
compound 2a in water. (B) idem of a dispersion of compound 2b in water
(Pt-shadowing, bars represent 200 nm).
The aggregation behaviour of compounds 2a,b and 3a,b
2
was studied in water. The critical aggregation concen-
8

B. Escuder et al. / Tetrahedron 60 (2004) 291–300
293
Figure 4. (A) Transmission electron micrographs of a 1% wt dispersion of
compound 3b in water 1 h after sonication. (B) idem after one day. (Pt-
shadowing, bars represent 400 nm). (C) Schematic drawing of the
aggregates onto the grid.
tubes, flat tapes, see Figure 4A). Upon standing for 1 day,
they reorganized to give more uniform lamellar structures.
As can be seen in Figure 4B, compound 3b self-assembles to
form flat disks that can fold or roll up to various degrees to
give semicircular flat objects and extended structures (Fig.
4C). The same grid was investigated by scanning electron
microscopy and it was revealed that flat disk-like objects of
ca. 550 nm diameter and 50 nm thickness occur at various
angles, along with folded semicircular structures and
extended structures which we propose to arise from
2
9
completely rolled up disks. The same aggregates were
also observed when the experiment was carried out in 0.1 M
HCl (Fig. 5A and B). It is logical to assume that all the
amino groups would be protonated in water as well as in the
acidic solution. When the assemblies were made in a 0.1 M
NaOH aqueous solution, different aggregates were found
(
2
Fig. 5C). Flat tapes with lengths between 400 nm and
.5 mm and widths between 200 and 400 nm were observed
together with tubes of 50 nm diameter and similar lengths.
Figure 3. (A) Transmission electron micrographs of a 1% wt dispersion of
compound 3a in water showing tubular architectures (Pt-shadowing,). (B)
and (C) Cryo-scanning electron micrographs of a 1% wt dispersion of
compound 3a in water showing hollow tubes. The inset shows the side view
of the tubes.
As amphiphiles constructed from lysine and its analogues
contain many moieties that can be involved in intermole-
cular hydrogen bonding, we considered it of interest to also
study the aggregation of compounds 2a,b and 3a,b in a
solvent that would allow the formation of such hydrogen
bonds, like chloroform. The strong aggregation of these
retained at the top of the column. In contrast, compound 2a
at the same concentration formed similar but slightly less-
defined assemblies in water, only a mixture of round and flat
lamellae could be observed (Fig. 2A).
compounds was already evidenced by the broadening of the
1
H NMR signals in CDCl , in contrast to the sharp
3
resonances observed for solutions of the previously reported
N-functionalised hosts of this type, and was further
confirmed by electron microscopy. Compound 2b self-
assembled into well-defined thin tube-like structures
(diameter ca. 5 nm) in chloroform as can be seen in Figure
6A inset. After few hours, these tubular structures further
aggregate to give a flat array of aligned and superimposed
layers of tubes (Fig. 6A).
The effect of pH was studied for compound 2b by
transmission electron microscopy (not shown). The dis-
persion was prepared in 1 mL of 20 mM sodium mono-
phosphate (pH¼4.5) and the electron micrographs revealed
the presence of the vesicles and their preference for further
assembly. The vesicles clustered forming elongated aggre-
gates of various micrometers of length and less than 100 nm
width. In contrast to 2b, compound 3a aggregates to form
thin lamellae in water that roll up to form large tubular
objects of several micrometers of length (Fig. 3A). Cryo-
SEM micrographs clearly revealed that these tubular objects
are in fact hollow tubes of ca. 1 mm diameter and more than
In contrast to the previously reported receptors these
compounds possess amino acid arms which are the potential
sites for additional H-bonding or electrostatic interactions.
This feature is thought to be responsible for the observed
aggregation behaviour since intermolecular hydrogen bonds
can now also be formed between the amide functions
leading to extended structures. The aggregation behaviour
of molecular receptors 3a and 3b, possessing two alkyl tails
1
0 mm in length (Fig. 3B).
Transmission electron microscopy of host 3b in water
revealed a mixture of different aggregates (vesicle-like,

294
B. Escuder et al. / Tetrahedron 60 (2004) 291–300
Figure 6. Transmission electron micrographs of 1% wt. CHCl
A) compound 2b, (B) compound 3b and (C) compound 3a. (Pt-shadowing,
bars represent 100 nm).
3
solutions of
(
Figure 5. (A) SEM picture of a dispersion of compound 3b in 0.1 N HCl.
B) Transmission electron micrographs of a dispersion of compound 3b in
.1 N HCl (Pt-shadowing). (C) idem in 0.1 N NaOH.
(
0
2a,b possess the chiral functionalities required to perform
such enantioselectivity. The hosts present a pocket-like
geometry and also their aggregation behaviour suggests that
they could be incorporated easily into a membrane as a
carrier or membrane-bound receptor mimic.
in its structure, was also studied in chloroform by
transmission electron microscopy revealing curved bilayer
aggregates (Fig. 6B and C).
2
.3. Binding studies
UV–Vis spectroscopy was used for the determination of the
binding constants. The absorption signals of both the host
and the guest unfortunately overlapped and here a
competition experiment was carried out using Magneson
(4, 4-(4-nitrophenylazo)resorcinol) or HABA (5, 2-(4-
hydroxyphenylazo)benzoic acid) as a competing dye
(Chart 1). The binding of Magneson by compound 2b was
studied in water at 20 8C by UV–Vis spectroscopy above its
The binding of adrenaline and other catecholamines and
amino acids by their natural receptors is thought to occur via
a combination of non-covalent interactions (viz. electro-
static interactions between aspartate or glutamate protein
residues and the ammonium groups of the substrates,
H-bonding between the substrate and serine or lysine
residues, p–p interactions between the substrate aromatic
moiety and the aromatic residues of the protein, and the
p–cation interaction between the ammonium group and the
electron-rich aromatic system of tyrosine and tryptophan
4
0
CAC. When a sample containing the dye (4) was titrated
with 2b an increase in the absorbance at 450 nm was
observed reaching almost complete saturation when ca.
2 equiv. of the host molecule were added. It was recently
shown that for similar host molecules, this behaviour is in
accordance with the formation of vesicles in which only half
5
residues). On the other hand, an important goal for the
design of synthetic receptors is the enantioselective
recognition of guests. Considerable work has been devoted
to the design and syntheses of chiral hosts for amino acids
2
5
of the sites are available for binding the guest. The
addition of excess host induced small changes in the UV
spectra probably because the vesicle bilayer is not a
3
0–39
and other biologically relevant guests.
Compounds

B. Escuder et al. / Tetrahedron 60 (2004) 291–300
295
Table 2. Estimated binding parameters for 2b and guests 6–12 determined
by UV–Vis competition with Magneson 4 and HABA 5 at 20 8C
3
21 a
)
21 a
)
21 b
DDG (kJ mol )
Probe Guest
K
ass (£10 M
DG (kJ mol
c
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
7
8
L-9
6
7
4(1)
0.9(0.5)
4.7(0.4)
10(3)
1.9(0.3)
1.7(0.4)
4.7(0.5)
2.2(0.5)
16(2)
12.8(0.5)
26(5)
12(1)
17(2)
56(7)
220.2(0.6)
216.5(1.3)
220.6(0.2)
222.4(0.7)
218.4(0.4)
218.1(0.6)
220.6(0.3)
218.7(0.6)
223.6(0.3)
223.0(0.1)
224.7(0.5)
222.9(0.2)
223.7(0.3)
226.6(0.3)
—
—
—
—
—
—
—
4.9
c
c
c
d
d
d
d
d
d
d
d
d
d
8
L-9
D-9
10
L-11
D-11
L-12
D-12
—
21.8
2.9
a
b
c
d
Errors given between parentheses.
D
Value for the difference DG 2DG .
Magneson (4), pH 8, 450 nm.
HABA (5), pH 4.5, 350 nm.
L
For guests 6–9 moderate values for the binding constant
were found (Table 2). Upon comparison of the relative guest
structures the differences in Kass can be attributed to either
the presence or absence of the hydroxyl functions in the
aromatic ring and the carboxylate group in the side chain.
The highest binding constant was observed for L-tyrosine
3
21
(
when either the phenolic group is not present as in the case
L-9), K ¼(10^3)£10 M . This value slightly decreases
ass
3
21
of ephedrine (8), K ¼(4.7^0.4)£10 M , or when an
ass
extra hydroxyl is introduced, as in L-dopa (L-6),
K ¼(4^1)£10 M21. The effect of repulsion of a negative
3
ass
charge in the guest by the negatively charged carboxylates
in the host is noticeable in the in Kass going from dopamine
Chart 1.
(
7), which lacks the carboxylate, to L-dopa (6), where it is
present. As is well known, one of the main forces
responsible for the binding of guests in hosts in water is
the hydrophobic interaction. The release of water molecules
from the cavity when the guest is bound and the decrease of
the apolar surface in contact with water in the complex both
favour binding. This will be a common factor for all the
guests that we studied. All of them possess a hydrophobic
aromatic moiety by which the molecule is pulled into the
sufficiently impermeable barrier and some molecules of
Magneson (4) can cross it and be bound by the inner binding
sites. The binding constant for Magneson (4) was calculated
by fitting the data assuming only half of the hosts bound in a
4
21
1
:1 host–guest ratio, K ¼(4.4^1.2)£10 M (Table 1).
ass
The binding of the different guests (Chart 1) with compound
b was first studied above its CAC at pH¼8 by UV–Vis
competition by adding different amounts of the guest to a
2
41
solution containing the host 2b and Magneson (4). The
Magneson absorption band at 450 nm decreases with the
addition of the guest and the presence of an isosbestic point
at ca. 290 nm agrees with the presence of a single
complexation equilibrium which can be fitted to a 1:1
complex stoichiometry.
Table 1. Binding parameters of compound 2b with the probes 4 and 5 at
20 8C
2
1 a,b
)
b,c
21 b
DG (kJ mol )
Probe
Solvent
K
ass (M
D1
d
4
4
4
5
pH 8 buffer
pH 4.5 buffer
(4.4^1.2)£10
17,150
3724
226.0(0.7)
224.7(0.2)
d
(2.5^0.2)£10
a
All the experiments were carried out in duplo.
Errors are given between parentheses.
Difference between the extinction coefficients of guest and complex in
absolute value.
b
c
Figure 7. UV/VIS binding competition curves for the 2b:2-(4-hydro-
xyphenylazo)benzoic acid (5) complex with L-dopa (6) at pH¼4.5.
d
0
.02 M phosphate buffer solutions.

296
B. Escuder et al. / Tetrahedron 60 (2004) 291–300
case the binding of guests 6–12 with host 2b was studied
Table 2). The absorption spectra for the competition
(
experiments with HABA (5) and L-dopa (6) in host 2b are
shown in Figure 7. In this case, the absorption band of
HABA at ca. 350 nm increases upon the addition of the
guest.
The binding constants for host 2b with L-dopa (L-6),
dopamine (7) and ephedrine (8) measured at pH 4.5 were
similar to those observed at higher pH. In contrast to binding
at pH 8, the binding of L-tyrosine (L-9) is reduced at pH 4.5
Figure 8. Calculated minimum energy geometry for the 1:1 complex of 2b
and adrenaline (10), hydrogen atoms have been omitted for clarity.
(
Table 2). The binding of adrenaline (10) as a guest was also
21
3
studied giving a binding constant of (12.8^0.5)£10 M
,
cavity while the differences in binding between them are
caused by differences in solvation and the differing
complementarity between the guests and the host. Thus,
an increase in the number of hydroxyl functions will, in
principle, increase the solvation of the guest in water and
then decrease slightly the binding constant as is obvious
from a comparison of the binding constants of dopamine (7)
with ephedrine (8). The later which lacks OH groups would
be more poorly solvated by water and then bound deeper in
the cavity. Although it is a more polar group, the presence of
a carboxylate also exerts a favourable effect, either by
orientating the guest inside the cavity or by electrostatic
charge–charge or dipole–charge interactions.
one order of magnitude larger than those obtained for the
former guests, and considerably higher than the values
1
0,16
found in literature in aqueous solution.
In this case, the
presence of the methyl group increases the hydrophobicity
of the guest resulting in a strong binding spite of the
presence of OH groups in the molecule. In order to prove the
role played by the carboxylate groups in the binding of this
kind of guests both enantiomers of the amino acids
phenylalanine (L-11, D-11) and tryptophan (L-12, D-12) as
well as the D-tyrosine (D-9) were also studied. In all cases,
the binding constants were one order of magnitude higher
than those obtained for dopamine (7) and ephedrine (8). It is
clear that electrostatic or charge–dipole interactions play an
important role in the binding process.
The important role that p–cation interactions play in
2,43
4
biological binding sites is well documented.
The
2.4. Enantioselective binding
protonated nitrogen atoms on the aza-crown moiety could
play a role in the binding process. In order to ensure the full
protonation of the tertiary amino functions of the azacrown
moieties the binding studies were also carried out at slightly
acidic conditions in a 0.02 M NaH PO buffer solution of
Enantioselective binding was observed when the pair of
enantiomers of tyrosine (L-9, D-9), phenylalanine (L-11,
D-11) and tryptophan (L-12, D-12) were studied (see Table
2). In the first case, a DDG (DG
L
2DG ) of 4.9 kJ/mol was
D
2
4
4
4
pH 4.5. For these studies, HABA (2-(4-hydroxyphenyl-
azo)benzoic acid, 5), with a binding constant of
found in favour of the D-enantiomer (D-9). In a racemic
guest solution almost 90% of the binding sites would be
occupied by the D-enantiomer. For the pair of enantiomers
of phenylalanine (L-11, D-11) there is 1.8 kJ/mol more loss
of free energy for the binding of the L-enantiomer (L-11) and
in the case of tryptophan (L-12, D-12) the difference was of
4
21
(
2.5^0.2)£10 M
for the 1:1 complex with 2b, was
used as a UV–Vis probe instead of Magneson (4), because
the later had inconvenient additional absorption bands at
low pH. Following a similar procedure as for the former
Figure 9. Calculated minimum energy complexation geometries for the 1:1 complexes of 2b and (A) D-tyrosine (D-9), (B) L-phenylalanine (L-11) and (C)
D-tryptophan (D-12), hydrogen atoms have been omitted for clarity.

B. Escuder et al. / Tetrahedron 60 (2004) 291–300
297
2
.9 kJ/mol in favour of the D-enantiomer (D-12). Although
the cavity and seems to be clipped-on by a H-bond between
the indole NH group and the carbonylic oxygen from the uril
moiety whereas L-tryptophan (L-12) is more flexible. The
distance between the uril carbonylic oxygen and the indole
NH of tryptophan as well as the phenolic OH of the tyrosine
was monitored during the Molecular Dynamics simulation.
As can be seen in Figure 10C, the 80% of the population of
conformers showed an average distance in the range of 1.5–
the enantioselectivities observed are moderate they are
significant. The binding geometries for the three pairs of
enantiomers at first glance seem to be quite similar and
independent of the different aromatic moieties they present.
The reverse enantioselectivity of the complex with 11 as
compared to 9 and 12, which contain a phenolic hydroxy
group and an indole nitrogen, respectively, must be due to
H-bonding interactions inside the cavity, as confirmed by
our modelling studies (see Section 2.5).
˚
3.5 A in the case of D-12 and no conformation in a distance
˚
above 7 A was found, whereas for the diastereomeric
complex L-12 the 70% of the population was found within
4
5
˚
this average distance, 25% between 3.6 and 7 A and 5%
2
.5. Modelling studies
˚
above 7 A. The majority of the conformers show for both
˚
enantiomers distances below 4 A and agrees with the guest
Computational studies have been widely used in the last
years for the modelling of chiral recognition systems such as
cyclodextrins, proteins and synthetic receptors.
being confined into the cavity, the presence of confor-
mations with larger distances in the L-enantiomer being
related to the larger looseness of the complex. In the case of
4
6–50
Here
we used a Monte Carlo/Molecular dynamics mixed
approach for an approximation of the structure of the
host–guest complexes coherent with the experimental
results (see Section 4 for details). In most cases, the
minimum energy conformers found showed the guest placed
inside the cavity. For the same host–guest complex, the
main difference between the calculated energy of the
collected structures was in the Van der Waals and solvation
terms, in agreement with the hydrophobic effect being the
driving force of the inclusion process. In fact, the minimum
energy conformations always showed the smallest molecu-
lar surface area exposed to the solvent. On the other hand,
the largest favourable energetic term seems to correspond to
the electrostatic interactions. This could explain the higher
binding observed when the guest contained a carboxylate
group in its structure as in tyrosine (9), phenylalanine (11),
and tryptophan (12) compared with dopamine (7) and
ephedrine (8).
In general, the lowest energy conformations of the
complexes between 2b and the different guests always
showed the guest bound inside the cavity. As an example,
Figure 8 shows the calculated structure for the complex
between adrenaline (10) and 2b. It can be seen that the
binding again involves electrostatic interactions, this time
between the ammonium group of the guest and one of the
carboxylates of the host, in addition to the extra hydro-
phobicity provided by the methyl group, as mentioned
earlier (cf. previous section) as well as H-bonding and
aromatic hydrophobic interactions. The complex structures
found for the guests 9, 11 and 12 are also in agreement with
the enantioselectivity observed in the experimental binding
studies (Fig. 9). For the complex between 2b and D-tyrosine
(
D-9) all the minimum energy conformers have the guest
placed inside the cavity. The main difference between the
different minima is the possibility of H-bonding between the
phenolic OH of the guest and the CvO of the uril moieties.
In the case of the binding of the enantiomer L-tyrosine (L-9)
the conformations found for its complex with 2b showed a
more flexible geometry. Thus, the guest L-9 appeared in
different dispositions inside the cavity and bound by the
lysine arms in a smaller energy range of less than 20 kJ/mol.
Nevertheless, a clear energetic preference was calculated
for those complexes with the guest located inside the cavity,
and no structures were found with the guest outside the
cavity. Even more clear results were calculated in the case
of tryptophan (12) guests. The D-enantiomer (D-12) fits into
Figure 10. Distribution of complexation geometries (133,333 structures)
during 2000 ps of molecular dynamics simulation vs the selected host–
guest distances for 2b and (A) tyrosine (9), (B) phenylalanine (11) and (C)
tryptophan (12).

298
B. Escuder et al. / Tetrahedron 60 (2004) 291–300
tyrosine, there is a clear difference between the observed
behaviour of the two complexes (Fig. 10A). Whereas the
complex with D-9 shows a major population of conformers
(m, 4H); 6.7 (m, 4H); 7.0 (s, 10H). ESI-MS m/z¼626.5
þ
64 82 8 18
2
[Mþ2H] . Anal. Calcd for C H N O : C. 61.43; H.
6.60; N. 8.95. Found: C. 61.01; H. 6.82; N. 8.64. 2b (62%).
Mp 128 8C. [a] ¼þ5.858 (c¼0.65, CHCl ). H NMR
˚
20
D
1
between 1.6 and 3.5 A the diastereomeric complex with L-9
3
˚
presents two clearly different populations, the first at 2–3 A
(CDCl ): 1.2–2.0 (mþs, broad, 32H); 2.3–3.0 (m, 12H);
3
˚
and the second at 5–7 A. In this case the guest D-9 is much
3.5–4.6 (m, 30H); 5.5 (m, 4H); 6.71 (m, 4H); 7.10 (s, broad,
10H). FAB-MS m/z¼1335.8 [MþH] . Anal. Calcd for
þ
C H N O : C. 62.95; H. 7.09; N. 8.39. Found: C. 62.71;
H. 6.78; N. 8.17.
better accommodated into the cavity than the guest L-9. In
the case of phenylalanine (11), the conformational popu-
lation also points to a more defined complex with the
experimentally preferred enantiomer, L-11 (Fig. 10B).
7
0 94 8 18
4.1.2. Amphiphilic receptor 3 (a,b). Compound 2(a,b)
(0.34 mmol) and N-hydroxysuccinimide (0.69 mmol) were
3
. Conclusions
dissolved in dry dimethoxy ethane and cooled at 0 8C. Then
DCC was added and the mixture was kept in the refrigerator
for 24 h. Afterwards, the white precipitate was filtered off
and to the resulting solution hexadecylamine (0.69 mmol)
was added at 0 8C. The reaction was stirred overnight at
room temperature and the white solid formed was filtered-
off and washed with a small amount of diethyl ether.
Finally, the solid was suspended in a 4 M HCl solution in
ethyl acetate and after stirring at room temperature during
2 h compound 3(a,b) precipitated as the hydrochloride.
Further slightly basic work-up lead to the free amines. 3a:
We have presented a new series of diphenylglycoluril-based
amino acid-tethered cavity containing receptors with
remarkably versatile aggregation behaviour in both water
and chloroform solutions. In particular, compound 2b forms
thin tubules in chloroform, whereas it forms vesicles in
water. The long tail derivatives 3a,b form large aggregates
that can also be tuned with the pH of the medium.
Compounds 2a,b present a variety of interaction sites that
allow the strong binding of biologically relevant molecules
such as amino acids and catecholamines via a combination
of several non-covalent interactions. In general, the binding
constants calculated for 2b with these guests are moderate to
high and a remarkable enantioselectivity for D-tyrosine
2
0
1
(63%). Mp 235 8C dec. [a]
NMR (CDCl
): 0.87 (t, 6H); 1.2–1.4 (sþm, 56H); 2.5–3.3
(m, 6H); 3.5–4.3 (m, 40H); 5.6 (m, 4H); 6.7 (m, 4H); 7.1 (s,
¼26.118 (c¼0.4, CH
OH). H
D
3
3
2
2
10H). ESI-MS m/z¼896.5 [M22(C16
H
33NHCOCH )þ
þ
9.35. Found: C. 69.71; H. 9.08; N. 9.11. 3b:(58%). Mp
(
D-9), L-phenylalanine (L-11) and D-tryptophan (D-12) with
Na] . Anal. Calcd for C86
H
132
N
10
O
12: C. 69.84; H. 8.88; N.
respect to their antipodes is shown at pH 4.5. The
aggregation and binding features described here will be
exploited in the future with the incorporation of these
receptors into membranes for their study as cell-surface
adrenergic receptor mimics, as well as their potential use in
drug delivery or catalysis.
2
0
1
250 8C dec. [a]
¼þ3.048 (c¼0.23, CH
OH). H NMR
3
(CDCl ): 0.87 (t, 6H); 1.24 (s broad, 52H); 1.5–2.2 (m
3
D
broad, 16H); 3.2–4.2 (m broad, 42H); 5.6 (m, 4H); 6.7 (m,
þ
4H); 7.1 (s, 10H). FAB-MS m/z¼1583.2 [MþH] . Anal.
Calcd for C92H N O12: C. 69.84; H. 9.17; N. 8.85.
144 10
Found: C. 69.51; H. 9.42; N. 8.44.
4. Experimental
4.2. Electron microscopy
4
.1. General remarks
Aggregates preparation. The compounds were dissolved in
ca. 50 mL of methanol, injected into the aqueous solution at
room temperature to a final concentration of 0.1–1 mg/mL
and sonicated at 40 8C for 30 min. Different samples were
studied between two hours and few days after sonication.
Chloroform samples were made by dissolving 1 mg or less
of the compound in 1mL of the solvent.
1
NMR spectra were recorded on a Bruker AC-300 ( H NMR
1
3
spectrometers in CDCl with TMS as internal standard.
00 MHz) and Bruker FAMX500 ( H NMR 500 MHz)
3
UV–Vis spectra were recorded with a Varian Cary 50Conc
spectrophotometer. Compound 1 was prepared as reported
before and the N -Boc protected amino acids were
obtained from FLUKA. Magneson (4) and HABA (5)
were purchased from Aldrich.
5
1
a
Transmission electron microscopy. TEM was carried out
with a JEOL JEM.1010 electron microscope. The aqueous
samples were prepared by adding a drop of the solution over
a copper grid covered with a thin layer of formvar. After a
few seconds (depending on the concentration of the sample
and its affinity for the polymer surface) the sample was
drained and left to dry at room temperature overnight. When
chloroform was used as a solvent, the samples were
prepared in a similar way but over a hydrophobic carbon
coated copper grid. Before observation, the samples were
shadowed with Pt at 458.
4
1
0
.1.1. Molecular receptor 2(a,b). A mixture of compound
(4.05 mmol), NaI (40 g, 0.27 mol) and Na CO (13.3 g,
.13 mol) in 500 mL of acetonitrile was refluxed under
a
2
3
nitrogen for 4 h. The N -Boc-protected amino acid
12.2 mmol) was then added in small portions over a period
(
of 2 days and the mixture was refluxed for one week. After
filtration and evaporation of the solvent the crude white
solid was suspended in CHCl and washed with 10% citric
3
acid and water. The organic layer was dried (Na SO ) and
2
4
concentrated under vacuum. After column chromatography
neutral alumina, eluent 0.1–0.5% MeOH/CHCl , v/v) pure
Scanning electron microscopy. The samples where studied
with a Field Emission Scanning Electron Microscope Jeol
JSM-6330F. They were prepared in a similar way as before
on a copper grid covered with a thin layer of formvar and
then they were sputtered with 1.5 nm of Au/Pd.
(
3
2
(a,b) was obtained as a solid. 2a (87%). Mp 180 8C dec.
1
2
0
[
a] ¼þ8.728 (c¼1, CHCl ). H NMR (CDCl /CD OD, 8/
D
3
3
3
2
): 1.3 (m, 18H); 2.5–2.8 (m, 6H); 3.5–4.1 (m, 36H); 5.4

B. Escuder et al. / Tetrahedron 60 (2004) 291–300
299
Cryo-scanning electron microscopy. The samples where
studied with a Field Emission Scanning Electron Micro-
scope Jeol JSM-6330F. A drop of the sample solution was
placed in a stub and it was quickly cooled down at 2220 8C
with under-cooled nitrogen as slush. The sample was then
introduced into the microscope cooling pre-chamber and it
was allowed to warm up until 295 8C. At this temperature
the upper part of the drop was fractured with a cool knife
and etched for 2 min. Then, the pre-chamber was cooled
down until 2120 8C and the sample was sputtered in situ
with 1.5 nm of Au/Pd. Finally, it was transferred into the
microscope chamber were the temperature was kept below
of complexation, using the following equations in an Excel
spreadsheet
A ¼ 1 ½DŠ þ 1 ½HDŠ ¼ 1 D þ ð1 2 1 Þ½HDŠ
ð2Þ
ð3Þ
ð4Þ
ð5Þ
i
D
i
DH
i
D
t
DH
D
i
A 2 1 D
t
i
D
½
HDŠ ¼
i
1DH 2 1D
½DŠi ¼ Dt 2 ½HDŠi
½
HDŠ
i
½HŠ ¼
i
K ½DŠ
D
i
i
½HGŠ
H 2 ½HŠ 2 ½HDŠ
i
t
i
i
K ¼
¼
ð6Þ
2
130 8C to avoid the formation of ice crystals.
G
½
HŠ ½GŠ
½HŠ ðG 2 ðH 2 ½HŠ 2 ½HDŠ ÞÞ
i
i
t
t
i
i
4
.3. Conductivity measurements
where A is the absorbance at the studied wavelength for the
i
experiment i, [X] the concentration of the host (H), guest
i
Conductivity measurements were performed in duplo with a
Schott Ger a¨ te CG 852 Conductimeter with a platinum
electrode at room temperature. Stock solutions of the
receptors were diluted several times until very low
concentrations were achieved while continuously monitor-
ing the conductivity.
(G), dye (D), host–probe complex (HD) and host–guest
complex (HG) for the experiment i, X
tration of host (H), guest (G) and dye (D) in the experiment
i, K the known binding constant of the probe (D), and K is
the calculated binding constant for the guest G.
the total concen-
t
D
G
The binding constant values estimated were finally averaged
and the standard deviations calculated. The extinction
coefficients were calculated in separate experiments and
blank experiments were carried out to check possible
interferences of the guest on the absorption band of the dye
in the absence of host.
4
.4. Encapsulation of ethydium bromide
The encapsulation of ethydium bromide was measured by
gel permeation chromatography (GPC) in combination with
fluorescence and UV–Vis spectroscopy. The desired
vesicular dispersion was prepared in water containing
1
G25 column with water as eluent. The fluorescence intensity
2
4
0
M ethydium bromide and passed over a Sephadex
4.6. Modelling studies
of ethydium bromide (l ¼480 nm, l ¼630 nm) as well
Molecular mechanics docking calculations were performed
for the complex structures using the Monte Carlo confor-
mational search method implemented in the Macromodel
ex
em
as the absorbance of the host molecule at 288 nm was
monitored.
5
4
V7.0 program. The AMBER* force field was used in a
5
5
4
.5. UV binding studies
water continuous solvent simulation (GB/SA). Energy
minima were found and the conformational space close to
them was explored by performing molecular dynamics
simulations. Starting structures were drawn with different
disposition of the guest inside and outside the cavity of the
host. In each case, the conformational search was performed
with 3000 iterations for each step and structures were
collected in a 50 kJ/mol energy window. Then, molecular
dynamics simulations were performed on the energy
minima obtained with a total simulation time of 2000 ps.
All the experiments were carried out using double distilled
water or freshly distilled chloroform to prepare the
solutions. In a typical experiment a conveniently buffered
2
5
solution of the dye (ca. 2£10 M) was titrated by adding
small amounts of a solution containing the host and the dye
(
2
5
ca. 2£10 M). The change in the absorbance (450 nm for
Magneson (4) and 350 nm for HABA (5)) was plotted
against the total host concentration. The data were fitted
using Eq. 1 with an Excel spreadsheet.
Acknowledgements
½
HGŠ ¼ 0:5ððH þ G þ 1=K Þ
i
t
t
ass
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The authors thank H. P. M. Geurts for technical support in
the electron microscopy studies. B.E. also thanks the
European Community for a postdoctoral Marie Curie
TMR grant.
2
ðH þ G þ 1=K Þ 2 ð4H G ÞÞ
ð1Þ
t
t
ass
t
t
The fitting was compared with the double-reciprocal plot
2,53
5
graphical method,
giving a linear plot for the data,
corresponding to the first 0.5 equiv. of guest in the case of
Magneson, in agreement with the values estimated before
with non-linear fitting of expression 1.
References and notes
In a typical competition experiment, an appropriate buffer
2
1. Supramolecular Reactivity and Transport: Bioorganic Sys-
tems. Comprehensive Supramolecular Chemistry; Murakami,
Y., Ed.; Elsevier: Amsterdam, 1996; Vol. 3. Lehn, J.-M., ed.;
Atwood, J. L.; Davies, J. E. D.; Macnicol, D. D.; V o¨ gtle, F.
executive eds.
5
solution containing the host and the probe in ca. 3£10 and
2
5
1
.6£10 M concentration, respectively, was titrated with a
2
5
guest solution which was also 1.6£10 M in the probe. The
binding constant for each guest was calculated at several
total guest concentrations, always in the range of 20 to 80%
2. (a) Kikuchi, J. I.; Murakami, Y. J. Incl. Phen. Mol. Rec. Chem.

300
B. Escuder et al. / Tetrahedron 60 (2004) 291–300
1
998, 32, 209–221. (b) Kikuchi, J. I.; Ariga, K.; Ikeda, K.
Chem. Commun. 1999, 547–548.
. Stryer, L. Biochemistry. 3rd ed. W. H. Freeman: New York,
988.
28. The pH values measured upon dispersion in unbuffered water
are as follows: 2a, pH¼6.5; 2b, pH¼7.9; 3a, pH¼7.2; 3b,
pH¼7.9.
3
4
5
6
1
29. This observation can be due to a sample preparation artefact
because of a vacuum effect.
. Silverman, R. B. The Organic Chemistry of Drug Design and
Drug Action; Academic: New York, 1992.
30. Lipkowitz, K. B.; Raghothama, S.; Yang, J. Am. Chem. Soc.
. Trumpp-Kallmeyer, S.; Hoflack, J.; Bruinvels, A.; Hilbert, M.
J. Med. Chem. 1992, 35, 3448–3462.
. Schrader, T. Angew. Chem. Int. Ed. Engl. 1996, 35,
1992, 114, 1554–1562.
1. Konishi, K.; Yahara, K.; Toshishige, H.; Aida, T.; Inoue, S.
3
3
3
3
3
3
3
3
3
4
J. Am. Chem. Soc. 1994, 116, 1337–1344.
2649–2651.
2. Asakawa, M.; Brown, C. L.; Pasini, D.; Stoddart, J. F.; Wyatt,
P. G. J. Org. Chem. 1996, 61, 7234–7235.
3. Chen, H.; Ogo, S.; Fish, R. H. J. Am. Chem. Soc. 1996, 118,
7
8
. Schrader, T. J. Org. Chem. 1998, 63, 264–272.
. Bioorganic Chemistry. Highlights and New Aspects; Dieder-
ichsen, U., Lindhorst, T. K., Westermann, B., Wessjohann,
L. A., Eds.; Wiley-VCH: Weinheim, 1999; p 215.
4
993–5001.
4. Zhang, X. X.; Bradshaw, J. S.; Izatt, R. M. Chem. Rev. 1997,
7, 3313–3361.
9
. Herm, M.; Schrader, T. Chem. Eur. J. 2000, 6, 47–53.
9
1
1
1
1
1
1
1
1
1
1
0. Herm, M.; Molt, O.; Schrader, T. Angew. Chem. Int. Ed. 2001,
0, 3148–3151.
5. Gavin, J. A.; Garc ´ı a, M. E.; Benesi, A. J.; Mallouk, T. E.
J. Org. Chem. 1998, 63, 7663–7669.
6. Liu, Y.; You, C. C.; Wada, T.; Inoue, Y. J. Org. Chem. 1999,
4
1. Grawe, T.; Schrader, T.; Finocchiaro, P.; Consiglio, G.; Failla,
S. Org. Lett. 2001, 3, 1597–1600.
6
4, 3630–3634.
7. Rekharsky, M.; Inoue, Y. J. Am. Chem. Soc. 2000, 122,
418–4435.
2. Rensing, S.; Arendt, M.; Springer, A.; Grawe, T.; Schrader, T.
J. Org. Chem. 2001, 66, 5314–5821.
3. Behr, J. P.; Lehn, J. M.; Vierling, P. Helv. Chim. Acta 1982,
4
8. Ram ´ı rez, J.; Ahn, S.; Grigorean, G.; Lebrilla, C. B. J. Am.
Chem. Soc. 2000, 122, 6884–6890.
9. Stevenson, C. D.; Cashion, D. K. J. Org. Chem. 2000, 65,
65, 1853–1867.
4. Kimura, E.; Fujioka, H.; Kodama, M. J. Chem. Soc., Chem.
Commun. 1986, 1158–1159.
7
588–7594.
5. Hayakawa, K.; Kido, K.; Kanematsu, K. J. Chem. Soc., Perkin
Trans. 1 1988, 511–519.
0. Titration of compound 2a was not successful above its CAC
because the concentrations required were too large for the
UV–Vis experimental conditions.
6. Bernardo, A. R.; Stoddart, J. F.; Kaifer, A. E. J. Am. Chem.
Soc. 1992, 114, 10624–10631.
7. Ishizu, T.; Hirayama, J.; Noguchi, S. Chem. Pharm. Bull.
4
1. Scatchard plots showed no aggregation induced cooperativity
upon binding of compound 2b by the dyes showing a straight
line for host–guest ratios till 1:1. Larger excess of host leads to
a non linear plot probably due to a change in the stoichometry
of the complexes formed.
1994, 42, 1146–1148.
8. Paugam, M. F.; Valencia, L. S.; Boggess, B.; Smith, B. D.
J. Am. Chem. Soc. 1994, 116, 11203–11204.
9. Paugam, M. F.; Bien, J. T.; Smith, B. D.; Chrisstoffels, L. A. J.;
de Jong, F.; Reinhoudt, D. H. J. Am. Chem. Soc. 1996, 118,
4
2. Sussman, J. L.; Harel, M.; Frolow, F.; Oefner, C.; Goldman,
A.; Toker, L.; Silman, L. Science 1991, 253, 872–879.
3. Ma, J. C.; Dougherty, D. A. Chem. Rev. 1997, 97, 1303–1324.
4. In a similar way as described for amino acids compound 2b
should exists as bis-zwitterionic species at neutral or slightly
acidic pH.
9820–9825.
4
4
2
2
2
0. Campayo, L.; Bueno, J. M.; Navarro, P.; Ochoa, C.; Jim e´ nez-
Barbero, J.; P e` pe, G.; Samat, A. J. Org. Chem. 1997, 62,
2684–2693.
1. Lamarque, L.; Miranda, C.; Navarro, P.; Escart ´ı , F.; Garc ´ı a-
Espa n˜ a, E.; Latorre, J.; Ram ´ı rez, J. A. Chem. Commun. 2000,
45. Since we have not found any aggregation induced binding
cooperativity we assume for the modelling studies a single
host–guest complex.
1337–1338.
2. Lamarque, L.; Navarro, P.; Miranda, C.; Ar a´ n, V. J.; Ochoa,
C.; Escart ´ı , F.; Garc ´ı a-Espa n˜ a, E.; Latorre, J.; Luis, S. V.;
Miravet, J. F. J. Am. Chem. Soc. 2001, 123, 10560–10570.
3. Rowan, A. E.; Elemans, J. A. A. W.; Nolte, R. J. M. Acc.
Chem. Res. 1999, 32, 995–1006.
46. McDonald, D. Q.; Still, W. C. J. Am. Chem. Soc. 1996, 118,
2073–2077.
2
2
2
47. Lipkowitz, K. B.; Pearl, G.; Coner, B.; Peterson, M. A. J. Am.
Chem. Soc. 1997, 119, 600–610.
4. Schenning, A. P. H. J.; de Bruin, B.; Feiters, M. C.; Nolte, R. J.
M. Angew. Chem. Int. Ed. 1994, 33, 1662–1663.
48. Kolossv a´ ry, L. J. Am. Chem. Soc. 1997, 119, 10233–10234.
49. Lipkowitz, K. B.; Coner, R.; Peterson, M. A.; Morreale, A.;
Shakelford, J. J. Org. Chem. 1998, 63, 732–745.
50. Lipkowitz, K. B. Acc. Chem. Res. 2000, 33, 555–562.
51. Sijbesma, R. P.; Nolte, R. J. M. J. Org. Chem. 1991, 56,
5. Schenning, A. P. H. J.; Escuder, B.; van Nunen, J. L. M.; de
Bruin, B.; L o¨ wik, D. W. P. M.; Rowan, A. E.; Van der Gaast,
S. J.; Feiters, M. C.; Nolte, R. J. M. J. Org. Chem. 2001, 66,
1538–1547.
3
122–3124.
2
6. For earlier reports on lysine based amphiphiles see: Reichel,
F.; Roelofsen, A. M.; Geurts, H. P. M.; van der Gaast, S. J.;
Feiters, M. C.; Boons, G. J. J. Org. Chem. 2000, 65,
5
5
2. Connors, K. A. Binding Constants. Wiley: New York, 1987.
3. Schneider, H.-J.; Yatsimirsky, A. Principles and Methods in
Supramolecular Chemistry; Wiley: Chichester, 2000.
4. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C.
J. Comput. Chem. 1990, 11, 440.
3
357–3366. Zarif, L.; Gulik-Krzwycki, T.; Riess, J. G.;
Pucci, B.; Guedj, C.; Pavia, A. A. Colloids Surf. A 1994, 84,
07–112.
5
1
2
7. Part of this work has been published as a communication:
Escuder, B.; Rowan, B.; Feiters, A. E.; Nolte, M. C.
Tetrahedron Lett. 2001, 42, 2751–2753.
5
5. Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T.
J. Am. Chem. Soc. 1990, 61, 4439–4449.