Sequence-dependent binding of dipeptides by an artificial receptor in water

Article abstract of DOI:10.1002/chem.200600573

An artificial dipeptide receptor (1) was designed and observed to bind the deprotonated dipeptide AC-D-Ala-D-Ala-OH in buffered water with K = 33100M ^-1, whereas other dipeptides such as Ac-Gly-Gly-OH or Ac-D-Val-D-Val-OH were bound less effici

Full text of DOI:10.1002/chem.200600573

DOI: 10.1002/chem.200600573
Sequence-Dependent Binding of Dipeptides by an Artificial Receptor in
Water
Carsten Schmuck,* Daniel Rupprecht, and Wolfgang Wienand[a]
Abstract: An artificial dipeptide recep-
tor (1) was designed and observed to
bind the deprotonated dipeptide Ac-d-
Ala-d-Ala-OH in buffered water with
K = 33100mÀ1, whereas other dipepti-
des such as Ac-Gly-Gly-OH or Ac-d-
Val-d-Val-OH were bound less effi-
ciently, by factors of more than 10 (K
< 3000mÀ1). The efficient binding and
the pronounced sequence selectivity
are the result of a combination of
strong electrostatic contacts and size-
discriminating hydrophobic interac-
tions. To provide such a combination, a
guanidiniocarbonylpyrrole cation was
attached to a novel cyclotribenzylene-
substituted alanine derivative 5, to pro-
vide a hydrophobic bowl-shaped cavity
just large enough to bind a methyl
group but not any larger alkyl chains,
thus causing the receptor to prefer ala-
nine to valine. We describe the synthe-
sis of 1 and the evaluation of its com-
plexation properties in UV and fluores-
cence titration studies.
Keywords: artificial receptors
guanidinium cations molecular
recognition peptides
supramolecular chemistry
·
·
·
·
Introduction
did therefore prefer dipeptides over amino acids, the selec-
tivity between dipeptides of different sequences was only
modest, since complexation did not invoke any direct inter-
actions with the amino acid side chains.[6] The receptor had
a modest preference for large and bulky amino acids over
smaller ones—Val-Val, for example, was bound better than
Ala-Ala, which was in turn bound better than Gly-Gly—but
this trend in affinities most probably just reflected the over-
all hydrophobic characters of the corresponding dipeptides.
In general, hydrophobic interactions simply increase with
the areas of the interacting surfaces or residues.[7]
Artificial receptors that selectively bind to a given peptide
sequence are of considerable interest not only for study of
the underlying principles of such molecular recognition
events but also for the construction of biosensors or as tools
with which to probe cellular processes.[1,2] Only a very few
artificial receptors that allow the complexation of an oligo-
peptide in aqueous solvents (e.g., under physiological condi-
tions) exist.[3] Quite often, however, the substrate selectivi-
ties of these receptors cannot be predicted a priori, especial-
ly when they have been identified by the screening of com-
binatorial libraries,[4] so the design of a specific receptor for
a given peptide sequence still remains an open and challeng-
ing task. We recently designed a receptor that was capable
of binding dipeptides in aqueous solvents with association
constants up to Kass = 50000mÀ1, but this receptor was not
able to distinguish between different dipeptides, since the in-
teractions between receptor and substrate were limited to
the amide backbone of the dipeptide.[5] While the receptor
We now present here the synthesis and the binding prop-
erties of an artificial receptor 1 (Figure 1) specifically de-
signed to bind alanine-containing dipeptide carboxylates.
Such dipeptide sequences are of considerable biological
[a] Prof. Dr. C. Schmuck, Dipl.-Chem. D. Rupprecht, Dr. W. Wienand
Institut für Organische Chemie, Universität Würzburg
Am Hubland, 97074 Würzburg (Germany)
Fax : (+49)931-888-4626
E-mail: schmuck@chemie.uni-wuerzburg.de
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Figure 1. Design of an artificial receptor 1 with a preference for alanine-
containing dipeptide carboxylates.
9186
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 9186 – 9195

FULL PAPER
relevance: the dipeptide se-
quence d-Ala-d-Ala, for exam-
ple, represents the C-terminal
part of a bacterial peptidogly-
can and plays an important
role in cell wall maturation in
Gram-positive bacteria, whilst
also being the target of the
vancomycin group antibiot-
ics,[8] which bind to this dipep-
tide sequence through a com-
bination of hydrogen bonds
and electrostatic and hydro-
phobic interactions. Our newly
designed receptor 1 is capable
of selectively binding d-Ala-d-
Ala even in water, with an as-
Scheme 1. Synthesis of receptor 1.
sociation constant of Kass
=
33100mÀ1, in preference to other dipeptides such as d-Val-d-
Val (Kass < 3000mÀ1). The selectivity is most probably the
function of a size-discriminating cyclotribenzylene cavity
that provides a shallow hydrophobic pocket large enough
only for an alanine methyl group.[9]
lyzed Negishi coupling between racemic (P/M)-iodocyclotri-
benzylene 3[18] and the zinc derivative of d-iodoalanine
(4).[19] We modified Jacksonꢁs initial protocol, which gave
yields of about 40%, by increasing the reaction scale and by
using a prolonged reaction time, thanks to which we were
able to improve the yield significantly, to an excellent 75%.
Because of the planar chirality of the iodocyclotribenzylene
3, which was used as a racemic mixture, the amino acid de-
rivative 5 was obtained as a mixture of two diastereoisom-
ers, denoted P/d and M/d, respectively, with the configura-
tion of the stereogenic center of the amino acid described
by d/l nomenclature and the configuration of the cyclotri-
benzylene unit by P/M descriptors. These two diastereomers
differ only in the absolute curvature of the cyclotribenzylene
bowl, as the ring flip at room temperature is rather
slow,[15b,20] or looked at another way, the cyclotribenzylene
unit can be viewed as being attached to the amino acid
either through C3 or C4 in the two diastereomers (see
Figure 1). This produces only a very small difference in the
three-dimensional structures of the two diastereomers, and
the diastereomeric (P,M)/d mixture of 5 accordingly could
not be separated even by HPLC on a chiral column. We
therefore used the diastereomeric mixture (P,M)/d-5 in the
synthesis of receptor 1, since modeling studies also suggest-
ed that the binding affinity should not be affected signifi-
cantly by the absolute configuration of the cyclotribenzylene
unit (see below). The ester group in 5 was then transformed
into the amide by treatment with ammonia, and after depro-
tection with HCl the free amine 6 was coupled with the pyr-
rolecarboxylic acid (CBS) 7 in the presence of PyBOP in
DMF to give the N-Boc-protected receptor 8. Removal of
the Boc group with HCl provided the free cationic receptor
(P,M)/d-1 (chloride salt).
Results and Discussion
Design and synthesis of the receptor: The design of 1 was
based on some general features taken from the natural anti-
biotic Vancomycin.[8] In vancomycin a protonated secondary
amine serves as a cationic anchor point for the dipeptideꢁs
carboxylate, and several hydrogen bonds to the peptide
backbone, together with hydrophobic interactions, then fur-
ther strengthen the complex.[10] In addition, several aromatic
rings form a rather rigid aromatic cavity in which the dipep-
tide is embedded. This provides the selectivity for alanine,
as no larger alkyl chains fit into this bowl-shaped structure,
so the efficient binding features of Vancomycin are due to a
combination of electrostatic interactions (H-bonds and
charge interactions) and a size-selecting aromatic cavity. We
have now developed an artificial receptor 1 (Figure 1) by
following the same guidelines.[11] Receptor 1 consists of two
major building blocks: a carboxylate binding site attached to
a size-selecting aromatic cavity. As the binding site for the
dipeptideꢁs carboxylate we used a cationic guanidiniocarbo-
nylpyrrole moiety.[12] This modified guanidinium cation has
already proven useful in the context of peptide recognition
even in water.[13,14] To mimic at least a part of the aromatic
cavity formed by Vancomycinꢁs several aromatic rings we
chose a cyclotribenzylene unit. This moiety provides a hy-
drophobic bowl just large enough to accommodate a methyl
but not any other alkyl group,[15,16] thus hopefully providing
a selectivity for alanine.[17] Molecular mechanics calculations
(see below) confirmed that receptor 1 should be able to
bind d-Ala-d-Ala preferentially.
Additionally, a triethyleneglycol-substituted analogue 2
was also synthesized in the hope that the triethyleneglycol
chains would increase the solubility of this receptor in water
and hence facilitate the binding studies later on. Receptor 2
was synthesized as shown in Scheme 2. In the first step the
dibromide 9 was treated with triethyleneglycol (TEG) in the
The key intermediate for the synthesis of 1 (Scheme 1)
was the nonnatural amino acid 5, obtained from a Pd-cata-
Chem. Eur. J. 2006, 12, 9186 – 9195
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9187

C. Schmuck et al.
receptor 2, again as a mixture of the two (P,M)/d diaster-
eoisomers. Separation of the two diastereomers was at-
tempted by MPLC purification of 13, of the N-Boc-protect-
ed precursor Boc-2 (after TBAF cleavage of the TBS
ethers), and of the free receptor 2, but no conditions to sep-
arate the two isomers could be found either by silica-MPLC
or by RP18-HPLC, and all subsequent studies were conse-
quently carried out with the (P,M)/d-diastereoisomeric mix-
tures of receptors 1 or 2.
Binding studies: We first investigated the complexation
properties of the unsubstituted receptor 1 by UV and fluo-
rescence titration studies in aqueous buffer solution.[22] For
these experiments, samples of 1 were purified by preparative
HPLC before use. Control experiments confirmed that up
to a concentration of [1] ꢀ8·10À5 m ([bis-tris buffer]
=
0.032m in 10% DMSO in water) the UV absorption of the
pyrrole band at lmax = 300 nm displayed linear Lambert–
Beer behavior (Figure 2), but at higher concentrations of 1
Scheme 2. Synthesis of the triethyleneglycol-substituted receptor 2.
presence of triethylamine to provide the TEG-substituted
diester 10, which was isolated from the reaction mixture di-
rectly by RP18-MPLC. In this way the yield could be signifi-
cantly increased over that of the aqueous workup described
by us previously (64 versus 46%, respectively).[21] Transes-
terification of the ethyl ester to the benzyl ester 10b was
achieved with sodium benzylate in benzyl alcohol under re-
duced pressure to remove the liberated EtOH. When we
tried to cleave the tBu ester in 10b by a standard method
(20% TFA in CH2Cl2 at RT) we also observed significant
decomposition of the product before all starting material
had reacted, resulting in low overall yields, and so we limit-
ed the reaction time to 2 h, when decomposition was just
starting according to TLC monitoring. In this way the free
acid 11 was obtained in a yield of 40%, together with 35%
of reisolated and reusable starting material. The free acid
group in 11 was then coupled with Boc-protected guanidine
in the presence of 1H-benzotriazolium-1-[bis(dimethylami-
no)methylene]-5-chloro-hexafluorophosphate-1,3-oxide
(HCTU) as the coupling reagent.
Because of increasing difficulties with the isolation of the
TEG-substituted products we decided to protect the two al-
cohol functions as silyl ethers. In the presence of an excess
of NMM (4-methylmorpholine) and triethylamine, TBS-Cl
was added in portions with frequent TLC monitoring to pro-
vide the optimal reaction time needed for the maximum
yield of the O-diprotected product (without N-protection of
the pyrrole NH), and in this way the disilylated product
TBS-12 could be obtained in 67% yield. Hydrogenolysis of
the benzyl ester and subsequent coupling of the resulting
free carboxylic acid with 6 gave the protected receptor 13,
which could be easily purified by flash chromatography on
silica gel. TBS removal with TBAF and Boc removal with
TFA provided the free cationic triethyleneglycol-substituted
Figure 2. Comparison of the experimentally observed absorption of re-
ceptor 1 at 300 nm with the theoretical prediction according to the Lam-
bert–Beer rule. The deviation at larger concentrations (>0.1 mm) indi-
cates the start of self-aggregation of receptor 1.
the solution became foamy and a significant deviation from
linearity was observed, most probably due to the start of
self-aggregation of the receptor. Binding studies were there-
fore performed in water with 10% DMSO (added to in-
crease the solubility of the receptor) at a receptor concen-
tration of [1] ꢀ3·10À5 m to prevent any interference due to
self-aggregation of 1.
Upon addition of Ac-d-Ala-d-Ala-OH (14) to a solution
of 1 at pH 6.1 (3 mm bis-tris buffer[23]) a significant decrease
in the absorbance of the pyrrole band was observed
(Figure 3). From this binding isotherm an association con-
stant of K = 33100mÀ1 (logK = 4.52) was calculated by a
nonlinear curve-fitting procedure taking dilution into ac-
count (estimated error in K is Æ30%).[24] The 1:1 complex
stoichiometry was determined by a Job plot extracted from
the titration.[25] The high affinity of receptor 1 for this dipep-
tide Ac-d-Ala-d-Ala-OH (14) was independently confirmed
by fluorescence titration under the same conditions with
monitoring of the decrease in intensity at 331 nm
9188
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 9186 – 9195

Artificial Dipeptide Receptor
FULL PAPER
Figure 3. Binding isotherm (logK = 4.52) as obtained from a UV titra-
tion of 1 with Ac-d-Ala-d-Ala-OH (14) in aqueous buffer solution with
monitoring of the decrease in absorbance of 1 at l = 300 nm. The solid
line represents the curve fitting a 1:1 complexation, whereas the dotted
line indicates the expected absorbance change due to simple dilution of
the sample if no complexation were to take place.
Figure 5. At mm concentrations receptor 2 caused gelation of an aqueous
solution during attempts to perform NMR titrations in D2O.
already described UV and fluorescence titration results be
obtained.[29]
However, the formation of a 1:1 complex between recep-
tor 2 and dipeptide 14 could also be confirmed by ESI-MS
experiments (Figure 6). Whereas the unsubstituted receptor
1 did not give any signals due to complex formation in the
ESI-MS, the TEG-substituted receptor 2 allowed the detec-
tion of the complex in the negative ion mode.
Figure 4. Fluorescence titration of 1 with Ac-d-Ala-d-Ala-OH (14) in
aqueous buffer. The data were analyzed through the decrease in intensity
at l = 331 nm to minimize any interference by the background fluores-
cence of the buffer (g).
(Figure 4).[26] Analysis of the binding isotherm gave a bind-
ing constant of K = 39800mÀ1 (logK = 4.60).
As already mentioned above, the solubility of receptor 1
is limited by its self-aggregation to rather low concentrations
(<0.1 mm), so no binding studies at higher concentrations—
as might be needed for NMR titrations, for example—were
possible. This was our initial reason for also synthesizing the
TEG-substituted receptor 2, its solubility in water being
much better. A UV titration in pure water (without any
added DMSO) was therefore performed ([bis-tris buffer] =
4 mm, pH 5.8)[27] and provided a similar affinity of logK =
4.14 for dipeptide 14. Unfortunately, though, gelation of the
solution occurred when we tried to perform NMR binding
studies with the TEG-substituted receptor 2 at mM concen-
trations (Figure 5), thus preventing any further investigation
by NMR spectroscopy. This is most probably due to the am-
phiphilic nature of 2: gelation might be caused by self-aggre-
gation and stacking of the nonpolar cyclotribenzylene units,
together with the attached hydrophilic TEG groups, which
are capable of strongly interacting with water molecules.[28]
No further experimental binding studies could thus be per-
formed and nor could any other structural data besides the
Figure 6. ESI-MS spectrum (negative ion mode) of a 1:1 mixture of 2 and
14 in water. Besides the signals for the receptor, an intense signal at m/z
1059.50 a.u. indicates the formation of a 1:1 complex.
Substrate selectivity: The binding studies had so far con-
firmed that receptor 1 is indeed capable of strong binding to
the dipeptide Ac-d-Ala-d-Ala-OH (14) in buffered water,
with a binding affinity even larger than that observed with
our previously reported dipeptide receptor.[5] Whereas the
latter receptor preferred larger and more hydrophobic resi-
dues, resulting in a greater affinity for Val-Val than for Ala-
Ala, it was believed that our new receptor 1 should now
show the opposite binding affinity, due to the size-selecting
hydrophobic cavity, so various other dipeptides in which one
or both of the alanines had been exchanged for glycine or
valine were studied in order to examine the substrate selec-
tivity of 1. The results obtained from UV titration studies
Chem. Eur. J. 2006, 12, 9186 – 9195
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9189

C. Schmuck et al.
under the same conditions as described above are summar-
ized in Table 1 and are shown in Figure 7.
(20). This most probably reflects the loss of any hydrophobic
interactions and the increasing flexibility of the glycine-rich
dipeptides in relation to their alanine counterparts. Interest-
ingly, the depsipeptide Ac-d-Ala-d-Lac-OH (15) (Lac =
lactate) is also bound (logK = 4.3), though slightly less effi-
ciently than Ac-d-Ala-d-Ala-OH (14) or Ac-l-Ala-l-Lac-
OH (ent-14). This could be due to the exchanging of an at-
tractive hydrogen bond with the amide NH of the dipeptide
in the complex for a repulsive dipole interaction with the
ester oxygen of the depsipeptide; similar effects in artificial
systems have been observed before both by us and by
others.[30] The same effect is also responsible for the reduced
affinity of the antibiotic Vancomycin to this depsipeptide se-
quence in resistant bacterial strains.[31]
Table 1. Calculated binding constants for the complexation of various di-
peptides by receptor 1 obtained from UV titrations in buffered water.
Dipeptide
K[a]
logK
Ac-d-Ala-d-Ala-OH (14)
Ac-l-Ala-l-Ala-OH (ent-14)
Ac-d-Ala-d-Lac-OH (15)
Ac-d-Ala-Gly-OH (16)
Ac-Gly-d-Ala-OH (17)
Ac-d-Ala-d-Val-OH (18)
Ac-d-Val-d-Ala-OH (19)
Ac-Gly-Gly-OH (20)
33100
30900
18600
4911
4260
5020
6300
2900
–
4.52
4.49
4.27
3.69
3.63
3.70
3.80
3.47
Ac-d-Val-d-Val-OH (21)
Ac-l-Ala-OH (22)
<3.0
<3.0
–
[a] K in mÀ1, estimated error limit in K < Æ30%.
Molecular modeling studies: The binding studies in aqueous
buffer confirmed that receptor 1 indeed shows sequence-de-
pendant dipeptide binding, preferring alanine over valine or
glycine. Hence, even though hydrophobic interactions in
general increase simply with the areas of the interacting sur-
faces, the use of the size-discriminating shallow cyclotriben-
zylene cavity in 1 allowed the hydrophobic interactions to
be restricted to small alkyl residues. As no further structural
data could be obtained experimentally because of solubility
problems we turned to molecular mechanics calculations
(Macromodel V8.0, Amber* force field, GBSA water solva-
tion).[32] A Monte Carlo conformational search with 100000
steps was performed until the resulting energy-minimized
structure was found several times, and this structure was
then further subjected to a MD simulation at 300 K for
100 ps. The obtained complex structure is shown in Figure 8.
The dipeptide 14 is bound in a conformation that allows for
ion pairing with the guanidinium cation of receptor 1 and
hydrogen-bond formation between the backbone amides.
Furthermore, the N-terminal methyl group is directly situat-
ed above the cyclotribenzylene bowl, which should allow for
favorable hydrophobic interactions between those two
groups in water, so substrates lacking this methyl group
(Gly) or possessing a too bulky group (Val) that does not fit
into the aromatic bowl are bound less efficiently. Their bind-
ing constants are all around K ꢁ 103 mÀ1, similar to a value
expected for simple ion pairing between an amino acid car-
boxylate and our guanidiniocarbonylpyrrole moiety.[5]
Figure 7. Binding constants (Kass) for complex formation between recep-
tor 1 and various dipeptides in buffered water.
Inspection of the data in Figure 7 and Table 1 shows that
receptor 1 indeed favors alanine-containing peptides such as
d-Ala-d-Ala (14) or its enantiomer l-Ala-l-Ala (ent-14)
over dipeptides with other side chains. The receptor does
not show any stereoselectivity between Ac-d-Ala-d-Ala-OH
(14) and Ac-l-Ala-l-Ala-OH (ent-14), although this should
not be unexpected, at least from the results of modeling
studies (see below). However, a significant drop in affinity is
observed if Ala is exchanged for Val: compound 14 is bound
about six times more strongly than d-Val-d-Ala (19) or d-
Ala-d-Val ( 18), for example, whilst binding of d-Val-d-Val
(21) was not even detectable by UV titration under these
conditions, suggesting a value of logK = 3.0, similarly to the
situation with simple amino acids. Receptor 1 thus has an af-
finity for d-Ala-d-Ala over 10 times larger than that for d-
Val-d-Val. Obviously, the large isopropyl side chain in valine
is too bulky and does not allow direct favorable interactions
with the shallow hydrophobic cyclotribenzylene cavity of
the receptor. The binding affinity also drops when alanine is
exchanged for glycine: 14 is bound about eight times more
strongly than d-Ala-Gly (16) or Gly-d-Ala (17), for exam-
ple, and again about ten times more strongly than Gly-Gly
The modeling studies also reproduced the observed lack
of stereoselectivity with regard to the two dipeptides d-Ala-
d-Ala and l-Ala-l-Ala, as well as the nearly identical bind-
ing features of the two P/d and M/d-diastereoisomers of re-
ceptor 1. The linkage between the cyclotribenzylene moiety
À
and the rest of the receptor is a C C single bond, allowing
for easy rotation, so the bowl can adjust to the individual
position of the alanine methyl group; the calculated struc-
tures for the two complexes of the two diastereomers of re-
ceptor 1 with the dipeptide 14, for example, differ only in a
tilt of the cyclotribenzylene cavity relative to the alanine
methyl group and are energetically virtually identical within
the error limit of the calculation (DE < 2 kJmolÀ1), so no
significant difference in binding affinity are to be expected,
9190
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 9186 – 9195

Artificial Dipeptide Receptor
FULL PAPER
Experimental Section
General remarks: Solvents were dried and distilled under argon before
use. All other reagents were used as commercially obtained. 1H and
13C NMR shifts are reported relative to the deuterated solvents. Peak as-
signments are based on either DEPT or 2D NMR studies and/or compar-
ison with literature data. IR spectra were recorded with samples pre-
pared as tablets (KBr) or with ATR technique. Melting points are uncor-
rected.
N-Boc-(P/M)-d-Ctb-OMe (5): A suspension of zinc (900 mg, 13.8 mmol)
and 1,2-dibromoethane (58 mL, 675 mmol) in dry DMF (1.5 mL) was stir-
red and heated to 1108C with a heat gun under argon five times. Chloro-
trimethylsilane (18 mL, 142 mmol) was added and the suspension was stir-
red for 30 min at room temperature. A solution of N-Boc-d-iodoalanine
methyl ester (4, 741 mg, 2.25 mmol) in dry DMF (1.0 mL) was added by
syringe and the reaction mixture was stirred for 1 h at room temperature.
Iodocyclotribenzylene 3 (Ctb)[17] (893 mg, 2.25 mmol) and then [Pd2-
AHCTRE(UNG dba)3] (60 mg, 66 mmol) and tri-o-tolylphosphine (66 mg, 217 mmol) were
added and the reaction mixture was stirred at room temperature for
10 min. After addition of further DMF (1.5 mL) the mixture was stirred
for 2 h, AcOEt (75 mL) was added, and the solid was separated by filtra-
tion and washed with AcOEt. The organic filtrate was washed with H2O
(2”50 mL) and brine (50 mL) and the solvent was removed in vacuo.
The brown solid was purified by flash column chromatography to afford
5 (791 mg, 1.68 mmol, 75%) as a white solid. M.p. 157–1588C; Rf = 0.47
(CH2Cl2/AcOEt 25:1 + 1% NEt3); 1H NMR (300 MHz, CDCl3; since 5
is a mixture of two diastereomers, each of which exists in the form of two
conformational isomers about the carbamate protecting group, up to four
sets of signals can be observed for each proton in the NMR): d = 1.36,
Figure 8. Energy-minimized structure for the complex formed between
receptor 1 (green) and the dipeptide d-Ala-d-Ala 14 (gray) from the top,
highlighting the hydrogen bonds (red dotted lines), and from the side,
showing the interaction between the methyl group and the cyclotribenzy-
lene cavity.
1.41, 1.45, 1.48 (4”s, 9H; CACTHER(UGN CH3)3), 2.96–3.05 (m, 2H; benzyl-CH2), 3.60,
3.66, 3.73, 3.75 (4”s, 3H; CO2CH3), 3.69–3.78 (m, 3H; benzyl-CHH),
4.45–4.55 (m, 1H; CH), 4.84–4.89 (m, 3H; benzyl-CHH), 6.83–7.44 ppm
(m, 11H; aryl-CH); 13C NMR (75 MHz, CDCl3; since 5 is a mixture of
two diastereomers, each of which exists in the form of two conformation-
al isomers about the carbamate protecting group, up to four signals can
be observed for each carbon in the NMR): d = 28.2, 28.3, 30.9 (C-
AHCTREGNU(CH3)3), 36.8, 37.1, 37.2 (benzyl-CH2), 52.1, 52.2 (CH; CO2CH3), 126.9,
at least from these calculations. The flexibility of the recep-
tor and the resulting structural similarity of the two diaster-
eoisomers is also most probably the reason for the failure,
mentioned above, to accomplish any separation by chroma-
tography. Nevertheless, even with this diastereomeric mix-
ture of receptor 1, sequence-dependant binding of dipepti-
des in aqueous solution with a selectivity of up to a factor of
10 is possible.
127.0, 127.8, 127.9, 129.9, 130.0, 130.1, 130.2, 130.3, 130.9, 131.0, 134.4,
134.5, 138.1, 138.2, 139.2, 139.3, 139.4, 139.5, 139.6 (aryl-C), 172.3 ppm
(C=N; C=O); IR (ATR): n˜ = 3336 (bm), 2982 (bw), 1738 (m), 1692,
1593, 1527 (s), 1488, 1444 (w), 1364 (m), 1251 (w), 1168 (m), 1061, 1018,
746 (w), 719 (s), 630 (w), 612, 605 cmÀ1 (s); MS (EI): m/z (%): 471 (5)
[M]+, 415 (10) [MÀC4H10]+, 397 (10), 354 (75), 312 (20), 283 (90), 191
(30), 179 (75), 88 (35), 57 (100); HR-MS: m/z: calcd for C30H33NO4:
471.241; found: 471.240 [M]+; elemental analysis calcd (%) for
C30H33NO4 (471.6): C 76.41, H 7.05, N 2.97; found: C 76.32, H 7.00, N
3.01.
N-Boc-(P,M)-d-Ctb-NH2
through solution of
C
A stream of ammonia was passed
a
Conclusion
(150 mL) for 1 h at 08C. This solution was stirred for 4 d at room temper-
ature, during which ammonia was passed through the solution for 10 min
at 08C once a day. After complete conversion (TLC monitoring) a strong
stream of argon was passed through the solution for 1 h at room temper-
ature to remove the ammonia. After removal of the solvent in vacuo,
Boc-6 (870 mg, 1.91 mmol, quant.) was obtained as a colorless solid. M.p.
268–2698C; Rf = 0.48 (CH2Cl2/MeOH 20:1 + 0.1% NEt3); 1H NMR
(300 MHz, [D6]DMSO; since Boc-6 is a mixture of two diastereomers,
each of which exists in the form of two conformational isomers about the
carbamate protecting group, up to four sets of signals can be observed
for each proton/carbon (see below) in the NMR): d = 1.12, 1.18, 1.29,
In conclusion, we have demonstrated here that an artificial
receptor (1) for d-Ala-d-Ala can be rationally designed by
the same binding principles as found in the natural antibiot-
ic Vancomycin: strong electrostatic interactions in combina-
tion with a size-selecting shallow hydrophobic cavity deter-
mine the binding features of receptor 1. Even the rather un-
specific and nondirectional hydrophobic interactions can
hence be used to discriminate between related substrates
such as d-Ala-d-Ala and d-Val-d-Val. A more deliberate
use of hydrophobic interactions might be a promising ap-
proach to improve the performance of supramolecular re-
ceptors for aqueous solvents in the future.
1.31 (4”s, 9H; C
3H; benzyl-CHH), 3.98–4.2 (m, 1H; CH), 4.86–4.95 (m, 3H; benzyl-
CHH), 6.67, 6.76 (d, 4J
(H,H) = 7.95 Hz, 1H; NH), 6.93–7.44 ppm (m,
11H; aryl-CH); 13C NMR (75 MHz, [D6]DMSO): d
28.3, 28.4 (C-
(CH3)3), 35.9, 36.2, 36.4, 37.0, 37.1, 37.2 (benzyl-CH2), 55.5, 55.6 (CH),
78.0, 78.1 (C(CH3)3), 126.7, 126.8, 126.9, 127.5, 127.8, 129.9, 130.0, 130.1,
130.2, 130.3, 130.6, 131.0, 136.6, 136.8, 137.8, 139.2, 139.4, 139.8, 139.9,
140.0 (aryl-C), 155.3, 155.4, 173.7, 173.9 ppm (C=O); IR (ATR): n˜
ACHTRE(UGN CH3)3), 2.56–2.85 (m, 2H; benzyl-CH2), 3.59–3.71 (m,
AHCTREUNG
=
AHCTREUNG
AHCTREUNG
=
Chem. Eur. J. 2006, 12, 9186 – 9195
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9191

C. Schmuck et al.
3343, 3188 (bw), 2955 (m), 2920 (s), 2850 (m), 2359 (w), 1659 (s), 1520,
1488, 1455, 1365, 1249, 1167, 1047, 744, 719 cmÀ1 (w); MS (ESI): m/z
(%): 479 (100) [M+Na]+, 401 (25), 357 (35), 312 (5); HR-MS: m/z: calcd
for C29H32N2NaO3Na: 479.2311; found: 479.231 [M+Na]+.
was obtained as a yellowish oil. MPLC (RediSep C-18, reverse phase,
130 g): 7 min H2O
0.1% TFA, then 20 min to MeOH + 0.1% TFA, 40 mLminÀ1, tR
30.6 min; 1H NMR (300 MHz, CDCl3): d = 1.32 (t, 3J
3H; CO2CH2CH3), 1.53 (s, 9H; C(CH3)3), 3.43 (s, 2H; OH), 3.50–3.75
(H,H) = 7.0 Hz, 2H; CO2CH2CH3), 4.75
+
0.1% TFA, then 18 min to MeOH/H2O 2:1
+
=
AHCTRE(UNG H,H) = 7.0 Hz,
AHCTREUNG
N-Boc-CBS-NH-(P/M)-d-Ctb-NH2 ACTHRE(UNG Boc-1): A stream of dry HCl was
3
(m, 24H; CH2OCH2), 4.30 (q, JACHTREUNG
passed through a solution of Boc-6 (557 mg, 1.22 mmol) in CHCl3/MeOH
2:1 (75 mL) for 1 h at 08C. A stream of argon was then passed through
the solution for 15 min at room temperature to remove the HCl. The sol-
vent was removed in vacuo and the remaining solid (6) was dried in high
vacuum. N-Boc-CBS-H (7, 360 mg, 1.22 mmol), PyBOP (632 mg,
1.22 mmol), and 4-methylmorpholine (1.25 mL, 11.4 mmol) were added,
and the mixture was dissolved in DMF (25 mL) under argon and stirred
for 2 d at room temperature. Brine (50 mL) was added to the yellowish
solution and the mixture was extracted with chloroform (5”100 mL).
The combined organic phase was washed with sat. NaHCO3 (200 mL),
H2O (200 mL), and brine (200 mL) and dried in vacuo. The remaining
yellow oil was purified by flash column chromatography, and the ob-
tained oil was digerated in a few mL of dry diethyl ether and treated in
an ultrasonic bath. The precipitating solid was crystallized from CH2Cl2
to afford Boc-1 (379 mg, 600 mmol, 49%) as a white solid. M.p. 2318C
(s, 2H; pyrrole-CH2), 4.78 (s, 2H; pyrrole-CH2), 9.57 ppm (s, 1H; pyr-
role-NH); 13C NMR (75 MHz, CDCl3): d = 14.2 (CO2CH2CH3), 28.2 (C-
AHCTREUNG
AHCTREUNG
124.3, 126.5, 127.4 (pyrrole-Cq), 159.7, 160.3 ppm (CO2Et, CO2tBu); IR
(ATR): n˜ = 3441 (bw), 2920, 2865 (m), 1698 (s), 1455, 1365 (w), 1282,
1142, 1077 cmÀ1 (s); MS (ESI): m/z (%): 586 (55) [M+Na]+, 530 (100)
[MÀC4H8+Na]+; HR-MS: m/z: calcd for C26H45NNaO12Na: 586.2839;
found: 586.284 [M+Na]+.
2-Benzyl 5-tert-butyl 3,4-bis(TEG-methyl)-1H-pyrrole-2,5-dicarboxylate
(10b): Sodium (510 mg, 22.2 mmol) was treated with benzyl alcohol
(40 mL) under argon at 08C. After complete conversion the solution was
allowed to warm to room temperature and dropped by transfer cannula
onto 10 (931 mg, 1.65 mmol). The resulting solution was stirred at
1.2 mbar and heated to reflux for 3 h. The solvent was removed by distil-
lation at 0.1 mbar and the resulting oil was purified by MPLC to afford
10b (720 mg, 1.15 mmol, 70%) as a yellowish oil. Rf = 0.56 (CH2Cl2/
MeOH 5:1 + 0.1% TFA); MPLC (RediSep C-18 reverse phase, 43 g):
2 min MeOH/H2O 1:1 + 0.1% TFA, then 33 min to MeOH + 0.1%
(decomp); Rf
= 0.24 (CH2Cl2/acetone 5:1 +
1% NEt3); 1H NMR
(300 MHz, [D6]DMSO, T = 558C; since Boc-1 is a mixture of two dia-
stereomers, two sets of signals are observed for some protons/carbons): d
= 1.45 (s, 9H; CACHTREUNG(CH3)3), 2.72–3.01 (m, 2H; benzyl-CH2), 3.48–3.69 (m,
3H; benzyl-CHH), 4.45–4.66 (m, 1H; CH), 4.82–4.92 (m, 3H; benzyl-
CHH), 6.80–7.43 (m, 13H; aryl-CH, pyrrole-CH), 7.53 (brs, 1H; NH),
8.42–8.50 (m, 1H; NH), 8.56 (brs, 1H; NH), 9.32 (brs, 1H; NH), 10.85
(brs, 1H; NH), 11.50 ppm (brs, 1H; NH); 13C NMR (75 MHz,
1
TFA, 25 mLminÀ1, tR = 28.2 min; H NMR (400 MHz, CDCl3): d = 1.57
(s, 9H; CCATHER(UGN CH3)3), 2.50 (s, 2H; OH), 3.48–3.72 (m, 24H; CH2OCH2), 4.80
(s, 4H; pyrrole-CH2), 5.33 (s, 2H; CO2CH2Ph), 7.36–7.42 (m, 5H; aryl-
CH), 9.58 ppm (brs, 1H; pyrrole-NH); 13C NMR (100 MHz, CDCl3): d =
[D6]DMSO, T = 558C): d = 27.9 (CACTHREU(NG CH3)3), 31.1, 35.8, 35.9, 36.1, 36.2,
36.4, 36.5, 36.6 (benzyl-CH2), 53.5, 54.5 (CH), 112.9, 119.0, 119.1, 122.3,
126.6, 126.7, 126.8, 127.4, 127.7, 129.9, 130.0, 130.1, 130.2, 130.5, 130.6,
130.8, 136.5, 136.9, 137.8, 139.3, 139.5, 139.6, 139.7, 139.8, 139.9, 140.0
(pyrrole-C, aryl-C), 158.6, 158.7, 159.4, 159.5, 159.6, 173.4, 173.5 ppm (C=
N, C=O); IR (ATR): n˜ = 3375 (bm), 2977 (bw), 1727 (w), 1673 (m),
1625, 1547 (s), 1474, 1446 (m), 1393, 1368 (w), 1292, 1241, 1146 (s), 1094,
839, 781, 753 (w), 721, 611 cmÀ1 (m); MS (ESI): m/z (%): 673 (30)
[M+K]+, 657 (100) [M+Na]+, 567 (50), 537 (100), 321 (35), 312 (30), 280
(15), 258 (15); HR-MS: m/z: calcd for C36H38N6NaO5Na: 657.2802;
found: 657.280 [M+Na]+.
28.3 (C
ACHTREUNG
pyrrole-CH2, benzyl-CH2), 82.6 (C
AHCTREUNG
role-Cq),128.4, 128.5, 128.7 (aryl-CH), 135.4 (aryl-Cq), 159.7, 160.2 ppm
(CO2Bn, CO2tBu); MS (ESI): m/z (%): 648 (90) [M+Na]+, 592 (100)
[MÀC4H8+Na]+; HR-MS: m/z: calcd for C31H47NNaO12Na: 648.2991;
found: 648.2996 [M+Na]+.
5-Benzyloxycarbonyl-3,4-bis(TEG-methyl)-1H-pyrrole-2-carboxylic acid
(11): Compound 10b (460 mg, 735 mmol) was dissolved in dry CH2Cl2
(4 mL) and the reaction mixture was cooled to À208C. TFA (1 mL) was
added and the solution was allowed to warm to 58C and stirred for 2 h
with frequent TLC monitoring. Although the conversion was not yet
complete the reaction was stopped due to an appearing spot representing
a by-product (Rf = 0.72). The mixture was cooled to À208C, and the sol-
vent and the acid were removed under high vacuum. RP18 MPLC pro-
vided 11 (168 mg, 295 mmol, 40%) as a yellowish oil, together with reiso-
lated 10b (160 mg, 256 mmol, 35%). Rf = 0.36 (CH2Cl2/MeOH 5:1 +
0.1% TFA); MPLC (RediSep C-18 reverse phase, 43 g): 2 min H2O +
H-CBS-NH-(P,M)-d-Ctb-NH2 (1): A stream of HCl was passed through a
solution of Boc-1 (107 mg, 169 mmol) in MeOH/CHCl3 2:1 (25 mL) for
1 h at 08C. Argon was then passed through this solution for 15 min and
the solvent was removed in vacuo to afford 1 (92 mg, 161 mmol, 95%) as
a white solid. For UV and fluorescence measurements parts of this white
solid were purified by semipreparative RP18 HPLC. M.p. 1768C
(decomp); RP-HPLC (Supelcosil LC-18, 25 cm”10 mm, 5 mm): MeOH/
H2O 3:2
+
0.1% TFA, 4.0 mLminÀ1
,
tR
=
26.2 min; 1H NMR
0.1% TFA, then 21 min to MeOH + 0.1% TFA, 40 mLminÀ1, tR
16.2 min; 1H NMR (400 MHz, CDCl3):
3.57–3.80 (m, 24H;
=
(400 MHz, [D6]DMSO; since 1 is a mixture of two diastereomers, two
sets of signals are observed for some protons): d = 2.65–2.92 (m, 2H;
benzyl-CH2), 3.60–3.70 (m, 3H; benzyl-CHH), 4.53, 4.64 (m, 1H; CH),
4.84–4.93 (m, 3H; benzyl-CHH), 6.81–7.42 (m, 13H; aryl-CH, pyrrole-
CH), 7.59 (brs, 1H; NH), 8.43 (brs, 4H; NH2), 11.7 (s, 1H; NH),
12.4 ppm (s, 1H; NH); IR (KBr): n˜ = 3393, 3322 (bs), 3184 (bm), 3015,
2977, 2923, 2878 (w), 1698, 1663 (s), 1558, 1475 (m), 1446, 1404 (w), 1282
(m), 1260, 1202, 1092, 748, 721 cmÀ1 (m); MS (ESI): m/z (%): 535 (100)
[M+H]+, 287 (30); HR-MS: m/z: calcd for C31H31N6O3Na: 535.2457;
found: 535.244 [M+Na]+.
d =
CH2OCH2), 4.20 (brs, 2H; OH), 4.78 (s, 2H; pyrrole-CH2), 4.86 (s, 2H;
pyrrole-CH2), 5.32 (s, 2H; CO2CH2Ph), 7.36–7.42 (m, 5H; aryl-CH),
10.0 ppm (brs, 1H; pyrrole-NH); 13C NMR (100 MHz, CDCl3): d = 61.7,
61.8, 62.7, 64.4, 67.1, 69.1, 69.9, 70.2, 70.4, 70.5, 70.6, 70.7, 72.67, 72.70
(CH2OCH2, pyrrole-CH2, benzyl-CH2), 122.7, 124.0, 125.9, 126.6 (pyrrole-
Cq), 128.75, 128.79, 128.86 (aryl-CH), 135.3 (aryl-Cq), 159.9, 161.4 ppm
(CO2Bn, CO2H); MS (ESI): m/z (%): 608 (40) [M+K]+, 592 (100)
[M+Na]+, 470 (25); HR-MS: m/z: calcd for C27H39NNaO12Na: 592.2365;
found: 592.2370 [M+Na]+.
2-tert-Butyl 5-ethyl 3,4-bis(TEG-methyl)-1H-pyrrole-2,5-dicarboxylate
(10): The dibromide 9 (1.8 g, 4.23 mmol) was dissolved in dry CH2Cl2
(20 mL) and the solvent was removed in vacuo under argon to remove
traces of moisture in the starting material. The remaining solid was dis-
solved in dry triethyleneglycol (50 mL), and NEt3 (3 mL, 21.3 mmol) was
added under argon. The solution was stirred for 24 h at 1008C. After
cooling down to room temperature the solution was diluted with H2O
(200 mL) and filtration over a RP18 column (100% H2O to 100%
MeOH + 0.1% TFA) was carried out. All fractions exhibiting absorp-
tion at 300 nm were collected and the solvent was evaporated. The re-
maining oil was purified with MPLC and 10 (1.53 g, 2.71 mmol, 64%)
Benzyl
3,4-bis(TEG-methyl)-5-(Boc-guanidino)carbonyl-1H-pyrrole-2-
carboxylate (12): Compound 11 (1.21 g, 2.12 mmol), Boc-guanidine
(1.28 g, 6.63 mmol), HCTU (1.01 g, 2.44 mmol), and DMAP (609 mg,
4.98 mmol) were suspended under argon in dry CH2Cl2 (25 mL) and
DMF was added until all components were dissolved. The solution was
stirred for 3 d at room temperature, acetic acid (5 mL) was added, and
the solvent was removed in vacuo. Filtration over a RP18 column (100%
H2O to 100% MeOH + 0.1% AcOH) was carried out, all fractions ex-
hibiting absorption at 300 nm being collected and the solvent being
evaporated. The remaining oil was purified by MPLC, and 12 (680 mg,
957 mmol, 45%) was obtained as a yellowish oil. MPLC (RediSep C-18
9192
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 9186 – 9195

Artificial Dipeptide Receptor
FULL PAPER
reverse phase, 130 g): 2 min MeOH/H2O 2:1 + 0.1% TFA, then 25 min
MeOH in CH2Cl2, then 15 min to 5% MeOH in CH2Cl2, 30 mLminÀ1, tR
= 7.1 min; 1H NMR (400 MHz, CDCl3; since 13 is a mixture of two dia-
stereomers, each of which exists in the form of two conformational iso-
mers about the carbamate protecting group, up to four sets of signals can
be observed for each proton/carbon): d = 0.041, 0.044, 0.055, 0.058 (4”s,
to MeOH
+
0.1% TFA, 50 mLminÀ1
,
tR
=
13.8 min; 1H NMR
(400 MHz, CDCl3): d = 1.51 (s, 9H; CAHCTRE(UGN CH3)3), 1.82 (brs, 2H; OH), 3.11–
3.73 (m, 24H; CH2OCH2), 4.80 (s, 2H; pyrrole-CH2), 4.99 (s, 2H; pyr-
role-CH2), 5.32 (s, 2H; CO2CH2Ph), 7.32–7.43 (m, 5H; aryl-CH), 8.57 (s,
1H; NH), 9.10 (s, 1H; NH), 9.87 ppm (s, 2H; NH); 13C NMR (62.5 MHz,
12H; Si
A
U
ACHTREUNG(CH3)2C-
CDCl3): d = 28.2 (C
70.5, 70.6, 70.8, 72.8, 72.9 (CH2OCH2, pyrrole-CH2, benzyl-CH2), 77.4 (C-
ACHTREUNG(CH3)3), 121.3, 125.4, 128.3 (pyrrole-Cq), 128.59, 128.62, 128.8 (aryl-CH),
135.7 (aryl-Cq), 159.2, 160.4 ppm (C=O, C=N); MS (ESI): m/z (%): 733
(100) [M+Na]+, 711 (70) [M+H]+; HR-MS: m/z: calcd for
C33H50N4NaO13Na: 733.3267; found: 733.3257 [M+Na]+.
A
N
U
ACHTREU(NG CH3)3), 18.5 (SiAHCETR(UNG CH3)2CACHRTE(UNG CH3)3), 26.1 (Si-
N
CH2), 54.8, 55.0 (CH), 62.4, 62.72, 62.77, 62.79, 62.80, 63.0, 68.3, 68.5,
68.6, 69.5, 69.8, 69.9, 70.2, 70.4, 70.5, 70.6, 70.83, 70.85, 72.6, 72.7, 72.78,
72.80 (CH2OCH2, pyrrole-CH2), 122.6, 122.8, 124.6, 124.8, 125.9, 127.0,
127.1, 127.3, 128.2, 128.5, 129.9, 130.03, 130.06, 130.11, 130.20, 130.23,
130.4, 130.5, 131.3, 131.5, 135.9, 138.2, 138.3, 139.39, 139.48, 139.54, 139.6,
139.7, 139.8 (aryl-C), 158.3, 160.5, 160.7, 172.7, 173.5, 173.7, 174.4 ppm
(C=O, C=N); IR (film): n˜ = 3388, 3268 (bm), 2952, 2928, 2858, 1772,
1683 (m), 1635, 1556, 1541 (s), 1447 (m), 1298, 1246, 1148, 1098 (s), 836
(m); MS (ESI): m/z (%): 1226 (1) [M+K]+, 1210 (100) [M+Na]+, 1188
(3) [M+H]+; HR-MS: m/z: calcd for C62H94N6NaO13Si2Na: 1209.631;
found: 1209.633 [M+Na]+.
N-Boc-(TBS-TEG-CBS)-Bn
(TBS-12):
Compound
12
(510 mg,
717 mmol), TBS-Cl (217 mg, 1.44 mmol), and NEt3 (300 mL, 2.13 mmol)
were dissolved under argon in dry THF (30 mL) and the reaction mixture
was stirred overnight at room temperature. As TLC indicated 12 to be
still the major component of the solution, NMM and TBS-Cl were added
gradually with further TLC monitoring. Mono-and disilylation can be
monitored very well by TLC [silica gel, CH2Cl2/MeOH 10:1, Rf = 0.19
(12), 0.33 (mono-TBS adduct), 0.73 (TBS-12)]. After 3 d of stirring at
room temperature and overall addition of ꢀ5 equiv of TBS-Cl and
NMM the mixture was diluted with methyl tert-butyl ether (MTBE)
(200 mL) and washed with sat. NaHCO3 (200 mL), brine (200 mL), 5%
AcOH in H2O (80 mL), and brine (2”150 mL). The organic phase was
dried with Na2SO4 and the solvent was removed in vacuo. Flash column
chromatography (silica gel, cyclohexane/AcOEt 2:1) provided TBS-12
(450 mg, 479 mmol, 67%) as a colorless oil. Rf = 0.73 (CH2Cl2/MeOH
N-Boc-(TEG-CBS)-NH-(P,M)-d-Ctb-NH2
ACHTRE(UNG Boc-2): Compound 13
(34 mg, 28.6 mmol) was dissolved in THF (10 mL) and the reaction mix-
ture was cooled to 08C. TBAF·3H2O (20 mg, 63.0 mmol) was added and
the solution was stirred for 1 h at 08C. After the reaction mixture had
been allowed to warm to room temperature and further stirred for
20 min, AcOH (0.2 mL) was added and the solvent was removed in
vacuo. The crude product was purified by MPLC to afford Boc-2 (16 mg,
16.7 mmol, 56%) as a clear oil. After attempts to separate the diastereo-
meric mixture by HPLC had failed, the obtained oil was directly convert-
ed into 2 (next reaction). MPLC (RediSep C-18 peverse phase, 2”4.3 g):
2 min MeOH/H2O 1:1 + 0.1% AcOH, then 24 min to MeOH + 0.1%
10:1); 1H NMR (400 MHz, CDCl3): d
(CH3)2C(CH3)3), 0.879, 0.884 (2”s, 18H; Si
OC(CH3)3), 3.52–3.76 (m, 24H; CH2OCH2), 4.78 (s, 2H; pyrrole-CH2),
=
0.047, 0.053 (2”s, 12H; Si-
(CH3)2C(CH3)3), 1.51 (s, 9H;
A
N
A
ACHTREUNG
ACHTREUNG
4.93 (s, 2H; pyrrole-CH2), 5.33 (s, 2H; CO2CH2Ph), 7.32–7.43 (m, 5H;
aryl-CH), 8.52 (s, 2H; NH), 9.86 ppm (s, 1H; NH); 13C NMR (100 MHz,
CDCl3): d = À5.13, À5.11 (Si
26.1 (Si(CH3)2C(CH3)3), 28.2 (OC
70.59, 70.61, 70.7, 70.82, 70.84, 70.9, 72.77, 72.80 (CH2OCH2, pyrrole-
CH2, benzyl-CH2), 83.2 (OC(CH3)3), 121.4, 126.1, 128.3 (pyrrole-Cq),
A
ACHTERU(GN CH3)3), 18.5 (SiACHERT(UGN CH3)2CACHTRE(NUG CH3)3),
A
N
ACHTREUNG
AcOH, 10 mLminÀ1
, tR = 18.2 min; RP-HPLC (Supelcosil LC-18,
G
25 cm”4.6 mm, 5mm): 5 min MeOH/H2O 1:1 + 0.1% TFA, then 45 min
to MeOH + 0.1% TFA, 1.5 mLminÀ1, tR = 21.5, 22.1 min; MS (ESI):
m/z (%): 997 (5) [M+K]+, 981 (100) [M+Na]+, 959 (6) [M+H]+; HR-
MS: m/z: calcd for C50H66N6NaO13Na: 981.4580; found: 981.4581
[M+Na]+.
128.52, 128.55, 128.8 (aryl-CH), 130.0 (pyrrole-Cq), 135.7 (aryl-Cq), 154.6,
159.2, 160.4, 170.1 ppm (C=O, C=N); MS (ESI): m/z (%): 977 (4)
[M+K]+, 961 (100) [M+Na]+, 939 (37) [M+H]+; HR-MS: m/z: calcd for
C45H79N4O13Si2: 939.5177; found: 939.5179 [M+H]+.
N-Boc-(TBS-TEG-CBS)-H (23): TBS-12 (420 mg, 449 mmol) was dis-
solved in dry THF (25 mL) and a few mg of Pd/C were added. A stream
of H2 was passed through the solution for 2 h at room temperature, the
catalyst was filtered off, and the solvent was evaporated to afford 23
(381 mg, 449 mmol, quant.) as a white wax. Rf = 0.28 (CH2Cl2/MeOH
H-(TEG-CBS)-NH-(P,M)-d-Ctb-NH2 (2): Boc-2 (16 mg, 16.7 mmol) was
dissolved in MeOH/TFA 1:1 (10 mL) and the reaction mixture was stir-
red at room temperature for 1 h. The solvent was removed in vacuo and
the crude product was purified by MPLC. The obtained solid was lyophi-
lized twice with HCl (0.1m, 5 mL) to afford 2·HCl (9 mg, 10.1 mmol,
60%) as a white solid. m.p. 1078C; MPLC (RediSep C-18 reverse phase,
2”4.3 g): 3 min MeOH/H2O 3:2 + 0.1% TFA, then 35 min to MeOH +
0.1% TFA, 10 mLminÀ1, tR = 11.8 min; RP-HPLC (Supelcosil LC-18,
25 cm”4.6 mm, 5mm): 5 min MeOH/H2O 1:1 + 0.1% TFA, then 45 min
to MeOH + 0.1% TFA, 1.5 mLminÀ1, tR = 16.0, 16.6 min; 1H NMR
(600 MHz, CD3OD; since 2 is a mixture of two diastereomers, two sets of
signals are observed for some protons/carbons (see below)): d = 2.90–
3.83 (m, 31H, CH2OCH2; benzyl-CH2, pyrrole-CH2, CH), 4.34–5.01 (m,
6H; CH2OCH2, benzyl-CH2, pyrrole-CH2, CH), 6.63–7.46 ppm (m, 11H;
aryl-CH); DEPT135 (150 MHz, CD3OD): d = 35.0, 36.9, 37.11, 37.15,
37.33, 37.35, 37.39 (benzyl-CH2), 54.4, 55.9 (CH), 61.89, 61.91, 63.5, 69.1,
69.5, 70.4, 70.56, 70.63, 70.67, 70.74, 70.76, 70.89, 70.94, 70.97, 71.62,
71.65, 73.1, 73.3 (pyrrole-CH2, CH2OCH2), 127.3, 127.46, 127.53, 127.59,
127.62, 127.8, 128.6, 129.0, 130.59, 130.63, 130.68, 130.76, 130.84, 130.99,
131.05, 131.09, 131.2, 131.8, 132.4 ppm (aryl-CH); IR (KBr): n˜ = 3431
(bs), 2922 (w), 2872 (w), 1700 (m), 1686 (m), 1637 (s), 1285 (m), 1076 (s),
723 cmÀ1 (m); MS (ESI): m/z (%): 881 (36) [M+Na]+, 859 (100)
[M+H]+; HR-MS: m/z: calcd for C45H59N6O11Na: 859.4236; found:
859.4239 [M+Na]+.
1
10:1 + 0.1% AcOH); H NMR (400 MHz, CDCl3): d = 0.048, 0.051 (2”
s, 12H; Si
A
ACHTREUNG(CH3)3), 0.879, 0.881 (2”s, 18H; SiACHETR(UGN CH3)2CACHTREU(NG CH3)3),
1.59 (s, 9H; OCACHTREUNG
role-CH2), 4.95 (s, 2H; pyrrole-CH2), 9.03 (s, 1H; NH), 9.68 (s, 1H; NH),
10.64 (s, 1H; NH), 12.89 (s, 1H; NH), 15.84 ppm (s, 1H; CO2H);
13C NMR (100 MHz, CDCl3; quaternary signals in 1J, 2J, and 3J positions
next to N do not appear, due to prolonged relaxation times): d = À5.1
(Si
A
(CH3)3), 18.5 (Si
U
(CH3)3), 26.1 (Si
G
G
(OC
ACHTREUNG(CH3)3), 62.8, 70.7, 70.9, 72.8 ppm (CH2OCH2, pyrrole-CH2); MS
(ESI): m/z (%): 893 (24) [M+2NaÀH]+, 871 (100) [M+Na]+, 849 (22)
[M+H]+; HR-MS: m/z: calcd for C38H72N4NaO13Si2Na: 871.45266;
found: 871.45274 [M+Na]+.
N-Boc-(TBS-TEG-CBS)-NH-(P/M)-d-Ctb-NH2 (13): Compound 23
(104 mg, 123 mmol), 6·HCl (40 mg, 102 mmol), HCTU (90 mg, 218 mmol),
and DMAP (88 mg, 720 mmol) were suspended under argon in dry
CH2Cl2 (10 mL) and dissolved by addition of dry DMF, and the reaction
mixture was stirred for 24 h at room temperature. The mixture was dilut-
ed with MTBE (150 mL) and washed with sat. NaHCO3 (100 mL), brine
(150 mL), NaHSO4 (100 mL), and again brine (150 mL). To avoid Boc
cleavage a few drops of NEt3 were added. The solution was dried with
Na2SO4 and the solvent was removed in vacuo. The crude product was
purified by MPLC to afford 13 (32 mg, 26.9 mmol, 26%) as a clear oil.
MPLC (RediSep silica gel, 12 g, deactivated with NEt3): 1 min 2.5%
General procedure for UV and fluorescence titrations: Samples of the re-
ceptors were purified by preparative HPLC before use. A solution of the
receptor (1 or 2, c = (2.5–7.6)”10À5 m, 10 mL) in bis-tris buffer (c =
(1.1–3.3)”10À3 m, pH 5.77–6.29) was prepared and the pH was deter-
Chem. Eur. J. 2006, 12, 9186 – 9195
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9193

C. Schmuck et al.
mined. A solution of the substrate (c = (2.65–12.7)”10À4 m) was pre-
pared in the same buffer and adjusted to exactly the same pH
(DpH 0.02) as the solution of the receptor with HCl or NaOH. The solu-
tion of the receptor (1.8–2.2 mL) was placed in a cuvette and the spec-
trum was measured. The solution of substrate was subsequently added in
small steps (10–50 mL) with stirring and temperature control (T = 258C).
After every addition a spectrum was recorded. To rule out the spectral
changes being due simply to pH effects, the pH of the mixture was meas-
ured again after completion of addition: in all cases the pH had stayed
constant during the course of the titration (DpH < 0.02). A mathemati-
cal curve-fitting for a 1:1 complexation taking dilution into account was
performed by use of the decrease in absorption of the pyrrole band of
the receptor at l = 300 nm for the UV titrations and l = 331 nm for the
c) N. T. Southall, K. A. Dill, A. D. J. Haymet, J. Phys. Chem. B 2002,
106, 521–533.
[8] a) R. D. Süssmuth, ChemBioChem 2002, 3, 295–298; b) K. C. Nico-
laou, C. N. C. Boddy, S. Bräse, N. Winssinger, Angew. Chem. 1999,
111, 2230–2287; Angew. Chem. Int. Ed. 1999, 38, 2096–2152;
c) D. H. Williams, B. Bardsley, Angew. Chem. 1999, 111, 1264–1286;
Angew. Chem. Int. Ed. 1999, 38, 1173–1193; d) C. T. Walsh, S. L.
Fisher, I.-S. Park, M. Prahalad, Z. Wu, Chem. Biol. 1996, 3, 21–28.
[9] Cyclodextrins and calixarenes are also often used as hydrophobic
pockets in supramolecular receptors but these are too large to allow
discrimination between different alkyl groups such as methyl and
isopropyl.
[10] a) P. J. Loll, J. Kaplan, B. S. Selinsky, P. H. Axelsen, J. Med. Chem.
1999, 42, 4714–4719; b) D. H. Williams, M. S. Westwell, Chem. Soc.
Rev. 1998, 27, 57–64.
fluoACHTREUNGrescence titration. The data were also analyzed over the whole spec-
tral range with the aid of SpecFit analysis software. Both methods gave
the same binding constants within the experimental error of the method,
which is estimated to be Æ30% in K.
[11] For selected examples of other artificial receptors that bind d-Ala-
d-Ala, but mostly only in organic and not in aqueous solvents, see:
a) J. Shepherd, T. Gale, K. B. Jensen, J. D. Kilburn, Chem. Eur. J.
2006, 12, 713–720; b) C. Chamorro, R. M. J. Liskamp, Tetrahedron
2004, 60, 11145–11157; c) K. A. Ahrendt, J. A. Olsen, M. Wakao, J.
Trias, J. A. Ellman, Bioorg. Med. Chem. Lett. 2003, 13, 1683–1686;
d) A. Casnati, F. Sansone, R. Ungaro, Acc. Chem. Res. 2003, 36,
246–254; e) G. Chiosis, I. G. Boneca, Science 2001, 293, 1484–1486;
f) R. Xuo, G. Greiveldinger, L. E. Marenus, A. Cooper, J. A.
Ellman, J. Am. Chem. Soc. 1999, 121, 4898–4899.
Acknowledgements
Financial support by the Fonds der Chemischen Industrie and by the
Deutsche Forschungsgemeinschaft (SFB 630) is gratefully acknowledged.
[12] Carboxylate-binding pockets more closely related to the natural an-
tibiotic have also been used in the past, but these are synthetically
much more challenging without providing any improvement in
terms of affinity; see for example: a) H. T. ten Brink, D. T. S. Rijk-
ers, R. M. J. Liskamp, J. Org. Chem. 2006, 71, 1817–1824; b) R. J. Pi-
eters, Tetrahedron Lett. 2000, 41, 7541–7545; c) N. Pant, A. Hamil-
ton, J. Am. Chem. Soc. 1988, 110, 2002–2003.
[1] For review articles on artificial peptide receptors see: a) M. W.
Peczuh, A. D. Hamilton, Chem. Rev. 2000, 100, 2479–2494; b) H.-J.
Schneider, Adv. Supramol. Chem. 2000, 7, 185–216; c) H.-J.
Schneider, Angew. Chem. 1993, 105, 890–892; Angew. Chem. Int.
Ed. Engl. 1993, 32, 848–850; d) T. H. Webb, C. S. Wilcox, Chem.
Soc. Rev. 1993, 22, 383–395.
[2] Two recent examples for the application of artificial peptide recep-
tors can be found in: a) X. Salvatella, M. Martinell, M. Gairí, M. G.
Mateu, M. Feliz, A. D. Hamilton, J. de Mendoza, E. Giralt, Angew.
Chem. 2004, 116, 198–200; Angew. Chem. Int. Ed. 2004, 43, 196–
198; b) P. Breccia, M. vanGool, R. PerØz-Fernandez, S. Martín-San-
tamaria, F. Gago, P. Prados, J. de Mendoza, J. Am. Chem. Soc. 2003,
125, 8270–8284.
[3] For examples of peptide recognition in polar solvents see: a) C.
Schmuck, M. Heil, Chem. Eur. J. 2006, 12, 1339–1348; b) C.
Schmuck, M. Heil, J. Scheiber, K. Baumann, Angew. Chem. 2005,
117, 7374–7379; Angew. Chem. Int. Ed. 2005, 44, 7208–7212; c) M.
Kruppa, C. Mandl, S. Miltschitzky, B. Kçnig, J. Am. Chem. Soc.
2005, 127, 3362–3365; d) P. Krattinger, H. Wennemers, Synlett 2005,
706–708; e) C. Schmuck, M. Heil, ChemBioChem 2003, 4, 1232–
1238; f) A. T. Wright, E. V. Anslyn, Org. Lett. 2004, 6, 1341–1344;
g) T. Mizutani, K. Wada, S. Kitagawa, Chem. Commun. 2002, 1626–
1627; h) S. Rensing, T. Schrader, Org. Lett. 2002, 4, 2161–2164; i) M.
Sirish, H.-J. Schneider, Chem. Commun. 1999, 907–908; j) M.
Davies, M. Bonnat, F. Guillier, J. D. Kilburn, M. Bradley, J. Org.
Chem. 1998, 63, 8696–8703; k) R. Breslow, Z. Yang, R. Ching, G.
Trojandt, F. Odobel, J. Am. Chem. Soc. 1998, 120, 3536–3537.
[4] For review articles on the use of combinatorial receptor libraries in
supramolecular studies see: a) N. Srinivasan, J. D. Kilburn, Curr.
Opin. Chem. Biol. 2004, 8, 305–310; b) B. Linton, A. D. Hamilton,
Curr. Opin. Chem. Biol. 1999, 3, 307–312; c) Y. R. DeMiguel,
J. M. K. Sanders, Curr. Opin. Chem. Biol. 1998, 2, 417–421; d) W. C.
Still, Acc. Chem. Res. 1996, 29, 155–163.
[13] a) C. Schmuck, L. Geiger, Curr. Org. Chem. 2003, 7, 1485–1502;
b) C. Schmuck, Chem. Eur. J. 2000, 6, 709–718.
[14] For recent reviews on carboxylate binding by artificial receptors
see: a) K. A. Schug, W. Lindner, Chem. Rev. 2005, 105, 67–113;
b) M. D. Best, S. L. Tobey, E. V. Anslyn, Coord. Chem. Rev. 2003,
240, 3–15; c) P. A. Gale, Coord. Chem. Rev. 2003, 240, 191–221;
d) R. J. Fitzmaurice, G. M. Kyne, D. Douheret, J. D. Kilburn, J.
Chem. Soc. Perkin Trans. 1 2002, 841–864; e) T. S. Snowden, E. V.
Anslyn, Curr. Opin. Chem. Biol. 1999, 3, 740–746; f) P. D. Beer, P.
Schmitt, Curr. Opin. Chem. Biol. 1997, 1, 475–482; g) A. Bianchi,
K. Bowman-James, E. Garcia-EspaÇa, Supramolecular Chemistry of
Anions, Wiley-VCH, New York, 1997; h) F. P. Schmidtchen, M.
Berger, Chem. Rev. 1997, 97, 1609–1646; i) C. Seel, A. Galꢂn, J. de
Mendoza, Top. Curr. Chem. 1995, 175, 101–132.
[15] a) A. Collet, J.-P. Dutata, B. Lozach, J. Canceill, Top. Curr. Chem.
1993, 165, 103–129; b) A. Collet, Tetrahedron 1987, 43, 5725–5759.
[16] The more readily available cyclotriveratrylenes (= symmetrically
tris-substituted cyclotribenzylenes) have already been used in supra-
molecular chemistry but mostly only as rigid scaffolds to attach
other binding groups and not as binding pockets in themselves. For
some examples see: a) C. J. Sumby, J. Fisher, T. J. Prioir, M. J.
Hardie, Chem. Eur. J. 2006, 12, 2945–2959; b) S. Zhang, L. Eche-
goyen, J. Am. Chem. Soc. 2005, 127, 2006–2011; see also ref. [9b].
[17] A related receptor derived from simple phenylalanine instead of
this unnatural cyclotribenzylene amino acid had already been stud-
ied in the past but failed to show efficient and selective dipeptide
binding in aqueous solvents: C. Schmuck, W. Wienand, unpublished
results.
[5] C. Schmuck, L. Geiger, J. Am. Chem. Soc. 2004, 126, 8898–8899.
[6] For another interesting approach to sequence-specific peptide recog-
nition by designed receptors see: a) M. Wehner, D. Janssen, G. Schä-
fer, T. Schrader, Eur. J. Org. Chem. 2006, 138–153; b) M. Wehner,
T. Schrader, Angew. Chem. 2002, 114, 1827–1831; Angew. Chem.
Int. Ed. 2002, 41, 1751–1754.
[7] For reviews on hydrophobic interactions see: a) B. Widom, P. Bhi-
malapuram, K. Koga, Phys. Chem. Chem. Phys. 2003, 5, 3085–3093;
b) L. R. Pratt, A. Pohorille, Chem. Rev. 2002, 102, 2671–2691;
[18] C. Schmuck, W. Wienand, Synthesis 2002, 655–663.
[19] S. Gair, R. F. W. Jackson, Curr. Org. Chem. 1998, 2, 527–550.
[20] For the iodocyclotribenzylene 3 we did not observe any coalescence
due to ring flip in the NMR up to a temperature of 1208C. For data
on the conformational behavior of cyclotriveratrylenes and the un-
substituted parent hydrocarbon see also: T. Sato, K. Uno, J. Chem.
Soc. Perkin Trans. 1 1973, 895–900.
[21] C. Schmuck, W. Wienand, J. Am. Chem. Soc. 2003, 125, 452–459.
9194
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 9186 – 9195

Artificial Dipeptide Receptor
FULL PAPER
[22] a) K. Hirose, J. Inclusion Phenom. Macrocyclic Chem. 2001, 39,
193–209; b) K. A. Connors, Binding Constants, Wiley, New York,
1987.
[28] N. M. Sangeetha, U. Maitra, Chem. Soc. Rev. 2005, 34, 821–836.
[29] Other solvents such as methanol, in which the receptor is more
readily soluble, are not reasonable here as they would drastically
reduce the importance of hydrophobic interactions between the cy-
clotribenzylene moiety and the dipeptide, thereby destroying any
substrate selectivity.
[30] For similar conclusions in different systems see: a) C. Schmuck, U.
Machon, Chem. Eur. J. 2005, 11, 1109–1118; b) S.-Y. Chang, H. S.
Kim, K.-J. Chang, K.-S. Jeong, Org. Lett. 2004, 6, 181–184; c) G. M.
Kyne, M. E. Light, M. B. Hursthouse, J. de Mendoza, J. D. Kilburn,
J. Chem. Soc. Perkin Trans. 1 2001, 1258–1263; d) K. Kavallieratos,
C. M. Bertao, R. H. Crabtree, J. Org. Chem. 1999, 64, 1675–1683.
[31] a) B. M. Crowley, D. L. Boger, J. Am. Chem. Soc. 2006, 128, 2885–
2892; b) C. C. McComas, B. M. Crowley, D. L. Boger, J. Am. Chem.
Soc. 2003, 125, 9314–9315.
[23] Initial binding studies were carried out at buffer concentrations of
10À2 as well as at 10À3 m. It was observed that a buffer concentration
of 10À3 m is sufficient to ensure a stable pH during the titration
course (DpH 0.02). As an increasing concentration of buffer caused
a significant decrease in complex stability due to the salt effect
(DlogK ꢀ0.5 for the complex between 1 and 14), subsequent titra-
tions were performed at the lower buffer concentration of [buffer]
= 10À3 m.
[24] For the equations and procedures used for data analysis see: C.
Schmuck, M. Schwegmann, Org. Biomol. Chem. 2006, 4, 836–838.
[25] P. McCarthy, Anal. Chem. 1978, 50, 2165–2165.
[26] To avoid detection of the Raman band of water the spectra were re-
corded with use of synchronous wavelength screening with a Dl of
20 nm between excitation and emission.
[32] F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M.
Lipton, C. Caufield, G. Chang, T. Hendrickson, W. C. Still, J.
Comput. Chem. 1990, 11, 440–467.
[27] Titrations were carried out in a pH range from 5.77 to 6.29 to
ensure that the acyl guanidine is fully protonated while the dipep-
tide is still deprotonated. In this range we did not observe any varia-
tion in the association constant K as a function of pH for different
titrations with the same substrate.
Received: April 25, 2006
Published online: September 12, 2006
Chem. Eur. J. 2006, 12, 9186 – 9195
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9195