Large increase in cation binding affinity of artificial cyclopeptide receptors by an allosteric effect

Article abstract of DOI:10.1021/ja983970j

The receptor properties of a cyclopeptide composed of L-glutamic acid and 3-aminobenzoic acid in an alternating sequence are described, ¹H NMR, NOESY NMR, and FT-IR spectroscopic investigations show that this cyclic peptide is relatively flexib

Full text of DOI:10.1021/ja983970j

5846
J. Am. Chem. Soc. 1999, 121, 5846-5855
Large Increase in Cation Binding Affinity of Artificial Cyclopeptide
Receptors by an Allosteric Effect
Stefan Kubik
Contribution from the Institut fu¨r Organische Chemie und Makromolekulare Chemie,
Heinrich-Heine-UniVersita¨t, UniVersita¨tsstrasse 1, D-40225 Du¨sseldorf, Germany
ReceiVed NoVember 17, 1998. ReVised Manuscript ReceiVed March 8, 1999
Abstract: The receptor properties of a cyclopeptide composed of L-glutamic acid and 3-aminobenzoic acid in
an alternating sequence are described. 1H NMR, NOESY NMR, and FT-IR spectroscopic investigations show
that this cyclic peptide is relatively flexible in solution. Still, it is able to bind cations by cation-π interactions.
For the n-butyltrimethylammonium iodide complex, for example, an association constant of 300 M-1 has
been determined in chloroform. Besides cations, the cyclopeptide is also able to bind certain anions, such as
sulfonates or phosphonates, at a second binding site. NMR and FT-IR spectroscopic investigations show that
these anions are hydrogen bonded to the peptidic NH groups. Anion complexation results in an increase of the
cyclic peptide’s cation affinity by a factor of 103-104. The cyclopeptide-tosylate complex structure in solution
was assigned by FT-IR , 1H NMR, and NOESY NMR spectroscopic methods as well as molecular modeling.
This structure shows that the drastic increase in cation binding affinity can be correlated with a preorganization
of the cyclic peptide by the anion as well as electrostatic interactions between anion and cationic substrates in
the final complex. Therefore, the influence of the anions on the complexing behavior of the cyclopeptide can
be regarded as an allosteric effect. Association constants of the K+-18-crown-6, Na+-15-crown-5, and
n-butyltrimethylammonium cation complexes have been determined by dilution and competitive NMR titrations.
Introduction
to enzymes, where effector and substrate binding sites are
usually at remote positions in the molecule, in small artificial
hosts both guests are bound in relatively close proximity.
Especially when effector and substrate are charged, possible
interactions between the two guests can influence the receptor
cooperativity. For example, because of unfavorable Coulomb
interactions, a positive cooperativity is relatively unlikely when
effector and substrate are equally charged. We are aware of only
a single artificial receptor with a positive cooperativity in the
complexation of two metal cations.2l When, on the other hand,
effector and substrate have opposite charges, one would expect
that attractive electrostatic interactions between these compounds
should even promote binding. Ditopic receptors which bind both
components of an ion pair simultaneously have been described.3
Yet, to our knowledge, so far there are no examples for systems
in which the complexation of an ion results in a preorganization
of the receptor as well as the introduction of a new binding site
with which the substrate can interact. We have found that such
unusual receptor properties can be realized with cyclopeptides
One of the fundamental principles in biochemistry is the
regulation of enzyme activities by allosteric effects.1 The
interaction of an effector with the allosteric binding site of an
enzyme causes a conformational change at the active center and,
as a consequence, greatly improves the enzyme’s substrate
affinity or catalytic activity. The same principle has been
transferred to a variety of different artificial receptors in recent
years.2 Many of these compounds have in common that they
contain two conformationally coupled binding sites. When they
bind a guest molecule, a conformational reorganization at the
second binding site occurs and, similar to natural systems, the
affinity toward another guest is altered. However, in contrast
(1) Herve´, G. Allosteric enzymes; CRC Press: Boca Raton, FL, 1989.
Perutz, M. F. Mechanisms of cooperatiVity and allosteric regulation in
proteins; University Press: Cambridge, UK, 1990.
(2) (a) Traylor, T. G.; Mitchell, M. J.; Ciccone, J. P.; Nelson, S. J. Am.
Chem. Soc. 1982, 104, 4986-4989. (b) Rebek, J., Jr. Acc. Chem. Res. 1984,
17, 258-264. (c) Rebek, J., Jr.; Costello, T.; Marshall, L.; Wattley, R.;
Gadwood, R. C.; Onan, K. J. Am. Chem. Soc. 1985, 107, 7481-7487. (d)
Beer, P. D.; Rothin, A. S. J. Chem. Soc., Chem. Commun. 1988, 52-54.
(e) Hunter, C. A.; Meah, M. N.; Sanders, J. K. M. J. Chem. Soc., Chem.
Commun. 1988, 694-696. (f) Tabushi, I. Pure Appl. Chem. 1988, 60, 581-
586. (g) Ebmeyer, F.; Rebek, J., Jr. Angew. Chem., Int. Ed. Engl. 1990, 29,
1148-1150. (h) Petter, R. C.; Salek, J. S.; Sikorski, C. T.; Kumaravel, G.;
Lin, F. J. Am. Chem. Soc. 1990, 112, 3860-3868. (i) Schneider, H.-J.;
Ruf, D. Angew. Chem., Int. Ed. Engl. 1990, 29, 1159-1160. (j) Kobuke,
Y.; Sumida, Y.; Hayashi, M.; Ogoshi, H. Angew. Chem., Int. Ed. Engl.
1991, 30, 1496-1498. (k) Sijbesma, R. P.; Nolte, R. J. M. J. Am. Chem.
Soc. 1991, 113, 6695-6696. (l) Kobuke, Y.; Satoh, Y. J. Am. Chem. Soc.
1992, 114, 789-790. (m) Pierre, J.-L.; Gagnaire, G.; Chautemps, P.
Tetrahedron Lett. 1992, 33, 217-220. (n) Schneider, H.-J.; Werner, F. J.
Chem. Soc., Chem. Commun. 1992, 490-491. (o) Inouye, M.; Konishi, T.;
Isagawa, K. J. Am. Chem. Soc. 1993, 115, 8091-8095. (p) Haino, T.;
Katsutani, Y.; Akii, H.; Fukazawa, Y. Tetrahedron Lett. 1998, 39, 8133-
8136. (q) Lustenberger, P.; Welti, R.; Diederich, F. HelV. Chim. Acta 1998,
81, 2190-2200. (r) Takeuchi, M.; Imada, T.; Shinkai, S. Angew. Chem.,
Int. Ed. Engl. 1998, 37, 2096-2099.
(3) (a) Reetz, M. T.; Niemeyer, C. M.; Harms, K. Angew. Chem., Int.
Ed. Engl. 1991, 30, 1474-1476. (b) Arafa, E. A.; Kinnear, K. I.; Lockhart,
J. C. J. Chem. Soc., Chem. Commun. 1992, 61-64. (c) Flack, S. S.;
Chaumette, J. L.; Kilburn, J. D.; Langley, G. J.; Webster, M. J. Chem.
Soc., Chem. Commun. 1993, 399-401. (d) Rudkevich, D. M.; Verboom,
W.; Reinhoudt, D. N. J. Org. Chem. 1994, 59, 3683-3686. (e) Rudkevich,
D. M.; Brzozka, Z.; Palys, M.; Visser, H. C.; Verboom, W.; Reinhoudt, D.
N. Angew. Chem., Int. Ed. Engl. 1994, 33, 467-468. (f) Savage, P. B.;
Holmgren, S. K.; Gellman, S. H. J. Am. Chem. Soc. 1994, 116, 4069-
4070. (g) Rudkevich, D. M.; Mercer-Chalmers, J. D.; Verboom, W.; Ungaro,
R.; de Jong, F.; Reinhoudt, D. N. J. Am. Chem. Soc. 1995, 117, 6124-
6125. (h) Scheerder, J.; van Duynhoven, J. P. M.; Engbersen, J. F. J.;
Reinhoudt, D. N. Angew. Chem., Int. Ed. Engl. 1996, 35, 1090-1093. (i)
Deetz, M. J.; Smith, B. D. Tetrahedron Lett. 1998, 39, 6841-6844. (j)
Pelizzi, N.; Casnati, A.; Friggeri, A.; Ungaro, R. J. Chem. Soc., Perkin
Trans. 2 1998, 1307-1311.
10.1021/ja983970j CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/11/1999

Receptor Properties of a Cyclopeptide
J. Am. Chem. Soc., Vol. 121, No. 25, 1999 5847
Scheme 1. Synthesis of the Cyclic Hexapeptide 2
subunits was still relatively low, sufficiently high solubility could
be achieved by using the 5-isopropyl ester instead. We have
performed the synthesis of the corresponding cyclopeptide 2 in
solution because, in this way, the product could be obtained in
higher purity and in better yields than by synthesis on solid
support. The repeating dipeptide subunit was prepared by
coupling the BOC-protected 5-isopropyl ester of L-glutamic acid
with the benzyl ester of 3-aminobenzoic acid. This dipeptide
was elongated successively, first to the tetrapeptide and then to
the hexapeptide by using conventional peptide-coupling meth-
ods. The fully deprotected linear hexapeptide was cyclized under
high dilution conditions (Scheme 1).
Figure 1. Structures of cyclic hexapeptides 1 and 2.
composed of natural amino acids and 3-aminobenzoic acid in
an alternating sequence (Figure 1).
These compounds are interesting candidates for a new class
of artificial receptors. They resemble conventional host mol-
ecules such as calixarenes or cyclophanes and can be easily
obtained by a simple sequential synthesis. As a consequence, it
is possible to vary their basic structure in a wide range by
deliberate exchange of individual subunits. The rigid aromatic
subunits not only cause a reduction of the conformational
flexibility of these cyclic peptides, but they also allow easy
modification of the peptide by introduction of functional groups
with which substrates can interact. In addition, the side chains
of the amino acid subunits may also serve as binding sites.
The complexation of phosphate by 1 has already been
described.4 Here, we will show for the first time that such cyclic
peptides are also able to bind various cations. We have found
that their cation affinity depends on the anion present in solution.
With anions, such as phosphonates and sulfonates, which bind
to NH groups of the cyclopeptides, allosteric effects have been
observed. NMR and FT-IR spectroscopic investigations show
that these anions stabilize a conformation of the originally
relative flexible host which is ideally suited for an interaction
with cations. Binding of the positively charged guests with the
cyclopeptide anion complexes is due to cation-π interactions
with the aromatic subunits of the peptide as well as electrostatic
interactions with the anion. We will show that the combination
of both binding mechanisms leads to an increase of the complex
stability by a factor of 103-104 in comparison to that of cation
complexes without the influence of effector anions.
Cation Binding of the Cyclopeptide in the Absence of
Effector Anions. Cation-π interactions of quaternary am-
monium ions and, e.g., calixarenes can be conveniently detected
by NMR spectroscopy.6 When cations are included into a cavity
which is lined by aromatic subunits, they come in close contact
with the faces of the aromatic systems. As a result, complexation
1
causes an upfield shift of the guest protons in the H NMR.
The larger these shifts are, the better is the interaction of the
guests with the aromatic systems or the fit of the guests inside
the receptor cavity. In contrast, complexation by hydrogen
bonding usually results in downfield shifts of the protons
involved in binding.
We tested the binding affinity of 2 toward n-butyltrimethyl-
ammonium (BTMA+) iodide. When this quaternary ammonium
salt is added to a solution of 2 in CDCl3, very small upfield
shifts of the guest’s N-methyl protons in the order of -0.02
ppm are observed. An interaction between BTMA+ and the
cyclopeptide obviously takes place, but the influence of the
aromatic rings on the resonances of the guest molecule protons
is relatively weak. An even smaller upfield shift can also be
observed for all other protons of BTMA+. Yet the more remote
they are from the cationic headgroup, the less pronounced the
effect is. This suggests a preferable inclusion of the trimethyl-
ammonium group into the host cavity. Rotation of the guest
inside the cavity seems to be possible to a certain degree, which
would account for the small shifts of the n-butyl protons. The
resulting shift pattern is comparable to the one observed for
cation-π interactions between similar quaternary ammonium
ions and calix[5]arenes,6l homooxacalixarenes,6i or cryptophanes.6d
No significant change in the NMR spectrum of 2 was detected
during complexation.
Results and Discussion
Design and Synthesis. Compound 1 has previously been
synthesized on solid support, and its binding to disodium
4-nitrophenyl phosphate in DMSO-d6 was demonstrated.4 In this
solvent, complexation of cations was not observed. On the other
hand, it is known that artificial receptors such as, for example,
calixarenes or cyclophanes are able to bind cations by cation-π
interactions.5 Since 1 can, in principle, be regarded as a hybrid
between a calixarene and a conventional cyclic peptide, one
can speculate that it might also be able to bind positively charged
guest molecules by the same binding mechanism. Investigations
on the interactions of cations with uncharged artificial receptors
are usually carried out in organic solvents of low polarity such
as CDCl3.6 Because 1 is only poorly soluble in this solvent, we
had to prepare derivatives with improved solubility. By replacing
the natural amino acid L-alanine in 1 with esters of L-glutamic
acid, it was possible to control the solubility of the products by
varying the type of ester used. Whereas the chloroform solubility
of a cyclic peptide with three L-glutamic acid 5-methyl ester
(6) (a) Stauffer, D. A.; Dougherty, D. A. Tetrahedron Lett. 1988, 29,
6039-6042. (b) Araki, K.; Shimizu, H.; Shinkai, S. Chem. Lett. 1993, 205-
208. (c) De Iasi, G.; Masci, B. Tetrahedron Lett. 1993, 34, 6635-6638.
(d) Garel, L.; Lozach, B.; Dutasta, J.-P.; Collet, A. J. Am. Chem. Soc. 1993,
115, 11652-11653. (e) Me´ric, R.; Lehn, J. M.; Vigneron, J. P. Bull. Soc.
Chim. Fr. 1994, 131, 579-83. (f) Takeshita, M.; Nishio, S.; Shinkai, S. J.
Org. Chem. 1994, 59, 4032-4034. (g) Casnati, A.; Jacopozzi, P.; Pochini,
A.; Ugozzoli, F.; Cacciapaglia, R.; Mandolini, L.; Ungaro, R. Tetrahedron
1995, 51, 591-598. (h) Cattani, A.; Dalla Cort, A.; Mandolini, L. J. Org.
Chem. 1995, 60, 8313-8314. (i) Masci, B. Tetrahedron 1995, 51, 5459-
5464. (j) Arduini, A.; McGregor, W. M.; Paganuzzi, D.; Pochini, A.; Secchi,
A.; Ugozzoli, F.; Ungaro, R. J. Chem. Soc., Perkin Trans. 2 1996, 839-
846. (k) Magrans, J. O.; Ortiz, A. R.; Molins, M. A.; Lebouille, P. H. P.;
Sa´nchez-Quesada, J.; Prados, P.; Pons, M.; Gago, F.; de Mendoza, J. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1712-1715. (l) Arnecke, R.; Bo¨hmer, V.;
Cacciapaglia, R.; Dalla Cort, A.; Mandolini, L. Tetrahedron 1997, 53,
4901-4908.
(4) Ishida, H.; Suga, M.; Donowaki, K.; Ohkubo, K. J. Org. Chem. 1995,
60, 5374-5375.
(5) (a) Lhotak, P.; Shinkai, S. J. Phys. Org. Chem. 1997, 10, 273-285.
(b) Ma, J. C.; Dougherty, D. A. Chem. ReV. 1997, 97, 1303-1324.

5848 J. Am. Chem. Soc., Vol. 121, No. 25, 1999
Kubik
Figure 2. 1H NMR titration curve of the complex of 2 with
n-butyltrimethylammonium iodide.
Figure 3. Rotating bonds in a peptide chain between the aromatic
residues of cyclopeptide 2.
A 1:1 stoichiometry of the complex of 2 with BTMA+ iodide
was determined by Job’s method of continuous variations,7,8
and an association constant Ka of 300 M-1 was determined by
1H NMR titration (Figure 2).7,9 The association constant is in
the order of calixarene complexes with related quaternary
ammonium ions.6b,f,g,j,l Despite the small influence of the
receptor on the proton resonances of the guest, 2 obviously still
forms reasonably stable complexes with BTMA+.
Figure 4. 1H NMR spectra of 2 in DMSO-d6 (a) and in CDCl3 (c) as
well as intramolecular NOE effects in the NOESY NMR spectrum of
2 in DMSO-d6.
For the determination of the binding mechanism of 2, its
conformation in the complex is an important indication. In the
case of hydrogen bonding, all groups which are involved in
binding must converge in such a way that they are able to bind
the guest molecule at the suitable positions. In contrast,
cation-π interactions require that the aromatic subunits of 2
are arranged around a cavity so that the included guests are
able to interact with the faces of the aromatic systems.5b,10 In
contrast to the case with simple calix[n]arenes,11 possible
conformations of 2 are not only determined by the relative
orientations of the aromatic rings. Because there are six single
bonds between the aromatic subunits of the cyclopeptides, four
more than in calix[n]arenes, rotations around additional bonds
have to be taken into account. For the secondary amide groups,
energetically more favored trans conformations can be assumed.
Rotations around the remaining bonds should be possible and
cause different orientations of the protons at both nitrogens and
the C(R) with respect to the aromatic rings (Figure 3).
Figure 5. Intramolecular NOE effects in the NOESY NMR spectrum
of 2 in 5% DMSO-d6/CDCl3.
NOE effects with each of their adjacent protons (Figure 4b).
This large number of NOE effects shows that fast rotations
around all bonds indicated in Figure 3 occur in DMSO at room
temperature. The NOESY spectrum thus illustrates the high
flexibility of cyclopeptide 2 in this solvent. Temperature-
1
dependent H NMR spectra of 2 have been recorded in DMF-
d7. Only broadening of the signals was observed down to -60
°C. Even at this temperature, the flexibility of 2 is obviously
still too high for individual conformers to be resolved by NMR
spectroscopy.
1
In CDCl3, all signals in the H NMR spectrum of 2 are
somewhat broadened already at room temperature (Figure 4c).
An explanation of this line broadening could be that the
conformational flexibility of 2 is reduced in CDCl3, possibly
because some of its conformers are stabilized by intramolecular
hydrogen bonds. Indeed, the NOESY NMR spectrum of 2 in
5% DMSO-d6/CDCl3 is much simpler than the one in DMSO-
d6. Increasing line broadening precluded measuring suitable
spectra in solutions containing less DMSO-d6. But even in 5%
DMSO-d6/CDCl3, certain conformers of 2 seem to be favored.
Cross-peaks can be detected in the NOESY NMR spectrum
between the aromatic NH and H(4) of the adjacent aromatic
residue as well as between the glutamic acid NH and H(2)
(Figure 5). No cross-peak between an aromatic and a glutamic
acid NH group is visible. This shows that the orientations of
the amide groups with respect to the aromatic rings are relatively
fixed, and the aromatic NH groups point “up” toward H(4) and
The simple 1H NMR (Figure 4a) and 13C NMR spectra of 2
in DMSO-d6 give no information whether these spectra represent
a single C3-symmetrical conformer of 2 or an average of dif-
ferent conformational isomers which rapidly interconvert in solu-
tion. In this respect, the NOESY NMR spectrum of 2 in DMSO-
d6 is more instructive. Both NH groups and H(R) possess strong
(7) Connors, K. A. Binding Constants; Wiley: New York, 1987.
(8) Gil, V. M. S.; Oliveira, N. C. J. Chem. Educ. 1990, 67, 473-478.
(9) Macomber, R. S. J. Chem. Educ. 1992, 69, 375-378.
(10) Mecozzi, S.; West, A. P., Jr.; Dougherty, D. A. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 10566-10571.
(11) Gutsche, C. D. Inclusion compounds, Vol. 4; Atwood, J. L., Davies,
J. E. D., MacNicol, D. D., Eds.; Oxford University Press: Oxford, UK,
1991; pp 27-63. Shinkai, S. Tetrahedron 1993, 49, 8933-8968. Bo¨hmer,
V. Angew. Chem., Int. Ed. Engl. 1995, 34, 713-745.

Receptor Properties of a Cyclopeptide
J. Am. Chem. Soc., Vol. 121, No. 25, 1999 5849
an influence of complex formation on the solution structure of
2 or an interaction of the cyclic peptide with the iodide anions.
Therefore, the complexing properties of 2 can best be explained
when it is assumed that at least some of the conformers of 2
that are present in solution possess a cavity which is suitable
for the interaction with cations. The upfield shifts of the BTMA+
protons in the 1H NMR suggest that binding is due to cation-π
interactions.13 The small influence of complexation on the
chemical shifts of the guest protons may be attributed to the
receptor’s flexibility. Another reason for the small shift can also
be that the diameter of the cyclopeptide cavity is too large for
optimum interaction with BTMA+. For the unexpected high
stability of the 2-BTMA+ complex, electrostatic effects of the
amide carbonyl groups may be responsible.15 Besides the
aromatic rings, also the peptide’s carbonyl groups line the
cyclopeptide’s cavity and contribute to the overall negative
electron density of its inner surface. As a consequence, the
carbonyl groups might assist in the binding of the substrate.
On the other hand, several groups have shown that, in the case
of cation-π interactions, a tight fit of the guest inside the
receptor is not necessary for the formation of stable complexes.6d,15
Cation Binding of the Cyclopeptide in the Presence of
Tosylate Anions. We have found that cyclopeptide 2 behaves
distinctly differently toward BTMA+ cations when, instead of
iodide, tosylate is the counterion in NMR titrations. In contrast
to the iodide salt, a significant influence on the 1H NMR spectra
can be observed when BTMA+ tosylate is added to solutions
of 2 in CDCl3. Mixtures of BTMA+ tosylate and an excess of
2 possess two sets of signals in their NMR spectra: the broader
signals of the free cyclopeptide and sharper ones of a new
species composed of equimolar amounts of 2 and BTMA+
tosylate. The latter can be assigned to a complex between 2
and the tosylate anion. At equimolar concentrations of 2 and
BTMA+ tosylate, only the spectrum of the complex is visible.
A further increase of the salt concentration causes the appearance
of the free BTMA+ tosylate signals. The stability of the
2-tosylate complex is obviously so high that the exchange of
guest molecules is slow on the NMR time scale, and complexed
and uncomplexed species can be observed separately.
Figure 6. NH stretching region of the FT-IR spectra of free 2 (a) and
its complex with K+-18-crown-6 tosylate (b) (c ) 1 mM in CDCl3).
the glutamic acid NH “down” toward H(2). Indeed, only in such
alternate orientations of the two NH groups are intramolecular
hydrogen bonds possible. However, since both NH groups still
possess NOE effects to H(R), rotations around the bonds at C(R)
must still be possible, and the cyclopeptide is not completely
rigid.
The variation of the spectra of free cyclopeptide and its
tosylate complex is most noticeable in the aromatic region
(Figure 7). Complex formation between tosylate and 2 causes
An alternative explanation for the NMR signal broadening
may also be an intermolecular association of 2. We could rule
out this possibility, however, by FT-IR and NMR spectroscopic
measurements. The IR spectrum of a 1 mM solution of 2 in
CDCl3 shows two N-H stretching bands, a weaker one at 3408
cm-1 (free NH) and a more intensive band at 3285 cm-1
(hydrogen-bonded NH) (Figure 6a).12 The ratio of the areas
under both bands remains constant over a concentration range
from 2 to 0.2 mM. This, and the fact that the line broadening
(13) Complexation of BTMA+ iodide by 2 seems to be due to cation-π
interactions, but other binding mechanisms may also be responsible for the
complexation of this guest. For example, Reetz has shown that the
R-methylene groups of a tetrabutylammonium cation can interact with
carbonyl groups of enolates by hydrogen bonding.14 In a recent investigation
on the complexation of quaternary ammonium ions by uncharged macro-
cyclic “phane” esters, no interaction of the receptor’s carbonyl groups with
the guests was detected in CDCl3.15 The presented data suggest, however,
that the ester functions may possess a cooperative role in cation-π
interactions. Furthermore, one has to consider that amide carbonyls are
stronger hydrogen bond acceptors than ester carbonyls.16 FT-IR spectro-
scopic investigations on the complexes of 2 with BTMA+ iodide showed
no influence of complex formation on the vibration frequencies of the
peptide’s carbonyl groups. We take this as a further indication that the
complexation of cations by 2 is due to cation-π interactions. More
importantly, we recently were able to crystallize a complex of a quaternary
ammonium ion and a cyclic hexapeptide which contains L-proline subunits
instead of glutamic acid. We could detect no directed hydrogen bonds
between the carbonyl groups of the peptide and the ammonium ion in the
crystal structure. The properties of the L-proline-containing cyclopeptide
will be published shortly.
1
in the H NMR spectrum of 2 is not affected upon dilution of
the solution down to 0.1 mM, indicate that hydrogen bonding
occurs intramolecularly. No significant intermolecular associa-
tion could be detected in CDCl3 at concentrations relevant for
our investigations on the complexing behavior of 2.
In summary, our results indicate that 2 is flexible in DMSO,
whereas in CDCl3 its flexibility is reduced and the conforma-
tional equilibrium is shifted toward conformers with intramo-
lecular hydrogen bonds. The addition of BTMA+ iodide does
(14) Reetz, M. T.; Hu¨tte, S.; Goddard, R. J. Am. Chem. Soc. 1993, 115,
9339-9340.
1
not affect the cyclopeptide’s H NMR spectrum or the NH
(15) Roelens, S.; Torriti, R. J. Am. Chem. Soc. 1998, 120, 12443-12452.
(16) Arnett, E. M.; Mitchell, E. J.; Murty, T. S. S. R. J. Am. Chem. Soc.
1974, 96, 3875-3891. Spencer, J. N.; Garrett, R. C.; Mayer, F. J.; Merkle,
J. E.; Powell, C. R.; Tran, M. T.; Berger, S. K. Can. J. Chem. 1980, 58,
1372-1375. Gellman, S. H.; Dado, G. P.; Liang, G.; Adams, B. R. J. Am.
Chem. Soc. 1991, 113, 1164-1173.
vibration bands in its FT-IR spectrum, which would indicate
(12) Vinogradov, S. N.; Ninnell, R. H. Hydrogen Bonding; Van Nostrand
Reinhold Co.: New York, 1971. Ne´el, J. Pure Appl. Chem. 1972, 31, 201-
225.

5850 J. Am. Chem. Soc., Vol. 121, No. 25, 1999
Kubik
Figure 8. Saturation curves of NMR dilution titrations of the BTMA+
2-tosylate complex (a) and of the K+-18-crown-6 2‚tosylate complex
(b).
Figure 7. 1H NMR spectra of 2 in CDCl3 (a) and after addition of
0.25 (b) and 1.0 equiv (c) of K+-18-crown-6 tosylate.
downfield shifts of the signals of both NH groups by ca. 0.25
ppm, but also all the signals of the aromatic ring protons of 2
are shifted in the same direction. The coupling constant
3JNH-C(R)H in the 2-tosylate complex amounts to 9.5 Hz, which
is significantly larger than the one of free 2 in DMSO-d6
(3JNH-C(R)H ) 7.9 Hz). From the coupling constant of the tosylate
complex, a dihedral angle θ of (158° can be calculated, whereas
3JNH-C(R)H of the free cyclopeptide corresponds to the angles
(24° and (143°.17 It has to be considered, however, that
3JNH-C(R)H of free 2 represents only a weighted averaged value
of the coupling constants of all conformers present in solution.
Besides complexation of the tosylate, also that of the BTMA+
cation can be observed in the NMR spectra from the charac-
teristic upfield shift of the corresponding protons. In the
spectrum of an equimolar mixture of 2 and BTMA+ tosylate (1
mM in CDCl3), the resonance of, e.g., the N-methyl protons of
BTMA+ is shifted by -0.34 ppm in comparison to that of the
free salt. This shift is significantly larger than the one which is
observed when 2 binds BTMA+ iodide (-0.02 ppm). When
more than 1 equiv of BTMA+ tosylate is added to 2, only
averaged signals for the cations are observed. This fast guest
exchange makes possible the NMR spectroscopic determination
of the Ka of the BTMA+ 2-tosylate complex by dilution
titration. No dissociation of 2-tosylate could be detected in the
Figure 9. Intra- and intermolecular NOE effects in the NOESY NMR
spectrum of the K+-18-crown-6 2-tosylate complex in CDCl3.
of the receptor. It is, therefore, reasonable to assume that the
cyclopeptide conformation in the 2-tosylate complex is much
better preorganized for the complexation of cations than that
of the more flexible free macrocycle. The preorganization is
furthermore reflected in the increase of ∆δmax. Preorganization
of 2 causes a better interaction of the guests with the aromatic
subunits and accordingly a larger value for ∆δmax
.
Again, NOESY NMR spectroscopy gives valuable informa-
tion about the structure of the 2-tosylate complex in solution.
We used potassium tosylate instead of the BTMA+ salt in these
investigations. The K+ salt is soluble in CDCl3 when equimolar
amounts of 18-crown-6 are added to the solution to bind the
cations. The change of the cations does not seem to influence
the complex conformation of 2-tosylate because, apart from
the cation signals, the spectra of BTMA+ 2-tosylate and K+-
18-crown-6 2-tosylate are identical.
The NOESY spectrum of the K+-18-crown-6 2-tosylate
complex shows cross-peaks between H(2) and both NH groups
as well as between the two NH groups themselves (Figure 9).
These effects show that 2 adopts a C3-symmetrical conformation
in the tosylate complex, with all NH groups as well as the side
chains of the glutamic acids pointing “down” in the same
direction as the H(2) of the aromatic rings. Clearly, the influence
of the tosylate ion causes orientations of the NH groups and
the amino acid side chains which are distinctly different from
the ones in free 2. Intermolecular NOE effects are visible
between the aromatic ring of the tosylate and the glutamic acid
side chains (Figure 9). Therefore, the aromatic ring of the
tosylate must be arranged parallel to these residues.
The downfield shift of the NH groups in the 1H NMR
spectrum of the complex suggests that these groups interact with
the tosylate anion by hydrogen bonding. This type of interaction
can also be derived from the FT-IR spectrum. In this spectrum,
two N-H vibration bands are visible which both possess the
characteristic frequencies and intensities of hydrogen-bonded
NH groups (Figure 6b).12 A band for free NH groups as in
Figure 6a cannot be observed. Therefore, all six NH groups of
2 must be involved in the binding of the tosylate anion. The
1H NMR spectra upon dilution of a solution up to 10-6 mol‚L-1
.
Therefore, the Ka of this complex is at least in the order of 109
M-1, and 2-tosylate can be safely regarded as a stable
compound in the investigated concentration region. On the other
hand, dilution causes the dissociation of the complex between
the BTMA+ cation and the 2-tosylate anion, as indicated by a
downfield shift of the cation protons. When the variation of
this shift is plotted against the total concentration of the complex,
Ka can be calculated from the resulting saturation curve (Figure
8a).7,9 An association constant of 3.88 × 106 M-1 has thus been
determined, which shows that the change from an iodide to a
tosylate salt causes an increased BTMA+ complex stability by
a factor of ca. 104. The tosylate anion behaves obviously highly
cooperatively in the complexation of cations by 2. The 1H NMR
spectrum of the 2-tosylate complex clearly shows that complex
formation is accompanied by a conformational reorganization
(17) Bystrov, V. F. Prog. Nucl. Magn. Reson. Spectrosc. 1976, 10, 41-
81.

Receptor Properties of a Cyclopeptide
J. Am. Chem. Soc., Vol. 121, No. 25, 1999 5851
Table 1. Association Constants Ka and Maximum Chemical Shifts
∆δmax of Complexes of Preorganized and Non-preorganized 2 with
Different Cations in CDCl3 at 298 K
∆δmax
(ppm)
anion
cation
Ka (M-1
)
iodide
K+-18-crown-6
K+-18-crown-6
Na+-15-crown-5
Li+-12-crown-4
BTMA+
nd
<0.01
0.17
0.07
<0.01
0.05
tosylate
tosylate
tosylate
iodide
8.09 × 105 ( 3.91 × 104
2.13 × 105 ( 2.07 × 104
nd
3.00 × 102 ( 0.40 × 102
tosylate
BTMA+
3.88 × 106 ( 1.95 × 105
0.37
Ka(BTMA+)
Ka(K+-18-crown-6)
Ka(BTMA+)
Ka(BTMA+) ) 1.70 × 105 0.49
) 0.210
tosylate
tosylate
K (BTMA+) ) 1.65 × 105 0.28
) 0.773
a
Ka(Na+-15-crown-5)
1.68 ×105
0.39
In addition, negative electron density of the aromatic cyclo-
peptide subunits and of the peptide’s carbonyl groups is
distributed around the rim of the cavity. In this region of the
receptor, binding should be due to cation-π interactions. The
hydrogen bonding of the tosylate anions to the NH groups of
the cyclopeptide may even increase the electron density of the
aromatic rings and thus provide stronger cation-π interaction.
This assumption is consistent with theoretical calculations, which
showed that cation-π interactions between phenol and Na+ are
considerably enhanced when the phenolic OH is hydrogen-
bonded to a formamide,10 and with experimental results on an
artificial receptor containing phenol subunits whose affinity
toward cations also increases when anions are hydrogen-bonded
at the OH groups.19 Hydrogen bonding of the guest cations to
the carbonyl groups of 2-tosylate is unlikely. Figure 10 shows
that all carbonyl groups of the receptor diverge from the center
of the cavity and should not be able to cooperatively form
defined 1:1 complexes with bound guest molecules.
Figure 10. Energy-minimized structure of the 2-tosylate complex (for
reasons of clarity, the glutamic acid side chains have been replaced by
methyl groups).
fact that two NH vibration bands are present indicates that, in
the complex, two types of hydrogen bonds between anion and
NH groups exist. Most probably, the six hydrogen bonds can
be subdivided into three bonds between the anion and the
glutamic acid NH and three between the anion and the aromatic
NH. All of these hydrogen bonds are weaker than the intramo-
lecular H-bonds in free 2, as indicated by the higher NH
absorption frequencies. A possible explanation for this observa-
tion may be that, in the 2-tosylate complex, binding of three
anion oxygens is distributed over six NH groups, and the
corresponding interactions may therefore not have an optimal
geometry. In contrast, each NH has one partner for a hydrogen
bond in free 2.
The sum of this structural information was used to calculate
the structure of the 2-tosylate complex by molecular model-
ing.18 The energy-minimized conformation is depicted in Figure
10. In this structure, all six NH groups of 2 bind the three
tosylate oxygens simultaneously, as predicted by IR spectro-
scopy. It has to be considered, however, that Figure 10 represents
only a static symmetrical structure of the complex. Although
the exchange of bound anions in solution is slow on the NMR
time scale, a rotation of the guests in the complex may certainly
be possible, and as a result, different orientations of the anion
inside the cyclopeptide rapidly interconvert. The angle θ in the
modeled cyclopeptide conformation amounts to -173°, which
is in acceptable agreement with the one calculated from the
observed coupling constant 3JNH-C(R)H ((158°). The difference
between calculated and observed values may be attributed to
dynamical processes in the complex. The almost coplanar
orientations of the carbonyl groups to the adjacent aromatic rings
account for the downfield shift of the aromatic H(4) and H(6)
upon complexation of tosylate by 2. The downfield shift of H(2)
can be attributed to the electric field of the anion which is bound
in close proximity.
In our NMR experiments with the K+-18-crown-6 2-tosyl-
ate complex, we found a considerable upfield shift of the crown
ether protons by -0.16 ppm in comparison to the resonance of
free K+-18-crown-6 tosylate (1 mM in CDCl3). Obviously, also
this cation, as possibly many others, can interact with the
cyclopeptide. In this respect, the quaternary ammonium ions
or the K+-18-crown-6 only possess the advantage that their
complexation can be easily followed by NMR spectroscopy.
The interaction of K+-18-crown-6 with 2-tosylate can be
rationalized by assuming that the whole cationic crown ether
complex is embedded in the 2-tosylate cavity, where it is
attracted by the tosylate anions. This inclusion brings the crown
ether protons in close proximity to the aromatic walls of the
cyclopeptide, where they experience the shielding of their
protons. No dissociation of the K+-18-crown-6 complex has
to occur for this type of complex formation. In fact, the K+-
18-crown-6 complex is so stable in CDCl3 (Ka ) 1.6 × 108
M-1 20
that, under the chosen conditions, its complexation by
)
2-tosylate can almost be regarded as an interaction between
two stable compounds. Consequently, it was possible to
determine the stability of this complex by dilution titration
analogously to the BTMA+ complex (Figure 8b). By choosing
other alkali cation crown ethers pairs, e.g., sodium tosylate and
15-crown-5, we could furthermore study the influence of the
size of the guest on the Ka of the complexes systematically
(Table 1). In all dilution titrations, the dissociation of the cation-
crown ether complexes at low concentrations was taken into
Figure 10 demonstrates that 2-tosylate should, indeed, be
ideally suited for the complexation of cations. The complex has
the shape of a shallow dish with the anion located at its bottom.
From there, the anion can attract cationic guests into the cavity.
(18) Molecular modeling was performed on a Silicon Graphics worksta-
tion using the program Cerius2 (Molecular Simulations Inc.) with a Dreiding
2.21 force field (Mayo, S. L.; Olafson, B. D.; Goddard, W. A., III. J. Phys.
Chem. 1990, 94, 8897-8909).
(19) Jeong, K.; Hahn, K.; Cho, Y. Tetrahedron Lett. 1998, 39, 3779-
3782.
(20) Iimori, T.; Still, W. C. J. Am. Chem. Soc. 1989, 111, 3439-3440.

5852 J. Am. Chem. Soc., Vol. 121, No. 25, 1999
Kubik
complex is now less stable than the cation-crown ether
complexes. In our eyes, this indicates that the Ka of the BTMA+
2-tosylate complex has been overestimated in the dilution
titration. Under the conditions of this method, a dissociation of
the complex is accompanied by an ion pair separation. Such a
separation is not so problematic when the cations are stabilized
by interactions with, e.g., crown ethers but is especially
unfavorable when they are not sufficiently solubilized.22 There-
fore, the Ka which resulted from the competitive titrations seems
to be a more realistic value for the stability of the BTMA+
2-tosylate complex.
It appears as if the high stability of the cation 2-tosylate
complexes is mainly due to strong electrostatic interactions
between the ionic guests. The Ka of the BTMA+ 2-tosylate
complex (1.68 × 105 M-1), for example, corresponds to a free
energy of complex formation ∆G of -29.6 kJ‚mol-1. But
Figure 11. Saturation curves of a competitive NMR titration of the
K+-18-crown-6 2-tosylate complex and BTMA+ iodide.
already the complex of 2 with BTMA+ iodide (Ka ) 300 M-1
)
possesses a considerable binding energy of ∆G ) -14.0
kJ‚mol-1. Cation-π interactions are certainly weaker than
electrostatic interactions between two permanent charges.5b Still,
the cooperativity of the subunits in 2 significantly contributes
to the overall stabilities of the resulting complexes.
account by including the corresponding stability constants20 in
the calculations. With Li+ and 12-crown-4, the upfield shift upon
complexation by 2-tosylate was not large enough to be followed
quantitatively. Electrostatic interactions between Li+-12-
crown-4 and 2-tosylate certainly also cause a complexation of
this cation. But compared to the cyclopeptide cavity, the Li+-
12-crown-4 complex is probably too small for a significant
influence of the aromatic systems on the resonance of the crown
ether protons. Table 1 furthermore shows that a complexation
of K+-18-crown-6 by 2 can be also detected in the absence of
effector anions, as indicated by the small upfield shift of the
crown ether protons. Because of the small variation in the
chemical shifts (<0.01 ppm), a quantitative determination of
the Ka of this complex was difficult.
Cation Binding of the Cyclopeptide in the Presence of
Phosphonate Anions. The complexation of phosphate by the
cyclic peptide 1 has been reported.4 Analogously to the
2-tosylate complex, the binding of dipotassium 4-nitrophenyl
phosphate by 1 in DMSO-d6 causes a change of the cyclopep-
1
tide’s H NMR spectrum. Besides significant downfield shifts
of the two NH signals, which were also interpreted in terms of
hydrogen bonding of the phosphate anion to 1, the aromatic
protons shift similarly to the ones in the spectrum of the
3
2-tosylate complex, and the coupling constants JNH-C(R)H of
Knowing the cation-crown ether 2-tosylate stability con-
stants, we were able to determine the Ka of the corresponding
BTMA+ complex by an alternative method, namely by competi-
tive NMR titrations.21 For this, increasing amounts of BTMA+
iodide were added to solutions of, e.g., K+-18-crown-6
2-tosylate in CDCl3. Under these conditions, complexes of
2-tosylate with BTMA+ and the cation-crown ether are formed
simultaneously in relative amounts that depend on the ratio of
their association constants. The relative concentrations of both
complexes in solution are determined by following the upfield
shift of the protons of both guests NMR spectroscopically. From
the resulting pair of saturation curves (Figure 11), the Ka ratio
of the complexes can be calculated and, by multiplying this
ratio with the previously determined Ka of the corresponding
cation-crown ether complex, a new association constant for
the BTMA+ complex is obtained. Such a titration has been
performed with K+-18-crown-6 as well as Na+-15-crown-5.
Table 1 shows that the association constants for the BTMA+
complexes from the two experiments are in excellent agreement
but more than 1 order of magnitude smaller than the Ka obtained
in the dilution titration.
both complexes even have the same value. Therefore, the side
chains of the natural amino acid subunits do not seem to have
a significant effect on the complexation of anions by 1 or 2,
and the cyclopeptide conformation of the phosphate and tosylate
complexes should be comparable. Consequently, also phosphate
anions might act as effectors for the complexation of cations
by 2.
To study the influence of the type of effector anion on the
cation complex stabilities of 2, we tested phosphate and related
anions in our investigations, too. The monoesters of phosphate
were only poorly soluble in CDCl3, so we were unable to
perform investigations with these anions. Instead, we could use
dipotassium phenylphosphonate which dissolved in CDCl3 when
2 equiv of 18-crown-6 was added to the solution for K+
complexation. In the 1H NMR spectrum of an equimolar mixture
of (K+-18-crown-6)2 phenylphosphonate and 2 in 10% DMSO-
(22) In nonpolar solvents, salts are known to exist as close ion pairs
even at low concentrations.23 In a series of papers, Wilcox et al. have shown
that, in nonpolar solvents, the rate of various reactions is significantly
enhanced when a charged group is attached covalently close to the reactive
site of a substrate.24 This rate enhancement was attributed mainly to
influences of the local electric field of the ion pair localized in close
proximity to the substrate’s reaction center. The effect was called the
intramolecular salt effect and is strongest in chloroform, because in this
solvent the most specific and most stable associations have been observed.
These results indicate that cation-anion separation in chloroform is, indeed,
unfavorable and can only be achieved, e.g., with ditopic receptors when
both ions are stabilized by suitable binding sites.3j Therefore, the assumption
that the dissociation of the BTMA+ 2-tosylate complex is less likely than
the one of K+-18-crown-6 2-tosylate seems to be reasonable.
(23) Davies, C. W. Ion Association; Butterworth: London, 1962.
(24) Smith, P. J.; Wilcox, C. S. J. Org. Chem. 1990, 55, 5675-5678.
Smith, P. J.; Wilcox, C. S. Tetrahedron 1991, 47, 2617-2628. Smith, P.
J.; Soose, D. J.; Wilcox, C. S. J. Am. Chem. Soc. 1991, 113, 7412-7414.
Smith, P. J.; Reddington, M. V.; Wilcox, C. S. Tetrahedron Lett. 1992, 33,
6085-6088.
The deviation of the two association constants can, to a certain
extent, be attributed to the different experimental conditions of
dilution and competitive NMR titrations. A decrease of the Ka
in competitive titrations could be expected because, due to the
higher salt concentration and, hence, the increased ionic
strengths of the solutions in these measurements, the electrostatic
interaction between receptor and substrate should be weakened.
However, the decrease is so pronounced that the BTMA+
(21) Martin, M.; Raposo, C.; Almaraz, M.; Crego, M.; Caballero, C.;
Grande, M.; Mora´n, J. R. Angew. Chem., Int. Ed. Engl. 1996, 35, 2386-
2388. Almaraz, M.; Martin, M.; Herna´ndez, J. V.; Caballero, M. C.; Mora´n,
J. R. Tetrahedron Lett. 1998, 39, 1811-1814.

Receptor Properties of a Cyclopeptide
J. Am. Chem. Soc., Vol. 121, No. 25, 1999 5853
The increased stability of the cation complexes of 2-phos-
phonate in comparison to that of 2-tosylate complexes is
reasonable and can be attributed to the higher charge of the
anion. Yet its 2-fold charge also has an influence on the complex
structure. In solution, 2-phosphonate interacts with two cations,
but, because of their size, most probably only inclusion of one
into the cyclopeptide cavity is possible. Such an equilibrium
complicates quantitative investigations on complexes of 2-phos-
phonate. Investigations in this respect are better carried out with
monoanions, e.g., tosylates. We have shown that their complex
equilibria can be analyzed in a straightforward manner by using
1:1 models.
Conclusion
In conclusion, we demonstrated the unusual allosteric receptor
properties of 2. This cyclopeptide is able to bind cations and
anions by two different binding mechanisms. Independent of
the type of the anion, cations can generally be bound by
cation-π interactions. Yet certain anions such as sulfonates or
phosphonates significantly increase the cyclopeptide affinity
toward positively charged substrates. These anions are hydrogen-
bonded to a second binding site of 2, where they induce a
preorganization of the receptor and are, furthermore, able to
interact with the cationic substrates by electrostatic interactions.
In contrast to many other artificial allosteric systems, the positive
cooperativity of the effectors is not only an effect of a
conformational reorganization of the receptor but also due to
the proximity of the substrate cations and the effector anions
in the final complex. The next step now should be to study
whether such allosteric binding mechanisms could be transferred
to more polar solvents. For this, we are synthesizing water-
soluble cyclopeptides with the aim to investigate whether, e.g.,
phosphate or sulfate ions act as effectors also in this solvent.
Furthermore, by introducing additional functional groups into
the aromatic subunits of the cyclopeptides, a variety of new
receptors for other substrates should be accessible. It should be
possible to regulate their receptor properties by taking advantage
of the allosteric effects of suitable anions.
Figure 12. 1H NMR spectra of (K+-18-crown-6)2 phenylphosphonate
(a), 2 (c), and the (K+-18-crown-6)2 2-phenylphosphonate complex
(b) in 10% DMSO-d6/CDCl3.
d6/CDCl3, the usual effects of anion complexation on the
cyclopeptide signals and of cation complexation on the crown
ether protons are visible (Figure 12). The most prominent effect
of phosphonate complexation is the pronounced downfield shift
of the NH protons by more than 3 ppm. This shift is significantly
larger than the one which resulted from the complexation of
tosylate by 2 (ca. 0.25 ppm) but is in the same order of
magnitude as the shift of the 1-phosphate complex.4 Obviously,
phosphates and phosphonates interact much more strongly with
the cyclopeptides than tosylates. This can be also demonstrated
by FT-IR spectroscopy. No defined NH stretching band is
observed in the spectrum of the 2-phosphonate complex in
CDCl3 above 3200 cm-1. Instead, only a very broad band is
visible between 2600 and 3400 cm-1 which overlaps with other
bands in this region. The absorption frequency and width of
this band account for the strong hydrogen bonding of the
phosphonate oxygens to the cyclopeptide’s NH groups. The
stronger interaction of 2 with phenylphosphonate in comparison
to tosylate can be attributed to the significantly larger basicity
of the phosphonate dianion (pKa2(phenylphosphonate) ) 7.07;25
pKa(tosylate) ) -1.326), with which a higher affinity of the
anion toward the NH protons is associated.
Experimental Section
General Methods. Analyses were carried out as follow: melting
points, Bu¨chi 510; optical rotation, Perkin-Elmer 241 MC digital
polarimeter (d ) 10 cm); NMR, Varian VXR 300, Bruker DRX 500
equipped with an automatic sampler (the chemical shifts in the NMR
titrations were calculated by using the CHCl3 signal at δ ) 7.27 ppm
as internal standard); FT-IR, Bruker Vector 22 FT-IR, cuvette d ) 5
mm (NaCl); elemental analysis, Pharmaceutical Institute of the Hein-
rich-Heine-University, Du¨sseldorf; mass spectrometry, Finnigan INCOS
50; chromatography, ICN silica gel 32-63 (ICN Biomedicals), MERCK
LiChroprep RP-8 (40-63 mm) prepacked column size B (310-25). The
following abbreviations are used: TsOH, toluene-4-sulfonic acid; Bn,
benzyl; BOC, tert-butoxy carbonyl; DIEA, N-ethyldiisopropylamine;
TFA, trifluoroacetic acid; PyCloP, chlorotripyrrolidinophosphonium
hexafluorophosphate; TBTU, O-(1H-benzotriazol-1-yl)-N,N,N′,N′-tet-
ramethyluronium tetrafluoroborate; Glu, glutamic acid; AB, 3-ami-
nobenzoic acid.
In the 1H NMR spectrum of the (K+-18-crown-6)2 2-phos-
phonate complex, an upfield shift of the crown ether protons
of -0.22 ppm was observed (1 mM in 10% DMSO-d6/CDCl3),
which accounts for the interaction of the cations with the
cyclopeptide moiety. However, we were unable to determine
the Ka of the corresponding complex quantitatively. It was not
possible to detect a dissociation of the complex NMR spectro-
scopically upon dilution of the solution. Obviously, the Ka of
the (K+-18-crown-6)2 2-phosphonate complex is even larger
than the one of the corresponding 2-tosylate complex and
cannot be determined by this method. Alternatively, we used
competitive NMR titrations because it can be expected that,
under the conditions of this method, the increased ionic strength
of the solutions should reduce the Ka. But the saturation curves
we obtained could not be fitted on the basis of a simple 1:1
complex formation.
Materials. All solvents were dried according to standard procedures
prior to use. DMF p.A. was purchased from Fluka and was used without
further purification. PyCloP27 and the K+-18-crown-6 4-toluene-
sulfonate salt28 were prepared according to the literature procedures.
TBTU was purchased from BACHEM.
Job Plot. Equimolar solutions (1 mM) of n-butyltrimethylammonium
iodide and cyclopeptide 2 in 1% DMSO-d6/CDCl3 were prepared and
(25) Jaffe´, H. H.; Freedman, L. D.; Doak, G. O. J. Am. Chem. Soc. 1953,
75, 2209-2211.
(26) Berkowitz, B. J.; Grunwald, E. J. Am. Chem. Soc. 1963, 83, 2956.
(27) Coste, J.; Fre´rot, E.; Jouin, P. J. Org. Chem. 1994, 59, 2437-2446.
(28) Dale, J.; Kristianssen, P. O. J. Chem. Soc., Chem. Commun. 1971,
670-671.

5854 J. Am. Chem. Soc., Vol. 121, No. 25, 1999
Kubik
mixed in various ratios. 1H NMR spectra of the solutions were recorded,
and the change in chemical shift of guest protons was analyzed.
Ka by Host-Guest Titrations. Stock solutions of the guest (0.1
µmol/200 µL) in 1% DMSO-d6/CDCl3 and of cyclopeptide 2 (2 µmol/
800 µL) in CDCl3 were prepared. Altogether, 11 NMR tubes were set
up by adding increasing amounts of the host solution (0-800 µL) to
200 µL of the guest solution. All samples were made up to 1 mL with
CDCl3, and their 1H NMR spectra were recorded. The chemical shifts
of prominent guest protons were plotted against the host concentration.
From the resulting saturation curves, Ka and ∆δmax were calculated by
a nonlinear least-squares fitting method for 1:1 complexes7,9 using the
SIGMA Plot 3.0 (Jandel Scientific) software package.
BOC-L-Glu(OiPr)-OH.30 L-Glutamic acid 5-isopropyl ester31 (7.56
g, 40 mmol) is suspended in a mixture of 1,4-dioxane (100 mL) and
water (60 mL). The resulting mixture is cooled with an ice bath, and
the pH is adjusted to 9-10 by addition of triethylamine (5.56 mL, 40
mmol). Afterward, di-tert-butyl dicarbonate (10.03 g, 46 mmol) is
added, and stirring is continued at room temperature for 4 h. During
this time, the starting material slowly dissolves. The dioxane is removed
in vacuo, and the remaining aqueous layer is extracted three times with
diethyl ether. The organic layers are discarded. The aqueous layer is
acidified to pH 3-4 by addition of aqueous KHSO4 and extracted with
ethyl acetate three times. The combined organic layers are washed with
water and dried, and the solvent is evaporated in vacuo. The product
slowly crystallizes in the refrigerator: yield 11.3 g (98%); mp 64-67
Ka by Dilution of Complex Solution. A stock solution containing
1 mmol/mL cyclopeptide 2, an equivalent amount of alkali 4-toluene-
sulfonate, and the corresponding crown ether in 0.1% DMSO-d6/CDCl3
was prepared. Altogether, 13 samples were prepared from this solution
by dilution with 0.1% DMSO-d6/CDCl3 up to a concentration of 0.001
°C; [R]25 ) -10.2 (c ) 2, methanol).
D
Dipeptide BOC-L-Glu(OiPr)-AB-OBn. 3-Aminobenzoic acid ben-
zyl ester toluene-4-sulfonate (2.39 g, 6.00 mmol) and BOC-L-Glu(OiPr)-
OH (2.43 g, 8.40 mmol) as well as PyCloP (3.54 g, 8.40 mmol) are
dissolved in CH2Cl2 (120 mL). At room temperature, DIEA (3.96 mL,
22.8 mmol) is added dropwise, and the reaction mixture is stirred
overnight. Afterward, the solvent is evaporated in vacuo, and the product
is isolated chromatographically (ethyl acetate/hexane 1:2) from the
residue. The product is triturated with petroleum ether 60/80 to afford
1
mmol/mL. H NMR spectra were recorded, and the chemical shift of
the crown ether protons was plotted against the concentration. From
the resulting curve, Ka and ∆δmax were calculated analogously to the
host-guest titration.
Ka by Competitive NMR Titrations. Stock solutions of n-
butyltrimethylammonium iodide (0.4 mmol/200 mL) in 1% DMSO-
d6/CDCl3 and of cyclopeptide 2 with an equivalent amount of alkali
4-toluenesulfonate and the corresponding crown ether (1 mmol/800 mL)
in CDCl3 were prepared. Altogether, 11 NMR tubes were set up by
adding increasing amounts of the cyclopeptide solution (0-800 mL)
to 200 mL of the n-butyltrimethylammonium iodide solution. All
samples were made up to 1 mL with CDCl3, and their 1H NMR spectra
were recorded. The chemical shifts of the protons of the trimethylam-
monium group and of the crown ether protons were plotted against the
host concentration. From the two resulting saturation curves, the ratio
of the association constants of the cation complexes and ∆δmax of the
ammonium ion complex were calculated by a nonlinear least-squares
fitting method based on eq 121 using SIGMA Plot 3.0. For the other
a white solid: yield 2.60 g (87%); mp. 97-98 °C; [R]25 ) -14.9 (c
D
1
) 2, methanol); H NMR (300 MHz, DMSO-d6, 25 °C, TMS) δ 1.15
3
+ 1.17 (2d, J(H,H) ) 6.3 Hz, 6H; iPrCH3), 1.39 (s, 9H; tBuCH3),
1.87 + 1.96 (2m, 2H; GluC(â)H2), 2.35 (m, 2H; GluC(γ)H2), 4.10 (m,
3
1H; GluC(R)H), 4.87 (sept, J(H,H) ) 6.3 Hz, 1H; iPrCH), 5.37 (s,
2H; PhCH2), 7.14 (d, 3J(H,H) ) 7.7 Hz, 1H; GluNH), 7.43 (m, b, 6H;
3
4
PhH + ABH(5)), 7.70 (dt, J(H,H) ) 8.0 Hz, J(H,H) ) 1.2 Hz, 1H;
ABH(6)), 7.94 (d, 3J(H,H) ) 8.0 Hz, 1H; ABH(4)), 8.28 (t, 4J(H,H) )
1.4 Hz, 1H; ABH(2)), 10.26 (s, b, 1H; ABNH). Anal. Calcd for
C27H34N2O7 (498.6): C, 65.04; H, 6.87; N, 5.62, Found: C, 64.99; H,
6.71; N, 5.55.
Tetrapeptide BOC-[L-Glu(OiPr)-AB]2-OBn. BOC-L-Glu(OiPr)-
AB-OBn (0.85 g, 1.70 mmol) is deprotected at the terminal amino group
according to the general method. An equivalent amount of the same
dipeptide is hydrogenated. Both components as well as PyCloP (1.2
equiv, 2.04 mmol, 0.86 g) are dissolved in CH2Cl2 (20 mL/mmol). At
room temperature, DIEA (3.4 equiv, 5.78 mmol, 1.00 mL) is added
dropwise, and the reaction mixture is stirred overnight. Afterward, the
solvent is evaporated in vacuo, and the product is isolated chromato-
graphically (ethyl acetate/hexane 1:1) from the residue. The product is
δ
max (alkali ion-crown ether), the value which resulted from the dilution
titration was used.
Ka(1) [δ(1) - δ0(1)] [δmax(2) - δ(2)]
)
(1)
Ka(2) [δ(2) - δ0(2)] [δmax(1) - δ(1)]
General Procedure for Hydrogenation of Benzyl Esters. The ester
is dissolved in methanol (50 mL/mmol). After addition of 10% Pd/C
(100 mg), the resulting reaction mixture is hydrogenated at 1 atm for
about 2 h. Completeness of reaction is checked by TLC. The catalyst
is filtered off through a layer of Celite and washed with methanol.
The solution is evaporated to dryness in vacuo.
triturated with hexane to afford a white solid: yield 1.17 g (88%); mp
1
67-70 °C; [R]25 ) -7.2 (c ) 2, methanol); H NMR (300 MHz,
D
CDCl3, 25 °C, TMS) δ 1.16 (m, 12H; iPrCH3), 1.39 (s, 9H; tBuCH3),
1.84-2.15 (m, b, 4H; GluC(â)H2), 2.38 (m, 4H; GluC(γ)H2), 4.11 (m,
1H; Glu1C(R)H), 4.59 (m, 1H; Glu3C(R)H), 4.85 (sept, 3J(H,H) ) 6.2
Hz, 2H; iPrCH), 5.37 (s, 2H; PhCH2), 7.12 (d, 3J(H,H) ) 7.9 Hz, 1H;
General Procedure for Cleavage of N-tert-Butoxycarbonyl Groups.
The carbamate is dissolved in CH2Cl2 (5 mL). The resulting solution
is cooled with an ice bath, and trifluoroacetic acid (5 mL) is added
dropwise. The reaction mixture is stirred for 1.5 h at 0-5 °C. Afterward,
the solvent is evaporated in vacuo. The residue is dissolved in ethyl
acetate, and the solution is extracted twice with 10% Na2CO3 solution
and three times with water. The organic layer is dried, 1 N HCl (1
mL/mmol) is added, and the mixture is evaporated to dryness in vacuo.
3
Glu1NH), 7.40 (m, b, 7H; PhH + 2ABH(5)), 7.64 (d, J(H,H) ) 7.7
Hz, 1H; AB2H(6)), 7.71 (dt, 3J(H,H) ) 8.0 Hz, 4J(H,H) ) 1.2 Hz, 1H;
AB4H(6)), 7.84 (d, 3J(H,H) ) 8.3 Hz, 1H; AB2H(4)), 7.96 (2dd, 3J(H,H)
4
) 8.2 Hz, J(H,H) ) 1.0 Hz, 1H; AB4H(4)), 8.05 (s, 1H; AB2H(2)),
4
3
8.31 (t, J(H,H) ) 1.8 Hz, 1H; AB4H(2)), 8.66 (d, J(H,H) ) 7.4 Hz,
1H; Glu3NH), 10.15 + 10.39 (2s, 2H; ABNH). Anal. Calcd for
C42H52N4O11 (788.9): C, 63.95; H, 6.64; N, 7.10. Found: C, 63.85; H,
6.55; N, 7.02.
3-Aminobenzoic Acid Benzyl Ester Toluene-4-sulfonate.29
A
Hexapeptide BOC-[L-Glu(OiPr)-AB]3-OBn. BOC-L-Glu(OiPr)-
AB-OBn (0.70 g, 1.40 mmol) is deprotected at the terminal amino group
according to the general method. An equivalent amount of BOC-[L-
Glu(OiPr)-AB]2-OBn (1.10 g, 1.40 mmol) is hydrogenated. Both
components as well as TBTU (1.1 equiv, 1.54 mmol, 0.49 g) are
dissolved in DMF (30 mL/mmol). At room temperature, DIEA (3.2
equiv, 4.48 mmol, 0.78 mL) is added dropwise, and stirring is continued
for 2 h. Afterward, the reaction mixture is poured into water (150 mL/
mmol). The pH is adjusted to ca. 4 with 1 N HCl, and the suspension
is stirred for another 10 min. The precipitate is filtered off, washed
with water, and dried. According to TLC, the product is usually obtained
mixture of 3-aminobenzoic acid (6.85 g, 50 mmol), dry toluene-4-
sulfonic acid (8.61 g, 50 mmol), and toluene-4-sulfonyl chloride (11.44
g, 60 mmol) in benzyl alcohol (100 mL) is heated to 80 °C for 4 h
under stirring. During this time, the benzoic acid completely dissolves.
The product is precipitated by pouring the hot reaction mixture into
diethyl ether (800 mL) under stirring. The suspension is kept in the
refrigerator overnight. Afterward, the crude product is filtered off,
washed with diethyl ether, and recrystallized from 2-propanol/diethyl
ether: yield 11.8 g (59%); mp 176-178 °C; 1H NMR (300 MHz,
methanol-d4, 25 °C, TMS) δ 2.30 (s, 3H; TsCH3), 5.04 (s, b, ∼3H;
NH), 5.36 (s, 2H; PhCH2), 7.15 (d, 3J(H,H) ) 8.5 Hz, 2H; TsH), 7.37
(m, 5H; PhH + ABH), 7.65 (m, 4H; PhH + ABH + TsH), 8.11 (m,
2H; ABH).
(30) El Marini, A.; Roumestant, M. L.; Viallefont, P.; Razafindramboa,
D.; Bonato, M.; Follet, M. Synthesis 1992, 1104-1108.
(31) Albert, R.; Danklmaier, J.; Ho¨nig, H.; Kandolf, H. Synthesis 1987,
635-636.
(29) Arai, I.; Muramatsu, I. J. Org. Chem. 1983, 48, 121-123.

Receptor Properties of a Cyclopeptide
J. Am. Chem. Soc., Vol. 121, No. 25, 1999 5855
in high purity by this procedure and can be used for the following step
without further purification. A small amount was purified chromato-
and subjected to a column conditioned with MeOH/H2O 1:1. Gradually
the eluent is changed to MeOH/H2O 2.5:1 and then to MeOH/H2O 5:1,
with which pure product is eluted. The product is finally recrystallized
graphically (methanol/CH2Cl2 1:15) for characterization: yield 1.31 g
1
(88%); mp 185-189 °C; [R]25 ) +19.5 (c ) 2, DMF); H NMR
from methanol/ethanol 1:2: yield 0.45 g (48%); mp 139-142 °C; [R]25
D
D
) +9.6 (c ) 2, DMF); 1H NMR (300 MHz, DMSO-d6, 25 °C, TMS)
(300 MHz, DMSO-d6, 25 °C, TMS) δ 1.15 (m, 18H; iPrCH3), 1.38 (s,
9H; tBuCH3), 1.80 + 2.20 (m, b, 6H; GluC(â)H2), 2.40 (m, 6H; GluC-
(γ)H2), 4.08 (m, 1H; Glu1C(R)H), 4.58 (m, 2H; Glu3+5C(R)H), 4.86
(m, 3H; iPrCH), 5.36 (s, 2H; PhCH2), 7.11 (d, 3J(H,H) ) 7.7 Hz, 1H;
3
δ 1.16 (d, J(H,H) ) 6.6 Hz, 18H; iPrCH3), 2.08 + 2.18 (2m, 6H;
GluC(â)H2), 2.41 (m, 6H; GluC(γ)H2), 4.60 (m, 3H; GluC(R)H), 4.88
3
3
(sept, J(H,H) ) 6.3 Hz, 3H; iPrCH), 7.43 (t, J(H,H) ) 7.9 Hz, 3H;
3
Glu1NH), 7.46 (m, b, 8H; PhH + 3ABH(5)), 7.63 (d, J(H,H) ) 8.2
3
3
ABH(5)), 7.53 (d, J(H,H) ) 7.7 Hz, 3H; ABH(6)), 7.63 (d, J(H,H)
) 8.4 Hz, 3H; ABH(4)), 8.31 (t, 3H; ABH(2)), 8.50 (d, 3J(H,H) ) 7.9
Hz, 3H; GluNH), 10.25 (s, 3H; ABNH); 13C NMR (75 MHz, DMSO-
d6, 25 °C, TMS) δ 21.5 (iPrCH3), 26.3 (GluC(â)), 30.5 (GluC(γ)), 53.3
(GluC(R)), 67.1 (iPrC), 118.7 (ABC(2)), 122.0 (ABC(4)), 122.5 (ABC-
(6)), 128.6 (ABC(5)), 135.1 (ABC(1)), 138.7 (ABC(3)), 167.0 (ABCO),
170.0 (GluC(R)CO), 171.7 (GluC(γ)CO); CI-MS (NH3) m/z (relative
intensity) 888 (100) [M + NH4+], 828 (4) [M - iPrOH + NH4+]. Anal.
Calcd for C45H54N6O12‚2H2O (907.0): C, 59.59; H, 6.45; N, 9.27.
Found: C, 59.89; H, 6.42; N, 9.20.
3
Hz, 2H; AB2+4H(6)), 7.70 (d, J(H,H) ) 7.8 Hz, 1H; AB6H(6)), 7.87
3
(m, 2H; AB2+4H(4)), 7.94 (d, J(H,H) ) 8.2 Hz, 1H; AB6H(4)), 8.04
4
+ 8.07 (2s, 2H; AB2+4H(2)), 8.29 (t, J(H,H) ) 1.7 Hz, 1H; AB6H-
(2)), 8.64 (2d, 2H; Glu3+5NH), 10.15 + 10.29 + 10.38 (3s, 3H; AB2+4+6
-
NH). Anal. Calcd for C57H70N6O15 (1079.2): C, 63.44; H, 6.54; N,
7.79. Found: C, 63.16; H, 6.50; N, 7.60.
Cyclopeptide cyclo-[L-Glu(OiPr)-AB]3 (2). The linear hexapeptide
BOC-[L-Glu(OiPr)-AB]3-OBn (1.19 g, 1.10 mmol) is first deprotected
at the terminal amino group and then hydrogenated according to the
general methods. For cyclization, the completely deprotected hexapep-
tide is dissolved in DMF (100 mL/mmol). Solid TBTU (1.1 equiv,
1.21 mmol, 0.39 g) is added, and the solution is stirred until complete
dissolution. DIEA (3.2 equiv, 3.52 mmol, 0.61 mL) is added dropwise,
and stirring is continued for 2 h. Afterward, the reaction mixture is
poured into water (500 mL/mmol) under stirring. The pH is adjusted
to ca. 4 with 1 N HCl, and the suspension is stirred for another 10
min. The precipitate is filtered off, washed with water, and dried. The
product is isolated from this mixture chromatographically. At first, an
initial purification step is carried out using a silica gel column (CH2-
Cl2/MeOH 10:1). The resulting material is further purified with a RP-8
column. For this, the product is dissolved in a small amount of DMF
Acknowledgment. I thank Ms. D. Kubik for her assistance
and Prof. G. Wulff for his interest and financial support.
Supporting Information Available: COSY and NOESY
NMR spectra of the 2-tosylate complex in CDCl3 and of free
2 in DMSO-d6 and in 5% DMSO-d6/CDCl3 (PDF). See any
current masthead page for ordering information and Web access
instructions.
JA983970J