A New Cyclic Pseudopeptide Composed of (L)-Proline and 3-Aminobenzoic Acid Subunits as a Ditopic Receptor for the Simultaneous Complexation of Cations and Anions

Article abstract of DOI:10.1021/jo991087d

The synthesis and receptor properties of a cyclic pseudopeptide composed of (L)-proline and the nonnatural amino acid 3-aminobenzoic acid in an alternating sequence are described. The structure of cyclo[(L)Pro-AB]₃ was determined in the solid state by X-ray crystallography and in solution by one- and two-dimensional NMR techniques and FT-IR spectroscopy. The cyclic peptide preferentially adopts conformations comparable with the cone conformation of calixarenes. Similar to calixarenes, cyclo[(L)Pro-AB]₃ is able to bind cations by cation-π interactions with its aromatic subunits. In some complexes, the peptide NH groups interact additionally with anions and the cyclic peptide thus behaves as a ditopic receptor. The structure of the ternary complex between cyclo[(L)Pro-AB]₃ and N-methylquinuclidinium iodide was determined by X-ray crystallography. Spectroscopic investigations show that this complex has a similar geometry in solution. Stability constants of complexes of the cyclopeptide with various ion pairs have been determined. Crossover experiments show that electrostatic interactions between cation and anion complexed by cycto[(L)Pro-AB]₃ result in cooperative effects of either ion on the complexation of the corresponding counterion. The binding properties of cyclo[(L)Pro-AB]₃ are correlated with its conformation in solution. The properties of related cyclic hexapeptides in which one or two (L)-proline subunits are replaced by (L)-glutamic acid are also described. In comparison with cyclo[(L)Pro-AB]₃, these peptides possess a reduced cation and anion affinity. Anion complexation is weakened because the amide NH groups at the glutamic acid subunits are involved in strong intramolecular hydrogen bonds and are not available for interactions with other partners. Consequently, the cation complex stabilities also decrease. The amino acid subunits obviously influence the receptor properties of such cyclopeptides by controlling their solution conformation.

Full text of DOI:10.1021/jo991087d

J . Org. Chem. 1999, 64, 9475-9486
9475
A New Cyclic P seu d op ep tid e Com p osed of (L)-P r olin e a n d
3-Am in oben zoic Acid Su bu n its a s a Ditop ic Recep tor for th e
Sim u lta n eou s Com p lexa tion of Ca tion s a n d An ion s
Stefan Kubik*,† and Richard Goddard‡
Institut fu¨r Organische Chemie und Makromolekulare Chemie, Heinrich-Heine-Universita¨t,
Universita¨tsstrasse 1, D-40225 Du¨sseldorf, Germany, and Max-Planck-Institut fu¨r Kohlenforschung,
Kaiser-Wilhelm-Platz 1, D-45470 Mu¨lheim/ Ruhr, Germany
Received J uly 7, 1999
The synthesis and receptor properties of a cyclic pseudopeptide composed of (L)-proline and the
nonnatural amino acid 3-aminobenzoic acid in an alternating sequence are described. The structure
of cyclo[(L)Pro-AB]3 was determined in the solid state by X-ray crystallography and in solution by
one- and two-dimensional NMR techniques and FT-IR spectroscopy. The cyclic peptide preferentially
adopts conformations comparable with the cone conformation of calixarenes. Similar to calixarenes,
cyclo[(L)Pro-AB]3 is able to bind cations by cation-π interactions with its aromatic subunits. In
some complexes, the peptide NH groups interact additionally with anions and the cyclic peptide
thus behaves as a ditopic receptor. The structure of the ternary complex between cyclo[(L)Pro-AB]3
and N-methylquinuclidinium iodide was determined by X-ray crystallography. Spectroscopic
investigations show that this complex has a similar geometry in solution. Stability constants of
complexes of the cyclopeptide with various ion pairs have been determined. Crossover experiments
show that electrostatic interactions between cation and anion complexed by cyclo[(L)Pro-AB]3 result
in cooperative effects of either ion on the complexation of the corresponding counterion. The binding
properties of cyclo[(L)Pro-AB]3 are correlated with its conformation in solution. The properties of
related cyclic hexapeptides in which one or two (L)-proline subunits are replaced by (L)-glutamic
acid are also described. In comparison with cyclo[(L)Pro-AB]3, these peptides possess a reduced
cation and anion affinity. Anion complexation is weakened because the amide NH groups at the
glutamic acid subunits are involved in strong intramolecular hydrogen bonds and are not available
for interactions with other partners. Consequently, the cation complex stabilities also decrease.
The amino acid subunits obviously influence the receptor properties of such cyclopeptides by
controlling their solution conformation.
In tr od u ction
The obvious choice for a cyclic compound accessible by
a sequential synthesis is a cyclic peptide. Cyclopeptides
have not often been used as artificial receptors so far,3
probably because they are usually relatively flexible and,
despite their cyclic structure, are rarely able to bind
substrates in a well-defined cavity. The flexibility of cyclic
peptides can be significantly reduced, however, when
they are composed of (L)- and (D)-amino acids in an
alternating sequence4 or when they contain rigid subunits
such as aromatic5 or alicyclic6 spacers. Such conforma-
tionally restrained cyclopeptides have been used for the
binding of, e.g., amino acids,6a aromatic compounds,5f or
dicarboxylic acids.5i Using the same approach, Ishida and
co-workers have achieved the catalytic acceleration of
4-nitrophenyl acetate hydrolysis5d and the complexation
of phosphate anions5c with cyclic pseudopeptides com-
posed of 3-aminobenzoic acid and natural amino acids
in an alternating sequence (Figure 1). We have recently
demonstrated that such peptides (e.g. 1) are also able to
An important aspect of supramolecular chemistry is
the design of artificial receptors.1 One strategy for the
synthesis of such compounds consists of the attachment
of additional substituents to preformed cyclic host mol-
ecules such as crown ethers, cyclophanes, calixarenes, or
cyclodextrins. In this approach, substituents are selected
so that they participate in substrate binding and thus
improve the receptor selectivity or induce altered behav-
iors or properties. This strategy has recently been termed
the modular approach of receptor design.2 Although
many highly selective host molecules have been synthe-
sized in this way, we believe that another approach,
namely a sequential synthesis, may have several advan-
tages. The sequential assembly allows the deliberate
introduction and exchange of individual subunits at
defined positions in a cyclic receptor. The sequence of the
subunits can be varied systematically in order to study
its influence on substrate binding. Furthermore, hits
identified by a combinatorial receptor synthesis can be
resynthesized easily.
(3) Ozeki, E.; Kimura, S.; Imanishi, Y. Int. J . Pept. Protein Res. 1989,
34, 111-117. Ishizu, T.; Hirayama, J .; Noguchi, S. Chem. Pharm. Bull.
1994, 42, 1146-1148. Leipert, D.; Nopper, D.; Bauser, M.; Gauglitz,
G.; J ung, G. Angew. Chem., Int. Ed. Engl. 1998, 37, 3311-3314.
(4) Ghadiri, M. R.; Granja, J . R.; Milligan, R. A.; McRee, D. E.;
Khazanovich, N. Nature 1993, 366, 324-327. Granja, J . R.; Ghadiri,
M. R. J . Am. Chem. Soc. 1994, 116, 10785-10786. Ghadiri, R. M. Adv.
Mater. 1995, 7, 675-677. Hartgerink, J . D.; Granja, J . R.; Milligan,
R. A.; Ghadiri, M. R. J . Am. Chem. Soc. 1996, 118, 43-50. The natural
potassium receptor valinomycin, a cyclic dodecadepsipeptide, is com-
posed of (D)-hydroxycarboxylic acids and (L)-amino acids.
† Heinrich-Heine-Universita¨t. E-mail: kubik@uni-duesseldorf.de.
‡ Max-Planck-Institut fu¨r Kohlenforschung.
(1) Lehn, J .-M. Supramolecular chemistry; VCH: Weinheim, Ger-
many, 1995.
(2) Higler, I.; Timmermann, P.; Verboom, W.; Reinhoudt, D. N. Eur.
J . Org. Chem. 1998, 2689-2702.
10.1021/jo991087d CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/01/1999

9476 J . Org. Chem., Vol. 64, No. 26, 1999
Kubik and Goddard
F igu r e 2. Rotating bonds in a peptide chain between the
aromatic residues of 1 (a). Preferred conformation of 1 in 5%
CDCl3/DMSO-d6 (b).
only the rotations around the secondary amides are
hindered and the energetically more favored trans-amide
conformation is more probable. Rotations around all
remaining bonds are in principle possible (Figure 2a).
Our investigations have shown that, in nonpolar solvents
such as CDCl3, 1 preferentially adopts conformations that
are stabilized by intramolecular hydrogen bonds. In these
conformers, the protons on the aromatic nitrogen atoms
point “up” toward H(4) of the adjacent aromatic ring and
those on the glutamic acid nitrogen atoms point “down”
toward H(2) (Figure 2b). The peptide is nevertheless not
completely rigid because rotations around the bonds at
C(R) still occur.
One can speculate that it should be possible to improve
the receptor properties of 1 by increasing its rigidity.
Figure 2a shows that a rotation around the C(R)-N bonds
can be prevented by replacing the (L)-glutamic acid
subunits with (L)-proline. Here, we report on the synthe-
sis and properties of the new cyclopeptide 2. We show
how the conformational equilibrium and the binding
properties of this peptide are affected by the proline
subunits. As predicted, 2 possesses improved receptor
properties toward cations. But we also find that the
cation complex stabilities of 2 depend on the counterions
used. In fact, 2 behaves like a ditopic receptor with a
variety of ion pairs and binds cation and anion simulta-
neously.9 Finally, complexation studies with cyclopeptides
in which one (3) or two (4) proline subunits are replaced
by glutamic acid show that, analogously to natural
receptors, the binding properties of our artificial peptides
depend on the sequence of the subunits from which they
are composed.
F igu r e 1. Structures of the cyclic hexapeptides 1-4.
bind cations by cation-π interactions.7 In this respect,
they possess receptor properties comparable with those
of the structurally related calixarenes.8 Certain anions,
such as tosylate or phenylphosphonate, behave highly
cooperatively in the cation complexation of 1. They bind
to the cyclopeptide NH groups and stabilize a receptor
conformer that is ideally suited for interactions with
positively charged substrates. This behavior can be
regarded as an allosteric effect.
The stability of cation complexes of 1 is much lower in
the presence of anions that cannot act as effectors.7 This
is partly due to the flexibility of the host in solution.
Conformations of cyclic peptides such as 1 result from
rotations around the single bonds between the aromatic
subunits. For the six bonds between two aromatic rings,
(6) (a) Miyake, H.; Yamashita, T.; Kojima, Y.; Tsukube, H. Tetra-
hedron Lett. 1995, 36, 7669-7672. (b) Haubner, R.; Schmitt, W.;
Ho¨lzemann, G.; Goodman, S. L.; J onczyk, A.; Kessler, H. J . Am. Chem.
Soc. 1996, 118, 7881-7891. (c) Karle, I. L.; Ranganathan, D.; Haridas,
V. J . Am. Chem. Soc. 1996, 118, 10916-10917. (d) Albert, D.; Feigel,
M. Helv. Chim. Acta 1997, 80, 2168-2181. (e) Ranganathan, D.;
Haridas, V.; Madhusudanan, K. P.; Roy, R.; Nagaraj, R.; J ohn, G. B.
J . Am. Chem. Soc. 1997, 119, 11578-11584. (f) Karle, I. L.; Ranga-
nathan, D.; Haridas, V. J . Am. Chem. Soc. 1998, 120, 6903-6908.
(7) Kubik, S. J . Am. Chem. Soc. 1999, 121, 5846-5855.
(8) (a) Araki, K.; Shimizu, H.; Shinkai, S. Chem. Lett. 1993, 205-
208. (b) Inokuchi, F.; Araki, K.; Shinkai, S. Chem. Lett. 1994, 1383-
1386. (c) Takeshita, M.; Nishio, S.; Shinkai, S. J . Org. Chem. 1994,
59, 4032-4034. (d) Casnati, A.; J acopozzi, P.; Pochini, A.; Ugozzoli,
F.; Cacciapaglia, R.; Mandolini, L.; Ungaro, R. Tetrahedron 1995, 51,
591-598. (e) Arduini, A.; McGregor, W. M.; Paganuzzi, D.; Pochini,
A.; Secchi, A.; Ugozzoli, F.; Ungaro, R. J . Chem. Soc., Perkin Trans. 2
1996, 839-846. (f) 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. (g)
Arnecke, R.; Bo¨hmer, V.; Cacciapaglia, R.; Dalla Cort, A.; Mandolini,
L. Tetrahedron 1997, 53, 4901-4908. (h) Lhotak, P.; Shinkai, S. J .
Phys. Org. Chem. 1997, 10, 273-285.
(5) (a) Bach, A. C., II; Eyermann, C. J .; Gross, J . D.; Bower, M. J .;
Harlow, R. L.; Weber, P. C.; DeGrado, W. F. J . Am. Chem. Soc. 1994,
116, 3207-3219. (b) J ackson, S.; DeGrado, W.; Dwivedi, A.; Parthasa-
rathy, A.; Higley, A.; Krywko, J .; Rockwell, A.; Markwalder, J .; Wells,
G.; Wexler, R.; Mousa, S.; Harlow, R. J . Am. Chem. Soc. 1994, 116,
3220-3230. (c) Ishida, H.; Suga, M.; Donowaki, K.; Ohkubo, K. J . Org.
Chem. 1995, 60, 5374-5375. (d) Ishida, H.; Donowaki, K.; Suga, M.;
Shimose, K.; Ohkubo, K. Tetrahedron Lett. 1995, 36, 8987-8990. (e)
Bach, A. C., II; Espina, J . R.; J ackson, S. A.; Stouten, P. F. W.; Duke,
J . L.; Mousa, S. A.; DeGrado, W. F. J . Am. Chem. Soc. 1996, 118, 293-
294. (f) Garcia, M. E.; Gavin, J . A.; Deng, N.; Andrievsky, A. A.;
Mallouk, T. E. Tetrahedron Lett. 1996, 37, 8313-8316. (g) Mu, L.;
Cheng, J .; Huang, H.; Lu, J .; Hu, X. Synth. Commun. 1998, 28, 3091-
3096. (h) Ranganathan, D.; Haridas, V.; Gilardi, R.; Karle, I. L. J . Am.
Chem. Soc. 1998, 120, 10793-10800. (i) Ranganathan, D.; Haridas,
V.; Karle, I. L. J . Am. Chem. Soc. 1998, 120, 2695-2702.

A New Cyclic Pseudopeptide
J . Org. Chem., Vol. 64, No. 26, 1999 9477
Sch em e 1. Syn th esis of th e Cyclic Hexa p ep tid e 2
F igu r e 3. Rotating bonds in a peptide chain between the
aromatic residues of 2.
Resu lts a n d Discu ssion
Syn th esis a n d Con for m a tion a l Beh a vior of 2. We
prepared 2 in the same manner as the previously
prepared cyclopeptide 1.7 First, the repeating dipeptide
subunit was synthesized by coupling BOC-protected (L)-
proline with benzyl 3-aminobenzoate. This dipeptide was
successively elongated in solution, first to the tetrapep-
tide and then to the hexapeptide. After complete depro-
tection, the hexapeptide was cyclized under high-dilution
conditions (Scheme 1). The product yield (30-40%) was
much better when the cyclization was carried out at 80
°C, which is probably a consequence of the rigidity of the
linear precursor.
The binding properties of receptors are determined to
a large extent by their conformation in solution. The
cavity shape of cyclic host molecules and hence their
complexation behavior depends on the number and
conformational flexibility of the bonds around the ring.
All conformationally nonrigid bonds between two aro-
matic rings of 2 are indicated in Figure 3. A comparison
with the structure of 1 shows that, although the rotation
around the N-C(R) bonds is prevented in 2, the total
number of rotatable bonds is identical in both peptides.
Because rotations around tertiary amides possess sig-
nificantly lower energy barriers than those around
secondary amides, proline amides are not rigid and cis-
amides have also to be taken into account.10
F igu r e 4. Crystal structure of 2‚3CH3OH, viewed onto the
plane through the three N atoms N1, N3, and N5, showing
the O-H‚‚‚O and N-H‚‚‚O hydrogen-bonding interactions.
Selected interatomic distances (Å): O2‚‚‚O8 2.734(2), O5‚‚‚O7
2.755(2), O6‚‚‚C9 2.713(2), N1‚‚‚O4′ 2.889(2), N3‚‚‚O7 2.961-
(2), N5‚‚‚O9′ 2.814(2).
methanol with three solvent molecules (O-H‚‚‚O) hy-
drogen bonded to three carbonyl O atoms of each cyclo-
peptide unit. In addition, each NH group in the cyclo-
peptide ring forms further intermolecular (N-H‚‚‚O)
hydrogen bonds to adjacent oxygen atoms but only one
NH group forms an intercyclopeptide hydrogen bond [N1‚
‚‚O4′ 2.889(2) Å]. Figure 5 shows that all the proline
amides adopt trans conformations in the solid state.
Moreover, the three aromatic rings in 2 are all tilted in
the same direction and thus line the walls of a dish-
shaped cavity. In this respect, the structure resembles
the cone conformation of the calixarenes.11 The cavity of
the cyclopeptide is certainly shallower, but the capability
for guest molecules to interact with faces of the aromatic
rings should still be present. In contrast, guest molecules
bound by the “cystinophanes” described by Ranganathan
and Karle et al. can only interact with the edges of the
aromatic subunits.5i
Initial information about the conformational behavior
of the cyclopeptide 2 in solution was provided by its
crystal structure (Figure 4). 2 crystallizes as needles from
(9) For examples of other artificial ditopic receptors that bind ion
pairs see: Reetz, M. T.; Niemeyer, C. M.; Harms, K. Angew. Chem.,
Int. Ed. Engl. 1991, 30, 1474-1476. Arafa, E. A.; Kinnear, K. I.;
Lockhart, J . C. J . Chem. Soc., Chem. Commun. 1992, 61-64. Flack,
S. S.; Chaumette, J . L.; Kilburn, J . D.; Langley, G. J .; Webster, M. J .
Chem. Soc., Chem. Commun. 1993, 399-401. Rudkevich, D. M.;
Verboom, W.; Reinhoudt, D. N. J . Org. Chem. 1994, 59, 3683-3686.
Rudkevich, D. M.; Brzozka, Z.; Palys, M.; Visser, H. C.; Verboom, W.;
Reinhoudt, D. N. Angew. Chem., Int. Ed. Engl. 1994, 33, 467-468.
Savage, P. B.; Holmgren, S. K.; Gellman, S. H. J . Am. Chem. Soc. 1994,
116, 4069-4070. Rudkevich, D. M.; Mercer-Chalmers, J . D.; Verboom,
W.; Ungaro, R.; de J ong, F.; Reinhoudt, D. N. J . Am. Chem. Soc. 1995,
117, 6124-6125. Scheerder, J .; van Duynhoven, J . P. M.; Engbersen,
J . F. J .; Reinhoudt, D. N. Angew. Chem., Int. Ed. Engl. 1996, 35, 1090-
1093. Pelizzi, N.; Casnati, A.; Friggeri, A.; Ungaro, R. J . Chem. Soc.,
Perkin Trans. 2 1998, 1307-1311.
Cyclopeptide 2 does not, however, adopt an ideal C3-
symmetrical structure in the crystal. Figure 6 (left) shows
the result of superimposing the structure of the cyclo-
peptide with itself under the operations of idealized C3
symmetry. The fit (root-mean-square deviation) at 1.301
(10) See for example: Richardson, J . S.; Richardson, D. C. In
Prediction of protein structure and the principles of protein conforma-
tion; Fasman, G. D., Ed.; Plenum Press: New York, 1989; pp 1-98.
(11) Gutsche, C. D. Calixarenes; RSC: London, 1989. Shinkai, S.
Tetrahedron 1993, 49, 8933-8968. Bo¨hmer, V. Angew. Chem., Int. Ed.
Engl. 1995, 34, 713-745.

9478 J . Org. Chem., Vol. 64, No. 26, 1999
Kubik and Goddard
F igu r e 7. 1H NMR spectra of 2 in DMSO-d6 (a), CDCl3 (b),
and of its complex with BTMA+ iodide in CDCl3 (c) (G denotes
the guest signals).
F igu r e 5. Views of the cyclopeptide ring in 2‚3CH3OH,
showing selected proton-proton relationships.
The conformational behavior of 2 in solution was
investigated by one- and two-dimensional NMR tech-
1
niques and FT-IR spectroscopy. The H NMR spectra of
2 in CDCl3 and DMSO-d6 indicate an average C3-
symmetrical structure of the cyclic peptide in these
solvents (Figure 7a,b). 13C NMR spectroscopy allows an
assignment of the proline amide conformations in solu-
tion. The resonances of the â- and γ-carbon atoms in
proline rings are particularly sensitive to cis/trans isomer-
ization of the tertiary amides.12 A comparison of the
observed chemical shifts of C(â) and C(γ) with charac-
teristic resonances of these atoms in proline rings with
cis- or trans-amides12 indicates that trans-amides are
more probable at the proline subunits of 2 than cis-
amides. NOESY NMR spectroscopy gives further infor-
mation about the solution conformation of the peptide.
In the spectrum of 2 in CDCl3, cross-peaks between NH-
H(2), NH-H(4), and NH-H(R) are observed. Other cross-
peaks are visible between H(2)-H(δ) and H(6)-H(δ)
(Figure 8). Apart from additional NOE effects between
NH-H(â), an almost identical spectrum is observed in
DMSO-d6. Spatial proximities of H(2) and H(δ) as well
as H(6) and H(δ) that would give rise to NOE effects are
indeed visible in the crystal structure of 2 (Figure 5). It
is therefore reasonable to assume that the proline
subunits in 2 are similarly oriented in both solution and
the solid state. NOE effects between the NH and all of
its adjacent protons illustrate the predicted rotation of
the secondary amide groups in solution.
F igu r e 6. Superposition of solid-state conformations of cy-
clopeptide in 2‚3CH3OH (left) and 2‚N-methylquinuclidinium
iodide (right) with those resulting from rotation by 120° about
the normal to the mean plane through the N-H groups, to
show the approximation to idealized C3 symmetry [rms fit: 2,
1.301 Å; 2‚N-methylquinuclidinium iodide, 0.492 Å].
Å is rather bad. Whereas the three proline subunits are
similarly orientated with respect to their adjacent aro-
matic rings, the secondary amide groups are not. The
most significant deviation from idealized C3 symmetry
involves the secondary amide at N3, which results from
a N-H‚‚‚O hydrogen-bonded interaction with one of the
methanol molecules located at the center of the cyclo-
peptide ring. The different geometries at the three amide
nitrogens N1, N3, and N5 cause two of the aromatic NH
protons to be oriented toward H(4) of the nearest aro-
matic group whereas one is pointed away (Figure 5). It
is reasonable to assume that in solution the three
secondary amide groups are flexible and in fact able to
rotate relatively freely. In contrast, the orientations of
the proline groups and aromatic rings would appear to
be more fixed.
(12) Deber, C. M.; Madison, V.; Blout, E. R. Acc. Chem. Res. 1976,
9, 106-113. Wu¨thrich, K. NMR in biological research: peptides and
proteins; North-Holland Publ. Co.: Amsterdam, 1976.

A New Cyclic Pseudopeptide
J . Org. Chem., Vol. 64, No. 26, 1999 9479
then it has also been verified for a variety of other
synthetic receptors such as calix[n]arenes,8 homo-
oxacalixarenes,17 and cryptophanes.18 It has also been
emphasized that cation-π interactions are important
factors for the complexation of cationic substrates, e.g.
acetylcholine, by their natural receptors.15,19
An assignment of cation-π interactions to the binding
mode in complexes between cations and artificial recep-
tors is simple when the receptor contains few other
functional groups apart from aromatic rings. However,
the additional amide groups in 1 and 2 may also be able
to interact with positively charged guests. We were not
able to detect any influence of cation binding on the
carbonyl groups of 1 by FT-IR spectroscopy. It is possible
that the carbonyls still assist in the complexation of the
substrates by increasing the overall negative charge
density inside the cyclopeptide cavity. This interpretation
is consistent with results of a recent publication, in which
a cooperative effect of ester carbonyls on interactions
between cations and aromatic esters was described.20
For the complex between 1 and BTMA+ iodide, we have
determined a stability constant Ka of 300 M-1 and a
maximum shift of the N-methyl groups of the guest in
the complex (∆δmax) of +0.05 ppm by NMR titration.7 This
method is now a standard tool for Ka measurements in
molecular recognition studies and has also been applied
in the investigations we present here.21 Titrations were
carried out in CDCl3 with a water content of 0.01% in
order to suppress effects of weak intra- or intermolecular
hydrogen-bonding interactions. All saturation curves of
the investigated complexes of 2 could be fitted on the
F igu r e 8. Intramolecular NOE effects in the NOESY NMR
spectrum of 2 in CDCl3.
A facile rotation of this group can additionally be
inferred from the FT-IR spectrum of 2 in 1% DMSO-d6/
CDCl3. In this spectrum, a band for NH vibrations is
observed at 3408 cm-1 with a distinct shoulder at ca. 3425
cm-1. The shoulder indicates that the three NH groups
of 2 have slightly different environments, possibly be-
cause their orientations to the aromatic subunits vary.
A weak band between 3325 and 3225 cm-1 may be
assigned to NH groups involved in intramolecular hy-
drogen bonds with carbonyl group of adjacent subunits.13
This interaction cannot be strong, however, because
usually hydrogen-bonded NH groups have a significantly
larger IR absorption than free ones.14 Finally, the amide
II vibration band is also split and possesses maxima at
1553 and 1553 cm-1, again indicating a rotation of the
secondary amide groups. The fact that no significant
hydrogen bonding was detected in the FT-IR spectrum
indicates that self-association of the cyclopeptide is not
1
very probable in solution. Accordingly, a H NMR dilution
experiment in CDCl3 in the concentration range of 2-0.01
mM revealed no significant dependence of the NH
resonances of 2 on the cyclopeptide concentration. Self-
association of the receptor in solution can therefore be
neglected in studies on its binding properties.
basis of simple 1:1 equilibria, hence Ka ) ccomplex
/
(chost‚cguest). The stability constants in Tables 1, 2, and 4
have been determined repeatedly, and the experimental
errors in our measurements were usually below 20%.
In summary, cyclopeptide 2 adopts an averaged C3-
symmetrical conformation in solution that resembles the
averaged solid-state structure in Figure 4. In the pre-
ferred conformer of 2, the N-C(δ) bonds at the proline
subunits point “down” toward H(2) of the adjacent
aromatic ring. The same orientation has been found for
the corresponding NH groups in cyclopeptide 1 (Figure
2b). In this respect, the conformation of peptides 1 and
2 are comparable. A conformational stabilization of the
secondary amide group by intramolecular hydrogen
bonds observed for 1 does not occur in 2. These amides
rotate and the cyclic peptide is not completely rigid.
However, since conformers other than those of peptide 1
dominate in solution, the proline subunits have a pro-
nounced influence on the complexing properties of 2.
Com p lexa tion P r op er ties of 2. We have shown that
1 binds cationic substrates, e.g. quaternary ammonium
ions, in chloroform.7 The complexation was demonstrated
Cyclopeptide 2 behaves completely differently from 1
in the complexation of BTMA+ iodide. The addition of this
salt to solutions of 2 in CDCl3 results in marked upfield
shifts of all BTMA+ protons in the 1H NMR spectrum
with a maximum shift of, e.g., the N-methyl protons of
up to -1.1 ppm. Complex formation furthermore causes
a significant change of the host spectrum (Figure 7c).
Besides pronounced shifts of the aromatic as well as the
aliphatic protons of 2, all signals become very broad. This
line broadening is most probably due to a reduced
flexibility of the host in the complex. The stability
constant of the complex is almost 2 orders of magnitude
larger than the corresponding one of 1 and amounts to
(16) Petti, M. A.; Shepodd, T. J .; Barrans, R. E., J r.; Dougherty, D.
A. J . Am. Chem. Soc. 1988, 110, 6825-6840. Stauffer, D. A.; Dougherty,
D. A. Tetrahedron Lett. 1988, 29, 6039-6042. Kearney, P. C.; Mizoue,
L. S.; Kumpf, R. A.; Forman, J . E.; McCurdy, A.; Dougherty, D. A. J .
Am. Chem. Soc. 1993, 115, 9907-9919. Mecozzi, S.; West, A. P., J r.;
Dougherty, D. A. J . Am. Chem. Soc. 1996, 118, 2307-2308. Also other
groups have described the complexation of cations by cyclophanes:
Cattani, A.; Dalla Cort, A.; Mandolini, L. J . Org. Chem. 1995, 60,
8313-8314. Me´ric, R.; Lehn, J . M.; Vigneron, J . P. Bull. Soc. Chim.
Fr. 1994, 131, 579-83.
1
by the small but significant upfield shift of the H NMR
resonances of the guest upon addition of 1. The upfield
shift was explained by an inclusion of the ammonium ion
into the cyclopeptide cavity, the result of which brings
the guest protons in close proximity to the π-systems of
the aromatic rings. Similar effects are usually observed
when positively charged guest molecules interact with
artificial hosts by cation-π interactions.15 Dougherty first
reported this type of interaction for cyclophanes.16 Since
(17) (a) De Iasi, G.; Masci, B. Tetrahedron Lett. 1993, 34, 6635-
6638. (b) Masci, B. Tetrahedron 1995, 51, 5459-5464. (c) Masci, B.;
Finelli, M.; Varrone, M. Chem. Eur. J . 1998, 4, 2018-2030.
(18) Garel, L.; Lozach, B.; Dutasta, J .-P.; Collet, A. J . Am. Chem.
Soc. 1993, 115, 11652-11653.
(19) Dougherty, D. A.; Stauffer, D. A. Science 1990, 250, 1558-1560.
Dougherty, D. A. Science 1996, 271, 163-168.
(13) Ne´el, J . Pure Appl. Chem. 1972, 31, 201-225.
(14) Vinogradov, S. N.; Ninnell, R. H. Hydrogen Bonding; Van
Nostrand Reinhold Co.: New York, 1971.
(20) Roelens, S.; Torriti, R. J . Am. Chem. Soc. 1998, 120, 12443-
12452.
(21) Connors, K. A. Binding Constants; Wiley: New York, 1987.
Macomber, R. S. J . Chem. Educ. 1992, 69, 375-378.
(15) Ma, J . C.; Dougherty, D. A. Chem. Rev. 1997, 97, 1303-1324.

9480 J . Org. Chem., Vol. 64, No. 26, 1999
Kubik and Goddard
Ta ble 1. Sta bility Con sta n ts Ka a n d Ma xim u m Ch em ica l
Sh ifts ∆δm a x of Com p lexes of 2 w ith Va r iou s Am m on iu m
Iod id es (CDCl3, T ) 298 K)
Ka(cation)
Ka(iodide)
ammonium
iodide
Ka(iodide)/
a
a
Ka
-∆δmax
Ka
+∆δmax Ka(cation)
acetylcholine
BTMA+
N,N-dimethyl-
pyrrolidinium
N-methyl-
11 000
21 100
35 700
1.15
1.11
1.15
18 700
31 500
55 700
0.84
0.83
0.83
1.7
1.5
1.6
42 200
1.16
82 800
0.85
2.0
quinuclidinium
a
Error limits of Ka < 20%.
21 100 M-1 when the upfield shift of the N-methyl group
is followed by a NMR titration. Surprisingly, a smaller
Ka of 7320 M-1 results when the shift of the n-butyl CH3
group is analyzed. This effect has not been observed for
the complex between BTMA+ iodide and 1 where the two
stability constants calculated from the shifts of the
N-methyl and the n-butyl CH3 group of the guest are of
the same order of magnitude (Ka(N-methyl) ) 300 M-1
;
Ka(n-butyl CH3) ) 380 M-1). The fact that the N-methyl
protons and the n-butyl CH3 protons of BTMA+ both shift
upon complexation by 1 or 2 suggests that the geometry
of the complexes is disordered with the N-methyl group
of the guest pointing either into or toward the exterior
of the hosts. Similar results have been obtained for
complexes of related quaternary ammonium ions with
calixarenes.8g,17b,18 But in the case of peptide 2, the
geometry with the cationic headgroup included into the
cavity obviously also leads to the more stable complex.
To evaluate the influence of guest structure on the
stability of complexes of 2, NMR titrations were carried
out with a number of other quaternary ammonium
iodides. With the exception of acetylcholine, all the
investigated iodides have simple aliphatic or alicyclic
cations so that their interactions with 2 are not compli-
cated by binding mechanisms such as π-π interactions
or hydrogen bonding. The results are summarized in
Table 1. The maximum upfield shift ∆δmax is of the same
order for all of the investigated ammonium ion com-
plexes. The stability constants range from 11 000 M-1 for
acetylcholine iodide to 42 200 M-1 for N-methylquinu-
clidinium iodide when the shifts of the N-methyl protons
were used for the nonlinear regression. Again, smaller
values for Ka resulted when the shifts of protons remote
from the cationic headgroup of the guest were analyzed.
The stability constants of the ammonium iodide com-
plexes of 2 are unusually large. For cation-π interactions
between cations and neutral artificial receptors in CDCl3,
stability constants <103 M-1 have been reported so
far,8,16,17 with only very few exceptions.17a,18 It is therefore
unlikely that cation-π interactions are solely responsible
for complex formation. Most probably, additional effects
contribute to the stability of these complexes.
F igu r e 9. Crystal structure of 2‚N-methylquinuclidinium
iodide, view as in Figure 4, showing the N-H‚‚‚I- hydrogen-
bonding interactions between the neutral cyclopeptide and the
I- anion [N1‚‚‚I 4.360(5), N3‚‚‚I 3.956(5), N5‚‚‚I 4.081(5) Å].
We were able to crystallize the complex of 2 with
N-methylquinuclidinium iodide and determine its crystal
structure (Figure 9). The structure provides valuable
information about the interactions between the cyclo-
peptide and the substrate. It demonstrates that 2 binds
both the cation and the anion simultaneously. As pre-
dicted, the cation is held in the shallow dish-shaped
cavity of 2 where it can interact with the aromatic
subunits of the cyclopeptide (Figure 10). Clearly, not only
is the N-methyl headgroup of the ammonium ion able to
enter the cavity but the whole bicyclic ring can be
F igu r e 10. Side and top views of the 2‚N-methylquinuclid-
inium iodide complex, showing the arrangement of the cation
and anion with respect to the cyclopeptide ring.
accommodated. Such a configuration would explain the
upfield shift of all the guest protons on complexation. In
solution, one would certainly expect a fast equilibrium
between different orientations of the cation inside the
cavity.
Reetz has found that R-methylene protons of tetrabu-
tylammonium cations interact with carbonyl groups of

A New Cyclic Pseudopeptide
J . Org. Chem., Vol. 64, No. 26, 1999 9481
enolates by hydrogen bonding.22 In the crystal structures
of such complexes, a distance of ca. 3.30 Å between the
negatively charged oxygen atoms and the R-C atoms of
the ammonium ions was found. In our case, the shortest
distance between a neutral peptide carbonyl oxygen atom
and an R-C atom is 3.75 Å. Moreover, all carbonyl groups
point outward away from the center of the complex. We
therefore believe that directed hydrogens bonds between
the guest and the neutral carbonyl groups of the host are
not very probable. Nevertheless, we cannot rule out that
the carbonyl groups cooperatively contribute to guest
binding by electrostatic interactions. It should also be
noted that all the proline amides have trans conforma-
tions in the crystal. The torsion angles involving these
amides deviate slightly from 180°, with a maximum
deviation of 13° (C26-N6-C30-C31 -167.1(4)°). Con-
formations of proline amides seem not to be restricted to
either syn or anti, which may be one of the reasons for
the unusual complexing properties of 2.
In the ternary complex of 2, the iodide anion is bound
at the smaller opening of the cyclopeptide. There it is
surrounded by the proline rings and appears to interact
with all three NH groups, which are all directed toward
the I- anion (Figure 10). Such N-H‚‚‚I- interactions
between the iodide ion and NH groups of secondary
amides are not unknown, e.g., lunarine hydroiodide (N-
H‚‚‚I- 3.73-3.66 Å)23 or 4′-N-methylstaurosporine me-
thiodide (N-H‚‚‚I- 3.78 Å).24 The N-H‚‚‚I- distances in
2‚N-methylquinuclidinium iodide at 3.96-4.36 Å are
somewhat longer, but this is probably a result of the
steric repulsive influence of the neighboring proline rings.
Figure 10 shows views of the complex which illustrate
this point. The anion sits 4.40-4.96 Å from the centroids
of the proline rings and approximately in the plane
defined by them (perpendicular distance: 0.19 Å) and
1.41 Å above the plane through the three NH groups.
As a result of anion binding, the cyclopeptide adopts
an almost C3-symmetrical conformation, with all three
NH groups pointing toward the center of the cavity.
Figure 6 (right) shows the result of superimposing the
structure of the cyclopeptide with itself under the opera-
tions of idealized C3 symmetry. As the comparison with
the uncomplexed 2 (Figure 6, left) clearly shows, the C3
point symmetry of the cyclopeptide in 2‚N-methylquinu-
clidinium iodide is almost exact. This occurs despite an
unsymmetrical crystal environment, caused in part by
the presence of one molecule of CDCl3 per cyclopeptide
in the crystal.
around the secondary amides of 2, which is observed for
free 2, does not occur anymore. As in the crystal, a C3-
symmetrical conformation of 2 seems to be stabilized by
the simultaneous interaction of all three NH groups with
the iodide anion.
Additional information on the interactions of 2 with
the anion can be derived from NMR spectroscopy. An
assignment of the peptide conformation in the complex
by NOESY NMR spectroscopy was not possible because
broadening of the host signals upon addition of BTMA+
iodide prevented the measurement of suitable spectra.
None the less, important information can be gained from
1
the H NMR spectrum itself. Clearly, the electric field of
anions bound at the narrower opening of 2 should
influence the resonances of only those proline protons
that point toward the center of the cavity. Indeed, a
significant downfield shift of only one of the two proline
H(δ) signals is observed upon complexation of ammonium
iodides by 2. NOESY NMR spectroscopic investigations
of the related ammonium tosylate complex, which will
be discussed below, show that this signal indeed corre-
sponds to the inner H(δ) protons of 2. Its shift can be
used to determine the Ka of the iodide complex of 2
quantitatively. The iodide complex stabilities of the salts
that have also been used for the determination of the
cation stability constants are summarized in Table 1.
Evidently, the stability of the iodide complex is not
independent of the counterion. A stronger complexation
is found for the more stable ammonium complexes, and
the iodide complex is generally 1.5-2 times more stable
than the corresponding cation complex. It can be specu-
lated that the iodide complexation is influenced by subtle
effects caused by the complexation of the cation on the
conformation of 2. For example, inclusion of the bulkier
ammonium ions in the cyclopeptide cavity is expected to
result in the aromatic subunits at the upper rim being
pushed somewhat apart. This would slightly decrease the
diameter of the cavity at its narrower opening and lead
to the observed stronger interaction of the NH groups of
the cyclopeptide with iodide.
The high stabilities of the ammonium iodide complexes
in Table 1 can now be easily explained by electrostatic
interactions between the bound ion pairs that, in addition
to cation-π interactions, stabilize the cation complexes
of 2. Similar effects have been observed for cation
complexes of 1 in the presence of tosylate anions,7 but in
contrast to 2, no cooperative effect of iodide on the
complexing behavior of 1 could be detected. The strong
intramolecular hydrogen bonds in 1 seem to prevent a
complexation of this anion. A simultaneous anion and
cation complexation can only be achieved with ions that
can form stronger hydrogen bonds to NH. In the case of
2, the amide groups of the cyclopeptide are not signifi-
cantly involved in intramolecular hydrogen bonds and
an interaction with iodide is possible.
Spectroscopic investigations show that 2 behaves as a
ditopic receptor also in solution. The interaction of 2 with
cations is evident in the upfield shift of the guest protons
1
in the H NMR spectra. The interaction with iodide can
be demonstrated by FT-IR spectroscopy. In the spectrum
of the BTMA+ iodide complex, one broad band is visible
in the NH stretching region at 3348 cm-1. Compared with
the spectrum of free 2, this band is shifted toward lower
frequencies by 60 cm-1, which can be attributed to
interactions between NH groups and the anions.14 Fur-
thermore, besides minor shifts of all CdO vibrations, only
one amide II band is observed indicating that the rotation
It can consequently be expected that 2 is also able to
bind other anions. Table 2 illustrates that the BTMA+
complex stability indeed depends on the nature of the
anion. The cyclopeptide forms the most stable BTMA+
complexes when tosylate is the counterion. A similar
behavior has been detected for 1. Stability constants >106
are difficult to determine accurately by NMR titration,
but Table 2 certainly reflects the expected trend in the
dependence of Ka on the nature of the anion. FT-IR and
1H NMR spectroscopic investigations of the complex
between BTMA+ tosylate and 2 indicate strong hydrogen
(22) Reetz, M. T.; Hu¨tte, S.; Goddard, R. J . Am. Chem. Soc. 1993,
115, 9339-9340.
(23) J effreys, J . A. D.; Ferguson, G. J . Chem. Soc. B 1970, 826-
829.
(24) Funato, N.; Takayanagi, H.; Konda, Y.; Toda, Y.; Harigaya, Y.;
Iwai, Y.; Omura, S. Tetrahedron Lett. 1994, 35, 1251-1254.

9482 J . Org. Chem., Vol. 64, No. 26, 1999
Kubik and Goddard
Ta ble 2. Sta bility Con sta n ts Ka a n d Ma xim u m Ch em ica l
Sh ifts ∆δm a x of Com p lexes of 2 w ith Va r iou s BTMA+ Sa lts
(CDCl3, T ) 298 K)
Ka(R-N(CH3)3)
Ka(1)a -∆δmax Ka(2)a -∆δmax Ka(1)/Ka(2)
1550
Ka(n-butyl CH3)
BTMA+
TFPB-
0.17
0.70
1.11
0.99
1560
1140
0.32
0.28
0.35
0.33
1.0
1.1
picrate
1260
21 100
205 000
iodide
7320
17200
2.9
11.9
tetrafluoro-
borate
tosylate
5 050 000
1.16
204 000
0.26
24.8
a
Error limits of Ka < 20%.
F igu r e 12. Intra- and intermolecular NOE effects in the
NOESY NMR spectrum of the complex of 2 with BTMA+
tosylate in CDCl3.
Ta ble 3. NH a n d CdO Vibr a tion Ba n d s in th e F T-IR
Sp ectr a of Com p lexes of 2 w ith Va r iou s BTMA+ Sa lts
(c ) 2 m M in 1% DMSO-d 6/CDCl3)
anion
ν(NH)
ν(CdO)pro ν(CdO)amide
ν(CdO)amide
I
II
3425/3408
3421
3420/3409
3348
1622
1622
1622
1629
1630
1690
1691
1689
1687
1685
1553/1533
1558/1535
1554/1534
1535
TFPB-
picrate
iodide
tetrafluoro- 3374
1541
borate
tosylate
3292
1627
1683
1549
cyclopeptide. The outer H(δ) protons of 2 only experience
a small upfield shift upon complexation of BTMA+
tosylate. Cross-peaks between the BTMA+ N-methyl and
the aromatic and NH protons of 2 show that, in solution,
the cationic head of the guest is inserted deeply into the
dish-shaped cavity of the receptor.
F igu r e 11. 1H NMR spectra of 2 (a), BTMA+ tosylate (c), and
the complex between 2 and BTMA+ tosylate (b) in CDCl3.
The BF4- anion seems to interact with the NH groups
of 2 by NH‚‚‚F bonds. Results of the FT-IR-spectroscopic
characterization of this and all other anion complexes of
2 are summarized in Table 3. In three complexes of 2,
the ones with tosylate, tetrafluoroborate, and iodide
counterions, a higher Ka was observed for the N-methyl
than for the n-butyl CH3 group of the cation (Table 2).
This effect is stronger for anions that interact better with
the cyclic peptide. Indeed, it is not surprising that
electrostatic interactions between the anion located at
the bottom of the cyclopeptide cavity and the guest favor
a complex geometry with the cationic headgroup of the
guest closer to the anion. With two of the investigated
anions, picrate and tetrakis[3,5-bis(trifluoromethyl)phe-
nyl]borate (TFPB-), this effect was not observed, how-
ever. In both cases, the stability constants of the N-me-
thyl and n-butyl CH3 groups of BTMA+ are of the same
order of magnitude, and even those of the N(R)CH2
groups of the cations (Ka(N(R)CH2) for picrate 1270 M-1
and for TFPB- 1340 M-1) are comparable. Furthermore,
the BTMA+ complex stabilities of 2 are similar in the
presence of either anion.
bonds between the tosylate oxygen atoms and the NH
groups of the cyclopeptide. The NH protons of 2, for
1
example, are shifted downfield by +0.62 ppm in the H
NMR spectrum, and the NH bond vibration is located at
only 3292 cm-1 in the IR spectrum. In fact, the complex
is so stable that, in an equimolar mixture of 2 and
BTMA+ tosylate, the conformational equilibrium is shifted
almost completely toward the side of the complex and
sharp signals are observed in the 1H NMR spectrum
(Figure 11). This enables the determination of the
complex structure in solution by NOESY NMR spectros-
copy. The NOE effects that are important for the assign-
ment of the complex geometry are depicted schematically
in Figure 12. These effects are consistent with a confor-
mation of 2 similar to the one in the crystal structure of
the complex (Figure 10). An intermolecular cross-peak
is visible between the tosylate aryl protons and the same
H(δ) protons of 2 that shift downfield upon addition of
the guest. On one hand, this result shows that tosylate
is bound to the narrower opening of the cyclopeptide and
is arranged parallel to the proline rings. On the other
hand, the signal of the H(δ) protons can now be easily
assigned to protons that point toward the center of the
Since the NMR resonance of the inner H(δ) protons of
2 (∆δ < +0.03 ppm) and the IR spectrum of the cyclic

A New Cyclic Pseudopeptide
J . Org. Chem., Vol. 64, No. 26, 1999 9483
Ta ble 4. Sta bility Con sta n ts Ka a n d Ma xim u m Ch em ica l
Sh ifts ∆δm a x of Com p lexes of 1-4 w ith BTMA+ Iod id e
(CDCl3, T ) 298 K)
Ka(R-N(CH3)3)
Ka(n-butyl CH3)
Ka(1)/
cyclopeptide Ka(1)a -∆δmax -∆G Ka(2)a -∆δmax -∆G Ka(2)
2
3
4
1
21100 1.11 24.7 7320
3860 0.82 20.5 3360
0.35
0.89
0.56
0.02
22.0
20.1
16.6
14.7
2.9
1.1
1.1
0.8
850 0.49 16.7
300 0.05 14.1
800
380
a
Error limits of Ka < 20%.
peptide are not significantly affected by the presence of
picrate or TFPB- (Table 3), specific interactions between
these anions and the peptide are not visible. From other
molecular recognition studies it is in fact known that the
large lipophilic anion TFPB- does not interfere in inter-
actions between a cation and a receptor.25 The splitting
of the N-H vibration and amide II bands in the FT-IR
spectrum indicates that 2 is conformationally flexible in
solution in the presence of both anions. The cyclopeptide
is nevertheless able to interact with BTMA+, and complex
formation must now be primarily due to cation-π inter-
actions. Since the peptide is not preorganized by interac-
tions with an anion, the maximum chemical shifts ∆δmax
of the BTMA+ N-methyl groups are reduced. We have
detected a similar behavior for 1 in the absence of effector
anions.7
The stability of the BTMA+ picrate or the TFPB-
complex of 2 is 3-4 times larger than that of the BTMA+
iodide complex of 1. Despite the conformational flexibility
of 2, the proline-containing cyclopeptide obviously forms
significantly more stable complexes with BTMA+ than
1, which may certainly be attributed to the rigidity of
the proline subunits.
The different ∆δmax values of the BTMA+ protons in
the presence of TFPB- or picrate show that the two
anions still influence cation complex formation of 2
differently. Possibly the geometry with which BTMA+ is
included into the cyclopeptide cavity is not the same. We
could therefore not completely eliminate the effects of
anions on the cation complex formation of 2. Especially
in nonpolar solvents a relatively strong ion pair aggrega-
tion has to be considered and the cation may still be
oriented in close proximity of the cation even after
complexation by the host. In fact, the dependence of the
cation affinity of artificial receptors such as calixarenes
on the nature of the anion has often been attributed
solely to ion pair aggregation.8e,g,17a In the case of cyclo-
peptide 2, however, it is evident that specific interactions
of the anions with the receptor may also be responsible
for an anion effect on cation binding.26
F igu r e 13. Plots of stability constants Ka (a), free enthalpy
of complex formation (∆G) (b), and ∆δmax (c) of complexes of
1-4 with BTMA+ iodide vs number of proline residues n (Ka
in M-1, ∆G in kJ mol-1, ∆δmax in ppm, T ) 298 K).
was indeed observed. In Figure 13, the dependencies of
the stability constants Ka, the free enthalpy of complex
formation ∆G, and the maximum shift ∆δmax on the
number of proline residues in the ring are plotted. The
graphs show an exponential increase of Ka when the
number of proline residues in the receptor rises from n
) 0 to n ) 3. Consequently, a linear relationship is found
when ∆G is plotted against n. From the slope of this
graph one can calculate that the stability of the BTMA+
iodide complexes increases on average by -3.6 kJ mol-1
/
additional proline residue in the cyclopeptide. Figure 13c
also shows a relatively good linear relationship between
∆δmax and n. This dependence indicates that the more
proline subunits the cyclopeptide receptors contain, the
better they are preorganized for the complexation of
cations. We have shown that the preorganization of 2 is
not only due to the rigidity of the proline subunits but
also to interactions of the cyclopeptide NH groups with
the anion. Interactions with iodide have not been detected
in the glutamic acid containing cyclopeptide 1 because
the amides in this peptide are involved in intramolecular
hydrogen bonds. In principle, such hydrogen bonds
should always be possible when a receptor of the general
structure depicted in Figure 1 contains natural amino
acids other than proline. Indeed, significant hydrogen
bonding of the NH groups was detected in the FT-IR
spectra of 3 and 4. The number of possible intramolecular
hydrogen bonds is directly proportional to the number
Com p lexa t ion P r op er t ies of 3 a n d 4. The mixed
cyclopeptides 3 and 4 have been obtained by cyclization
of linear hexapeptides with the appropriate subunit
sequences. It can be expected that their binding affinities
toward, e.g., BTMA+ iodide will range between those of
1 and 2. Table 4 and Figure 13 show that such a behavior
(25) Nishida, H.; Takada, N.; Yoshimura, M.; Sonoda, T.; Kobayashi,
H. Bull. Chem. Soc. J pn. 1984, 57, 2600-2604. Bell, D. A.; Anslyn, E.
V. Tetrahedron 1995, 51, 7161-7172.
(26) Doxsee, K. M. J . Org. Chem. 1989, 54, 4712-4715. Olsher, U.;
Hankins, M. G.; Kim, Y. D.; Bartsch, R. A. J . Am. Chem. Soc. 1993,
115, 3370-3371. In these papers it was speculated that the influence
of anions on the affinity of certain crown ethers toward cations may
be attributed to specific effects of the anions on the conformation of
the host in the final complex. In this respect, these publications are
related to this work.

9484 J . Org. Chem., Vol. 64, No. 26, 1999
Kubik and Goddard
Du¨sseldorf, Germany. The syntheses of AB-OBn‚TsOH, Boc-
Glu(OiPr)-OBn, Boc-[AB-Glu(OiPr)-AB]2-OBn, Boc-[AB-Glu-
(OiPr)-AB]3-OBn, and cyclo[AB-Glu(OiPr)-AB]2 have been
described elsewhere.7 The following abbreviations are used:
TsOH, toluene-4-sulfonic acid; Bn, benzyl-; BOC, tert-butoxy-
carbonyl-; DIEA, N-ethyldiisopropylamine; TFA, trifluoroacetic
acid; PyCloP, chlorotripyrrolidinophosphonium hexafluoro-
phosphate; TBTU, O-(1H-benzotriazol-1-yl)-N,N,N′,N′-tetram-
ethyluronium tetrafluoroborate; Glu, glutamic acid; Pro, pro-
line; AB, 3-aminobenzoic acid.
Host-Gu est Titr a tion s. Stock solutions of the cyclopeptide
(2, 1 µmol/800 µL; 3, 4, 2 µmol/800 µL) in CDCl3 (water content
0.01%) and of the corresponding guest (0.4 µmol/200 µL for 2,
3; 0.2 µmol/200 µL for 4) in 1% DMSO-d6/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. Chemical shifts were
calculated by using the CHCl3 signal at 7.27 ppm as internal
standard. The shifts of, e.g., the N-methyl protons of the guest
were plotted against chost/cguest. From the resulting saturation
curve, Ka and ∆δmax were calculated by a nonlinear least-
squares fitting method for 1:1 complexes using the SIGMA Plot
3.0 (J andel Scientific) software package.21 The same set of
NMR spectra could also be used for the determination of the
stability of the cyclopeptide iodide complexes. For this, the
chemical shifts of the inner H(δ) protons of 2 were plotted
against cguest/chost. For the determination of the chemical shift
at cguest ) 0 a spectrum of the pure cyclopeptide in 0.2% DMSO-
d6/CDCl3 was recorded. The stability constant could be calcu-
lated from the resulting saturation curve by taking into
account that this time protons of the host and not of the guest
have been monitored.
Gen er a l P r oced u r e for Hyd r ogen a tion of Ben zyl Es-
ter s. 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.
Gen er a l P r oced u r e for Clea va ge of N-ter t-Bu toxyca r -
bon yl Gr ou p s. Meth od A. The carbamate is dissolved in CH2-
Cl2 (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.
of glutamic acid subunits in the ring. Therefore, the
decrease of cation complex stability with increasing
number of glutamic acid subunits can be attributed to a
successive loss of the iodide affinity of the cyclopeptides.
The more glutamic acid subunits these peptides contain,
the fewer NH groups are free to interact with the anion.
The reduced iodide complex stabilities of 3 and 4 can be
indirectly derived from the decreasing stability differ-
ences when different BTMA+ protons are monitored in
the titrations (Table 4).
Our investigations illustrate that the choice of subunits
has a significant effect on the conformational behavior
of the cyclopeptides and, consequently, their binding
properties. Knowledge of the influence of individual
subunits should in principle enable one to modulate the
properties of such host molecules deliberately by choosing
an appropriate subunit sequence.
Con clu sion s a n d Ou tlook
In certain aspects, the cyclopeptides presented in this
paper have properties similar to calixarenes. In solution,
they adopt conformations comparable to the cone con-
formation of calixarenes. Moreover, they are able to bind
cations by cation-π interactions with their aromatic
subunits. However, in contrast to simple calixarenes, the
aromatic rings are not their only binding sites. The
presence of NH groups allows additional interactions
with anions by hydrogen bonding. Simultaneous cation
and anion complexation has already been described for
cyclopeptide 1.7 In this paper we have extended this
approach with the introduction of the new cyclopeptide
2. Receptor 2 is able to bind a variety of ion pairs and
can therefore be regarded as a new ditopic receptor. We
have found that the nature of the cationic guest affects
the affinity of the cyclopeptide toward the corresponding
anion and vice versa. The cooperativities of various
cations and anions have been quantified independently
and could be furthermore correlated with specific effects
of the ions on the conformation of 2 in the complexes.
We could also demonstrate that the binding properties
of the presented cyclopeptides are determined by influ-
ences of their subunits on the preferred solution confor-
mation of the host. A detailed knowledge of the influence
of individual subunits may facilitate the syntheses of
receptors with predictable properties. This approach is
highly flexible with regard to the nature of the subunits.
In addition, the introduction of functional groups either
in the aromatic or the aliphatic subunits of the cyclopep-
tides which are able to serve as additional binding sites
can be realized easily. Studies on the design of new and
efficient artificial host molecules on the basis of similar
cyclopeptides are in progress.
Meth od B. The carbamate is suspended in 1,4-dioxane (20
mL). The resulting suspension is cooled with an ice bath, and
a 6 N solution of HCl in 1,4-dioxane (40 mL) is added dropwise.
The reaction mixture is stirred for 2 h at 0-5 °C. Afterward,
the solution is evaporated to dryness in vacuo.
Dip ep tid e BOC-(L)P r o-AB-OBn . 3-Aminobenzoic acid
benzyl ester toluene-4-sulfonate (2.39 g, 6.00 mmol) and BOC-
(L)-proline (1.81 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 chromato-
graphically (1:1 ethyl acetate/hexane) from the residue. The
product solidifies during drying in vacuo: yield 2.49 g (98%);
Exp er im en ta l Section
Ma ter ia ls. All solvents were dried according to standard
procedures prior to use. DMF pA was purchased from Fluka
and was used without further purification. PyCloP was
prepared according to the literature procedure.27 TBTU was
purchased from Bachem. For chromatographic separations
either ICN silica gel 32-63 (ICN Biomedicals) or a Merck
LiChroprep RP-8 (40-63 µm) prepacked column, size B (310-
25), was used. Elemental analyses were carried out at the
Pharmaceutical Institute of the Heinrich-Heine-University,
mp 52-55 °C; [R]25 ) -59.0 (c ) 2, methanol); 1H NMR (300
D
MHz, DMSO-d6, 100 °C, TMS) δ 1.33 (s, 9H), 1.86 (m, 3H),
2.19 (m, 1H), 3.38 (m, 2H), 4.30 (dd, J ) 8.3/4.1 Hz, 1H), 5.35
(s, 2H), 7.43 (m, b, 6H), 7.67 (td, J ) 7.7/1.2 Hz, 1H), 7.88 (2
× dd, J ) 8.1/1.1 Hz, 1H), 8.22 (t, J ) 1.9 Hz, 1H), 9.86 (s,
1H). Anal. Calcd for C24H28N2O5 (424.5): C, 67.91; H, 6.65; N,
6.60. Found: C, 67.70; H, 6.81; N, 6.67.
Tetr a p ep tid e BOC-[(L)P r o-AB]2-OBn . BOC-(L)Pro-AB-
OBn (0.89 g, 2.10 mmol) is deprotected at the terminal amino
group according to method A. An equivalent amount of the
same dipeptide is hydrogenated. Both components as well as
(27) Coste, J .; Fre´rot, E.; J ouin, P. J . Org. Chem. 1994, 59, 2437-
2446.

A New Cyclic Pseudopeptide
J . Org. Chem., Vol. 64, No. 26, 1999 9485
PyCloP (1.2 equiv, 2.52 mmol, 1.06 g) are dissolved in CH2Cl2
(20 mL/mmol). At room temperature, DIEA (3.4 equiv, 7.14
mmol, 1.24 mL) 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) from the residue. The product is triturated with
hexane to afford a white solid: yield 1.26 g (94%); mp 133 °C
resulting material is dissolved in a small amount of DMF and
further purified by using a RP-8 column.
cyclo[(L)P r o-AB]3 (2). For the RP-8 column, initially 1:10
1,4-dioxane/H2O is used as eluent. Gradually it is changed to
1:5 1,4-dioxane/H2O and then to 1:2.5 1,4-dioxane/H2O, with
which the pure product is eluted. The product is recrystallized
from methanol: yield 43%; mp >250 °C (softening from 243
(softening from 105 °C); [R]25 ) -106.0 (c ) 2, methanol); 1H
°C); [R]25 ) -66.8 (c ) 2, methanol); 1H NMR (300 MHz,
D
D
NMR (300 MHz, DMSO-d6, 100 °C, TMS) δ 1.34 (s, 9H), 1.90
(m, 6H), 2.16 (m, 1H), 2.27 (m, 1H), 3.39 (m, 2H), 3.58 (m,
2H), 4.23 (dd, J ) 8.3/3.8 Hz, 1H), 4.58 (m, 1H), 5.36 (s, 2H),
7.17 (d, J ) 6.8 Hz, 1H), 7.39 (m, b, 7H), 7.64 (d, J ) 7.2 Hz,
1H), 7.67 (td, J ) 7.7/1.3 Hz, 1H), 7.75 (s, b, 1H), 7.87 (d, J )
8.2 Hz, 1H), 8.20 (s, b, 1H), 9.72 (s, 1H), 9.95 (s, 1H). Anal.
Calcd for C36H40N4O7‚H2O (658.7): C, 65.64; H, 6.43; N, 8.50.
Found: C, 65.49; H, 6.46; N, 8.24.
Hexa p ep tid es. A dipeptide is deprotected at the terminal
amino group according to method A. An equivalent amount of
a tetrapeptide is hydrogenated. Both components as well as
TBTU (1.1 equiv) are dissolved in DMF (30 mL/mmol). At room
temperature, DIEA (3.2 equiv) 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 in high
purity by this procedure and can be used for the following step
without further purification. A small amount is purified
chromatographically (1:15 methanol/CH2Cl2) for characteriza-
tion.
BOC-[(L)P r o-AB]3-OBn : yield 92%; mp 186 °C (softening
from 147 °C); [R]25D ) -144.8 (c ) 2, methanol); 1H NMR (300
MHz, DMSO-d6, 100 °C, TMS) δ 1.34 (s, 9H), 1.93 (m, 9H),
2.39 (m, 3H), 3.38 (m, 2H), 3.57 (m, 4H), 4.23 (dd, J ) 8.3/3.7
Hz, 1H), 4.59 (m, 2H), 5.36 (s, 2H), 7.17 (d, J ) 6.6 Hz, 2H),
7.43 (m, b, 8H), 7.65 (m, 3H), 7.76 (s, b, 2H), 7.86 (d, J ) 7.7
Hz, 1H), 8.21 (s, b, 1H), 9.73 (s, 1H), 9.83 (s, 1H), 9.95 (s, 1H).
Anal. Calcd for C48H52N6O9‚H2O (875.0): C, 65.89; H, 6.22; N,
9.60. Found: C, 65.88; H, 6.15; N, 9.51.
DMSO-d6, 25 °C, TMS) δ 1.82-2.03 (m, 9H), 2.28 (m, 3H), 3.39
(m, 3H), 3.46 (m, 3H), 4.64 (dd, J ) 8.2/3.6 Hz, 3H), 7.11 (d, J
) 7.5 Hz, 3H), 7.22 (d, J ) 8.1 Hz, 3H), 7.38 (t, J ) 7.9 Hz,
3H), 8.38 (s, 3H), 10.31 (s, 3H); 13C NMR (75 MHz, DMSO-d6,
25 °C, TMS) δ 24.5 C(γ), 29.4 C(â), 49.3, 59.8, 118.1/118.2,
119.5/119.6, 120.4, 128.5, 138.1, 138.8/138.9, 168.7, 170.2/
170.3. Anal. Calcd for C36H36N6O6‚2H2O (684.7): C, 63.15; H,
5.89; N, 12.27. Found: C, 63.44; H, 5.86; N, 12.22. CI-MS (NH3)
[m/z (relative intensity)]: 666 (100) [M + NH4+], 649 (2) [M +
H+].
Cr ystal Str u ctu r e An alysis of2.28 Data: C39H48N6O9‚(CH4O)3
Mr ) 744.83, colorless prism, crystallized from methanol,
crystal size 0.24 × 0.36 × 0.40 mm, a ) 8.490(1) Å, b ) 19.624-
(1) Å, c ) 22.681(2) Å, V ) 3779.0(5) Å3, T ) 100 K,
orthorhombic, space group P212121 (No. 19), Z ) 4, Dc ) 1.31
g cm-3, µ ) 0.09 mm-1; Siemens SMART diffractometer, Mo
KR X-radiation, λ ) 0.710 73 Å, 42307 measured reflections,
analytical absorption correction (Tmin ) 0.964 96, Tmax 0.986 52),
14 253 unique (θmax ) 33.9°), 9608 (gt) with I > 2.0σ(Fo2). The
structure was solved by direct methods29 and refined by full-
matrix least squares30 on F2 for all data with Chebyshev
weights, with R ) 0.052 (gt), Rw ) 0.123 (all data), 514
parameters, S ) 0.979, H atoms riding, disordered methylene
group [0.66(1):0.34(1)], max shift/error 0.001, and residual Fmax
) 0.362 e Å-3
.
Cr ysta l Str u ctu r e An a lysis of 2‚N-Meth ylqu in u clid -
in iu m Iod id e.28 Data: C36H36N6O6[I]-[C8H16N]+‚CDCl3, Mr )
1022.19, colorless prism, crystallized from toluene/deuterio-
chloroform, crystal size 0.11 × 0.14 × 0.24 mm, a ) 13.7350-
(2) Å, b ) 14.7780(2) Å, c ) 22.3104(3) Å, V ) 4528.5(1) Å3, T
) 100 K, orthorhombic, space group P212121 (No. 19), Z ) 4,
Dc ) 1.50 g cm-3, µ ) 0.94 mm-1, Nonius KappaCCD diffrac-
tometer, Mo KR X-radiation, λ ) 0.710 73 Å, 41 057 measured
reflections, analytical absorption correction (Tmin ) 0.827 23,
Tmax ) 0.922 37), 9888 unique (θmax ) 27.5°), 7241 (gt) with I
> 2.0σ(Fo2). The structure was solved by direct methods29 and
refined by full-matrix least squares30 on F2 for all data with
Chebyshev weights, with R ) 0.056 (gt), Rw ) 0.158 (all data),
559 parameters, S ) 1.04, H atoms riding, absolute configu-
ration determined [Flack parameter -0.03(2)], max shift/error
0.001, and residual Fmax ) 2.317 e Å-3 (1.615 Å from Cl1).
cyclo[(L)Glu (OiP r )-AB-[(L)P r o-AB]2] (3). For the RP-8
column, initially 1:2 MeOH/H2O is used as eluent. Gradually
it is changed to 1:1 MeOH/H2O and then to 2:1 MeOH/H2O,
with which the pure product is eluted. In case it is not obtained
white after this column, another one on silica gel can be carried
BOC-(L)Glu (OiP r )-AB-[(L)P r o-AB]2-OBn : yield 92%; mp
136 °C (softening from 127 °C); [R]25 ) -76.1 (c ) 2,
D
methanol); 1H NMR (300 MHz, DMSO-d6, 100 °C, TMS) δ 1.16
(d, J ) 6.2 Hz, 6H), 1.34 (s, 9H), 1.79-2.47 (m, b, 12H), 3.39
(m, 2H), 3.59 (m, 2H), 4.24 (dd, J ) 8.3/3.6 Hz, 1H), 4.64 (m,
2H), 4.88 (sept, J ) 6.3 Hz, 1H), 5.36 (s, 2H), 7.17 (d, J ) 7.2
Hz, 1H), 7.35 (m, b, 8H), 7.58 (dt, J ) 8.0/1.4 Hz, 1H), 7.65 (d,
J ) 8.3 Hz, 1H), 7.69 (dt, J ) 8.2/1.6 Hz, 1H), 7.76 (m, 2H),
7.89 (2 × dd, J ) 8.2/1.1, 1H), 8.02 (s, 1H), 8.81 (d, J ) 7.4
Hz, 1H), 8.24 (t, J ) 1.8 Hz, 1H), 9.73 (s, 1H), 9.85 (s, 1H),
10.01 (s, 1H). Anal. Calcd for C51H58N6O11‚H2O (949.1): C,
64.54; H, 6.37; N, 8.86. Found: C, 64.50; H, 6.42; N, 8.94.
BOC-[(L)Glu (OiP r )-AB]2-(L)P r o-AB-OBn : yield 94%; mp
1
108-113 °C; [R]25 ) -56.1 (c ) 2, methanol); H NMR (300
D
MHz, DMSO-d6, 100 °C, TMS) δ 1.17 (m, 12H), 1.39 (s, 9H),
1.86-2.44 (m, b, 12H), 3.60 (m, 2H), 4.14 (m, 1H), 4.64 (m,
2H), 4.89 (sept, J ) 6.3 Hz, 2H), 5.35 (s, 2H), 6.61 (d, J ) 7.6
Hz, 1H), 7.18 (d, J ) 8.2 Hz, 1H), 7.39 (m, b, 8H), 7.59 (dt, J
) 8.0/1.2 Hz, 1H), 7.65 (m, 2H), 7.78 (m, 2H), 7.86 (d, J ) 7.9
Hz, 1H), 8.01 (t, J ) 1.7 Hz, 1H), 8.20 (m, 2H), 9.76 (s, 1H),
out with acetone as eluent: yield 34%; 206-214 °C; [R]25
)
D
-33.0 (c ) 2, methanol); 1H NMR (300 MHz, DMSO-d6, 25
°C, TMS) δ 1.15 + 1.16 (2 × d, J ) 6.3 Hz, 6H), 1.85-2.49 (m,
b, 12H), 3.51 (m, 4H), 4.67 (m, 3H), 4.86 (sept, J ) 6.2 Hz,
1H), 7.12 (d, J ) 7.6 Hz, 1H), 7.20 (m, 2H), 7.40 (m, b, 4H),
7.50 (m, 2H), 8.04 (t, 1H), 8.41 (t, 1H), 8.73 (t, 1H), 8.46 (d, J
) 7.9 Hz, 1H), 10.28 (s, 1H), 10.33 (s, 1H), 10.35 (s, 1H); 13C
NMR (75 MHz, DMSO-d6, 25 °C, TMS) δ 21.5, 24.4, 24.6, 26.6,
29.3, 29.4, 30.5, 49.3, 49.4, 53.1, 59.7, 60.1, 67.1, 117.9, 118.0,
119.2, 119.6, 120.5, 120.7, 121.5, 122.6, 128.5, 128.7, 135.7,
137.8, 137.9, 138.4, 138.8, 139.1, 166.9, 168.6, 168.7, 169.9,
9.90 (s, 1H), 9.96 (s, 1H). Anal. Calcd for
C54H64N6O13
(1005.1): C, 64.53; H, 6.42; N, 8.36. Found: C, 64.38; H, 6.64;
N, 8.37.
Cyclop ep tid es. The linear hexapeptides are first hydro-
genated. Afterward, they are deprotected at the terminal
amino group according to method B. For cyclization, a com-
pletely deprotected hexapeptide is dissolved in DMF (100 mL/
mmol), and DIEA (3.2 equiv) is added. The resulting mixture
is heated to 80 °C, and a solution of TBTU (1.1 equiv) in DMF
(20 mL) is added dropwise. Stirring is continued for 2 h at 80
°C. The solvent is then evaporated in vacuo. The residue is
stirred with diethyl ether overnight (2) or triturated with water
(3, 4). The crude product is filtered off, dried, and purified
chromatographically. At first, an initial purification step is
carried out using a silica gel column (10:1 CH2Cl2/MeOH). The
(28) The crystallographic data (without structure factors) for 2 and
2‚N-methylquinuclidinium iodide have been deposited as supplemen-
tary publications CCDC-126020 and CCDC-126021 at the Cambridge
Crystallographic Data Centre. Copies of the data can be obtained from
the following: The Director, CCDC, 12 Union Road, Cambridge CB2
1EZ, U.K. (Telefax +44-1223/336-033. E-mail: deposit@ccdc.cam.ac.uk.)
(29) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, A46, 467-
473.
(30) Sheldrick, G. M. SHELXL-97, Program for Crystal Structure
Refinement, Universita¨t Go¨ttingen, 1997.

9486 J . Org. Chem., Vol. 64, No. 26, 1999
Kubik and Goddard
170.3, 170.4, 171.7. Anal. Calcd for C39H42N6O8‚H2O (740.8):
C, 63.23; H, 5.99; N, 11.34. Found: C, 63.51; H, 5.98; N, 11.33.
CI-MS (NH3) [m/z (relative intensity)]: 740 (100) [M + NH4+],
723 (2) [M + H+], 680 (4) [M - iPrOH + NH4+].
128.6, 128.7, 135.1, 135.2, 137.7, 138.3, 138.7, 139.1, 166.9,
167.3, 168.4, 169.9, 170.0, 170.4, 171.7. Anal. Calcd for
C42H48N6O10‚H2O (814.9): C, 61.91; H, 6.18; N, 10.31. Found:
C, 61.87; H, 6.26; N, 10.26. CI-MS (NH3) [m/z (relative
intensity)]: 814 (100) [M + NH4+], 797 (2) [M + H+].
cyclo[(L)Glu (OiP r )-AB]2-(L)P r o-AB] (4). For the RP-8
column, initially 1:2 MeOH/H2O is used as eluent. Gradually
it is changed to 1:1 MeOH/H2O, 2:1 MeOH/H2O, and then to
3:1 MeOH/H2O, with which the pure product is eluted. In case
it is not obtained white after this column, another one on silica
gel can be carried out with acetone as eluent: yield 43%; 172-
Ack n ow led gm en t. S.K. thanks Mrs. D. Kubik for
her assistance in the preparative work and Prof. G.
Wulff for his financial support. Thanks are also due to
Dr. U. Matthiesen for the CI mass spectra and Mrs. M.
Beuer for the majority of the NMR spectra.
177 °C; [R]25 ) -28.2 (c ) 2, methanol); 1H NMR (300 MHz,
D
DMSO-d6, 25 °C, TMS) δ 1.16 (m, 12H), 1.85-2.45 (m, b, 12H),
3.45 (m, 1H), 3.54 (m, 1H), 4.58 (m, 1H), 4.67 (m, 2H), 4.87 +
4.89 (2 × sept, J ) 6.3 Hz, 2 × 1H), 7.17 (dt, J ) 7.5 Hz, 1H),
7.32 (dt, J ) 7.9 Hz, 1H), 7.42 (m, b, 6H), 7.68 (d, J ) 8.0 Hz,
1H), 8.01 (t, 1H), 8.35 (t, 1H), 8.73 (t, 1H), 8.46 (d, J ) 8.0 Hz,
1H), 8.53 (d, J ) 7.7 Hz, 1H), 10.25 (s, 1H), 10.31 (s, 1H), 10.33
(s, 1H); 13C NMR (75 MHz, DMSO-d6, 25 °C, TMS) δ 21.5, 24.6,
26.2, 26.5, 29.4, 30.4, 30.6, 49.4, 53.1, 53.5, 59.9, 67.0, 67.1,
117.8, 118.8, 119.1, 120.6, 121.0, 121.4, 121.9, 122.5, 122.6,
Su p p or tin g In for m a tion Ava ila ble: FT-IR spectrum of
2 in 1% DMSO-d6/CDCl3, NOESY NMR spectra of 2 and its
BTMA+ tosylate complex in CDCl3, and crystal data, bond
lengths and angles, atomic coordinates, and anisotropic ther-
mal parameters for 2 and its complex with N-methylquinu-
clidinium iodide. This material is available free of charge via
the Internet at http://pubs.acs.org.
J O991087D