Folding control in cyclic peptides through N-methylation pattern selection: Formation of antiparallel β-sheet dimers, double reverse turns and supramolecular helices by 3α,γ cyclic peptides

Article abstract of DOI:10.1002/chem.200701059

Peptide foldamers constitute a growing class of nanomaterials with potential applications in a wide variety of chemical, medical and technological fields. Here we describe the preparation and structural characteristics of a new class of cyclic peptide foldamers (3α,γ-CPs) that, depending on their backbone N-methylation patterns and the medium, can either remain as flat rings that dimerize through arrays of hydrogen bonds of antiparallel β-sheet type, or can fold into twisted double reverse turns that, in the case of double γ-turns, associate in nonpolar solvents to form helical supramolecular structures. A 3α,γ-CP consists of a number of multiples of a repeat unit made up of four amino acid residues of alternating chirality: three corresponding to examino acids and one to a γ-amino acid (a cis-3-aminocycloalkanecarboxylic acid).

Full text of DOI:10.1002/chem.200701059

DOI: 10.1002/chem.200701059
Folding Control in Cyclic Peptides through N-Methylation Pattern Selection:
Formation of Antiparallel b-Sheet Dimers, Double Reverse Turns and
Supramolecular Helices by 3a,g Cyclic Peptides
Manuel Amorín, Luis Castedo, and Juan R. Granja*[a]
Abstract: Peptide foldamers constitute
a growing class of nanomaterials with
potential applications in a wide variety
of chemical, medical and technological
fields. Here we describe the prepara-
tion and structural characteristics of a
new class of cyclic peptide foldamers
(3a,g-CPs) that, depending on their
backbone N-methylation patterns and
the medium, can either remain as flat
rings that dimerize through arrays of
hydrogen bonds of antiparallel b-sheet
type, or can fold into twisted double re-
verse turns that, in the case of double
g-turns, associate in nonpolar solvents
to form helical supramolecular struc-
tures. A 3a,g-CP consists of a number
of multiples of a repeat unit made up
of four amino acid residues of alternat-
ing chirality: three corresponding to a-
amino acids and one to a g-amino acid
(a
cis-3-aminocycloalkanecarboxylic
Keywords: beta sheets · foldamers ·
acid).
helical structures
self-assembly
·
peptides
·
Introduction
vidually or by association—to form tertiary or even quater-
nary structures known as tyligomers.[2b,7] In this respect, al-
though the most intensively studied foldamers have been
helical, much attention has recently focused on sheet-form-
ing foldamers, because of their relevance to pathological
amyloid plaques[8] and their potential use for the manufac-
ture of nanotapes or nanotubes.[9] Sheet-forming foldamers
include certain types of cyclic peptides (CPs) designed to
stack spontaneously through arrays of hydrogen bonds simi-
lar to those found in b-sheets, to form nanotubes [SPNs
(self-assembling peptide nanotubes) such as CP-A Q SPN-
A (Figure 1)].[10]
Changes in the structure, number and distribution of the
CPs allow for the modulation of the properties of the result-
ing SPNs and their subsequent development for chemical
studies and technological applications.[10–13] For example, al-
ternating N-methylation on a CP (CP-B) makes it form
dimers DCP-B (CP-B Q DCP-B, Figure 1), the shortest possible
SPNs, which have been successfully employed to estimate
the relative stabilities of b-sheetsꢀ parallel and antiparallel
arrangements.[14]
In pursuit of biomaterials with improved biocompatibility,
rigidity, responsiveness, specificity and other properties that
are desirable in medical devices and drug delivery systems,
increasing attention is being paid to nanostructures.[1] These
entities consist of hierarchies of substructures, the higher-
ranking members of which are formed by the same kinds of
self-assembly process as are involved in the formation of
natural biopolymers that carry out sophisticated chemical
operations such as catalysis or electron transport. To gain
deeper insight into how to control the three-dimensional
structures of such materials, chemists have in recent years
devoted increasing research effort to synthetic foldamers,
non-natural oligomers that exhibit the same kinds of secon-
dary structural motifs as natural biomolecules (helices,
sheets, loops etc.).[2–6]
A major goal of foldamer chemistry is to develop oligo-
mers that spontaneously adopt higher-order structure—indi-
[a] Dr. M. Amorín, Prof. Dr. L. Castedo, Prof. Dr. J. R. Granja
Departamento de Química Orgµnica
Laboratorios del CSIC
In recent years we have been exploring the possibilities of
CPs composed of a-amino acids (a-Aas) alternating with a
g-amino acid, more specifically a cis-3-aminocycloalkanecar-
boxylic acid (g-Aca).[15,16] In these a,g-CPs (e.g., CP-1–3 in
Figure 1) the projection of one of the cycloalkane methyl-
enes into the lumen (space within the tube) creates a hydro-
phobic region, and the relative rigidity of the cyclohexane
Facultad de Química, Universidad de Santiago
15782 Santiago de Compostela (Spain)
Fax : (+34)981-595012
E-mail: qojuangg@usc.es
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
2100
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2100 – 2111

FULL PAPER
dues of the same chirality (Figure 2), or l-g-Aca residues for
only half the l-a-Aa residues. The overall symmetry of the
resulting a,g-CP, when flat, is Cn, where n is the number of
[l-g-Aca-d-a-Aa] or [d-g-Aca-l-a-Aa] units in the peptide
ring, and thus also depends on the size of the ring.[15] How-
ever, the cavity properties of SPNs composed of CPs with g-
Aca residues, and hence their technological applications,
may well depend not only on size and number or type of
substituents on their inner surfaces but also on their symme-
tries. It is therefore of interest to be able to control the over-
all symmetry of CPs independently of their size. As part of
an extensive program directed towards the design, synthesis
and structural and functional evaluation of novel homo- and
heterosupramolecular entities involving CPs,[15,16] but also
with aim of checking the possibility described above, we
have now synthesized a new class of CPs in which the a-
amino acids (a-Aas) and cis-3-aminocyclohexanecarboxylic
acid (g-Ach) are combined in a three to one ratio (“3a,g-
CPs”), resulting from substitution of g-Ach residues for only
half the a-Aa residues of identical chirality to the formal
d,l-a-CP parent. These CPs should retain the b-sheet-form-
ing propensity and at the same time would be a simple
model of novel SPNs with partial hydrophobic cavities of C2
symmetry.
Surprisingly, the full structural characterization of the new
CPs showed that their folding is dictated by their methyla-
tion patterns.[19,20] These novel 3a,g-CPs can either retain
their flat ring structures through hydrogen bonds in b-sheet-
like arrays, or they can fold into double reverse turn motifs,
which may associate as helical aggregates.[21]
Figure 1. SPN and dimer formation by stacking of cyclic peptides, and the
structures of a,g-CPs.
ring ensures that the g-amino
acid segments of the peptide
backbone have the all-trans
conformation necessary for the
peptide ring to be flat and
stackable.[15,17] This conforma-
tional rigidity means that in
CPs the cycloalkane ring of a g-
Aca can be regarded as a su-
peratom, and (1S,3R)- and
(1R,3S)-g-Aca
equivalent to d- and l-a-amino
acid residues, respectively
residues
as
(Figure 2).[18] Hereafter we avail
ourselves of this equivalence by
using d and l to describe the
stereochemistry of g-Aca resi-
dues.
In terms of the above equiva-
lence, the a,g-CPs with which
we have been working in recent
years can each be viewed as the
result of taking a parent nano-
tube-forming CP consisting of
alternating d- and l-a-Aas (d,l-
a-CP) and replacing alternate
Figure 2. Equivalence between d-(l)-a-amino acids and (1S,3R)-[(1R,3S)]-g-aminocycloalkanecarboxylic acids
(a-Acas), and the generation of a,g- or 3a,g-CPs by substitution of d-g-Acas for selected d-g-Aas in a d,l-a-
a-Aa residues with g-Aca resi- CP. Solid dishes indicate cycloalkane superatoms.
Chem. Eur. J. 2008, 14, 2100 – 2111
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2101

J. R. Granja et al.
Results and Discussion
our studies with peptide CP-4 in which the l-Aas, both the
g-Ach and Ala, were methylated. Their NMR spectra in
polar organic solvents showed the presence of several slowly
interconverting conformers, presumably the results of cis–
trans isomerization about peptide bonds. Under this assump-
tion, the temperature dependence of the NMR data allowed
the activation barrier for cis–trans isomerization in
[D6]DMSO be estimated as 16–17 kcalmolꢀ1 at the coales-
cence temperature (358 K).
The replacement of a-Aa by g-Aca residues has the result
that in the flat conformation of the resulting CP (3a,g-
CPs)—as in a,g-CPs, in one of which all the amide NH and
C=O groups belong to a-Aas and on the other to g-Acas—
the two faces are distinct: at one face (a,a) all the amide
NH and C=O groups belong to a-Aas, while at the other
(a,g) they belong to equal numbers of a-Aa and g-Aca resi-
dues. All the residues with NH and C=O groups pointing to
the same side have the same chirality. Because of the differ-
ent spacings of the NH and C=O groups on the two faces,
the association of CPs in SPNs through b-sheet-like hydro-
gen bonding requires successive
1
At room temperature the H NMR spectrum of CP-4 in
deuterochloroform showed a single set of signals that lacked
amide proton signals because of coalescence, but spectra re-
corded at below 253 K showed two species that intercon-
CPs to lie antiparallel,[22] so that
like faces face each other: a,g-
faces bond to a,g-faces (AG-
type bonding) and a,a-faces
bond to a,a-faces (AA-type
bonding) (Figure 3). To investi-
gate these two types of associa-
tion separately and to avoid the
insolubility of multi-CP SPNs,
we adopted the same strategy
as in our work on a,g-CP-based
SPNs,[14,15] using CPs in which
the backbone NH groups of
either one CP face or the other
were methylated so as to limit
the b-sheet-like association of
CPs to the b-strand dimer stage
(Figure 4).
Suspecting
that
Figure 3. AA- and AG-type hydrogen bonding between cyclo-[(l-g-Ach-d-Aa-l-Aa-d-Aa)2-] units (for clarity,
methylation of a-Aas might be
affecting dimerization efficiency
by preventing the CP from
adopting a sufficiently flat con-
formation, we also carried out
experiments in which only half
the NH groups of one face
were methylated. It was in
these latter experiments that
the alternative folding proper-
ties of 3a,g-CPs were manifest-
ed.
most amino acid side chains have been omitted in the representation of the nanotube).
CPs with full amide N-methyla-
tion of one face: The C2-sym-
metric CPs cyclo-[(l-g-Ach-
d-MeN-Ala-l-Leu-d-MeN-Ala)2-]
(CP-4) and cyclo-[(l-MeN-g-
Ach-d-Phe-l-MeN-Ala-d-Phe)2-]
(CP-5), in which g-Ach is cis-3-
aminocyclohexanecarboxylic
acid, were prepared by standard
procedures and characterized
by NMR spectroscopy and
Figure 4. Cyclic peptides CP-4 and CP-5 and their corresponding dimers (for clarity, most amino acid side
mass spectrometry. We started chains have been omitted from the representations of the dimers).
2102
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2100 – 2111

Folding Control in Cyclic Peptides
FULL PAPER
verted slowly on the NMR timescale. The assumption that
one of the two species was the monomer CP-4 and the
other the AG-bonded dimer D-4 is supported by the fact
that their relative concentrations depended on total peptide
concentration and by their interconversion rate, which was
moderately slow on the NMR timescale, as befits a process
involving the making or breaking of eight hydrogen
bonds.[23] Also, the presence of dimers was confirmed by
NOE cross-peaks between the a-carbon protons of the ala-
nines (at 5.47 and 5.95 ppm), which are in close proximity to
species was monomeric and the other dimeric was again
supported by the dependence of their relative concentra-
tions on total peptide concentration, which allowed the asso-
ciation constant in CDCl3 at 303 to be estimated as 62mꢀ1
(Table S1, Supporting Information).[24] Van’t Hoff plots over
the 263–303 K range afforded values of ꢀ47.9 kJmolꢀ1 for
DHA298 and 123.8 JKꢀ1 molꢀ1 for DSA298. These values are
consistent with dimerization being essentially an enthalpy-
driven[27] hydrogen-bonding process.[14] The b-sheet hydro-
3
gen-bonding network is supported by the JNa values of the
dimer, which are substantially greater than those of the mo-
nomer, as well as by the downfield shifts (Table 1) of the
1
each other in the dimer but not in the monomer. H NMR
experiments in which total peptide concentration was varied
afforded a value of at least of 10mꢀ1 for the association con-
stant of the dimerization process at 243 K (see Table S1,
Supporting Information).[24] Although the poor solubility of
CP-4 in chloroform at low temperature prevented reliable
Van’t Hoff analysis, the strength of the dimerizing hydrogen
bonds is reflected by the downfield shifts of the Leu and g-
ꢀ
Phe N H (8.31 and 8.80 ppm in the dimer as against 7.49
and 7.10 ppm, respectively, in the monomer) and a-carbon
signals (5.26–5.89 ppm in the dimer as against 4.86–5.16 ppm
in the monomer),[25] as well as by the amide bands at 1525,
1627 and 3310 cmꢀ1 in the IR spectrum (Table 2).[15,22,26]
ꢀ
Ach N H signals from 7.02 and 7.66 ppm, respectively, in
the monomer (CP-4) to 8.75 and 8.21 ppm, respectively, in
the dimer (D-4). Their antiparallel b-sheet nature is attested
Table 2. Summary of IR data for 3a,g-peptides CP-4–10.
[a]
Peptide CP-1[a]
Amide I?
Amide IIII
1542
Amide A[b]
CP-4/D-4
1632 (1673)
3310
3413
3310
3296
3403
3324
3312
3417
3326
3432
3327
ꢀ
to by the downfield shifts of the Ca H signals (5.51, 5.15
and 5.95 ppm in the dimer, as against 5.22, 4.95 and
4.38 ppm in the monomer; see Table 1),[25] and also by the
IR spectrum recorded in CHCl3, which shows amide I?, IIII
and A bands at 1542, 1632 and 3310 cmꢀ1, respectively, close
to those of previously reported peptide nanotubes (the band
at 3413 cmꢀ1 corresponds to the non-hydrogen-bonded
amide proton).[14,15,22,26] Unfortunately no crystals suitable
for X-ray diffractometry were obtained, which prevented
crystallographic confirmation of dimerization.
CP-5/D-5
CP-6/D-6
1627 (1669)
1642 (1681)
1525
1515
CP-7
CP-8/D-8
1652, 1687
1616, 1651 (1683)
1537
1543
CP-9
1629 (1671)
1655, 1684
1521
1530
CP-10
[a] Values in parentheses are for putative amide I? bands arising from
non-hydrogen-bonding N-methylated residues. [b] Upper values corre-
spond to hydrogen-bonded residues, lower values non-hydrogen-bonded
residues.
Peptide CP-5 behaved in the same way as CP-4 in polar
solvents such as deuterated DMSO, its 1H NMR spectra
showing the presence of several slowly interconverting con-
formers. Spectra run in nonpolar solvents such as CDCl3
showed the presence of two slowly interconverting species,
3
and their JN,H coupling constants of 7.1–8.2 Hz were consis-
Conclusive confirmation of the AA-type dimerization of
CP-5 was provided by X-ray crystallography of the colour-
less prismatic crystals obtained by equilibration of a solution
of the peptide in dichloromethane with hexane vapour.
These crystals consisted of b-sheet dimers composed of es-
sentially flat, slightly elliptical, antiparallel monomers linked
by a b-sheet-like array of eight hydrogen bonds with N···O
distances of 2.87–3.02 Š (Figure 5). The lumen of one of the
monomers has approximate maximum and minimum van
der Waals diameters of 8.83 Š (between Ala Cas) and
4.34 Š (between Ach Cb protons), respectively, while those
of the other are 9.16 and 3.81 Š. The cavity of the dimer has
a van der Waals volume of approximately 167 Š3. The crys-
tals are body-centred, with the axes of the dimers approxi-
mately parallel to the a axis and each dimer making van der
Waals contacts with its neighbours in this direction through
its a,g-face carbonyl and N-methyl groups. This body-cen-
tred structure creates channels along the a axis that are
composed of two alternating types of segment with a period
of about 13.8 Š: the dimer pore, and the space surrounded
by the external surfaces of four coplanar dimers (Figure 5b).
tent with the all-trans geometry required for dimerization.
As in the case of CP-4, the conjecture that one of the two
Table 1. Summary of NMR data for 3a,g-peptides CP-4 and CP-5.
Peptide[a]
Aa
H
ddimer (3JNa Hz)
Dmon (3JNa Hz)
CP-4
MeN-Ala
Ha
5.51
3.26–3.23
8.75
5.15
5.95
3.26–3.23
8.21
3.70
5.22
3.07–2.91
7.02
4.95
4.38
3.07–2.91
7.66
3.83
ꢀ
(233 K)
N Me
ꢀ
Leu
N H
Ha
Ha
MeN-Ala
Ach
ꢀ
N Me
ꢀ
N H
Hg
ꢀ
CP-5
Phe
N H
Ha
Ha
8.80 (7.5)
5.48
5.89 (5.9)
3.06
8.31 (8.2)
5.26
4.39
7.10
4.86
5.16
2.61
7.49 (7.1)
4.86
4.39
MeN-Ala
Phe
ꢀ
N Me
ꢀ
N H
Ha
Hg
MeN-Ach
ꢀ
N Me
2.72
2.52
Chem. Eur. J. 2008, 14, 2100 – 2111
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2103

J. R. Granja et al.
Effect of incomplete methylation on dimerization: In our
previous studies with eight-residue a,g-CPs, we found that
the CPs with N-methylated g-Achs dimerized through the
formation of eight hydrogen bonds between carbonyl and
ꢀ
N H components of a-Aas with association constants typi-
cally greater than 105 mꢀ1,[28] at least a thousand times great-
er than those of eight-residue a,g-CPs with N-methylated a-
Aas.[15] This has tentatively been attributed to steric interac-
tions between the N-methyl and carbonyl groups of the
latter, which might prevent the CP from adopting as flat a
conformation as is required for strong dimerization in the b-
strand form (effects of this kind have been observed in com-
putational studies of b-conformations of model peptides).[29]
The fact that in this work both CP-4 and CP-5 had associa-
tion constants of the same order as eight-residue a,g-CPs
with N-methylated a-Aas was therefore tentatively attribut-
ed to their both having residues of this type. To investigate
this, we prepared two 3a,g-CPs that would both “theoreti-
cally” undergo AA-type dimerization, but which both had
just two N-methylated residues on their a,g-faces: cyclo-[(l-
Ser(Bn)-d-MeN-g-Ach-l-Phe-d-Ala)2-] (CP-6), in which the
g-Achs are methylated, and cyclo-[(l-g-Ach-d-Phe-l-MeN-
Ala-d-Phe)2-] (CP-7), in which the N-methylated Aas are
the two a-Aas (Figure 7).[30] Compound CP-6 was expected
to undergo strong AA-type dimerization to b-band D-6,
since it has no N-methylated a-AAs, while compound CP-7
was expected to dimerize more weakly to D-7, because it
does have such residues.
Our expectations for compound CP-6 were totally fulfil-
led. That it adopts a flat, all-trans conformation in chloro-
Figure 5. a) Side view, and b) view down the axis, of dimer D-5, with the
1
3
form was corroborated by its H NMR JNa values (9.1, 7.6
and 6.5 Hz for Phe, Ser and Ala, respectively). The b-sheet-
type hydrogen-bonded interaction to form D-6 was con-
firmed by the positions of its Phe and Ser NH signals (8.66
and 8.58 ppm, respectively) and by its IR bands at 1642,
1515 and 3296 cmꢀ1.[15,22,26] The strength of this bonding was
shown by the chemical shifts of all its NH groups being un-
changed by addition of up to 30% of methanol or by dilu-
tion with chloroform to a concentration below 1”10ꢀ3 m,
which suggests total dimerization and an association con-
stant of at least 104 mꢀ1. Also, there appeared to be no addi-
tional association through the unmethylated a,g-face Ala
dimers of the next layer in gold.
As noted above, the spacing of the NH and C=O groups
of CP-4 and CP-5 (and their methylation) means that their
monomers must be antiparallel in their homodimers. By the
same token, heterodimers of these two species must be
formed as parallel monomers (Figure 6). To investigate the
possible parallel dimer formation and its relative stability
versus the antiparallel b-sheet-like dimer, we recorded the
1H NMR spectrum of an equimolecular mixture of CP-4 and
CP-5 in chloroform at 230 K. The spectrum showed the
presence of both D-4 and D-5, but no sign of heterodimer
D-4–5. This result clearly confirms the preference of SPN-
forming CPs for antiparallel dimerization.[22]
1
NH, the H NMR signal and IR stretching band of which
lay at 6.49 ppm and 3403 cmꢀ1, respectively.
Comparison of the results for CP-5 and CP-6 showed that
when the g-Ach was N-methylated, N-methylation of the
a,g-face a-Aas reduced the dimerization constant, as hy-
pothesized. There was no sign that the g-Ach N-methyl
groups of CP-6 hampered b-sheet formation. Quite different
results were obtained with CP-7 and no evidence of dimer
D-7 was found, suggesting that the g-Ach N-methyl groups
might not be irrelevant to CP conformation and dimeriza-
1
tion.[31] The H NMR spectrum of this compound in chloro-
form shows broad signals that are in some cases concentra-
tion-dependent (for example, the Phe CaH signal shifts from
5.00 to 4.60 ppm upon raising of the concentration from 3.75
to 15.0 mm). Notably, one of the concentration-dependent
Figure 6. Theoretical heterodimer resulting from parallel b-sheet-like hy-
drogen bonding of cyclic peptides CP-4 and CP-5.
2104
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2100 – 2111

Folding Control in Cyclic Peptides
FULL PAPER
ent, but the position of its Ser
NH signal—6.59 ppm—showed
that it was not involved in hy-
drogen bonding as it would be
in the AG-bonded dimer D-9.
Also, the unmethylated a-Aa
on its a,a-face, like the unme-
thylated
(7.97 ppm), appeared to be in-
volved in weak hydrogen
g-Ach
of
CP-7
a
bond, its NH signal lying at
7.3 ppm. An additional indica-
tion that CP-9 had not dimer-
ized was the location of its me-
thylated Ala CaH signal at sur-
prisingly high field: 3.54 ppm.
In contrast, the well defined
1H NMR spectrum of CP-8 in
chloroform showed that, like
CP-4, and apparently unlike
CP-9, it underwent AG-type di-
merization: Ser and g-Ach NH
signals appeared at positions in-
Figure 7. Compounds CP-6 and CP-7 and the corresponding theoretical AA-bonded dimers. Only peptide CP-
6 is able to form a stable dimer (D-6) in nonpolar solvents.
signals is that of Ach NH, which shows that the a,g-face of
CP-7, which would be expected to be solvent-exposed, is
not fully protected against hydrogen bonding by its N-me-
thylated alanines and suggests that the peptide does not
adopt the flat conformation required for the b-sheet forma-
tion. CP-7 also exhibited unexpected behaviour in polar sol-
vents capable of acting as hydrogen bond acceptors. Though
dicative of hydrogen bonding (8.70 and 7.91 ppm, respec-
tively), the corresponding JNa coupling constants (9.5 and
3
8.9 Hz, respectively) suggested that the hydrogen bonding
3
was of the b-sheet type, and the position and JNa value of
the NH signal of the unmethylated Ala (6.55 ppm and
6.5 Hz) furthermore showed that this a,a-face proton was
not involved in a hydrogen bond. The dimerization of CP-8
to D-8 was corroborated by its ROESY spectrum in CDCl3,
1
its H NMR spectrum is well defined and concentration-in-
dependent, the signal of only two of its four Phe NH pro-
tons lies in the region expected for strong hydrogen bonding
(at 8.81 ppm in deuteromethanol), and the position of the g-
Ach NH signal (7.68 ppm in deuteromethanol) shows that
this proton is also involved in hydrogen bonding. Addition-
ally, the chemical shifts of both amide NH protons do not
significantly shift to higher magnetic field relative to the
other NH protons with elevation of temperature.[32] All
these data suggested that the species present appeared not
to be the result of the expected almost flat CPs and raised
the question of the identity of the CP structure. Unfortu-
nately, two-dimensional NMR experiments failed to throw
further light on the structure of CP-7 in either polar or non-
polar solvents, and the failure of attempts to crystallize this
CP precluded crystallographic characterization. We shall
return below to the issue of the conformation and interac-
tions of this CP.
The suspicion that the effects of N-methylation went
beyond protection against multi-CP stacking or slight distor-
tion of more or less flat CPs led us to extend our investiga-
tion to 3a,g-CPs that were partially N-methylated on their
a,a-faces. Specifically, we prepared cyclo-[(l-g-Ach-d-Ala-l-
Ser(Bn)-d-MeN-Ala)2-] (CP-8) and cyclo-[(l-g-Ach-d-MeN-
Ala-l-Ser(Bn)-d-Ala)2-] (CP-9) (Figure 8).
Figure 8. Compounds CP-8 and CP-9 and the corresponding theoretical
AG-bonded dimers (D-8 and D-9). Only peptide CP-8 is able to form a
stable dimer (D-8) in nonpolar solvents. Inset: Theoretical double reverse
turn-type structure (F-9) proposed for CP-9.
1
CP-9 behaved similarly to CP-7. Its H NMR spectrum in
chloroform was well defined and concentration-independ-
Chem. Eur. J. 2008, 14, 2100 – 2111
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2105

J. R. Granja et al.
which showed cross-peaks indicative of close spatial proxim-
ity between the Ca protons of the N-methylated and unme-
thylated Ala residues (Figure 8), while the presence of b-
sheet-like hydrogen bonding was corroborated by the results
of IR studies in chloroform, which showed amide I, IIII and
A bands at 1543, 1616, 1651 and 3312 cmꢀ1 (as well as a
band at 3417 cmꢀ1 attributed to the stretching mode of the
unmethylated Ala NH).[15,22,26] The bonding in D-8 was
much stronger than in D-4: the chemical shifts of the Ser, g-
Ach and unmethylated Ala NH protons were unchanged by
dilution down to 2”10ꢀ4 m, showing that the association con-
stant for AG-type dimerization
tration-independent signals in methanol, in which the Ser
NH signal lay at 8.56 ppm and the g-Ach NH signal at
7.49 ppm. In fact, as with compound CP-7, neither the NMR
spectrum nor NOE experiments with CP-10 afforded suffi-
cient information to allow unequivocal inference of its struc-
ture from the data. Fortunately, however, we were able to
crystallize compound CP-10 from methanol and chloroform.
In crystals of CP-10 obtained from a solution in methanol
Figure 11a, the peptide adopts a structure akin to a pair of
type II’ b-turns,[31c,33,34] with a hydrogen bond between each
g-Ach carbonyl and the nonadjacent Ser(Bn) NH, and with
by CP-8 is at least 104 mꢀ1.
Methylation of the a-Aa nitro-
gen adjacent to the serine C=O
thus appeared not to impede
AG-type dimerization, while in
the presence of this methyl
group, additional methylation
of the a-Aa nitrogen adjacent
to the g-Ach C=O reduced the
strength of dimerization in the
same way as N-methylation of
Figure 9. CP-10 and its folded structure (F-10).
the a,g-face a-Aas appeared to
weaken the AA-type dimeriza-
tion of g-Ach-methylated CPs
such as CP-6.
Twisted CPs: Two-dimensional
NMR experiments carried out
to throw further light on the
structures adopted by CP-7 and
CP-9 failed to achieve their ob-
jective, and in both cases the
failures of attempts at crystalli-
zation also precluded crystallo-
graphic characterization. We
therefore investigated the be-
haviour of cyclo-[(l-Ser(Bn)-d-
g-Ach-l-Phe-d-MeN-Ala)2-]
(CP-10, Figure 9), in which two
of the Phe residues of CP-7 had
been replaced with benzylated
serines in the hope that it
would behave in the same way
as CP-7 but also provide struc-
tural information through the
chemical shifts of the protons
on Ser Ca and Cb and through
NOE cross-peaks.
In both polar and nonpolar
1
solvents the H NMR behaviour
of CP-10 was indeed similar to
that of CP-7 (Figure 10), with
broad, poorly resolved, concen-
tration-dependent signals in
Figure 10. Top: 1H NMR spectrum of a solution of peptide CP-10 in CDCl3 (7.4 mm) with the 4.0–5.5 ppm
region of the spectra of 74.0, 25.0 and 7.4 mm solutions, showing the downfield shift of Phe Ha. Bot-
tom: 1H NMR spectrum of a solution of peptide CP-10 in CD3OH (10 mm). [* CHCl3 (7.89 ppm), ** CH2Cl2
chloroform and sharp, concen- (5.48 ppm)].
2106
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2100 – 2111

Folding Control in Cyclic Peptides
FULL PAPER
Ramachandran plot coordinates (f,y) of (60.4,ꢀ107.0) or
(55.3,ꢀ107.4) for Phe (the two turns are not exactly identi-
cal) and (ꢀ110.0,47.9) or (ꢀ105.4,38.0) for MeN-Ala.[35] This
double reverse turn structure avoids steric interactions be-
tween the N-methyl and carbonyl groups of MeN-Ala (the
equilibrium between F-10 and fragments of this helical
structure in chloroform may explain the concentration de-
pendence of NMR spectra of CP-10 in this solvent
(Figure 10); fitting of a LaPlanche model[37] of the relation-
ship between total peptide concentration and the chemical
shift of Phe CaH afforded an value of 420mꢀ1 for the associ-
ation constant at 298 K, and Van’t Hoff plots for the 273–
323 K range afforded values of ꢀ27.5 kJmolꢀ1 for DHA298
and ꢀ42.0 JKꢀ1 molꢀ1 for DSA298 (Table S1 in the Supporting
Information).
ꢀ
N Me and C=O bonds making an angle of 528 with each
other) and is stabilized not only by the b-turn hydrogen
bonds but also by a pair of hydrogen bonds between the two
g-Achs. Layers of these double reverse turn structure are
formed, in which methanol and water molecules form hy-
drogen bonds with carbonyl groups to stabilize the crystal
structure (See Figure S2 in the Supporting Information).
When crystallized from chloroform (10% CS2), CP-10
adopts a structure like a couple of g-turns (F-10), with a hy-
drogen bond between each Phe CO and the NH group of
the Ser(Bn) on the other side of the adjacent MeN-Ala,
which has (f,y) coordinates of (93.7,ꢀ75.0) (Fig-
ure 11b).[33,36] As in the crystals obtained from methanol,
there are also hydrogen bonds between the two g-Ach resi-
dues. The receptor of the Phe NH is a Ser(Bn) belonging to
another molecule of CP-10, giving rise, together with a
water molecule, to the formation of a helical supramolecular
structure with four CPs per turn (Figure 11c and d). A fast
The likelihood that CP-7 may adopt a double reverse turn
structure similar to that of CP-10 in chloroform is supported
1
by the similar concentration dependence of its H NMR sig-
nals, and by its IR spectrum in chloroform having a band at
1687 cmꢀ1, a position typical of reverse turns and close to
the band at 1684 cmꢀ1 in the IR spectrum of CP-10. The
possibility that CP-9 may also adopt a double reverse turn
conformation (F-9) is suggested by the positions (in chloro-
form solution) of its g-Ach NH signal (7.97 ppm) and its
a,g-face exposed Ala NH (7.3 ppm), which suggest that both
are involved in hydrogen bonding, while the Ser NH with its
signal at 6.59 ppm is not, which is consistent with the double
turn conformation adopted by CP-10. Interestingly, CP-9
1
does not display concentration dependence of its H NMR
signals in nonpolar solvent, suggesting that the differences
in the methylation patterns of those peptides (the methyl
group in peptide CP-9 is placed in the a-Aa after the g-Ach,
while in CP-10 it is in the middle of the three a-Aas) might
prevent the reverse turn structure self-assembling into a tyli-
gomeric structure. Additionally all three peptides (CP-7, 9
and 10) display dramatic conformational changes when dis-
solved in polar solvents such as methanol, denoted by
1H NMR spectra with sharp and well defined signals, with
some of the NH groups involved in strong hydrogen-bond-
ing interactions (Phe NH at 8.81 ppm for peptide CP-7, Ser
NH at 8.56 ppm and g-Ach NH at 7.49 ppm for peptide CP-
10, g-Ach NH at 9.20 ppm for peptide CP-9) and in all three
the g-Ach NH (7.68 ppm for 7, 7.49 ppm for CP-10, and
9.20 ppm for CP-9) are also involved in hydrogen-bonding
interactions. All these data are consistent with the double b-
turn conformation found for CP-10. These more closely tied
structures in polar solvents must result from hydrophobic in-
teraction between the two cyclohexyl moieties assisted by
the formation of hydrogen bonds by the C=O and NH
groups of the g-Ach residues.
Conclusion
In this work we studied the behaviour of a new class of
cyclic peptide—3a,g-CPs—that were designed with overall
C2 symmetry with a view to the formation of SPNs featuring
two functionalizable hydrophobic bands on opposite sides of
their cavities. NMR, IR, MS, and X-ray diffraction data
show that, depending on their backbone N-methylation pat-
terns and the medium, monomers of eight-residue 3a,g-CPs
can remain as flat rings or fold into double reverse turns,
Figure 11. a) A monomer of CP-10 as crystallized from methanol. b) A
monomer of CP-10 as crystallized from chloroform. c) Axial and d) lat-
eral views of the helical supramolecular structure of CP-10 as crystallized
from chloroform, showing each monomer of the four-monomer pitch in a
different colour, with the amide groups in blue (nitrogen), red (oxygen)
and white (hydrogen).
Chem. Eur. J. 2008, 14, 2100 – 2111
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2107

J. R. Granja et al.
cyclo-[(l-g-Ach-d-MeN-Ala-l-Leu-d-MeN-Ala)2-] (CP-4):
A solution of
that flat monomers dimerize through sets of hydrogen bonds
of antiparallel b-sheet type, forming a “b-band”, and that
monomers forming double g-turns associate to form helical
supramolecular structures.
Boc-[(l-Leu-d-MeN-Ala-l-g-Ach-d-MeN-Ala)2-]-OFm (124 mg, 0.10 mmol)
in piperidine in CH2Cl2 (20%, 1 mL) was stirred at RT for 30 min. After
removal of the solvent, the residue was dissolved in CH2Cl2 and washed
with citric acid (10%), dried over Na2SO4, filtered, and concentrated.
The resulting material was dissolved in TFA/CH2Cl2 (1:1) and stirred at
RT for 10 min. After removal of the solvent, the residue was dried under
high vacuum for 2 h and used without further purification. The linear
peptide was dissolved in CH2Cl2 (165 mL, 0.6 mm) and treated with
TBTU (35 mg, 0.11 mmol), followed by dropwise addition of DIEA
(75 mL, 0.44 mmol). After 12 h, the solvent was removed under reduced
pressure and the crude product was purified by HPLC (Phenomenex
N-Methylation just of their a,g-face g-Achs (as in CP-6),
or of their a,a-face a-Aas with backbone nitrogens that are
not adjacent to a g-Ach C=O (as in CP-8), appears to
ensure that eight-residue 3a,g-CP monomers are flat and
form tight dimeric b-bands. Additional N-methylation of the
other residues on the same face (as in CP-4 and CP-5)
weakens dimerization, probably due to deviation from flat-
ness caused by steric interaction between the N-methyl and
C=O groups of the “newly” methylated a-Aas. N-Methyla-
tion of these other residues in the absence of the flatness-
enforcing methyls (as in CP-7, 9 and 10) appears to result in
the monomers forming twisted, double reverse turn struc-
tures comprising a pair of b-turns (in methanol) or g-turns
(in chloroform). Monomers of CP-10 with g-turns associate
in helical supramolecular structures. Given the structural
and functional importance of reverse turns in molecular
biology,[38] it is possible that the reverse-turn-forming pro-
pensity of these 3a,g-CPs may find application in studies of
biologically significant peptides.
Maxsil 10, CH2Cl2/MeOH), affording CP-4 as
a white solid (27 mg,
33%). 1H NMR (CDCl3, 500 MHz, 233 K): d=8.75 (brs, 1–0H; NH-
Leudimer), 8.21 (brs, 1–0H; NH-g-Achdimer), 7.66 (brs, 0–1H; NH-g-
Achmon), 7.02 (brs, 0–1H; NH-Leumon), 5.95 (m, 0–1H; HaMeN-Aladimer),
5.51 (m, 0–1H; Ha-MeN-Aladimer), 5.22 (m, 0–1H; Ha-MeN-Alamon), 5.15
(m, 0–1H; Ha-Leudimer), 4.95 (m, 0–1H; Ha-MeN-Leumo), 4.38 (m, 0–1H;
Ha-Alamon), 3.83 (m, 0–1H; Hg-Achmon), 3.70 (m, 0–1H; Hg-Achdimer),
3.26 (s, 0–3H; NMedimer), 3.23 (s, 0–3H; NMedimer), 3.07 (s, 0–6H;
NMemon), 2.91 ppm (s, 0–6H; NMemon); FTIR (293 K, CHCl3): n˜ = 3413
and 3310 (amide A), 2955, 2937, 2860, 1673, 1632 (amide I), 1542 cmꢀ1
(amide IIII); MALDI-TOF: m/z (%): 817.5 (40) [M+H]+, 839.6 (49)
[M+Na]+, 855.6 (100) [M+K]+; MALDI-TOF-HRMS: m/z (%): calcd
for: 817.55461 [M+H]+; found: 817.55054.
cyclo-[(l-MeN-g-Ach-d-Phe-l-MeN-Ala-d-Phe)2-] (CP-5): This compound
was prepared in the same way as CP-4, from Boc-[(l-MeN-g-Ach-d-Phe-
l-MeN-Ala-d-Phe)2-]-OFm (210 mg, 0.158 mmol). Yield after HPLC pu-
rification: 35 mg (21%). 1H NMR (CDCl3, 500 MHz): d=8.80 (d, J=
7.5 Hz, 1–0H; NH-Phedimer), 8.31 (d, J=8.2 Hz, 1–0H; NH-Phedimer), 7.49
(d, J=7.1 Hz, 0–1H; NH-Phemon), 7.32–7.19 (m, 5H; Ar), 7.10 (0–1H;
NH-Phemon), 5.89 (d, J=5.9 Hz, 0–1H; Ha MeN-Aladimer), 5.48 (m, 0–1H;
Ha Phedimer), 5.26 (m, 0–1H; Ha Phedimer), 5.16 (m, 0–1H; Ha MeN-
Alamon), 4.86 (m, 0–2H; Ha Phemon), 4.39 (m, 0–1H; Hg g-Ach), 3.06 (s,
0–3H; MeN-Aladimer), 2.92 (m, 4H; CH2b Phe), 2.72 (s, 0–3H; MeN-g-
Achdimer), 2.61 (s, 0–3H; MeN-Alamon), 2.52 ppm (s, 0–3H; MeN-g-Achmon);
FTIR (293 K, CHCl3): n˜ =3310 (amide A), 2934, 2860, 1669, 1627
(amide I), 1525 cmꢀ1 (amide IIII); MS-FAB+: m/z (%): 1037.7 (100)
[M+H]+; HRMS: m/z (%): calcd for: 1037.59454 [M+H]+; found:
1037.59087.
Experimental Section
General: 2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluor-
ophosphate (HBTU), 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluroni-
um tetrafluoroborate (TBTU), 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetrame-
thyluronium hexafluorophosphate (HATU), Boc-phenylalanine, Boc-
serine and Boc-N-methylalanine were purchased from Novabiochem or
Advanced ChemTech. All reagents and solvents were used as received
unless otherwise noted. Proton nuclear magnetic resonance (1H NMR)
spectra were recorded on Varian Inova 750 MHz, Varian Mercury
300 MHz, Bruker AMX 500 MHz, or Bruker WM 250 MHz spectrome-
ters. Chemical shifts are reported in parts per million (ppm, d) relative to
cyclo-[(l-Ser(Bn)-d-MeN-g-Ach-l-Phe-d-Ala)2-] (CP-6): This compound
was prepared in the same way as CP-4, from Boc-[(l-Ser(Bn)-d-MeN-g-
Ach-l-Phe-d-Ala)2-]-OFm (130 mg, 0.09 mmol). Yield after HPLC purifi-
1
1
cation: 46 mg (35%). H NMR (CDCl3, 750 MHz): d=8.66 (d, J=7.6 Hz,
tetramethylsilane (d=0.00 ppm). H NMR splitting patterns are designat-
1H; NH-Phe), 8.58 (d, J=9.1 Hz, 1H; NH-Ser), 7.23–7.11 (m, 10H; Ar),
6.49 (d, J=6.5 Hz, 1H; NH-Ala), 5.36 (q, J=7.7 Hz, 1H; Ha Ser), 4.68
(m, 1H; Ha Phe), 4.50 (m, 1H; Hg g-Ach), 4.43 (m, 1H; Ha Ala), 4.36
(d, J=12.3 Hz, 1H; CH2-Bn), 4.21 (d, J=12.3 Hz, 1H; CH2-Bn), 3.31 (m,
1H; CH2-Ser), 3.26 (m, 1H; CH2-Ser), 3.08 (m, 1H; CH2-Phe), 2.95 (s,
3H; NMe), 2.82 (m, 1H; Ha-Ach), 2.76 (m, 1H; CH2-Phe), 0.95 ppm (d,
J=5.4 Hz, 3H; Me-Ala); 13C NMR (CDCl3, 62.83 MHz): d=175.0 (C=
O), 171.5 (C=O), 170.3 (C=O), 168.9 (C=O), 137.7 (C), 136.6 (C), 129.4
(CH), 128.4 (CH), 128.3 (CH), 127.5 (CH), 126.8 (CH), 72.7 (CH2), 71.1
(CH2), 54.6 (CH), 51.6 (CH), 47.3 (CH), 47.3 (CH), 43.0 (CH), 41.1
(CH2), 32.5 (CH2), 29.9 (CH3), 28.9 (CH2), 27.5 (CH2), 24.3 (CH2),
21.3 ppm (CH3); FTIR (293 K, CHCl3): n˜ =3403 and 3296 (amide A),
2934, 2863, 1681, 1642 ppm (amide I), 1515 cmꢀ1 (amide IIII); MALDI-
TOF: m/z (%): 1069.6 (13) [M+H]+, 1091.6 (100) [M+Na]+, 1107.5 (46)
[M+K]+; MALDI-TOF-HRMS: m/z (%): calcd for: 1091.55768
[M+Na]+; found: 1091.55636.
cyclo-[(l-g-Ach-d-Phe-l-MeN-Ala-d-Phe)2-] (CP-7): This compound was
prepared in the same way as CP-4, from Boc-[(l-g-Ach-d-Phe-l-MeN-
Ala-d-Phe)2-]-OFm (354 mg, 0.27 mmol). Yield after HPLC purification:
130 mg (48%). 1H NMR (CD3OH, 750 MHz): d=8.81 (s, 1H; NH-Phe),
7.68 (d, J=8.0 Hz, 1H; NH-Ach), 7.3–7.0 (m, 11H; NH-Phe and Ar),
4.90 (m, 1H; Ha-MeN-Ala), 4.66 (m, 1H; Ha-Phe), 4.32 (m, 1H; Ha-
Phe), 3.77 (m, 1H; Hg-Ach), 3.2–2.90 (m, 4H; CH2-Phe), 2.59 (m, 1H;
Ha-Ach), 2.49 (s, 3H; NMe), 2.15 (d, J=11.9 Hz, 1H; Ach), 2.03 (d, J=
ed as singlet (s), doublet (d), triplet (t), or quartet (q). All first-order
splitting patterns were assigned on the basis of the appearance of the
multiplet. Splitting patterns that could not easily be interpreted are desig-
nated as multiplet (m) or broad (br). Carbon nuclear magnetic resonance
(13C NMR) spectra were recorded on Varian Mercury 300 MHz or Bru-
ker WM 250 MHz spectrometers. Carbon resonances were assigned by
use of distortionless enhancement by polarization transfer (DEPT) spec-
tra obtained with phase angles of 1358. Mass spectra (MS) were obtained
with Bruker Autoflex MALDI-TOF and Micromass Autospec mass spec-
trometers. Crystallographic data were collected in
a Bruker–Nonius
FR591-Kappa CCD2000 diffractometer or a Bruker SMART 1000 fitted
with a sealed tube (Mo Ka) and a highly ordered graphite monochroma-
tor. FTIR measurements with 5–10 mm solutions in CHCl3 placed in
NaCl solution IR cells were made on a JASCO FT/IR-400 spectropho-
tometer. Column chromatography was performed on EM Science silica
gel 60 (230–400 mesh). Solvent mixtures for chromatography are report-
ed as v/v ratios. HPLC was carried out on Phenomenex Maxsil 10 silica
columns with CH2Cl2/MeOH gradients between 100:0 and 90:10. CH2Cl2
and pyridine used as reaction solvents were distilled from CaH2 over
argon immediately prior to use.
Peptide synthesis: Unless otherwise noted, linear peptide synthesis was
carried out by Boc solution-phase synthesis as previously described.[15]
The resulting fully protected linear peptides were purified by flash
column chromatography (CH2Cl2/MeOH).
2108
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2100 – 2111

Folding Control in Cyclic Peptides
FULL PAPER
11.7 Hz, 1H; Ach), 1.92 (d, J=13.1 Hz, 1H; Ach), 1.77 (d, J=10.0 Hz,
1H; Ach), 1.72 (d, J=12.2 Hz, 1H; Ach), 1.50 (d, J=13.1 Hz, 1H; Ach),
1.44 (d, J=12.1 Hz, 1H; Ach), 1.29 (d, J=10.1 Hz, 1H; Ach), 0.74 ppm
(d, J=7.1 Hz, 3H; Me-Ala); 13C NMR (CD3OD, 75.4 MHz): d=178.1
(C=O), 175.6 (C=O), 173.7 (C=O), 172.4 (C=O), 137.4 (C), 136.9 (C),
130.4 (CH), 129.9 (CH), 129.6 (CH), 129.5 (CH), 128.2 (CH), 128.0
(CH), 57.9 (CH), 53.7 (CH), 52.6 (CH), 48.9 (CH), 42.9 (CH), 40.9
(CH2), 38.3 (CH2), 35.5 (CH2), 32.4 (CH2), 31.9 (CH2), 31.4 (CH3), 25.8
(CH2), 12.6 ppm (CH3); FTIR (293 K, CHCl3): n˜ =3324 (amide A), 2939,
2860, 1687, 1652 (amide I), 1537 cmꢀ1 (amide IIII); MS-FAB+: m/z (%):
1009.5 (100) [M+H]+, 1031.5 (11) [M+Na]+; HRMS: m/z (%): calcd for:
1009.55514 [M+H]+; found: 1009.55409.
[M+Na]+, 1107.5 (48) [M+K]+; MALFI-TOF-HRMS: m/z (%): calcd
for: 1091.5577 [M+Na]+; found: 1091.5527.
1H NMR assignments of cyclic peptides: The 1H NMR spectra of pep-
tides were identified with the aid of the corresponding double-quantum-
filled 2D COSY (2QF-COSY) and/or NOESY and ROESY spectra,
which were recorded at the indicated concentration and temperature.
Mixing times (ꢁ250 ms and 500 ms) were not optimized. Because of con-
formational averaging on the NMR timescale, CPs with C2 sequence sym-
metry generally have 1H NMR spectra that reflect the C2 symmetry of
the monomeric species and the D2 symmetry of the dimeric species.
Measurement of association constants by variable-concentration
1H N MR: The HPLC-purified peptides CP-4, CP-5 and CP-10 were dis-
solved in CDCl3 and their association constants (Ka) were estimated as
described in refs. [24,37].
cyclo-[(l-g-Ach-d-Ala-l-Ser(Bn)-d-MeN-Ala)2-] (CP-8): This compound
was prepared in the same way as CP-4, from Boc-[(l-Ser(Bn)-d-MeN-Ala-
l-g-Ach-d-Ala)2-]-OFm (140 mg, 0.116 mmol). Yield after HPLC purifi-
Van’t Hoff analysis of dimerization: The HPLC-purified cyclic peptides
CP-4, CP-5 and CP-10 were each dissolved in CDCl3 at concentrations
ranging from 1 to 74 mm. 1H NMR spectra were recorded at intervals of
5–10 K over the 223–323 K temperature range. Ka was estimated at each
temperature as described in refs. [24,37]. Analysis of a plot of 1/T (K) vs.
1
cation: 74 mg (70%). H NMR (CDCl3, 250 MHz): d=8.70 (d, J=9.5 Hz,
1H; NH Ser), 7.91 (d, J=8.9 Hz, 1H; NH g-Ach), 7.26–7.15 (m, 10H;
Ar), 6.55 (d, J=6.5 Hz, 1H; NH-Ala), 5.54 (dd, J=7.0 Hz, 1H; Ha-Ser),
5.34 (q, J=7.1 Hz, 1H; Ha-MeN-Ala), 5.06 (t, J=6.6 Hz, 1H; Ha-Ala),
4.41 and 4.28 (d, J=12.5 Hz, 2H; CH2-Ser), 3.72 (m, 1H; Hg-Ach), 3.39
lnKa then afforded the values DHA298 =ꢀ41.5 kJmolꢀ1 and DSA298
ꢀ152.5 JKꢀ1 molꢀ1 for CP-4, DHA298 =ꢀ47.8 kJmolꢀ1 and DSA298
ꢀ126.1 JKꢀ1 molꢀ1 for CP-5, and DHA298 =ꢀ27.5 kJmolꢀ1 and DSA298
ꢀ42.0 JKꢀ1 molꢀ1 for CP-10.
=
=
=
(5H; CH -Ser and MeN-Ala), 2.09 (m, 1H; Ha-Ach), 1.30 (s, 3H; Me-
2
Ala) and 1.28 ppm (s, 3H; Me-MeN-Ala); 13C NMR (CDCl3, 62.83 MHz):
d= 173.9 (C=O), 172.3 (C=O), 171.5 (C=O), 169.9 (C=O), 137.5 (C),
128.4 (CH), 127.7 (CH), 127.1 (CH), 72.9 (CH2), 71.0 (CH2), 51.1 (CH),
48.9 (CH), 47.4 (CH), 46.4 (CH), 45.1 (CH), 34.4 (CH2), 32.7 (CH2), 31.4
(CH3), 29.0 (CH2), 24.1 (CH2), 20.0 (CH3), 17.2 ppm (CH3); FTIR
(293 K, CHCl3): n˜ =3417 and 3312 (amide A), 2935, 2862, 1683, 1651,
1616 (amide I), 1543 cmꢀ1 (amide IIII); MS-FAB+: m/z (%): 917.4 (100)
[M+H]+, 1833.4 (3) [M+H]+; HRMS: m/z (%): calcd for: 917.51367
[2M+H]+; found: 917.51484.
Preparation of single crystals for X-ray analysis: HPLC-purified peptide
CP-5 (3 mg) was dissolved in CH2Cl2 (1 mL) and equilibrated by vapour-
phase diffusion against hexanes (5 mL), which resulted in spontaneous
crystallization after 2–4 d.
HPLC-purified peptide CP-10 (3 mg) was dissolved in a mixture of
CHCl3 and CS2 (10:1, 1 mL) and equilibrated by vapour-phase diffusion
against hexanes (5 mL), which resulted in spontaneous crystallization
after 3–4 d.
cyclo-[(l-g-Ach-d-MeN-Ala-l-Ser(Bn)-d-Ala)2-] (CP-9): This compound
was prepared in the same way as CP-4, from Boc-[(l-Ser(Bn)-d-Ala-l-g-
Ach-d-MeN-Ala)2-]-OFm (71.6 mg, 0.060 mmol). Yield after HPLC purifi-
cation: 15 mg (28%). 1H NMR (CDCl3, 500 MHz): d=7.97 (d, J=
8.80 Hz, 1H; NH-Ach), 7.37–7.27 (m, 6H; NH-Ala and Ar), 6.59 (d, J=
8.3 Hz, 1H; NH-Ser), 4.61 (m, J=6.4 Hz, 2H; CH2-Bn), 4.58–4.51 (m,
2H; Ha-Ala and Ser), 4.25 (m, 1H; Hg-Ach), 3.96 (dd, J=6.1 and
10.1 Hz, 1H; CH2-Ser), 3.78 (dd, J=5.1 and 10.1 Hz, 1H; CH2-Ser), 3.54
(q, J=7.0 Hz, 1H; Ha-MeN-Ala), 3.21 (s, 3H; NMe), 2.91 (m, 1H; Ha-
Ach), 1.50 (d, J=7.0 Hz, 3H; Me-MeN-Ala), 1.35 ppm (d, J=6.9 Hz, 3H;
Me-Ala); 13C NMR (CDCl3, 62.83 MHz): d= 179.2 (C=O), 173.2 (C=O),
170.5 (C=O), 169.2 (C=O), 137.7 (C), 128.5 (CH), 127.9 (CH), 127.8
(CH), 73.4 (CH2), 66.9 (CH2), 61.8 (CH3), 52.7 (CH), 49.0 (CH), 42.9
(CH), 38.7 (CH), 31.7 (CH), 29.5 (CH2), 29.2 (CH2), 26.7 (CH2), 17.6
(CH3), 16.7 (CH2), 13.6 ppm (CH3); FTIR (293 K, CHCl3): n˜ =3432 and
3326 (amide A), 2935, 2855, 1671, 1629 (amide I), 1521 cmꢀ1 (amide IIII);
MALDI-TOF: m/z (%): 917 (9) [M+H]+, 939 (100) [M+Na]+; MALDI-
TOF-HRMS: m/z (%): calcd for: 939.4952 [M+Na]+; found: 939.4951.
HPLC-purified peptide CP-10 (5 mg) was dissolved in MeOH (0.5 mL),
resulting in spontaneous crystallization after 2 d.
X-ray crystallographic analysis: Data were collected either on a Bruker
Nonius FR591 KappaCCD2000 fitted with an FR591 rotating anode
(CuKa) and Osmic multilayer confocal optics (CP-5 and CP-10 from
CHCl3) or on a Bruker SMART 1000 fitted with a sealed tube (MoKa
)
and a highly ordered graphite monochromator at low temperature (CP-5
at 100 K, CP-10 (CHCl3) at 100 K and CP-10 (MeOH) at 120 K). All cal-
culations were performed on an IBM-compatible PC with use of the pro-
grams COLLECT,[40] HKL Denzo and Scalepack,[41] ORTEP3,[42]
PARST,[43] PLATON (SQUEEZE),[44] SORTAV,[45] SADABS and
SHELX-97,[46] SIR2002,[47] SMART[48] and WinGx.[49]
CCDC 612329 (CP-5), 612330 [CP-10·CHCl3] and 612331 [CP-10·MeOH]
contain the supplementary crystallographic data for this paper. These
data can be obtained free of charge from the Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
cyclo-[(l-Ser(Bn)-d-g-Ach-l-Phe-d-MeN-Ala)2-] (CP-10): This compound
was prepared in the same way as CP-4, from Boc-[(l-Ser(Bn)-d-g-Ach-l-
Phe-d-MeN-Ala)2-]-OFm (71 mg, 0.05 mmol). Yield after HPLC purifica-
tion: 20 mg (37%). 1H NMR (CD3OH, 750 MHz): d=8.56 (s, 1H; NH-
Ser), 7.49 (d, J=8.5 Hz, 1H; NH-Ach), 7.4–7.2 (m, 11H; NH Ar), 5.15
(q, J=7.0 Hz, 1H; Ha-MeN-Ala), 4.63 (m, 1H; Ha-Ser), 4.50 (d, J=
12.7 Hz, 1H; CH2-Bn), 4.40 (m, 1H; Ha-Phe), 4.22 (d, J=12.7 Hz, 1H;
CH2-Bn), 3.66 (m, 1H; Hg-Ach), 3.42–3.35 (m, 4H; CH2-Phe), 3.12 and
3.00 (m, 2H; CH2-Ser), 2.53 (s, 3H; NMe), 2.33 (m, 1H; Ha-Ach), 1.95
(d, J=11.9 Hz, 1H; Ach), 1.83 (d, J=12.8 Hz, 1H; Ach), 1.78 (d, J=
12.5 Hz, 1H; Ach), 1.62 (d, 2H; Ach), 1.39 (d, 1H; Ach), 1.16 (d, J=
12.9 Hz, 1H; Ach), 0.86 ppm (d, J=7.0 Hz, 3H; Me-Ala); 13C NMR
(CD3OH, 75.4 MHz): d=178.0 (C=O), 175.7 (C=O), 172.9 (C=O), 171.3
(C=O), 138.9 (C), 137.2 (C), 130.4 (CH), 129.7 (CH), 129.5 (CH), 128.8
(CH), 128.7 (CH), 128.2 (CH), 73.8 (CH2), 70.0 (CH2), 56.4 (CH), 53.6
(CH), 52.5 (CH), 48.9 (CH) 43.1 (CH), 38.2 (CH2), 35.2 (CH2), 32.3
(CH2), 31.0 (CH3), 25.9 (CH2), 12.6 ppm (CH3); FTIR (293 K, CHCl3):
n˜ =3327 (amide A), 2941, 2862, 1684, 1655 (amide I), 1530 cmꢀ1 (ami-
de IIII); MALDI-TOF: m/z (%): 1069.5 (5) [M+H]+, 1091.5 (100)
Acknowledgement
This work was supported by the Spanish Ministry of Education and Sci-
ence (SAF2004-01044), by the Xunta de Galicia (projects PGIDIT06P-
XIB209018PR and GRC2006/132) jointly with the European Regional
Development Fund, and through a Spanish Ministry of Education and
Science FPU grant to M.A. We also thank Antonio L. Llamas Saíz for X-
ray crystallography.
[1] a) R. Langer, D. A. Tirrell, Nature 2004, 428, 487–492; b) D. Cui, H.
Gao, Biotechnol. Prog. 2003, 19, 683–692; c) T. M. Allen, P. R.
Cullis, Science 2004, 303, 1818–1822; d) R. Langer, Science 2001,
293, 58–59; e) J. P. Vacanti, R. Langer, Lancet 1999, 354, SI32–34;
f) S. Fçster, M. Konrad, J. Mater. Chem. 2003, 13, 2671–2688; g) S.
Fçster, T. Plantenberg, Angew. Chem. 2002, 114, 712–739; Angew.
Chem. Int. Ed. 2002, 41, 688–714.
Chem. Eur. J. 2008, 14, 2100 – 2111
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2109

J. R. Granja et al.
[2] a) S. H. Gellman, Acc. Chem. Res. 1998, 31, 173–180; b) D. J. Hill,
M. J. Mio, R. B. Prince, T. S. Hughes, J. S. Moore, Chem. Rev. 2001,
101, 3893–4011; c) R. P. Cheng, S. H. Gellman, W. F. DeGrado,
Chem. Rev. 2001, 101, 3219–3232; d) A. R. Sanford, B. Gong, Curr.
Org. Chem. 2003, 7, 1649–1659.
[3] For examples of reversible folding properties of foldamers induced
by changes in temperature, solvent, pH or light, see: a) A. Khan, C.
Kaisher, S. Hecht, Angew. Chem. 2006, 118, 1912–1915; Angew.
Chem. Int. Ed. 2006, 45, 1878–1881; b) A. Khan, C. Kaisher, S.
Hecht, Angew. Chem. 2003, 115, 6203–6206; Angew. Chem. Int. Ed.
2003, 42, 6021–6024; c) C. Dolain, V. Maurizot, I. Huc, Angew.
Chem. 2003, 115, 2844–2846; Angew. Chem. Int. Ed. 2003, 42, 2738–
2740; ; d) E. Kolomiets, V. Berl, I. Odriozola, A.-M. Stadler, N. Kyr-
itsatas, J. M. Lehn, Chem. Commun. 2003, 2868–2869; e) M. Bar-
boiu, J. M. Lehn, Proc. Natl. Acad. Sci. USA 2002, 99, 5201–6206;
f) R. B. Prince, T. Okada, J. S. Moore, Angew. Chem. 1999, 111,
245–249; Angew. Chem. Int. Ed. 1999, 38, 233–236; .
[12] For an example of biotechnological applications of nanotubes, see:
C. R. Martin, P. Kohli, Nature Reviews Drug Discovery 2003, 2, 29–
37.
[13] a) For transmembrane transport of molecules or ions by peptide
nanotubes, see: M. R. Ghadiri, J. R. Granja, L. K. Buehler, Nature
1994, 369, 301–304; J. R. Granja, M. R. Ghadiri, J. Am. Chem. Soc.
1994, 116, 10785–10786; J. Sanchez-Quesada, H. S. Kim, M. R. Gha-
diri, Angew. Chem. 2001, 99, 2571–2574; Angew. Chem. Int. Ed.
2001, 40, 2503–2506; J. Sanchez-Quesada, H. S. Kim, M. R. Ghadiri,
Angew. Chem. 2001, 113, 2571–2574; Angew. Chem. Int. Ed. 2001,
40, 2503–2506; b) for biological application of peptide nanotubes,
see: S. Fernµndez-Lꢃpez, H.-S. Kim, E. C. Choi, M. Delgado, J. R.
Granja, A. Khasanov, K. Kraehenbuehl, G. Long, D. A. Weinberger,
K. Wilcoxen, M. R. Ghadiri, Nature 2001, 412, 452–455; V. Dartois,
J. Sanchez-Quesada, E. Cabezas, E. Chi, C. Dubbelde, C. Dunn, C.
Gritzen, D. Weinberger, J. R. Granja, M. R. Ghadiri, T. R. Parr Jr. ,
Antimicrob. Agnt. Chemother. 2005, 49, 3302–3310; W. S. Horne,
C. M. Wiethoff, C. Cui, K. M. Wilcoxen, M. Amorín, M. R. Ghadiri,
G. R. Nemerow, Bioorg. Med. Chem. 2005, 13, 5145–5153; c) for
photoswitchable materials based on self-assembled peptides, see: C.
Steinem, A. Janshoff, M. S. Vollmer, M. R. Ghadiri, Langmuir 1999,
15, 3956–3964; M. S. Vollmer, T. D. Clark, C. Steinem, M. R. Gha-
diri, Angew. Chem. 1999, 111, 1703–1706; Angew. Chem. Int. Ed.
1999, 38, 1598–1601; d) for size-selective ion sensing based on
SAM-supported peptide nanotubes, see: K. Motesharei, M. R. Gha-
diri, J. Am. Chem. Soc. 1997, 119, 11306–11312; e) for hybrids of
peptide nanotubes with polymers, see: J. Couet, J. D. J. S. Samuel, A.
Kopyshev, S. Santer, M. Biesalski, Angew. Chem. 2005, 117, 3361–
3365; Angew. Chem. Int. Ed. 2005, 44, 3297–3301; f) for potential
applications in molecular electronics, see: N. Ashkenasy, W. S.
Horne, M. R. Ghadiri, Small 2006, 2, 99–102.
[14] a) M. R. Ghadiri, K. Kobayashi, J. R. Granja, R. K. Chadha, D. E.
McRee, Angew. Chem. 1995, 107, 76–78; Angew. Chem. Int. Ed.
Engl. 1995, 34, 93–95; b) D. T. Bong, M. R. Ghadiri, Angew. Chem.
2001, 113, 2221–2224; Angew. Chem. Int. Ed. 2001, 40, 2163–2166;
c) M. Saviano, A. Lombardi, C. Pedone, B. Di Blasio, X. C. Sun,
G. P. Lorenzi, J. Incl. Phenom. 1994, 18, 27–36; d) X. C. Sun, G. P.
Lorenzi, Helv. Chim. Acta 1994, 77, 1520–1526; e) W. S. Horne, N.
Ashkenasy, M. R. Ghadiri, Chem. Eur. J. 2005, 11, 1137–1144.
[15] a) M. Amorín, L. Castedo, J. R. Granja, J. Am. Chem. Soc. 2003,
125, 2844–2845; b) for eight-residue a,g-CPs, see: M. Amorín, L.
Castedo, J. R. Granja, Chem. Eur. J. 2005, 11, 6539–6547; c) for
four-residue a,g-CPs, see: M. Amorín, R. J. Brea, L. Castedo, J. R.
Granja, Org. Lett. 2005, 7, 4681–4684; d) for larger-size a,g-CPs,
see: R. J. Brea, L. Castedo, J. R. Granja, Chem. Com. 2007, 3267–
3269.
[16] For a,g-CP heterodimer formation and applications in electron- and
energy-transfer processes, see: R. J. Brea, M. Amorín, L. Castedo,
J. R. Granja, Angew. Chem. 2005, 117, 5856–5859; Angew. Chem.
Int. Ed. 2005, 44, 5710–5713; R. J. Brea, L. Castedo, J. R. Granja,
M. ꢄ. Herranz, L. Sanchez, N. Martín, W. Seitz, D. M. Guldi, Proc.
Natl. Acad. Sci. USA 2007, 104, 5291–5294.
[17] The flexibility of its 30-membered ring has been blamed for the
non-association of an N-methylated cyclic decapeptide composed of
alternating d- and l-a-amino acids: T. D. Clark, J. M. Buriak, K. Ko-
bayashi, M. P. Isler, D. E. McRee, M. R. Ghadiri, J. Am. Chem. Soc.
1998, 120, 8949–8962.
[18] For peptides made of alternating a-amino acid and 3-aminobenzoic
acid derivatives: as ion channels, a) H. Ishida, Z. Qi, M. Sokabe, K.
Donowaki, Y. Inoue, J. Org. Chem. 2001, 66, 2978–2989; as ion pair
receptors, b) S. Kubik, J. Am. Chem. Soc. 1999, 121, 5846–5855.
[19] a) Alkylation of the amide nitrogen reduces the energy difference
between the cis and trans forms: D. E. Stewart, A. Sarkar, J. E.
Wampler, J. Mol. Biol. 1990, 214, 253–260; b) N-modified amino
acids would influence polypeptides containing them by adopting an
extended conformation and ensuring the adoption of a similar con-
formation by their N-terminal neighbours: P. Manavalan, F. A.
Momany, Biopolymers 1980, 19, 1943–1973; but also promoting the
[4] For some examples of foldamer-based reactivity, see: a) R. A. Smal-
done, J. Moore, J. Am. Chem. Soc. 2007, 129, 5444–5450; b) T. J.
Peelen, C. Yonggui, S. H. Gellman, J. Am. Chem. Soc. 2005, 127,
11598–11599; c) C. Dolain, C. Zhan, J.-M. Daniels, I. Huc, J. Am.
Chem. Soc. 2005, 127, 2400–2401; d) J. M. Heemstra, J. Moore, J.
Am. Chem. Soc. 2004, 126, 1648–1649.
[5] a) T. Beke, I. G. Csizmadia, A. Perczel, J. Am. Chem. Soc. 2006, 128,
5158–5167; b) D. Seebach, A. K. Beck, D. J. Bierbaum, Chemistry &
Biodiversity. 2004, 1, 1111–1239; Biodiversity. 2004, 1, 1111–1239.
[6] For some examples of g-peptide foldamers, see: a) S. Hanessian, X.
Lou, R. Schaum, Tetrahedron Lett. 1999, 40, 4925–4929; b) M. G.
Woll, J. R. Lai, I. A. Guzei, S. J. C. Taylor, M. E. B. Smith, S. H.
Gellman, J. Am. Chem. Soc. 2001, 123, 11077–11078; c) D. Seebach,
A. Beck, M. Brenner, C. Gaul, A. Heckel, Chimia 2001, 6, 831–838;
d) D. Seebach, L. Schaeffer, M. Brenner, D. Hoyer, Angew. Chem.
2003, 99, 800–802; Angew. Chem. Int. Ed. 2003, 42, 776–778; e) J.
Farrera-Sinfreu, L. Zaccaro, D. Vidal, X. Salvatella, E. Giralt, M.
Pons, F. Albericio, M. Royo, J. Am. Chem. Soc. 2004, 126, 6048–
6057; f) G. V. M. Sharma, V. B. Jadhav, V. S. Ramakrishna, P. Jayap-
rakash, K. Narsimulu, V. Subash, A. C. Kunwar, J. Am. Chem. Soc.
2006, 128, 14657–14668.
[7] a) J. L. Price, W. S. Horne, S. H. Gellman, J. Am. Chem. Soc. 2007,
129, 6376–6377; b) D. S. Daniels, E. J. Petersson, J. X. Qiu, A. Sche-
partz, J. Am. Chem. Soc. 2007, 129, 1532–1533; c) N. Delsuc, J.-M.
Lꢂger, S. Massip, I. Huc, Angew. Chem. 2007, 99, 218–221; Angew.
Chem. Int. Ed. 2007, 46, 214–217; d) B.-C. Lee, R. N. Zuckermann,
K. A. Dill, J. Am. Chem. Soc. 2005, 127, 10999–11009; e) R. P.
Cheng, W. F. J. DeGrado J. Am. Chem. Soc. 2002, 124, 11564–
11565.
[8] M. Stefani, C. M. Dobson, J. Mol. Med. 2003, 81, 678–699, and ref-
erences therein.
[9] a) J. M. Smeenk, M. B. J. Otten, J. Thies, D. A. Tirrell, H. G. Stun-
nenberg, J. C. M. van Hest, Angew. Chem. 2005, 117, 2004–2007;
Angew. Chem. Int. Ed. 2005, 44, 1968–1971; b) M. R. Ghadiri, J. R.
Granja, R. A. Milligan, D. E. McRee, N. Khazanovich, Nature 1993,
366, 324–327; c) C. Whitehouse, J. Fang, A. Aggeli, M. Bell, R.
Brydson, C. W. G. Fishwick, J. R. Henderson, C. M. Knobler, R. W.
Owens, N. H. Thomson, D. A. Smith, N. Boden, Angew. Chem. 2005,
117, 2001–2004; Angew. Chem. Int. Ed. 2005, 44, 1965–1968.
[10] a) N. Khazanovich, J. R. Granja, D. E. McRee, R. A. Milligan, M. R.
Ghadiri J. Am. Chem. Soc. 1994, 116, 6011–6012; b) J. D. Hartger-
ink, J. R. Granja, R. A. Milligan, M. R. Ghadiri, J. Am. Chem. Soc.
1996, 118, 43–50; c) K. Rosenthal-Aiman, G. Svensson, A. Undꢂn,
J. Am. Chem. Soc. 2004, 125, 3372–3373.
[11] a) R. J. Brea, J. R. Granja, Self-Assembly of Cyclic Peptides in Hy-
drogen-Bonded Nanotubes, In Dekker Encyclopedia of Nanoscience
and Nanotechnology, Marcel Dekker Inc., 2004, pp. 3439–3457;
b) G. R. Patzke, F. Krumeich, R. Nesper, Angew. Chem. 2002, 114,
2554–2571; Angew. Chem. Int. Ed. 2002, 41, 2446–2461; c) D. T.
Bong, T. D. Clark, J. R. Granja, M. R. Ghadiri, Angew. Chem. 2001,
113, 1016–1041; Angew. Chem. Int. Ed. 2001, 40, 988–1011; d) P. M.
Ajayan, O. Z. Zhou, Top. Appl. Phys. 2001, 80, 391–425.
2110
www.chemeurj.org
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 2100 – 2111

Folding Control in Cyclic Peptides
FULL PAPER
formation of b-turns: c) Y. Takeuchi, G. R. Marshall J. Am. Chem.
Soc. 1998, 120, 5363–5372.
ing involved) change with the temperature. T. Cierpici, J. Otlewski,
J. Biomol. NMR 2001, 21, 249–261.
[33] a) J. A. Smith, L. G. Pease, CRC Crit. Rev. Biochem. 1980, 8, 315–
399; b) G. D. Rose, W. B. Young, L. M. Gierasch, Nature 1983, 304,
654–657; c) G. D. Rose, L. M. Gierasch, J. A. Smith, in Turns in Pep-
tides and Proteins, Vol. 37 (Eds.: C. B. Anfinsen, J. T. Edsall, F. M.
Richards), Academic Press, New York, 1985, pp. 1–109; d) J. S. Ri-
chardson, D. C. Richardson, in Prediction of Protein Structure and
Principles of Protein Conformation (Ed.: G. D. Fasman), Plenum,
New York, 1989, pp. 1–98; e) C. K. Smith, L. Regan, Acc. Chem.
Res. 1997, 30, 153–166; f) E. Lacroix, T. Kortemme, M. L. de la Paz,
L. Serrano, Curr. Opin. Struct. Biol. 1999, 9, 487–493.
[34] It has been shown that d-a-amino acids induce type II’ b-turns in so-
lution, see: a) V. Bobde, C. Sadasivan, S. Durani, Int. J. Peptide Pro-
tein Res. 1994, 44, 209–218; b) V. Bobde, S. Durani, V. Pattabhi, C.
Sadasivan, Int. J. Peptide Protein Res. 1995, 45, 332–336; c) Y. Yan,
B. W. Erickson, A. Tropsha, J. Am. Chem. Soc. 1995, 117, 7592–
7599.
[20] For a recent study of structural behaviour of N-methylated cyclic
peptide, see: J. Chatterjee, D. Mierke, H. Kessler, J. Am. Chem. Soc.
2006, 128, 15164–15172, and references therein.
[21] For a recent study on b-turns structures by insertion of b, g or d-
amino acids, see: R. Rai, P. G. Vasudev, K. Ananda, S. Raghothama,
N. Shamala, I. L. Karle, P. Balaram, Chem. Eur. J. 2007, 13, 5917–
5926.
[22] Even in the case of CPs formed of alternate l- and d-a-Aas, hydro-
gen bonding between antiparallel CPs is thermodynamically more
favoured than hydrogen bonding between parallel CPs: K. Kobaya-
shi, J. R. Granja, M. R. Ghadiri, Angew. Chem. 1995, 107, 79–81;
Angew. Chem. Int. Ed. Engl. 1995, 34, 95–98.
[23] J. K. M. Sanders, B. K. Hunter, Modern NMR Spectroscopy, Oxford
University Press, Oxford, 1987, p. 210.
[24] The dimerization constant (Ka) was estimated by fitting the equa-
tion: [P]t =Iobs/Ka
A
[35] b Hairpins design by insertion of b-, g- and d-amino acid residues at
the position i+2, see: R. S. Roy, H. N. Gopi, S. Raghothama, I. L.
Karle, P. Balaram, Chem. Eur. J. 2006, 12, 3295–3302; R. Rai, P. G.
Vaused, K. Anada, S. Raghothama, N. Shamala, I. L. Karle, P. Balar-
am, Chem. Eur. J. 2007, 13, 5917–5926.
(the ratio of the integral of the Leu (4) NH or Phe (6) CaH signal of
the dimer to the sum of those of dimer and monomer) at various
total peptide concentrations [P]t in the range (1.0–30.0)”10ꢀ3 m,
where Isat, the integral ratio at saturation, was also optimized in the
fitting process.
[36] a) B. L. Sibanda, J. M. Thornton, Nature, 1985, 116, 170–174;
b) C. M. Wilmot, J. M. Thornton J. Mol. Biol. 1988, 203, 221–223;
c) E. G. Hutchinson, J. M. Thornton, Prot. Sci. 1994, 3, 2207–2216;
d) C. Mattos, G. A. Petsko, M. Karplus, J. Mol. Biol. 1994, 238, 733–
747; e) R. L. Dunbrack, M. Karplus, Nat. Struct. Mol. Biol. 1994, 1,
334–340.
[37] A. M. Andrianov, Y. A. Sokolov, J. Biomol. Struct. Dyn. 2003, 20,
603–614.
[38] The chemical shift (dobs) of the CaH group was measured at concen-
trations c ranging from 7.4”10ꢀ3 to 7.5”10ꢀ2 m, and the association
constant (Ka), together with the shifts in the (dmonomer) and dimer
(ddimer), were estimated by using a nonlinear regression program to
[25] A. Pardi, M. Billeter, K. Wutrich, J. Mol. Biol. 1984, 180, 741–751.
[26] a) P. I. Haris, D. Chapman, Biopolymers (Peptide Sci.) 1995, 37,
251–263; b) S. Krimm, J. Bandekar, Adv. Protein Chem. 1986, 38,
181–364; c) J. Bandekar, Biochim. Biophys. Acta 1992, 1120, 123–
143; d) J. Kubelka, T. A. Keiderling, J. Am. Chem. Soc. 2001, 123,
12048–12058.
[27] a) M. S. Searle, M. S. Westwell, D. H. Williams, J. Chem. Soc. Perkin
Trans. 2 1995, 141–151; b) J. D. Dunitz, Chem. Biol. 1995, 2, 709–
712.
[28] For a recent study on dimerization constants of a,g-CPs by fluores-
cence measurements, see: R. J. Brea, M. E. Vµzquez, M. Mosquera,
L. Castedo, J. R. Granja, J. Am. Chem. Soc. 2007, 129, 1653–1657.
[29] a) D. Mçhle, H.-J. Hofman, J. Pept. Res. 1998, 51, 19–28; b) B.
Vitoux, A. Aubry, M. T. Cung, M. Marraud, Int. J. Pept. Protein Res.
1986, 27, 617.
[30] Because of the unavailability of d-Boc-Ser(Bn)-OH, we prepared
compound CP-7 as the enantiomer of the CP strictly corresponding
to peptides CP-6 and CP-10, using d-MeN-g-Ach instead of l-MeN-g-
Ach.
[31] a) Alkylation of the amide nitrogen reduces the energy difference
between the cis and trans forms to approximately 2.1 kJmolꢀ1 as
against around 10.9 kJmolꢀ1 for conventional amino acids: D. E.
Stewart, A. Sarkar, J. E. Wampler, J. Mol. Biol. 1990, 214, 253–260;
b) N-modified amino acids would influence the polypeptides con-
taining them by adopting an extended conformation and ensuring
the adoption of a similar conformation by their N-terminal neigh-
bours: P. Manavalan, F. A. Momany, Biopolymers 1980, 19, 1943–
1973; but also by c) promoting the formation of b-turns: W. Braun,
J. Kallen, V. Mikol, M. D. Walkinshaw, K. Wꢅthrich, FASEB 1995,
9, 63–72; N. Fusetani, S. Matsunaga, Chem. Rev. 1993, 93, 1793–
1806; Y. Takeuchi, G. R. Marshall, J. Am. Chem. Soc. 1998, 120,
5363–5372.
fit the (d, c) data with the equation dobs =ddimer+(dmonomerꢀddimer
)
1
=
2
[(1+8Kac) ꢀ1]/
(4Kac), L. A. LaPlanche, H. B. Thompson, M. T.
Rogers, J. Phys. Chem. 1965, 69, 1482.
[39] a) R. L. Stanfield, T. M. Fieser, R. A. Lerner, I. A. Wilson, Science
1990, 248, 712–719; b) G. R. Marshall, Curr. Opin. Struct. Biol.
1992, 2, 904–919.
[40] Nonius BV, Delft (The Netherlands) 1997–2000.
[41] Z. Otwinowski, W. Minor, Methods Enzymol. 1997, 276, 307–326.
[42] L. J. Farrugia J. Appl. Crystallogr. 1997, 30, 565.
[43] M. J. Nardelli, Appl. Crystallogr. 1995, 28, 659.
[44] A. L. Spek, PLATON, University of Utrecht (The Netherlands)
(2001); P. van der Sluis, A. L. Spek, Acta Crystallogr. Sect. A 1990,
46, 194–201.
[45] R. H. Blessing, Acta Crystallogr. Sect. A 1995, A51, 33–38.
[46] G. M. Sheldrick, Institute fꢅr Anorganische Chemie, Universitꢆt
Gçttingen, Gçttingen (Germany).
[47] M. C. Burla, M. Camalli, B. Carrozzini, G. L. Cascarano, C. Giaco-
vazzo, G. Polidori, R. Spagna, J. Appl. Crystallogr. 2003, 36, 1103.
[48] Bruker AXS, Madison (USA).
[49] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[32] Chemical shifts of NH protons involved in strong hydrogen-bonding
interactions stay unchanged with temperature change, while the
chemical shifts of solvent-exposed NH protons (no hydrogen bond-
Received: July 10, 2007
Revised: October 23, 2007
Published online: December 28, 2007
Chem. Eur. J. 2008, 14, 2100 – 2111
ꢁ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2111