The N-terminal nonapeptide of cephaibols A and C: A naturally occurring example of mismatched helical screw-sense control

Article abstract of DOI:10.1002/chem.201302648

The N-terminal nonapeptide domain of the fungal nonribosomal peptide antibiotics cephaibol A and cephaibol C (AcPheAib₄LeuIvaGly- Aib) is reported to adopt a right-handed helical conformation in the crystalline state. However, this conformation is at odds with the left-handed helicity observed in solution in related synthetic oligomers capped with Ac-L-PheAib₄ fragments. We report the synthesis of four diastereoisomers of the cephaibol N-terminal nonapeptide, and show by NMR and CD spectroscopy that the peptide containing the chiral amino acids Phe and Leu in the naturally occurring relative configuration exists in solution as an interconverting mixture of helical screw-sense conformers. In contrast, the nonapeptide containing the unnatural relative configuration at Phe and Leu adopts a single, stable helical screw-sense, which is left handed when the N-terminal Phe residue is L and right-handed when the N-terminal Phe residue is D.

Full text of DOI:10.1002/chem.201302648

FULL PAPER
DOI: 10.1002/chem.201302648
The N-Terminal Nonapeptide of Cephaibols A and C: A Naturally Occurring
Example of Mismatched Helical Screw-Sense Control
Ugo Orcel,[a] Matteo De Poli,[a] Marta De Zotti,[b] and Jonathan Clayden*[a]
Abstract: The N-terminal nonapeptide
domain of the fungal nonribosomal
peptide antibiotics cephaibol A and
thesis of four diastereoisomers of the
cephaibol N-terminal nonapeptide, and
show by NMR and CD spectroscopy
that the peptide containing the chiral
amino acids Phe and Leu in the natu-
rally occurring relative configuration
exists in solution as an interconverting
mixture of helical screw-sense con-
formers. In contrast, the nonapeptide
containing the unnatural relative con-
figuration at Phe and Leu adopts
cephaibol C
(AcPheAib4LeuIvaGly-
Aib) is reported to adopt a right-
handed helical conformation in the
crystalline state. However, this confor-
mation is at odds with the left-handed
helicity observed in solution in related
synthetic oligomers capped with Ac-l-
PheAib4 fragments. We report the syn-
a
single, stable helical screw-sense,
which is left handed when the N-termi-
nal Phe residue is l and right-handed
when the N-terminal Phe residue is d.
Keywords: conformation
mers · NMR spectroscopy · peptai-
biotics · peptides
· foldACHTNUTGNRUEaGN -
AHCTUNGTRENNUNG
Introduction
At first sight, it seems reasonable to assume that the
right-handed screw-sense preference in the solid state struc-
ture of this achiral pentapeptide segment arises from the ab-
solute configuration of the flanking l-Phe and l-Leu resi-
dues. Peptide helices containing l-amino acids typically
adopt a right-handed screw sense, and the two chiral resi-
dues may induce the unbiased remainder of the N-terminal
domain to adopt the same conformation. However, recent
studies have revealed that protected chiral amino acids such
as Ac-l-Phe, when located at the N-terminus of a peptide,
typically induce a left-handed screw sense in nearby achiral
residues.[17] This observation calls into question the origin of
the screw-sense preference at the N-terminus of the cephai-
bols, a preference that may play a crucial role in the antibac-
terial and anthelmintic activity of these and related antibiot-
ics.
The cephaibols belong to a broader class, known as the
peptaibols, of fungal non-ribosomal peptides containing be-
tween five and twenty amino acid residues.[18] Peptaibols are
characterized by a high content of non-proteinogenic, a,a-
dialkylated amino acids such as Aib or Iva (isovaline, or 2-
amino-2-methylbutyric acid), along with a C-terminal 1,2-
amino alcohol and (usually) acylation at the N-terminus.[19,20]
The broad spectrum of antibacterial activities of peptaibols
derives from their interaction with biological phospholipid
membranes.[21] Because these antimicrobial peptides per-
meabilize membranes without specificity towards a protein
target, the likelihood that bacterial resistance may develop
is low. Understanding their mechanism of action is thus cru-
cial for the development of peptaibols or their analogues as
useful anti-microbial agents, which could address the in-
creasingly serious, worldwide health problem of multidrug
antibiotic resistance.
Peptide chains of achiral amino acids, even though they con-
tain no asymmetric centres, may be induced into either
a left- or a right-handed helical conformation by an external
chiral influence. For example, homo-oligomers of the helico-
genic[1–3] quaternary amino acid Aib[4,5] (2-aminoisobutyric
acid), or hetero-oligomers of Aib and other achiral residues
(such as Gly,[6–8] Ac6c,[8–10] or dehydroamino acids[9,11]), may
be induced to adopt a preferred screw sense by chiral resi-
dues bonded either covalently[6,8–13] or non-covalently[13,14] at
their N or C terminus. A remarkable naturally occurring
achiral peptide motif of this type occurs towards the N-ter-
minus of the antibiotic fungal metabolites cephaibol A, C, D
and E (Figure 1).[15] These non-ribosomal peptides contain
a string of no less than five achiral amino acids (Aib-Aib-
Aib-Aib-Gly) which adopt, in the solid state, a right-handed
helical conformation.[16] The achiral pentapeptide segment is
capped at the N- and C-terminus by the chiral residues Ac-
l-Phe and l-Leu, respectively, with one or two further achi-
ral residues linking this AcPheAib4GlyLeu oligomer to the
proline/hydroxyproline-rich C-terminal domain of the 16-
residue chain.
[a] U. Orcel, Dr. M. De Poli, Prof. J. Clayden
School of Chemistry, University of Manchester
Oxford Road, Manchester M13 9PL (UK)
Fax : (+44)161-275-4939
E-mail: clayden@man.ac.uk
[b] Dr. M. De Zotti
Dipartimento di Scienze Chimiche, Universitꢀ di Padova
via Marzolo 1, 35131 Padova (Italy)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302648.
Chem. Eur. J. 2013, 19, 16357 – 16365
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
16357

Peptaibols fold into helical structures both in solution and
when membrane-bound.[19] The most studied member of this
family is alamethicin, a 20-residue pore-forming peptide
from the fungus Trichoderma viridans.[22,23] Depending on
peptide concentration and/or transmembrane potential, ala-
methicin forms pores by the so-called “barrel-stave” mecha-
nism, in which peptide helices insert in a transmembrane
sub-class, so the insertion of these additional methylene
groups seems to play a major role in the antimicrobial
action of these peptides.[15]
The three-dimensional structures of various peptaibols
have been determined both in the solid state and in solu-
tion.[29,30,32] These studies indicate that many of these mole-
cules adopt a helical conformation consisting of an N-termi-
nal domain followed by a bend at a proline or a hydroxypro-
line residue located close to the middle of the helix. While
the crystal structures of cephaibols A, B and C have already
been reported,[16] no structural information in solution is
available for this group of peptaibols.[33]
Our earlier studies on the screw-sense preference induced
by chiral amino acids in achiral oligomers made use of com-
pounds whose structure was inspired by the N-terminal se-
quence of the cephaibols, and we showed that a series of
orientation and aggregate in
a cylindrical superstruc-
ture.[23,24] This pore-forming mechanism has been unambigu-
ously demonstrated only for alamethicin and melittin, the
lytic toxin from honeybee venom.[25] However, in view of
both the closely related amino acid composition and physi-
co-chemical properties of other long-chain (i.e. ca. 20-resi-
due) peptaibols, it is conceivable that they too could form
pores in a similar way. Support for the “barrel-stave” mode
of action is also provided by the fact that the length of ala-
methicin means it can almost exactly span the thickness of
bacterial membranes. By contrast, another important sub-
family of membrane-active peptaibols has less than 15 resi-
dues:[26] representative members include antiamoebins,[27–29]
cephaibols[15] and zervamicins.[27,30] Their membrane-perturb-
ing mechanism of action is still under debate.
ꢂcephaibologuesꢃ Cbz-l-PheACTHNUTRGNE(UNG Aib4Gly)nAib4AibCH2OH (n=
1
0–3) displayed features in their H NMR spectra indicating
that the entire achiral segment of the oligomers adopts to
some extent a preferred screw sense.[6] We developed NMR-
based methods to quantify screw-sense preference in similar
achiral helices,[34] and used related Aib oligomers to relay in-
formation from a switchable centre or binding site to
a remote isotopically labelled enantiomerically enriched
NMR reporter.[4,6–8,34,35] This work revealed that, contrary to
The cephaibols were isolated from the soil fungus Acre-
monium tubakii, DSM 127741.[15] In addition to their consid-
erable antibacterial potency, they exhibit pronounced an-
thelmintic activity against ectoparasites.[15] The amino acid
sequence of cephaibols is highly conserved among the mem-
bers of the group (Figure 1) and the principal points of var-
iation are residues 5, 8 and 12, where either Aib or its one-
carbon homologue d-Iva[31] may be found. A 10-fold varia-
tion in the antibacterial potency is observed in this peptaibol
expectations, Cbz-l-PheACTHNUTGRNE(UNG Aib4Gly)nAib4AibCH2OH and re-
lated (Aib)n chains carrying N-terminal l-amino acids adopt
left-handed helicity in solution, unless the N-terminal l-
amino acid is quaternary, in which case right-handed helicity
is preferred.[35] We further showed that the preference for
left- or right-handed helicity arises from the detailed struc-
ture of the b-turn at the N terminus: tertiary l-amino acids
induce a left-handed screw sense by favouring a Type II b-
turn, while quaternary l amino acids induce a right-handed
screw sense by favouring a Type III b-turn.[17]
Despite the similarity of the N-terminal Ac-l-PheAib4Gly
sequence of cephaibols A and C to the sequences of the
left-handed
helical
oligomers
Cbz-l-Phe-
AHCTNUGERTN(GNNU Aib4Gly)nAib4AibCH2OH, the N-terminal helices of both
cephaibols A and C are right-handed in the solid state
(Figure 2).[16] Moreover, the usual strong propensity of Aib-
rich peptides to populate 310 helical conformations[1] is not
evident in the X-ray crystal structure of the cephaibolsꢃ N-
terminal domain, which is folded into an a-helix until the
ninth residue.
We therefore set out to synthesise the four configurational
diastereoisomers 1 of the N-terminal domain of cephaibols
A and C (Figure 3) and to explore the dependence of their
conformation in solution on the absolute and relative config-
urations of the three chiral amino acids that they contain.
We hoped to achieve deeper insights into the origin of
screw-sense preference in these bioactive peptides and to
clarify whether their solid state conformational preferences
are representative of their solution structures.
Figure 1. Chemical structure of cephaibols and comparison with anti-
ACHTUNGTRENNUNGamoebin.
16358
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16357 – 16365

N-terminal Nonapeptide of Cephaibols A and C
FULL PAPER
ed) residues, which precluded the possibility of using solid-
phase peptide-synthesis methods. We therefore used a frag-
ment-based solution-phase approach, shown in Scheme 1,
starting from l-Iva.
We reasoned that the presence of a Gly residue in the
middle of the target peptides 1a–d would facilitate the syn-
thesis by allowing the azlactone of Cbz-Phe-Aib4-OH[36] to
be ring opened by the reactive Gly amino group of tetrapep-
tide 5. The synthesis began with the sterically demanding
coupling between Aib tert-butyl ester and the acyl fluoride
derivative[37] of Cbz-l-Iva-OH, giving dipeptide 3[38] in 80%
yield. Standard peptide coupling and deprotection using
EDC/HOBt with Cbz-l- or Cbz-d-Leu-OH, then with Cbz-
Gly-OH, each followed by deprotection, provided the two
diastereoisomers 5a and 5b, each in 47% overall yield from
Cbz-l-Iva-OH. Conversion of each enantiomer of Cbz-Phe-
Aib4-OH[36] to its azlactone with acetic anhydride and ring
opening with each of the amines 5 afforded each of the dia-
stereoisomers of 2. Final Cbz deprotection and N-acetyla-
tion to yield 1a–d was achieved by hydrogenation in a solu-
tion of EtOAc/Ac2O. The four peptides were purified by
preparative HPLC.
Figure 2. Superimposition of the solid state structures of cephaibols A
(black; PDB entry: 1OB4) and
ref. [16]).
C (grey; PDB entry: 1OB7) (see
Conformational studies
CD spectroscopy: CD spectra[39] of the four diastereoisomers
of 2 were acquired in MeOH (a solvent that prevents aggre-
gation of related peptides) and are shown in Figure 4.
The CD spectra of the two pairs 2a,d and 2b,c are near
mirror images, despite all four compounds containing the
same l-Iva residue. This observation indicates that the con-
figuration of the Iva residue relative to the rest of the pep-
tide exerts little effect on the peptideꢃs overall conforma-
tion. Ivaꢃs Ca Me and Et groups are similar in size, and Iva
is typically seen to be screw-sense indifferent.[16,35,40] Thus
2a,d and 2b,c appear to be essentially pseudoenantiomeric
pairs.
Figure 3. The diastereoisomeric nonapeptide targets 1a–d (R=Ac) and
2a–d (R=Cbz).
Results and Discussion
Synthesis of the nonapeptides: The primary sequences of
the four peptides 1 include several quaternary (Ca-dialkylat-
Importantly, however, the shape of the spectra of the 2a,d
pair is very different from that of the 2b,c pair. 2b and 2c,
Scheme 1. Synthesis of the diastereoisomers of 1, the N-terminal domain of cephaibols A and C.
Chem. Eur. J. 2013, 19, 16357 – 16365
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16359

J. Clayden et al.
significant change is evident in the CD curves of the confor-
mation of 1b,c in MeOH (Figure 5a): again, they appear to
adopt well-structured 310-helices. Indeed, the increased in-
tensity of the CD signal is consistent with an even higher
population of folded structures.[42] The same 310-helical char-
acter is likewise evident in CH3CN (Figure 5b) and 2,2,2-tri-
fluoroethanol (TFE) (Figure 5c).
However, the CD curves of the N-acetylated peptide pair
1a,d reveal a few differences from those of their Cbz-pro-
tected counterparts 2a,d. As illustrated in Figure 5, the con-
formation of the less well-structured peptides 1a,d appears
to be more affected by solvent than that 1b,c. Remarkably,
an inversion in the Cotton effect seems to take place for
1a,d when moving from CH3CN to TFE.[43]
These CD results indicate that the relative configuration
of the Phe and Leu residues of peptides 1b and 1c work to-
gether in solution to maintain a well-defined, solvent-inde-
pendent, mostly 310-helical conformation. In other words,
they are stereochemically ꢂmatchedꢃ, despite having opposite
absolute configurations. This conclusion is fully consistent
with our previous observations that an N-terminal l-amino
acid induces a left-handed screw sense, even though l-
amino acids incorporated within a helical motif typically
favour right-handed screw sense. The left-handed screw
sense evident in the CD spectrum of 1b is induced by the
N-terminal l-Phe and reinforced by the left-handed screw
sense favoured by the d-Leu7 of the helix. Likewise the
right-handed screw-sense evident in the CD spectrum of 1c
is induced by the N-terminal d-Phe and reinforced by the
right-handed screw-sense favoured by the l-Leu7. The rela-
tive configuration of the l-Iva8 has no effect on the overall
screw sense.
Figure 4. CD spectra of the four configurational isomers of Cbz-protected
2 in MeOH. Peptide concentration: 1 mm.
in which the absolute configurations of Phe and Leu are op-
posite, display a profile typical of a highly populated 310-
helix,[41] with 2b and 2c being left- and right-handed, respec-
tively. 310-Helical conformations are characteristic of pep-
tides rich in quaternary amino acids such as Aib and Iva.
On the other hand, no clear conclusions can be drawn from
the CD on the secondary structure adopted by peptides
2a,d, where both Phe and Leu have the same configuration,
despite the same high proportion of helix-promoting Aib
and Iva residues.
When the Cbz group of 2 is replaced with the acetyl
group of 1, as found at the N-terminus of the cephaibols, no
In contrast, the corresponding CD spectra of 1a and 1d
show that these peptides may either fold into a mixture of
Figure 5. CD spectra of the four configurational isomers of 1 in MeOH (a), CH3CN (b) and TFE (c). Peptide concentration: 1 mm.
16360
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16357 – 16365

N-terminal Nonapeptide of Cephaibols A and C
FULL PAPER
interconverting left- and right-handed helices or may not
adopt any organized conformation at all. The configurations
of the Phe and Leu residues in 1a and 1d are mismatched,
and work against one another to frustrate the adoption of
a well-defined structure. Remarkably, it is one of these two
peptides, 1d, that has the relative stereochemistry (though
the opposite absolute stereochemistry) of natural cephaibols
A and C.
NMR spectroscopy: CD spectroscopy provides evidence of
helicity in peptides 1b and 1c, but fails to give detailed in-
formation on the preferred conformation of peptides 1a and
1d. Detailed structural characterization of the peptides was
therefore carried out using 2D NMR spectroscopy. We fo-
cused our analysis on peptides 1a (with the two principal
chiral residues conformationally ꢂmismatchedꢃ) and its
epimer 1b (whose two principal chiral residues are
ꢂmatchedꢃ). Experiments were performed in CD3OH solu-
tion in order to preserve amide NH proton signals, to avoid
complications arising from aggregation in chlorinated sol-
vents, and to allows us to correlate results with those ob-
tained by CD spectroscopy.
Overall assignment of proton resonances of both peptides
1a and 1b was made by means of homo- and heteronuclear
2D NMR spectra (see the Experimental Section). The
amide proton region of the ROESY spectrum of ꢂmatchedꢃ
peptide 1b is shown in Figure 6. All NH···NH sequential in-
teractions are visible (apart from those obscured by diagonal
peaks), consistent with the presence of a helical conforma-
tion.
Figure 7. CaH–NH fingerprint region of the ROESY spectrum (tm
200 ms, 298 K, 500 MHz) of 1b in CD3OH (5 mm). Medium range
CaHi!NHi+2 (typical for a 310-helix) and CaHi!NHi+3 (310- or a-helix)
cross-peaks are underlined.
(Figure 7 and 8). Many diagnostic
a(b)Hi–NHi+2 cross peaks are visible, confirming the pres-
C
a(b)Hi–NHi+3 and
C
ence of a mainly 310 helix throughout the sequence of pep-
tide 1b. In particular, the two
C
a(b)Hi–NHi+3 [2!5
(Figure 8) and 6!9 (Figure 7)] connectivities support the
hypothesis that peptide 1b is folded into a helical conforma-
tion. Moreover, the two CaHi–NHi+2 [1!3 and 7!9
(Figure 7)] cross-peaks are diagnostic of a mainly 310-helical
secondary structure throughout the peptide sequence.[44] The
a-helical contribution, if present, is not significant, since no
C
a(b)Hi–NHi+4 are detectable. The 2D NMR study supports
the conclusions drawn from the CD spectra that this nona-
peptide displays a strong preference to fold into a 310-helix
in methanol.
In the case of the conformationally ꢂmismatchedꢃ peptide
1a, 3D structural NMR analysis is less straightforward.
Further relevant 3D structural information is obtained
from the Ca(b)H–NH regions of the ROESY spectrum
Figure 8. CbH–NH correlation region of the ROESY spectrum (tm
200 ms, 298 K, 500 MHz) of peptide 1b in CD3OH (5 mm). Medium-
range interactions CbHi!NHi+3 diagnostic for helical structures, and in-
traresidue Iva cross-peaks, are underlined. (HB1*: Aib pro-R methyl
group; HB2*: Aib pro-S methyl group).
Figure 6. Amide proton region of the ROESY spectrum (tm 200 ms,
298 K, 500 MHz) of peptide 1b in CD3OH (5 mm).
Chem. Eur. J. 2013, 19, 16357 – 16365
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16361

J. Clayden et al.
Figure 10. CaH–NH region of the ROESY spectrum (tm 200 ms, 298 K,
500 MHz) of 1a in CD3OH (5 mm). The medium range cross-peak
CaHi!NHi+2 (310-helix) is underlined.
helical nature of both peptides 1a and 1b, though the mag-
nitude of the coefficient of Aib3 NH also suggests that they
exist as a mixture of 310/a-helical structures in MeOH solu-
tion.
Figure 9. Amide proton region of the ROESY spectrum (tm 200 ms,
298 K, 500 MHz) of peptide 1a in CD3OH (5 mm).
Figure 9 shows the amide proton region of the ROESY
spectrum of peptide 1a.
Another valuable indicator of helical character in Aib-
rich peptides is the anisochronicity of the 13C NMR signals
of a pair of diastereotopic methyl groups in each Aib resi-
due.[46] The 13C NMR signals of such gem-dimethyl pairs in
unfolded structures, even when directly adjacent to a stereo-
genic centre, are usually separated by less than 0.5 ppm.
However, within a stable helical conformation, this aniso-
chronicity value (also called “chemical nonequivalence” or
CNE[46]) can rise to more than 2 ppm. Table 1 reports the
anisochronicities in the 13C NMR spectrum of the methyl
groups of the five Aib residues of peptides 1a and 1b.
Despite their difference being just one chiral centre, only
peptide 1b (the ꢂmatchedꢃ structure) shows high (>2 ppm)
anisochronicity values in the main body of the Aib4 se-
As with peptide 1b, all the sequential, inter-residue NH-
NH interactions are also detectable in the case of peptide
1a. Furthermore, inspection of the fingerprint region of its
ROESY spectrum (Figure 10 and 11) reveals a few medium-
range cross-peaks accounting for the presence of a helical
structure (CbHi–NHi+3, 2!5 and 4!7, Figure 11) of the 310-
type (CaHi–NHi+2, 1!3, Figure 10). This seems to suggest
that peptide 1a is helical, despite its inconclusive CD spec-
trum.
In order to identify the presence of a- or 310 helical con-
formations in 1a and 1b, the variation in chemical shift with
temperature of each NH signal was measured. In the Sup-
porting Information, Figure S1
shows the different NMR spec-
tra for peptides 1a and 1b be-
tween the range 278–308 K.
Figure 12 reports the temper-
ature coefficients of the chemi-
cal shifts of the amide NH sig-
nals (Dd(HN)/DT) measured
for peptides 1a and 1b, indicat-
ing that both peptides behave
similarly with variation of tem-
perature. The highest values are
exhibited by amide NHs locat-
ed at the N-terminus, which are
exposed to the solvent and
therefore are more affected
than the rest of the intramolec-
ular H-bond network by
Figure 11. CbH–NH correlation region of the ROESY spectrum (tm 200 ms, 298 K, 500 MHz) of peptide 1a in
a change in temperature. This
VT-NMR analysis confirms the CD3OH (5 mm). Medium-range interactions CbHi!NHi+3 diagnostic for helical structures are underlined.
16362
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16357 – 16365

N-terminal Nonapeptide of Cephaibols A and C
FULL PAPER
shows smaller values for the anisochronicities of the methyl
groups in each of its Aib residues. These values indicate that
this peptide is not well structured as a helix with a single
screw sense, and may instead be undergoing rapid (on the
NMR time scale) conformational exchange, possibly be-
tween right- and left-handed helical conformations.
It has been reported that the chemical-shift difference be-
tween the diastereotopic b-CH2 protons of an Iva residue
embedded in a peptide chain may be correlated to the screw
sense adopted by the helix, provided that the configuration
of the Iva residue is known (Table 2).[47]
Indeed, when embedded into the diastereoisomeric pep-
tides 1a and 1b, the l-Iva8 residues behave differently, as
shown in Table 3:
Table 3. NMR parameters relative to the l-Iva8 residues in the two pep-
tides 1a and 1b.
NMR parameter
1a [ppm]
1b [ppm]
d gCH3 (1H NMR)
Dd bCH2
0.870
0.148
30.98
0.851
0.393
28.30
d bCH2 (13C NMR)
By comparing the two sets of values with those reported
in Table 2, it is fair to conclude that while the parameters
obtained from l-Iva8 of peptide 1b are consistent with
a left-handed (M) helix, those obtained from peptide 1a do
not show a clear propensity for only one helical screw sense
over the other, but rather supports the view of a system un-
dergoing fast conformational interconversion.
Figure 12. Temperature coefficients for amide H-atom chemical shifts
(temperature range: 278–308 K) of peptide 1a (top) and 1b (bottom).
The dashed line at ꢀ4.5 ppbKꢀ1 illustrates the proposed level (see
refs. [3] and [45]) that divides intramolecularly hydrogen-bonded from
free-amide NHs.
2D NMR analysis of peptide 1a indicates that in solution
it adopts principally a helical structure. Conversely, the CD
data and the data in Tables 1 and 3 are consistent with the
interpretation either that peptide 1a is unfolded, or that if it
is helical it does not adopt a single favoured screw sense.
We therefore deduce that while peptide 1b is a stable left-
handed helix in solution, peptide 1a exists as a rapidly inter-
converting mixture of left- and right-handed helices. The dif-
ference results from either a match or a mismatch between
the effects of the Phe and Leu residues. Both peptides are
rich in quaternary amino acids, and so favour the population
of helical conformations. In both 1a and 1b, the N-terminal
l-Phe induces a left-handed screw sense, and in 1b this con-
formational preference is reinforced by d-Leu7, giving a ther-
modynamically stable left-handed helical structure. In 1a,
however, the left-handed screw-sense favoured by the N-ter-
minal l-Phe is opposed by the right-handed screw sense
preference of l-Leu7. Both screw senses are therefore popu-
lated, and rapid conformational exchange results. The same
Table 1. Anisochronicity or CNE values (ppm) of 13Cb of methyl groups
of the Aib residues of peptides 1a and 1b.
residue
1a [ppm]
1b [ppm]
Aib2
Aib3
Aib4
Aib5
Aib9
1.61
1.08
0.75
<1
2.85
3.16
2.95
2.91
0.54
<1
quence, lending further support to the hypothesis that this
molecule adopts in solution a well-developed 310 helix. On
the other hand, peptide 1a (the ꢂmismatchedꢃ structure)
Table 2. Summary of the NMR parameters used to assess the absolute configuration of Iva residues in a right-handed peptide (see ref. [47]).[a]
NMR parameter
Value for d-Iva [ppm]
Value for l-Iva [ppm]
1H NMR chemical shift of the g-methyl protons
d<0.89
Dd>0.28
d<29
d>0.91
Dd<0.20
d>33
Difference between the chemical shifts (Dd) of the two b-methylene protons
13C NMR chemical shift of the b-methylene carbon atom
[a] For a left-handed helical peptide, the parameters must be reversed.
Chem. Eur. J. 2013, 19, 16357 – 16365
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16363

J. Clayden et al.
(256 experiments of 200 scans each, spectral width in F1 is 200 ppm, cen-
tred at 95 ppm).[50] HMBC experiments[51] were performed by using
a long-range coupling constant of 7.5 Hz, a spectral width in F1 of
220 ppm centred at 100 ppm, 128 t1 experiments of 500 scans, and 1 K
points in F2. ROESY experiments were used for sequence-specific assign-
ment. After analysis of the build-up curves, a mixing time of 240 ms was
employed to acquire the ROESY experiments. The splitting patterns (see
the Supporting Information) are abbreviated as follows: (s) singlet, (d)
doublet, (t) triplet, (m) multiplet.
conclusions can be drawn for peptides 1c and 1d, whose CD
spectra suggest almost identical behaviour to 1b and 1a, re-
spectively, given the weakness of the role played by inverted
relative stereochemistry at the Iva reside.
Conclusion
The conformational analysis carried out for the four diaste-
reoisomers of the N-terminal domain of cephaibols A and C
indicates that while all four adopt helical conformations,
only in those (1b, 1c) with the unnatural relative configura-
tion between the residues Phe1 and Leu7 is this conforma-
tion a stable, single screw sense (left handed for 1b; right
handed for 1c). In peptides 1a and 1d, the helical confor-
mation is fluxional in its screw sense, and the population of
different conformers varies in a solvent-dependent manner.
Intriguingly, the relative configuration displayed by the
natural products cephaibols A and cephaibols C corresponds
to (though is enantiomeric with) that of 1a. This suggests
that the N-terminal domain of the natural compound may
also be characterised by conformational fluxionality. The ap-
parently well-folded right-handed helix observed for this
domain in the X-ray crystal structure of these cephaibols
may therefore be a crystallographic artefact, or may possibly
be stabilised by interactions with the C-terminal domain
missing from 1a. Detailed studies of the conformation of
the intact cephaibols in solution have never been carried
out: 2D NMR work was used to confirm their primary se-
quence but not their secondary structure.[15,16]
Acknowledgements
We acknowledge funding from the ERC (Advanced Grant ROCOCO),
the Accademia dei Lincei of Rome (fellowship to M.D.Z). Jonathan
Clayden is the recipient of a Royal Society Wolfson Research Merit
Award.
[1] a) J. Venkatraman, S. C. Shankaramma, P. Balaram, Chem. Rev.
2001, 101, 3131; b) C. Toniolo, E. Benedetti, Trends Biochem. Sci.
1991, 16, 350.
[2] a) V. Moretto, M. Crisma, G. M. Bonora, C. Toniolo, H. Balaram, P.
Balaram, Macromolecules 1989, 22, 2939; b) I. L. Karle, P. Balaram,
Biochemistry 1990, 29, 6747; c) C. Toniolo, M. Crisma, G. M.
Bonora, E. Benedetti, B. Di Blasio, V. Pavone, C. Pedone, A. Santi-
ni, Biopolymers 1991, 31, 129; d) C. Toniolo, M. Crisma, F. Formag-
gio, C. Peggion, Biopolymers 2001, 60, 396; e) R. Banerjee, G. Basu,
ChemBioChem 2002, 3, 1263; f) J. R. Kumita, C. J. Weston, L.-P.
Choo-Smith, G. A. Woolley, O. S. Smart, Biochemistry 2003, 42,
4492; g) B. V. Prasad, P. Balaram, CRC Crit. Rev. Biochem. 1984, 16,
307; h) C. Toniolo, G. M. Bonora, V. Barone, A. Bavoso, E. Bene-
detti, B. Di Blasio, P. Grimaldi, F. Lelj, V. Pavone, C. Pedone, Mac-
romolecules 1985, 18, 895; i) C. Toniolo, G. M. Bonora, A. Bavoso,
E. Benedetti, B. Di Blasio, V. Pavone, C. Pedone, Macromolecules
1986, 19, 472.
Whether this conformational mobility plays a role in the
bioactivity of the cephaibols remains unclear, but it is an in-
triguing thought that the incorporation of the N-terminal l-
Phe residue by the fungus is the result of evolutionary selec-
tion for organisms producing antibiotics with a disordered
N-terminal helix. Similar proposals have been made for
other peptaibols whose conformational mobility contributes
to their bioactivity.[29,48]
[3] J. D. Augspurger, V. A. Bindra, H. A. Scheraga, A. Kuki, Biochem-
istry 1995, 34, 2566.
[4] J. Solꢀ, S. P. Fletcher, A. Castellanos, J. Clayden, Angew. Chem.
2010, 122, 6988; Angew. Chem. Int. Ed. 2010, 49, 6836.
[5] a) R. A. Brown, V. Diemer, S. J. Webb, J. Clayden, submitted; b) B.
Pengo, F. Formaggio, M. Crisma, C. Toniolo, G. M. Bonora, Q. B.
Broxterman, J. Kamphius, M. Saviano, R. Iacovino, F. Rossi, E. Ben-
edetti, J. Chem. Soc. Perkin Trans. 2 1998, 1651.
[6] J. Clayden, A. Castellanos, J. Solꢀ, G. A. Morris, Angew. Chem.
2009, 121, 6076; Angew. Chem. Int. Ed. 2009, 48, 5962.
[7] J. Solꢀ, M. Helliwell, J. Clayden, J. Am. Chem. Soc. 2010, 132, 4548.
[8] T. Boddaert, J. Solꢀ, M. Helliwell, J. Clayden, Chem. Commun.
2012, 48, 3397.
[9] N. Ousaka, Y. Takeyama, E. Yashima, Chem. Eur. J. 2013, 19, 4680.
[10] N. Ousaka, Y. Takeyama, H. Iida, E. Yashima, Nat. Chem. 2011, 3,
856.
[11] a) Y. Inai, Y. Kurokawa, A. Ida, T. Hirabayashi, Bull. Chem. Soc.
Jap. 1999, 72, 55; b) Y. Inai, Y. Kurokawa, T. Hirabayashi, Biopoly-
mers 1999, 49, 551.
[12] a) J. Solꢀ, M. Helliwell, J. Clayden, Biopolymers 2011, 95, 62; b) Y.
Inai, Y. Kurokawa, N. Kojima, J. Chem. Soc. Perkin Trans. 2 2002,
1850.
[13] Y. Inai, K. Tagawa, A. Takasu, T. Hirabayashi, T. Oshikawa, M. Ya-
mashita, J. Am. Chem. Soc. 2000, 122, 11731.
[14] a) Y. Inai, Y. Ishida, K. Tagawa, A. Takasu, T. Hirabayashi, J. Am.
Chem. Soc. 2002, 124, 2466; b) Y. Inai, N. Ousaka, T. Okabe, J. Am.
Chem. Soc. 2003, 125, 8151; c) N. Ousaka, Y. Inai, J. Org. Chem.
2009, 74, 1429.
[15] M. Schiell, J. Hofmann, M. Kurz, F. R. Schmidt, L. Vertesy, M.
Vogel, J. Wink, G. Seibert, J. Antibiot. 2001, 54, 220.
Experimental Section
Synthesis and characterization of peptides: The synthetic strategy used in
this work is shown in Scheme 1. Characterization details are reported in
the Supporting Information.
Circular dichroism: CD spectra were obtained on a Jasco (Tokyo, Japan)
J-815 spectropolarimeter. A quartz cell (Hellma) of 0.1 mm path length
was used. The values are expressed in terms of [q]M, molar ellipticity
(degcm2 dmolꢀ1). Spectrograde MeOH, CH3CN and TFE (Sigma Al-
drich) were used as solvents.
NMR spectroscopy: NMR experiments were carried out on a Bruker
500 MHz spectrometer. The peptide concentration was 5 mm in CD3OH.
The alcohol OH signal was suppressed by presaturation during the relax-
ation delay or applying a WATERGATE gradient program. All homonu-
clear spectra were acquired by collecting 256 experiments, each one con-
sisting of 32 scans and 2 K data points. The spin systems of protein amino
acid residues were identified by using standard DQF-COSY and TOCSY
spectra (spin-lock pulse: 70 ms).[49] The stereospecific assignment of the
Aib residues methyl groups was achieved through the HMQC spectra
[16] G. Bunkꢄczi, M. Schiell, L. Vꢅrtesy, G. M. Sheldrick, J. Pept. Sci.
2003, 9, 745.
16364
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16357 – 16365

N-terminal Nonapeptide of Cephaibols A and C
FULL PAPER
[17] M. De Poli, M. De Zotti, J. Raftery, J. A. Aguilar, G. A. Morris, J.
Clayden, J. Org. Chem. 2013, 78, 2248.
[35] R. A. Brown, T. Marcelli, M. De Poli, J. Solꢀ, J. Clayden, Angew.
Chem. 2012, 124, 1424; Angew. Chem. Int. Ed. 2012, 51, 1395.
[36] S. J. Pike, M. De Poli, W. Zawodny, J. Raftery, S. J. Webb, J. Clayden,
Org. Biomol. Chem. 2013, 11, 3168.
[37] L. A. Carpino, D.-A. Aalaee, H. G. Chao, R. H. DeSelms, J. Am.
Chem. Soc. 1990, 112, 9651.
[38] M. De Zotti, F. Damato, F. Formaggio, M. Crisma, E. Schievano, S.
Mammi, B. Kaptein, Q. B. Broxterman, P. J. Felock, D. J. Hazuda,
S. B. Singh, J. Kirschbaum, H. Brꢆckner, C. Toniolo, Chem. Eur. J.
2010, 16, 316.
[39] N. Berova, L. Di Bari, G. Pecsitelli, Chem. Soc. Rev. 2007, 36, 914.
[40] a) R. P. Hummel, C. Toniolo, G. Jung, Angew. Chem. 1987, 99, 1180;
Angew. Chem. Int. Ed. Engl. 1987, 26, 1150; b) M. Kubasik, A.
Blom, ChemBioChem 2005, 6, 1187; c) M. A. Kubasik, J. Kotz, C.
Szabo, T. Furlong, J. Stace, Biopolymers 2005, 78, 87.
[41] C. Toniolo, A. Polese, F. Formaggio, M. Crisma, J. Kamphuis, J. Am.
Chem. Soc. 1996, 118, 2744.
[42] We have previously noted that an N-terminal amide provides a great-
er level of conformational uniformity than an N-terminal carbamate
in related structures.[35]
[43] M. Goodman, A. S. Verdini, C. Toniolo, W. D. Phillips, F. A. Bovey,
Proc. Natl. Acad. Sci. USA 1969, 64, 444.
[44] K. Wꢆthrich, in NMR of Proteins and Nucleic Acids, Wiley, New
York, NY, 1986.
[18] a) J. F. Daniel, E. R. Filho, Nat. Prod. Rep. 2007, 24, 1128; b) S.
Ayers, B. M. Ehrmann, A. F. Adcock, D. J. Kroll, E. J. Carcache de
Blanco, Q. Shen, S. M. Swanson, J. O. Falkinham III, M. C. Wani,
S. M. Mitchell, C. J. Pearce, N. H. Oberlies, J. Pept. Sci. 2012, 18,
500.
[19] C. Toniolo, H. Brꢆckner, Peptaibiotics, Wiley-VCH, 2009.
[20] E. Benedetti, A. Bavoso, B. Di Blasio, V. Pavone, C. Pedone, C. To-
niolo, G. M. Bonora, Proc. Natl. Acad. Sci. USA 1982, 79, 7951.
[21] W. C. Wimley, ACS Chem. Biol. 2010, 5, 905.
[22] B. Leitgeb, A. Szekeres, L. Manczinger, C. Vꢇgvçlgyi, L. Kredics,
Chem. Biodiversity 2007, 4, 1027.
[23] R. O. Fox, F. M. Richards, Nature 1982, 300, 325.
[24] a) H. W. Huang, F. Y. Chen, M. T. Lee, Phys. Rev. Lett. 2004, 92,
198304; b) S. Qian, W. Wang, L. Yang, H. W. Huang, Biophys. J.
2008, 94, 3512; c) H. Duclohier, H. Wrꢄblewski, J. Membr. Biol.
2001, 184, 1; d) S. Ye, H. Li, F. Wei, J. Jasensky, A. P. Boughton, P.
Yang, Z. Chen, J. Am. Chem. Soc. 2012, 134, 6237; e) G. Menestrina,
K. P. Voges, G. Jung, G. Boheim, J. Membr. Biol. 1986, 93, 111.
[25] a) K. Matsuzaki, S. Yoneyama, K. Miyajima, Biophys. J. 1997, 73,
831; b) D. Sengupta, H. Leontiadou, A. E. Mark, S. J. Marrink, Bio-
chim. Biophys. Acta Biomembr. 2008, 1778, 2308.
[26] J. K. Chugh, B. A. Wallace, Biochem. Soc. Trans. 2001, 29, 565.
[27] K. L. Rinehart Jr., L. A. Gaudioso, M. L. Moore, R. C. Padney, J. C.
Cook Jr., J. Am. Chem. Soc. 1981, 103, 6517.
[28] A. Jaworski, H. Brꢆckner, J. Pept. Sci. 2000, 6, 149.
[29] Z. O. Shenkarev, A. S. Paramonov, K. D. Nadezhdina, E. D. Bochar-
ov, I. A. Kudelina, D. A. Skladnev, A. A. Tagaev, Z. A. Yakimenko,
T. V. Ovchinnikova, A. S. Arseniev, Chem. Biodiversity 2007, 4,
1219.
[45] a) N. J. Baxter, M. P. Williamson, J. Biomol. NMR 1997, 9, 359;
b) G. M. Bonora, C. Mapelli, C. Toniolo, R. R. Wilkening, E. S. Ste-
vens, Int. J. Biol. Macromol. 1984, 6, 179.
[46] G. Jung, H. Brꢆckner, R. Bosch, W. Winter, H. Schaal, J. Strꢈhle,
Liebigs Ann. Chem. 1983, 1096.
[47] M. De Zotti, B. Biondi, M. Crisma, C. U. Hjørringgaard, A. Berg, H.
Brꢆckner, C. Toniolo, Biopolymers 2012, 98, 36.
[30] I. L. Karle, J. L. Flippen-Anderson, S. Agarwalla, P. Balaram, Proc.
Natl. Acad. Sci. USA 1991, 88, 5307.
[31] Peptaibols may contain either l-Iva or d-Iva: X-ray crystal struc-
tures of the cephaibols (ref. [16]) indicate that these residues have
d-configuration in the cephaibols.
[32] a) I. L. Karle, M. A. Perozzo, V. K. Mishra, P. Balaram, Proc. Natl.
Acad. Sci. USA 1998, 95, 5501; b) R. Anders, O. Ohlenschlager, V.
Soskic, H. Wenschuh, B. Heise, L. R. Brown, Eur. J. Biochem. 2000,
267, 1784; c) T. P. Galbraith, R. Harris, P. C. Driscoll, B. A. Wallace,
Biophys. J. 2003, 84, 185; d) T. A. Balashova, Z. O. Shenkarev, A. A.
Tagaev, T. V. Ovchinnikova, J. Raap, A. S. Arseniev, FEBS Lett.
2000, 466, 333; e) T. Degenkolb, H. Brꢆckner, Chem. Biodiversity
2008, 5, 1817; f) S. Aravinda, N. Shamala, P. Balaram, Chem. Biodi-
versity 2007, 4, 1238; g) C. F. Snook, G. A. Woolley, G. Oliva, V. Pat-
tabhi, S. P. Wood, T. L. Blundell, B. A. Wallace, Structure 1998, 6,
783.
[48] a) P. A. Grigoriev, B. Schlegel, M. Kronen, A. Berg, A. Hꢈrtl, U.
Grꢈfe, J. Pept. Sci. 2003, 9, 763; b) M. De Zotti, B. Biondi, C. Peg-
gion, Y. Park, K.-S. Hahm, F. Formaggio, C. Toniolo, J. Pept. Sci.
2011, 17, 585; c) S. Bobone, D. Roversi, L. Giordano, M. De Zotti,
F. Formaggio, C. Toniolo, Y. Park, L. Stella, Biochemistry 2012, 51,
10124.
[49] a) A. Bax, D. G. Davis, J. Magn. Reson. 1985, 65, 355; b) M. Piotto,
V. Saudek, V. Sklenar, J. Biomol. NMR 1992, 2, 661.
[50] a) A. Bax, S. Subramanian, J. Magn. Reson. 1986, 67, 565; R.
Anders, H. Wenschuh, V. Soskic, S. Fischer-Frꢆhholz, O. Ohlenschl-
ꢈger, K. Dornberger, L. R. Brown, J. Pept. Res. 2009, 52, 34; b) M.
Bellanda, E. Peggion, R. Bꢆrgi, W. van Gunsteren, S. Mammi, J.
Pept. Res. 2001, 57, 97.
[51] A. Bax, M. F. Summers, J. Am. Chem. Soc. 1986, 108, 2093.
[33] Antiamoebin has already been studied in solution (ref. [19]).
[34] J. Solꢀ, G. A. Morris, J. Clayden, J. Am. Chem. Soc. 2011, 133, 3712.
Received: July 8, 2013
Published online: October 9, 2013
Chem. Eur. J. 2013, 19, 16357 – 16365
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16365