Chemoselective cyclization of unprotected linear peptides by α-ketoacid-hydroxylamine amide-ligation

Article abstract of DOI:10.1039/c2ob25129a

Cyclic peptides are important synthetic targets due to their constrained conformation, enhanced metabolic stability and improved bioavailability, which combine to make them promising lead compounds for drug candidates. They are typically synthesized by a multi-step sequence of carefully orchestrated protecting group manipulations and cyclization of side-chain protected linear precursors. In the present manuscript we disclose an alternative approach to the synthesis of peptide macrocycles by the α-ketoacid-hydroxylamine (KAHA) ligation. This reaction allows readily prepared linear peptides to be cyclized without reagents or side-chain protecting groups and delivers a native backbone amide bond at the ligation site. The precursors are prepared with Fmoc-based solid phase peptide synthesis using reagents that we have previously disclosed. No post-cyclization manipulations or deprotections other than purification are required. This protocol was applied to five different cyclic peptide natural products of varying ring sizes and side chain functionalities.

Full text of DOI:10.1039/c2ob25129a

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
This article is part of the
OBC 10th anniversary
themed issue
All articles in this issue will be gathered together
online at
www.rsc.org/OBC10

Dynamic Article Links
Organic &
View Article Online
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 5837
www.rsc.org/obc
PAPER
Chemoselective cyclization of unprotected linear peptides by
α-ketoacid–hydroxylamine amide-ligation†‡
Takeo Fukuzumi,a Lei Jua and Jeffrey W. Bode*b
Received 17th January 2012, Accepted 21st February 2012
DOI: 10.1039/c2ob25129a
Cyclic peptides are important synthetic targets due to their constrained conformation, enhanced metabolic
stability and improved bioavailability, which combine to make them promising lead compounds for drug
candidates. They are typically synthesized by a multi-step sequence of carefully orchestrated protecting
group manipulations and cyclization of side-chain protected linear precursors. In the present manuscript
we disclose an alternative approach to the synthesis of peptide macrocycles by the α-ketoacid–
hydroxylamine (KAHA) ligation. This reaction allows readily prepared linear peptides to be cyclized
without reagents or side-chain protecting groups and delivers a native backbone amide bond at the
ligation site. The precursors are prepared with Fmoc-based solid phase peptide synthesis using reagents
that we have previously disclosed. No post-cyclization manipulations or deprotections other than
purification are required. This protocol was applied to five different cyclic peptide natural products of
varying ring sizes and side chain functionalities.
synthesis including a late-stage global deprotection. The need to
retain protecting groups on the side chains also induces steric
demands than can hamper cyclizations. Furthermore, the use of
coupling reagents limits, to some extent, the utility of high
dilution methods that favour cyclic peptide formation. Native
chemical ligation allows chemoselective intramolecular amide
bond formation between C-terminal thioester and N-terminal
cysteine to directly deliver unprotected cyclic peptides.6 The
primary constraints of NCL are the requirement for a cysteine
residue at the cyclization site and the relative difficulties of
obtaining the C-terminal thioesters. To address these limitations
while maintaining the attractive properties of NCL, chemists
have continually sought to apply novel ligation reactions to the
preparation of cyclic peptides. Recent progress includes the use
of traceless Staudinger ligation, which works most efficiently
with glycine as the ligation residue,7 and cyclizations of thioa-
cids.8 The synthesis of cyclic peptides with non-amide connec-
tions by azide–alkyne cyclization9 or Ugi-type condensations10
are also attractive routes to peptide-based macrocycles. Bio-
engineering approaches to the production of libraries of cyclic
peptides are also of great contemporary interest.11
Introduction
Cyclic peptides are widely recognized as powerful platforms for
the development of biologically active molecules.1 In addition to
their widespread occurrence as natural products, cyclic peptides
are now common motifs for the design of unnatural compounds
as therapeutic agents. Their success is largely attributed to
improved biological properties relative to their linear forms.2
Confining a peptide into a cyclic structure results in increased
resistance to proteolytic cleavage, enhanced stability and bio-
availability, and reduced conformational freedom. For these
reasons, cyclic peptides remain important synthetic targets for
both chemists and biochemists.3
Current synthetic strategies towards cyclic peptides rely on the
use of orthogonally protected precursors, which are selectively
unmasked at specific functional groups.4 Cyclization of these
semi-protected linear peptides with excess amounts of coupling
reagents can be performed either in solution or on solid
support.5 This is a successful method but requires several
additional manipulations beyond standard solid-phase peptide
Our group has developed a chemoselective amide-forming
ligation reaction between C-terminal peptide α-ketoacids and
N-terminal hydroxylamine (KAHA ligation).12 This reaction has
been demonstrated to proceed in the presence of unprotected
side chain functionalities,13 which would allow the direct cycli-
zation of unprotected linear oligopeptides. A final deprotection
step is not necessary and cyclization can be performed on side-
chain unprotected substrates. Importantly, the amide-formation
proceeds over a wide range of concentrations (0.2 M–0.0001 M)
aRoy and Diana Vagelos Laboratories, Department of Chemistry,
University of Pennsylvania, Philadelphia, Pennsylvania, USA, 19104
bLaboratorium für Organische Chemie, ETH-Zürich, HCl F-315
Wolfgang Pauli Strasse 10, 8093 Zürich, Switzerland.
E-mail: bode@org.chem.ethz.ch
†This article is part of the Organic & Biomolecular Chemistry 10th
Anniversary issue.
‡Electronic supplementary information (ESI) available: experimental
procedures, HPLC traces, compound characterization, and determination
of resin loading. See DOI: 10.1039/c2ob25129a
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5837–5844 | 5837

View Article Online
and does not require any added reagents or catalysts. In the
context of peptide cyclization, reagent-less conditions would
allow the use of high dilution without diminishing the cycliz-
ation rate, thereby favouring cyclization over oligomerization
and potentially diminishing racemization. We envisioned that
applying this strategy toward the cyclization of side-chain un-
protected seco-peptides that could be prepared with standard
Fmoc-based solid-phase peptide synthesis in combination with
reagents and resins that we have previously reported. This article
describes the successful implementation of this strategy and its
application to several examples of cyclic peptide natural products
(Scheme 1).14
via our sulphur ylide linker (Scheme 2), cleavage from the resin
would afford a linear peptide sequence with an α-ketoacid–
hydroxylamine precursor residing on each terminus. Provided
that the nitrone-protected N-hydroxylamine survived the sub-
sequent sulphur ylide oxidation, deprotection and cyclization
would result in the desired cyclic peptide target.
We began our investigations with the synthesis of the prototy-
pical cyclic peptide natural product Gramacidin S.18 The prep-
aration of linear peptide precursor 2 was accomplished using a
MBHA resin-bound sulfonium salt using standard 9-fluorenyl-
methoxycarbonyl (Fmoc)-based solid phase peptide synthesis in
analogy to our previously described protocols (Scheme 3).15 The
protected hydroxylamine 3 was prepared from the corresponding
amino acid and introduced by standard peptide coupling proto-
cols as previously described. Cleavage of the resin bound
peptide (0.22–0.32 mmol g−1) was performed with neat TFA at
room temperature to give nitrone–sulphur ylide 4. To avoid pre-
mature hydrolysis during the cleavage step, we found it essential
to exclude moisture from the reaction by using only CH2Cl2
and TFA (no added H2O). When 3% of water was added in the
cleavage cocktail, rapid hydrolysis of the nitrone to the hydroxy-
lamine occurred along with subsequent formation of an oxime
side product. If desired, the cleaved peptide could be purified by
reversed phase preparative HPLC prior to cyclization. Although
subsequent cyclizations using purified precursors were somewhat
cleaner, as assayed by HPLC, partial hydrolysis of the nitrone-
protecting group during the purification was often observed and
we therefore elected to perform the oxidation and cyclization of
the unpurified peptides obtained directly from resin cleavage.
We have previously reported that oxidation of sulphur ylides
to α-ketoacids occurs easily and cleanly with Oxone under
aqueous solution.16 When using N-benzylidene protected hydro-
xylamines, however, the acidic Oxone solution induced hydro-
lysis of the nitrone and led to the formation of
hydroxylamine–α-ketoacids and the oxidized oxime–ketoacids
as side products. Dimethoxyldioxirane (DMDO) was therefore
Results and discussion
The greatest obstacle to implementing the KAHA ligation for the
peptide cyclization is the installation of the requisite reactive
partners into the linear peptide sequence prior to the cyclization
step. The preparation of peptide-derived C-terminal α-ketoacids
has been previously demonstrated in our laboratory with a solid-
supported sulphur ylide linker that can be used with standard
Fmoc-based solid phase peptide synthesis.15 This process affords
side-chain unprotected peptide sulphur ylides, which serve as
the immediate precursors to unprotected peptide α-ketoacids.16
Our group has also described general methods to synthesize
N-terminal, N-hydroxyamino acids protected as N-benzylidene
nitrones.17 Deprotection of the nitrones can be performed to
afford either side-chain protected or unprotected N-terminal
peptide hydroxylamines and is suitable for the preparation of
several enantiomerically pure N-hydroxylamino acid residues.
The nitrone-protection also provides the flexibility of unmasking
the N-hydroxyamino acid at different stages and enables us to
carry the protected hydroxylamine through resin cleavage and
side-chain deprotection. We envisioned that if a masked hydro-
xylamine was introduced at the end of a peptide bound to a resin
Scheme 1 Cyclization of unprotected linear peptides by α-ketoacid–hydroxylamine amide ligation.
Scheme 2 General strategy for preparation of unprotected linear peptides with N-terminal hydroxylamines and C-terminal α-ketoacids for direct,
reagent-less cyclization.
5838 | Org. Biomol. Chem., 2012, 10, 5837–5844
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 3 Representative synthesis of N-terminal nitrone protected peptide α-ketoacids; the synthesis of seco-gramicidin S is shown. Other cyclic
peptide precursors were prepared in an analogous fashion (see ESI‡ for details).
Scheme 4 Chemoselective oxidation of sulphur ylide 4 without nitrone deprotection using aqueous DMDO.
chosen as an alternative oxidant that does not require acidic con-
ditions.19 DMDO was prepared as a solution in acetone and
effected rapid oxidation of the sulphur ylide to the α-ketoacid.
Dimethylsulphide was added at the end of the oxidation to
quench any excess oxidant. The nitrone was found to be stable
through the oxidation process. The Gramicidin S precursor is a
representative example (Scheme 4): treatment of the crude
nitrone–sulphur ylide 4 with DMDO in acetone and water
mixture for 10 min yielded nitrone–ketoacid 5 with only small
amounts of the nitrone–carboxylic acid and oxime–α-ketoacid as
side products.
The crude peptide incorporating C-terminal α-ketoacid and
N-terminal nitrone (5) was directly subjected to cyclization in
degassed DMF and water mixture (Scheme 5). To suppress the
formation of oxime, which arises from oxidation of the hydroxy-
lamine,17 and increase the ligation yield, we examined different
solvents and acid additives. The presence of water is crucial;
with only anhydrous DMF as solvent, very little or no ligation
product was observed by LCMS analysis. This was attributed to
the need for a small amount of water to hydrolyse the nitrone
and reveal the hydroxylamine. The nitrone hydrolysis was aided
by the addition of oxalic acid and we found that only a trace
amount of the cyclized product was formed when no oxalic acid
was used as an additive. The use of hydrochloric acid as an addi-
tive also produced the cyclized products, albeit in slightly dimin-
ished yields (Table 1).
The formation of dimerized products or oligomers by intermo-
lecular cyclization is closely related to substrate concentration.
As a model compound to examine the effect of concentration
on the peptide cyclization, we selected the formation of semi-
gramicidin S (9), which is well-known to prefer dimerization to
cyclization under typical reaction conditions.20 The semi-grami-
cidin S precursor 11 was subjected to three separate reactions
under different concentrations (Scheme 6). Under dilute con-
ditions (0.001 M), only trace amount of the dimer, gramicidin
S was generated (47 : 1). As the concentration was increased to
0.01 M and 0.1 M, the amount of dimer formed also increased
(17 : 1 and 2 : 1).
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5837–5844 | 5839

View Article Online
Scheme 5 Cyclization of unprotected seco-peptides by ketoacid–hydroxylamines.
Table 1 Cyclization of linear peptides by KAHA ligation
HPLC yield
(%)b
Isolated yield
(%)c
Average yield per step
(%)e
Entry Cyclic peptide
Cyclization conditiona
1
2
3
4
5
6
7
8
Gramicidin S (6) cyclo
0.1 M (COOH)2, 0.001 M
DMF : H2O (50 : 1)
0.1 M (COOH)2, 0.001 M
DMF : H2O (5 : 1)
0.001 M DMF : 1 M HCl : DMS
(5 : 1 : 0.1)
0.1 M (COOH)2, 0.001 M
DMF : H2O (50 : 1)
13
10
11
36
9
13
ndd
ndd
22
4
66
63
64
81
62
68
70
(-DFPVOLDFPVOL-)
Tyrocidine A (7) cyclo-
(-DFPFDFNQYVOL-)
0.001 M DMF : 1 M HCl : DMS
(5 : 1 : 0.1)
Hymenamide B (8) cyclo(-FPPNFVE-) 0.1 M (COOH)2, 0.001 M
15
17
13
8
DMF : H2O (50 : 1)
0.1 M (COOH)2, 0.001 M
DMF : H2O (50 : 1)
semi Gramicidin S (9) cyclo
15
7
(-DFPVOL-)
Stylostatin Af (10) cyclo(-AISNDFPL) 0.1 M (COOH)2, 0.001 M
DMF : H2O (50 : 1)
a Reactions were carried out at 40 °C for 48 h. b The crude yield is determined from standard curve of purified cyclic peptide. c Isolated yield after
purification on reversed phase preparative HPLC. d Not determined. e Average yield per step for hydroxylamine coupling, resin cleavage and
deprotection, oxidation, nitrone hydrolysis, and cyclization. f A p-nitrobenzylidene nitrone was used instead of the benzylidene nitrone.
Epimerization is known to be one of the major concerns in
peptide chemistry, particularly during the formation of cyclic
peptides. In previous studies we have shown that some empimer-
ization of the α-ketoacid can occur during oxidation of the
sulphur ylide to the α-ketoacid. In our efforts to investigate the
epimerization issue during intramolecular cyclization, cyclic
peptide containing both epimeric amino acids at both the
α-ketoacid as well as the nitrone/hydroxylamine sides of the
cyclization point were prepared separately. Co-injection
on HPLC demonstrated that a small amount (approx. 5%) of epi-
merization occurred at the α-ketoacid residue (see ESI‡). The
nitrone–hydroxylamine residue was configurationally stable
throughout the sequence. In all cases studied, the epimerized
macrocyclic product had a significantly different retention time
and the desired product was easily purified by preparative
HPLC.
moeities, and acid functionalities. The peptide sulphur ylides
were prepared on a resin support with standard Fmoc-based
SPPS. The nitrone-protected hydroxylamines were introduced
during the final coupling step and cleavage of the Rink amide
resin provided the side-chain unprotected nitrone–sulphur ylides.
After oxidation and cyclization, the cyclic peptides were isolated
as pure products in respectable yields by preparative
HPLC. In all cases, the oxime formation was the most significant
side product. In the case of longer peptides (>5 residues), dimer
formation was not observed under the standard cyclization
conditions (0.001 M).
We have previously noted that alanine and glycine derived
nitrones are more resistant to hydrolysis.17 To address this chal-
lenge, we chose Leu–Ala as the disconnection site during syn-
thesis of stylostatin A. Under the standard cyclization conditions,
the benzylidene nitrone of alanine was resistant to hydrolysis; no
cyclized products were observed. Experiments performed using
TFA instead of oxalic acid in different solvents, as well as with
hydrochloric acid and p-toluenesolfonic acid provided only
the starting nitrone and decomposition. Previously studies have
suggested that nitrone hydrolysis is often kinetically feasible but
thermodynamically unfavourable;21 therefore, sequestering the
The established protocol was applied to formation of other
cyclic peptides of several ring sizes and with various unprotected
side chains (Fig. 1 and Table 1). The target peptides were chosen
to demonstrate the compatibility of a variety of functional group
to the method. The cyclic peptides selected include unprotected
side chains with primary and secondary amines, hydroxyl
5840 | Org. Biomol. Chem., 2012, 10, 5837–5844
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 6 Effect of substrate concentration on cyclization vs. dimerization–cyclization of a seco-gramicidin S peptide. HPLC analysis was per-
formed directly from the reaction mixtures without workup or prior purification.
benzaldehyde byproduct can facilitate the hydrolysis. We
attempted to achieve deprotection by passing the protected
α-ketoacid–nitrone peptide through a column of solid supported
hydroxylamine and C-18 silica gel, a protocol we have pre-
viously reported,17 but without success. When water was added
during cleavage of the peptide from the resin, the alanine nitrone
could be hydrolysed to yield the hydroxylamine–sulphur ylide.
However, the unprotected hydroxylamine does not tolerate the
subsequent oxidation and an alternative nitrone protecting group
was sought.
A variety of derivatized benzylidene nitrones (Fig. 2) were
prepared by condensing O-tert-butyl protected Ala-hydroxyl-
amine with the corresponding aldehydes. After tert-butyl
removal, the acids were loaded onto the end of the hexapeptide
sulphur ylide bound to the resin. After cleavage, trace amounts
of hydroxylamines were observed and the nitrone–sulphur ylides
were obtained as the major products. The cleaved nitrone–
sulphur ylide peptides were subjected to the standard DMDO
oxidation to give the stable nitrone–α-ketoacids. We found
that 4-methoxy, 2,4-dimethoxy, and 3-dimethylamino-substituted
benzylidene nitrones were difficult to deprotect. Stirring in
degassed hydrolysis conditions at 40–60 °C provided only the
nitrone–α-ketoacids, nitrone-carboxylic acid and their oxime
degradation products. Treating the cleaved sulphur ylide directly
with acid and water revealed that these nitrones were robust
enough to survive these conditions.
We found that 4-nitrobenzylidene nitrone 17, which is an elec-
tron-deficient nitrone, was more prone to hydrolysis and release
of the hydroxylamine. After peptide assembly and installation of
the nitrone protected hydroxylamine as the last residue on solid
phase, the peptide could be cleaved from the resin and purified
on reversed phase preparative HPLC (Scheme 7). Although
cyclization of the purified precursor was cleaner when monitored
on HPLC, nitrone hydrolysis on the column was observed
during the purification process; the crude peptide was therefore
used directly. Using N-terminal para-nitrobenzylidene nitrone
with a C-terminal α-ketoacid, the cyclized product stylostatin
A (10) was isolated in 6% yield along with recovered nitrone–α-
ketoacid and 13% of the oxime–ketoacid side product. These
results suggest that modulation of the nitrone protecting group
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5837–5844 | 5841

View Article Online
Fig. 1 Cyclic peptides prepared by KAHA-ligation of unprotected linear peptides and isolated yields (preparative HPLC and based on resin loading)
following Fmoc-solid phase peptide synthesis, resin cleavage and deprotection, sulphur ylide oxidation, and nitrone cleavage–cyclization. The ligation
sites are indicated in color.
may facilitate nitrone hydrolysis, but our recent mechanistic
studies into this amide-formation suggest that this may be detri-
mental to the amide-formation using these reaction partners.22
The method for the synthesis of cyclic peptides described
above requires the cyanosulphur ylide linker and the appropriate
nitrone-protected N-terminal amino acids. The latter must be pre-
pared by a several step chemical synthesis from the correspond-
ing protected α-amino acids. To avoid the need for the prior
preparation of the protected N-hydroxyamino acid, we have
Fig. 2 Various benzylidene nitrones investigated to facilitate hydrolysis
reported the design and development of reagent 20 for the direct
of alanine hydroxylamines.
oxidation of solid-supported N-terminal peptide amines to
nitrones, which can be hydrolysed to give the N-terminal peptide
hydroxylamines.23 The use of this reagent should allow for the
preparation of the linear peptide precursors without the need for
can be used to improve the cyclization conditions and adjust
the rate at which the nitrone moiety undergoes deprotection.
Increasing the amount of water present in the reaction mixture
Scheme 7 Successful nitrone cleavage and cyclization of para-nitrobenzylidene protected N-terminal alanine hydroxylamine.
5842 | Org. Biomol. Chem., 2012, 10, 5837–5844 This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 8 N-terminal hydroxylamine synthesis using oxaziridine reagent 20. In this approach only the sulphur ylide linker and reagent 20 are
needed to prepare cyclic peptides under standard Fmoc-SPPS conditions.
Synthesis of tyrocidine Awith reagent 20
independent preparation of the hydroxylamine monomers. To
test the suitability of this reagent for cyclic peptide synthesis, we
performed the synthesis of tyrocidine A using 20 to introduce
the N-terminal hydroxylamine from the linear peptide 19 with an
N-terminal D-phenyl alanine residue (Scheme 8). Following resin
cleavage, sulphur ylide oxidation, and cyclization we obtained
tyrocidine A in 17% yield based on the resin loading. The yield
is somewhat less than the route using the independently
synthesized, benzylidene nitrone protected N-hydroxy-D-phenyl
alanine. This route, however, utilizes a common resin for the
α-ketoacid and a single reagent for hydroxylamine formation to
be applicable to a larger variety of linear peptides as no separate
synthesis of the hydroxylamine monomers is required.
To resin bound peptide (H2N–D-Phe-Pro-Phe-D-Phe-Asn-(Trt)
Gln-Tyr-Val-(Boc)Orn-Leu-sulphur ylide-NH2, 50.3 mg, approx.
0.20 mmol g−1) in DMF–H2O (10 : 1, 0.5 ml) was added enan-
tiopure (R)-20 (14.3 mg, 0.0498 mmol, 5 equiv) in DMF–H2O
(10 : 1, 0.5 ml) at 0 °C. The reaction mixture was gently stirred
and slowly allowed to warm to rt over 15 h. The resin was
filtered and washed with MeOH and CH2Cl2. The resin bound
peptide was cleaved from the resin by TFA (1 ml) at rt for 1 h.
The mixture was filtered through a small plug of cotton wool
into a centrifuge tube and rinsed with a small volume of TFA.
The filtrate was reduced to minimum volume by a nitrogen
stream and Et2O was added. The white precipitate was collected
by centrifugation and dried under reduced pressure to give the
crude peptide. To a solution of the crude peptide in 1.0 mL
acetone–H2O (1 : 1) at rt was added DMDO (0.175 M in
acetone, 0.23 ml, 0.040 mmol, 4 equiv) and the mixture was
stirred. After 10 min, the reaction was quenched by the addition
of dimethylsulfide (4 drops). The solvent was removed under
reduced pressure to give the crude peptide. The crude peptide
was diluted to 0.001 M with degassed 10 mL 0.1 M (COOH)2
DMF–H2O (50 : 1) and the temperature was raised to 40 °C for
48 h. The solvent was removed under reduced pressure and the
crude material was diluted with MeOH (1 mL). Analysis of this
sample against a standard curve showed that tyrocidine A (7)
was formed in 17% yield for the overall process. For the iso-
lation of tyrocidine A and other peptide, as well as characteriz-
ation data and HPLC traces, see the ESI.‡
Conclusions
In summary, we have developed a new and convenient method
for the synthesis of cyclic peptides by reagentless cyclization of
peptides by α-ketoacid–hydroxylamine (KAHA) ligation. This
strategy allows for the specific cyclization of linear peptides
without the need for side chain protecting groups. Our present
studies demonstrate that unprotected amines, carboxylic acids,
phenols, and alcohols are tolerated. By using a generic sulphur
ylide resin for introducing the α-ketoacid and oxaziridine reagent
20 for introducing the hydroxylamine, cyclic peptides can be
prepared using standard Fmoc-based peptide synthesis. Isolation
of the intermediate linear peptides is not necessary and no other
regents, orthogonal protecting groups, or post-cyclization manip-
ulations are needed. We anticipate that this mechanistically
distinct amide-forming cyclization will find application for the
ever-growing number of natural and unnatural cyclic peptide
targets.
Acknowledgements
This work was supported by the Arnold and Mabel Beckman
Foundation, the David and Lucille Packard Foundation and the
Swiss National Science Foundation (200021–131957). T.F. was a
fellow of the Uehara Memorial Foundation.
Experimental
For experimental procedures, HPLC traces, compound character-
ization, and determination of resin loading please see the ESI.‡
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 5837–5844 | 5843

View Article Online
132, 2889–2891; (b) B. H. Rotstein, V. Rai, R. Hili and A. K. Yudin,
Synthesis of peptide macrocycles using unprotected amino aldehydes,
Nat. Protoc., 2010, 5, 1813–1822.
References
1 (a) M. Zasloff, Antimicrobial peptides of multicellular organisms,
Nature, 2002, 415, 389–395; (b) M. Trabi and D. J. Craik, Circular pro-
teins – no end in sight, Trends Biochem. Sci., 2002, 27, 132–138;
(c) C. Jennings, J. West, C. Waine, D. Craik and M. Anderson, Biosyn-
thesis and insecticidal properties of plant cyclotides: the cyclic knotted
proteins from Oldenlandia affinis, Proc. Natl. Acad. Sci. U. S. A., 2001,
98, 10614–10619.
11 (a) J. W. Trauger, R. M. Kohli, H. D. Mootz, M. A. Marahiel and
C. T. Walsh, Peptide cyclization catalysed by the thioesterase domain of
tyrocidine synthetase, Nature, 2000, 407, 215–218; (b) R. M. Kohli,
M. D. Burke, J. Tao and C. T. Walsh, Chemoenzymatic route to macro-
cyclic hybrid peptide/polyketide-like molecules, J. Am. Chem. Soc.,
2003, 125, 7160–7161.
12 J. W. Bode, R. M. Fox and K. D. Baucom, Decarboxylative
2 (a) D. J. Craik, Seamless proteins tie up their loose ends, Science, 2006,
331, 1563–1564; (b) H. Kessler, Conformation and biological activity of
cyclic peptides, Angew. Chem., Int. Ed. Engl., 1982, 21, 512–523.
3 (a) P. Li, P. P. Roller and J. Xu, Current synthetic approaches to peptide
and peptidomimetic cyclization, Curr. Org. Chem., 2002, 6, 411–440;
(b) S. Jiang, Z. Li, K. Ding and P. P. Roller, Recent progress of synthetic
studies to peptides and peptidomimetic cyclization, Curr. Org. Chem.,
2008, 12, 1502–1542; (c) J. S. Davies, The cyclization of peptides and
depsipeptides, J. Pept. Sci., 2003, 9, 471–501.
condensations of α-ketoacids and N-alkylhydroxylamines:
a
new amide ligation reaction, Angew. Chem., Int. Ed., 2006, 45, 1248–
1252.
13 J. Wu, J. Ruiz-Rodríguez, J. M. Comstock, J. Z. Dong and J. W. Bode,
Synthesis of human GLP-1 (7–36) by chemoselective α-ketoacid–
hydroxylamine peptide ligation of unprotected fragments, Chem. Sci.,
2011, 2, 1976–1979.
14 We have recently reported the synthesis of epi-aza-Surfactin by cycliza-
tion of a linear peptide using the KAHA-amide formation. This work
involved the use of an N-terminal beta-amino acid-derived hydroxyla-
mine; Y. L. Huang, R. Frey, M. E. Juarez-Garcia and J. W. Bode, Syn-
thesis of aza-surfactin and 3-epi-aza-surfactin, Heterocycles, 2012, 84,
1179–1191.
15 L. Ju and J. W. Bode, A general strategy for the preparation of C-terminal
peptide α-ketoacids by solid phase peptide synthesis, Org. Biomol.
Chem., 2009, 7, 2259–2264.
16 L. Ju and A. R. Lippert, Stereoretentive synthesis and chemoselective
amide-forming ligations of C-terminal peptide α-ketoacids, J. Am. Chem.
Soc., 2008, 130, 4253–4255.
4 C. J. White and A. K. Yudin, Contemporary strategies for peptide macro-
cyclization, Nat. Chem., 2011, 3, 509–524.
5 (a) M. Katsara, Round and round we go: cyclic peptides in disease, Curr.
Med. Chem., 2006, 13, 2221–2232; (b) W. F. DeGrado and E. T. Kaiser,
Solid phase synthesis of protected peptides on a polymer-bound oxime,
J. Org. Chem., 1982, 47, 3258–3261; (c) G. T. Bourne,
W. D. Meutermans, P. F. Alewood, R. P. McGeary, M. Scanion,
A. A. Watson and M. L. Smythe, A backbone linker for Boc-based
peptide synthesis and on-resin cyclization: synthesis of stylostatin 1,
J. Org. Chem., 1999, 64, 3095–3101.
6 (a) J. A. Camarero and T. W. Muir, Chemoselective backbone cyclization
of unprotected peptides, Chem. Commun., 1997, 1369–1370;
(b) J. A. Camarero, J. Pavel and T. W. Muir, Chemical synthesis of a
circular protein domain: evidence for folding-assisted cyclization, Angew.
Chem., Int. Ed., 1998, 37, 347–349.
17 S. I. Medina, J. Wu and J. W. Bode, Nitrone protecting groups for enan-
tiopure N-hydroxyamino acids: synthesis of N-terminal peptide hydroxy-
lamines for chemoselective ligations, Org. Biomol. Chem., 2010, 8,
3405–3417.
18 G. F. Gause and M. G. Brazhnikova, Gramicidin S and its use in the treat-
ment of infected wounds, Nature, 1944, 154, 703.
19 M.-K. Wong, C.-W. Yu, W.-H. Yuen and D. Yang, J. Org. Chem., 2001,
66, 3606–3609.
20 M. Tamaki, S. Akabori and I. Muramatsu, CD spectra and cyclisation of
linear pentapeptides as gramicidin S precursors with Orn(Z), Bull. Chem.
Soc. Jpn., 1993, 66, 3113–3115.
21 S. M. Turega, C. Lorenz, J. W. Sadownik and D. Philip, Target-driven
selection in a dynamic nitrone library, Chem. Commun., 2008, 4076–
4078.
22 I. Pusterla and J. W. Bode, The mechanism of the α-ketoacid-hydroxyla-
mine (KAHA) amide-forming ligation, Angew. Chem., Int. Ed., 2012, 51,
513–516.
23 T. Fukuzumi and J. W. Bode, A reagent for the convenient, solid-phase
synthesis of N-terminal peptide hydroxylamines for chemoselective
ligations, J. Am. Chem. Soc., 2009, 131, 3864–3865.
7 (a) O. David, W. J. N. Meester, H. Bieräugel, H. E. Schoemaker,
H. Hiemstra and J. H. van Marrseveen, Intramolecular Staudinger
ligation: a powerful ring-closure method to form medium-sized lactams,
Angew. Chem., Int. Ed., 2003, 42, 4373–4375; (b) R. Kleineweischede
and C. P. R. Hackenberger, Chemoselective peptide cyclization by trace-
less Staudinger ligation, Angew. Chem., Int. Ed., 2008, 47, 5984–5988.
8 K. Sasaki and D. Crich, Cyclic peptide synthesis with thioacids, Org.
Lett., 2010, 12, 3254–3257.
9 (a) R. A. Turner, A. G. Oliver and S. R. Lokey, Click chemistry as a
macrocyclization tool in the solid-phase synthesis of small cyclic
peptides, Org. Lett., 2007, 9, 5011–5014; (b) V. D. Bock, D. Speijer,
H. Hiemstra and J. H. van Maarseveen, 1,2,3-triazoles as peptide bond
isosteres: synthesis and biological evaluation of cyclotetrapeptide mimics,
Org. Biomol. Chem., 2007, 5, 971–975.
10 (a) R. Hili, V. Rai and A. K. Yudin, Macrocyclization of linear
peptides enabled by amphoteric molecules, J. Am. Chem. Soc., 2010,
5844 | Org. Biomol. Chem., 2012, 10, 5837–5844
This journal is © The Royal Society of Chemistry 2012