Exploration of an imide capture/N,N-acyl shift sequence for asparagine native peptide bond formation

Article abstract of DOI:10.1016/j.bmc.2013.02.053

Imide capture of a C-terminal peptidylazide with a side-chain thioacid derivative of an N-terminally protected aspartyl peptide leads to the formation of an imide bond bringing the two peptide ends into close proximity. Unmasking of the N^α protecting group and intramolecular acyl migration results in the formation of a native peptide bond to asparagine.

Full text of DOI:10.1016/j.bmc.2013.02.053

Bioorganic & Medicinal Chemistry 21 (2013) 3479–3485
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Exploration of an imide capture/N,N-acyl shift sequence for
asparagine native peptide bond formation
Reda Mhidia a, Emmanuelle Boll a, Fabien Fécourt b, Mikhail Ermolenko b, Nathalie Ollivier a,
Kaname Sasaki b, David Crich b, Bernard Delpech b, Oleg Melnyk a,
⇑
a CNRS UMR 8161, Univ. Lille Nord de France, Institut Pasteur de Lille, 59021 Lille, France
b Institut de Chimie des Substances Naturelles, CNRS, 91198 Gif-sur-Yvette Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 24 March 2013
Imide capture of a C-terminal peptidylazide with a side-chain thioacid derivative of an N-terminally pro-
tected aspartyl peptide leads to the formation of an imide bond bringing the two peptide ends into close
proximity. Unmasking of the N protecting group and intramolecular acyl migration results in the forma-
a
Keywords:
tion of a native peptide bond to asparagine.
Peptide bond
Imide capture step
N,N-Acyl shift
Asparagine
Ó 2013 Elsevier Ltd. All rights reserved.
Thioacid
Acylazide
1. Introduction
Sanger and Mukaiyama reagents,17 or arylsulfonamides18 enables
the block synthesis of peptides. C-terminal peptide thioacids have
The design of novel amide bond forming reactions is one of the
greatest challenges in synthetic organic chemistry today.1 In par-
ticular, methods enabling the formation of native peptide bonds
between peptide segments are of the utmost importance for
assembling complex peptide scaffolds or proteins.
been used also in the design of elegant capture/intramolecular
rearrangement sequences leading to native or peptidomimetic
amide bond formation.19 In particular, the capture step in these se-
quences exploits the nucleophilic properties of the thioacid func-
tion
toward
bromoalkyl,20
disulfide,21
isocyanate
or
The peptide bond can be formed by the direct reaction of a pep-
tide segment featuring a C-terminal activated carbonyl group with
isothiocyanate,22 or aziridine functionalities.23
The reaction of thioacids with azides is another route to amides.
The reaction is particularly effective with electron-poor azides
such as sulfonyl azides,24–30 and gives access to stable acylsulfona-
mide-linked peptide28 or protein30 conjugates of the type fre-
quently used as safety catch linkers.31–35 We have also recently
reported on the potential of the imide ligation, that is, on the reac-
tion of a C-terminal peptidyl thioacid with an azidocarbonyl deriv-
ative for the assembly and enzyme-triggered disassembly of
peptide-drug conjugates.36 We describe here our exploration on
the potential of a related Asparagine Native Peptide Ligation
(AsnNPL), which relies on an imide capture/intramolecular N,N-
acyl shift reaction sequence.
the free a
-amino group of the second segment.2–4 Alternately, the
peptide bond forming process can be initiated by the selective for-
mation of a non-peptidic bond between the two segments in a first
step, which brings the C- and N-terminal ends into close proximity.
The native peptide bond is then formed in a subsequent intramo-
lecular rearrangement step.5 Several native peptide bond forming
processes rely on such a capture/intramolecular rearrangement se-
quence, among which are the native peptide ligation methods used
for protein total synthesis starting from unprotected peptide seg-
ments, that is, native chemical ligation,6,7 bis(2-sulfanyleth-
yl)amido (SEA) native peptide ligation,8–12 the traceless
Staudinger ligation,13,14 or the decarboxylative condensation of
N-alkylhydroxylamines and a
-ketoacids.15
Thioacids are of interest in this area due to their unique reactiv-
ity compared to other functional groups present within polypep-
tides.16 Activation of C-terminal peptide thioacids by silver ion,3,4
2. Results and discussion
2.1. General strategy
AsnNPL requires a C-terminal peptide azide 1 and an N-terminal
thioaspartyl peptide 2 as starting peptide segments (Scheme 1).
⇑
Corresponding author. Tel.: +33 320871214.
E-mail address: Oleg.Melnyk@ibl.fr (O. Melnyk).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.02.053

3480
R. Mhidia et al. / Bioorg. Med. Chem. 21 (2013) 3479–3485
O
nium imide intermediate 8. The modest yield for this step was due
O
H
N
HS
to the formation of other imide products which were not isolated.
This aspect is discussed in more detail later. This compound proved
to be stable in mildly acidic aqueous solution due to the protection
of the bAla amino group by protonation, and therefore could be
purified by HPLC and fully characterized.38 In particular, its 1H
NMR spectrum showed a peak at d 11 in accordance with the pro-
posed imide structure.
+
peptide
peptide
N3
HN
PG
R
O
1
2
imide ligation capture step
O
O
H
N
peptide
NH
HN
PG
peptide
R
2.3. Intramolecular N,N-acyl shift reaction
O
3
Imides are mildly activated derivatives of carboxylic acids and
therefore have been used as amide prodrugs due to their propen-
sity to be hydrolyzed in biological fluids.39 In another application,
Tsunoda et al., developed a method for the hydrolysis of N-mono-
substituted amides based on the formation of acyclic imides fol-
lowed by an intramolecular N,O-acyl migration reaction.40
Recently, we reported on the design of a novel alcohol prodrug sys-
tem in which the unmasking step involved the intramolecular at-
- PG
O
O
H
N
NH
peptide
peptide
R
H2N
O
4
N,N-acyl shift
O
NH2
O
tack of
a primary amine on an imide carbonyl moiety
H
N
(Scheme 3).36 In this system, the alcohol moiety acted as a leaving
group during the intramolecular O,N-imide carbonyl migration
reaction. To the best of our knowledge, the involvement of imides
in intramolecular N,N-acyl migration reactions resulting in the
departure of a primary amide has not been reported in the
literature.
peptide
N
H
peptide
O
R
5
Scheme 1. Asparagine native peptide ligation (AsnNPL) relies on an imide capture/
N,N-acyl shift sequence.
The capture step is an imide ligation between the acyl azide and
the thioacid moieties, which results in the formation of imide-
linked intermediate 3. Removal of the temporary protecting group
from the N-terminal position of the thioaspartyl segment after the
capture step enables the migration of the acyl segment derived
In a first approach, we used imide 8 to probe the feasibility of
the intramolecular N,N-acyl shift (Scheme 2). The pKa of the imide
group in compound 8 was estimated to be 7.8 using the SPARC41
algorithm. Consequently, this group was expected to be in its
protonated form at pH 7 and thus to act as a mildly activated car-
boxylic acid set up to enable the N,N-acyl shift to occur through a
6-membered cyclic intermediate.
In agreement with this hypothesis, solubilization of imide 8 in
water at pH 7 resulted within 1 min in the clean formation of
dipeptide 10 (Fig. 1), which was found to be identical by 1H and
13C NMR spectroscopy with a reference compound prepared by
standard Fmoc-solid phase peptide synthesis (SPPS) (see Supple-
mentary data).
from the acyl azide 1 from the carboxamide nitrogen to the a-ami-
no group. This rearrangement is an N,N-acyl shift which occurs
presumably through a 6-membered cyclic intermediate,37 and is
favored by the mild electrophilicity of the imide carbonyl. The rear-
rangement results in the formation of a native peptide bond to Asn.
The carboxamide nitrogen of the final asparagine side-chain de-
rives from the C-terminal azido group of acyl azide segment 1.
2.2. Proof-of-concept of the imide capture step using model
amino acid derivatives
2.4. Proof-of-concept of AsnNPL using model peptide segments
To illustrate the method we first used simple amino acid deriv-
atives, that is, Fmoc-Ala-N3 6 and Boc-bAla-SH 7 (Scheme 2), with
the latter mimicking the N-terminal thioaspartyl residue of seg-
ment 2 (Scheme 1). The alanine derivative was protected by a 9-
fluorenylmethoxycarbonyl (Fmoc) group to facilitate the UV detec-
tion of the compounds during HPLC. The b-alanine derivative was
protected by a tert-butyloxycarbonyl (Boc) group since the inter-
mediate imide group was expected to be stable in trifluoroacetic
acid (TFA).
2.4.1. Imide capture step
We next examined the imide capture step using model N-termi-
nal thioaspartyl peptide 12 and C-terminal peptide azide 13
(Scheme 4). In a first approach, peptide azide 13a was prepared
using diphenylphosphoroazidate (DPPA) in the presence of trieth-
ylamine. DPPA allows the prompt and straightforward preparation
of C-terminal peptide azide, but results in the partial racemization
of the C-terminal residue (ꢀ30%
D-Ala for 13a). Once formed, pep-
tidylazide 13a was reacted with N-terminal thioaspartyl peptide
12 in DMF42 at room temperature resulting in the formation of
the target imide 14a, but also unexpectedly of two other products
Imide ligation of Fmoc-Ala-N3 6 with Boc-bAla-SH 7 followed by
removal of the Boc group in TFA successfully furnished the ammo-
O
O
O
1) DMF,DIEA, rt
25 %
R
R
O
H
N
Fmoc-NH
HS
HN
Boc
N
H
N3
O
PG
SH
PG
O
O
Fmoc-NH
+
R
N3
N
H
N
H
R
+
H3N+
CF3CO3-
2) TFA, rt, 70%
O
O
O
O
6
7
8
O
R
O
H
N
R
O
N,N-acyl
shift
- PG
O
O
Fmoc-NH
pH > 7
N
H
ROH
+
NH
H2N
R
Fmoc-NH
HN
N
H
NH2
O
O
NH2
O
9
10
Scheme 3. An intramolecular O,N-imide carbonyl shift reaction used in the design
Scheme 2. Proof-of-concept study using b-alanine as an aspartic acid mimetic.
of an alcohol prodrug system.36

R. Mhidia et al. / Bioorg. Med. Chem. 21 (2013) 3479–3485
3481
Abs (215 nm)
16
14a
10
15a
t = 5 h
(a)
8
13a
O
O
O
O
H
H
N
12
- H+
H+
N
N
Fmoc
N
Fmoc
H
H
t = 0
H2N
NH3+
CF3CO2-
RP-HPLC
8
8
C18 column
pH > 7
O
UVdetection
at215 nm
20.0
25.0
30.0
35.0
time (min)
60
50
40
30
20
10
0
O
H
N
16
14a
15a
NH2
N
Fmoc
H
10
(b)
20.0
25.0
30.0
35.0
time (min)
Figure 1. HPLC analysis of the rearrangement of imide
8 into dipeptide 10.
Continuous trace: imide 8, dashed trace: dipeptide 10 obtained after dissolving
imide 8 in water at pH 7.
0.3
0,3
0.8
0,8
1.3
1,3
1.8
1,8
N-terminal thioaspartyl peptide 12 (eq)
O
Figure 2. Imide ligation of peptides 12 and 13a. (a) HPLC analysis of the reaction
mixture immediately after mixing (t = 0) or after 5 h at rt (12/13a: 1/1 by mol). (b)
Influence of the stoichiometry of the reaction on the proportion of imides 14a, 15a
and 16.
O
HO
H2N
PAFAQGA-NH2
11
1) Boc2O, Et3N
water/acetonitrile
2) NMM, ClCOOEt
H2S (60 % overall)
anion then gives the scrambled thioacid and azido peptides. A mix-
ture of scrambled imides was also obtained on mixing thiobenzoic
acid and 3-phenylpropanoyl azide (see Supplementary data), and it
was demonstrated that the reaction of authentic benzoic thioanhy-
dride or 3-phenylpropanoic thioanhydride with sodium azide fur-
nished the corresponding symmetrical imides in 68–75% yield, in
support of the proposed reaction pathway. The occurrence of a
scrambling process during the reaction of a thiocarboxylic acid
with an acyl azide has not been previously discussed to our knowl-
edge. In their pioneering studies, Williams and co-workers de-
scribed the reaction of thioacids with azides such as
benzylazidocarbonate or benzenesulfonylazide.24,25 Unlike the acyl
azides examined in this study, no scrambling reaction was ob-
served with these electron deficient azides. More recently, we
examined the reaction of C-terminal peptide thioacids with azido-
formates as described in Scheme 3.36 Here again, no scrambling
reaction was observed and imides were isolated in good yield.
In an attempt to improve the proportion of dissymmetrical
imide 14a in the ligation mixture, we varied the stoichiometry of
the reaction (Fig. 2B). Unfortunately, the proportion of dissymmet-
rical imide 14a did not vary significantly over the stoichiometry
range examined. In summary, the imide ligation step proceeds as
anticipated but is complicated by a background reaction involving
the scrambling of thioacid and azido peptide components.
O
O
Boc-GFGQGFG-NH
O
HS
BocHN
PAFAQGA-NH2
N3
13
12
a : L+D-Ala, b : L-Ala, c : D-Ala
O
O
NH
Boc-GFGQGFG-NH
PAFAQGA-NH2
HN
O
Boc
O
14 (15%)
PAFAQGA-NH2
O
O
BocNH
O
Boc-GFGQGFG-NH
NH
Boc-GFGQGFG-NH
NH
HN
PAFAQGA-NH2
O
O
Boc
15
16
Scheme 4. Study of imide capture step using model peptides. GFGQGFG is the one-
letter code for GlyPheGlyGlnGlyPheGly. PAFAQGA is the one-letter code for
ProAlaPheAlaGlnGlyAla.
corresponding to the symmetrical imides 15a and 16 (Fig. 2A).
Interestingly, monitoring of the ligation reaction by HPLC revealed
the scrambling of the thioacid and azido functionalities, that is, the
in situ formation of Boc-GlyPheGlyGlnGlyPheGlyAla-SH and Boc-
Asp(N3)ProAlaPheAlaGlnGlyAla-NH2 (see Supplementary data)
within a few seconds after mixing peptides 12 and 13a. This scram-
bling of the thioacid and azido moieties, which explains the forma-
tion of symmetrical imides 15a and 16, might proceed through the
formation of a thioanhydride by nucleophilic attack of the thio-
carboxylate group of peptide 12 on the C-terminal carbonyl group
of azido peptide 13a. Subsequent reaction with the released azide
The imide ligation step also was performed with epimerization
free C-terminal azido peptides 13b (L-Ala) and 13c (D-Ala) to probe
the issue of partial racemization of the C-terminal residue of the
peptide azide segment during AsnNPL. Compounds 13b and 13c
were prepared by reacting the corresponding hydrazides Boc-
GlyPheGlyGlnGlyPheGlyAla-NHNH2 or Boc-GlyPheGlyGlnGlyPhe-
Gly-D-Ala-NHNH2 with tert-butyl nitrite (see Supplementary data).
This method is known to be non-racemizing.43–45 The structure of
imides 14a–c was confirmed by 1H and 13C NMR spectroscopy (see

3482
R. Mhidia et al. / Bioorg. Med. Chem. 21 (2013) 3479–3485
Abs (215 nm)
17a
O
O
NH
Boc-GFGQGFG-NH
PAFAQGA-NH2
HN
O
Boc
14
(a)
14a (L+D-Ala)
14b (L-Ala)
18a
18a
(b)
(c)
14c (D-Ala)
11.8
11.6
11.4
11.2
11.0
10.8
10.6
10.4
10.2 ppm
Figure 3. Imide region of the 1H NMR spectra for imides 14a–c (300 MHz, DMF-d7,
20 °C, 11 mM).
17a
O
O
PAFAQGA-NH2
NH
Boc-GFGQGFG-NH
HN
O
Boc
15.0
20.0
time (min)
14a,b a : L+D-Ala, b : L-Ala
TFA
Figure 4. HPLC analysis of: (a) imide 17a. The peak eluting after 17a is isobaric to
17a and might be due to the presence of a minor imide conformer; see main text.
(b) Rearranged peptide 18a formed after dissolving 17a in water at pH 7 (1 min, rt).
Both compounds were co-injected in (c).
O
O
PAFAQGA-NH2
NH
H-GFGQGFG-NH
NH2
O
17a,b
pH 7
imide 17a proceeded cleanly in less than 1 min to give peptide
18a (Fig. 4), which was found to be identical by HPLC and capillary
electrophoresis to a reference peptide synthesized by standard
Fmoc-SPPS. Moreover, automated Edman microsequencing of pep-
tide 18a showed the signature of an Asn residue at position 9, in
between Ala 8 and Pro 10. Importantly, chiral GC–MS analysis of
H2N
O
O
O
PAFAQGA-NH2
NH
H-GFGQGFG-NH
18a,b (63 % overall)
peptide 18b after acid hydrolysis revealed a D-Ala content of only
0.28%,48 showing that the imide capture/N,N-acyl shift sequence
proceeded with no significant racemization of C-terminal Ala
residue.
Scheme 5. N,N-Acyl shift with model imide peptides 14a,b.
Fig. 3 and Supplementary data), and the imide region of the 1H
NMR spectrum showed two main peaks for 14a at d 10.81 (0.3
H) and 10.86 (0.7 H), whereas 14b (L-Ala) and 14c (D-Ala) showed
3. Conclusions
only one main peak at d 10.86 and 10.81, respectively. Note that
peaks of low intensity relative to the main imide proton are pres-
ent in the 1H NMR spectra of imides 14b and 14c. These minor
peaks, which are also present in the 1H NMR spectrum of 14a,
might be due to minor imide conformers.46,47 Acyclic imides such
as imide 14 may adopt three planar conformations, cis–cis, cis–
trans, and trans–trans, which differ by the relative orientations of
the two carbonyl groups to the central NH bond. Of the three pos-
sible conformers, the cis–trans is the most favored in simple ali-
phatic imides. The trans–trans form can be present as a minor
conformer, whereas the cis–cis conformer is usually not observed.
Note that unsymmetrical imides such as imide 14 can adopt poten-
tially two cis–trans conformations. Consequently, the presence of
several isobaric species for imide 14 is not surprising given the
known conformational properties of unsymmetrical imides.
Overall, these data show that the imide capture step is non-
racemizing and that the presence of two species in 14a is due to
the partial racemization of C-terminal Ala residue during the prep-
aration of C-terminal azido peptide 13a.
In conclusion, we have used imide ligation, that is, the reaction
of a thioacid with an acyl azide, as a capture step in the design of a
novel peptide bond forming reaction. The peptide bond is formed
after an intramolecular N,N-acyl shift reaction involving the nucle-
ophilic attack of an
a-amino group on the imide group. The N,N-
acyl shift occurs through a 6-membered cyclic intermediate. An
asparagine residue is formed at the ligation junction. The whole
process proceeds with no detectable racemization of the preceding
residue (Ala in this study). While the scope of the substrates tested
in this study in DMF has been limited to those bearing unfunction-
alized side chains, the established compatibility of peptidyl thioac-
ids and imides with more complex side chains and aqueous media
suggests that the method may have broader scope if and when the
issue of thioacid scrambling is solved. Further work in this direc-
tion is in progress.
4. Experimental section
4.1. Proof-of-concept of AsnNPL using model amino acids
2.4.2. N,N-Acyl rearrangement step
Next, the isolated imide-linked peptides 14a and 14b were
deprotected in TFA and rearranged in water at pH 7 (Scheme 5).
As for model imide 8 (Scheme 2), rearrangement of deprotected
4.1.1. Synthesis of thioacid 7
To a solution of Boc-bAla-OH (1.89 g, 10 mmol) dissolved in
10 mL of CH2Cl2 was added N-methylmorpholine (NMM, 11 mmol,

R. Mhidia et al. / Bioorg. Med. Chem. 21 (2013) 3479–3485
3483
1.22 mL). The reaction mixture was immersed in an ice bath and
ethyl chloroformate (12 mmol, 1.15 mL) was added to the mixture
which was stirred for 5 min. Hydrogen sulfide was then bubbled
for 5 min and 2 equivalents of NMM (22 mmol, 2.44 mL) were
added. The reaction mixture was stirred at rt for 30 min. The sol-
vent was then evaporated in vaccuo and the yellow residue ob-
tained was redissolved in 30 mL of deionized water, acidified
with 0.1 N HCl to pH 2–3. The aqueous solution was extracted with
3 Â 30 mL of ethyl acetate and the organic phase was dried with
anhydrous MgSO4, filtered and evaporated in vacuo. Thioacid 7
was isolated as a yellow oil with a yield of 92% (1.97 g) and used
directly in the next experiment without further purification.
ESI-MS [M+H]+ m/z calcd (monoisotopic) 206.0, found 206.2.
1H NMR (300 MHz, CDCl3, sodium trimethylsilylpropionate
used as internal reference for chemical shifts) d 4.87 (s, 1H), 3.32
(q, J = 12.1, 6.1 Hz, 2H), 2.82 (t, J = 5.9 Hz, 2H), 1.37 (s, 9H).
13C NMR (75 MHz, CDCl3) d 197.10 (s), 155.87 (s), 46.03 (s),
36.27 (s), 28.49 (s).
Peptide 10 was identical by RP-HPLC and 1H or 13C NMR to a ref-
erence compound produced by conventionnal Fmoc-SPPS (see Sup-
plementary data).
ESI-MS calcd for [M+H]+ m/z = 382.18 (monoisotopic); obsd m/
z = 382.2.
HR-MS calcd for [M+H]+ m/z = 382.17613, found 382.17542.
1H NMR (300 MHz, DMF-d7, sodium trimethylsilylpropionate
used as internal reference) d 7.94 (d, J = 7.3 Hz, 2H), 7.77 (t,
J = 6.3 Hz, 2H), 7.55–7.29 (m, 5H), 4.29 (s, 2H), 4.18 (t, J = 14.5,
7.2 Hz, 1H), 3.41 (d, J = 6.1 Hz, 2H), 2.39 (t, J = 6.7 Hz, 2H), 1.38–
1.29 (m, 3H).
13C NMR (75 MHz, DMF-d7, sodium trimethylsilylpropionate
used as internal reference) d 173.21, 158.69–155.88 (m), 146.70–
144.77 (m), 141.75, 128.80, 128.24, 126.54, 121.16, 67.30, 51.80,
48.18, 36.60, 36.16, 19.18.
4.2. Proof-of-concept of AsnNPL using model peptide segments
The procedure is illustrated with the synthesis of peptide imide
14b and Asn peptide 18b. Other experimental procedures and
characterization are described in the Supplementary data.
4.1.2. Synthesis of imide 8
Preparation of Fmoc-Ala-N3 6. Fmoc-Ala-OH (600 mg,
1.9 mmol) was dissolved in 1.8 mL of anhydrous DMF. The solution
was cooled in an ice bath. Triethylamine (TEA, 352
lL, 2.50 mmol)
4.2.1. Synthesis of peptide 12
and diphenylphosphoroazidate (DPPA, 541 L, 2.50 mmol) were
l
Preparation of peptide 11. The solid phase synthesis of peptide
H-DPAFAQGA-NH2 was performed using the Fmoc/tert-butyl strat-
egy on NovaSyn TGR resin (0.2 mmol/g, Novabiochem, 0.5 mmol
scale) using an automated peptide synthesizer. The coupling of
amino acids was carried out with 2.5 mmol of each amino acid,
2.25 mmol of HBTU activator, and 5 mmol of DIEA. Final deprotec-
tion and cleavage of the peptide from the resin were performed
with a mixture of TFA/triisopropylsilane (TIS)/H2O: 95/2.5/2.5 by
vol. The resin beads were filtered and the filtrate was precipitated
in diethyl ether/heptane: 1/1 by vol. The peptide was then dis-
solved in water and lyophilized. The peptide was obtained as a
white powder with a yield of 87% (386 mg).
added to the solution. The reaction mixture was kept under mag-
netic stirring for 4 h. This solution was used directly in the next
step.
The above solution of Fmoc-Ala-N3 in DMF (1.6 mmol) was di-
luted in 18 mL of DMF. To this solution was added Boc-bAla-SH 7
(328 mg, 1.60 mmol) and TEA (224 lL, 1.60 mmol). The reaction
mixture was colored yellow accompanied by gas evolution. The
reaction mixture was stirred at rt for 6 h, after which the DMF
was evaporated in vacuo. The yellow residue was redissolved in
50 mL of water/acetonitrile: 4/1 by vol containing 0.05% TFA by
vol, purified by RP-HPLC and lyophilized. The compound was ob-
tained as a white powder (184 mg, 24%). LC–ESI MS analysis of
Boc-protected imide intermediate gave [M+H]+ calcd (monoiso-
topic) 482.23, found 482.31
MALDI-Tof (positive mode). Matrix:
a-cyano-4-hydroxycin-
namic acid (HCCA). Calcd for [M+H]+ 775.3 (monoisotopic), found
775.2.
Because RP-HPLC purification led to a partial removal of Boc
group, compound was further deprotected with TFA. For this,
Boc-protected imide intermediate (100 mg, 0.2 mmol) was treated
with 20 mL of TFA/CH2CL2: 1/1 by vol at rt for 30 min. The solvent
was evaporated to dryness in vaccuo. The residue obtained was
redissolved in 20 mL of water/acetonitrile: 9/1 by vol containing
0.05% of TFA by vol and purified by RP-HPLC. The fractions contain-
ing imide 8 were combined and lyophilized to give the product
72 mg of a white powder (70%).
The peptide H-DPAFAQGA-NH2 (115 mg, 0.13 mmol) was next
dissolved in 1.3 mL of a mixture of water/acetonitrile: 1/1 by vol.
Boc2O (41 lL, 0.19 mmol) and TEA (54 lL, 0.39 mmol) was added
to this solution. The reaction mixture was stirred for 3 h at rt.
The peptide was purified by RP-HPLC and lyophilized. Peptide
Boc-DPAFAQGA-NH2 11 was obtained as a white powder with a
yield of 70% (m = 79 mg).
MALDI-Tof (positive mode). Matrix:
a-cyano-4-hydroxycin-
namic acid (HCCA). Calcd for [M+H]+ 875.4 (monoisotopic), found
875.4; [M+Na]+ 897.4.
1H NMR (300 MHz, DMF) d 11.07 (s, 1H); 8.50 (s, 1H); 7.96 (d,
J = 7.5 Hz, 2H); 7.84 (d, J = 7.3 Hz, 1H); 7.79 (dd, J = 7.2, 4.5 Hz,
2H); 7.47 (t, J = 7.5 Hz, 2H); 7.37 (td, J = 7.5, 1.1 Hz, 2H); 4.56 (p,
J = 7.2 Hz, 1H); 4.38–4.23 (m, 3H); 3.65 (s, 1H); 3.39 (t, J = 6.6 Hz,
2H); 3.24 (dd, J = 10.2, 3.9 Hz, 2H); 1.43 (d, J = 7.2 Hz, 3H).
13C NMR (75 MHz, CDCl3) d 173.94; 172.01; 156.31; 144.25;
141.19; 127.77; 127.18; 125.45; 120.14; 66.41; 51.42; 47.08;
35.09–34.76 (m); 16.86 .
Preparation of peptide 12. Peptide Boc-DPAFAQGA-NH2 11
(102 mg, 0.12 mmol) was dissolved in 1.3 mL of DMF/acetonitrile:
1/1 by vol. The solution was cooled to À45 °C under argon, and
NMM (14 lL, 0.13 mmol) and ethyl chloroformate (25 lL,
0.24 mmol) were added. After 5 min, the reaction mixture was
bubbled with hydrogen sulfide for 5 min. The flow of hydrogen sul-
fide was maintained during the addition of NMM (29 lL,
HR-MS analysis of imide 8. Calcd for [M+H]+ m/z = 382.17613,
found 382.17553.
0.26 mmol). The reaction mixture was stirred for an additional
15 min still with hydrogen sulfide bubbling. The bubbling of
hydrogen sulfide was then stopped and the reaction mixture was
allowed to warm at rt and was stirred for 1 h.
The solvents were then evaporated in vacuum and the residue
obtained was purified by RP-HPLC using triethylamine acetate buf-
fer (TEAA) buffer. The peptide Boc-D(SH)PAFAQGA-NH2 12 was
lyophilized twice to remove excess TEAA. It was isolated as a white
powder with a yield of 60% (69 mg).
4.1.3. Rearrangement of imide 8 into dipeptide 10
Imide 8 (16 mg) was dissolved in acetonitrile/0.1 M pH 7.3 so-
dium phosphate buffer: 7/3 by vol (16 mL). After 15 min, the reac-
tion mixture was extracted with ethyl acetate (2 Â 20 mL). The
organic phase was dried over sodium sulphate, evaporated to dry-
ness, dissolved in 50% by vol aqueous acetonitrile and lyophilized
to give 16 mg (ꢀquantitative) of dipeptide 10.
LC–MS analysis of peptide 12. MS trace (positive mode). Calcd
for [M+H]+ 891.4 (monoisotopic), found 891.9.

3484
R. Mhidia et al. / Bioorg. Med. Chem. 21 (2013) 3479–3485
MALDI-Tof (negative mode). Matrix: 2,5-dihydroxybenzoic acid
13C NMR (75 MHz, DMF-d7) d 174.87 (d, J = 6.4 Hz), 174.53 (s),
173.92 (s), 173.47 (s), 173.12 (s), 172.70 (s), 172.50 (s), 172.33
(s), 172.21 (s), 172.07 (s), 171.72 (s), 170.65 (s), 170.05 (s),
169.94 (s), 169.42 (s), 169.08 (s), 156.60 (s, J = 20.5 Hz), 155.99–
155.78 (m), 138.46 (s), 138.35 (s), 138.28–138.23 (m), 129.60 (d,
J = 7.1 Hz), 128.50 (s), 126.65 (s), 78.93 (s), 78.59 (d, J = 8.5 Hz),
61.73 (s), 55.61 (s, J = 19.1 Hz), 55.35 (s), 54.15 (s), 53.69 (s),
50.88 (s), 50.05 (s), 49.07 (s), 43.77 (s), 43.12 (s), 42.90 (s),
42.72–42.49 (m), 37.62 (s), 37.34 (s), 31.92 (s), 28.06 (s), 27.76
(d, J = 12.7 Hz), 25.11 (s), 17.95 (s), 17.13 (s), 16.73 (s).
(DHB). Calcd for [MÀH]À 889.3 (monoisotopic), found 889.4.
4.2.2. Synthesis of peptide 13b
Peptide 13b was obtained by nitrosation of peptide hydrazide
Boc-GFGQGFGA-NHNH2 with tert-butylnitrite in acidic media. For
this, synthesis of peptide hydrazide H-GFGQGFGA-NHNH2 was
performed using the Fmoc/tert-butyl strategy on a Fmoc-Ala-Sasrin
resin (0.25 mmol scale, Bachem) in an automated peptide synthe-
sizer. The coupling of amino acids was carried out with 1.25 mmol
of each amino acid, 1.12 mmol of HBTU activator, and 2.5 mmol of
DIEA.
4.2.4. Synthesis of peptide 17b
The cleavage of the peptide from the resin was performed using
of hydrazine hydrate (20 mL, 20% in N,N-dimethylacetamide
(DMA), 2 Â 24 h). The resin was filtered and washed after each
treatment with 10 mL of DMA. The fractions obtained were pooled
and evaporated to dryness. The residue obtained was treated for
1 h with 40 mL TFA/TIS: 95/5 by vol. The reaction mixture was
evaporated to dryness. The residue was redissolved in 30 mL of
water containing 1% TFA by vol and purified by RP-HPLC. The frac-
tions containing the target peptide were pooled and lyophilized.
The peptide H-GFGQGFGA-NHNH2 was obtained as a white pow-
der with a yield of 60% (143 mg).
The deprotection of peptide 14b (11 mg, 6.4 lmol) was per-
formed with a mixture of TFA/CH2Cl2/TIS: 50/49/1 by vol. The reac-
tion mixture was stirred for 40 min at rt. The solvent was then
evaporated in vaccuo and the residue obtained was redissolved
in deionized water (30 mL), desalted with SPE C18 Column and
lyophilized. Peptide 17b was obtained as a white powder with a
yield of 98 % (11 mg).
MALDI-Tof (positive mode). Matrix:
a-cyano-4-hydroxycin-
namic acid (HCCA). Calcd for [M+H]+ 1495.71 (monoisotopic),
found 1495.7.
MALDI-Tof analysis of peptide H-GFGQGFGA-NHNH2. Matrix:
2,5-dihydroxybenzoic acid (DHB). Calcd for [M+H]+ 754.3, found
754.2.
4.2.5. Synthesis of peptide 18b
Peptide 17b (8 mg, 4.7 lmoL) was dissolved in phosphate buf-
fered saline (0.01 M, pH 7.4, 10 mL). The reaction mixture was kept
at rt for 1 h and then purified by RP-HPLC. The fractions containing
peptide 18b were combined and lyophilized to give 5.1 mg of a
white powder (63%).
Next, the peptide H-GFGQGFGA-NHNH2 (143 mg, 147
was dissolved in 5 mL of water/acetonitrile: 1/1 by vol. To this
solution was added TEA (206 L, 1.47 mmol) and then (Boc)2O
(33 L, 147 mol). The reaction mixture was stirred at rt for
lmol)
l
l
l
MALDI-Tof (positive mode). Matrix: a-cyano-4-hydroxycin-
20 h and then evaporated to dryness. The peptide was solubi-
lized in an aqueous solution containing 0.05% TFA by volume,
purified by RP-HPLC and lyophilized. The peptide Boc-
GFGQGFGA-NHNH2 was obtained as a white powder with a yield
of 46% (66 mg).
namic acid (HCCA). Calcd for [M+H]+ 1495.71 (monoisotopic),
found 1495.6.
Chiral GC–MS analysis of peptide 18b after acid hydrolysis re-
vealed a D-Ala content of 0.28% (CATT, Germany).
The same imide capture/N,N-acyl shift sequence allowed the
synthesis of peptide 18a. Peptide 18a was identical by HPLC and
capillary electrophoresis to a reference compound produced by
conventional Fmoc-SPPS. Edman microsequencing showed the
presence of Asn 9 residue in between Ala 8 and Pro 10, in accord
with the proposed structure.
ESI-MS calcd for [M+H]+ 854.4 (monoisotopic), found 854.7.
MALDI-Tof analysis of peptide Boc-GFGQGFGA-NHNH2. Matrix:
DHB. Calculated for [M+H]+ 854.4 (monoisotopic), found 854.0.
Finally, the peptide Boc-GFGQGFGA-NHNH2 (30 mg, 32
was dissolved in 195 L of anhydrous DMF. The solution was
placed under argon at À25 °C. The solution was acidified with
40 L of 4 N HCl in 1,4-dioxane. tert-Butylnitrite (39 L of a 10%
solution in DMF, 35 mol) was added. The reaction mixture was
lmol)
l
l
l
Acknowledgments
l
stirred for 15 min and added dropwise to 20 mL of diethyl ether
to precipitate peptide azide 13b. The white suspension was centri-
fuged, washed twice with 10 mL of diethyl ether and dried under
vacuum. The peptide 13b was stored at À20 °C and used immedi-
ately without further purification.
We acknowledge financial support from Région Nord Pas de Ca-
lais, ANR Grant ‘click–unclick’ (11-BS07-02201) and the European
Community.
Supplementary data
4.2.3. Synthesis of imide 14b
Supplementary data (experimental procedures and character-
ization data for all compounds) associated with this article can
be found, in the online version, at http://dx.doi.org/10.1016/
j.bmc.2013.02.053.
The C-terminal azido peptide 13b (35
DMF (200 L). The N-terminal thioaspartyl peptide 12 (31 mg,
35 mol) was dissolved in DMF (400 L) and added to this solu-
lmol) was dissolved in
l
l
l
tion. The reaction mixture was stirred for 24 h at rt, then diluted
with 20 mL of water containing 0.05% TFA, and purified by RP-
HPLC to give peptide imide 14b with a yield of 19% (11.4 mg).
References and notes
MALDI-Tof (positive mode). Matrix:
a-cyano-4-hydroxycin-
1. Pattabiraman, V. R.; Bode, J. W. Nature 2011, 480, 471.
2. Aimoto, S. Biopolymers 1999, 51, 247.
3. Blake, J.; Li, C. H. Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 4055.
4. Blake, J. Int. J. Pept. Protein Res. 1981, 17, 273.
5. Coltart, D. M. Tetrahedron 2000, 56, 3449.
6. Kent, S. B. Chem. Soc. Rev. 2009, 38, 338.
7. Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. Science 1994, 266, 776.
8. Raibaut, L.; Ollivier, N.; Melnyk, O. Chem. Soc. Rev. 2012, 41, 7001.
9. Ollivier, N.; Vicogne, J.; Vallin, A.; Drobecq, H.; Desmet, R.; El-Mahdi, O.;
Leclercq, B.; Goormachtigh, G.; Fafeur, V.; Melnyk, O. Angew. Chem., Int. Ed.
2012, 51, 209.
namic acid (HCCA). Calcd for [M+Na]+ 1717.8 (monoisotopic),
found 1717.9.
1H NMR (300 MHz, DMF-d7) d 10.71 (s, 1H), 8.38–8.29 (m, 1H),
8.23–7.97 (m, 6H), 7.96–7.82 (m, 8H), 7.82–7.74 (m, 1H), 7.72–7.58
(m, 2H), 7.30 (s, 1H), 7.28–6.98 (m, 23H), 6.87 (s, 1H), 6.81–6.60
(m, 3H), 4.70 (d, J = 7.0 Hz, 1H), 4.57–4.31 (m, 4H), 4.29–4.06 (m,
6H), 4.06–3.95 (m, 1H), 3.88–3.52 (m, 17H), 3.23–2.96 (m, 5H),
2.91–2.78 (m, 4H), 2.24–2.07 (m, 5H), 2.07–1.68 (m, 9H), 1.28–
1.03 (m, 40H).
10. Boll, E.; Dheur, J.; Drobecq, H.; Melnyk, O. Org. Lett. 2012, 14, 2222.

R. Mhidia et al. / Bioorg. Med. Chem. 21 (2013) 3479–3485
3485
11. Hou, W.; Zhang, X.; Li, F.; Liu, C. F. Org. Lett. 2011, 13, 386.
12. Ollivier, N.; Dheur, J.; Mhidia, R.; Blanpain, A.; Melnyk, O. Org. Lett. 2010, 12,
5238.
35. Ingenito, R.; Bianchi, E.; Fattori, D.; Pessi, A. J. Am. Chem. Soc. 1999, 121, 11369.
36. Mhidia, R.; Beziere, N.; Blanpain, A.; Pommery, N.; Melnyk, O. Org. Lett. 2010,
12, 3982.
13. Saxon, E.; Armstrong, J. I.; Bertozzi, C. R. Org. Lett. 2000, 2, 2141.
14. Nilsson, B. L.; Kiessling, L. L.; Raines, R. T. Org. Lett. 1939, 2000, 2.
15. Bode, J. W.; Fox, R. M.; Baucom, K. D. Angew. Chem., Int. Ed. 2006, 45, 1248.
16. Sasaki, K.; Aubry, S.; Crich, D. Phosphorus, Sulfur, Silicon 2011, 186, 1005.
17. Crich, D.; Sharma, I. Angew. Chem., Int. Ed. 2009, 48, 2355.
18. Crich, D.; Sharma, I. Angew. Chem., Int. Ed. 2009, 48, 7591.
19. Tam, J. P.; Yu, Q.; Miao, Z. Biopolymers 1999, 51, 311.
20. Tam, J. P.; Lu, Y. A.; Liu, C. F.; Shao, J. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 12485.
21. Liu, C.-F.; Rao, C.; Tam, J. P. Tetrahedron Lett. 1996, 37, 933.
22. Crich, D.; Sasaki, K. Org. Lett. 2009, 11, 3514.
23. Assem, N.; Natarajan, A.; Yudin, A. K. J. Am. Chem. Soc. 2010, 132, 10986.
24. Kolakowski, R. V.; Shangguan, N.; Sauers, R. R.; Williams, L. J. J. Am. Chem. Soc.
2006, 128, 5695.
25. Shangguan, N.; Katukojvala, S.; Greenberg, R.; Williams, L. J. J. Am. Chem. Soc.
2003, 125, 7754.
26. Barlett, K. N.; Kolakowski, R. V.; Katukojvala, S.; Williams, L. J. Org. Lett. 2006, 8,
823.
27. Merkx, R.; Brouwer, A. J.; Rijkers, D. T.; Liskamp, R. M. Org. Lett. 2005, 7, 1125.
28. Rijkers, D. T.; Merkx, R.; Yim, C. B.; Brouwer, A. J.; Liskamp, R. M. J. Pept. Sci.
2010, 16, 1.
29. Merkx, R.; van Haren, M. J.; Rijkers, D. T.; Liskamp, R. M. J. Org. Chem. 2007, 72,
4574.
30. Zhang, X.; Li, F.; Lu, X. W.; Liu, C. F. Bioconjugate Chem. 2009, 20, 197.
31. Kenner, G. W.; McDermott, J. R.; Sheppard, R. C. J. Chem. Soc. D: Chem. Commun.
1971, 636.
37. Note that the nucleophilic attack of the N-alpha amino group on the other
imide carbonyl would proceed through a 4-membered ring intermediate and
thus is not favored.
38. Imide ligation proceeds efficiently at acidic pH, typically at pH 2.5 as shown in
Ref. 36. The reaction has been shown to be compatible with the presence of N-
alpha or N-epsilon amino groups, which are protected by protonation at pH
below 5, and therefore unable to react with thioacid, acylazide or imide groups.
39. Kahns, A. H.; Bundgaard, H. Int. J. Pharm. 1991, 71, 31.
40. Tsunoda, T.; Sasaki, O.; Takeuchi, O.; Itô, S. Tetrahedron 1991, 47, 3925.
41. Lee, P. H.; Ayyampalayam, S. N.; Carreira, L. A.; Shalaeva, M.; Bhattachar, S.;
Coselmon, R.; Poole, S.; Gifford, E.; Lombardo, F. Mol. Pharm. 2007, 4, 498.
42. Native peptide ligations are often carried out in aqueous buffers. While DMF
was used as solvent in this proof-of-concept study, imide ligation is known to
be compatible with the use of aqueous media as shown in Ref. 36. Native
peptide ligations are also often carried out in the presence of denaturants such
as guanidinium hydrochloride or urea. The effect of these additives on the
imide capture step remains to be established. However, they are poor
nucleophiles and thus are potentially compatible with the imide capture step
and the AsnNPL process.
43. Lutz, J.; Musiol, H. J.; Moroder, L. In Houben–Weyl, Methods of Organic Chemistry,
Synthesis of Peptides and Peptidomimetics; Goodman, M., Felix, A., Moroder, L.,
Toniolo, C., Eds.; Georg Thieme: Stuttgart, 2002; Vol. E22a, p 427.
44. Fang, G.-M.; Wang, J.-X.; Liu, L. Angew. Chem., Int. Ed. 2012, 124, 10493.
45. Fang, G.-M.; Li, Y.-M.; Shen, F.; Huang, Y.-C.; Li, J.-B.; Lin, Y.; Cui, H.-K.; Liu, L.
Angew. Chem., Int. Ed. 2011, 50, 7645.
32. Shin, Y.; Winans, K. A.; Backes, B. J.; Kent, S. B. H.; Ellman, J. A.; Bertozzi, C. R. J.
Am. Chem. Soc. 1999, 121, 11684.
33. Backes, B. J.; Ellman, J. A. J. Org. Chem. 1999, 64, 2322.
34. Backes, B. J.; Virgilio, A. A.; Ellman, J. A. J. Am. Chem. Soc. 1996, 118, 3055.
46. Raban, M.; Keintz, R.; Haritos, D. P.; Greenblatt, J. J. Org. Chem. 1980, 45, 2672.
47. Etter, M. C.; Reutzel, S. M. J. Am. Chem. Soc. 1991, 113, 2586.
48. Amelung, W.; Brodowski, S. Anal. Chem. 2002, 74, 3239.