Chemoselective peptide bond formation using formyl-substituted nitrophenylthio ester

Article abstract of DOI:10.1016/S0040-4039(03)00471-4

A novel method for peptide bond formation utilizing amino acid 2-formyl-4-nitrophenylthio ester has been developed. The reaction can be performed in water-containing media and is compatible with various types of amino acid side-chain functional groups. Use of N-methylmaleinimide as an additive is essential for the reaction to proceed with high efficiency. It captures liberated formyl-substituted thiophenol through 1,4-addition followed by aldol cyclization.

Full text of DOI:10.1016/S0040-4039(03)00471-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 3187–3190
Chemoselective peptide bond formation using formyl-substituted
nitrophenylthio ester
Akihiro Ishiwata, Tsuyoshi Ichiyanagi, Maki Takatani and Yukishige Ito*
RIKEN (The Institute of Physical and Chemical Research), 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan
Received 15 January 2003; revised 10 February 2003; accepted 14 February 2003
Abstract—A novel method for peptide bond formation utilizing amino acid 2-formyl-4-nitrophenylthio ester has been developed.
The reaction can be performed in water-containing media and is compatible with various types of amino acid side-chain functional
groups. Use of N-methylmaleinimide as an additive is essential for the reaction to proceed with high efficiency. It captures
liberated formyl-substituted thiophenol through 1,4-addition followed by aldol cyclization. © 2003 Elsevier Science Ltd. All rights
reserved.
Thioester-mediated peptide bond formation has
recently attracted attention and been shown to be par-
ticularly suitable for block condensation of large pep-
tide fragments. In particular, Aimoto’s pioneering work
demonstrated that Ag+-aided thioester activation is
highly efficient.1 Kent et al. also developed an ingenious
method called native chemical ligation, which offers
exceptional chemoselectivity.2 Kent’s approach allows
for the use of completely unprotected peptide for frag-
ment condensation, although application of the method
is restricted by the inherent limitation of the require-
ment of a cystein residue at the ligation site of the
C-terminal fragment. Taking this limitation into
account, further innovative approaches including
tion of hemiaminal (1), which should be facile for acyl
transfer to create peptide linkage. Although amide
bond formation through hemiaminal via NꢀO acyl
transfer has been reported by Kemp et al.,7 their
approach seems suitable only for reactive amine due to
the competing imine formation. However, since
thioester has higher reactivity toward aminolysis, acyl
transfer is expected to predominate over imine forma-
tion. A nitro group at the para position enhances the
activity of the leaving group of the thiol component.8
Initial concern was focused on the non-productive
interaction of liberated aldehyde (2) with amine, which
may retard peptide bond formation. This interaction
was fortuitously eliminated by the use of N-methyl-
maleimide as a scavenger (vide infra).
Sta¨udinger
ligation,3
phenylmercaptoethylamine-
assisted ligation,4 and mercaptobenzylamine-assisted
ligation5 have been reported, although the generality of
these recent proposals has yet to be scrutinized.
Fmoc-valine-derived 3 was used as the thioester,
which in turn was prepared as depicted in Scheme 2.
Specifically, commercially available 2-chloro-5-nitro-
benzaldehyde† was converted to thiolate, which was
In this paper, the authors report a novel entry to
thioester-based peptide bond formation, proceeding by
amine capture followed by acyl transfer (Scheme 1).
The technique is shown to be highly efficient, chemose-
lective, and compatible with aqueous media.
Specifically, a 2-formyl-4-nitrophenylthio ester (3) was
designed based on the following considerations. A for-
myl group placed at the ortho position can be expected
to capture amine,6 proceeding to the transient genera-
Scheme 1. Plausible pathway of the reaction.
* Corresponding author. Tel.: +81-48-467-9430; fax: +81-48-462-4680;
e-mail: yukito@postman.riken.go.jp
† Purchased from Tokyo Kasei Kogyo Co. (Yꢁ 6,750/25 g).
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00471-4

3188
A. Ishiwata et al. / Tetrahedron Letters 44 (2003) 3187–3190
then reacted with Fmoc-Val-Cl9 to give 3 in high
yield.‡
Reactions with glycine or valine ester were examined
initially, the results of which are summarized in Table
1. In the majority of cases, diisopropylethylamine
(DIPEA) was added to liberate free amine from the
corresponding hydrochloride or acetate. In the absence
of DIPEA, the reaction proceeded with impractically
slow kinetics (entry 1). Use of N-methylmaleimide
(NMM) was found to be essential for attaining a clean
result, and afforded much higher yield than without
(entry 3 and 7).§ The loss of 2-formyl-4-nitroben-
zenethiol was accounted for by the formation of tri-
cyclic product 5, which was isolated in high yield as a
4:1 mixture of diastereomers and dipeptide 4 (Scheme
3).¶ Thus, NMM appeared to be functionally driving
the reaction, not only by capturing thiol, but also by
Scheme 2. Synthesis of formyl-substituted nitrophenylthio
ester.
Table 1. Reaction of 3 under various conditionsa
‡ Preparation of Fmoc-valine 2-formyl-4-nitrophenyl-thioester 3: To a
solution of 2-chloro-4-nitrobenzaldehyde (2.38 g, 12.8 mmol) in
THF (45 mL) and DMF (4.5 mL) was added Na2S·9H2O (3.08 g,
12.8 mmol) in one portion at room temperature, and the reaction
mixture was stirred at room temperature for 12 h. After cooling
down to −78°C, a solution of Fmoc-Val-Cl (3.00 g, 8.55 mmol) in
THF (5 mL) was added to the mixture, the reaction mixture was
stirred for 3 h during which time the reaction temperature was
slowly going up to −50°C. Then the reaction was quenched by sat.
KHSO4 aq. and THF was removed in vacuo from the resulted
mixture. Product was extracted with ethyl acetate and combined
organic layer was washed with H2O, sat. NaHCO3 aq., H2O, and
brine, then dried over Na2SO4, and concentration in vacuo. The
product mixture was purified by flash column chromatography
using a gradient solvent system (hexane/ethyl acetate=10/1 to 5/1
to 2/1 to 1/1 to 1/2) to give title compound 3 (4.15 g, 96%): 1H
NMR (CDCl3, 400 MHz) l 0.93 (3H, d, J=6.8 Hz, Me2CH), 1.00
Entry H-AA-OR
Solvent
Time
3 days
2.5 h
2.5 h
Yield (%)
1b
2
AcOH·H-Gly-OtBu DMF–H2O
90
91
(9:1)
DMF–H2O
(9:1)
DMF–H2O
(9:1)
DMSO–H2O 2.5 h
(9:1)
DMF
3
HCl·H-Val-OtBu
100
84
4
5
2.5 h
2.5 h
2.5 h
83
99
32
(3H, d, J=6.8 Hz, Me2CH), 2.21–2.41 (1H, m, Me2CH
dd, J=6.8, 6.4 Hz, Fmoc), 4.44 (1H, dd, J=8.4, 4.4 Hz, Val-a-CH
4.51 (1H, dd, J=10.8, 6.8 Hz, Fmoc), 4.65 (1H, dd, J=10.8, 6.4
Hz, Fmoc), 5.26 (1H, d, J=8.4 Hz, NH), 7.24–7.41 (4H, m, Fmoc),
6
), 4.24 (1H,
6
CH2Cl2
DMF–H2O
(9:1)
7c
6
),
6
a DIPEA (1.1 equiv.) and NMM (1.5 equiv.).
b Without DIPEA.
7.59–7.61 (2H, m, Fmoc), 7.64 (1H, dd, J=10.8 Hz, Ar), 7.71–7.77
(2H, m, Fmoc), 8.39 (1H, dd, J=8.4, 2.4 Hz, Ar), 8.81 (1H, d,
J=2.4 Hz, Ar), 10.08 (1H, s, CHO); 13C NMR (CDCl3, 100 MHz)
l 17.30, 19.38, 30.75, 47.32, 66.51, 67.11, 119.98, 120.01, 123.78,
124.77, 124.86, 127.03, 127.48, 127.71, 127.75, 137.49, 137.88,
137.99, 141.30, 141.33, 143.38, 143.43, 148.49, 156.03, 187.97,
196.46; MALDI-TOF MS calcd for C27H24N2NaO6S ([M+Na]+)
527.1, found 527.1.
c Without NMM.
§ A typical experimental procedure for peptide bond formation:
Preparation of Fmoc-Val-Val-O-tBu 4a: To a solution of 3 (50.0
mg, 99.1 mmol), valine t-butyl ester hydrochloride (18.9 mg, 90.1
mmol) and N-methylmaleinimide (15.0 mg, 135 mmol) in 1.0 mL of
DMF–H2O (9:1) was added 1.0 M solution of diisopropylethyl-
amine (100 mL, 0.10 mmol) in dioxane at room temperature. The
mixture was stirred for 2.5 h at room temperature and concentrated
in vacuo. The product mixture was purified by flash column chro-
matography using a gradient solvent system (hexane/ethyl acetate=
20/1 to 10/1 to 5/1 to 2/1 to 1/1) to give title compound 4a (44.6
mg, 100%).
¶ Spectroscopic data of the tricyclic product 5 (major diastereomer):
1H NMR (CDCl3, 400 MHz) l 3.11 (3H, s, MeN), 3.87 (1H, d,
J=2.4 Hz, OH), 3.96 (1H, dd, J=9.6, 8.8 Hz, CH
6
CH(OH)), 4.31
OH),
(1H, d, J=9.6 Hz, SCH), 4.92 (1H, dd, J=8.8, 2.4 Hz, CH
6
6
7.59 (1H, d, J=8.4 Hz, Ar), 8.15 (1H, dd, J=8.4, 2.0 Hz, Ar), 8.56
(1H, d, J=2.0 Hz, Ar); 13C NMR (CDCl3, 100 MHz) l 25.69,
41.68, 49.79, 69.33, 120.70, 123.02, 129.10, 130.17, 140.08, 173.41,
176.91; FAB MS 295 ([M+H]+).
Scheme 3. Trapping of resulting formyl-substituted nitrothio-
phenol by NMM.

A. Ishiwata et al. / Tetrahedron Letters 44 (2003) 3187–3190
3189
blocking the non-productive interaction of aldehyde with
amine. When conducted in DMF, reactions were sub-
stantially more efficient in the presence of H2O
(entry 3 and 5), presumably by governing hemiami-
nallimine equilibrium in favor of the former. The use
of aq. DMSO (entry 4) was also successful, yet at
somewhat lower efficiency. CH2Cl2 afforded high yield
under anhydrous conditions (entry 6).
1
The reaction course was monitored by H NMR (400
MHz, DMF-d6–D2O, 9:1) to confirm the mechanistic
supposition, revealing the formation of transient species
that could be assigned to hemiaminal (l 6.2, s),ꢀꢀ which
was disappeared at the end of the reaction. Under
identical
conditions,
regioisomeric
2-nitro-4-
formylphenylthio ester**,10 gave corresponding peptide
in substantially reduced yield (74%). These results indi-
cated that the fomyl group at ortho-position is obviously
effective to give high yielding reaction, and that hemiami-
nal is a major reactive intermediate in the case of
2-formyl derivative 3. Energy minimization of the hemi-
aminal, as calculated by MOPAC-PM3, showed that its
hydroxyl group forms an intramolecular hydrogen bond
with the carbonyl oxygen of the thioester, and that a
nitrogen atom orients toward the carbonyl carbon to
form a favorable configuration for nucleophilic attack
(Fig. 1).
Figure 1. The energy-minimized structure of the hemi-aminal
intermediate as calculated by MOPAC-PM3. The dotted line
shows the intramolecular hydrogen bond between the hydroxyl
group of the hemi-aminal and the carbonyl oxygen of the
,
thioester (1.891 A). The bold arrow indicates the direction of
nucleophilic attack of a nitrogen atom of the hemi-aminal
toward the carbonyl carbon.
Peptide bond formation in aq. DMF was also examined
using other amino acid esters (Table 2), including side-
chain functionalized (entries 3 and 6), cyclic (entry 4) and
sterically hindered (entry 7) esters. In most cases, just 1.1
equivalent of thioester was sufficient to give products in
high yield.†† No racemization was detectable by HPLC
Table 2. Synthesis of various peptides
analysis of Fmoc-L-Val-L-Val-OtBu and Fmoc-D-Val-L-
Val-OtBu, both of which were formed in a mutually
exclusive manner (entries 1 and 2). Mechanistic consid-
erations, as well as the observed compatibility with
aqueous reaction environments, strongly suggested that
chemoselective amide formation might well be possible
using amino acids with unprotected hydroxy group. In
fact, the reaction of serine methyl ester (entry 8) gave
desired dipeptide in 82% yield and tyrosine t-butyl ester
(entry 9) gave Fmoc-Val-Tyr(OH)-OtBu in nearly quan-
titative yield.
Entry
Fmoc-Val-AA-OR
Yield (%)
1
2
3
4
5
6
7
8
9
Fmoc-Val-Val-OtBu (4a)
Fmoc-D-Val-Val-OtBu (4a%)
100
91
97
100
91
Fmoc-Val-Lys(Z)-OtBu (4b)
Fmoc-Val-Pro-OtBu (4c)
Fmoc-Val-Gly-OtBu (4d)
Fmoc-Val-Thr(OtBu)-OtBu (4e)
Fmoc-Val-Aib-OBn (4f)
95
72a
82
Fmoc-Val-Ser(OH)-OMe (4g)
Fmoc-Val-Tyr(OH)-OtBu (4h)
98
The suitability of this method for glycopeptide synthesis
was verified by reaction using a glycosylated asparagine
derivative with unprotected sugar moiety (Scheme 4).
Fmoc-Asn(N-acetylglucosamine)-OtBu (6)11 was sub-
jected to Fmoc deprotection, and the resultant amine was
reacted with thioester 3 (1.1 equiv.) in aq. DMF. The
desired acylated product (4i) was obtained in high yield.‡‡
a 2.2 equiv. of thioester.
ꢀꢀ The NMR experiment was performed by using thioester 3 and
glycine t-butyl ester HCl salt (1:1) in the presence of an equimolar
amount of DIPEA.
** Half-life of 3 and regioisomeric 2-nitro-4-formylphenylthio ester in
DMF-d6–pH 9.0 phosphate buffer (10:1) at ꢀ25°C were ca. 7 h and
over 2 days, respectively.
Scheme 4. Synthesis of glycopeptide 4i. Reagents and condi-
tions: (a) 20% piperidine–DMF; (b) Fmoc-Val-SAr (3) (1.1
equiv.), NMM (1.5 equiv.), DMF–H2O (9:1), 98% in two
steps.
†† The reaction of Fmoc-Pro-SAr with Val-OtBu also gave corre-
sponding peptide in 98% yield.
‡‡ Details of the synthesis of glycopeptide will be reported in a
separate paper.

3190
A. Ishiwata et al. / Tetrahedron Letters 44 (2003) 3187–3190
Table 3. Synthesis of peptides under buffered media
Entry
Time (h)
Buffer or base
Product
Yield (%)a
1
2
3
4
5
6
7
8
2.5
2.5
2.5
2.5
2.5
8
Phosphate (pH 8.0)
Phosphate (pH 6.8)
NaHCO3
Fmoc-Val-Val-OtBu (4a)
Fmoc-Val-Val-OtBu (4a)
Fmoc-Val-Val-OtBu (4a)
Fmoc-Val-Val-OtBu (4a)
Fmoc-Val-Gly-OtBu (4d)
Fmoc-Val-Aib-OtBu (4j)
Fmoc-Val-Asp-OtBu (4k)
Fmoc-Val-Gly-Gly-OH (4l)
Fmoc-Val-Gly-Gly-OH (4l)
Fmoc-Val-Trp-OtBu (4m)
Fmoc-Val-Arg-OEt (4n)
93 (100)
71
97
KOAc
56
Phosphate (pH 8.0)
Phosphate (pH 8.0)
Phosphate (pH 8.0)
Phosphate (pH 8.0)
NaHCO3
100 (93)
75 (61b)
92 (57)
77 (33)
53
8
2.5
2.5
6
9
10
11
Phosphate (pH 8.0)
Phosphate (pH 8.0)
89 (77)
81 (42)
6
a Yields in parenthesis obtained using DIPEA (1.1 equiv.).
b 2.2 equiv. of thioester.
Being aware of the limited stability of the thioester
linkage under basic conditions, buffered media were
examined in place of DIPEA in aq. DMF. As shown in
Table 3, phosphate buffer (pH 8.0) was highly
effective§§ and proved to be particularly suitable for
substrates having ionizable functionalities (entries 7, 8,
and 11) and Trp residue (entry 10). Also the reaction
with highly hindered amino acid (4j) was quite success-
ful (entry 6). A system containing 1 M NaHCO3 also
gave excellent results for Val-OtBu (entry 3), while
more basic conditions (entry 4) and slightly acidic
buffer (entry 2) were less satisfactory.
Kawakami, T.; Yoshimura, S.; Aimoto, S. Tetrahedron
Lett. 1998, 39, 7901–7904; (e) Mizuno, M.; Haneda, K.;
Iguchi, R.; Muramoto, I.; Kawakami, T.; Aimoto, S.;
Yamamoto, K.; Inazu, T. J. Am. Chem. Soc. 1999, 121,
284–290; (f) Kawakami, T.; Hasegawa, K.; Aimoto, S.
Bull. Chem. Soc. Jpn. 2000, 73, 197–203.
2. Recent review: Dawson, P. E.; Kent, S. B. H. Annu. Rev.
Biochem. 2000, 69, 923–960.
3. Nilsson, B. L.; Kiessling, L. L.; Raines, R. T. Org. Lett.
2000, 2, 1939–1941; 2001, 3, 9–12.
4. (a) Botti, P.; Carrasco, M. R.; Kent, S. B. H. Tetrahedron
Lett. 2001, 42, 1831–1833; (b) Low, D. W.; Hill, M. G.;
Carrasco, M. R.; Kent, S. B. H.; Botti, P. Proc. Natl.
Acad. Sci. USA 2001, 98, 6554–6559; (c) Marinzi, C.;
Bark, S. J.; Offer, J.; Dawson, P. E. Bioorg. Med. Chem.
2001, 9, 2323–2328.
5. (a) Offer, J.; Dawson, P. E. Org. Lett. 2000, 2, 23–26; (b)
Kawakami, T.; Akaji, K.; Aimoto, S. Org. Lett. 2001, 3,
1403–1405; (c) Offer, J.; Boddy, C. N. C.; Dawson, P. E.
J. Am. Chem. Soc. 2002, 124, 4642–4646.
In conclusion, a novel peptide formation reaction using
Fmoc-amino acid 2-formyl-5-nitrophenylthio ester has
been developed. The reaction was demonstrated to be
highly chemoselective12 and compatible with unpro-
tected carbohydrate as well as a negatively or positively
charged amino acid side chain.
6. Recent review: Koltart, D. M. Tetrahedron 2000, 56,
3449–3491.
7. Kemp, D. S.; Vellaccio, F., Jr. J. Org. Chem. 1975, 40,
3003–3004.
8. Farrington, J. A.; Hextall, P. J.; Kenner, G. W.; Turner,
J. M. J. Chem. Soc. 1957, 1407–1413.
9. Carpino, L. A.; Cohen, B. J.; Stephens, Jr., K. E.;
Sadat-Aalaee, S. Y.; Tien, J.-H.; Langridge, D. C. J. Org.
Chem. 1986, 51, 3734–3736. Preparation from corre-
Acknowledgements
Financial support from RIKEN President’s Fund (AI),
Mizutani Foundation for Glycoscience, and a Grant-in-
Aid for Scientific Research from the Japan Society for
the Promotion of Science (Grant No. 13480191) are
acknowledged. The authors thank Ms. Akemi Taka-
hashi for technical assistance.
sponding fluoride can be performed in
a similar
efficiency.
10. (a) Shalitin, Y.; Bernhard, S. A. J. Am. Chem. Soc. 1964,
86, 2292; (b) Bender, M. L.; Reinstein, J. A.; Silver, M.
S.; Mikulak, R. J. Am. Chem. Soc. 1965, 87, 4545–4553.
11. (a) Wen, S.; Guo, Z. Org. Lett. 2001, 3, 3773–3776; (b)
Ishiwata, A.; Takatani, M.; Nakahara, Y.; Ito, Y. Synlett
2002, 634–636.
12. Very recently, a peptide coupling of unprotected amino
acids through in situ p-nitrophenyl ester formation were
reported: Gagnon, P.; Huang, X.; Therrien, E.; Keillor, J.
W. Tetrahedron Lett. 2002, 43, 7717–7719.
References
1. (a) Hojo, H.; Aimoto, S. Bull. Chem. Soc. Jpn. 1991, 64,
111–117; (b) Hojo, H.; Kwon, Y.; Kakuta, Y.; Tsuda, S.;
Tanaka, I.; Hikichi, K.; Aimoto, S. Bull. Chem. Soc. Jpn.
1993, 66, 2700–2706; (c) Kawakami, T.; Kogure, S.;
Aimoto, S. Bull. Chem. Soc. Jpn. 1996, 69, 3331–3338; (d)
§§ Half-life of 3 in DMF–pH 8.0 phosphate buffer (10:1) was ca. 20 h
at ꢀ25°C.