A traceless approach to amide and peptide construction from thioacids and dithiocarbamate-terminal amines

Article abstract of DOI:10.1039/c2sc21317f

A novel and traceless strategy has been devised that allows a coupling of thioacids and dithiocarbamate-terminal amines. This strategy had been assumed to be dependent on the attachment of a functional equivalent of a cysteine side chain in earlier native chemical ligation approaches. This approach enables the traceless removal of CS₂ to directly generate the desired amide bond and is compatible with a range of unprotected side chains of amino acid. The ability to produce amide or peptides by a traceless removal of the auxiliary is a significant virtue of the method. Meanwhile, the application of this new peptide-bond-forming reaction to the synthesis of novel endomorphin (EM) derivatives with various binding potencies was realized.

Full text of DOI:10.1039/c2sc21317f

Chemical Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
A traceless approach to amide and peptide construction
from thioacids and dithiocarbamate-terminal amines†
Cite this: Chem. Sci., 2013, 4, 970
Wenteng Chen,‡a Jiaan Shao,‡a Miao Hu,a Wanwan Yu,a Marc A. Giulianotti,b
b
ab
*
*
Richard A. Houghten and Yongping Yu
A novel and traceless strategy has been devised that allows a coupling of thioacids and dithiocarbamate-
terminal amines. This strategy had been assumed to be dependent on the attachment of a functional
equivalent of a cysteine side chain in earlier native chemical ligation approaches. This approach enables
the traceless removal of CS2 to directly generate the desired amide bond and is compatible with a range
of unprotected side chains of amino acid. The ability to produce amide or peptides by a traceless
removal of the auxiliary is a significant virtue of the method. Meanwhile, the application of this new
peptide-bond-forming reaction to the synthesis of novel endomorphin (EM) derivatives with various
binding potencies was realized.
Received 21st August 2012
Accepted 21st November 2012
DOI: 10.1039/c2sc21317f
www.rsc.org/chemicalscience
implementation of dithiocarbamates. This design is shown in
Fig. 1. The dithiocarbamate-terminal amine or peptide (1-B) was
Introduction
The amide bond is one of the most prevalent structural
elements which appears in medicine and materials, especially
proteins, conjugates and small molecule pharmaceuticals.1 In
recent years, many powerful methods of amide synthesis have
appeared in the literature, such as conventional amide
synthesis via coupling reagents,2 Staudinger ligation,3 the ami-
dation of thioacids with azides,4 hydrative amide synthesis
through alkyne–azide coupling,5 oxidative amidation of alco-
hols,6 aldehydes,7 or alkynes,8 ketoacid–hydroxylamine liga-
tion,9 amide synthesis by isocyanate/isothiocyanate–thioacid
coupling,10 activated thioacid-mediated coupling,11 amide
formation by isonitrile intermediates12a,b and acyl disulde-
mediated ligation.12c One of the very best methods to demon-
strate efficacy is peptide ligation,13 that is polypeptide synthesis
by segment coupling in buffered aqueous solutions of neutral
pH without epimerization. While considerable progress has
been made in this area over the past years, there is still certainly
room for improvement.
obtained by reacting the desired amine or peptide (1-A) with
CS2. Thereaer, thioacid (1-C) reacts with dithiocarbamate (1-B)
to afford an amide bond (1-D) by a traceless loss of a simple,
volatile byproduct, carbon disulde.10a The central feature of
this strategy represents a novel, successful and traceless
approach for amide and peptide synthesis. Additionally, we
believe that this is the rst example of the incorporation of the
dithiocarbamate template into amide formation.
Results and discussion
This design was guided by the knowledge that thioacids are
poorly reactive with amines to form amides.14 We found that
treatment of benzylamine in MeOH followed by addition of
thiobenzoic acid gave the amide product (<10%) aer 24 h in N2
atmosphere at room temperature. Our study of the amide bond-
forming reaction began with the model reaction of the thio-
benzoic acid and potassium of N-benzyl dithiocarbamate15 in
MeOH. We are pleased to nd that the conversion (>95%) and
isolated yield (90%) were improved substantially aer 8 h. This
observation described above led to the hypothesis that dithio-
carbamate might be the key one in the desired transformation.
A standard experimental protocol was developed in order to
ascertain an initial scope for the reaction. These conditions
included thioacid (2.0 equiv.) and dithiocarbamate salt
(1.0 equiv.) in MeOH, operating at room temperature for a
standard time. Encouraged by the initial result, we set out to
explore the thioacid–dithiocarbamate coupling in greater
details (Table 1). The thioacids were either commercially avail-
able or prepared using sodium sulde and carbonyldiimidazole
(CDI) and used without further purication.16 Amines bearing
In this article, we describe a traceless and convenient route
to the preparation of amides and peptides. This amide
synthesis uses thioacids as the acyl donor, but explores a new
chemical template for the amine feedstock through the
aInstitute of Materia Medica, College of Pharmaceutical Sciences, Zhejiang University,
Hangzhou 310058, P. R. China. E-mail: yyu@zju.edu.cn; Fax: +86-571-88208452; Tel:
+86-571-88208452
bTorrey Pines Institute for Molecular Studies, 11350 Village Parkway, Port St. Lucie, FL
34987, USA. E-mail: houghten@tpims.org; Fax: +1-772-345-3649; Tel: +1-772-345-
4590
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2sc21317f
‡ These authors contributed equally.
970 | Chem. Sci., 2013, 4, 970–976
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
Fig. 1 Amide and peptide synthesis by thioacid–dithiocarbamate coupling.
Table 1 Substrate scope of amide formationa
a
The reaction conditions: dithiocarbamate salt (1.0 equiv.), thioacid (2.0 equiv.) in MeOH for 8–17 h at room temperature.
aliphatic and aromatic substituents performed well, delivering possible. (Table 1, 1i) The ability to produce such products
the desired amide in good to excellent yield (Table 1). Amines under mild conditions is a strong virtue of the method.
with increasing steric hindrance provided comparable levels of
Attempts to extend this methodology to the construction of
efficiency (Table 1, 1b and 1c). Piperidine and N-methyl ben- peptides were not initially encouraging (Fig. 2). It was
zylamine were evaluated as representative secondary amines, as found that dithiocarbamate-terminal peptide (A) formation
these are oen more difficult substrates for dehydrative amide underwent conversion to their corresponding 2-thio-
synthesis. This protocol provided the desired tertiary amide 1f xoimidazolidin-4-ones (C) when KOH was utilized as the base
(56%), 1j (58%) and 1k (65%). 1g, 1h and 1i of Table 1 estab- during the preparation process. The corresponding 2-thio-
lished the viability of the method and nicely illustrated the xoimidazolidin-4-ones (C) cannot undergo a coupling reaction
potential for application to a hindered thioacid. The formation with the peptide thioacids. Accordingly, we investigated the
of anilides 1e and 1h, notwithstanding the relatively weak optimal conditions for the formation of dithiocarbamates
reactivity of aniline nucleophiles, was encouraging. Amide (Table 2). An array of bases was assessed for their ability to
formation using phenylalanine ethyl ester provided a promising prevent this side reaction. Importantly, pyridine was found to
indication that applications in peptide synthesis might be signicantly favor the reaction and was superior to other bases,
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci., 2013, 4, 970–976 | 971

View Article Online
Chemical Science
Edge Article
Fig. 2 Challenge: thioacid–dithiocarbamate coupling in peptide synthesis.
Table 2 Screening the optimal condition for the formation of dithiocarbamatesa
Entry
Base
Bb%
Cb%
Db%
1c
2
3
KOH
Et3N
Imidazole
Pyridine
10
10
40
88
>85
>85
20
Trace
Trace
40
4
6
6
a
A (1.0 mmol), 0.1 M CS2 and 0.1 M base were stirred in DCM (2.0 ml) in ice-bath. b The conversion of B, C and D was monitored by HPLC. c Entry 1
was carried out in MeOH–H2O (10 : 1).
including KOH, Et3N and imidazole. Our premise that the base L-Phe–L-Phe–COOMe derived dithiocarbamate. The acetamide
would be an important factor was validated by the yield of the was obtained in excellent yield (88%, Table 3, 3d). A series of
dithiocarbamates. The data (in Table 2) indicates that pyridine substrates were then assessed in DCM (Table 3). As shown in
favors the formation of dithiocarbamates with limited side Table 3, C-terminal Phe, Aib, Gln and Val residue derived thi-
reactions.
oacids were readily accommodated (Table 3, 3d–3p). Impor-
With these conditions in hand, a more detailed study of the tantly, even sterically hindered peptide-derived thioacids such
peptide synthesis was undertaken. A series of thioesters was as Val and Aib could be coupled effectively in good to excellent
prepared by the coupling of suitably protected amino acids with yields under these conditions. Moreover, coupling of a range of
9-uorenylmethanethiol,17 or triphenylmethanethiol.10a The hindered amino acid derived dithiocarbamates, including Phe,
thioacids were released with piperidine or triuoroacetic acid Trp, Val, Ile, Tyr and His, were efficiently prepared (Table 3, 3a–
and triethylsilane, depending on which thioesters were 3p). The experimental results in Table 3 conrm the compati-
employed. The corresponding dithiocarbamates were formed in bility of this methodology with functionalities, such as
situ from the amino ester in the presence of pyridine and CS2.18 phenolic, indole, and imidazole (Table 3, 3g, 3h, 3l and 3n). The
The rst coupling was investigated by using thioacetic acid with question of N-terminal epimerization during the course of
972 | Chem. Sci., 2013, 4, 970–976
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci., 2013, 4, 970–976 | 973

View Article Online
Chemical Science
Edge Article
Table 4 Substrate scope of peptide synthesis in aqueous mediaa
a
General procedure: 1.0 equiv. of amino acid, 0.1 M CS2 and 0.1 M pyridine were stirred at room temperature in NMP:1 M HEPES buffer (V/V ¼ 4 : 1,
pH ¼ 7.0) and then 2.0 equiv. of thioacid was added. The reaction mixture was stirred for 48 h.
coupling was also addressed here. 3a and 3b, 3e and 3f of peptides with a carboxylic acid at the C-terminal (Table 4).18
A
Table 3 veried the absence of epimerization.19 We next series of peptides were smoothly prepared in a mixed organic
addressed the question of C-terminal epimerization during the and buffer system with acceptable yields (Table 4, 4a–4j). As the
course of ligation. The preparation of the L,D and D,D isomer of same in Table 3, the inuence of functionalities and steric
Boc–Val–Phe–COOMe (3o and 3p in Table 3) veried the effects are shown. Similarly, the hindered thioacids (Table 4, 4b,
absence of racemization, as also determined by LC-MS.19 4c, 4e–4g) coupled with dithiocarbamate-terminal peptides
Furthermore, this methodology was also applicable for the under mild conditions. Meanwhile, we also demonstrated the
coupling of short peptide fragments (Table 3, 3k, 3l and 3m). application of this approach to peptide fragment synthesis,
We further explored the efficiency of our reaction in the pres- such as tripeptides (Table 4, 4f and 4i), tetrapeptides (Table 4,
ence of an internal glutamic thioacid (Table 3, 3q) as well as an 4j), hexapeptides (Table 4, 4b, 4e and 4g), and heptapeptides
unprotected glutamine-derived thioacid (Table 3, 3n and 3o). As (Table 4, 4c and 4d). In fact, some of the fragments are termi-
expected, the reaction proceeded smoothly to give the desired nated by the incorporation of a further acid, which enable the
products in moderate yields.
“le to right” peptide synthesis strategy.
To further extend the scope of this reaction, we have also
Finally, we applied this new peptide-bond-forming reaction
applied it in buffered aqueous media for the synthesis of to the synthesis of a novel endomorphin (EM) derivative (5a)
974 | Chem. Sci., 2013, 4, 970–976
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
Fig. 3 Synthesis of endomorphin derivative 5a.
Fig. 4 Proposed mechanism.
(Fig. 3). This tetrapeptide, containing an a-hydroxyl-b-phenyl- protected.21 This methodology holds considerable promise for a
alanine in the sequence, with the binding affinities for the m supplement to current methods for peptide synthesis.
and k opioid receptor are 10.37 nM and 2087 nM, respectively
(unpublished work). The unprotected residues of the
both coupling fragments can be tolerated to this reaction
Acknowledgements
condition.
Support for this research was provided by National Science
The proposed mechanism of the reaction is shown in Fig. 4. Foundation of China (no. 81072516 to Y. Yu) and Natural
The amine (I) rst reacts with CS2 to form the corresponding Science Foundation of Zhejiang Province (no. Z2110655 to
dithiocarbamate (II). Then the dithiocarbamate (II) is reacted Y. Yu) and provided in part by the National Institute of Drug
with thioacid (III) to give the intermediate (IV).20 Next, this Abuse USA (1R01DA031370-01 to R. A. Houghten). We appre-
intermediate undergoes a spontaneous intramolecular rear- ciate Greg Welmaker (Torrey Pines Institute for Molecular
rangement (1,3-S / N-acyl transfer) with loss of CS2 to form a Studies, USA) for critical reading of the manuscript.
desired amide bond. The target product (V) is obtained in the
desired nal form without further manipulation.
References:
In summary, we have demonstrated a novel and straight-
forward amide and peptide synthesis approach. The success of
this strategy had been assumed to be dependent on the
attachment of a functional equivalent of a cysteine side chain in
earlier native chemical ligation approaches. The ability to
produce amides or peptides by a traceless removal of the
auxiliary is a signicant virtue of our described method. Simple
and complex amide products can be obtained in moderate to
excellent yields with this methodology. Furthermore, this
method is tolerant to a range of naturally occurring amino
acid side chains, except lysine and cysteine, which must be
1 (a) A. Graul and J. Castaner, Drugs Future, 1997, 22, 956; (b)
A. A. Patchett, J. Med. Chem., 1993, 36, 2051; (c) M. de
Gasparo and S. Whitebread, Regul. Pept., 1995, 59, 303; (d)
V. S. Ananthanarayanan, S. Tetreault and A. Saint-Jean,
J. Med. Chem., 1993, 36, 1324.
2 S.-Y. Han and Y.-A. Kim, Tetrahedron, 2004, 60, 2447.
3 (a) B. L. Nilsson, L. L. Kiessling and R. T. Raines, Org. Lett.,
2000, 2, 1939; (b) E. Saxon, J. I. Armstrong and
C. R. Bertozzi, Org. Lett., 2000, 2, 2141; (c) B. L. Nilsson,
L. L. Kiesling and R. T. Raines, Org. Lett., 2001, 3, 9; (d)
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci., 2013, 4, 970–976 | 975

View Article Online
Chemical Science
Edge Article
¨
M. Kohn and R. Breinbauer, Angew. Chem., Int. Ed., 2004, 43, 13 P. E. Dawson, T. W. Muir, I. Clarklewis and S. B. H. Kent,
3106; (e) L. Liu, Z.-Y. Hong and C.-H. Wong, ChemBioChem, Science, 1994, 266, 776.
2006, 7, 429; (f) R. Merkx, D. T. S. Rijkers, J. Kemmink and 14 (a) R. Liu and L. E. Orgel, Nature, 1997, 389, 52; (b) The direct
R. M. J. Liskamp, Tetrahedron Lett., 2003, 44, 4515; (g)
C. Grandjean, A. Boutonnier, C. Guerreiro, J.-M. Fournier
and L. A. Mulard, J. Org. Chem., 2005, 70, 7123.
reaction of thioacid with amines is highly dependent on the
purity of the thioacids, see ref. 11a; (c) S. M. Mali, S. V. Jadhav
and H. N. Gopi, Chem. Commun., 2012, 48, 7085–7087
4 (a) N. Shangguan, S. Katukojvala, R. Greenberg and 15 (a) A. K. Mishra, S. B. Mishra, N. Manav, D. Saluja,
L. J. Williams, J. Am. Chem. Soc., 2003, 125, 7754; (b)
R. V. Kolakowski, N. Shangguan, R. R. Sauers and
L. J. Williams, J. Am. Chem. Soc., 2006, 128, 5695; (c)
X. Zhang, F. Li, X.-W. Lu and C.-F. Liu, Bioconjugate Chem.,
2009, 20, 197; (d) K. N. Barlett, R. V. Kolakowski,
S. Katukojvala and L. J. Williams, Org. Lett., 2006, 8, 823.
R. Chandrac and N. K. Kaushika, Bioorg. Med. Chem., 2006,
14, 6333; (b) J. Zhang, X. Lin, J. Ren, J. Liu and X. Wang,
Appl. Radiat. Isot., 2010, 68, 101. Potassium of N-benzyl
1
dithiocarbamate: H NMR (500 MHz, DMSO) d 7.94 (s, 1H),
7.32–7.25 (m, 5H), 4.69 (s, 2H); 13C NMR (125 MHz, DMSO)
d 215.2, 140.7, 128.3, 127.9, 126.7, 58.2.
5 (a) S. H. Cho, E. J. Yoo, L. Bae and S. Chang, J. Am. Chem. Soc., 16 N. Assem, A. Natarajan and A. K. Yudin, J. Am. Chem. Soc.,
2005, 127, 16046; (b) M. P. Cassidy, J. Raushel and
V. V. Fokin, Angew. Chem., Int. Ed., 2006, 45, 3154.
2010, 132, 10986.
17 D. Crich, K. Sana and S. Guo, Org. Lett., 2007, 9, 4423.
6 (a) C. Gunanathan, Y. Ben-David and D. Milstein, Science, 18 The experimental result showed that the dithiocarbamate-
2007, 317, 790; (b) L. U. Nordstrom, H. Vogt and
R. Madsen, J. Am. Chem. Soc., 2008, 130, 17672.
7 (a) W. J. Yoo and C. J. Li, J. Am. Chem. Soc., 2006, 128, 13064;
(b) J. Gao and G. W. Wang, J. Org. Chem., 2008, 73, 2955.
8 W. K. Chan, C. M. Ho, M. K. Wong and C. M. Che, J. Am.
Chem. Soc., 2006, 128, 14796.
9 J. W. Bode, R. M. Fox and K. D. Baucom, Angew. Chem., Int.
Ed., 2006, 45, 1248.
10 (a) D. Crich and K. Sasaki, Org. Lett., 2009, 11, 3514; (b)
K. Sasaki and D. Crich, Org. Lett., 2011, 13, 2256.
11 (a) W. Wu, Z. Zhang and L. S. Liebeskind, J. Am. Chem. Soc.,
2011, 133, 14256; (b) J. Pan, N. O. Devarie-Baez and M. Xian,
terminal peptide is not stable. It was found that
dithiocarbamate could decompose slowly to the free amine
and form the corresponding thiourea (detailed
experimental results was showed in ESI†). Based on the
above results, we chose to prepare the dithiocarbamate-
terminal peptide in situ in the presence of excess of CS2
without further purication. The dithiocarbamate-terminal
peptide formed in situ was reasonably stable toward
decomposition under the reaction conditions.
19 NMR spectra and chiral HPLC spectra of 3a and 3b of Table 3
are listed in ESI.† LC-MS spectra of 3e and 3f, 3o and 3p in
Table 3 are listed in ESI†.
Org. Lett., 2011, 13, 1092; (c) P. Wang and S. J. Danishefsky, 20 Caution: H2S is highly toxic and the reaction must be
J. Am. Chem. Soc., 2010, 132, 17045. performed in a well-ventilated fume hood.
12 (a) Y. Yuan, J. Zhu, X. Li, X. Wu and S. J. Danishefsky, 21 From consideration of the reactivity of amine
a
Tetrahedron Lett., 2009, 50, 2329; (b) Y. Rao, X. Li and
S. J. Danishefsky, J. Am. Chem. Soc., 2009, 131, 12924; (c)
C. Liu, C. Rao and J. P. Tam, Tetrahedron Lett., 1996, 37, 933.
nucleophiles toward CS2, the lysine and cysteine residues
in a sequence must be protected when the ligation
reaction is performed.
976 | Chem. Sci., 2013, 4, 970–976
This journal is ª The Royal Society of Chemistry 2013