In situ carboxyl activation using a silatropic switch: A new approach to amide and peptide constructions

Article abstract of DOI:10.1021/ja2065158

The novel reactivity of O-silylthionoesters with amine nucleophiles to generate oxoamides (rather than thioamides) is described. A straightforward first-generation trimethylsilylation protocol using bistrimethylsilylacetamide (BSA) combined with the unique reactivity of the O-silylthionoesters toward 1° and 2° amines to generate oxoamides provides the simplest means of activating a thiol acid for peptide bond formation at neutral pH. Excellent stereoretention is observed.

Full text of DOI:10.1021/ja2065158

COMMUNICATION
pubs.acs.org/JACS
In Situ Carboxyl Activation Using a Silatropic Switch: A New Approach
to Amide and Peptide Constructions
Wenting Wu, Zhihui Zhang,† and Lanny S. Liebeskind*
Sanford S. Atwood Chemistry Center, Emory University, 1515 Dickey Drive, Atlanta, Georgia 30322, United States
S Supporting Information
b
activation reaction run under an argon atmosphere (64%) and
one conducted open to air (70%).
ABSTRACT: The novel reactivity of O-silylthionoesters
with amine nucleophiles to generate oxoamides (rather than
thioamides) is described. A straightforward first-generation
trimethylsilylation protocol using bistrimethylsilylaceta-
mide (BSA) combined with the unique reactivity of the
O-silylthionoesters toward 1° and 2° amines to generate
oxoamides provides the simplest means of activating a thiol
acid for peptide bond formation at neutral pH. Excellent
stereoretention is observed.
The amide yields observed from the silylative activation
protocol open to air are a function of the reactivity of the O-
silylthionoester and are not caused by oxidation of the thiol acid
to its corresponding diacyldisulfide, which then reacts with the
amine.9 A variety of observations support this analysis: (1) a
CDCl3 solution of thiobenzoic acid (1.0 equiv), BSA (1.1 equiv),
and Et3N (1.3 equiv) exposed to air showed no evidence of the
formation of dibenzoyldisulfide after 2 days at room temperature,
and (2) the reaction rates were essentially identical for experi-
ments run under argon after freezeꢀthaw degassing or open to
air (1.0 equiv thiobenzoic acid, 1.3 equiv i-PrNH2, 1.0 equiv BSA
in CDCl3). In other control studies, PhCOSEt remained intact
after treatment with i-PrNH2 in THF for 24 h at room tempera-
ture, both in the presence and in the absence of BSA, reaction
conditions under which the O-silylthionoester is converted
completely to the amide. Furthermore, treating 0.5 equiv of
PhCOSEt and 0.5 equiv of 4-CH3C6H4COSH with 0.5 equiv
of BSA and 1 equiv of i-PrNH2 in THF at room temperature
provided the “unscrambled” reaction product (4-CH3C6H4CON-
Hi-Pr) in 88% isolated yield with recovery of PhCOSEt in 99%
yield, further supporting the mechanistic framework suggested
herein. Of additional mechanistic interest, an authentic sample of
bisbenzoyldisulfide did react rapidly with i-PrNH2, but the rate of
amide formation stalled at 50% conversion (see the Supporting
Information).
With this background in hand, a study of the BSA activation of
thiol acids for amide and peptide bond formation was under-
taken. Results for amide formation are gathered in Table 1. The
thiol acids were either commercially available or prepared from
the corresponding 9-fluorenylmethyl thiol esters via a standard
piperidine deprotection/acidification protocol10 and were used
without further purification. When the silylation of the thiol acid
was carried out at room temperature in the presence of a primary
or secondary amine, amide linkages were generated in very good
yields within a matter of hours. Both aromatic thiol acids (entries
1ꢀ5) and aliphatic thiol acids (entries 6ꢀ10) reacted effectively
with primary and secondary amines to produce secondary and
tertiary amides, respectively. The hydroxyl group was well-
tolerated under the reaction conditions (entry 4). Even aniline,
which has low nucleophilicity, reacted smoothly with N-Boc-Glu
thiol acid to produce the corresponding anilide (entry 7).
Sterically hindered thiol acids and amines also reacted to provide
the sterically congested amides in quite good yields (entries
The mild and chemoselective formation of amide bonds, in
particular from the context of racemization-free peptide ligations,
remains a focus of much modern research activity.1 In fact, the
development of new amide-forming reactions tops a recent
priority list of process improvements desired by the pharmaceu-
tical industry.2
We disclose herein a novel, pH-neutral method for in situ
activation of the carboxyl function that underpins a surprisingly
simple system for room-temperature amide and peptide bond
construction. This unique carboxyl activation relies upon the
known spontaneous formation of O-silylthionoesters from their
S-silylthiol ester isomers by a thermodynamically driven tauto-
merization of the triorganosilicon group from sulfur to oxygen
(Scheme 1).3 Surprisingly, the efficient reaction of O-silylthionoesters
with amines to generate amide rather than thioamide linkages is
unknown.4 The latter are formed exclusively when the analogous
O-alkylthionoesters are substrates (Scheme 2).5 Since thionoe-
sters are significantly more reactive toward nucleophiles than
thiol esters,5c the S-to-O silatropy and the attendant formation of
oxoamides serves as a novel and mild in situ activation of the
carboxyl function for nucleophilic addition.
Our first efforts focused on the stoichiometric preparation of
O-silylthionoesters through direct silylation of thiol acids.6 The
silylation of thiol acids occurs kinetically at sulfur to generate the
S-trimethylsilylthiol ester, but a rapid silatropic equilibration to
the more stable O-silylthionoester ensues.3 With this simple
silylative chemistry, the key reactivity principles underscoring the
amidic and peptidic coupling of O-silylthionoesters with amines
were defined and are described herein.
With thiobenzoic acid serving as a model, in situ silylation with
bistrimethylsilylacetamide (BSA) generated the O-trimethylsi-
lylthionoester within 4 min at room temperature (Scheme 3).7
Upon addition of i-PrNH2, a reaction ensued at room tempera-
ture to provide the corresponding amide in 70% yield after 3 h.
The free thiol acid itself is poorly reactive with i-PrNH2.8 Almost
identical isolated yields of the amide resulted from a silylative
Received: July 20, 2011
Published: August 16, 2011
r
2011 American Chemical Society
14256
dx.doi.org/10.1021/ja2065158 J. Am. Chem. Soc. 2011, 133, 14256–14259
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. In Situ Carboxyl Activation Using a Silicon Switch
were required at room temperature. No racemization was
observed for entries 7 and 10.
Peptide bond formation was also easily accomplished using
the silylative activation of N-protected R-amino thiol acids
(Table 2). The peptidic thiol acids were generated from the
corresponding 9-fluorenylmethyl thiol esters by the method of
Crich10 and used without further purification. As shown in
Table 2, Gly, Met, Phe, Glu, and Pro thiol acid residues reacted
easily to give the corresponding dipeptides (entries 1ꢀ9). It is
noteworthy that sterically hindered R-amino thiol acids such as
Val (entries 10 and 11) and even 2-aminoisobutyric thiol acid
(entry 12) were effectively coupled using this method, although
longer reaction times were required to achieve acceptable yields
at room temperature. The amino acids, Phe, Tyr, Val, Ala, Gly,
Met, Trp, and Pro were all equally effective as N-terminal
coupling partners (entries 1ꢀ15). The formation of Cbz-Gly-
L-Tyr-OMe indicates that phenolic residues do not interfere with
the coupling reaction (entry 2). Cbz-Gly-L-Phe-OMe, which was
formed as a single dipeptide in 78% yield from Cbz-Gly-SH and
L-Phe-OMe under the general coupling conditions (entry 1), was
generated in an almost identical isolated yield (71%) when an
equimolar amount of Cbz-L-Arg-OH was added to the reaction
mixture. This simple experiment confirms the compatibility of this
method with both carboxylic acid and guanidine functionalities.
Even prior to any extensive studies of the influence of
electronic and steric effects of the silicon reagent, epimerization
of sensitive stereocenters using the pH-neutral “silylative switch”
protocol for peptide coupling was found to be competitive with
or better than existing technologies. For example, the absence of
epimerization at both coupling partners was verified by HPLC
analysis for the dipeptides shown in entries 1 and 3ꢀ5 of Table 2
and for the tripeptide Boc-L-Phe-L-Pro-L-Ala-OEt in entry 13.
Furthermore, both the Anderson (entry 14) and Anteunis (entry 15)
tests were conducted to evaluate epimerization-prone linkages
during the peptide formation process.1a Of significance for future
systematic studies of the influence of the silylating agent,
PhSiH2Cl gave lower levels of epimerization than BSA in the
few cases preliminarily investigated. In the Anteunis test, this
method gave less than 5% epimerization, which is superior to
results using N,N0-dicyclohexylcarbodiimide (18.8% DL) and
PyBOP (6.6% DL) to facilitate the coupling between Z-Gly-L-
Scheme 2. Differential Reactivity of Thionoesters
Scheme 3. Exploratory Reactions Using Thiobenzoic Acid
Table 1. S-to-O Silylative Switch-Induced Amide Formation
Phe-OH and L-Val-OMe HCl, as reported in the literature.11
3
A comparison of the reaction of Boc-L-Glu(O-tBu)-SH and
Gly-OMe using a traditional peptide-coupling protocol with the
new silylative activation is both illustrative and compelling,
demonstrating the unique reactivity of the O-silylthionoester
approach to peptide construction. A 1:1.3 mixture of Boc-L-
Glu(O-tBu)-SH and Gly-OMe HCl was first exposed to Et3N to
3
liberate the amine and then to 1 equiv of BSA in THF at room
temperature to produce the peptide in 74% yield within 8 h. In
contrast, traditional activation of the thiol acid with PyBop and
diisopropylethylamine (DIEA) gave a mixture of the desired
peptide (40%) and the thioamide (28%), as depicted in Scheme 4.
The origin of the different reactivities of O-alkylthionoesters
and O-silylthionoesters toward amines is suggested in Scheme 5.
The mechanism of the known reaction of O-alkylthionoesters
with amines to give thioamides is straightforward.5c Attack by the
nucleophilic amine at the CdS bond of the thionoester generates
tetrahedral intermediate A, which can only collapse to create a
thioamide or revert back to starting materials. Without cleavage
of the strong RꢀO bond of A in Scheme 5, an oxoamide cannot
be formed from the tetrahedral intermediate generated from the
a General procedure: 1 equiv of thiol acid, 1.3 equiv of amine, and 1 equiv
of BSA were stirred at room temperature in THF. Isolated yields
are shown. b No racemization was observed.
8ꢀ10). Remarkably, doubly hindered amides were also obtained
in good yields (entries 9 and 10), although longer reaction times
14257
dx.doi.org/10.1021/ja2065158 |J. Am. Chem. Soc. 2011, 133, 14256–14259

Journal of the American Chemical Society
COMMUNICATION
Table 2. Activation of Thiol Acids for Peptide Synthesis via the “Silylative Switch”a
entry
1
amino thiol acid
Cbz-Gly-SH
amino ester
di- or tripeptide
time (h)
yield (%)b
epimerizationc
L/D = 100:0
L-Phe-OMe HCl
Cbz-Gly-L-Phe-OMe
10
10
10
10
12
10
8
78
71d
69
83
72
71
76
68
74
65
70
67
74
65e
80
60e
73f
74
50e
55f
3
2
3
Cbz-Gly-SH
L-Tyr-OMe HCl
Cbz-Gly-L-Tyr-OMe
3
Boc-L-Met-SH
L-Val-OEt HCl
Boc-L-Met-L-Val-OEt
LL/DL > 99:1
LL/DL > 99:1
LL/DL > 99:1
3
4
Boc-L-Phe-SH
L-Ala-OEt HCl
Boc-L-Phe-L-Ala-OEt
3
5
Boc-L-Glu(OtBu)-SH
Boc-L-Glu(OBn)-SH
Boc-L-Glu(OBn)-SH
Boc-L-Glu(OBn)-SH
Boc-L-Pro-SH
L-Val-OMe HCl
Boc-L-Glu(OtBu)-L-Val-OMe
Boc-L-Glu(OBn)-Gly-OEt
Boc-L-Glu(OBn)-L-Met-OMe
Boc-L-Glu(OBn)-L-Trp-OMe
Boc-L-Pro-L-Val-OMe
3
6
Gly-OEt HCl
3
7
L-Met-OMe HCl
10
8
3
8
L-Trp-OMe HCl
3
9
L-Val-OMe HCl
10
48
63
54
8
3
10
11
12
13
14
Boc-L-Val-SH
L-Pro-OMe HCl
Boc-L-Val-L-Pro-OMe
3
Boc-L-Val-SH
L-Phe-OMe HCl
Boc-L-Val-L-Phe-OMe
Boc-Aib-L-Trp-OMe
3
Boc-Aib-SH
L-Trp-OMe HCl
3
Boc-L-Phe-L-Pro-SH
Cbz-Gly-L-Phe-SH
L-Ala-OEt HCl
Boc-L-Phe-L-Pro-L-Ala-OEt
Cbz-Gly-L-Phe-Gly-OEt
LLL/LDL > 99:1
L/D > 96:4
3
Gly-OEt HCl
15
3
L/D > 96:4
L/D = 97:3
15
Cbz-Gly-L-Phe-SH
L-Val-OMe HCl
Cbz-Gly-L-Phe-L-Val-OMe
15
LL/DL > 95:5
LL/DL > 96:4
LL/DL = 97:3
3
a General procedure: A solution containing 1 equiv of the N-protected R-amino thiol acid (generated from the corresponding 9-fluorenylmethyl thiol
ester via piperidine deprotection/HCl acidification) and 1 equiv of BSA in THF was added to a solution containing 1.3 equiv of the amino acid
hydrochloride salt and 1.3 equiv of triethylamine or DIEA in THF. The mixture was then stirred at room temperature for 8ꢀ63 h. b Isolated yields.
c Epimerization ratios were determined by HPLC. d A solution containing 1 equiv of Cbz-Gly-SH and 1 equiv of BSA in THF was added to a solution
containing 1 equiv of Cbz-L-Arg-OH, 1.3 equiv of L-Phe-OMe HCl, and 1.3 equiv of triethylamine. The reaction mixture was stirred at room
3
temperature for 10 h. e To a CH3CN solution containing 1 equiv of dipeptidic thiol acid, 1.1 equiv of PhSiH2Cl, and 1.3 equiv of amino ester
hydrochloride salt was added 2.3 equiv of DIEA, and the mixture was stirred at room temperature for 8 or 15 h. f To a CH3CN solution containing 1 equiv
of dipeptidic thiol acid, 1.1 equiv of PhSiH2Cl, and 2 equiv of amino acid ester hydrochloride salt was added 3.0 equiv of DIEA, and the mixture was
stirred at room temperature for 15 h.
O-alkylthionoester. In contrast, however, the tetrahedral inter-
mediate formed by attack of an amine on an O-silylthionoester
can participate in a tautomeric migration of silicon from oxygen
to sulfur, possibly via the pentavalent Si intermediate B.
Scheme 4. Silylative versus Traditional Activation of a Thiol
Acid
In conclusion, O-silylthionoesters, generated in situ from thiol
acids via a straightforward and simple sequence of S-silylation
followed by S-to-O silatropy, have been shown to react with 1°
and 2° amines in a mild, ambient temperature construction of
amides and mono-, di-, and tripeptides. Both unhindered and
highly hindered substrates participated in room-temperature
amide and peptide bond formation, although longer reaction
times were required for the sterically encumbered entities.
Stereocenter epimerization was absent for most substrates and
very low for substrates known to be problematic in traditional
peptide-coupling protocols. Although a thorough study of the
influence of the nature of the silane on the reaction rate and
epimerization level of the amide/peptide coupling has yet to be
undertaken, a single example probing variation of the silane
structure suggested that unwanted epimerization can be posi-
tively influenced by the nature of the silylating agent. At present,
this straightforward first-generation trimethylsilylation proto-
col combined with the unique reactivity of O-silylthionoesters
toward 1° and 2° amines to generate oxoamides provides the
simplest means of activating a thiol acid for peptide bond forma-
tion at neutral pH.12 The tolerance of hydroxylic and phenolic
groups to the reaction conditions holds promise for the extension
of the silylative protocol to glycopeptide synthesis. This and
experiments to understand how the nature of the triorganosilyl
moiety influences the reactivity of the O-triorganosilylthionoester
toward various nucleophiles are planned. Finally, in its most
compelling conception, the formation of amides would take place
directly from carboxylic acids and amines via in situ-generated O-
silylthionoesters using only catalytic quantities of a thiosilylating
14258
dx.doi.org/10.1021/ja2065158 |J. Am. Chem. Soc. 2011, 133, 14256–14259

Journal of the American Chemical Society
COMMUNICATION
Scheme 5. S-to-O Silylative Switch Amidation Mechanism
agent. Experiments to achieve this forward-looking goal are in
progress.
footnote 24 within that manuscript, false coupling results obtained with
unpurified and unsparged thiol acid were assumed to be caused by
impurities and/or incomplete formation of the thionoester.
(5) (a) Glass, R. S. Sci. Synth. 2005, 22, 85. (b) Yeo, S. K.; Choi, B. G.;
Kim, J. D.; Lee, J. H. Bull. Korean Chem. Soc. 2002, 23, 1029. (c) Castro,
E. A. Chem. Rev. 1999, 99, 3505.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures, syn-
b
(6) Tautomerization of an S-silylthioester to an O-silylthionoester
functions as a novel in situ activation of the carboxyl function for reaction
with nucleophiles. Therefore, any synthetic method that can directly
generate an S-silylthioester under mild reaction conditions can serve as
an entry to the O-silylthionoester.
thesis and characterization of all new compounds, and scanned
spectra. This material is available free of charge via the Internet at
http://pubs.acs.org.
(7) Thiobenzoic acid: 1H NMR: δ 7.91 (Hortho, dd, J = 1.2, 8.4 Hz,
2H), 7.61 (Hpara, dt, J = 1.2, 8.0 Hz, 1H), 7.47 (Hmeta, t, J = 7.8 Hz, 2H),
4.58 (SꢀH, br s, 1H). IR: 2567(sh), 1660, 1594, 1580 cmꢀ1. O-
’ AUTHOR INFORMATION
Corresponding Author
Lanny.Liebeskind@emory.edu
1
Trimethylsilylthionoester (see ref 12h): H NMR: δ 8.21 (Hortho, d,
J = 8.4 Hz, 2H), 7.54 (Hpara, t, J = 7.2 Hz, 1H), 7.37 (Hmeta, t, J = 7.8 Hz,
2H), 0.5 (Me3Si, s, 9H). 13C NMR: δ 212.6 (SdCꢀO), 139.6, 133.0,
128.9, 128.1, 0.3.
Present Addresses
†Department of Chemistry, University of Chicago, 5735 South
Ellis Avenue, Chicago, IL 60637.
(8) From the recent focus on reagent-based methods to activate thiol
acids for reaction with amines to give amides (see refs 10 and 12 in this
manuscript), one might infer that thiol acids are unable to directly
acylate amines. Examples of amine acylation by thiol acids without
special activation have appeared in the literature for decades (see:
Cronyn, M. W.; Jiu, J. J. Am. Chem. Soc. 1952, 74, 4726. Sheehan,
J. C.; Johnson, D. A. J. Am. Chem. Soc. 1952, 74, 4726. Hawkins, P. J.;
Tarbell, D. S.; Noble, P. J. Am. Chem. Soc. 1953, 75, 4462. Hirabayashi,
Y.; Mizuta, M.; Kojima, M.; Horio, Y.; Ishihara, H. Bull. Chem. Soc. Jpn.
1971, 44, 791. ). However, some caution in interpreting the published
results is appropriate, since the direct reaction of thioacids with amines is
highly dependent on the purity of the thioacid.
’ ACKNOWLEDGMENT
This project originated with partial funding from the National
Institute of General Medical Sciences, DHHS, through Grant
GM066153. We acknowledge the exceptionally helpful advice of
Dr. Hao Li and Dr. Kasinath Sana and the critical insights and
thoughtful control experiments offered by Dr. Angus Lamar.
’ REFERENCES
(9) Liu, R.; Orgel, L. E. Nature 1997, 389, 52.
(1) (a) Valeur, E.; Bradley, M. Chem. Soc. Rev. 2009, 38, 606. (b)
Kent, S. B. H. Chem. Soc. Rev. 2009, 38, 338. (c) Al-Zoubi, R. M.; Marion,
O.; Hall, D. G. Angew. Chem., Int. Ed. 2008, 47, 2876. (d) Maki, T.;
Ishihara, K.; Yamamoto, H. Tetrahedron 2007, 63, 8645. (e) Bode, J. W.
Curr. Opin. Drug Discovery Dev. 2006, 9, 765. (f) Tam, J. P.; Xu, J.; Eom,
K. D. Biopolymers 2001, 60, 194.
(2) Constable, D. J. C.; Dunn, P. J.; Hayler, J. D.; Humphrey, G. R.;
Leazer, J. L.; Linderman, R. J.; Lorenz, K.; Manley, J.; Pearlman, B. A.;
Wells, A.; Zaksh, A.; Zhang, T. Y. Green Chem. 2007, 9, 411.
(3) (a) Martel, B.; Duffaut, N. C. R. Acad. Sci. Paris, Ser. IIc 2: Sci.
Chim. 1966, 263, 74. (b) Kricheldrof, H. R.; Leppert, E. Synthesis
1971, 435. (c) Kricheldorf, H. R.; Leppert, E. Makromol. Chem. 1972,
158, 223. (d) Kato, S.; Akada, W.; Mizuta, M.; Ishii, Y. Bull. Chem. Soc.
Jpn. 1973, 46, 244. (e) Kricheldorf, H. R.; Leppert, E. Makromol. Chem.
1973, 167, 47. (f) Ishii, A.; Nakayama, J. In Comprehensive Organic
Functional Group Transformations II; Katritzky, A. R., Taylor, R. J. K.,
Eds.; Elsevier: Amsterdam, 2005; Vol. 5, pp 459ꢀ491. (g) Kato, S.;
Niyomura, O. Top. Curr. Chem. 2005, 251, 13. (h) Wojnowski, W.;
Wojnowska, M.; Grubba, R.; Konitz, A.; Pikies, J. Z. Anorg. Allg. Chem.
2008, 634, 730.
(10) Crich, D.; Sana, K.; Guo, S. Org. Lett. 2007, 9, 4423.
(11) Chen, S.; Xu, J. Tetrahedron Lett. 1992, 33, 647.
(12) In the search for effective new peptide ligations, thiol acids and
their derivatives have attracted much interest and recently resulted in
impressive synthetic advances, most notably from the following labora-
tories: Crich: (a) Sasaki, K.; Crich, D. Org. Lett. 2010, 12, 3254. (b)
Crich, D.; Sharma, I. Angew. Chem., Int. Ed. 2009, 48, 2355. (c) Crich, D.;
Sasaki, K. Org. Lett. 2009, 11, 3514. (d) Reference 10. Danishefsky:(e)
Wu, X.; Stockdill, J. L.; Wang, P.; Danishefsky, S. J. J. Am. Chem. Soc.
2010, 132, 4098. (f) Wang, P.; Danishefsky, S. J. J. Am. Chem. Soc. 2010,
132, 17045. (g) Rao, Y.; Li, X.; Danishefsky, S. J. J. Am. Chem. Soc. 2009,
131, 12924. Williams:(h) Kolakowski, R. V.; Shangguan, N.; Sauers,
R. R.; Williams, L. J. J. Am. Chem. Soc. 2006, 128, 5695. (i) Barlett, K. N.;
Kolakowski, R. V.; Katukojvala, S.; Williams, L. J. Org. Lett. 2006, 8, 823.
(j) Shangguan, N.; Katukojvala, S.; Greenberg, R.; Williams, L. J. J. Am.
Chem. Soc. 2003, 125, 7754. Others:(k) Pan, J.; Devarie-Baez, N. O.;
Xian, M. Org. Lett. 2011, 13, 1092. (l) Zhang, X.; Li, F.; Lu, X.-W.; Liu,
C.-F. Bioconjugate Chem. 2009, 20, 197. (m) Hackenberger, C. P. R.;
Schwarzer, D. Angew. Chem., Int. Ed. 2008, 47, 10030.
(4) As a control reaction to assess the mechanism of the reaction of thiol
acids with azides to generate amides, Williams and co-workers investigated
the reaction of O-trimethylsilylthionobenzoate with benzyl azide, which
generated only trace amounts of corresponding amide (see ref 12h). In
14259
dx.doi.org/10.1021/ja2065158 |J. Am. Chem. Soc. 2011, 133, 14256–14259