Efficient Site-Specific Incorporation of Thioamides into Peptides on a Solid Support

Article abstract of DOI:10.1021/acs.orglett.5b01484

Designing bioactive peptides containing "thioamide" functionality to modulate their pharmacological properties has been thwarted so far because of various synthetic challenges. The fast, efficient, and inexpensive synthesis and incorporation of a wide range of thionated amino acids into a growing peptide chain on a solid support is reported using standard Fmoc-based chemistry. The commonly employed methodology is comprehensively investigated and optimized with significant improvements regarding the quantity of reagents and reaction conditions. The utility of the protocol is further demonstrated in the synthesis of dithionated linear and monothionated cyclic peptides, which has been a daunting task.

Full text of DOI:10.1021/acs.orglett.5b01484

Letter
pubs.acs.org/OrgLett
Efficient Site-Specific Incorporation of Thioamides into Peptides on a
Solid Support
Somnath Mukherjee, Hitesh Verma, and Jayanta Chatterjee*
Molecular Biophysics Unit, Indian Institute of Science, Bangalore 560012, India
S
* Supporting Information
ABSTRACT: Designing bioactive peptides containing “thioamide” functionality to modulate their pharmacological properties
has been thwarted so far because of various synthetic challenges. The fast, efficient, and inexpensive synthesis and incorporation
of a wide range of thionated amino acids into a growing peptide chain on a solid support is reported using standard Fmoc-based
chemistry. The commonly employed methodology is comprehensively investigated and optimized with significant improvements
regarding the quantity of reagents and reaction conditions. The utility of the protocol is further demonstrated in the synthesis of
dithionated linear and monothionated cyclic peptides, which has been a daunting task.
eplacement of an amide bond in a peptide by structurally
analogous functionalities has opened up a whole new area of
such as thiobenzotriazolides,8 thioacylfluorobenzimidazolinone,6
and the azirine/oxazolone method.9 Thionating agents are often
required in excess under long refluxing conditions in anhydrous
solvents.7,10 Such treatment usually furnishes an inseparable
mixture of products, which restricts their usage for the site-
specific installation of the thioamide unit. Their utility is mostly
limited to peptides with hydrophobic side chains.7 Usually,
thioacylating reagents are superior to thionating reagents in
terms of achieving site-specific incorporation of the thioamide
moiety. However, use of thioacylating reagents can suffer from
oxo-amide formation and their epimerization during formation
under basic conditions.11
R
peptide bond surrogates. Such replacement of a peptide bond
consequently alters the spectroscopic, physicochemical, struc-
tural, and biological properties of the native peptide, giving rise to
enhanced biological activities and intriguing structural features.
Of the various peptide bond surrogates known, a “thioamide”,
where the peptide bond “O” atom is replaced with “S” (Figure 1),
Following the work of Rapoport et al.,8 we sought to optimize
the reaction conditions used to incorporate thioamide moieties
into a growing peptide on a solid support. Herein, we present the
systematic and comprehensive optimization study of such site-
specific thioamide incorporation using Fmoc chemistry. We have
made significant improvements over the conventional method-
ology (Table 2) regarding the amount of reagents used, time
required, conversion, purity, substrate scope, etc. Additionally,
we have elaborated our optimized protocol to incorporate two
consecutive thioamide linkages into a growing peptide chain.
To obtain the thioamide precursors of the commercially
available orthogonally protected Fmoc amino acids (2a−l), we
modified the protocols reported by Degrado4b and Rapoport et
al.8 We chose the HCTU/DIPEA method over the conventional
IBCF/NMM method as it furnished 2a−l in higher yields. It is
worth noting that in all the cases except the Arg-aminoanilide
precursor (1l) (Table 1) only a single purification step after the
Figure 1. Comparison of bond lengths between amide and thioamide.
is perhaps the closest and simplest one.1 Although a thioamide is
an isosteric replacement of an amide bond, it manifests markedly
different structural and biological properties.2 The CS bond in
thioamide is significantly longer (∼1.65 Å) than a CO bond in
the peptide (∼1.23 Å) due to the increased size of sulfur.3
Thioamide N−H acts as a stronger H-bond donor than the “oxo”
amide N−H, and S acts as a weaker H-bond acceptor than O.
This aspect has, in fact, been utilized to site-specifically study the
dynamics of hydrogen bond formation in β-sheets and α-
helices.4,5 Thiopeptides have also demonstrated enhanced in
vivo activity by virtue of their enzymatic stability as compared to
their “oxo” congeners.6
One of the major hurdles toward the study of thionated
peptides has been the installation of thioamide linkage into
peptides in a site-specific and controlled manner. Available
literature on incorporation of the thioamide moiety into a
peptide chain has been limited to either using thionating agents
such as Lawesson’s reagent (LR)7 or using thioacylating agents
Received: May 21, 2015
Published: June 10, 2015
© 2015 American Chemical Society
3150
DOI: 10.1021/acs.orglett.5b01484
Org. Lett. 2015, 17, 3150−3153

Organic Letters
Letter
Table 1. Synthesis of Benzotriazolide Precursors from Fmoc
Amino Acids
Table 2. Optimization of Reaction Conditions for Thioamide
Incorporation on a Solid Support
a
a)
entry
solvent
THF
4 (%)
4a (%)
5f (%)
1
2
3
4
5
6
7
trace
trace
43
24
76
94
39
25
79
98
99
THF (anhyd)
DMF
6
18
a
NMP
42
33
entry
compd
Xaa
yield (%)
DCE
21
ND
ND
ND
1
2
2a
2b
2c
2d
2e
2f
Ala
Val
Leu
Pro
Gly
53
76
60
83
80
81
66
53
34
63
25
92
DCM
2
DCM (anhyd)
1
3
3f
DIPEA
(equiv × times)
time
(min)
4/5f
(%)
b
4
b) entry
(equiv × times)
5
1
2
5 × 2
3.5 × 2
2 × 2
2 × 2
2 × 2
10 × 2
7 × 2
4 × 2
4 × 2
2 × 2
1 × 2
3 × 2
1.5 × 2
0.75 × 2
NIL
90 × 2
90 × 2
90 × 2
45 × 2
45 × 2
45 × 2
45 × 2
45 × 2
45 × 2
45 × 2
45 × 2
2:98
3:97
7:93
2:98
1:99
1:99
2:98
1:99
1:99
2:98
2:98
6
D-Phe
7
2g
2h
2i
D-Trp(Boc)
Ser(t-Bu)
Cys(Trt)
Asp(t-Bu)
Lys(Boc)
Arg(Pbf)
3
8
4
9
5
10
11
12
2j
6
2 × 2
2k
2l
7
1.5 × 2
1.5 × 2
1.5 × 2
1.5 × 2
1 × 2
8
a
Yields of 2a−l after chromatographic purification; 1l was purified
9
prior to thionation.
10
11
2 × 2
a
Optimization done with 5 equiv × 2 of 3f. 4a: Fmoc-fVR(Pbf)G-OH.
Optimization done with DCM; % conversion calculated by HPLC.
b
thionation with Lawesson’s reagent was sufficient to obtain 2a−k
with high purity (for details, refer to Supporting Information).
Furthermore, the thionation of the aminoanilide precursor of
Fmoc-Lys(Boc)-OH (1k) into the corresponding thioacyl
precursor 2k with Lawesson’s reagent led to partial deprotection
of the Boc-protecting group. Thus, we reprotected the free
primary amine side chain of lysine using (Boc)2O/NaHCO3
(Figure S11). The benzotriazolides (3a−l) were generated prior
to the coupling onto the resin-bound peptide and were used
directly without purification due to their instability.
To identify the suitable solvent and reaction conditions for the
incorporation of thioamides into a peptide on a solid support, we
chose the tripeptide sequence VRG loaded on 2Cl-TCP resin to
react with D-phenylalanine benzotriazolide (3f) in the presence
of a non-nucleophilic base, DIPEA, as our model system. 3f was
chosen because of its ease of synthesis and handling. We carried
out our experiments on 0.03 mmol scale (with respect to 2Cl-
TCP resin). We observed that the solvent had a significant
impact on the conversion and on the undesired formation of the
oxo-amide 4a via a water-mediated pathway (Table 2a). Our
results indicated anhydrous dichloromethane (DCM) to be the
solvent of choice, but the comparable efficacy of regular DCM
prompted us to use it throughout the entire study, circumventing
the laborious process of solvent drying. Next, we optimized the
quantity of the benzotriazolide, base, and time necessary to
achieve a quantitative conversion of the N-term deprotected
tripeptide (4) to the corresponding thionated tetrapeptide (5f)
(Table 2b). It was surprising that the reduction in the amount of
the benzotriazolide and the base from 5 to 1 equiv and from 10
equiv to none, respectively, had almost no effect on the
conversion of 4 to 5f.
Thus, conventional use of 5 equiv of benzotriazolide and 10
equiv of DIPEA twice12 was found to be unnecessary, as
suggested by our data (Table 2b). In fact, the excessive use of
thioacylating agents along with base led to further complications,
such as degradation and conversion into the oxo-amide peptide,
while attempting the insertion of two consecutive thioamide
moieties in a peptide sequence. Although the importance of the
base was unclear at small-scale (0.03 mmol) couplings, we
observed its necessity while scaling up (0.18 mmol) the
benzotriazolide coupling for preparative purposes, to neutralize
the residual acetic acid (Figure S32). It is important to note here
that, although for our particular model scheme, we could achieve
comparable conversions using even lesser amounts of reagents
(entry 11, Table 2b), the substrate generality was lacking, as we
obtained lower yields with other sequences.
Subsequently, the optimized condition (entry 8, Table 2b) was
utilized to synthesize the monothionated peptides 5a−l using
3a−l with the common peptide sequence VRG loaded onto 2Cl-
TCP resin to demonstrate the substrate scope with a wide variety
of orthogonally protected benzotriazolides (Table 3b). Most of
the conversions, except for 5b, 5e, and 5i, were quantitative with
high purity (Table 3a). We speculated that the lower conversion
of 5b and 5i was presumably due to the β-branch in Val (3b) and
the bulky trityl side chain present in Cys (3i). An increase in the
amount of 3i (2.5 equiv × 2) also did not improve the conversion
to 5i. While we tried to install 3b and 3i on a different sequence,
fVRG (henceforth, the lower case letter would refer to a D-amino
acid, while the symbol “prime” would refer to the thioamide
moiety), there was hardly any improvement on the conversion
3151
DOI: 10.1021/acs.orglett.5b01484
Org. Lett. 2015, 17, 3150−3153

Organic Letters
Letter
using HCTU and DIPEA. Therefore, we resynthesized 1h using
the IBCF and NMM method12 and performed the coupling of 3h
with 4, which significantly suppressed the epimerization, yielding
5h (entry 8, Table 3b) in an epimeric ratio 15:1 (Figure S38).
We further applied our optimized protocol to incorporate two
consecutive thioamide linkages (Table 4) into a peptide
Table 3. Synthesis of Monothionated Peptides Using the
Optimized Protocol
Table 4. Synthesized Dithionated Peptides with the
Optimized Protocol
a
entry
compd
sequence
conv (%)
M − 34 (%)
1
2
6
7
A′V′VRG
f′V′VRG
A′f′VRG
V′f″VRG
G′f′VRG
f′f′VRG
w′f′VRG
D′f′VRG
R′f′VRG
f′D′VRG
D′D′VRG
f′R′VRG
D′R′VRG
89
77
80
0
0
23
18
76
0
3
8
4
9
5
10
11
12
13
14
15
16
17
18
60
95
73
95
98
90
99
80
100
6
0
7
0
8
0
a
b)
entry
compd
sequence
conv (%)
9
0
1
2
5a
5b
5c
5d
5e
5f
A′VRG
V′VRG
L′VRG
P′VRG
G′VRG
fVRG
96
83
95
95
69
99
87
92
72
97
92
99
80
80
81
74
10
11
12
13
10
0
3
16
0
4
b
5
a
Fmoc and orthogonally protected; % conversion calculated by HPLC.
6
7
5g
5h
5i
w′VRG
S′VRG
C′VRG
D′VRG
K′VRG
R′VRG
V′fVRG
G′fVRG
G′DfVR
C′fVRG
8
sequence, which is rather difficult to introduce via solution-
phase thionation. With the exception of the V′f′VRG sequence
(entry 4), we obtained reasonably good conversion and crude
purity (Figures S57−S82) of the synthesized dithionated
peptides.7b However, V′f′VRG and a few others displayed the
formation of a compound with 34 Da lower mass than the
expected peptide. We envisioned that this sequence-dependent
phenomenon could be attributed to the formation of a thiazole
moiety via an acid-catalyzed “Edman degradation”11a type
transformation with the elimination of H2S (M = 34) (see
Supporting Information for details). Note that the synthesized
dithionated peptides with our optimized protocol were
absolutely free of any oxo-amide counterpart.13
To further broaden the scope of this protocol, we attempted
the synthesis of a monothionated cyclic peptide (Scheme 1) as
peptide macrocycles are gaining attention over the past few
years.14 Synthesis of the linear peptide 21 with an internal
thioamide was accomplished using the optimized protocol, and
even after repetitive Fmoc deprotections and amino acid
couplings, the crude peptide showed a remarkably clean profile
with no traces of epimerization or oxo-amide formation (Scheme
1a and Figure S83). Finally, 21 was cleaved off of the resin and
cyclized via HOBt/DIC (Scheme 1b) and DPPA/NaHCO3
(Scheme 1c) methods in solution to form 22 with significant
purity and conversion, suggesting the compatibility of
thioamides with the cyclization reagents. Unfortunately, the
final global deprotection of 22 led to spontaneous degradation,
and we failed to characterize the degradation products.
Therefore, we further synthesized cyclo(-RGD′fV-) using
DPPA/NaHCO3-mediated cyclization (Figures S88−S91) and
deprotected the orthogonal protecting groups with a milder
cleavage cocktail of TFA/DCM/TIPS/H2O (20:75:2.5:2.5) in
contrast to the reported one (Figure S92).15
9
10
11
12
13
14
15
16
5j
5k
5l
5m
5n
5o
5p
b
b
a
Fmoc and orthogonally protected; % conversion calculated by HPLC.
Reaction with 5 equiv of benzotriazolide (1×), 30 min without base.
b
(Table 3b and Figures S48 and S55). This suggests that, for 3b
and 3i, the steric demand of the side chains dictates the final
conversion into the thionated peptides.
Coupling of glycine benzotriazolide (3e) onto 4 to yield 5e
turned out to be rather challenging. The optimized condition, as
discussed above, led to only 56% conversion into 5e. We
discovered that 3e was unstable in the presence of DIPEA,
leading into its slow degradation, consequently lowering the
product conversion. Thus, we performed a series of reactions to
optimize the coupling efficiency of 3e onto 4 (Table S2).
However, we could achieve a maximum of 69% conversion into
5e with 5 equiv of 3e and a shorter reaction time of 30 min.
Nevertheless, when this optimized condition was tested against
two other sequences, 5n and 5o, we could achieve higher
conversions of 80 and 81%, respectively (Figures S50 and S53),
suggesting that the coupling efficiency of 3e onto peptides is
sequence-dependent. The HPLC chromatogram of 5h revealed
the presence of two peaks of identical mass in the ratio of 2:1,
suggesting an epimerization during the incorporation of 3h
(Figure S36). To circumvent this problem, we performed the
reaction in the absence of DIPEA, which slightly improved the
ratio (2.5:1) but significantly reduced the conversion to 5h (78%
for both diastereomers), re-emphasizing the robustness of our
optimized protocol (Figure S37). These results suggested that
the epimerization of 3h took place during the formation of 1h
However, it is important to note that in our hands multiple
orthogonally protected thiopeptides yielded the final depro-
tected peptide with a cleavage cocktail of TFA/DCM/TIPS/
H2O (47.5:47.5:2.5:2.5) without substantial loss due to cleavage
3152
DOI: 10.1021/acs.orglett.5b01484
Org. Lett. 2015, 17, 3150−3153

Organic Letters
Letter
Scheme 1. Synthesis of Monothionated Cyclic Peptide 22
ACKNOWLEDGMENTS
■
This work was supported by Science and Engineering Research
Board (SERB) through the project SB/YS/LS-220/2013 to J.C.
We thank Sunita Prakash and Raghu Tadala, proteomics facility
MBU, IISc, for recording mass spectra.
REFERENCES
■
(1) (a) Fischer, G. Chem. Soc. Rev. 2000, 29, 119−127. (b) Choudhary,
A.; Raines, R. T. ChemBioChem 2011, 12, 1801−1807.
(2) (a) Frank, R.; Jakob, M.; Thunecke, F.; Fischer, G.; Schutkowski,
M. Angew. Chem., Int. Ed. 2000, 39, 1120−1122. (b) Goldberg, J. M.;
Batjargal, S.; Petersson, E. J. J. Am. Chem. Soc. 2010, 132, 14718−14720.
(c) Wang, Y. J.; Szantai-Kis, D. M.; Petersson, E. J. Org. Biomol. Chem.
2015, 13, 5074−5081.
(3) Seebach, D.; Ko, S. Y.; Kessler, H.; Kock, M.; Reggelin, M.;
̈
Schmieder, P.; Walkinshaw, M. D.; Bolsterli, J. J.; Bevec, D. Helv. Chim.
Acta 1991, 74, 1953−1990.
̈
(4) (a) Miwa, J. H.; Patel, A. K.; Vivatrat, N.; Popek, S. M.; Meyer, A.
M. Org. Lett. 2001, 3, 3373−3375. (b) Culik, R. M.; Jo, H.; DeGrado, W.
F.; Gai, F. J. Am. Chem. Soc. 2012, 134, 8026−8029.
(5) (a) Miwa, J. H.; Pallivathucal, L.; Gowda, S.; Lee, K. E. Org. Lett.
2002, 4, 4655−4657. (b) Reiner, A.; Wildemann, D.; Fischer, G.;
Kiefhaber, T. J. Am. Chem. Soc. 2008, 130, 8079−8084.
(6) Zacharie, B.; Lagraoui, M.; Dimarco, M.; Penney, C. L.; Gagnon, L.
J. Med. Chem. 1999, 42, 2046−2052.
(7) (a) Kessler, H.; Matter, H.; Geyer, A.; Diehl, H.-J.; Kock, M.; Kurz,
̈
of the peptide bond following thioamide linkage. We observed
that these thionated cyclic peptides are stable over a period of
months without any traceable degradation or oxidation. To the
best of our knowledge, this is the first report synthesizing a cyclic
thioamide-containing peptide on a solid support in a controlled,
regiospecific way using thioacylating agent under mild conditions
(Figure S94).
In summary, we have successfully incorporated one and two
consecutive thioamide moieties into a growing peptide chain on a
solid support in a site-specific manner, avoiding any side
reactions that complicate the synthesis. Additionally, we have
shown the broad substrate scope of our optimized protocol by
manifesting examples of aliphatic, aromatic, hydrophilic, and
hydrophobic amino acids with diverse orthogonal protecting
groups. Finally, we applied our protocol for the synthesis of cyclic
peptides containing a thioamide linkage and demonstrated the
compatibility of thioamide linkage with base, two major
cyclization agents, and acidic conditions necessary for global
deprotection. We strongly believe that, with this thoroughly
optimized synthetic strategy, mono- to multiply thionated linear
and cyclic peptides would be accessible and can pave the way
toward the long sought use of thioamides as a tool to modulate
pharmacological properties of peptides.
G.; Opperdoes, F. R.; Callens, M.; Wierenga, R. K. Angew. Chem., Int. Ed.
Engl. 1992, 31, 328−330. (b) Formaggio, F.; Crisma, M.; Toniolo, C.;
Peggion, C. Eur. J. Org. Chem. 2013, 2013, 3455−3463.
(8) Shalaby, M. A.; Grote, C. W.; Rapoport, H. J. Org. Chem. 1996, 61,
9045−9048.
(9) Lehmann, J.; Linden, A.; Heimgartner, H. Tetrahedron 1999, 55,
5359−5376.
(10) Morita, H.; Nagashima, S.; Takeya, K.; Itokawa, H. J. Chem. Soc.,
Perkin Trans. 1 1995, 2327−2331.
(11) (a) Wildemann, D.; Drewello, M.; Fischer, G.; Schutkowski, M.
Chem. Commun. 1999, 1809−1810. (b) Jensen, O. E.; Senning, A.
Tetrahedron 1986, 42, 6555−6564.
(12) Batjargal, S.; Wang, Y. J.; Goldberg, J. M.; Wissner, R. F.;
Petersson, E. J. J. Am. Chem. Soc. 2012, 134, 9172−9182.
(13) Newberry, R. W.; VanVeller, B.; Raines, R. T. Chem. Commun.
2015, 51, 9624−9627.
(14) White, C. J.; Yudin, A. K. Nat. Chem. 2011, 3, 509−524.
(15) Miwa, J. H.; Margarida, L. A.; Meyer, A. E. Tetrahedron Lett. 2001,
42, 7189−7191.
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed experimental procedures, HPLC traces, and character-
ization data. The Supporting Information is available free of
charge on the ACS Publications website at DOI: 10.1021/
acs.orglett.5b01484.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: jayanta@mbu.iisc.ernet.in.
Notes
The authors declare no competing financial interest.
3153
DOI: 10.1021/acs.orglett.5b01484
Org. Lett. 2015, 17, 3150−3153