R. Joseph et al. / Tetrahedron Letters 56 (2015) 4302–4304
4303
O
That no reduction of the thioester to aldehyde was observed in
this reaction was taken as initial support for our steric argument.
However, at this stage we reached an impasse. All attempts to cou-
ple the carboxylic acid 8 with even a simple amine to produce an
amide in good yield met with failure. The attempted reaction using
a carbodiimide coupling reagent was illustrative, producing a mix-
ture that indicated competitive formation of a reactive anhydride
(5, R0 = Fmoc, X = O) which triggered other undesired reactions.
We conjectured that this entropically-favored cyclization must
proceed through a carboxylate anion and that a weaker nucle-
ophile might shut down (or at least mitigate) this unwanted path-
way. Since the Pd-catalyzed de-allylation reaction had presumably
passed through a silyl ester intermediate without a problem, we
decided to explore the use of an analogous silyl ester as a neutral
carboxylate surrogate in a peptide coupling context.12 In the event,
trimethylsilyl ester 9, produced quantitatively by the action of
TMSCl on 8, underwent a very clean reaction with benzylamine
(DIC + HOBt in DCM) to give Fmoc-Asp(STrt)-NHBn (10) in 79%
yield. We were now poised to apply this reaction to the SPPS of
O
O
TrtSH
EDC, DMAP
STrt
OAll
OH
OAll
FmocHN
FmocHN
CH2Cl2
65%
O
7
6
PhSiH3
Pd(PPh3)4
CH2Cl2
90%
O
O
O
TMSCl
CH2Cl2
STrt
OH
STrt
OTMS
FmocHN
FmocHN
quant.
O
9
8
x
-aspartic thioacid-containing peptides.
Scheme 2. Synthesis of thioaspartic acid building block 9.
We chose the heptapeptide thioacid H-Val-Gln-Lys-Asp(SH)-
bulky yet readily removable using a standard acidic global depro-
tection cocktail. This line of reasoning eventually led us to consider
the use of a trityl thioester for this purpose. Our choice of this pro-
tecting group for the thioacid was guided by the knowledge that
tertiary thioesters are known to resist Pd-mediated Fukuyama
reduction8 as well as nucleophilic acyl substitution during
standard Fmoc-deprotection.9 Thus, Fmoc-Asp(STrt)-OH (8) was
synthesized from the known aspartic acid derivative 610 via
Steglich thioesterification with TrtSH11 followed by Pd-mediated
deallylation to give 8 (Scheme 2).
Val-Thr-Ser-NH2 (11) as our first target since it contains the
progenitor to typical N-glycopeptide consensus sequence
(Asn(glycan)-Xaa-Thr/Ser). Starting with Rink amide resin
(Scheme 3), the resin-bound tetrapeptide Fmoc-Asp(STrt)-Val-
Thr(tBu)-Ser(tBu)-NH-Rink (12) was assembled without problem
using standard Fmoc removal and HATU coupling conditions for
the first three amino acids followed by DIC-HOBt-DIEA coupling
of the TMS ester 9. LC/MS analysis after micro-cleavage from the
resin confirmed formation of the expected tetrapeptide thioacid.
No aspartimide was detected. However, the situation changed
a
a
O
O
t
N
CONH-Ser( Bu)-NH-Rink
CONH-Ser(tBu)-Gly-O-2-Cl-Trt
N
H2N
17
H2N
O
20
O
iPr
iPr
1. 9, DIC, HOBt, DIEA, DCM
2. 6% piperazine, DMF
STrt
O
N
STrt
O
N
t
O
CONH-Ser( Bu)-NH-Rink
CONH-Ser(tBu)-Gly-O-2-Cl-Trt
O
H
N
H
N
18
21
H2N
O
H2N
O
O
iPr
O
iPr
1. Fmoc-AA-OH, HBTU, DIEA, DMF
2. 6% piperazine, DMF
3. Boc-Val-OH, HBTU, DIEA, DMF
STrt
O
STrt
O
N
CONH-Ser(tBu)-NH-Rink
19
O
N
CONH-Ser(tBu)-Gly-O-2-Cl-Trt
O
H
N
H
N
Boc-Val-Gln(Trt)-Lys(Boc)N
H
22
O
Boc-Val-Gln(Trt)-Lys(Boc) N
H
O
O
iPr
O
iPr
92.5% TFA, 5% DMB, 2.5% TES
H-Val-Gln-Lys-Asp(SH)-Val-Thr-Ser-NH2
H-Val-Gln-Lys-Asp(SH)-Val-Thr-Ser-Gly-OH
23 (62%)
11 (81%)
Scheme 3. Synthesis of
x-thioaspartic acid containing peptides.