Angewandte
Communications
Chemie
We therefore decided to explore a modified Boc SPPS
route to peptide thioesters that used a milder acid for final
cleavage. This approach would have the following benefits:
the direct synthesis of peptide thioesters with a single, safer,
cleavage step; the possibility of parallel synthesis and small-
scale cleavages and greater compatibility with post-transla-
tional modifications. The superior quality of peptides syn-
thesised by Boc SPPS was the main factor.[13,17]
There have been many attempts to substitute HF with
trifluoromethanesulfonic acid (TFMSA).[18] However, in
contrast to HF, it is difficult to remove as it is not volatile,
and residual reagent can cause degradation of the peptide
product. A two-step deprotection has been evolved to address
this problem.[19]
TFA/TMSBr (trimethylsilyl bromide) mixtures can cleave
benzyl-based side-chain protection from Asp/Glu/Ser/Thr/
Lys/Tyr/Cys and Mts from arginine.[20] As demonstrated in the
elegant work of the Yajima group TFA/TMSBr has many
notable properties as a cleavage reagent: it is volatile, more so
than TFA, and therefore easily removed by sparging; it
reduces any methionine sulfoxide formed during synthesis;
and benzyl deprotection of aspartyl residues is not accom-
panied by aspartimide formation.[20] It is routinely used in
Fmoc SPPS when stronger deprotection conditions are
required.[21]
Unfortunately, TFA/TMSBr is not a sufficiently strong
acid to achieve complete peptide cleavage from methylben-
zylhydrylamine (MBHA) and 4-(hydroxymethyl)phenylace-
tamidomethyl (PAM) resins and deprotect all the standard
side-chain protection of classical Boc SPPS. However, these
limitations could potentially be overcome by combining TFA/
TMSBr cleavage with Merrifield hydroxymethyl resin, Boc
in situ neutralization protocols[14] and changing some of the
side-chain protection for that identified by the Yajima group
as most compatible: Arg to Mts, Cys to Mob and Merrifieldꢀs
original choice of benzyl for Asp and Glu (Table S1 in the
Supporting Information (SI)). Cyclohexyl had been intro-
duced primarily because of concerns over aspartimide
formation during HF cleavage with benzyl protection,[15b]
not so problematic with TFA/TMSBr cleavage.[20]
The use of Merrifield resin for Boc SPPS was largely
abandoned when 4-(Hydroxymethyl)phenylacetamidomethyl
(PAM) and 4-methylbenzylhydrylamine (MBHA) resins were
introduced, primarily because of reported peptide cleavage
by TFA over the prolonged (20 min) TFA cleavage cycles.[22]
However, contemporary Boc SPPS in situ neutralization
protocols with their shorter, typically, 2 ꢁ 1 min treatment
with TFA per cycle[14] are more suitable. Merrifield resin had
also been associated with other side reactions, notably
capping by trifluoroacetylation and formation of deletion
sequences caused by the presence of aldehyde impurities on
the resin.[23] However, these problems were identified many
years ago, before improvements in the chemical purity of
commercial resins and before the adoption of HBTU and
other uronium coupling agents that do not couple trifluoro-
acetic acid. Consequently, a re-examination of Boc SPPS on
modern preparations of Merrifield resin, with in situ neutral-
ization protocols was timely.
First, the stability of Merrifield resin linked peptides to
TFA was reinvestigated. For this, we used hydroxymethyl
resin derivatized with Fmoc-Gly because Gly is one of the
more acid labile residues[22] and Fmoc provides a good
reporter to monitor loss. The rate of Fmoc-Gly cleavage by
neat TFA was monitored directly by following the release of
Fmoc by UV (Figure 1B). Less than 5% amino acid cleavage
from the resin was observed after 500 min TFA treatment.
The experiment was repeated with Fmoc-Gly attached to
mercaptopropionicacidleucine (MPAL) linker (Figure 1B)
used for thioester synthesis. Chain loss was considerably
slower with the MPAL linker presumably because of the
comparably more hindered terminal leucine residue. In both
cases the loss of amino acid was in general agreement with
previous measurements.[22] Although there was chain loss with
the MPAL linker, at 1% over 500 min, it was suitable for
in situ neutralization cycles (Figure 1B).
Next, we investigated the use of Boc in situ neutralization
cycles on Merrifield resin to see if the notorious capping and
deletion reactions occur under these conditions. Elastin
1 (Table 1) has been synthesised recently by Kent and co-
workers and gave us a direct comparison of yield and purity to
current, optimized, Boc-SPPS.[24] The resin was derivatized
with PBr3 and Fmoc-valine to obviate racemization.[25] Syn-
thesis was carried out manually with in situ neutralization Boc
cycles and the peptide cleaved with TFA/TMSBr/thioanisole/
EDT (1:0.05:0.05:0.025) at room temperature for 1 h. The
Table 1: Yields of peptides synthesized via modified Boc SPPS.
Peptide Sequence
Crude
Isolated
yield [%][a] yield [%][b]
1
2
3
4
5
6
7
8
H-PGVGPGVGV-OH
H-GCCSDPRCRYRCR-OH
H-LAPAV-MPAL
H-LAPAA-MPAL
H-LAPAG-MPAL
H-LYRAF-MPAL
H-LAPAG-MPAA
H-LAPAA-MPAA
H-LAPAQ-MPAA
H-LAPAV-MPAA
H-LAPAT-MPAA
H-LAPAW-MPAA
H-LYRAI-MPAA
H-LYRAL-MPAA
98
77
92
90
86
81
68
84
83
89
86
83
91
82
57
66
81
21
54
50
22
34
16
31
25
36
36
26
28
23
23
24
16
9
10
11
12
13
14
15
16
17
H-LETVSpTQELY-MPAA
H-LKAQADIYKA-MPAA
H-dAlaArgArgArgdNalArgPhe(4-F) 84
dNleGlnTrpThr-MPAA
18
19
H-CdYVYNTRSGWRWYT-MPAA
H-AEQH(DNP)KIVMETVPLKAQA
DIYKA-MPAA
88
51
28
14
20
21
22
H-LEDLRQQLQQAEEALVAKQE
LI-MPAA
H-LEDLRQQLQQAEEALVAKQELI
DKL-MPAA
H-LEDLRQQLQQAEEALVAKQELI
DKLKEEA-MPAA
80
68
64
19
13
10
[a] Yield calculated from resin loading and weight of crude, unpurified
peptide. [b] Isolated yields calculated on dried weight of purified peptide
(including TFA salts).
2
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 7
These are not the final page numbers!