Angewandte
Communications
Chemie
through enzymatic trans-glycosylation after the ligation step.
In the sugar-assisted ligation approach (Scheme 1b), Wongꢀs
group[13] demonstrated that a non-cysteine terminated glycan-
linked peptide fragment can be used to join another standard
thioester peptide. This allows less fragmentation of the
peptide synthetic plan, accelerating the assembly of the N-
glycopeptide. However, the sulfur appendage needs to be
installed on the sugar unit, followed by glycosylation with Asn
residue. Preparation of this crucial species is no trivial task
which requires laborious work and careful planning, espe-
cially with more complex glycan chains.[13c] The modified
sugar is also resistant to enzymatic trans-glycosylation and an
additional desulfurization step is required. Understandably,
the sulfhydryl group only participated in ligating the two
peptide fragments and not in the glycosylation of the amino
acid itself. Nevertheless, the method elegantly overcomes the
requirement for cysteine residue at the ligation junction.
We studied the scope of NCL with modified glycosyl
amine 1a (Scheme 1c). During our preliminary study, the
trimethoxybenzyl auxiliary was found to be the most profi-
cient in the ligation and in its removal, in line with previous
studies.[14] The 3-nitro-2-pyridinesulfenyl group is well-known
for rapid disulfide exchange under NCL condition.[15] This
scaffold is therefore capable of performing ligation to
produce the glycan-linked N-terminal asparagine oligopep-
tide 3, followed by a second ligation with C-terminal thioester
oligopeptide 4 to provide the glycopeptide 5. Subsequently,
the auxiliary is readily cleaved following treatment with TFA
and deacetylation to furnish glycopeptide 6.
We conducted our first ligation of compound 1a with
oligopeptide sequence H-(COSPh)DATGVS-CONH2 12a in
a phosphate buffer (pH 7.5) at 378C. Reaction progress was
monitored by reverse-phase HPLC, and a ligation conversion
of more than 70% was observed after 1 hour. Optimization of
additives and buffer revealed the best ligation efficiency was
achieved when the reaction was conducted in a phosphate
buffer of pH 8.0 with 6m Gn·HCl, 60 mm TCEP, and 1% v/v
MeSNa as additives at 378C. The desired ligation product 13a
was formed in 92% conversion within 1 hour with less than
3% of peptide racemization at N-terminal Asp residue
(Figure 1). A control experiment was carried out by mixing
thioesters 12a and additives in our buffer condition. Less than
3% racemization was observed after 12 hours and only 7%
racemization was detected with 72 hours of incubation time.
To examine the effect of different residues at the ligation
junction, a series of oligopeptides (12a–12h) with thioesteri-
fied aspartic acid on the N-terminal were subjected to the
optimized conditions. As seen from Table 1, all native
chemical ligation proceeded smoothly and high to excellent
conversions could be achieved. Slightly lower conversions
Table 1: First native chemical ligation of oligopeptide thioester 12 and
sugar auxiliaries 1a–c.[a]
Entry
Peptide 1[b]
Product
Conv [%][c]
1[d]
2[d]
3[d]
4[d]
5[d]
6[d]
7[d]
8[d]
9[d]
10[e]
11[f]
12a ATGVS-CONH2
12b GTGVS-CONH2
12c I TGVS-CONH2
12d LTGVS-CONH2
12e FTGVS-CONH2
12 f STGVS-CONH2
12g TTGVS-CONH2
12h VTGVS-CONH2
12i GTGVS-CONHNH2
12i GTGVS-CONHNH2
12i GTGVS-CONHNH2
13a
13b
13c
13d
13e
13 f
13g
13h
13i
92
91
89
90
87
94
91
90
93
87
81
13j
13k
[a] Conditions: 6m Gn·HCl, 60 mm TCEP, 1% MeSNa, pH 8.0, 378C, 1 h.
[b] ratio of 1 to 12 is 1.3:1. Amino acids adjacent to aspartic acid are
highlighted in bold. [c] Conversion based on RP-HPLC analysis upon
complete consumption of 1. [d] 1a: Glc(OAc)NAc. [e] 1b: unprotected
GlcNAc. [f] 1c: unprotected maltotriose.
were observed when Ile, Leu, Val, and Phe were present next
to the ligation site, which is most likely due to increase in
steric bulk. Liuꢀs group[16] recently developed a facile method
for in situ formation of peptide thioester from a C-terminal
hydrazide, allowing the convergent synthesis of several
proteins.[17] In our experiment, the hydrazide peptide 12i
gave the first ligation product in excellent conversion
(entry 9).
The b-N-glycosidic bond of 13i remained intact based on
1D-TOCSY NMR analysis (see the Supporting Information).
Evaluation of the reaction with unprotected GlcNAc and
unprotected maltotriose as glycan auxiliaries showed excel-
lent conversions without compatibility issue (entries 10,11).
However, we found unprotected sugar auxiliary more prone
to hydrolysis compared to peracetylated form. Anomeriza-
tion of the glycosidic benzyl amine was observed to occur at
room temperature after several hours. Nevertheless, glyco-
peptide 13j and 13k were stable during isolation and
purification. We recommend carrying out the first ligation
reaction immediately after furnishing the auxiliary from
unprotected glycosylamine.
We then investigated the second native chemical ligation
of glycopeptide 13 with a C-terminal oligopeptide thioester.
Under similar conditions, glycopeptide 13a was treated with
oligopeptide thioester H-ISTVS-COS(CH2CH2CONH2) 14a
(Table 2, entry 1). A final ligation conversion of 78% was
Figure 1. Analytical RP-HPLC (l=280 nm) of the first ligation (Table 1,
entry 1). Reaction at 1 h: peak a: ligation product 13a with calcd
M=1087.4 Da, obs. M=1088.1 Da (see inset, ESI-MS); peak b: excess
of 1a (disulfide-bond cleaved).
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 6
These are not the final page numbers!