.
Angewandte
Communications
crambin as the synthetic target. Crambin is a 46-residue
protein containing three disulfide bonds.[10] Crambin with
a V15A mutation has been frequently used as a standard to
test new methods for chemical protein synthesis.[3,4b,11] Our
synthetic strategy is shown in Figure 2A, where [V15A]-
crambin was divided into four peptide segments (10, 8, 5’’’,
and 6). All the peptide thioesters are prepared through
activation and thiolysis of the corresponding peptide hydra-
zides.[12] The one-pot experiment was carried out in the
ligation buffer containing phosphate (0.2m), urea (8m),
MPAA (0.2m), and TCEP (0.16m). The whole process was
conducted at 378C and monitored by LC–MS (Figure 2).
The first ligation between 6 (1.0 equiv, 4 mm) and 5’’’
(1.0 equiv, 4 mm) went to completion at pH 6.5 within 4 h to
generate 7. Then the pH value was raised to 11.5 for 40 min to
remove Tfacm and produce 7a. After the pH was adjusted
back to 6.5, 8 (1.0 equiv, 2.7 mm) was added to the reaction to
initiate the second ligation, which was finished within 2 h to
generate 9. The pH value was again raised to 11.5 for 40 min
to convert 9 into 9a, followed by readjustment of the
pH value to 6.5. Finally, we added 10 (1.0 equiv, 2 mm) to
the reaction. The last ligation went to completion within 20 h.
HPLC analysis of the whole experiment showed that in each
step of the one-pot process, the desired intermediate was
cleanly produced as the major peak (Figure 2B). The one-pot
experiment was completed within 30 h to produce the full-
length peptide which was isolated in 40% yield.
Figure 1. A) Ligation of 5 with 6 and monitoring of the reaction by
analytical HPLC (l=214 nm): a) before addition of NaNO2; b) after
addition of NaNO2 for 15 min, pH 3.0; c) after addition of MPAA and
adjustment of the pH value to 6.5 for 10 min; d) after addition of 6 for
4 h, pH 6.5. B) Deprotection of Tfacm. The solid lines represented
~
&
*
reactions at 378C and at pH 9 ( ), 10 (&), 11 ( ), 11.5 ( ), and 12
!
*
( ). The dotted line ( ) represented reaction at 258C and pH 11.5.
C) Analytical HPLC traces for the removal of Tfacm from 7 by adjust-
ment of the pH value to 11.5 at 378C. **: peak attributable to MPAA.
Note that addition of TCEP significantly improved the
reaction, as TCEP effectively prevented base-mediated
elimination of disulfides to form dehydroalanine (Dha).[13]
Indeed, in the absence of TCEP a byproduct with an
increased mass of 134 Da was detected during the depro-
tection step, which could be attributed to the addition of
MPAA to Dha (see the Supporting Information for details).
Synthetic [V15 A]crambin was folded (Figure 2C) and
then crystallized at 188C by the vapor-diffusion method. Its
X-ray crystal structure solved at 2.1 ꢀ resolution (Fig-
ure 2D,E) was fully consistent with the reported data.[14] To
further verify that all amino acids remained intact in the one-
pot process, we reduced the synthetic [V15 A]crambin by
dithiothreitol (DTT) and alkylated its cysteine residues with
iodoacetamide. After digestion by trypsin, we obtained three
peptide fragments which were analyzed by (LC–MS)/MS. The
mass spectrum of the internal segment “SNFNAC*R” was
shown in Figure 2F as a representative, which was identical to
that expected from the user database (for more information
see the Supporting Information; user database built by
Proteome Discoverer 1.4 Software (Thermo Scientific)
according to the sequence provided). Collectively, our X-ray
crystal data and (LC–MS)/MS analysis confirmed that
crambin prepared by the one-pot four-segment condensation
method was correct at the atomic level.
Next, we tried the ligation reaction between Tfacm-
protected peptide hydrazide 5 and a model Cys-peptide,
namely Cys-Pro-Gly-Asp-Tyr-Ala-Asn (6; Figure 1A). In
aqueous phosphate buffer, 5 was cleanly oxidized by NaNO2
at pH 3.0 and À158C for 15 min to generate a peptide azide 5’.
After addition of 4-mercaptophenylacetic acid (MPAA) and
adjustment of the pH value to 6.5, 5’ was fully converted into
a peptide thioester 5’’, which then reacted with 6 to afford the
ligation product 7 in quantitative conversion within 4 h.
To remove Tfacm, 5 was dissolved in aqueous phosphate
buffer (0.2m) containing urea (8m), MPAA (0.1m), and TCEP
(0.08m; tris(2-carboxyethyl)phosphine). According to HPLC
analysis (Figure 1B), deprotection at pH 9–11 was not
efficient, whereas clean and rapid removal of Tfacm could
be achieved at pH 11.5 and 12.
To test whether the deprotection involved racemization at
the N-terminal cysteine residue, we compared 5a, prepared
by deprotection of 5 at pH 11.5 for 40 min, with Cys(d)-Ile-
Ile-Ile-Pro-Gly-Ala-Thr-NHNH2 (5b). The experiment
showed that the extent of racemization was less than 1%
(Supporting Information).
Next, we performed the deprotection reaction by adjust-
ing the pH value in situ to 11.5 after the ligation between 5
and 6 was completed. As shown in Figure 1C, a quantitative
conversion from 7 into 7a was achieved within 40 min at
378C.
To examine the use of the new one-pot method for the
synthesis of more practically useful proteins, we chose human
chemokine hCCL21 as another synthetic target. hCCL21
contains 111 residues and six cysteines. Through activation of
G-protein-coupled receptor CCR7, hCCL21 recruits normal
immune cells and metastasizes tumor cells to lymph nodes.[15]
The details of hCCL21 biology have attracted increasing
The above findings laid the foundation for the new one-
pot ligation method. To test its efficiency, we first chose
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 6
These are not the final page numbers!