Ascorbate as an Alternative to Thiol Additives in Native Chemical Ligation

Article abstract of DOI:10.1002/cbic.201100179

Vitamin C for peptide ligation: Phosphine and thiol additives maintain reducing environments and increase the reactivities of peptide thioesters in native chemical ligation. Thiol additives such as thiophenol also act as radical scavengers that inhibit phosphine-induced desulfurization of cysteine. This role can be assumed by the odourless, nontoxic, highly water-soluble and inexpensive ascorbate, which can replace the usually added thiols.

Full text of DOI:10.1002/cbic.201100179

DOI: 10.1002/cbic.201100179
Ascorbate as an Alternative to Thiol Additives in Native Chemical Ligation
Heike Rohde, Josephine Schmalisch, Ziv Harpaz, Franziska Diezmann, and Oliver Seitz*[a]
Among the segment coupling methods used for the chemical
synthesis of proteins,[1] native chemical ligation is one of the
most powerful and most widely applied reactions.[2] This
method provided a major boost to chemical protein sciences,
as can be inferred from the increasing amount of publications
in which native chemical ligation is employed to construct bio-
logically active proteins. The approach involves a peptide thio-
ester such as 1 (Scheme 1) and an N-terminal cysteinyl peptide
ethylenephosphine (TCEP) is added to prevent the formation
of disulfides. Thiol additives such as 2-mercaptoethanesulfo-
nate sodium salt (MESNa), benzyl mercaptan, thiophenol or (4-
mercaptophenyl)acetic acid are commonly included to help
maintain reducing environments and to increase the reactivi-
ties of peptide thioesters.[4] Alternatively, native chemical liga-
tions have been performed by using combinations of benzyl
mercaptan/thiophenols, in which case there is no need for
TCEP as a reducing agent.[5] Despite their widespread use,
however, thiols are not among the most appreciated reagents.
Some thiols are characterized by unpleasant odours and toxici-
ties. Moreover, the thiol additives can co-elute with products
in HPLC analysis and form mixed disulfides, which can compli-
cate the analysis of native chemical ligation at long reaction
times. In addition, thiol-induced side reactions such as addition
to multiple carbon-carbon bonds have been reported.[6] The
unwanted effects can be avoided when the native chemical li-
gation is performed with TCEP as sole additive.[7]
In this communication we show that the use of TCEP as sole
additive can, however, lead to desulfurization both of starting
cysteinyl peptides and of formed ligation products. Our find-
ings suggest that the commonly used thiols such as thiophe-
nols act not only as reducing and thiol exchange agents but
also as radical scavengers to prevent the TCEP-induced desul-
furization. We describe a native chemical ligation format that
involves the use of TCEP and yet proceeds remarkably well
without the involvement of thiol additives. This reaction is
based on the use of the nontoxic, highly water-soluble radical
scavenger ascorbate, which is available at very low cost.
In a study involving investigation of the reactivities of vari-
ous peptide thioesters, we observed a side reaction. We per-
formed the native chemical ligation of the peptide thioester 1
with the cysteinyl peptide 2 in the presence of TCEP without
use of thiol additives (Scheme 1a). Two unexpected peaks ap-
peared in the HPLC trace together with the ligation product 3
(Figure 1A). MS analysis suggested the formation of the ala-
nine-containing peptides 4 and 5. HPLC-MS analysis of authen-
tic reference compounds (Figures S3 and S4 in the Supporting
Information) confirmed this finding.
Scheme 1. Synthesis of the peptide H-LYRAACRAEYSK-NH2 (3) by native
chemical ligation. The ligation reactions were performed in different buffer
systems (conditions a–c) with 1 (2 mm) and 2 (4 mm). Desulfurization of the
cysteinyl-peptide 2 and the ligation product 3 to afford the alanyl-peptides
4 and 5, respectively, was observed under conditions a and conditions b
(RSH=BnSH or NaO3SCH2CH2SH), but not with PhSH.
such as 2, which react in aqueous solutions at neutral pH to
furnish a native peptide bond as in 3.[3] Native chemical liga-
tion proceeds through a reversible thiol exchange reaction,
which eventually involves the side chain of the N-terminal cys-
teine in 2. The formed thioester-linked ligation intermediate
undergoes a spontaneous S!N acyl shift and provides the
thermodynamically favoured amide bond at the ligation site
in 3.
We assumed that the desulfurization of the cysteine residues
would have been induced by TCEP. This reaction has prece-
dence: it has been shown that phosphites and phosphines
trigger the radical-based desulfurization of alkyl mercaptans, a
reaction that proceeds slowly at ambient temperature but can
be driven to completion at elevated temperatures.[8–11] This
side reaction has also been observed in the TCEP-induced
cleavage of SꢀStBu-protected cysteines.[12] Very recently the oc-
currence of TCEP-induced desulfurization during sample prepa-
ration for protein characterization by mass spectrometry was
proposed.[13]
The aqueous buffer systems used in native chemical ligation
reactions include a number of additives. Typically, triscarboxy-
[a] H. Rohde, J. Schmalisch, Z. Harpaz, F. Diezmann, Prof. Dr. O. Seitz
Institut fꢀr Chemie der Humboldt Universitꢁt zu Berlin
Brook-Taylor-Strasse 2, 12489 Berlin (Germany)
Fax: (+49)30-2093-7266
E-mail: oliver.seitz@chemie.hu-berlin.de
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201100179.
1396
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1396 – 1400

Table 1. Yields of native chemical ligation and amount of desulfurization
of ligation products and unconsumed cysteinyl peptides.
Ligation conditions[a]
Yields[b,c] [reaction time]
LYRAACRAEYS (3)
LB A, pH 7.5
LB A, pH 7.5, PhSH (3 v/%)
3 (97%), 4 (3%), 5 (7%) [2 h]
3 (quant., 55% isol.),[d]
no desulfurization [2 h]
LB A, pH 7.5, BnSH (3 v/%)
LB A, pH 7.5, MESNa (50 mm)
LB A, pH 6.5
3 (96%), 4 (4%), 5 (9%) [2 h]
3 (98%), 4 (2%), 5 (13%) [2 h]
3 (81%), 4 (1%), 5 (2%) [2 h]
3 (quant., 51% isol.),[d] no desulfuri-
zation [2 h]
Figure 1. UPLC traces for the ligation of the peptide thioester 1 with the
peptide 2 performed in A) TCEP, B) TCEP/PhSH, and C) TCEP/ascorbate buf-
fers. Conditions: 1 (2 mm), 2 (4 mm), NaH2PO4 (100 mm), TCEP (20 mm),
pH 7.5, 258C, PhSH (3 vol% when added) or sodium ascorbate (50 mm when
added). Compound 6: disulfide (-CRAEYSK)2. TCEP=triscarboxyethylene-
phosphine.
LB A, pH 7.5, ascorbate (50 mm)
LYRAPCRAEYS (12)
LB B, pH 7.5
12 (7%), 5 (61%), desulfurized 12
(2%) [30 h]
12 (11%), 5 (21%), desulfurized 12
(1%) [30 h]
LB B, pH 7.5, BnSH (3 v/%)
We examined the reaction between 1 and 2 in commonly
used ligation buffers containing TCEP and thiol additives. No
desulfurization products were detectable when thiophenol was
added (Figure 1B). Benzyl mercaptan and MESNa were found
to be less effective than thiophenol in preventing desulfuriza-
tion (Table 1). Careful GC-MS and NMR analysis revealed that
thiophenol did not serve as surrogate of cysteine desulfuriza-
tion, because no formation of benzene could be detected.
Rather, the mercaptan probably acted as a radical scavenger.
Thiophenol readily forms a stabilized benzenethiyl radical (SꢀH
bond dissociation energies: thiophenol 322 kJmolꢀ1, cysteine
367 kJmolꢀ1).[14] It has been shown that thiophenol inhibits the
phosphite-induced radical desulfurization of mercaptoal-
kanes.[9]
LB B, pH 7.5, PhSH (3 v/%)
LB B, pH 7.5, ascorbate (50 mm)
LNELDADEQADLCESLHDHADELYRSCLARFGDDGENL (13)
LB B, pH 7.5
LB B, pH 7.5, BnSH (3 vol%)
LB B, pH 7.5, PhSH (3 vol%)
LB B, pH 7.5, MESNa (50 mm)
LB B, pH 7.5, ascorbate (50 mm)
LB B, pH 7.5, ascorbate
(50 mm)+PhSH (3 vol%)
12 (16%), no desulfurization [30 h]
12 (11%), no desulfurization [30 h]
13 (17%), desulfurization[e] [30 h]
13 (35%), desulfurization[e] [30 h]
13 (74%) [30 h]
13 (58%), desulfurization[e] [30 h]
13 (51%) [30 h]
84% 13 [30 h]
DVPLPAGWEMAKTSCGQRYFLNHIDQTTTWQDPRKAML (14)
LB B, pH 7.5
14 (37%, 26% isol.),[d] desulfuriza-
tion[e] [24 h]
14 (86%, 30% isol.)[d] [24 h]
14 (91%, 40% isol.)[d] [24 h]
14 (65%, 35% isol.),[d] desulfurized
11[e] [24 h]
LB B, pH 7.5, ascorbate (50 mm)
LB B, pH 7.5, PhSH (3 vol%)
LB B, pH 7.5, MESNa (50 mm)
The TCEP-induced desulfurization was found to be pH-de-
pendent and occurred with lower rates when performed at
slightly acidic pH 6.5 (Table 1). However, the native chemical
ligation is slower under these conditions (pH 7.5: 97%; pH 6.5:
81% after 2 h), which can limit the achievable yields in the
native chemical ligation of more challenging peptides.
[a] Ligation buffer A: NaH2PO4 (100 mm), TCEP (20 mm). Ligation buffer B:
NaH2PO4 (100 mm), Gn·HCl (6m), TCEP (20 mm). [b] The peptide 12 was
synthesized from the thioester 7 and the cysteinyl peptide 2, the pep-
tide 13 was synthesized from the thioester 8 and the cysteinyl peptide 9
and the peptide 14 was synthesized from the thioester 10 and the cys-
teinyl peptide 11. [c] The ligation yield was determined by integration of
the peak areas. [d] Yield obtained after HPLC purification. [e] Observed by
ESI mass spectrometry.
We considered the possibility of using other, odourless radi-
cal scavengers and examined sodium ascorbate as a water-
soluble, nontoxic and inexpensive alternative to thiols. We
were pleased to find that the reaction between the model
peptide thioester 1 and the model cysteinyl peptide H-
CRAEYSK-NH2 2 occurred without any desulfurization when as-
corbate (50 mm) was included (Figure 1C). The yield of ligation
product obtained after HPLC purification and the purity of the
crude product were comparable to those obtained in the “con-
ventional approach” based on addition of thiophenol (51 or
55%, respectively).
with the reaction product 12 being formed in 11% yield. A
similar ligation yield was obtained when ascorbate was used
as additive. This radical scavenger completely inhibited desul-
furization, reducing the spectrum of products detected by
HPLC analysis (Figure S12 in the Supporting Information).
To evaluate the usefulness of ascorbate as a thiol replace-
ment in native chemical ligation further, we studied the liga-
tion reactions between the longer peptides 8 and 9 and be-
We next turned our attention to a difficult ligation reaction.
The Pro-Cys bond is the most challenging ligation site. Here,
the problem of desulfurization is particular pressing because
ligation can proceed at a slower rate than desulfurization. The
reaction between the peptide proline thioester 7 and the cys-
teinyl peptide 2 was sluggish (Figure S12 in the Supporting
Information). In the absence of radical scavengers, only a 7%
yield of the ligation product 12 was obtained after 30 h reac-
tion time (Table 1). More than 60% of the cysteinyl peptide 2
was consumed in the desulfurization. The addition of benzyl
mercaptan reduced the amount of desulfurization to 21%,
ChemBioChem 2011, 12, 1396 – 1400
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1397

tween 10 and 11 in the syntheses of the C-terminal segment
(26–63) of the ColE1-Rop protein (13) and the 38-amino-acid-
long WW domain (14) of human Yes-associated protein (hYAP),
respectively. The ligation buffer included guanidinium hydro-
chloride (Gn·HCl, 6m) for denaturation and TCEP as reducing
agent. The reaction between the leucine thioester in 8[15,16] and
the 26-residue cysteinyl peptide 9 proceeded smoothly when
ascorbate was included in the buffer (Figure 2A). After 30 h
reaction time the ligation product 13 (Table 1) was obtained in
51% yield based on HPLC analysis. Similar yields were ob-
tained when MESNa was included as thiol additive.
Supporting Information). It is worth noting that the desulfuri-
zation was difficult to detect by HPLC analysis alone because
the desulfurization products had elution properties similar to
those of the cysteine-containing peptides before desulfuriza-
tion. The use of benzyl mercaptan led to improved ligation
yields, but not to the extents observed with ascorbate or
MESNa. Again, MS analysis revealed that benzyl mercaptan
failed in preventing desulfurization (Figure 2C). As was ob-
served previously, the use of thiophenol as additive (74% liga-
tion yield) proved superior to the use of benzyl mercaptan and
also completely prevented desulfurization (Figure S13 in the
Supporting Information).[17] Interestingly, the combination of
thiophenol with ascorbate conferred a further improvement to
84% yield. This might point to a potential activating effect of
ascorbate or simply indicate that the addition of ascorbate can
increase the concentration of thiophenol available for the
native chemical reaction.
We then investigated native chemical ligation reactions in
the synthesis of the 38-amino-acid-long WW domain 14 (Fig-
ure 3A). The reaction between the peptide thioester 10[16] and
the cysteinyl-peptide 11 was performed in the absence of thiol
additives and ascorbate. HPLC analysis showed that under
these conditions only a 37% yield of the ligation product 14
was formed after 24 h. The mass spectrometric characterization
showed that substantial amounts of the ligation product 14
and the unconsumed cysteinyl peptide 11 were desulfurized to
form the alanine-containing peptides 17 and 18, respectively.
In the presence of sodium ascorbate (50 mm) neither desulfuri-
zation of the cysteinyl peptide 11 to 17 nor desulfurization of
the ligation product 14 occurred and the ligation succeeded in
86% yield based on HPLC analysis (Figure 3B, Table 1). The
WW domain 14 was isolated in 30% yield after HPLC purifica-
tion. A similar yield (91%) was obtained when thiophenol was
included in a “conventional” ligation buffer (Table 1, Figure S15
in the Supporting Information). Again, desulfurization was in-
hibited. The MESNa additive also provided useful product
yields (65%), although MS analysis revealed a minor degree of
desulfurization of the cysteinyl peptide.
In previous studies we had noticed that the thiol reagents
commonly used in native chemical ligation chemistry can trig-
ger undesired side reactions.[6] For a project concerned with
the controlled spatial display of peptides on DNA architec-
tures[18] we prepared various peptide-oligonucleotide conju-
gates by making use of native chemical ligation of peptide
thioesters such as 1 with internally cysteine-modified oligonu-
cleotides such as 20 (Figure 4A). The aqueous ligation buffer
included MESNa. However, HPLC analysis showed a broad
product peak and MALDI-mass spectrometry revealed the oc-
currence of a mercaptoethanesulfonic acid (MESA) adduct (Fig-
ure 4B), which remained stable even after repeated treatment
with TCEP. The side reaction was prevented when light was ex-
cluded.[6] It was concluded that the thiol adduct was formed
upon free radical addition to the triple bond of the aminopro-
pagyl linkage in 20. We performed the reaction between 1 and
20 in the presence of light but included ascorbate as radical
scavenger. We found that the reaction proceeded smoothly
(Figure 4C). Neither desulfurization nor formation of adducts
Figure 2. UPLC traces for the ligation between the peptide thioester 8 and
the peptide 9 performed in A) TCEP/ascorbate, B) TCEP, or C) TCEP/BnSH buf-
fers. Left: UPLC traces after 30 h. Right: ESI-MS of the cysteinyl peptide after
30 h. Conditions: peptides (2 mm concentrations), TCEP (20 mm), Gn·HCl
(6m), NaH2PO4 (100 mm), pH 7.5, 258C, 30 h and A) sodium ascorbate
(50 mm), B) only TCEP (20 mm), and C) BnSH (3 vol%). Peptide 9: CESLHDH-
ADELYRSCLARFGDDGENL. Peptide 15: AESLHDHADELYRSALARFGDDGENL.
Peptide 15a: AESLHDHADELYRSCLARFGDDGENL. Peptide 16: LNELDADEQ-
ADLAESLHDHADELYRSALARFGDDGENL.
In the absence of ascorbate ligation was slow (17% yield,
Table 1). HPLC-MS analysis revealed that the desulfurization
products 15 and 16 were formed (Figures 2B and S13 in the
1398
www.chembiochem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1396 – 1400

Figure 4. A) Synthesis of the peptide-DNA conjugate 21 by native chemical
ligation. Conditions: DNA and peptide (1 mm concentrations), TRIS·HCl
(20 mm), TCEP (20 mm), NaCl (200 mm), pH 8.0 and a) MESNa (1%), 62% or
b) sodium ascorbate (50 mm), 54%. HPLC analyses of reactions performed in
B) TCEP/MESNa or C) TCEP/ascorbate buffer.
radical desulfurization of the cysteinyl peptide and the ligation
product. This reaction was readily detected by HPLC analysis
when reactions involved small peptide fragments. However,
we noticed that careful MS analysis is required to detect desul-
furization of longer peptides.
In the majority of the published reactions, arylthiols such as
thiophenol or 4-mercaptophenylacetic acid (MPAA) are includ-
ed. In fact, these thiols are amongst the most efficient addi-
tives used for native chemical ligation. Our findings suggest
that thiophenol not only accelerates the ligation but also
serves the purpose of preventing TCEP-mediated radical desul-
furization both of starting materials and of ligation products.
The latter role has not been reported previously. It is plausible
to assume that MPAA also forms a stable thiyl radical and so is
able to act as radical scavenger. In our hands, the commonly
used benzyl mercaptan and MESNa also increased the ligation
rates but provided only a partial remedy to the desulfurization
problem.
Figure 3. Synthesis of the WW domain of human Yes-associated protein
(hYAP) by native chemical ligation between the peptide thioester 10 and
the peptide 11 in TCEP-containing buffer A) in the absence or B) in the pres-
ence of sodium ascorbate (50 mm) and in the presence either of C) PhSH
(3 vol%) or of D) MESNa (50 mm). Left: UPLC traces after 24 h. Right: ESI-MS
of: A), B) and C) the ligation product, or D) the starting material. Conditions:
peptides (2 mm concentrations), TCEP (20 mm), Gn·HCl (6m), NaH2PO4
(100 mm), pH 7.5, 258C (* thioesterification of the internal cysteine).
were detected. The desired product 21 was isolated after HPLC
purification in a yield of 54%. This result demonstrates the ver-
satility of the TCEP/ascorbate-promoted native chemical liga-
tion, which can extend beyond the scope of peptide-peptide
conjugation.
We have demonstrated that sodium ascorbate can substi-
tute for the thiol additives used in native chemical ligation
chemistry and shown that sodium ascorbate completely inhib-
its TCEP-mediated desulfurization. It is worth noting that reac-
tions that were performed in TCEP/ascorbate-containing buf-
fers furnished similar rates as reactions in the “conventional”
Native chemical ligation is abundantly used in protein
chemistry. Typically, the ligation reactions are performed in
aqueous denaturating buffer systems containing TCEP as well
as thiol additives. Our results suggest that TCEP can induce
ChemBioChem 2011, 12, 1396 – 1400
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1399

[6] F. Diezmann, H. Eberhard, O. Seitz, Biopolymers 2010, 94, 397–404.
[7] a) J. Offer, P. E. Dawson, Org. Lett. 2000, 2, 23–26; b) J. Offer, C. N. C.
Boddy, P. E. Dawson, J. Am. Chem. Soc. 2002, 124, 4642–4646; c) H.
Hojo, C. Ozawa, H. Katayama, A. Ueki, Y. Nakahara, Y. Nakahara, Angew.
Chem. 2010, 122, 5446–5449; Angew. Chem. Int. Ed. 2010, 49, 5318–
5321.
[8] F. W. Hoffmann, R. J. Ess, T. C. Simmons, R. S. Hanzel, J. Am. Chem. Soc.
1956, 78, 6414–6414.
[9] C. Walling, O. H. Basedow, E. S. Savas, J. Am. Chem. Soc. 1960, 82, 2181–
2184.
TCEP/thiophenol-containing buffer. Furthermore, thiol-free
native chemical ligation might open new conjugation opportu-
nities as exemplified by the synthesis of propargyl-linked pep-
tide-DNA conjugates.[18] Sodium ascorbate is inexpensive,
odourless, nontoxic and highly water-soluble. We propose
sodium ascorbate as a useful alternative to thiol additives in
native chemical ligation reactions.
[10] C. Walling, R. Rabinowitz, J. Am. Chem. Soc. 1957, 79, 5326–5326.
[11] Q. Wan, S. J. Danishefsky, Angew. Chem. 2007, 119, 9408–9412; Angew.
Chem. Int. Ed. 2007, 46, 9248–9252.
Acknowledgements
[12] J. C. D. Mꢁller, E. Graf von Roedern, F. Grams, H. Nagase, L. Moroder,
Biol. Chem. 1997, 378, 1475–1480.
[13] Z. Wang, T. Rejtar, Z. S. Zhou, B. L. Karger, Rapid Commun. Mass Spec-
trom. 2010, 24, 267–275.
This work was supported by the Deutsche Forschungsgemein-
schaft (SFB 765).
[14] a) A. Rauk, D. Yu, D. A. Armstrong, J. Am. Chem. Soc. 1998, 120, 8848–
8855; b) I. S. Lim, J. S. Lim, Y. S. Lee, S. K. Kim, J. Chem. Phys. 2007, 126,
034306–1–034306–10.
Keywords: desulfurization · DNA · ligation · peptides · radical
scavengers
[15] F. Mende, O. Seitz, Angew. Chem. 2007, 119, 4661–4665; Angew. Chem.
Int. Ed. 2007, 46, 4577–4580.
[16] F. Mende, M. Beisswenger, O. Seitz, J. Am. Chem. Soc. 2010, 132, 11110 –
11118.
[17] P. E. Dawson, M. J. Churchill, M. R. Ghadiri, S. B. H. Kent, J. Am. Chem.
Soc. 1997, 119, 4325–4329.
[18] H. Eberhard, F. Diezmann, O. Seitz, Angew. Chem. 2011, 123, 4232-–
4236; Angew. Chem. Int. Ed. 2011, 50, 4146-–4150.
[1] B. L. Nilsson, M. B. Soellner, R. T. Raines, Annu. Rev. Biophys. Biomol.
Struct. 2005, 34, 91–118.
[2] P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. H. Kent, Science 1994, 266,
776–779.
[3] T. Wieland, E. Bokelmann, L. Bauer, H. U. Lang, H. Lau, Justus Liebigs
Ann. Chem. 1953, 583, 129–149.
[4] E. C. B. Johnson, S. B. H. Kent, J. Am. Chem. Soc. 2006, 128, 6640–6646.
[5] a) T. M. Hackeng, J. H. Griffin, P. E. Dawson, Proc. Natl. Acad. Sci. USA
1999, 96, 10068–10073; b) L. Z. Yan, P. E. Dawson, J. Am. Chem. Soc.
2001, 123, 526–533; c) M. Villain, H. Gaertner, P. Botti, Eur. J. Org. Chem.
2003, 3267–3272; d) A. Brik, Y.-Y. Yang, S. Ficht, C.-H. Wong, J. Am.
Chem. Soc. 2006, 128, 5626–5627.
Received: March 17, 2011
Published online on May 6, 2011
1400
www.chembiochem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2011, 12, 1396 – 1400