A direct method for site-specific protein acetylation

Article abstract of DOI:10.1002/anie.201103754

Radicals at work: Radical-mediated thiol-ene addition of the thiol group of Cys to N-vinylacetamide gives acetyl-thialysine (K_SAc), a near-perfect mimic of acetyl-lysine (see picture). The reaction is highly efficient with near quantitative con

Full text of DOI:10.1002/anie.201103754

DOI: 10.1002/anie.201103754
Protein Acetylation
A Direct Method for Site-Specific Protein Acetylation**
Fupeng Li, Abdollah Allahverdi, Renliang Yang, Gavian Bing Jia Lua, Xiaohong Zhang,
Yuan Cao, Nikolay Korolev, Lars Nordenskiçld, and Chuan-Fa Liu*
Post-translational modification (PTM) is a fundamental
mechanism for modulating protein function. One such PTM
with increasingly recognized significance is protein lysine
acetylation.[1] A reversible biochemical process,[2] lysine
acetylation was initially discovered in histones.[3] Recently it
has also been observed in a very large number of other
proteins,[4] thus suggesting its diverse regulatory functions in
the cell.[5] There is mounting evidence that aberrant lysine
acetylation is implicated in many disease conditions such as
cancer and neurological disorders.[6] Therefore, the study of
lysine acetylation biology is of great importance and will lead
to continued therapeutic innovations.[7,8]
Although lysine acetylation has long been recognized as a
histone epigenetic mark affecting chromatin structure and
function,[2] the exact effects of most individual protein
acetylation events, especially those identified more recently
remain to be elucidated. A major difficulty in the study of
lysine acetylation biology lies in the limited availability of
homogeneous protein samples that contain the acetylated
lysine residue(s) of interest. Such materials would be invalu-
able reagents for discerning the structural and functional
effects of a particular Lys acetylation PTM by biophysical and
biochemical means.[9] Several methods can be used to prepare
site-specifically acetylated proteins, such as unnatural amino-
acid mutagenesis using the amber stop codon/suppressor
tRNA system[10] and protein chemical synthesis.[11] While the
stop codon suppression strategy is a powerful method, it is
currently not widely available. And significant technical
barriers exist for the adoption of chemical synthesis methods
by the large bioscience community. Another method com-
bines unnatural amino acid mutagenesis with chemical
modification to introduce an acetyl lysine analogue into a
protein; however, the chiral integrity of the modified amino
acid is compromised in the process.[12] Direct enzymatic Lys
acetylation is unrealistic given the often promiscuous as well
as inefficient and incomplete nature of such enzymatic
reactions. Chemical acetylation of a selected lysine among
many Lys residues in a protein is also obviously not feasible.
The unique reactivity of the thiol group of cysteine as a soft
nucleophile has been exploited extensively for selective
protein modification.[13] For instance, the classic reaction of
aminoethylation of Cys has long been used for converting a
cysteine residue to 4-thialysine as a functional equivalent to
lysine.[14a] This method was also extended recently to the
preparation of Ne-methyl-lysine analogues by using N-
methylaminoethyl halides as the alkylating agents.[14b] The
efficiency of this reaction is attributed to the formation of a
highly reactive aziridinium intermediate.[14c] Unfortunately,
this reaction system does not work for N-acetyl-thialysine as
seen in the failed attempts by others[15] and us to use N-acetyl-
aziridine and N-acetyl-aminoethyl bromide or iodide for
cysteine alkylation. More recently it was reported that the use
of methylthiocarbonyl-aziridine led to selective Cys alkyl-
ation.[15] The resultant methylthiocarbonyl-thialysine was
shown to mimic Ne-acetyl-lysine in certain functions,[15]
although the methylthio-carbamate moiety is electrosterically
rather different from the acetamide in Ne-acetyl-lysine.
Clearly, analogously to 4-thialysine and N-methyl-thialysine
being ideal mimics of lysine and Ne-methyl-lysine respec-
tively, N-acetyl-thialysine [sLys(Ac)] would also be an ideal
mimic of Lys(Ac), in which the only difference is the isosteric
thioether in lieu of the 4-methylene in natural Lys(Ac). As the
position of this substitution is rather far away—by 2 carbon
atoms—from the acetamide nitrogen, little difference is
expected between this Lys(Ac) mimic and native Lys(Ac) in
their exhibited physicochemical and biochemical properties.
However, since existing methods for cysteine modification
are not applicable here, a new method must be discovered to
obtain such a Lys(Ac) mimic.
In searching for ways of introducing an sLys(Ac) residue
into proteins, we came across a radical reaction known as
thiol-ene addition,[16] which might serve our needs. A classic
reaction discovered over a century ago,[17] radical thiol-ene
reaction gives an anti-Markovnikov addition thioether pro-
duct.[16a] Over the years this reaction has found extensive use
in polymer chemistry.[16] More recently, it has also emerged as
a useful click reaction for bioorganic functionalization.[18,19]
We realized that thiol-ene coupling between the cysteine thiol
and N-vinyl-acetamide (NVA) would directly generate the
desired acetyl-thialysine (Scheme 1). We first used a small
organic thiol compound, benzyl mercaptan (BzSH), as the
substrate and examined the alkylation reaction under differ-
ent conditions. We found that the free radical reaction
proceeded well in acetate buffer at pH 4 and in the presence
of VA-044 as the initiator under UV irradiation at 365 nm. At
a 1:1 ratio of NVA to BzSH and at low concentrations of the
two reactants (at 5 or 10 mm), a 30 min reaction gave over
70% conversion of BzSH with the expected thiol acetamido-
ethylation product. The reaction at 10 mm also produced a
[*] F. Li, A. Allahverdi, R. Yang, G. B. J. Lua, Dr. X. Zhang, Y. Cao,
Dr. N. Korolev, Prof. L. Nordenskiçld, Prof. C. F. Liu
Divisions of Chemical Biology and Biotechnology and of Structural
and Computational Biology, School of Biological Sciences, Nanyang
Technological University, Singapore 637551 (Singapore)
E-mail: cfliu@ntu.edu.sg
[**] This work is supported by grants from A*Star (C.F.L.) and Ministry
of Education (L.N.) of Singapore.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201103754.
Angew. Chem. Int. Ed. 2011, 50, 9611 –9614
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9611

Communications
the rate-limiting step of this thiol-ene coupling reaction[16a]
—
to effectively quench the carbon free radical intermediate
formed at the addition step of the thiyl radical to the NVA
ethylene double bond and to prevent it from reacting with
another molecule of NVA. As expected, glutathione was also
alkylated by NVA in the reaction. In fact, when peptide 4
(5 mm) was alkylated deliberately in the absence of gluta-
thione under otherwise identical conditions, products in
which the Cys residue was alkylated by di-, tri-, and tetrameric
NVAs were observed in addition to the desired product (see
the Supporting Information). No reaction occurred on a
control peptide with no Cys residue (Ac-FQPKSG); equally,
no reaction occurred in the absence of the initiator or without
UV irradiation. A weakly acidic pH and the presence of
glutathione also prevented oxidative disulfide formation of
the thiol in the substrates. Use of the reducing agent tris(2-
carboxyethyl)phosphine (TCEP) was detrimental to the
reaction as it led to desulfurization of the peptide substrate.
A noteworthy example to show the high efficiency of this
reaction is the alkylation of peptide 5, which contains four Cys
residues. Remarkably, when this peptide (1.25 mm) was
treated in the reaction mixture for 1 h, a very clean tetra-
alkylating reaction gave the desired product in 95% yield
(Figure 1, top).
Scheme 1. Radical-mediated addition of Cys thiol to N-vinyl-acetamide
(NVA) generates sLys(Ac), a close mimic of native Lys(Ac).
minor side product, which was due to telomerization,[16a] that
is, polymerization (in this case dimerization only) of NVA on
the thiol. When NVAwas used in 5-, 10- or even 20-fold excess
with respect to BzSH, the reaction was complete within
30 min but more telomerization products were formed (see
the Supporting Information). However, even under these
conditions, the desired product still formed in a predominant
amount. These encouraging results led us to further test this
reaction on synthetic peptides. We reasoned that the inclusion
of a second thiol compound in the reaction mixture might
help to suppress the side reaction of radical telomerization.
After numerous trials we found that glutathione (reduced)
could perfectly serve this purpose and that a reaction mixture
of 15 mm glutathione, 50 mm NVA, and 5 mm VA-044 in 0.2m
acetate buffer (pH 4) or 0.2m phosphate buffer (pH 6 or 7)
was well suited for the alkylation of peptide and protein
substrates. Using these conditions, the reaction on the
synthetic peptides was basically completed in 30 min and no
significant side reactions were found when extending the
reaction time to 1 h. For instance, when each of the peptides
1–4 (Table 1) was treated with this mixture in acetate buffer
(pH 4, final peptide concentration = 5 mm), the desired
alkylating product was obtained in near quantitative yield
based on HPLC and MS analysis. Under these conditions, no
or little telomerization products were detected. Therefore,
glutathione together with the peptide substrate participated
in the critical radical chain transfer step—which is known as
This reaction was also highly effective on protein sub-
strates. Firstly, a ubiquitin mutant that contains a Cys at
position 48 was prepared and subjected to this modification.
The protein (0.5 mm) was used in its native folded state for
Table 1: Peptide and protein substrates alkylated with NVA.[a]
N
Substrate
Yield[b] [%]
1
2
3
4
5
6
7
8
Ac-FQPKCG
VGCAEKSL
WACYKSL
Biotin-GKGGACRHRKVLRDN
Biotin-GCGGCGLGCGGACR
Ubiquitin K48C
Histone H4 K16C
Histone H3 K27C
>95
>95
95
>95
95
>95
90
90
Figure 1. Top: HPLC and MS analyses of the alkylation reaction on
peptide 5. A) peptide 5 (peak a) before alkyation. B) after alkylation.
Peak b is the alkylation product; peak c appears because of the
oxidation of one of the thioether linkages in the alkylated product.
C) MALDI-MS spectrum of the isolated product corresponding to peak
b in (B) (m/z [M+H]+ found: 1735.9, calcd: 1735.7). Bottom: HPLC
and MS analyses of histone H4 alkylation. D) HPLC profile of the H4
alkyation reaction at pH 4. Peak a’ is the alkylation product; peak b’
presumably represents product from oxidation of peak a’. E) ESI-MS of
the desired alkylation product with the insert showing the deconvo-
luted mass (MW found: 11295.5, MW calcd: 11296.2).
[a] Reaction conditions: Peptide or protein substrate (5 mm for peptides
1–4, 1.25 mm for peptide 5, 0.5 mm for ubiquitin and 1 mm for H3 and
H4) was treated in a reaction mixture containing 50 mm NVA, 5 mm VA-
044 and 15 mm glutathione in 0.2m acetate buffer at pH 4 (for H3 and
H4 the mixture also contained 6n Gdn·HCl) under UV (365 nm)
irradiation.[b] Yields based on HPLC and MALDI-MS analysis (reaction
time: 1 h for 1, 2, 4, and 5, 30 min for 3 and 2 h for the proteins 6–8). See
the Supporting Information for experimental details. Peptide 1, ubiquitin
(entry 6), and H4 (entry 7) were also reacted at pH 6 or 7 with similar
results obtained (see the Supporting Information).
9612
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9611 –9614

alkylation in the same reaction mixture at pH 4 or 7 (see the
Supporting Information). MS analysis clearly showed an
almost quantitative conversion of the Cys residue to sLys(Ac)
in the course of 2 h with the expected + 85 Da MW for the
alkylated product. The protein remained soluble during the
reaction, hence suggesting that no denaturing was occurring
and that the presence of 50 mm NVA and 5 mm VA-044 did
not affect the structure of the folded protein. It is worth
noting that it would be difficult to use a semisynthetic method
to prepare such a modified protein, because the modification
site is in the middle of the sequence. Two other proteins,
histone H4K16C and H3K27C, were also modified with
excellent results. In these cases, 6m Gdn·HCl was included in
the alkylation reaction mixture. Interestingly, for the alkyla-
tion of H4K16C at pH 4 or 7, in addition to the desired
product, a side product was formed in significant amount (see
the Supporting Information). The side product was more
hydrophilic with a molecular weight (MW) that was 16 Da
higher than that of the expected acetyl-thialysine product.
This side product appeared to result from oxidation of the
thioether linkage to sulfoxide and the inclusion of dimethyl-
sulfide in the reaction mixture minimized the formation of
this side product to about 5% (Figure 1, bottom and the
Supporting Information). No alkylation was detected when
wild-type H4 was subjected to the same treatment. From the
above results we can see that this free radical thiol reaction
can tolerate various reaction conditions, for example, native
or denatured buffers, to modify a protein.
The generated sLys(Ac) was shown to be a good func-
tional mimic of the natural Lys(Ac). Firstly, histone protein
H4 KS16Ac was recognized by a specific anti-H4K16Ac
antibody, whereas the unmodified H4K16C was not recog-
nized by the same antibody (Figure 2A). Next, an enzymatic
test was conducted to investigate whether the acetyl-lysine
mimic could be recognized by a histone deacetylase and be
substrate for deacetylation. SIRT2, a class III NAD-depen-
dent deacetylase, was used for the deacetylation reaction of
the alkylated peptide 1 and its native counterpart Ac-
FQPKK(Ac)G. Using an HPLC assay (Figure 2B), we
showed that the sLys(Ac) residue in the alkylated peptide 1
was susceptible to enzymatic deacetylation, albeit to a lesser
degree than its native counterpart (see the Supporting
Information), which might be due to the fact that the side
chain of sLys is slightly longer than that of native Lys. The
recently reported methylthiocarbamate modification, on the
other hand, was resistant to enzymatic deacetylation.[15]
Acetylation of Lys16 in histone H4 is known to inhibit the
folding of nucleosome arrays and hence the formation of the
compact 30 nm chromatin fibre.[9,20] H4 KS16Ac and three
other control H4 proteins (H4 K16Ac, H4 KS16, and H4-WT)
were incorporated respectively, together with H3, H2A, and
H2B into histone octamers. H4 K16Ac was prepared using a
semisynthetic approach[20] and H4 KS16 was synthesized from
alkylating H4 C16 with 2-bromoethylamine. The four differ-
ent octamers were then individually combined with the 12-
177-601 DNA to assemble into the 12-nucleosome arrays.
Using analytical ultracentrifugation (AUC), we demonstrated
that the KS16Ac of interest produced an identical effect as the
native K16Ac in abolishing Mg2+-induced folding of the
Figure 2. A) Western blots with anti-H4 K16Ac Ab on the H4 proteins:
H4 K16C and H4 KS16Ac; B) deacetylation assay of Ac-FQPKKs(Ac)G
by SIRT2 after 16 h; C) effects of H4 K16 acetylation on nucleosome
array folding as seen from sedimentation distributions of the nucleo-
some arrays before (no Mg2+, open symbols) and after Mg2+-induced
folding (with 1.0 mm MgCl2, solid symbols). See the Supporting
Information for details.
reconstituted nucleosome array (Figure 2C). The AUC data
clearly showed that, in the presence of 1 mm MgCl2, the
nucleosome array containing wild-type H4 or its equivalent
H4 KS16 folded into a significantly more compact state
(sedimentation coefficient S208C,w = 52–53S) than did the
K16Ac and KS16Ac arrays (S208C,w = 44–45S). Remarkably,
these results prove not only the functional equivalency
between sLys(Ac) and Lys(Ac) but also that between sLys
and Lys.
To summarize, a previously unexplored thiol-ene radical
addition reaction involving the commercially available NVA
is well suited for the S-acetamidoethylation of cysteine
residues in synthetic peptides and recombinant proteins.
The resultant N-acetyl-thialysine differs from natural acetyl-
lysine only isosterically at the g position of the amino acid
structure and is functionally equivalent or similar to the latter.
Although a limitation of the method is that the protein
substrates should not contain other cysteines, it nevertheless
has many potential applications, such as the histone epige-
netic study—an intense research area at present. The reaction
system is robust and gives near quantitative yields of site-
specifically acetylated proteins that can be purified in a simple
chromatography or dialysis step. The ease of implementation
of this method also makes it easily adoptable for researchers
from the bioscience research community. As such, this radical
reaction approach provides a convenient enabling tool for the
study of lysine acetylation biology and will help to advance
research in this important field.
Received: June 1, 2011
Revised: August 2, 2011
Published online: September 16, 2011
Angew. Chem. Int. Ed. 2011, 50, 9611 –9614
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9613

Communications
[8] T. K. Kelly, D. D. De Carvalho, P. A. Jones, Nat. Biotechnol.
2010, 28, 1069 – 1078.
[9] M. Shogren-Knaak, H. Ishii, J.-M. Sun, M. J. Pazin, J. R. Davie,
C. L. Peterson, Science 2006, 311, 844 – 847.
[10] H. Neumann, S. Peak-Chew, J. W. Chin, Nat. Chem. Biol. 2008, 4,
232 – 234.
Keywords: epigenetics · posttranslational modifications ·
protein modifications · radical reactions · thiol–ene reactions
.
[1] For reviews, see: a) T. Kouzarides, EMBO J. 2000, 19, 1176 –
1179; b) K. L. Norris, J.-Y. Lee, T.-P. Yao, Sci. Signaling 2009, 2,
pe76.
[2] a) M. D. Shahbazian, M. Grunstein, Annu. Rev. Biochem. 2007,
76, 75 – 100; b) M.-H. Kuo, C. D. Allis, BioEssays 1998, 20, 615 –
626.
[3] a) V. G. Allfrey, R. Faulkner, A. E. Mirsky, Proc. Natl. Acad. Sci.
USA 1964, 51, 786 – 794; b) E. L. Gershey, G. Vidali, V. G.
Allfrey, J. Biol. Chem. 1968, 243, 5018 – 5022.
[11] a) S. B. H. Kent, Chem. Soc. Rev. 2009, 38, 338; b) C. Chatterjee,
T. W. Muir, J. Biol. Chem. 2010, 285, 11045 – 11050.
[12] J. Guo, J. Wang, J. S. Lee, P. G. Schultz, Angew. Chem. 2008, 120,
6499 – 6501; Angew. Chem. Int. Ed. 2008, 47, 6399 – 6401.
[13] J. M. Chalker, G. J. L. Bernardes, Y. A. Lin, B. G. Davis, Chem.
Asian J. 2009, 4, 630 – 640.
[14] a) M. A. Raftery, R. D. Cole, J. Biol. Chem. 1966, 241, 3457 –
3461; b) M. D. Simon, F. Chu, L. R. Racki, C. C. de La Cruz,
A. L. Burlingame, B. Panning, G. J. Narlikar, K. M. Shokat, Cell
2007, 128, 1003 – 1012; c) C. E. Hopkins, G. Hernandez, J. P. Lee,
D. R. Tolan, Arch. Biochem. Biophys. 2005, 443, 1 – 10.
[15] R. Huang, M. A. Holbert, M. K. Tarrant, S. Curtet, D. R.
Colquhoun, B. M. Dancy, B. C. Dancy, Y. Hwang, Y. Tang, K.
Meeth, R. Marmorstein, R. N. Cole, S. Khochbin, P. A. Cole, J.
Am. Chem. Soc. 2010, 132, 9986 – 9987.
[16] For reviews, see a) K. Griesbaum, Angew. Chem. 1970, 82, 276 –
290; Angew. Chem. Int. Ed. Engl. 1970, 9, 273 – 287; b) C. E.
Hoyle, T. Y. Lee, T. Roper, J. Polym. Sci. Part A 2004, 42, 5301 –
5337.
[17] T. Posner, Ber. Dtsch. Chem. Ges. 1905, 38, 646.
[18] a) A. Dondoni, Angew. Chem. 2008, 120, 9133 – 9135; Angew.
Chem. Int. Ed. 2008, 47, 8995 – 8997; b) C. E. Hoyle, C. N.
Bowman, Angew. Chem. 2010, 122, 1584 – 1617; Angew. Chem.
Int. Ed. 2010, 49, 1540 – 1573; c) F. Floyd, B. Vijayakrishnan, J. R.
Koeppe, B. G. Davis, Angew. Chem. 2009, 121, 7938 – 7942;
Angew. Chem. Int. Ed. 2009, 48, 7798 – 7802.
[4] a) S. C. Kim, R. Sprung, Y. Chen, Y. Xu, H. Ball, J. Pei, T. Cheng,
Y. Kho, H. Xiao, L. Xiao, N. V. Grishin, M. White, X. J. Yang, Y.
Zhao, Mol. Cell 2006, 23, 607 – 618; b) J. Zhang, R. Sprung, J. Pei,
X.-H. Tan, S. Kim, H. Zhu, C.-F. Liu, N. V. Grishin, Y. Zhao,
Mol. Cell. Proteomics 2009, 8, 215 – 225; c) C. Choudhary, C.
Kumar, F. Gnad, M. L. Nielsen, M. Rehman, T. C. Walther, J. V.
Olsen, M. Mann, Science 2009, 325, 834 – 840; d) S. Zhao, W. Xu,
W. Jiang, W. Yu, Y. Lin, T. Zhang, J. Yao, L. Zhou, Y. Zeng, H.
Li, Y. Li, J. Shi, W. An, S. M. Hancock, F. He, L. Qin, J. Chin, P.
Yang, X. Chen, Q. Lei, Y. Xiong, K. L. Guan, Science 2010, 327,
1000 – 1004; e) Q. Wang, Y. Zhang, C. Yang, H. Xiong, Y. Lin, J.
Yao, H. Li, L. Xie, W. Zhao, Y. Yao, Z.-B. Ning, R. Zeng, Y.
Xiong, K.-L. Guan, S. Zhao, G.-P. Zhao, Science 2010, 327, 1004 –
1007.
[5] a) S. Spange, T. Wagner, T. Heinzel, O. H. Krꢀmer, Int. J.
Biochem. Cell Biol. 2009, 41, 185 – 198; b) K.-L. Guan, Y. Xiong,
Trends Biochem. Sci. 2011, 36, 108 – 116.
[6] a) S. B. Hake, A. Xiao, C. D. Allis, Br. J. Cancer 2004, 90, 761 –
769; b) X. L. Yang, Nucleic Acids Res. 2004, 32, 959 – 976; c) S.
Peleg, F. Sananbenesi, A. Zovoilis, S. Burkhardt, S. Bahari-
Javan, R. C. Agis-Balboa, P. Cota, J. L. Wittnam, A. Gogol-
Doering, L. Opitz, G. Salinas-Riester, M. Dettenhofer, H. Kang,
L. Farinelli, W. Chen, A. Fischer, Science 2010, 328, 753 – 756.
[7] P. A. Cole, Nat. Chem. Biol. 2008, 4, 590 – 597.
[19] Y.-X. Chen, G. Triola, H. Waldmann, Acc. Chem. Res. 2011,
DOI: 10.1021/ar200046h.
[20] A. Allahverdi, R. Yang, N. Korolev, Y. Fan, C. A. Davey, C.-F.
Liu, L. Nordenskiçld, Nucleic Acids Res. 2011, 39, 1680 – 1691.
9614 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9611 –9614