A cysteine-appended deoxyuridine for the postsynthetic DNA modification using native chemical ligation

Article abstract of DOI:10.1016/j.tetlet.2005.02.018

A new deoxyuridine derivative 6 bearing a cysteine group at the C5 position was synthesized and incorporated into oligodeoxynucleotide (ODN) by phosphoramidite chemistry. The postsynthetic DNA modification of the cysteine-containing ODN using native chemi

Full text of DOI:10.1016/j.tetlet.2005.02.018

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 2235–2238
A cysteine-appended deoxyuridine for the postsynthetic
DNA modification using native chemical ligation
Shuji Takeda, Shinya Tsukiji and Teruyuki Nagamune*
Department of Chemistry and Biotechnology, School of Engineering, The University of Tokyo, Tokyo 113-8656, Japan
Received 12 January 2005; revised 1 February 2005; accepted 3 February 2005
Abstract—A new deoxyuridine derivative 6 bearing a cysteine group at the C5 position was synthesized and incorporated into oligo-
deoxynucleotide (ODN) by phosphoramidite chemistry. The postsynthetic DNA modification of the cysteine-containing ODN using
native chemical ligation with thioesters of biotin and green fluorescent protein variant was successfully demonstrated.
ꢀ
2005 Elsevier Ltd. All rights reserved.
Modified oligonucleotides equipped with diverse func-
tional molecules have numerous applications in struc-
tural biology, biotechnology, materials science, and
medicine. The most common strategy for DNA modifi-
cation involves phosphoramidite chemistry that enables
the direct incorporation of functionalized nucleotides at
defined sites within DNA strands during solid-phase oligo-
nucleotide synthesis. A variety of phosphoramidite
for the postsynthetic DNA modification involve incor-
poration of thiol- or amine-functionalized phosphoro-
0
0
amidites into DNA, at the 5 –/3 -terminus or internal
positions, and its subsequent chemical modification
6
using traditional bioconjugation techniques. These
strategies have found general applicability for attaching
various compounds, ranging from small molecules to
large proteins such as antibodies, to the oligonucleotides
under physiological conditions. However, in prepara-
tion of protein–oligonucleotide conjugates, cross-reac-
tion involving native reactive groups present on
protein surfaces often generates conjugates with high
heterogeneity. Recently, Dey and Sheppard developed
the ketone-modified DNA as a new platform for DNA
1
building blocks bearing, for example, unnatural bases,
2
3
fluorescentprobes, electroactive compounds, and sac-
4
charides, have been synthesized and incorporated into
DNA. While this approach is general and widely used,
many molecules of interest to be incorporated into
DNA oligomers are not stable under the conditions used
in DNA synthesis. Each new modified nucleotide must
have sufficient stability against repetitive oxidation steps
during oligomer synthesis and basic treatment for full
7
modification. Although this approach offers a versatile
and convergent method for the chemoselective modifica-
tion of DNA with various aminooxy compounds by
oxime formation, this method is still limited to the
attachment of relatively small compounds synthetically
accessible, thus not applicable to large proteins.
5
deprotection and cleavage from a solid-phase support.
Thus, there remains a need for efficient methods for
modifying DNA with functional molecules, which is
unsuitable for the phosphoramidite chemistry, such as
peptides and proteins.
We have recently reported a convenient method for con-
jugating synthetic oligonucleotides to recombinant pro-
8
An attractive alternative strategy for DNA modification
is to incorporate a versatile building block bearing a
reactive functional group, which allows conjugation of
any desired molecule by chemical reaction after synthe-
sis of the DNA strand. Several strategies currently used
teins specifically at their C-terminus. Our approach
0
0
involves the modification of 5 - or 3 -amine-functional-
ized oligonucleotides with a cysteine, followed by native
9
chemical ligation (NCL) with a recombinantly ex-
pressed protein containing a C-terminal a-thioester to
form a stable amide bond. A similar approach has been
reported by Gait and coworkers to prepare peptide–
1
0
oligonucleotide conjugates. These successful demon-
strations motivated us to synthesize a new versatile
phosphoramidite building block, which can be incorpo-
rated at any position within DNA strands during DNA
Keywords: Native chemical ligation; Intein; Deoxyuridine; DNA.
*
Corresponding author. Tel.: +81 3 5841 7328; fax: +81 3 5841
657; e-mail: nagamune@bio.t.u-tokyo.ac.jp
8
0
040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.02.018

2236
S. Takeda et al. / Tetrahedron Letters 46(2005) 2235–2238
0
13
synthesis, for an efficient postsynthetic DNA modifica-
tion using NCL.
the 5 -O-DMT-protected nucleoside 5 in 15%. Sub-
0
sequent3 -O-phosphitylation with 2-cyanoethyl diisopro-
pylchlorophosphoramidite yielded phosphoramidite
6 in a quantitative yield.
1
4
On the basis of previous studies, we designed a new
deoxyuridine derivative 6 bearing a protected cysteine
group at ht e C5 posi it on via an allylamine linker. The
C5 position of (deoxy)uridine is an appropriate site to
introduce functionality without interfering with oligonu-
cleotides duplex stability and recognition by RNA and
With compound 6 in hand, a cysteine-containing ODN
(T6dU T6, 7), incorporating the modified base dU
cys
cys
at a central position, was readily prepared by standard
automated DNA synthesis methods. After synthesis,
the ODN was fully deprotected and cleaved from a
solid-phase support by ammonia hydroxide treatment
in good yield and purity. The crude ODN was purified
by a reverse phase cartridge, followed by a reverse phase
1
1
DNA polymerases. N-a-Fmoc-S-tert-butylthio-L-cys-
teine having its amino and thiol group protected with
Fmoc and tert-butylthiol, respectively, was chosen as a
protected cysteine to be introduced. Both protecting
groups are stable under the conditions for DNA synthe-
sis, but smoothly removed during full deprotection and
1
5
HPLC (RP-HPLC). The product was satisfactorily
1
characterized by MALDI-TOF-MS spectrometry.
6
8
NCL steps as reported previously.
To investigate the chemical reactivity of the cysteine-
containing ODN, we attempted NCL between the
ODN 7 and the biotin-thioester 8 (Scheme 2). A reac-
Phosphoramidite 6 was synthesized as shown in Scheme
0
00
17
1. 5-(3 -aminoallyl)-2 -deoxyuridine 3 was prepared in
two steps from 5-iodo-deoxyuridine according to the
tion mixture of ODN 7 and biotin-thioester was incu-
bated at room temperature in a ligation buffer
containing 2-mercaptoethanesulfonic acid (MESNA)
7
procedure reported by Dey and Sheppard. In brief,
the palladium-catalyzed Heck reaction was used to cou-
ple 5-iodo-deoxyuridine to N-allyltrifluoroacetamide to
give 2, which was followed by treatment with 28% aque-
ous ammonia to remove the trifluoroacetamide group,
affording 3. Condensation of 3 with the N-a-Fmoc-S-
tert-butylthio-L-cysteine pentafluorophenyl ester (Nova-
biochem) under standard conditions afforded the cys-
1
5
and was analyzed by RP-HPLC. Time course analysis
revealed nearly quantitative formation of the expected
biotin–ODN conjugate, verified by MALDI-TOF-MS
analysis, within 12 h without significant side reactions
1
8
(Fig. 1).
1
2
teine-appended deoxyuridine 4 in 49%. Nucleoside 4
0
Next we applied our strategy to modify oligonucleotides
with proteins in a site-specific manner. Expressed pro-
tein ligation was performed to ligate the ODN 7 to the
was further treated with 4,4 -dimethoxytrityl chloride
in the presence of 4-(dimethylamino)pyridine to give
Scheme 1. (a) CH
a-Fmoc-S-tert-butylthio-L-cysteine pentafluorophenyl ester, Et
CNCH CH O)PCl, DIEA, CH Cl
2
@CHCH
2
NHCOCF
3
, Na
2
[PdCl
4
], 50% DMF in NaOAc buffer (0.1 M, pH 5.2), 70 ꢁC, 2 h (2, 59%); (b) NH
4
OH (3, 100%); (c) N-
N)
3
N, DMF (4, 59%); (d) DMTCl, DMAP, dry pyridine (5, 15%); (e) (i-Pr
2
(
2
2
2
2
.
Scheme 2. Biotin-thioester 8 (90 equiv), ligation buffer (20 mM Tris–HCl, 500 mM NaCl, 1 mM EDTA, 50 mM MESNA, pH 8.5), rt.

S. Takeda et al. / Tetrahedron Letters 46(2005) 2235–2238
2237
clearly observed, however, the reaction yield was low
approximately 1%) (data not shown). This is probably
(
due to folding of the single strand ODN and thus less
accessibility of the cysteine group, which is a general
issue in handling single strand DNAs. Future study will
be aimed at optimizing the reaction condition for more
efficient chemical ligation, for instance by performing
the reaction in the presence of denaturant or comple-
mentary DNAs.
In conclusion, we have developed a new deoxyuridine
derivative to incorporate a cysteine group at any posi-
tion in oligonucleotides by automated DNA synthesis.
A cysteine-appended DNA should be a general scaffold
for the postsynthetic DNA modification with various
thioester-tagged molecules, ranging from small com-
pounds to recombinant proteins, using NCL.
Figure 1. RP-HPLC analysis of the biotinylation of ODN 7. (A) ODN
(B) ODN 7 after ligation reaction at 4 ꢁC for 72 h. The ODN and
7
ligation product are denoted by • and *, respectively.
Acknowledgements
We are grateful to Professor Kazuhiko Saigo and Dr.
Yasuhiro Ishida for their supports with the NMR mea-
surement, as well as Professors Makoto Komiyama and
Hiroyuki Asanuma for allowing us to use their DNA
synthesizer. Dr. Akinori Kuzuya is acknowledged for
his helpful assistance. We thank Mr. Satoshi Goda for
providing a synthetic intermediate.
C-terminus of an enhanced green fluorescent protein
1
9
(
EGFP) to give a T-shaped protein–DNA conjugate.
An a-thioester form of EGFP was prepared by intein-
fusion technique and was purified by affinity chromato-
8
graphy over 95% homogeneity, as previously reported.
Progress of NCL between the EGFP-thioester and the
ODN 7 was monitored by SDS-PAGE under a reducing
condition. As shown in Figure 2, the ligation proceeded
to approximately 80% completion after incubation at
References and notes
4
ꢁC for two days. In contrast, a control reaction of
Cys
the EGFP-thioester with an ODN lacking the dU
T ) under the same condition resulted in no change
1
. (a) Hirao, I.; Ohtsuki, T.; Fujiwara, T.; Mitsui, T.;
Yokogawa, T.; Okuni, T.; Nakashima, H.; Takio, K.;
Yabuki, T.; Kigawa, T.; Kodama, K.; Yokogawa, T.;
Nishikawa, K.; Yokoyama, S. Nat. Biotechnol. 2002, 20,
(
1
2
of the migrating band as expected (data not shown).
Nextwe prepared randomized ODN 10
CGTTATCCCdU GATTCTGTGG) as described
a
Cys
(
1
77; (b) Tanaka, K.; Tengeiji, A.; Kato, T.; Toyama, N.;
2
0
above, and performed the NCL with the EGFP-thio-
ester. Formation of the EGFP-ODN 10 conjugate was
Shionoya, M. Science 2003, 299, 1212; (c) Meggers, E.;
Holland, P. L.; Tolman, W. B.; Romesberg, F. E.; Schultz,
P. G. J. Am. Chem. Soc. 2000, 122, 10714.
2
3
. (a) Tyagi, S.; Kramer, F. R. Nat. Biotechnol. 1996, 4, 303;
b) Tong, A. K.; Li, Z.; Jones, G. S.; Russo, J. J.; Ju,
J. Nat. Biotechnol. 2001, 19, 756; (c) Heyduk, T.; Heyduk,
E. Nat. Biotechnol. 2002, 20, 171.
. (a) Anne, A.; Blanc, B.; Moiroux, J. Bioconjugate Chem.
(
2
001, 12, 396; (b) Yu, C. J.; Wang, H.; Wan, Y.; Yowanto,
H.; Kim, J. C.; Donilon, L. H.; Tao, C.; Strong, M.;
Chong, Y. J. Org. Chem. 2001, 66, 2937; (c) Navarro,
A.-E.; Spinelli, N.; Moustrou, C.; Chaix, C.; Mandrand,
B.; Brisset, H. Nucleic Acids Res. 2004, 32, 5310; (d) Mori,
K.; Subasinghe, C.; Cohen, J. S. FEBS Lett. 1989, 249,
2
13; (e) Whittemore, N. A.; Mullenix, A. N.; Inamati, G.
B.; Manoharan, M.; Cook, P. D.; Tuinman, A. A.; Baker,
D. C.; Chamber, J. Q. Bioconjugate Chem. 1999, 10, 261.
. (a) Sheppard, T. L.; Wong, C.-H.; Joyce, G. F. Angew.
Chem., Int. Ed. 2000, 39, 3662; (b) Matsuura, K.; Hibino,
M.; Ikeda, T.; Yamada, Y.; Kobayashi, K. Chem. Eur. J.
4
5
2
004, 10, 352; (c) Maier, M. A.; Yannopoulos, C. G.;
Mohamed, N.; Roland, A.; Fritz, H.; Mohan, V.; Just, G.;
Manoharan, M. Bioconjugate Chem. 2003, 14, 18.
. Caruthers, M. H.; Barone, A. D.; Beaucage, S. L.; Dodds,
D. R.; Fisher, E. F.; McBride, L. J.; Matteucci, M.;
Stabinsky, Z.; Tang, J.-Y. Methods Enzymol. 1987, 154,
287.
Figure 2. SDS-PAGE analysis. Lane 1, molecular weightmarker; 2,
EGFP-thioester; 3, EGFP-thioester mixed with ODN 7. The EGFP-
thioester and ligation product are denoted by • and *, respectively.
6. (a) MacMillan, A. M.; Verdine, G. L. Tetrahedron 1991,
47, 2603; (b) Shinozuka, K.; Kohgo, S.; Ozaki, H.; Sawai,

2238
S. Takeda et al. / Tetrahedron Letters 46(2005) 2235–2238
H. Chem. Commun. 2000, 59; (c) de la Torre, B. G.;
1H), 3.77 (s, 6H), 3.58 (m, 2H), 3.43 (dd, 1H), 3.34 (dd,
1H), 3.03 (m, 2H), 2.54 (m, 1H), 2.31 (m, 2H) 1.30 (s, 9H):
Morales, J. C.; Avi nˇ o´ , A.; Iacopino, D.; Ongaro, A.;
Fitzmaurice, D.; Murphy, D.; Doyle, H.; Redmond, G.;
Eritja, R. Helv. Chim. Acta 2002, 85, 2594.
. Dey, S.; Sheppard, T. L. Org. Lett. 2001, 3, 3983.
. Takeda, S.; Tsukiji, S.; Nagamune, T. Bioorg. Med. Chem.
Lett. 2004, 14, 2407.
MALDI-TOF-MS (Matrix: dithranol) calcd for
[M+Na] = 1021.35, obsd. 1021.67.
5
+
3
14. The phosphoramidites 6 was identified by P NMR, and
2
7
8
3
2
MALDI-TOF-MS. 6:
MALDI-TOF-MS (Matrix: dithranol) calcd for
P NMR d 149.56, 149.96:
6
+
[M+Na] = 1221.46, obsd. 1221.03.
9
. Dawson, P. E.; Kent, S. B. H. Annu. Rev. Biochem. 2000,
6
9, 923.
15. HPLC was carried outon a lBondasphere C18 column
(5 lm, 150 · 3.8 mm, Waters) eluted with 0.1 M ammo-
nium acetate buffer (pH 7.0) containing 10–20% acetoni-
trile in a linear gradient over 30 min at a flow rate of
0.5 mL/min, detected at 260 nm.
1
1
0. Stetsenko, D. A.; Gait, M. J. J. Org. Chem. 2000, 65, 4900.
1. (a) Ahmadian, M.; Zhang, P.; Bergstrom, D. E. Nucleic
Acids Res. 1998, 26, 3127; (b) Lee, S. E.; Sidorov, A.;
Gourlain, T.; Mignet, N.; Thorpe, S. J.; Brazier, J. A.;
Dickman, M. J.; Hornby, D. P.; Grasby, J. A.; Williams,
D. M. Nucleic Acids Res. 2001, 29, 1565; (c) Latham, J. A.;
Johnson, R.; Toole, J. J. Nucleic Acids Res. 1994, 22, 2817;
16. ODN 7 was identified by MALDI-TOF-MS (Matrix: 3-
hydroxypicolinic acid) ODN 7 calcd for [M+H] =
4141.43, obsd 4137.74.
+
(
d) Ahmadian, M.; Zhang, P.; Bergstrom, D. E. Nucleic
17. Biotin-thioester 8 was synthesized according to the
following paper: Tolbert, T. J.; Wong, C.-H. Angew.
Chem., Int. Ed. 2002, 41, 2171.
Acids Res. 1998, 26, 3127; (e) Sawai, H.; Ozaki, A. N.;
Satoh, F.; Ohbayashi, T.; Masud, M. M.; Ozaki, H. Chem.
Commun. 2001, 2604.
18. Biotinylation reaction was carried out in a total volume of
10 lL of ligation buffer (20 mM Tris–HCl, 500 mM NaCl,
1 mM EDTA, 50 mM MESNA, pH 8.5). Final concen-
trations of biotin-thioester 8 and ODN 7 were 9 mM and
100 lM, respectively. The ligation products were charac-
terized by MALDI-TOF-MS (Matrix: 3-hydroxypicolinic
1
1
2. The nucleoside 4 was satisfactorily characterized by
H
1
NMR: H NMR (d-DMSO) d 11.41 (s, 1H), 8.24 (s, 1H),
7
7
(
3
3
.98 (s, 1H), 7.89 (d, 2H), 7.68–7.75 (m, 2H), 7.42 (t, 2H),
.32 (t, 2H), 6.40–6.45 (m, 1H), 6.13–6.18 (m, 2H), 5.23
m, 1H), 5.05 (t, 1H), 4.34 (m, 1H), 4.22–4.30 (m, 5H),
+
.76–3.79 (m, 2H), 3.58–3.61 (m, 2H), 3.32 (m, 1H), 2.91–
.09 (m, 2H), 2.11 (t, 2H), 1.30 (s, 9H).
acid). ODN 9 calcd for [M+H] = 4261.76, obsd 4265.80.
19. Expressed protein ligation was carried out in a total
1
1
3. The nucleoside 5 was characterized by H NMR and
volume of 10 lL of ligation buffer. Final concentrations of
1
MALDI-TOF-MS: H NMR (CDCl ) d 9.82 (s, 1H),
7
EGFP-thioester and ODN
respectively.
7 were 9 and 100 lM,
3
.65–7.74 (m, 2H+1H), 7.58 (m, 2H), 7.34–7.39 (m, 2H),
7
.23–7.29 (m, 9H+2H), 6.82 (d, 4H), 6.73 (m, 1H), 6.47
20. ODN 10 was identified by MALDI-TOF-MS (Matrix: 3-
hydroxypicolinic acid) ODN 10 calcd for [M+H] =
6617.11, obsd 6614.46.
+
(
(
m, 1H), 6.22–6.27 (m, 1H), 5.90 (d, 1H), 5.54 (d, 1H), 4.54
m, 1H), 4.34–4.39 (m, 2H+1H), 4.20 (m, 1H), 4.10 (m,