Postsynthetic modification of DNA via threoninol on a solid support by means of allylic protection

Article abstract of DOI:10.1016/j.tetlet.2008.06.090

We have developed a facile but versatile method to introduce functional molecules into DNA on a CPG support. Threoninol, whose amino group was protected with an (allyloxy)carbonyl (Alloc) group, was introduced into DNA via the corresponding phosphoramidite monomer. After selective deprotection of the Alloc group by treatment with palladium(0), a dye with a carboxyl group could be introduced into the DNA on the CPG support through an amide bond.

Full text of DOI:10.1016/j.tetlet.2008.06.090

Tetrahedron Letters 49 (2008) 5144–5146
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Postsynthetic modification of DNA via threoninol on a solid support
by means of allylic protection
Hiroyuki Asanuma a,b, , , Yuichi Hara a, , Akira Noguchi a, Kanae Sano a, Hiromu Kashida a
*
a Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8603, Japan
b Core Research for Evolution Science and Technology (CREST), Japan Science and Technology Agency (JST), Kawaguchi, Saitama 332-0012, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We have developed a facile but versatile method to introduce functional molecules into DNA on a CPG
support. Threoninol, whose amino group was protected with an (allyloxy)carbonyl (Alloc) group, was
introduced into DNA via the corresponding phosphoramidite monomer. After selective deprotection of
the Alloc group by treatment with palladium(0), a dye with a carboxyl group could be introduced into
the DNA on the CPG support through an amide bond.
Received 13 May 2008
Revised 17 June 2008
Accepted 20 June 2008
Available online 25 June 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Recent advances in phosphoramidite chemistry have enabled a
wide variety of chemical modifications of oligonucleotides.1 The
growing field of DNA/RNA research requires further advances in
synthetic chemistry for DNA modification. Postsynthetic modifica-
tion methods on solid supports have been developed to facilitate
the functionalization of DNA:1c oligonucleotides involving specifi-
cally masked functional groups or a designed convertible unit
can be synthesized on an automated DNA synthesizer, followed
by introduction of functional molecules on the solid support.2,3
Postsynthetic modification is effective for the introduction of func-
tional molecules that are hard to convert to phosphoramidite
monomers. Furthermore, it should be also very effective in diversi-
fying the functionality of modified DNA, since we do not have to
synthesize each phosphoramidite precursor.
Seitz et al. reported on the solid-phase modification of PNA with
Alloc protection.8 Here, we report a facile and versatile method
to introduce functional molecules into oligonucleotides through
amide bonds on a solid phase via an Alloc-protected phosphorami-
dite sub-monomer as illustrated in Schemes 1 and 2: the sub-
monomer 3 is incorporated into DNA through the automated
DNA synthesizer. The obtained controlled pore glass (CPG) support
anchoring fully protected DNA is treated with palladium(0) to
remove the Alloc group selectively, followed by the coupling with
a functional molecule through an amide bond. The desired DNA is
finally obtained by deprotection with aqueous ammonia solution.
By this method, we can easily diversify the functionality of DNA.
Furthermore, functional molecules that are difficult to convert to
Recently, we found that D-threoninol is an excellent internucle-
osidic linker (base surrogate) for the introduction of functional
molecules.4 By tethering azobenzene or pyrene we successfully
synthesized photo-responsive oligonucleotides, fluorescent probes,
and molecular clusters.4,5 These modified DNAs were synthesized
via the corresponding phosphoramidite monomers. Our previous
study on the optimization of azobenzene required the syntheses
of as many as eight different monomers of azobenzene derivatives,
which was highly time-consuming.6 Although a postsynthetic ap-
proach is very effective in avoiding such problems, little research
has been reported on facile methods for introducing functional
molecules into oligonucleotides via amino groups on solid support,
except for the photo-labile strategy.2
DMT
OH
OH
O
H
N
H
N
ii
i
O
NH2
OH
O
O
O
OH
OH
1
2
DMT
O
H
N
O
iii
O
O
P
NC
O
N
Allyl and (allyloxy)carbonyl (Alloc) groups are useful protectors
for hydroxyl, carboxyl, and amino groups, and can be removed by a
palladium(0)-catalyzed reaction under mild conditions.7 Recently,
3
Scheme 1. Synthesis of Alloc-protected phosphoramidite monomer. Reagents and
conditions: (i) allyl chloroformate, triethylamine, THF, 0 °C, 15 min, then rt, 1.5 h,
97%; (ii) dimethoxytrityl chloride, DIPEA, pyridine, CH2Cl2, rt, 4.5 h, 79%; (iii)
(iPr2N)2PO(CH2)2CN, 1H-tetrazole, acetonitrile, rt, 1.5 h, >99%.
* Corresponding author. Tel.: +81 52 789 2488; fax: +81 52 789 2528.
E-mail address: asanuma@mol.nagoya-u.ac.jp (H. Asanuma).
They equally contributed to this work.
0040-4039/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2008.06.090

H. Asanuma et al. / Tetrahedron Letters 49 (2008) 5144–5146
5145
sufficient removal of Alloc group were determined. The CPG 4
binding 0.45 mol of DNA was put in a polyethylene syringe
ctatgg-5'
ctatgg-5'
O
l
H
N
O
O
H
N
equipped with a filter, and was exposed to the mixed solution of
Pd(PPh3)4 and N-methylaniline as a scavenger in CH2Cl2 or THF
(0.5 mL) for 3 h at a constant temperature. After the CPG-sup-
ported material was washed with 0.1 M N,N-diethyldithiocarba-
mate solution to remove contaminated palladium, it was cleaved
from the CPG support and deprotected by treatment with 25%
ammonia solution, followed by Poly-Pak treatment to evaluate
the removal efficiency from the results of HPLC analysis.4 Since
the Alloc group was not removed at all under the deprotection con-
ditions (ammonia treatment), the removal efficiency could be esti-
mated from the ratio of the peak area of 5 to that of 4 (see
Supplementary Fig. 1). We first applied the conditions reported
by Hayakawa et al. (Table 1, entry 1): CPG-4 was exposed with
3 equiv amounts of Pd(P(Ph)3)4 to Alloc group in THF at 50 °C for
O
O
iii
O
P
O
O
O
O
O
CN
O
P OH
c
taacg
N
H
3'-ctaacg
CPG-4
O
4
i
ctatgg-5'
ctatgg-5'
NH2
O
O
NH2
O
O P O
aacg
iii
O
O
O
P OH
CN
c
t
3'-ctaacg
N
H
3 h in the presence of
a large excess (1000 equiv/Alloc) of
O
CPG-5
5
N-methylaniline as a scavenger.7 However, only 38% was found
to be removed. We tried several conditions to raise the removal
efficiency and found that the amount of the catalyst Pd(P(Ph)3)4
was crucial. As can be seen from entries 4 and 5 in Table 1, addition
of 10 equiv amounts (5.2 mg of Pd(PPh3)4 under the present reac-
tion conditions employed) improved the efficiency (see Supple-
ii
ctatgg-5'
ctatgg-5'
O
H
N
O
H
R
N
R
iii
O
O
P OH
mentary Fig.
1 for the actual HPLC profile). Assumedly,
O
O
P
O
concentration of Pd(P(Ph)3)4 rather than its absolute amount in
the reaction mixture affected the removal efficiency.
O
O
O
3'-ctaacg
CN
-O
ct
aacg
N
H
After optimizing the conditions for the removal of the Alloc
group as above, we next tried to introduce a dye with a carboxyl
group on CPG-5 through an amide bond (CPG-6). Here, we used
thiazole orange (TO: see Scheme 2) as a target because TO, which
is an effective fluorophore in detecting oligonucleotides sequence
specifically,8,9 is rather difficult to convert to a phosphoramidite
monomer. Since TO hardly dissolved in DMF, pyridinium para-tol-
uenesulfonate (PPTS) was added to the reaction mixture as an
acidic additive.8 After the coupling reaction, the DNA was cleaved
from the support to evaluate the coupling efficiency by HPLC anal-
ysis. Among the coupling reagents we tried, PyBOP most efficiently
accelerated the coupling compared with other reagents. An in-
crease in the amount of PyBOP from 100 to 250 equiv/NH2 raised
the coupling efficiency from 28% to 60% (compare entry 5 with 6
in Table 2). The efficiency after three days coupling became 77%
(see Fig. 1 for the HPLC profile).10 Thus, we could attain the intro-
duction of a dye through threoninol on a CPG-support through an
amide bond. By this method, even two TO molecules could be effi-
ciently incorporated into DNA. We also tried another conventional
postsynthetic modification:11 fully deprotected DNA 5 was reacted
with N-hydroxysuccinimidyl (NHS) ester of merocyanine (Mero in
Scheme 2) in aqueous solution under several reaction conditions
according to the literature.12 However, since decomposition of this
activated ester proceeded much faster, the target DNA 6 could not
be obtained by all means (see Supplementary data).
6
O
CPG-6
N
S
R =
+
+
N
N
Thiazole Orange (TO)
Merocyanine (Mero)
Pyrene (Py)
N
N
+
N
N
N N
Methyl Stilbazole (MS) Azobenzene (Azo)
Methyl Red (MR)
Scheme 2. Deprotection of Alloc and amide coupling. Reagents and conditions: (i)
Pd(P(Ph)3)4, N-methylaniline, CH2Cl2, rt, 3 h; (ii) RCOOH, PyBOP, PPTS, N-methyl-
morpholine, DMF, rt 3 days; (iii) 25% aq ammonia, 50 °C, 12 h.
a phosphoramidite monomer can be incorporated into DNA via the
single phosphoramidite precursor 3.
The actual procedure for the preparation of the sub-monomer 3
is shown in Scheme 1: D-threoninol was reacted with allyl chloro-
Table 1
Removal of Alloc group from CPG-4 by palladium(0)-catalyzed allyl transfera
formate in THF in the presence of triethylamine at 0 °C to afford
compound 1. Then its primary hydroxyl group was protected with
dimethoxytrityl group (compound 2), followed by the conversion
to phosphoramidite monomer 3, which was then incorporated into
DNA on an automated DNA synthesizer. The coupling efficiency of
the obtained sub-monomer 3 was as high as those of conventional
phosphoramidite monomers, as estimated from the coloration of
the released trityl cation. We synthesized 13-mer DNA involving
Entry
Pd(PPh3)4 (equiv/Alloc)
Solvent
Temp (°C)
Removal efficiencyb (%)
1
2
3
4
5
3
3
3
3
THF
THF
CH2Cl2
CH2Cl2
CH2Cl2
50
25
25
35
35
38
13
24
70
>90
10
a
N-Methylaniline (1000 equiv/Alloc) was used as an allyl scavenger, and
deprotecting reaction was carried out for 3 h.
b
Alloc-protected
D-threoninol residues on CPG support (50 nm pore
The removal efficiency was evaluated by reversed phase HPLC analysis from the
peak areas of 4 and 5.
size, CPG-4 in Scheme 2), with which optimum conditions for the

5146
H. Asanuma et al. / Tetrahedron Letters 49 (2008) 5144–5146
Table 2
in DMF in which activated ester does not decompose, simply elon-
gating reaction time can raise coupling efficiency. In addition, we
can further elongate natural nucleotides and the threoninol units
on CPG-6 with a DNA synthesizer because protecting groups on
nucleotides and phosphorus are intact. Accordingly, second dye
that is also hard to convert to phosphoramidite monomer can be
incorporated by the subsequent deprotection of Alloc group and
amide coupling on CPG.13 This method could be also applied to
the modification of 3-amino-1,2-propanediol as well as threoninol,
demonstrating the versatility of the present Allylic protection
method.
In conclusion, a facile and versatile postsynthetic modification
method was developed with Allylic protection. This method al-
lowed diversification of the functionality of DNA via threoninol
with a single sub-monomer. By using this method, modification
of other functional nucleotides involving amino groups such as
20-amino-20-deoxyuridine is expected. We are now investigating
the spectroscopic properties of the Mero- and TO-tethered DNAs.
Incorporation of a dye to amine of threoninol residue of CPG-5a
Entry
Dye
Conditionsb
Coupling efficiencyc (%)
1
2
3
4
5
6
7
8
9
TO
TO
TO
TO
TO
TO
TO
Mero
Py
DCC/NHS (100 equiv), 1 day
7.3
1.3
3.0
6.4
28
60
77
44
>90
73
DCC/HOAt (100 equiv), 1 day
DCC/HOBt (100 equiv), 1 day
HATU (100 equiv), DMF, 1 day
PyBOP (100 equiv), DMF, 1 day
PyBOP (250 equiv), DMF, 1 day
PyBOP (250 equiv), DMF, 3 days
PyBOP (250 equiv), DMF, 3 days
PyBOP (250 equiv), DMF, 3 days
PyBOP (250 equiv), DMF, 3 days
PyBOP (250 equiv), DMF, 3 days
PyBOP (250 equiv), DMF, 3 days
10
11
12
MR
MS
Azo
50
82
a
The coupling reaction was conducted in DMF in the presence of PPTS and N-
methylmorpholine at rt.
b
Molar ratio of the added coupling reagent with respect to Alloc group on CPG
support.
c
The coupling efficiency was evaluated by reversed phase HPLC analysis from
the peak areas of 5 and 6 tethering corresponding dye.
Acknowledgments
This work was supported by Core Research for Evolution Sci-
ence and Technology (CREST), Japan Science and Technology
Agency (JST). Partial supports by a Grant-in-Aid for Scientific Re-
search from the Ministry of Education, Culture, Sports, Science
and Technology, Japan and The Mitsubishi Foundation (for H.A.)
are also acknowledged.
Supplementary data
Experimental procedures for the preparation of compounds
1–6, analytical data for 1, 2, 4–6, HPLC profile of Alloc-deprotec-
tion (Table 1, entry 5), and reaction conditions of the conven-
tional coupling in solution with Mero. Supplementary data
associated with this article can be found, in the online version,
at doi:10.1016/j.tetlet.2008.06.090.
References and notes
1. (a) Tawarada, R.; Seio, K.; Sekine, M. J. Org. Chem. 2008, 73, 383–390; (b) Oka,
N.; Wada, T.; Saigo, K. J. Am. Chem. Soc. 2003, 125, 8307–8317; (c) Virta, P.;
Katajisto, J.; Niittymäki, T.; Lönnberg, H. Tetrahedron 2003, 59, 5137–5174 and
references cited therein.
2. Hwang, J.-T.; Greenberg, M. M. J. Org. Chem. 2001, 66, 363–369.
3. (a) Shinozuka, K.; Kohgo, S.; Ozaki, H.; Sawai, H. Chem. Commun. 2000, 59–60;
(b) Ozaki, H.; Momiyama, S.; Yokotsuka, K.; Sawai, H. Tetrahedron Lett. 2001, 42,
677–680.
4. Asanuma, H.; Liang, X. G.; Nishioka, H.; Matsunaga, D.; Liu, M.; Komiyama, M.
Nat. Protocol. 2007, 2, 203–212 and references cited therein.
5. (a) Asanuma, H.; Shirasuka, K.; Takarada, T.; Kashida, H.; Komiyama, M. J. Am.
Chem. Soc. 2003, 125, 2217–2223; (b) Kashida, H.; Asanuma, H.; Komiyama, M.
Chem. Commun. 2006, 2768–2770.
Fig. 1. HPLC profiles of the modified DNAs before (solid line) and after (dotted line)
coupling with thiazole orange monitored at 260 nm. The peaks at 14 and 18.5 min
correspond to the modified DNAs without (compound 5 in Scheme 2) and with
thiazole orange (compound 6 conjugated with thiazole orange in Scheme 2) at
threoninol residue, respectively. These peaks were characterized by MALDI-TOFMS
after being fractionally collected. Only the peak at 18.5 min had orange color due to
the attached TO molecule. HPLC conditions: Reversed phase HPLC (Merck LiChro-
spher 100 RP-18(e) column) with acetonitrile/water containing 50 mM ammonium
formate (pH7.0) as mobile phase with the linear gradient 7.5–17.5% acetonitrile/
water (30 min, 0.5 mL/min).
6. Nishioka, H.; Liang, X. G.; Kashida, H.; Asanuma, H. Chem. Commun. 2007,
4354–4356.
7. Hayakawa, Y.; Wakabayashi, S.; Kato, H.; Noyori, R. J. Am. Chem. Soc. 1990, 112,
1691–1696.
8. Jarikote, D. V.; Köhler, O.; Socher, E.; Seitz, O. Eur. J. Org. Chem. 2005, 3187–
3195.
9. Constantin, T. P.; Silvia, G. L.; Robertson, K. L.; Hamilton, T. P.; Fague, K.;
Waggoner, A. S.; Armitage, B. A. Org. Lett. 2008, 10, 1561–1564.
10. Dimethoxytrityl (DMT) group was gradually cleaved from the DNA anchored
on CPG under these conditions.
11. (a) Reynolds, M. A.; Beck, T. A.; Hogrefe, R. I.; McCaffrey, A.; Arnold, L. J., Jr.;
Vaghefi, M. M. Bioconjugate Chem. 1992, 3, 366–374; (b) Fukui, K.; Morimoto,
M.; Segawa, H.; Tanaka, K.; Shimidzu, T. Bioconjugate Chem. 1996, 7, 349–355.
12. (a) Ami, T.; Fujimoto, K. Sci. Technol. Adv. Mater. 2006, 7, 249–254; (b) Viladkar,
S. M. Tetrahedron 2002, 58, 495–502.
Besides thiazole orange, all the dyes we tried in this study were
successfully incorporated into DNA on CPG support with a reason-
able efficiency as shown in Table 2. By this method, we could easily
diversify the functionality of DNA with a single sub-monomer 3 by
varying the functional molecules attached. It should be noted that
merocyanine was difficult to introduce into DNA via conventional
phosphoramidite chemistry. Even merocyanine that was hard to
be tethered to DNA through a conventional coupling in solution
could be incorporated into DNA with reasonable yield (Table 2, en-
try 8). This is one of the merits of the present postsynthetic mod-
ification on CPG support. As the present amide coupling proceeds
13. Introduction of two different dyes to threoninols at different positions in DNA
is very difficult with the conventional coupling with activated ester in solution.