A versatile modification of on-column oligodeoxynucleotides using a copper-catalyzed oxidative acetylenic coupling reaction

Article abstract of DOI:10.1021/ja036055t

We report herein a versatile postsynthetic modification of on-column oligodeoxynucleotides (ODNs) using a copper-catalyzed oxidative acetylenic coupling reaction. Hexamers supported on resins via a methylamino-modified linker were prepared, and on-column

Full text of DOI:10.1021/ja036055t

Published on Web 08/28/2003
A Versatile Modification of On-Column Oligodeoxynucleotides
Using a Copper-Catalyzed Oxidative Acetylenic Coupling
Reaction†
Noriaki Minakawa,* Yayoi Ono, and Akira Matsuda*
Contribution from the Graduate School of Pharmaceutical Sciences, Hokkaido UniVersity,
Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
Received May 10, 2003; E-mail: noriaki@pharm.hokudai.ac.jp; matuda@pharm.hokudai.ac.jp
Abstract: We report herein a versatile postsynthetic modification of on-column oligodeoxynucleotides
(ODNs) using a copper-catalyzed oxidative acetylenic coupling reaction. Hexamers supported on resins
via a methylamino-modified linker were prepared, and on-column modifications of ODNs were examined.
ArgoPore resin proved to be the best choice for the modification, and introduction of functional molecules,
such as anthraquinone, biotin, and fluorescein, resulted in good yields at not only the 5′-terminal but also
the internal 3′-end of the ODNs. This method is applicable to the modification of 12mer ODN consisting of
a random sequence. The resulting ODN9 possessing fluorescein at its 5′-terminal acts as a non-RI primer
for primer extension assays using the Klenow fragment.
Introduction
addition, the direct phosphoramidite approach cannot be applied
if the functional group is unstable under the basic deprotection
conditions in the solid-phase synthesis.
In recent years, synthetic oligodeoxynucleotides (ODNs) have
become one of the most important molecular tools in scientific
research. In addition to their role as antisense agents for
therapeutic applications,1 modified ODNs are now widely used
in genomic studies as primers for DNA sequencing and
polymerase chain reactions,2,3 and in biotechnology as molecular
beacons and probes for detection of gene expression in DNA
microarrays.4,5 However, to exploit these versatile techniques,
site-specific modifications of ODNs with functional groups
including fluorescent dyes and biotin are still needed.
The most straightforward method to introduce the functional
groups into ODNs, the so-called presynthetic method, can be
achieved by using an appropriate functionalized phosphoramidite
through automated solid-phase synthesis.6 It is reliable, and a
variety of functionalized phosphoramidites are currently com-
mercially available; however, these phosphoramidites are ex-
tremely expensive and sometimes difficult to prepare. In
A convenient alternative, the so-called postsynthetic method,
is currently being used.7 With this method, once a convertible
nucleoside unit is introduced into the ODN, the ODN can be
modified with a variety of functional groups. Therefore, the
postsynthetic method has been intensely investigated with the
hope of developing a new strategy using various organic
reactions. One typical approach is to react a primary amino
group introduced into the ODN and a functional molecule
possessing an isothiocyanate or an N-hydroxysuccinimidyl ester
moiety.8,9 However, because the coupling reaction is carried out
with an unprotected ODN in aqueous solution, a large excess
(7) Goodchild, J. Bioconjugate Chem. 1990, 1, 165-187.
(8) For examples: (a) Chehab, F. F.; Kan, Y. W. Proc. Natl. Acad. Sci. U.S.A.
1989, 86, 9178-9182. (b) Randolph, J. B.; Waggoner, A. S. Nucleic Acids
Res. 1997, 25, 2923-2929. (c) Smith, L. M.; Fung, S.; Hunkapiller, M.
W.; Hunkapiller, T. J.; Hood, L. E. Nucleic Acids Res. 1985, 13, 2399-
2412. (d) Jenkins, Y.; Barton, J. K. J. Am. Chem. Soc. 1992, 114, 8736-
8738. (e) Sigurdsson, S. T.; Eckstein, F. Nucleic Acids Res. 1996, 24, 3129-
3133.
† This paper constitutes Part 223 of Nucleosides and Nucleotides. Part
222: Minakawa, N.; Kojima, N.; Hikishima, S.; Sasaki, T.; Kiyosue, A.;
Atsumi, N.; Ueno, Y.; Matsuda, A. J. Am. Chem. Soc. 2003, 125, 9970-
9982.
(1) (a) Uhlmann, E.; Peyman, A. Chem. ReV. 1990, 90, 543-584. (b) Crooke,
S. T., Lebleu, B., Eds. Antisense Research and Application; CRC Press:
Boca Raton, FL, 1993.
(2) (a) Smith, L. M.; Sanders, J. Z.; Kaiser, R. J.; Hughes, P.; Dodd, C.; Connell,
C. R.; Heiner, C.; Kent, S. B. H.; Hood, L. E. Nature 1986, 321, 674-
679. (b) Chehab, F. F.; Kan, Y. W. Proc. Natl. Acad. Sci. U.S.A. 1989, 86,
9178-9182. (c) Ju, J.; Ruan, C.; Fuller, C. W.; Glazer, A. N.; Mathies, R.
A. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 4347-4351.
(3) Erlich, H. A.; Gelfand, D.; Sninsky, J. J. Science 1991, 252, 1643-1650.
(4) (a) Zhang, P.; Beck, T.; Tan, W. Angew. Chem., Int. Ed. 2001, 40, 402-
405. (b) Kuhn, H.; Demidov, V. V.; Coull, J. M.; Fiandaca, M. J.; Gildea,
B. D.; Frank-Kamenetskii, M. D. J. Am. Chem. Soc. 2002, 124, 1097-
1102. (c) Summerer, D.; Marx, A. Angew. Chem., Int. Ed. 2002, 41, 3620-
3622.
(9) Except for ref 8, various postsynthetic modifications of ODN were reported.
For examples: (a) Connolly, B. A. Nucleic Acids Res. 1985, 13, 4485-
4502. (b) Jones, D. S.; Barstad, P. A.; Feild, M. J.; Hachmann, J. P.; Hayag,
M. S.; Hill, K. W.; Iverson, G. M.; Livingston, D. A.; Palanki, M. S.;
Tibbetts, A. R.; Yu, L.; Coutts, S. M. J. Med. Chem. 1995, 38, 2138-
2144. (c) Rajur, S. B.; Roth, C. M.; Morgan, J. R.; Yarmush, M. L.
Bioconjugate Chem. 1997, 8, 935-940. (d) Zuckermann, R.; Corey, D.;
Schultz, P. Nucleic Acids Res. 1987, 15, 5305-5320. (e) Hanna, M. M.;
Meyer, K. L. Bioconjugate Chem. 1996, 7, 401-412. (f) Coleman, R. S.;
Arthur, J. C.; McCary, J. L. Tetrahedron 1997, 53, 11191-11202. (g) Ono,
A.; Haginoya, N.; Kiyokawa, M.; Minakawa, N.; Matsuda, A. Bioorg. Med.
Chem. Lett. 1994, 4, 361-366. (h) Haginoya, N.; Ono, A.; Nomura, Y.;
Ueno, Y.; Matsuda, A. Bioconjugate Chem. 1997, 8, 271-280. (i) Nomura,
Y.; Ueno, Y.; Matsuda, A. Nucleic Acids Res. 1997, 25, 2784-2791. (j)
MacMillan, A. M.; Verdine, G. L. J. Org. Chem. 1990, 55, 5931-5933.
(k) Xu, Y.-Z.; Zheng, Q.; Swann, P. Tetrahedron 1992, 48, 1729-1740.
(l) Gao, H.; Fathi, R.; Gaffney, B. L.; Goswami, B.; Kung, P.-P.; Rhee,
Y.; Jin, R.; Jones, R. A. J. Org. Chem. 1992, 57, 5931-5933. (m) Kim, S.
J.; Stone, M. P.; Harris, C. M.; Harris, T. M. J. Am. Chem. Soc. 1992, 114,
5480-5481.
(5) (a) Fodor, S. P. A.; Read, J. L.; Pirrung, M. C.; Stryer, L.; Lu, A T.; Solas,
D. Science 1991, 251, 767-773. (b) Schena, M.; Shalon, D.; Davis, R.
W.; Brown, P. O. Science 1995, 270, 467-470.
(6) Beaucage, S. L.; Iyer, R. P. Tetrahedron 1993, 49, 1925-1963.
9
10.1021/ja036055t CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 11545-11552
11545

A R T I C L E S
Minakawa et al.
Table 1. The Copper-Catalyzed Oxidative Acetylenic Coupling
Scheme 1 a
Reaction of EdU Derivatives
entry
substrate
solvent
time (h)
product
yield (%)
1
2
3
4
1
1
2
2
DMF
CH3CN
80% CH3CN/H2O
H2O
24
24
48
72
3
3
4
4
78
71
87
68
of the functional molecule is required and side reactions may
take place with the exocyclic amino groups of the nucleobases.
To overcome these problems, Greenberg et al. have developed
a method for an efficient on-column modification using a
photolabile phosphoramidite unit.10 By employing a site-specific
modification, their approach has the advantage of a facile
purification of the resulting ODN; however, a photochemical
apparatus is required. Grinstaff et al. reported a novel procedure
that combines the advantages of on-column modification and
Pd(0)-catalyzed cross-coupling reactions.11 Although this method
utilizes general Pd chemistry for an ODN modification, a rather
unstable Pd(0) catalyst, Pd(Ph3P)4, is required. In addition, the
coupling reaction proceeded well only when a convertible
nucleoside (2′-deoxy-5-iodouridine, IdU) unit was located at the
5′-terminal of the on-column ODN.11b Consequently, develop-
ment of a versatile and flexible procedure which is site-specific
and high yielding to give modified ODNs would be highly
desirable.12
a Reagents: (a) p-nitrophenylchloroformate, Et3N, dioxane; (b) Ac2O,
pyridine; (c) 1,8-diaminooctane, dioxane; (d) 6, EDC, DMF; (e) automated
ODN synthesis.
with 2 equiv of CuCl-TMEDA (N,N,N′,N′-tetramethyleth-
ylenediamine) complex in DMF under an O2 atmosphere, the
homo-coupling product 3 was obtained in 78% yield (entry 1).
The reaction also proceeded in CH3CN (entry 2). Although the
reaction rate slowed, the acetylenic coupling reaction of EdU
(2)14 in aqueous solvent gave 4 in good yield (entries 3 and 4).
Because the homo-coupling reaction proceeded well in the
liquid phase, the reaction of the on-column ODN possessing
an EdU unit with the EdU derivative to give the ODN with a
diacetylene was examined. As mentioned in the Introduction,
on-column modification by Pd(0)-catalyzed cross-coupling
reactions proceeded well only when the IdU unit was located
at the 5′-terminal of the ODN.11b This can probably be attributed
to the physical properties of the resin-supported ODN, including
its pore size. A 500 Å controlled pore glass (CPG500) resin is
generally used for ODN synthesis, and the Pd(0)-catalyzed cross-
coupling reaction was also carried out on this resin. Our thought
was that this resin and its pore size are not suitable for on-
column modification at the internal 3′-end of the on-column
ODNs. On the basis of these considerations, we decided to
employ 2000 Å controlled pore glass (CPG2000) and ArgoPore
resins along with CPG500 for the on-column modification. Our
proposed reaction was first examined with an on-column ODN
with a convertible nucleoside, that is, the EdU unit, at the 5′-
terminal. The synthesis of the resin-supported ODNs 12-14 is
shown in Scheme 1. The reaction of 5′-O-(4,4′-dimethoxytrityl)-
3′-O-succinylthymidine (6) with commercially available CPG2000
resin (23 µmol/g) under standard conditions gave the solid
support 8. For the synthesis of support 9, ArgoPore (280 µmol/
Herein, we report a versatile modification of on-column
ODNs using a copper-catalyzed oxidative acetylenic coupling
reaction. To achieve the desired modification at various sites
of the ODNs, we also examined resin-supported ODNs. As a
result, we have found a suitable resin to introduce the functional
molecules at not only the 5′-terminal but also at the internal
3′-end of on-column ODNs.
Results and Discussion
The copper-catalyzed oxidative acetylenic coupling reaction
is a classical C-C bond forming reaction, one of the few which
takes place under mild conditions in various solvents, including
aqueous solution.13 Therefore, the acetylenic coupling reaction
was chosen for the site-specific modification of ODNs, polar
macromolecules. Prior to using on-column ODNs, the reaction
conditions were first examined on a nucleoside level (Table 1).
When protected 5-ethynyl-2′-deoxyuridine (EdU) 1 was treated
(10) (a) Kahl, J. D.; Greenberg, M. M. J. Am. Chem. Soc. 1999, 121, 597-604.
(b) Hwang, J.-T.; Greenberg, M. M. Org. Lett. 1999, 1, 2021-2024. (c)
Hwang, J.-T.; Greenberg, M. M. J. Org. Chem. 2001, 66, 363-369.
(11) (a) Khan, S. I.; Grinstaff, M. W. J. Am. Chem. Soc. 1999, 121, 4704-
4705. (b) Beilstein, A. E.; Grinstaff, M. W. Chem. Commun. 2000, 509-
510.
(12) For examples: (a) Hill, K. W.; Taunton-Rigby, J.; Carter, J. D.; Kropp,
E.; Vagle, K.; Pieken, W.; McGee, D. P. C.; Husar, G. M.; Leuck, M.;
Anziano, D. J.; Sebesta, D. P. J. Org. Chem. 2001, 66, 5352-5358. (b)
Graham, D.; Grondin, A.; Mchugh, C.; Fruk, L.; Smith, W. E. Tetrahedron
Lett. 2002, 43, 4785-4788. (c) Okamoto, A.; Taiji, T.; Tainaka, K.; Saito,
I. Bioorg. Med. Chem. Lett. 2002, 12, 1895-1896. (d) Seo, T. S.; Li, Z.;
Ruparel, H.; Ju, J. J. Org. Chem. 2003, 68, 609-612.
(13) For a review, see: Siemsen, P.; Livingston, R. C.; Diederich, F. Angew.
Chem., Int. Ed. 2000, 39, 2632-2657.
(14) Crisp, G. T.; Flynn, B. L. J. Org. Chem. 1993, 58, 6614-6619.
9
11546 J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003

Modification of On-Column Oligodeoxynucleotides
A R T I C L E S
Scheme 2 a
Figure 1. HPLC profiles of (a) ODN1 (treatment of 12 with NH4OH); (b)
crude material after on-column reaction of 12 with 15 (entry 1).
and 2% yield of ODN3, a coupling product between on-column
ODNs17,18 (Table 2, entry 1 and Figure 1b). Although the
chemical yield of the desired ODN2 was slightly improved by
increasing the reaction time (entry 2), 15 was nearly completely
consumed after 48 h due to the formation of the homo-coupling
product. Therefore, two cycles of the coupling reaction (pro-
cedures 2-4) were carried out, and these conditions gave ODN2
in 51% yield (entry 3). With the reaction conditions fixed, the
reaction of 13 and 14 was next examined. Both substrates gave
ODN2 in better yield than did 12 (entries 4 and 5), which
suggests that the physical properties of the resins influence on-
column modifications in copper-catalyzed acetylenic coupling
reactions. Because ODN1 was not completely consumed in
every entry, the reaction was carried out at 50 °C. Consequently,
the desired ODN2 was obtained in 90% yield (entry 6), as
estimated from peak areas in the HPLC analysis. However, the
total peak areas of the HPLC under these conditions were
approximately 40% less than those under conditions carried out
at room temperature. Thus, cleavage of the succinyl linker
supporting the ODN seemed to take place with heating. This is
presumably due to a mechanism involving deprotonation of the
amide nitrogen by TMEDA followed by intramolecular nucleo-
philic displacement at the ester carbonyl group.
a Reagents: (a) 15, CuCl, TMEDA, DMF, O2; (b) NH4OH.
Table 2. The Copper-Catalyzed Oxidative Acetylenic Coupling
Reaction of On-Column ODNs 12-14 with 15
yields (%)a
entry
substrate
time (h)
24
temp
ODN1
ODN2
ODN3
1
2
3
4
5
6
12
12
12
13
14
13
room temp
room temp
room temp
room temp
room temp
50 °C
60
51
47
28
35
2
38
43
51
72
58
90
2
6
2
trace
6
8
48
24 × 2b
24 × 2b
24 × 2b
24 × 2b
a Yields were estimated from the ratio of peak areas. b The reaction was
carried out for 24 h, and the resin was washed. Another cycle of the same
coupling was then subjected for 24 h.
g) was first treated with p-nitrophenyl chloroformate in dioxane.
After treatment with acetic anhydride to cap the remaining free
amino groups, the resulting resin was reacted with 1,8-
diaminooctane to give 5 (145 µmol/g), possessing an appropriate
alkyl linker. Compound 6 was then coupled with 5 under the
same conditions to give 9 (50 µmol/g). The resin-supported
hexamers 12-14 containing the convertible unit at their
5′-termini were prepared on a DNA synthesizer using the
phosphoramidites 10 and 11.15 The average coupling yield of
the phosphoramidites was 99% using a 0.1 M solution of each
phosphoramidite in CH3CN and a coupling time of 300 s for
11 and 30 s for 10, even when support 9 was used for the
automated ODN synthesis.
With 12-14 in hand, on-column modification by the copper-
catalyzed acetylenic coupling reaction was next examined. The
reaction was conducted as follows: (1) 2 mg of the resin-
supported hexamer was poured into a reactive tube for the solid-
phase synthesis; (2) 6 mg of 1516 (large excess) and the CuCl-
TMEDA complex in DMF (0.1 M, 500 µL) were added to the
tube; (3) the mixture was shaken under an O2 atmosphere; (4)
the resin was washed with appropriate solvents; (5) the
remaining resin was treated with NH4OH to detach the ODNs
from the resin; and (6) the resulting ODNs were analyzed by
reversed-phase HPLC, and the yields were estimated from their
peak areas (see Experimental Section). The results are sum-
marized in Scheme 2, Table 2, and Figure 1. When 12 was
treated with 15 for 24 h, the desired product ODN2 was obtained
in 38% yield along with 60% yield of ODN1 (unreacted ODN)
In accordance with the results described above, a base-stable
linker,19 the methylamino-modified linker, was devised for the
on-column modification. The requisite resins were prepared as
shown in Scheme 3. According to the method reported by Brown
et al.,19a the primary amino groups of CPGs were coupled with
Boc-sarcosine,20 followed by treatment with trifluoroacetic acid
to give 16 and 17, respectively. For the ArgoPore resin, 18 was
prepared in a manner similar to that for 5.19b The resulting resins
were then coupled with 6 to give 19-21, respectively. The resin-
supported hexamers 24-26 were prepared on a DNA synthesizer
in high average coupling yield. To examine the on-column
modification at the internal 3′-end, the resin-supported hexamers
27 and 28 containing the convertible unit at their 3′-end were
also prepared via the functionalized supports 22 and 23,
respectively. Prior to the reaction of the on-column ODN, the
stability of the methylamino-modified linker under the reaction
conditions was examined. When 19 was subjected to the
acetylenic coupling conditions at 50 °C for 24 h in the absence
of 15, cleavage of the linker was not detected at all (data not
shown).21
(17) When the reaction was carried out in the absence of 15, ODN3 was obtained
in approximately 20% yield, which was analyzed by MALDI-TOF/MASS
measurement.
(18) ODN1-9 possess DMTr groups on their 5′-termini, except when the ODNs
are used for MALDI-TOF/MASS measurements and primer extension
assays.
(19) (a) Brown, T.; Pritchard, C. E.; Turner, G.; Salisbury, S. A. J. Chem. Soc.,
Chem. Commun. 1989, 891-893. (b) Weiler, J.; Hoheisel, J. D. Anal.
Biochem. 1990, 189, 218-227.
(15) Graham, D.; Parkinson, J. A.; Brown, T. J. Chem. Soc., Perkin Trans. 1
1998, 1131-1138.
(16) Hurley, D. J.; Tor, Y. J. Am. Chem. Soc. 2002, 124, 3749-3762.
(20) In the original paper by Brown et al., FMOC-sarcosine was used.
9
J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003 11547

A R T I C L E S
Minakawa et al.
Scheme 3 a
a Reagents: (a) Boc-sarcosine, EDC, CH2Cl2; (b) TFA, CH2Cl2; (c) p-nitrophenylchloroformate, Et3N, dioxane; (d) Ac2O, pyridine; (e) N,N′-dimethyl-
1,6-diaminohexane; (f) 6 or 5-ethyl-5′-O-(4,4′-dimethoxytrytyl)-3′-O-succinyl-2′-deoxyuridine, EDC, DMF; (g) automated ODN synthesis.
Table 3. The Copper-Catalyzed Oxidative Acetylenic Coupling
Reaction of On-Column ODNs 24-26 with 15
Table 4. The Copper-Catalyzed Oxidative Acetylenic Coupling
Reaction of On-Column ODNs 27 and 28 with 15
yields (%)a
entry
substrate
ODN1
ODN2
ODN3
1
2
3
24
25
26
27
4
6
65
89
73
8
7
21
yields (%)a
entry
substrate
ODN4
ODN5
ODN6
1
2
27
28
9
trace
59
84
32
16
a Yields were estimated from the ratio of peak areas.
Because this linker tolerates the acetylenic coupling condi-
tions, the reaction of on-column ODNs employing 24-26 was
first examined (Table 3). When 24 was treated with 15 in the
presence of the CuCl-TMEDA complex at 50 °C, the desired
ODN2 was obtained in 65% yield along with 27% of ODN1
(entry 1). As we anticipated, when 25 was subjected to the
coupling reaction under the same conditions, ODN1 nearly
completely disappeared, and the desired ODN2 was obtained
in much better yield (entry 2). Although a moderate yield of
ODN3 was obtained in the presence of a large excess of 15,
the reaction of 26 also resulted in a small amount of the
remaining ODN1 and a better yield of ODN2 than that of 24
(entry 3).
The advantage of using CPG2000 and ArgoPore resins was
clearly observed when the on-column acetylenic coupling
reaction was examined at the internal 3′-end. The requisite
reaction did not proceed when the hexamer on the CPG500 resin
was subjected to the reaction,22 and this result agreed well with
that reported by Grinstaff et al.11b As can be seen in Table 4,
when 27 and 28 were subjected to the reaction at 50 °C, ODN4
(unreacted ODN) almost disappeared in both cases, and the
desired ODN5 was obtained in good yields (59% and 84%).
Interestingly, contrary to the results using 25 and 26, the reaction
of 27 gave a moderate yield of ODN6 (32%), a coupling product
between two on-column ODNs, while that of 28 afforded ODN5
in better yield (84%) with less ODN6 being formed (16%).18
a Yields were estimated from the ratio of peak areas.
As mentioned above, we succeeded with the copper-catalyzed
acetylenic coupling reaction of the on-column ODN at not only
the 5′-terminal but also the internal 3′-end. This reaction was
definitely affected by the kinds of resins supporting the ODN,
which presumably depends on the physical properties of each
resin.23 A CPG resin is a silica-based support and is most
common and suitable for efficient ODN synthesis. The typical
property of this resin is nonswelling in organic solvents.
Therefore, the accessibility of a reactant is not dependent on
the swellability but probably on the pore size of the resin.23
CPG500 is typically used and prefers a rather shorter ODN
synthesis. The pore size of this resin tolerates the acetylenic
coupling reaction when the reaction center is located at the 5′-
terminal of the ODNs, whereas it is not suitable for the
modification at the internal 3′-end located at the inner surface
of the cavity because of poor accessibility of the reactant, such
as 15. This consideration is also supported by Grinstaff’s
results.11b In contrast, CPG2000 (a support preferring longer
ODN synthesis) has a wider pore size, and thus the reactant
can reach to the convertible unit even when it is located at the
internal 3′-end. Unlike CPGs, ArgoPore is a highly cross-linked
polystyrene resin, which was developed to facilitate the direct
transfer of reaction conditions from known solution-phase
methods to solid-phase methods. This resin offers the advantages
of permitting rapid reaction kinetics and diffusion of reactants
to the reaction center.24 These properties would be convenient
(21) When 7 was subjected to the acetylenic coupling reaction under the same
conditions, the product of cleavage of the succinyl linker was detected in
approximately 20% yield.
(22) Although we did not prepare a hexamer on CPG500 analogous to 27
(possessing a methylamino-modified linker), our preliminary attempt using
a hexamer on CPG500 possessing a succinyl linker did not give the coupling
product when the EdU unit was introduced at the internal 3′-end.
(23) Bellon, L.; Wincott, F. In Solid-Phase Synthesis. A Practical Guide; Kates,
S. A., Albericio, F., Eds.; Marcel Dekker Inc.: New York, 2000; pp 475-
528 and references therein.
9
11548 J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003

Modification of On-Column Oligodeoxynucleotides
A R T I C L E S
Table 5. The Copper-Catalyzed Oxidative Acetylenic Coupling
Reaction of On-Column ODNs 26 and 28 with Functional
Molecules
Figure 2. Structures of the functional molecules.
for modification of on-column ODNs, and the accessibility of
the reactant should be the same either at the 5′-terminal or at
the internal 3′-end of on-column ODNs. Although this consid-
eration does not explain the product distribution of the reaction
using 25 and 26 (the reaction at the 5′-terminal), and 27 and 28
(the reaction at the internal 3′-end), these physical and structural
properties of the resins would partly explain the above-
mentioned results. Thus far, although much effort has been
dedicated to elucidating the effects of the resin on the synthesis
of ODNs,23 little is known about the effect of the resin on on-
column modifications. To the best of our knowledge, only one
example has been reported by De Napoli et al.,25 in which an
on-column synthesis of cyclic ODNs proceeded well when an
ODN supported on Tentagel was used rather than the reaction
of an ODN supported on CPG. Tentagel is a highly cross-linked
polystyrene resin similar to ArgoPore, with a long poly(ethylene
glycol) linker (n ) 70).23,26 Although this resin is known to be
one of the most effective resins for ODN synthesis, our
preliminary experiment using a Tentagel supported ODN
resulted mainly in the formation of dimers such as ODN3
(Scheme 2) and ODN6 (Table 4) rather than the desired products
ODN2 and ODN5 (data not shown). These results imply that
the character of the linker and its length may also affect the
chemical reactions of on-column ODNs, and ArgoPore pos-
sessing short alkyl linkers was better than Tentagel in our
experiments.
To make this a versatile method, we next introduced
functional molecules such as fluorescent dye into ODNs. Our
results indicate that the CPG2000 resin also prefers ODN
modification at the 5′-terminal end, while ArgoPore resin prefers
the internal 3′-end. If the reaction between the interstrand ethynyl
groups to form undesired ODN3 could be avoided, ArgoPore
would be better than CPG2000 for the ODN modification at
either side because of its higher loadings. Therefore, we designed
functional molecules possessing a more reactive ethynyl group,
which would react readily with the convertible unit on the ODN
and thus would diminish the formation of ODN3. To this end,
the functional molecules 29a-c were prepared (Figure 2).27
When the on-column ODN 26 was subjected to the acetylenic
coupling reaction in the presence of 29a under optimized
entry
substrate
reactant
product
yields (%)a
1
2
3
26
26
26
29a
29b
29c
ODN7a
ODN7b
ODN7c
95
91
82
4
5
6
28
28
28
29a
29b
29c
ODN8a
ODN8b
ODN8c
91
88
76
a Yields were estimated from the ratio of peak areas.
Figure 3. HPLC profiles of the crude material after on-column reaction
of 26 (a) and 28 (b) with 29a.
conditions, the ODN7a was obtained in 95% yield (Table 5,
entry 1). As expected, formation of the undesired ODN3 was
negligible, and the HPLC trace after the reaction was quite
simple without any purification (Figure 3a). Introduction of
biotin and fluorescent dye using 29b and 29c also proceeded
well to give ODN7b and ODN7c in good yields, respectively
(entries 2 and 3). Reaction of 28 possessing the convertible unit
at the internal 3′-end with 29a-c gave ODN8a-c in good
yields, respectively, equal to those of 26 (entries 4-6). As can
be seen in Figure 3b, the HPLC trace was also unambiguous.
Consequently, we succeeded in developing an efficient method
to introduce functional molecules at arbitrary positions on on-
column ODNs supported on ArgoPore.
Our goal was to develop a versatile method for the introduc-
tion of functional molecules into ODNs which could be used
as primers and probes. Although we were successful in
introducing functional molecules such as fluorescent dyes into
ODNs, our attempts resulted in only a simple and short
sequence. In addition, the functional molecules are attached to
ODNs with a rigid and somewhat shorter linker than usual, when
the functional molecules are attached to ODNs via long alkyl
chains. Therefore, we examined the modification of a 12mer
ODN with a random sequence and evaluated the resulting ODN
for a fluorescent dye labeled primer. Starting with 21, a resin-
supported 12mer 30 possessing the convertible unit on its 5′-
terminal was prepared (Scheme 4). Treatment of 30 with 29c
in the presence of the CuCl-TMEDA complex in DMF for 48
h, followed by NH4OH, gave ODN9 possessing the fluorescent
dye at its 5′-terminal in 83% yield regioselectively (estimated
(24) Resins and Reagent Catalog, Technical Handbook; Argonaut Technologies,
Inc.; pp 144-148.
(25) (a) De Napoli, L.; Galeone, A.; Mayol, L.; Messere, A.; Piccialli, G.;
Santacroce, C. J. Chem. Soc., Perkin Trans. 1 1993, 747-749. (b) De
Napoli, L.; Galeone, A.; Mayol, L.; Messere, A.; Montesarchio, D.; Piccialli,
G. Bioorg. Med. Chem. 1995, 3, 1325-1329.
(26) Forns, P.; Fields, G. B. In Solid-Phase Synthesis. A Practical Guide; Kates,
S. A., Albericio, F., Eds.; Marcel Dekker Inc.: New York, 2000; pp 1-77
and references therein.
(27) The synthesis of 29a-c was presented in the Supporting Information.
9
J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003 11549

A R T I C L E S
Minakawa et al.
Scheme 4 a
most comprehensive modification methods of on-column ODNs
because of its simple technique of organic and solid-phase
synthesis. In addition, we would like to emphasize that our
findings could be applicable to on-column modification of
ODNs using other reaction systems.
Experimental Section
General Methods. Physical data were measured as follows. Melting
points are uncorrected. 1H and 13C NMR spectra were recorded at 270
and 100 MHz instruments, respectively, in CDCl3 or DMSO-d6 as the
solvent with tetramethylsilane as an internal standard. Chemical shifts
are reported in parts per million (δ), and signals are expressed as s
(singlet), d (doublet), t (triplet), q (quartet), m (multiplet), or br (broad).
Mass spectra were measured on a JEOL JMS-D300 spectrometer. TLC
was done on Merck Kieselgel F254 precoated plates. Silica gel used
for column chromatography was Merck silica gel 5715. CPG500 and
CPG2000 resins were purchased from Glen Research. ArgoPore-NH2
resin was purchased from Aldrich.
a Reagents: (a) automated ODN synthesis; (b) 29c, CuCl, TMEDA,
DMF, O2, 50 °C; (c) NH4OH, 55 °C; (d) 80% aq AcOH.
5-Ethynyl-5′-O-(4,4′-dimethoxytrityl)-3′-O-levulinyl-2′-deoxyuri-
dine (1). To a solution of 1516 (1.70 g, 3.0 mmol) in CH2Cl2 (30 mL)
were added levulinic acid (614 µL, 6.0 mmol), DMAP (730 mg, 6.0
mmol), and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydro-
chloride (EDC) (1.4 g, 73 mmol), and the mixture was stirred for 1 h
at room temperature. The reaction mixture was partitioned between
AcOEt and H2O, and the separated organic layer was washed with brine
and dried (Na2SO4). The organic layer was concentrated in vacuo, and
the residue was purified by a silica gel column, eluted with hexane/
AcOEt (1:1-1:2) to give 1 (1.85 g, 94%) as a pale yellow foam: 1H
NMR (270 MHz, CDCl3) δ 8.11 (br s, 1H), 7.42-7.21 (m, 10H), 6.86
and 6.83 (each s, each 2H), 6.30 (dd, 1H, J ) 5.3 and 8.6 Hz), 5.02
(m, 1H), 4.17 (m, 1H), 3.79 (s, 6H), 3.49 (dd, 1H, J ) 3.3 and 10.6
Hz), 3.33 (dd, 1H, J ) 2.6 and 10.6 Hz), 2.86 (s, 1H), 2.77-2.56 (m,
5H), 2.44-2.34 (m, 1H), 2.20 (s, 3H); 13C NMR (100 MHz, CDCl3) δ
205.95, 171.91, 160.72, 158.44, 148.74, 144.10, 143.30, 135.24, 134.95,
129.85, 127.95, 127.71, 126.89, 113.29, 113.26, 99.52, 87.31, 85.28,
84.53, 82.03, 77.20, 75.25, 73.77, 63.54, 55.29, 38.75, 37.88, 29.90,
28.04; FAB-LRMS (NaI) m/z 675 (MNa+); FAB-HRMS (NaI) calcd
for C37H36N2O9Na (MNa+) 675.2359; found 675.2339.
Bis[5′-O-(4,4′-dimethoxytrityl)-3′-O-levulinyl-2′-deoxyuridine-5-
yl]diacetylene (3). A solution of CuCl (91 mg, 0.91 mmol) and
TMEDA (140 µL, 0.91 mmol) in DMF (5 mL) was stirred for 30 min
under O2 atmosphere at room temperature. Compound 1 (300 mg, 0.46
mmol) was added to the solution, and the mixture was stirred for 24 h
at room temperature. The reaction mixture was partitioned between
AcOEt and H2O, and the separated organic layer was washed with 5%
EDTA aq, H2O, and brine, and was dried (Na2SO4). The organic layer
was concentrated in vacuo, and the residue was purified by a silica gel
column, eluted with 0-5% MeOH in CHCl3 to give 3 (230 mg, 78%)
as a pale yellow foam: 1H NMR (270 MHz, CDCl3) δ 8.20 (br s, 2H),
8.00 (s, 2H), 7.38-7.15 (m, 18H), 6.86 and 6.83 (each s, each 4H),
6.22 (dd, 2H, J ) 5.6 and 8.6 Hz), 5.46 (m, 2H), 4.20 (m, 2H), 3.74
and 3.73 (each s, each 6H), 3.53 (dd, 2H, J ) 3.0 and 10.6 Hz), 3.27
(dd, 1H, J ) 2.6 and 10.6 Hz), 2.79-2.56 (m, 10H), 2.40 (m, 1H),
2.20 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 206.16, 171.99, 160.53,
158.51, 148.95, 143.96, 135.20, 134.93, 129.88, 129.81, 128.02, 127.72,
127.07, 113.36, 99.27, 87.25, 86.03, 84.68, 78.28, 75.43, 73.13, 63.65,
55.25, 38.70, 37.82, 29.81, 27.97; FAB-LRMS m/z 1303 (MH+); FAB-
HRMS calcd for C74H71N4O18 (MH+) 1303.4717; found 1303.4740.
Bis(2′-deoxyuridine-5-yl)diacetylene (4). A solution of CuCl (16
mg, 0.16 mmol) and TMEDA (24 µL, 0.16 mmol) in H2O (1 mL) was
stirred for 30 min under O2 atmosphere at room temperature. Compound
2 (20 mg, 0.08 mmol) was added to the solution, and the mixture was
further stirred for 72 h at room temperature. The reaction mixture was
diluted with MeOH, and H2S gas was bubbled through the mixture.
The resulting precipitation was filtered, and the filtrate was concentrated
in vacuo. The residue was purified by C18-reversed phase column HPLC
Figure 4. Primer extension assay using ODN9.
from HPLC analysis, Supporting Information). With ODN9 in
hand, a primer extension assay using the Klenow fragment (KF)
was examined in the labeled ODN9-template duplex. The
experiments were carried out at 37 °C for 1 h in the presence
of varied mixtures of dNTPs. The primer extension reaction
was analyzed by an Image Analyzer after denaturing polyacryl-
amide gel electrophoresis. As can be seen in Figure 4, the
expected extension was observed in all conditions with 1.0 pmol
of the primer ODN9 (lanes 2-5). Consequently, it has been
demonstrated that the fluorescent labeled ODN9 acts as a non-
RI primer.
In conclusion, we have demonstrated the on-column modi-
fications of ODNs using a copper-catalyzed oxidative acetylenic
coupling reaction. This reaction was obviously affected by the
physical properties of the resin-supported ODN, and ArgoPore
proved to be the best choice of the modification. The ODN,
which includes the EdU unit, supported on ArgoPore reacts with
a substrate possessing a terminal acetylene regioselectively at
not only the 5′-terminal but also the internal 3′-end to give
functionalized ODNs. Although the functional group is attached
with an ODN through a rigid and somewhat short diacetylene
linker, the fluorescent dye labeled ODN effectively acts as a
non-RI primer. This method consists of (1) the readily available
phosphoramidite derivative of EdU; (2) the usual automated
ODN synthesis using ArgoPore resin; (3) preparation of
functional molecules possessing a terminal acetylene; and (4)
use of a stable CuCl complex. Our results have shown that the
copper-catalyzed acetylenic coupling reaction is amenable to
on-column modification of ODNs. This would be one of the
9
11550 J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003

Modification of On-Column Oligodeoxynucleotides
A R T I C L E S
and eluted with 5-50% MeOH in H2O to give 4 (14 mg, 68%) as a
pale yellow solid (crystallized from MeOH/H2O): mp 242 °C (colored);
1H NMR (270 MHz, DMSO-d6) δ 11.72 (br s, 2H), 8.46 (s, 2H), 6.06
(dd, 1H, J ) 6.6 and 5.9 Hz), 5.23 (d, 1H, J ) 4.6 Hz), 5.17 (t, 1H,
J ) 4.6 Hz), 4.21 (m, 1H), 3.78 (m, 1H), 3.59 (m, 2H), 2.12 (m, 2H);
13C NMR (100 MHz, CDCl3) δ 161.30, 148.97, 145.86, 96.54, 87.57,
85.09, 76.47, 75.39, 69.66, 60.64, 40.37; FAB-LRMS m/z 503 (MH+).
Anal. Calcd for C22H22N4O10‚0.5H2O: C, 51.67; H, 4.65; N, 10.77.
Found: C, 51.42; H, 4.45; N, 10.65. UV (H2O) λmax 356.6 nm (ꢀ
19 500), 333.6 nm (ꢀ 28 900), 314.0 nm (ꢀ 26 800), 238.8 nm (ꢀ
25 900), 333.6 nm (ꢀ 21 200), ꢀ260 17 900.
Preparation of ArgoPore Resin 5 Containing an Aminooctyl
Linker. To a mixture of ArgoPore-NH2 (1.00 g, 280 µmol/g), dioxane
(10 mL), and Et3N (1.0 mL) was added a solution of p-nitrophenyl-
chloroformate (560 mg, 2.80 mmol) in dioxane (5 mL) at 0 °C, and
the mixture was kept for 12 h at room temperature. The solid support
was filtered and washed with CH2Cl2, and the remaining amino groups
were capped by treatment with Ac2O/pyridine (1:1, 10 mL) for 3 h.
After the solid support was filtered and washed with CH2Cl2, a solution
of 1,8-diaminooctane (288 mg, 2.0 mmol) in dioxane (10 mL) was
added, and the mixture was kept for 48 h at room temperature. The
resulting solid support 5 was filtered and washed with MeOH and CH2-
Cl2 and was dried under reduced pressure. The amount of the reactive
amino group on the support was estimated by picrate assay26,28 to be
145 µmol/g.
Preparation of Methylamino-Modified CPG 500 Resin 16. To a
suspension of CPG500 (500 mg, 90.0 µmol/g) in CH2Cl2 (5 mL) were
added Boc-sarcosine (34 mg, 0.18 mmol) and EDC (34 mg, 0.18 mmol),
and the mixture was kept for 8 h at room temperature. The resin was
filtered and washed with CH2Cl2, and the remaining amino groups were
capped by treatment with Ac2O/pyridine (1:1, 4 mL) for 3 h. After the
resin was filtered and washed with CH2Cl2, the Boc group was removed
by treatment with 50% TFA in CH2Cl2 for 1 h. The resulting resin 16
was filtered and washed with 5% iPr2NEt in CH2Cl2 (10 mL) and CH2-
Cl2 and was dried under reduced pressure. The amount of the methyl-
amino group on the resin was estimated by picrate assay26,28 to be 90.0
µmol/g.
the extended coupling time (300 s) of EdU phosphoramidite 11. After
completion of the synthesis of fully protected ODNs supported by resins,
a small amount of the each resin (approximately 2 mg) was treated
with concentrated NH4OH (12-14 and 24-28, room temperature for
2 h; 30, 55 °C for 12 h). After filtration of the resin, the filtrate was
concentrated in vacuo to give ODN1, ODN4, and ODN9 possessing a
5′-DMTr group, respectively. The resulting ODN1, ODN4, and ODN9
were used as references for the subsequent on-column modification
using the copper-catalyzed oxidative acetylenic coupling reaction. The
remaining resin-supported ODNs were stored at -30 °C.
General Procedure of the Acetylenic Coupling Reaction Using
Resin-Supported ODN. A solution of CuCl (5 mg, 50 µmol) and
TMEDA (15 µL, 100 µmol) in DMF (500 µL) was stirred for 30 min
under O2 atmosphere at room temperature. To the resin-supported ODN
(2 mg, containing 20-100 nmol of ODN) in a Valiant bond elute
reservoir tube (5 mL, equipped with a polystyrene filter at the base)
were added a reactant (15, 29a, 29b, or 29c, 10 µmol) and the
CuCl-TMEDA/DMF solution (500 µL), and the mixture was shaken
under O2 atmosphere at 50 °C for 24 h. The resin was washed with
DMF (1 mL × 3), 5% EDTA aq (1 mL × 3), H2O (1 mL × 3), and
CH3CN (1 mL × 3). The resulting resin was subjected to another cycle
of the coupling reaction (24 h) and washed. The resin was then treated
with concentrated NH4OH (12-14 and 24-28, room temperature for
2 h; 30, 55 °C for 12 h). After filtration of the resin, the filtrate was
concentrated in vacuo, and the residue was dissolved in H2O. This
solution was analyzed by reversed-phase HPLC, using a J’sphere ODS-
M80 column (4.6 × 150 mm, YMC) with a linear gradient of CH3CN
(from 23% to 41% over 20 min) in 0.1 M triethylammonium acetate
(TEAA, pH 7.0), and the yields were estimated from peak areas of the
HPLC chart.
Purification of ODNs for Measurement of MALDI-TOF/MASS
and Primer Extension Assay.18 The resulting ODNs after the
acetylenic coupling reaction were purified by reversed-phase HPLC,
using a J’sphere ODS-M80 column (10 × 150 mm, YMC) with a linear
gradient of CH3CN (from 23% to 41% over 20 min) in 0.1 M
triethylammonium acetate (TEAA, pH 7.0). Fractions were concen-
trated, and the residue was treated with 80% AcOH aq for 20 min at
room temperature. After successive coevaporation with H2O, the residue
was dissolved in H2O and washed with AcOEt (three times). The water
layer was concentrated in vacuo, and the residue was dissolved in H2O
and purified by reversed-phase HPLC using a J’sphere ODS-M80
column (10 × 150 mm, YMC) with a linear gradient of CH3CN (ODN1
and ODN4, from 23% to 41%; ODN2, ODN3, ODN5, and ODN6,
from 7% to 20%; ODN7a-c, ODN8a-c, and ODN9, from 7% to 41%
over 20 min) in 0.1 M triethylammonium acetate (TEAA, pH 7.0) to
give the corresponding ODNs without 5′-DMTr groups.
Preparation of Methylamino-Modified CPG 2000 Resin 17. In
the same manner as described for 16, treatment of CPG 2000 (500
mg, 23 µmol/g) with Boc-sarcosine (9 mg, 46 µmol) gave 17 (23 µmol/
g).
Preparation of Methylamino-Modified ArgoPore Resin 18. In the
same manner as described for 5, ArgoPore-NH2 (1.00 g, 280 µmol/g)
was treated with p-nitrophenylchloroformate (560 mg, 2.80 mmol),
followed by N,N′-dimethyl-1,6-diaminohexane (500 µL, 2.8 mmol) to
give 18 (220 µmol/g).
MALDI-TOF/MASS Spectra of ODNs.18 Spectra were obtained
on a Voyager-DE pro (PerSeptive Biosystems), and the observed
molecular weights supported their structure. ODN2: calculated mass,
C72H86N14O45P5 2022.4 (M - H); observed mass, 2021.7. ODN3:
calculated mass, C122H151N24O80P10 3543.4 (M - H); observed mass,
3544.4. ODN5: calculated mass, C72H86N14O45P5 2022.4 (M - H);
observed mass, 2022.9. ODN6: calculated mass, C122H151N24O80P10
3543.4 (M - H); observed mass, 3542.6. ODN7a: calculated mass,
C85H89N13O43P5 2135.6 (M - H); observed mass, 2136.8. ODN7b:
calculated mass, C80H97N15O42P5S 2127.6 (M - H); observed mass,
2127.9. ODN7c: calculated mass, C91H93N13O46P5 2259.7 (M - H);
observed mass, 2260.3. ODN8a: calculated mass, C85H89N13O43P5
2135.6 (M - H); observed mass, 2135.3. ODN8b: calculated mass,
C80H97N15O42P5S 2127.6 (M - H); observed mass, 2127.6. ODN8c:
calculated mass, C91H93N13O46P5 2259.7 (M - H); observed mass,
2261.8. ODN9: calculated mass, C148H162N46O76P11 4141.9 (M - H);
observed mass, 4142.5.
General Procedure of Attachment of Leader 3′-Nucleoside to
Resins 8, 9, and 19-23. To a solution of 6 (34 mg, 53.5 µmol) in
DMF (2 mL) were added EDC (10 mg, 53.5 µmol) and 16 (150 mg,
90 µmol/g), and the mixture was kept for 48 h at room temperature.
The solid support was filtered and washed with pyridine. The remaining
amino groups were capped by treatment with 0.1 M DMAP and 10%
Ac2O in pyridine. The resulting 19 was filtered and washed with EtOH
and acetone and was dried under reduced pressure. The loading amount
of the leader 3′-nucleoside unit was estimated by DMTr cation assay
to be 58 µmol/g. In the same manner as described for 19, resin-
supported leader 3′-nucleoside units 8 (6 µmol/g), 9 (50 µmol/g), 20
(8 µmol/g), 21 (49 µmol/g), 22 (7 µmol/g), and 23 (39 µmol/g) were
prepared.
Synthesis of ODNs.18 Resin-supported ODNs 12-14, 24-28, and
30 were prepared on an Applied Biosystems 392 DNA/RNA synthesizer
utilizing standard phosphoramidite chemistry at a 1.0 µmol scale. The
phosphoramidite monomers were used at a concentration of 0.1 M in
dry CH3CN. The standard DNA synthesis cycle was used, except for
Primer Extension Assay.18 A solution of template ODN (2 pmol)
and 5′-fluorescent primer ODN9 (1 pmol) in H2O (4 µL) was heated
at 100 °C for 3 min and slowly cooled to room temperature. The primer
(28) Gisin, B. F. Anal. Chim. Acta 1972, 58, 248-249.
9
J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003 11551

A R T I C L E S
Minakawa et al.
extension assay was carried out in a buffer containing 10 mM Tris-
HCl, pH 7.5, 5 mM MgCl2, 7.5 mM dithiothreitol, and 1 unit of KF
(Takara) in the presence of dNTP(s) (10 nmol each) and template ODN/
ODN9 complex at 37 °C for 1 h. The reaction was stopped by adding
40 µL of loading solution (90% formamide in tris-borate-EDTA), and
the mixture was heated at 100 °C for 5 min. One-half of this mixture
was analyzed by 20% PAGE including 7 M urea (16 × 20 × 0.1 cm,
500 V, 2 h), and the extension products were detected by a Fluorescent
Image Analyzer FLA-2000 (FUJIFILM).
Education, Science, Sports, and Culture of Japan. We would
like to thank Ms. H. Matsumoto and Ms. A. Maeda (Center for
Instrumental Analysis, Hokkaido University) for elemental
analysis. We also would like to thank Ms. S. Oka (Center for
Instrumental Analysis, Hokkaido University) for the measure-
ment of mass spectra.
Supporting Information Available: Synthetic procedures of
29a-c and HPLC profiles of on-column modification of 30 with
29c (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
Acknowledgment. This work was supported in part by a
Grant-in-Aid for Scientific Research on Priority Areas and
Encouragement of Young Scientists from the Ministry of
JA036055T
9
11552 J. AM. CHEM. SOC. VOL. 125, NO. 38, 2003