Kinetics and Mechanism of Facile and Selective Dephosphorylation of 2′-Phosphorylated and 2′-Thiophosphorylated Dinucleotides: Neighboring 3′-5′ Phoshodiester Promotes 2′-Dephosphorylation

Article abstract of DOI:10.1021/jo970021k

2′-Phosphorylated and 2′-thiophosphorylated dinucleotides U(2′-p)pU (1) and U(2′-ps)pU (2) were found to undergo facile 2′-specific dephosphorylation at 90°C in neutral aqueous solution to give UpU, and the first-order rate constants of these reactions were determined by HPLC. Particularly, U(2′-ps)pU (2, k = 1.38 ± 0.4 × 10^-3 s^-1, t_comp = 1 h) was cleanly dephosphorylated ca. 100 times more rapidly than U(2′-p)pU (1, k = 1.41 ± 0.05 × 10^-5 s^-1, t_comp = 72 h). Dephosphorylations of 1 and 2 were faster than those of thymidine 3′-phosphate (8) and thymidine 3′-thiophosphate (9), respectively. The kinetic data observed were independent of the 2′- or 3′-position of the phosphate group and the kind of base moiety. The neighboring 3′-5′ phosphodiester function most probably promotes the 2′-dephosphorylation efficiently. A branched trimer, U(2′-pU)pU (3), and related compounds having a substituent on the 2′-phosphoryl group, such as U(2′-pp-biotin)pU (4) and U(2′-ps-bimane)pU (5), were rather resistant to hydrolysis. The addition of divalent metal ions (Mg²⁺, Mn²⁺, Zn²⁺, Ca²⁺, Co²⁺, and Cd²⁺) remarkably decreased the rate of 2′-de(thio)phosphorylation of 1 or 2. Among these metal ions, Zn²⁺ most significantly inhibited the dephosphorylation. On the contrary, trivalent metal ions considerably accelerated the 2′-de(thio)phosphorylation of 1 or 2. The mechanism of 2′-dephosphorylation in the presence and absence of various metal ions is also discussed.

Full text of DOI:10.1021/jo970021k

J . Org. Chem. 1997, 62, 2813-2822
2813
Kin etics a n d Mech a n ism of F a cile a n d Selective
Dep h osp h or yla tion of 2′-P h osp h or yla ted a n d
2′-Th iop h osp h or yla ted Din u cleotid es: Neigh bor in g 3′-5′
P h osp h od iester P r om otes 2′-Dep h osp h or yla tion
Hiroyuki Tsuruoka, Koh-ichiroh Shohda, Takeshi Wada, and Mitsuo Sekine*
Department of Life Science, Tokyo Institute of Technology, Nagatsuta, Midoriku, Yokohama 226, J apan
Received J anuary 2, 1997X
2′-Phosphorylated and 2′-thiophosphorylated dinucleotides U(2′-p)pU (1) and U(2′-ps)pU (2) were
found to undergo facile 2′-specific dephosphorylation at 90 °C in neutral aqueous solution to give
UpU, and the first-order rate constants of these reactions were determined by HPLC. Particularly,
U(2′-ps)pU (2, k ) 1.38 ( 0.4 × 10-3 s-1, tcomp ) 1 h) was cleanly dephosphorylated ca. 100 times
more rapidly than U(2′-p)pU (1, k ) 1.41 ( 0.05 × 10-5 s-1, tcomp ) 72 h). Dephosphorylations of
1 and 2 were faster than those of thymidine 3′-phosphate (8) and thymidine 3′-thiophosphate (9),
respectively. The kinetic data observed were independent of the 2′- or 3′-position of the phosphate
group and the kind of base moiety. The neighboring 3′-5′ phosphodiester function most probably
promotes the 2′-dephosphorylation efficiently. A branched trimer, U(2′-pU)pU (3), and related
compounds having a substituent on the 2′-phosphoryl group, such as U(2′-pp-biotin)pU (4) and
U(2′-ps-bimane)pU (5), were rather resistant to hydrolysis. The addition of divalent metal ions
(Mg2+, Mn2+, Zn2+, Ca2+, Co2+, and Cd2+) remarkably decreased the rate of 2′-de(thio)phosphorylation
of 1 or 2. Among these metal ions, Zn2+ most significantly inhibited the dephosphorylation. On
the contrary, trivalent metal ions considerably accelerated the 2′-de(thio)phosphorylation of 1 or
2. The mechanism of 2′-dephosphorylation in the presence and absence of various metal ions is
also discussed.
In tr od u ction
during tRNA splicing, which is one of several nuclear
RNA-processing events. Similar structures have been
reviewed by Culbertson and Winey.7 With respect to
another biochemical property of the 2′-phosphate, a 2′-
phosphotransferase playing an important role in tRNA
splicing was found to transfer the 2′-phosphoryl function
from 2′-phosphorylated pre-tRNA to NAD with elimina-
tion of nicotinamide to give an ADP-ribose 1′′,2′′-cyclic
phosphate.8
Szostak and co-workers reported that, when the selec-
tion of the ribozymes bearing a 5′-thiophosphorylation
activity was performed by the in vitro selection method,
some of the products obtained were unexpectedly phos-
phorylated at the 2′-position.9 Interestingly, they also
disclosed that reverse transcription using a RNA oligo-
mer having a 2′-thiophosphoryl group as the template
caused a pause of reverse transcriptase, which ultimately
read through the point.10
Quite recently, we have found on the basis of NMR
spectroscopy and computational molecular modeling that
diribonucleotides X(2′-p)pY (X, Y ) U or A) having an
inner 2′-phosphoryl group exist in a C2′-endo conforma-
tion like that seen in the DNA duplex.11 Since the
presence of the 2′-phosphoryl group in the spliced pre-
tRNA is expected to inhibit molecular recognition of ARS
at the amino acylation step, the chemical property of such
In eukaryotic cells, RNAs transcribed from DNAs are
modified at various sites, such as base residues,1 5′-
terminal phosphate,2 and hydroxyl groups,3 to give
mature functional molecules. Among these modifica-
tions, an unprecedented product that was phosphorylated
at the 2′-hydroxyl function was first discovered by Ko-
narska4,5 in ligation products derived from circular RNA
fragments of tobacco mosaic virus RNAΩ73. This 2′-
phosphorylated RNA has a unique structure bearing two
different neighboring phosphoryl groups on the 2′,3′-cis-
diol of the nucleoside. Later, Abelson6 also reported that
a 2′-phosphorylated RNA was found in pre-tRNALeu
X Abstract published in Advance ACS Abstracts, April 1, 1997.
(1) (a) McCloskey, J . A.; Nishimura, S. Acc. Chem. Res. 1977, 10,
403. (b) Stern, L.; Schulman, L. H. J . Biol. Chem. 1978, 253, 6132. (c)
Dunn, D. B.; Hall, R. H. In Handbook of Biochemistry and Molecular
Biology, 3rd ed.; Fasman, G., Ed.; CRC Press: Cleveland, OH, 1975;
Vol. 1, pp 65-217. (d) Sprinzl, M.; Hartmann, T.; Meissner, F.; Moll,
J .; Voderwu¨lbecke, T. Nucleic Acids Res. Seq. Suppl. 1987, 15, r53. (e)
Sprinzl, M.; Hartmann, T.; Weber, J .; Blank, J .; Zeidler, R. Nucleic
Acids Res. Seq. Suppl. 1989, 17, r1. (f) Sprinzl, M.; Dank, N.; Nock,
S.; Scho¨n, A. Nucleic Acids Res. 1991, 19, 2127. (g) Limbach, P. A.;
Crain, P. F.; McCloskey, J . A. Nucleic Acids Res. 1994, 22, 2183.
(2) (a) Furuichi, Y.; Miura, K. Nature 1975, 253, 374. (b) Busch, H.;
Reddy, R.; Rothbblum, L.; Choi, C. Y. Annu. Rev. Biochem. 1982, 51,
617. (c) Reddy, R.; Busch, H. Nucleic Acids Res. 1983, 30, 127.
(3) For the 2′-methylation, see: (a) Smith, J . D.; Dunn, D. B.
Biochim. Biophys. Acta 1959, 31, 573. (b) Hall, R. H. Biochemistry
1964, 3, 876. (c) Lane, B. G. Biochemistry 1965, 4, 212. (d) Hall, R. H.
The Modified Nucleosides in Nucleic Acids; Columbia University
Press: New York, 1971. (e) Perry, R. P.; Kelly, D. E. Cell 1974, 1, 37.
(f) Both, G. W.; Banerjee, A. K.; Shatkin, A. J . Proc. Natl. Acad. Sci.
U.S.A. 1975, 72, 1189. For 2′-ribosylation, see: (g) Desgress, J .; Keith,
G.; Kuo, K. C.; Gehrke, C. W. Nucleic Acids Res. 1989, 17, 865. (h)
Kiesewetter, S.; Ott, G.; Sprinzl, M. Nucleic Acids Res. 1990, 18, 4677.
(i) Keith, G.; Glasser, A.; Desgres, J .; Kuo, K.; Keith, G. Nucleic Acids
Res. 1991, 19, 5199.
(6) Greer, C. L.; Peebles, C. L.; Gegenheimer, P.; Abelson, J . Cell
1983, 32, 537.
(7) For a comprehensive review, see: Culbertson, R. M.; Winey, M.
Yeast 1989, 5, 405.
(8) Culver, G. M.; McCraith, S. M.; Zillmann, M.; Kierzek, R.;
Michaud, M.; LaReau, R. D.; Turner, D. H.; Phizicky, E. M. Science
1993, 261, 206.
(9) Lorsch, J . R.; Szostak, J . W. Nature 1994, 371, 31.
(10) Lorsch, J . R.; Bartel, D. P.; Szostak, J . W. Nucleic Acids Res.
1995, 23, 15.
(4) Konarska, M.; Filipowicz, W.; Domdey, H.; Gross, H. J . Nature
1981, 293, 112.
(5) Konarska, M.; Filipowicz, W.; Domdey, H.; Gross, H. J . Proc.
Natl. Acad. Sci. U.S.A. 1982, 79, 1474.
(11) (a) Tsuruoka, H.; Shohda, K.; Wada, T.; Sekine, M. Tetrahedron
Lett. 1996, 37, 6741. (b) Tsuruoka, H.; Shohda, K.; Wada, T.; Sekine,
M. Nucleic Acids Symp. Ser. 1996, 21.
S0022-3263(97)00021-2 CCC: $14.00 © 1997 American Chemical Society

2814 J . Org. Chem., Vol. 62, No. 9, 1997
Tsuruoka et al.
Sch em e 1
However, in spite of the importance of the 2′-phos-
phoryl group in molecular biology, no papers dealing with
the fundamental chemistry of 2′-phosphorylated RNAs
have been published. Moreover, a few are known about
the basic hydrolytic property of alkyl thiophosphates or
nucleoside thiophosphate derivatives. Breslow et al. first
described that 4-nitrophenyl phosphorothioate under-
went hydrolysis 63 times as rapidly as 4-nitrophenyl
phosphate.42 Eckstein observed that thymidine 5′-phos-
phorothioate was hydrolyzed to thymidine to a degree of
67% on heating in 80% acetic acid at 100 °C for 40 min.40a
Quite recently, Lo¨nnberg et al.19b revealed the dethio-
phosphorylation rate of nucleoside phosphorothioate
mono- and diesters within an acidic pH range at 90 °C
for the first time in nucleic acid chemistry, as far as the
kinetic study was concerned.
In this paper, we wish to report the detailed site-
specific dephosphorylation or dethiophosphorylation of 2′-
phosphorylated or 2′-thiophosphorylated uridylyl(3′-5′)-
uridine [U(2′-p)pU (1) or U(2′-ps)pU (2)] and related
compounds under thermal but neutral conditions. The
present paper also describes the kinetics and mechanism
of these reactions in the presence or absence of metal
ions.
a 2′-phosphoryl group as well as the 3D-structure around
the 2′-phosphorylated ligation site are of great impor-
tance. In our recent studies, an effective method for the
polymer-supported synthesis of 2′-phosphorylated RNAs
has already been established.12-15 To our surprise, these
studies disclosed that, when the Tm value of a RNA
duplex having a 2′-phosphoryl group was measured, the
2′-phosphate group was hydrolyzed more rapidly than
expected at neutral pH, 90 °C. (Scheme 1).
In the field of nucleic acid chemistry, several research
groups have reported the acid or base hydrolysis of DNA/
RNA as well as nucleotide monomers. More recently,
Lo¨nnberg and his co-workers16-20 have reported detailed
studies on the understanding of the behavior of internal
and terminal phosphate esters in hydrolysis or migration.
On the other hand, much interest has recently been paid
to the effect of metal ions on hydrolysis or transesteri-
fication of phosphoester bonds of nucleic acids since it is
well known that hydrolysis catalyzed by enzymes re-
quires metal ions as cofactors.21 Binding of metal ions
to biomolecules also induces higher order structures. For
example, the mature tRNA functions only in the presence
of Mg2+ ion.22 Hitherto, it has been known that the
addition of metal ions, especially trivalent lanthanide
ions, results in promoting the hydrolysis of phosphomono-
or phosphodiester bonds in both DNA and RNA20,23-26 as
well as cyclic nucleotides.26 A number of studies on site-
specific cleavage of DNA/RNA using metal binding
ligands have been reported.27
Resu lts a n d Discu ssion
Hyd r olytic 2′-Dep h osp h or yla tion of U(2′-p )p U (1).
In our continuing studies on chemical synthesis and
properties of 2′-phosphorylated RNAs, we found that the
2′-phosphorylated uridylate dimer U(2′-p)pU (1) under-
went facile site-specific dephosphorylation upon simple
heating in aqueous solution to give the corresponding
UpU dimer (12a ) (Scheme 2). The detailed study showed
that this reaction proceeded in 0.1 M ammonium acetate
(pH 7.0) at 90 °C to give 12a in ca. 80% yield along with
small amounts of hydrolyzed and isomerized products
[14a ,b (<5%) and 12b (ca. 10%)] derived from UpU after
49 h, as shown in Figure 1A2.
In this reaction, the 2′-phosphomonoester bond was
first cleaved, since products such as U(2′-p)p (16a )
directly hydrolyzed at the internal phosphodiester bond
could not be detected (see Supporting Information for
more detail). As the disappearance of U(2′-p)pU (1)
strictly obeyed first-order kinetics, the rate constant (k)
(12) Sekine, M.; Iimura, S.; Furusawa, K. J . Org. Chem. 1993, 58,
3204.
(13) Sekine, M. In Studies in Natural Product Chemistry; Rahmann,
A.-U., Ed.; Elsevier: Tokyo, 1994; Vol. 14, pp 283-331.
(14) Sekine, M.; Tsuruoka, H.; Iimura, S.; Wada, T. Nat. Prod. Lett.
1995, 5, 41.
(15) Sekine, M.; Tsuruoka, H.; Iimura, S.; Kusuoku, H.; Wada, T.
J . Org. Chem. 1996, 61, 4087.
(16) (a) Oivanen, M.; Lo¨nnberg, H. J . Org. Chem. 1989, 54, 2556.
(b) Oivanen, M.; Lo¨nnberg, H. Trends Org. Chem. 1991, 2, 183. (c)
J a¨rvinen, P.; Oivanen, M.; Lo¨nnberg, H. J . Org. Chem. 1991, 56, 5396.
(d) Mikhailov, S. N.; Oivanen, M.; Oskman, P.; Lo¨nnberg, H. J . Org.
Chem. 1992, 57, 4122.
(17) (a) Oivanen, M.; Shnell, R.; Pfleiderer, W.; Lo¨nnberg, H. J . Org.
Chem. 1991, 56, 3623. (b) Oivanen, M.; Lo¨nnberg, H. Acta Chem.
Scand. 1991, 45, 968. (c) Oivanen, M.; Mikhailov, S. N.; Florentiev, V.
L.; Vihanto, P.; Lo¨nnberg, H. Acta Chem. Scand. 1993, 47, 622. (d)
Kuusela, S.; Lo¨nnberg, H. J . Chem. Soc. Perkin Trans. 2 1994, 2109.
(18) (a) Kosonen, M.; Lo¨nnberg, H. J . Chem Soc. Perkin Trans. 2
1995, 1203. (b) Kosonen, M.; Oivanen, M.; Lo¨nnberg, H. J . Org. Chem.
1994, 59, 3704.
(19) (a) Oivanen, M.; Mikhailov, S. N.; Padyukava, N. S.; Lo¨nnberg,
H. J . Org. Chem. 1993, 58, 1617. (b) Oivanen, M.; Ora, M.; Almer, H.;
Stro¨mberg, R.; Lo¨nnberg, H. J . Org. Chem. 1995, 60, 5620. (c) Ora,
M.; Oivanen, M.; Lo¨nnberg, H. J . Chem Soc. Perkin Trans. 2 1996,
771.
(24) (a) Butcher, W. W.; Westheimer, F. H. J . Am. Chem. Soc. 1955,
77, 2420. (b) Dimroth, K.; Witzel, H.; Hu¨lsen, W.; Mirbach, H. Ann.
Chem. 1959, 620, 94. (c) Trapmann, H.; Devani, M. Hoppe-Seylers Z.
Physiol. Chem. 1965, 340, 81-86. (d) Eichhorn, G. L.; Butzow, J . J .
Biopolymers 1965, 3, 79. (e) Dimroth, K.; J aenicke, L.; Heinzel, D. Ann.
Chem. 1950, 566, 206. (f) Dimroth, K.; J aenicke, L.; Vollbrechtshausen,
I. Z. Physiol. Chem. 1952, 289, 71. (g) Dimroth, K.; Witzel, H. Ann.
Chem. 1959, 620, 109. (h) Witzel, H. Ann. Chem. 1959, 620, 122. (i)
Huff, J . W.; Sastry, K. S.; Gordon, M. P.; Wacker, W. E. C. Biochemistry
1964, 3, 501.
(25) (a) Komiyama, M.; Matsumoto, Y.; Hayashi, N.; Matsumura,
K.; Takeda, N.; Watanabe, K. Polym. J . 1993, 25, 1211. (b) Shiiba, T.;
Yonezawa, K.; Takeda, N.; Matsumoto, Y.; Yashiro, M.; Komiyama,
M. J . Mol. Catal. 1993, 84, L21. (c) Hayashi, N.; Takeda, N.; Shibata,
T.; Yashiro, M.; Watanabe, K.; Komiyama, M. Inorg. Chem. 1993, 32,
5899. (d) Sumaoka, J .; Shiiba, T.; Komiyama, M. J . Mol. Catal. 1993,
78, L53.
(26) (a) Breslow, R.; Huang, D. L. Proc. Natl. Acad. Sci. U.S.A. 1991,
88, 4080. (b) Sumaoka, J .; Yashiro, M.; Komiyama, M. J . Chem. Soc.,
Chem. Commun. 1992, 1707.
(27) (a) Han, H.; Dervan, P. B. Nucleic Acids Res. 1994, 22, 2837
and references cited therein. (b) Magda, D.; Miller, R. A.; Sessler, J .
L.; Iverson, B. J . Am. Chem. Soc. 1994, 116, 7439 and references cited
therein. (c) Inokawa, T.; Komiyama, M. J . Biochem. 1994, 116, 719.
(20) (a) Kuusela, S.; Lo¨nnberg, H. J . Phys. Org. Chem. 1992, 5, 803.
(b) Kuusela, S.; Lo¨nnberg, H. J . Phys. Org. Chem. 1993, 6, 347. (c)
Kuusela, S.; Lo¨nnberg, H. J . Chem. Soc. Perkin Trans. 2 1994, 2301.
(d) Kuusela, S.; Azhayev, A.; Guzaev, A.; Lo¨nnberg, H. J . Chem. Soc.
Perkin Trans. 2 1995, 1197.
(21) Gani, D.; Wilkie, J . Chem. Soc. Rev. 1995, 55.
(22) For a comprehensive review, see: Schimmel, P. R.; Redfield,
A. G. Annu. Rev. Biophys. Bioeng. 1980, 9, 181.
For a comprehensive review, see: Knorre, D. G.; Vlassov, V. V.;
Zarytova, V. F.; Lebedev, A. V.; Federova, O. S. Design and Targeted
Reactions of Oligonucleotide Derivatives; CRC Press, Inc.: Boca Raton,
FL, 1994.
(23) (a) Butzow, J . J .; Eichhorn, G. L. Biopolymers 1965, 3, 95.

Dephosphorylation of Phosphorylated Dinucleotides
J . Org. Chem., Vol. 62, No. 9, 1997 2815
Ch a r t 1
Sch em e 2
Sch em e 3
of this 2′-dephosphorylation was calculated to be 1.41 (
0.05 × 10-5 s-1 (tcomp ) 72 h) by HPLC analysis. To study
the mutual effect between the proximal 2′- and 3′-
bisphosphoryl groups of 1, the rate constant of thymidine
3′-phosphate (8, Tp) as a reference compound was also
determined under the same conditions to be 7.5 ( 0.2 ×
10-5 s-1. Consequently, it turned out that 2′-dephosphor-
ylation of U(2′-p)pU (1) occurred about 2 times faster
than that of Tp (8) and considerably faster than that of
any other simple monoalkyl phosphates and nucleoside
monophosphates reported previously.16,20b,28
Next, 3′-dephosphorylation of U(3′-p)pU (10), in which
the 2′- and 3′-substituents of 1 were interchanged, was
examined to study the effect of the regioisomer on the
hydrolysis of the 3′-phosphate group. U(3′-p)pU (10) was
prepared by the use of 3′-phosphorylation of appropri-
ately protected U(2′-5′)pU derivative (19b) having a 2′-
5′ phosphodiester linkage followed by deprotection. This
kind of phosphorylation was first reported by Caruthers
and Kierzek.29 In connection with the 2′-phosphorylation,
Chattopadhyaya and his co-workers reported the bis-
(cyanoethyl) phosphorylation as an intermediate for the
synthesis of branched RNA (Scheme 3).30 Quite recently,
Kierzek also reported the synthesis of U(2′-p)pU (1) by
this reaction.31
Table 3 shows that there was little regiospecificity in
hydrolysis between 1 and 10 since the rate constant of
(28) (a) Bunton, C. A.; Llewellyn, K. G.; Oldham, K. G.; Vernon, C.
A. J . Chem. Soc. 1958, 3574. (b) Bacher, J . E.; Allen, F. W. J . Biol.
Chem. 1950, 182, 701. (c) Hayes, D. H. J . Chem. Soc. 1960, 1184. (d)
Takami, A.; Imazawa, M.; Irie, M.; Ukita, T. J . Pharm. Soc. J pn. 1965,
85, 658. (e) Bacher, J . E.; Allen, F. W. J . Biol. Chem. 1951, 188, 59. (f)
Baddiley, J .; Bachanan, J . G.; Letters, R. J . Chem. Soc. 1958, 1000.
(g) Bamann, E.; Meisnheimer, M. Chem. Ber. 1938, 71, 1711. (h)
Bailley, M. C. Bull. Soc. Chim. 1942, 9, 421.
(29) Kierzek, R.; Kopp, D. W.; Edmonds, M.; Caruthers, M. H.
Nucleic Acids Res. 1986, 14, 4751.
(30) (a) Zhou, X.; Remand, G.; Chattopadhyaya, J . Tetrahedron 1988,
44, 6471. (b) Sund, C.; Agback, P.; Chattopadhyaya, J . Tetrahedron
1988, 47, 9659.
(31) Kierzek, R. Nucleosides Nucleotides 1994, 13, 1757.

2816 J . Org. Chem., Vol. 62, No. 9, 1997
Tsuruoka et al.
F igu r e 1. Reversed phase HPLC profiles of the mixtures
obtained by the thermolysis of U(2′-p)pU (1) (A, after 71 h),
U(3′-p)pU (10) (B, after 1 h), and U(2′-ps)pU (2) (C, after 72
h) in 0.1 M AcONH4 (pH 7.0) at 90 °C. The UV detection was
performed at 254 nm.
the dephosphorylation of 10 was similar to that of
U(2′-p)pU (1) (Figure 1B). The reason why the k value
of U(3′-p)pU (10) was slightly greater than that of 1
might be ascribed to the difference in basicity between
the 2′ and 3′-oxygens of 1 and 10.
F igu r e 2. Time-dependent product distribution for de(thio)-
phosphorylation of U(2′-p)pU (1) (A) and U(2′-ps)pU (2) (B) in
0.1 M AcONH4 (pH 7.0) at 90 °C. The concentration of 1 or 2
was 1.5 × 10-4 mol dm-1; (b) 1 or 2, (O) UpU (13a ).
Interaction between the uracil residue and the 2′-
phosphoryl group should also be taken into account. The
2′-phosphoryl group of U(2′-p)pU (1) may be sterically
associated with the 3′-downstream base moiety. How-
ever, the kinetic data of U(2′-p)pA (6), which was
prepared by our method reported previously15 (Scheme
4), did not support this possibility, as shown in Table 4.
Accordingly, the promoted dephosphorylation was only
due to the effect of a neighboring 3′-5′ phosphodiester.
The rate enhancement promoted by a proximal phos-
phodiester was most probably attributed to protonation
on the 2′-esterified oxygen atom through route B depicted
in Scheme 5. A similar mechanism has been quite
recently proposed by our laboratory, suggesting that the
3′-5′ phosphodiester promotes the hydrolytic deprotec-
tion of the 2′-tert-butyldimethylsilyl group in synthetic
RNA oligomers under acidic conditions due to the en-
hancement of protonation on the oxygen of the 2′-silyl
ether bond.32
detail, the related compounds bearing substituents on the
2′-phosphoryl group, such as 2′-phosphodiester (3), py-
rophosphate (4), and thiophosphodiester (5) derivatives,
were investigated in the same way (Chart 1).
First, to study the thermal stability of the 2′-phos-
phodiester derivative U(2′-pU)pU (3), it was synthesized
by the use of a bifunctional 2′,3′-bisphosphoramidite
derivative reported by Damha and Ogilvie.33 In fact,
U(2′-pU)pU (3) having two vicinal phosphodiester link-
ages was hardly hydrolyzed even after 1 week, similarly
to the case of d(UpU) (11) with an internucleotidic bond
which is very stable in neutral conditions (Figure 3A).
Furthermore, a 2′-dialkyl pyrophosphate derivative,
U(2′-pp-biotin)pU (4), was also unexpectedly fairly stable
as can be seen in Figure 3B. The 2′-thiophosphodiester
derivative 5 was not as stable as U(2′-pp-biotin)pU (4),
and its decomposition profile was complicated. (Figure
3C). However, it was more stable than U(2′-p)pU (1). The
mechanism of hydrolysis of these compounds was as-
In order to examine the relative effect of the 3′,5′-
phosphodiester function on the 2′-hydrolysis in more
(33) (a) Damha, M. J .; Pon, R. T.; Ogilvie, K. K. Tetrahedron Lett.
1995, 26, 4839. (b) Damha, M. J .; Ogilvie, K. K. J . Org. Chem. 1988,
53, 3710. (c) Damha, M. J .; Ganeshan, K.; Hudson, R. H. E.; Zabarylo,
S. V. Nucleic Acids Res. 1992, 20, 6565.
(32) (a) Kawahara, S.; Wada, T.; Sekine, M. Nucleic Acids Symp.
Ser. 1995, 34, 29. (b) Kawahara, S.; Wada, T.; Sekine, M. J . Am. Chem.
Soc. 1996, 118, 9461.

Dephosphorylation of Phosphorylated Dinucleotides
J . Org. Chem., Vol. 62, No. 9, 1997 2817
Sch em e 4
F igu r e 3. Reversed phase HPLC profiles of the mixtures
obtained by the thermolysis of 2′,3′-bisphosphorylated deriva-
tives U(2′-pU)pU (3) (A, after 168 h), U(2′-pp-biotin)pU (4) (B,
after 24 h), and U(2′-ps-bimane)pU (5) (C, after 24 h). The UV
detection was performed at 254 nm.
Ta ble 1. F ir st-Or d er Ra te Con sta n ts for
Dep h osp h or yla tion of 1 a n d 8 in th e P r esen ce a n d
Absen ce of Meta l Ion s a t p H 7.0, 90 °C
Sch em e 5
U(2′-p)pU (1)
k (10-6 s-1
14.1 ( 0.5c
Tp (8)
additives
)
t1/2 (h) k (10-6 s-1
)
t1/2 (h)
nonea
Ca2+
Mg2+
Co2+
Cd2+
Mn2+
Zn2+
La3+
Y3+
13.6
14.5
16.4
18.6
19.0
66.9
84.0
7.4
7.5 ( 0.2
25.7
13.3 ( 0.1
11.8 ( 1.0
10.3 ( 0.5
10.1 ( 0.4
2.9 ( 0.2
6.1 ( 0.2
31.6
4.3 ( 0.2
5.2 ( 0.8
5.4 ( 0.5
5.3 ( 0.5
44.4
36.9
35.9
36.8
2.3 ( 0.2
25.9 ( 0.4
16.9 ( 0.1
11.4
17.2
phosphate bufferb 11.2 ( 0.2
a
b
0.1 M AcONH4 (pH 7.0) was used. 10 mM sodium phosphate
buffer (pH 7.0) was used as the solvent. c Standard deviation.
Ta ble 2. F ir st-Or d er Ra te Con sta n ts for
Deth iop h osp h or yla tion of 2 a n d 9 in th e P r esen ce a n d
Absen ce of Meta l Ion s a t p H 7.0, 90 °C
U(2′-ps)pU (2)
k (10-6 s-1
1380 ( 40c
Tps (9)
additives
)
t1/2 (h) k (10-6 s-1
)
t1/2 (h)
nonea
Mg2+
0.14
0.27
0.76
3.5
230 ( 10
0.82
0.90
0.99
3.19
710 ( 60
250 ( 10
56 ( 8
210 ( 30
190 ( 10
60 ( 3
Mn2+
Zn2+
phosphate bufferb
820 ( 40
0.24
a
b
0.1M AcONH4 (pH 7.0) was used. 10 mM sodium phosphate
buffer (pH 7.0) was used as the solvent. c Standard deviation.
sumed to be a nucleophilic substitution on phosphorus
to give the usual pentacoordinated phosphorus interme-
diates.34 Accordingly, the rate of hydrolysis was consid-
erably slow. The substituted 2′-phosphoryl group of this
kind was different from that of the 2′-phosphomonoester
types as far as their stability and mechanism are
concerned.
Hyd r olytic 2′-Dep h osp h or yla tion of U(2′-p s)p U.
As a substrate for the present study, 2′-thiophosphory-
lated uridylate dimer U(2′-ps)pU (2) was also examined.
This compound was prepared according to the procedure
reported previously.15 Dephosphorylation of U(2′-ps)pU
(2) proceeded site-specifically like that of U(2′-p)pU (1)
and was completed in only 1 h at 90 °C.
Interestingly, we found that the dephosphorylation of
U(2′-ps)pU (2) was much more rapid (k ) 1.38 ( 0.4 ×
10-3 s-1) than that of U(2′-p)pU (1). In general, monoalkyl
thiophosphates are known to be more unstable than the
corresponding phosphates.41,42 Lo¨nnberg et al.19c also
reported that uridine 2′- and 3′-thiophosphate (2′/3′-
UMPS) underwent dethiophosphorylation to give uridine
200-300-fold more rapidly than dephosphorylation of the
corresponding phosphate. Moreover, they disclosed that
dethiophosphorylation of 2′/3′-UMPS did not accompany
(34) Cox, J . R.; Ramsay, O. B. Chem. Rev. 1964, 317.

2818 J . Org. Chem., Vol. 62, No. 9, 1997
Tsuruoka et al.
Ta ble 3. F ir st-Or d er Ra te Con sta n ts for
Dep h osp h or yla tion of 10 in th e P r esen ce a n d Absen ce of
Meta l Ion s a t p H 7.0, 90 °C
A
U(3′-p)pU (10)
additives
k (10-6 s-1
)
t1/2 (h)
nonea
Mg2+
Mn2+
Zn2+
15.2 ( 0.4b
10.4 ( 0.3
4.2 ( 0.3
3.8 ( 0.1
12.7
18.4
45.5
50.4
a
b
0.1 M AcONH4 (pH 7.0) was used. Standard deviation.
desulfurization at all over an acidic pH range (pH 2-7),19c
although UpsU was desulfurized over the hydrolytic
cleavage of the internucleotidic linkage in the thermal
hydrolysis within a similar pH range. No desulfurization
of (2) giving rise to 1 similarly occurred under these
conditions. It was also observed that U(2′-ps)pU (2) was
hydrolyzed ca. 6 times faster than thymidine 3′-thiophos-
phate (9, Tps) at pH 7.0 as depicted in Table 2 and ca.
22 times faster than uridine 2′-thiophosphate (k ) ca. 6
× 10-5) at pH 6.5 as reported by Lo¨nnberg et al.19c
Noteworthy is the very simple reaction shown in Figure
1C, which illustrates the HPLC profile of the reaction.
In fact, the present facile dephosphorylation of 2 pro-
duced UpU (12a ) in 98% yield after 1 h. This result
suggests a new possibility that the thiophosphoryl group
might serve as a tentative 2′-hydroxyl-protecting group
as well as a unique thermally removable group.
B
On the other hand, U(2′-ps)pA (7) in which the 3′-
downstream nucleoside of 2 was replaced with adenosine
was synthesized and subjected to hydrolysis. As a result,
there is only a slight difference in the rates of dethio-
phosphorylation between 2 and 7 as mentioned in the
dephosphorylation of 1 and 6.
p H Dep en d en cy of 2′-Dep h osp h or yla t ion of
U(2′-p )p U a n d U(2′-p s)p U. The kinetic rates of 2′-
dephosphorylation of 1 and 2 in the range of pH 2-8 are
depicted in Figure 4. As shown in Figure 4A, the
dephosphorylation rate of 1 at 90 °C was highest at near
pH 4 and almost in proportion to the concentration of
the monoanion species of 1. This result strongly sug-
gested that the rate of 2′-dephosphorylation should
depend on the population of this species as described in
the hydrolysis of phosphomonoesters.28a,34 A detailed
mechanism of 2′-dephosphorylation is described below.
Dethiophosphorylation of U(2′-ps)pU (2) at 90 °C was too
rapid to measure the kinetic rate (tcomp < 10 min at pH
5.8, 90 °C), so that the reaction was carried out at 70 °C.
As shown in Figure 4B, the pH dependency of dethio-
phosphorylation of U(2′-ps)pU (2) also gave a similar
result, showing the highest rate at pH 4.0. However, the
rate constant varied considerably compared with that of
1. For instance, the k value of dephosphorylation of 2 at
pH 4 (k ) 7.4 ( 0.4 × 10-4 s-1) was ca. 23 times greater
than that at pH 8 (k ) 3.2 ( 0.1 × 10-5 s-1), while 1
showed a lesser difference of 1.6 times.
Mech a n ism of 2′-Dep h osp h or yla tion of U(2′-p )p U
a n d U(2′-p s)p U. The dissociative mechanism proposed
for the typical hydrolysis of general phosphomonoesters
has been reported as follows:16a,b,28 Transfer of a proton
from the hydroxyl ligand of a phosphomonoester to the
esteric oxygen atom results in hydrolysis to an alcohol
and metaphosphate through the P-O bond fission. The
reactive species is a monoanionic phosphoester that is
in equilibrium with the esteric O-protonated form (24a )
via route A depicted in Scheme 3. The pH range most
effective for the intramolecular protonation is known to
F igu r e 4. pH dependency of the de(thio)phosphorylation rate
of U(2′-p)pU (1) at 90 °C (A) and U(2′-ps)pU (2) at 70 °C (B).
Conditions: 10 mM NaH2PO4 (pH 2.5 and 5.8), 0.1 M citric
acid-0.2 M Na2HPO4 (pH 4.0), 0.1 M NH4OAc (pH 7.0), 0.1 M
Tris-HCl (pH 8.0).
be near pH 4.28a However, it was noteworthy that the
present dephosphorylation of U(2′-p)pU (1) proceeded
smoothly even at pH 7.0. Lo¨nnberg and co-workers
reported that the rate constant of uridine 2′-phosphate
in its hydrolysis at pH 5.6 was 0.87 ( 0.02 × 10-5 s-1 20b
.
This finding and our present results suggested that the
facile dephosphorylation of 1 should involve a somewhat
different mechanism due to the neighboring group par-
ticipation of the 3′-5′ phosphodiester or the base moiety
and the substituent effect around the 2′-phosphate group.
Particularly, the proton transfer of the 2′-phosphate via
route B involving the help of the 3′-5′ phosphodiester
oxygen should be considered as a plausible explanation
of the dephosphorylation of 1 based on the above experi-
ment.
It is known that the sulfur atom in thiophosphate
esters stabilizes a transition state in the dissociative
mechanism and destabilizes that of the associative mech-
anism.43 Accordingly, the P-O split of the protonated
intermediate 24a to give the hydrolyzed product requires
more activation energy than that of 24b. Since this
knowledge is in good agreement with the present result
that 2 undergoes the 2′-dephosphorylation more rapidly
than 1, both the reactions were similarly explained in

Dephosphorylation of Phosphorylated Dinucleotides
J . Org. Chem., Vol. 62, No. 9, 1997 2819
terms of the mechanism via metaphosphate35 or meta-
thiophosphate.36,19c
Ta ble 4. F ir st-Or d er Ra te Con sta n ts of Hyd r olysis in
th e P r esen ce a n d Absen ce of Meta l Ion s a t p H 7.0, 90 °C
(k, 10-6 s-1
)
E ffect of Met a l Ion s on Hyd r olyt ic 2′-Dep h os-
p h or yla tion . As the 2′-phosphorylated and 2′-thiophos-
phorylated ribonucleotides 1 and 2 have two different
proximal phosphoryl groups bearing several binding sites
of various metal ions, the effect of these compounds on
the 2′-dephosphorylation rate in the presence of metal
ions was expected. We have chosen several metal ions
which can be completely dissolved in 0.1 M ammonium
acetate to avoid inherent heterogeneous effects. Table
2 summarizes the kinetic data of the 2′-dephosphoryla-
tion of these compounds in the presence of various metal
ions as additives. Similar hydrolytic dephosphorylations
using Tp (8) and Tps (9) were studied as control experi-
ments. Kuusela and Lo¨nnberg reported that some di-
and trivalent metal ions promoted the hydrolysis of
ribonucleoside phosphoesters under slightly acidic condi-
tions without enhancement of intramolecular phosphate
migration.20b In marked contrast to these observations,
it was interestingly found that, at the neutral pH we used
in this study, the rate constants of Tp (8) were slightly
decreased by addition of various divalent metal ions
(Mg2+, Mn2+, Zn2+, Ca2+, Co2+, and Cd2+), as can be seen
in Table 1. Compared with these results, the rate
constants of the 2′-dephosphorylation of U(2′-p)pU (1) in
the presence of metals considerably decreased under the
same conditions. Among the metal ions tested, Mn2+ and
Zn2+ ions inhibited the hydrolysis most effectively. It
might be concluded that apparently the proximal 3′-5′
phosphodiester anion strongly interferes with this hy-
drolysis.
additives
2
3
4
6
7
11
nonea
Mg2+
Mn2+
Zn2+
ndd
nd
nd
nd
nd
b
15.1 ( 0.1c
1450 ( 30
nd
nd
nd
nd
a
b
0.1 M AcONH4 (pH 7.0) was used. The reaction was compli-
cated. c Standard deviation. The term “nd” indicates that the
d
hydrolyzed product could not be detected.
the intermediate protonated on the 2′-esteric oxygen
atom. According to Schwing-Weil’s recent finding, a
phenol derivative, in which two phosphonate groups are
bound to the 2- and 5-positions, forms a rigid complex
with Zn2+ and Ni2+ ions.39
Con clu sion s
The results of dephosphorylation of Tp (8) and Tps (9)
as reference compounds under neutral conditions showed
that phospho- or thiophosphomonoester derivatives were
not essentially more stable than phosphodiesters. Es-
pecially, hydrolytic dephosphorylation of thiophospho-
monoester Tps (9) occurred rapidly. In the present study,
U(2′-p)pU (1) and U(3′-p)pU (10), having both the vicinal
phosphodiester and phosphomonoester bonds, were found
to undergo more facile dephosphorylation only at the
monoester site than the other usual phosphomonoesters
due to acid catalysis by the neighboring effect of the
phosphodiester function. The most rapid hydrolysis was
observed in the case of U(2′-ps)pU. One of the reasons
for the promoted de(thio)phosphorylation might be ex-
plained by the fact that the 2′,3′-bisphosphoryl groups
are fixed rigidly by the five-membered ring sugar in the
case of RNAs.
Although a lot of nucleoside phosphorothioates of the
monoester type have been synthesized40 and utilized in
biological applications such as inhibitors and enzyme-
resistant substrates,41 the chemical properties of these
materials have not been elucidated until quite recently,
especially as far as the basic kinetics of their hydrolytic
behavior are concerned. Since Breslow and Katz42 first
noted that thiophosphate monoesters undergo nucleo-
philic displacement reactions more rapidly than their oxy
counterparts, few studies on their displacement reactions
involving hydrolysis have appeared to date.34 Cullis and
co-workers reported the stereochemical thiophosphoryl-
transfer reaction of 18O-labeled 4-nitrophenyl phospho-
rothioate in ethanol36a and also noted the hydrolysis of
On the other hand, the 2′-phosphomonoester hydrolysis
of U(2′-p)pU (1) in the presence of trivalent metal
ions20a,b,24,25 (La3+ and Y3+) was markedly accelerated, but
the 3′-phosphomonoester hydrolysis of Tp (8) was not
affected significantly. In any case, the produced UpU
(12a ) was rapidly hydrolyzed to uridine 15 and Up 14a ,b
followed by further dephosphorylation of 14a ,b to uridine
15. Consequently, the HPLC profiles of dephosphoryla-
tion of U(2′-p)pU (1) observed showed a large amount of
uridine 15.
Furthermore, the 2′-dethiophosphorylation of U(2′-ps)-
pU (2) was more greatly suppressed by divalent metal
ions than the 2′-dephosphorylation of U(2′-p)pU (1).
Particularly, addition of Zn2+ ion to U(2′-ps)pU (2) led to
1
a severe repression with a
/
rate constant compared
25
with the hydrolysis of U(2′-ps)pU (2) in the absence of
metal ions, while there is a difference of only ca. 4 times
in the rate constant between the hydrolysis reactions of
Tps measured in the presence and absence of Zn2+ ion.
It is known that Zn2+ ion exists in the active site of
alkaline phosphatases as a cofactor.37 Kimura et al. have
recently reported enzyme models having Zn2+ chelates
for dephosphorylation of alkyl monophosphates.38
Divalent metal ions such as Mn2+ and Zn2+ could each
bind to two proximate phosphoryl groups. Therefore, the
formation of a relatively stable nine-membered ring
containing the two phosphoryl groups and a divalent
metal ion could probably interrupt the decomposition of
2,4-dinitrophenyl phosphorothioate at pH 12.7 giving the
rate constant of 6.9 × 10-4 s-1
.
Lo¨nnberg and his co-
36b
workers19b described that, in the thermal hydrolysis of
UpsU at pH 6-8 at 363.2 K, uridine 2′- and 3′-phos-
phorothioates accumulated in addition to the products
obtained under more acidic conditions, but dephosphor-
ylation of these compounds was not described probably
because of the very complicated distribution of products
as suggested in their paper.19b Later, they disclosed the
(39) Schwing-Weil, M.; Chafaa, S.; Meullemeestre, J .; Vierling, F.
J . Chem. Res. 1995, 258.
(35) Westheimer, F. H. Chem. Rev. 1981, 81, 313.
(36) (a) Cullis, P. M.; Iagrossi, A. J . Am. Chem. Soc. 1986, 108, 7870.
(b) Burgess, J .; Blundell, N.; Cullis, P. M.; Hubbard, C. D.; Misra, R.
J . Am. Chem. Soc. 1988, 110, 7900.
(37) Anderson, R. A.; Bosron, W. F.; Kennedy, F. S.; Vallee, B. L.
Natl. Proc. Acad. Sci. U.S.A. 1975, 72, 2989.
(40) (a) Eckstein, F. J . Am. Chem. Soc. 1966, 88, 4292. (b) Eckstein,
F. J . Am. Chem. Soc. 1970, 92, 4718. (c) Ozturk, D. H.; Park, I.;
Colman, R. F. Biochemistry 1992, 31, 10544.
(41) (a) Eckstein, F.; Sternbach, H. Biochim. Biophys. Acta 1967,
146, 618. (b) Eckstein, F. Annu. Rev. Biochem. 1985, 54, 367. (c)
Eckstein, F. Angew. Chem., Int. Ed. Engl. 1983, 22, 423.
(42) Breslow, R.; Katz, I. J . Am. Chem. Soc. 1968, 90, 7376.
(38) Koike, T.; Kimura, E. J . Am. Chem. Soc. 1991, 113, 8935.

2820 J . Org. Chem., Vol. 62, No. 9, 1997
Tsuruoka et al.
TMS) δ 1.50-1.72 (m, 12H), 2.62-2.85 (m, 6H), 3.44-3.59 (m,
2H), 3.76-3.81 (m, 6H), 4.16-4.37 (m, 2H), 4.49-4.76 (m, 4H),
5.29-5.40 (m, 2H), 5.47-5.52 (m, 1H), 6.03-6.26 (m, 2H), 6.33,
6.35 (2d, J ) 6.3, 5.3 Hz, 1H), 6.50, 6.52 (2d, J ) 5.3, 4.6 Hz,
1H), 6.84-6.89 (m, 4H), 7.16-7.69 (m, 24H), 7.83-8.00 (m,
11H), 8.36, 8.41 (2s, 1H), 8.68, 8.69 (2s, 1H); 13C NMR (67.8
MHz, CDCl3) δ 19.48, 19.57, 20.40, 26.67, 26.72, 26.76, 26.92,
26.96, 27.17, 27.23, 31.45, 31.50, 31.56, 31.74, 31.77, 55.22,
55.26, 62.79, 62.84, 62.88, 62.91, 63.00, 70.78, 73.87, 74.05,
81.24, 83.02, 83.06, 83.18, 83.29, 83.79, 83.85, 83.97, 86.61,
86.92, 88.00, 102.79, 113.50, 116.48, 116.60, 116.64, 116.91,
125.25, 127.33, 127.71, 128.10, 128.18, 128.23, 128.43, 128.52,
128.64, 128.77, 128.90, 128.99, 129.08, 129.42, 129.76, 129.85,
130.14, 130.23, 130.64, 131.32, 133.05, 133.80, 133.87, 133.94,
134.45, 134.48, 134.74, 135.04, 137.83, 140.14, 140.20, 143.49,
143.76, 149.29, 152.13, 152.51, 152.72, 158.83, 161.87, 165.00,
165.14, 165.21, 165.26, 168.43, 172.18, 172.25; 31P NMR (109
MHz, CDCl3, 85% H3PO4) δ -2.20, -1.92 (2s, 1P), 51.2 (s, 1P).
Anal. Calcd for C88H80N10O21P2S: C, 61.90; H, 4.72; N, 8.20;
S, 1.88. Found: C, 62.03; H, 5.02; N, 7.77; S, 1.61.
detailed and basic kinetic data of hydrolytic dethiophos-
phorylation of uridine 2′- and 3′-thiophosphates under
similar conditions in their subsequent paper.19c
It is anticipated that new methodologies using this
characteristic of the thiophosphate group will be devel-
oped especially in molecular biology, since thiophos-
phoryl-transfer reactions have been established by using
some enzymatic systems.41c,44 In consideration of these
facts, possibly in extreme thermophilic bacteria, such as
Thermus aquaticus and Thermus thermophilus, 2′-phos-
phorylated RNA species might not be found in cells, since
spliced tRNA containing a 2′-phosphate should be de-
phosphorylated by the above-mentioned thermal dephos-
phorylation as well as a tRNA-specific phosphatase. In
fact, the reason why there are only a few discoveries of
2′-phosphorylated RNA species to date may be supported
by the dephosphorylation described above.
Dep r otection of F u lly P r otected 2′-P h osp h or yla ted
Din u cleotid e U(2′-p )p A (6). The fully protected dinucleotide
U(2′-p)pA (23a ) (29 mg, 0.017 mmol) was dissolved in pyridine
(0.5 mL) after repeated coevaporation with dry pyridine (2 mL
× 3). To the solution were added BSA (0.1 mL, 0.4 mmol) and
DBU (0.05 mL, 0.05 mmol), and the mixture was stirred for
10 min at room temperature. To the mixture was added water
(0.05 mL). After being stirred for a few minutes, the solution
was treated with iodine (44 mg, 0.17 mmol). After being
stirred for an additional 48 h, the mixture was diluted with
water. Then, extraction was performed by use of H2O (2 mL)-
Et2O (2 mL), and the aqueous layer was concentrated under
reduced pressure. The residue was dissolved in concd NH3-
pyridine (9:1, v/v, 5 mL). After stirring for 20 h, the mixture
was coevaporated with water, and the resulting residue was
treated with 80% acetic acid (5 mL) for 15 min. The solution
was concentrated under reduced pressure, and then the
residue was dissolved in water (2 mL) and washed with Et2O
(2 mL). The aqueous layer was collected and chromatographed
on Whatman 3MM Chr papers with iPrOH-concd NH3-H2O
(6:1:3, v/v/v) to give pure material 6 (248 A260, 58%): 1H NMR
(400 MHz, D2O, DSS) δ 3.69 (dd, J ) 3.4, 12.8 Hz, 1H), 3.73
(dd, J ) 12.8 Hz, 1H), 4.17 (ddd, J ) 4.1, 5.9, 11.8 Hz, 1H),
4.21 (ddd, J ) 2.9, 5.2, 11.7 Hz, 1H), 4.25 (ddd, J ) 2.6, 3.4,
4.0 Hz, 1H), 4.37 (ddd, J ) 2.9, 4.1, 4.3 Hz, 1H), 4.56 (dd, J )
4.3, 5.2 Hz, 1H), 4.71 (ddd, J ) 2.6, 5.2, 7.3 Hz, 1H), 4.77 (m,
1H), 4.79 (dd, J ) 5.2, 5.6 Hz, 1H), 5.85 (d, J ) 8.0 Hz, 1H),
5.94 (d, 1H, J ) 6.7 Hz), 6.10 (d, 1H, J ) 5.6 Hz), 7.75 (d, 1H,
J ) 8.0 Hz), 8.24 (s, 1H), 8.45 (s, 1H); 13C NMR (100.6 MHz,
D2O, DSS) δ 63.80, 66.08, 72.93, 76.48, 76.56, 76.96, 86.60,
86.63, 89.80, 90.57, 105.17, 142.60, 145.38, 151.90, 154.40,
155.59, 158.32, 168.92; 31P NMR (109 MHz, D2O, 85% H3PO4)
δ 0.14, (brs, 1P), 0.34 (brs, 1P); MS (FAB-) calcd for
C19H23O15N7P2Na (M+ - H) 674.0625, found 674.0634; Rf 0.27
(solvent system: iPrOH-concd NH3-H2O, 6:1:3, v/v/v).
Syn t h esis of 2′-Th iop h osp h or yla t ed Din u cleot id e
U(2′-p s)p A (7). The fully protected dinucleotide U(2′-p)pA
(23a ) (29 mg, 0.017 mmol) was dissolved in 1% trifluoroacetic
acid (5 mL), and the solution was stirred for 15 min. The
reaction was quenched with CH3OH-pyridine (1:1, v/v, 2 mL).
After being concentrated under reduced pressure, the mixture
was extracted by use of CH2Cl2 (20 mL)-5% NaHCO3 (20 mL).
The organic layer was collected, washed with H2O (20 mL),
and dried over Na2SO4. The residue was rendered anhydrous
by repeated coevaporation with pyridine and finally dissolved
in pyridine (0.5 mL). To the solution were added successively
BSA (0.1 mL, 0.4 mmol) and DBU (0.05 mL, 0.05 mmol). After
the mixture stirred for 10 min at room temperature, the
reaction was quenched by addition of water (0.05 mL). Stirring
was continued for a few minutes, and then the mixture was
extracted with H2O (2 mL)-Et2O (2 mL). The aqueous layer
was concentrated under reduced pressure, and the residue was
dissolved in concd NH3-pyridine (9:1, v/v, 5 mL). After
stirring for 24 h, the mixture was coevaporated twice with
water, and the resulting residue was treated with 80% acetic
acid (5 mL) for 15 min. The solution was concentrated under
Exp er im en ta l Section
Gen er a l Met h od s. 1H NMR spectra were measured on
J EOL-EX 270 and Varian Unity 400 spectrometers at 270 and
400 MHz with TMS (for CDCl3) or DDS (for D2O) as internal
standard. 13C NMR spectra were obtained on J EOL-EX 270
and Varian Unity 400 spectrometers at 67.8 and 100.6 MHz,
respectively, with CHCl3 (for CDCl3) as internal standard or
dioxane (for D2O) as external standard. 31P NMR spectra were
recorded on a J EOL-EX 270 spectrometer at 109 MHz with
the external reference of 85% H3PO4. UV spectra were taken
on a HITACHI U-2000 spectrophotometer. Reversed phase
HPLC was performed on a Waters LC module 1 with a
µBondasphere 5 µm C18 100 Å (3.9 × 150 mm) column using
a linear gradient of acetonitrile (0-30%) in 0.1 M ammonium
acetate buffer (pH 7.0) for 30 min at flow rate of 1 mL/min at
50 °C. Ion exchange HPLC was carried out at a flow rate of
1 mL/min for 30 min at 50 °C on a Whatman Partisil 10 SAX
WCS analytical column (4.6 × 250 mm) using a linear gradient
of 0-100% solution B (20% CH3CN in 0.005 M KH2PO4) in
solution A (20% CH3CN in 0.5 M KH2PO4). Paper chroma-
tography was carried out by use of a descending technique with
Whatman 3MM Chr papers using iPrOH-concd NH3-H2O (6:
1:3, v/v/v). Thin-layer chromatography was performed using
Merck Kieselgel 60F-254 (0.25 mm) with developing solvent
of CH2Cl2-CH3OH (9:1, v/v) or iPrOH-concd NH3-H2O (7:1:
2, v/v/v). Column chromatography was performed with silica
gel C-200 purchased from Wako Co. Ltd., and a minipump for
a goldfish bowl was conveniently used to attain sufficient
pressure for rapid chromatographic separation. Elemental
analyses were performed by the Microanalytical Laboratory,
Tokyo Institute of Technology, Nagatsuta, J apan.
Syn th esis of F u lly P r otected 2′-P h osp h or yla ted Di-
n u cleotid e U(2′-p )p A (23a ). 2′,3′-O,4-N,4-N-tetrabenzoyl-
adenosine 22 (32 mg, 0.05 mmol) and 1H-tetrazole (7 mg, 0.1
mmol) were rendered anhydrous by repeated coevaporation
with dry pyridine (1 mL × 3) and toluene (1 mL × 1) and
finally dissolved in CH3CN (0.5 mL). To the mixture was
added 23 (61 mg, 0.055 mmol). After being stirred for 3 h,
the mixture was concentrated under reduced pressure and
treated with iodine (127 mg, 0.5 mmol) in pyridine-H2O (9:1,
v/v, 1 mL) for 1 h. The mixed solution was evaporated in
vacuo, and the residue was dissolved in CH2Cl2 (10 mL) and
washed with saturated Na2S2O3 (10 mL) and water (10 mL).
After the organic layer was dried over Na2SO4, the solution
was filtered and concentrated under reduced pressure. Silica
gel column chromatography was performed with hexane-ethyl
acetate (2:3, v/v) containing 0.5% pyridine to give the fully
protected dimer 23a (73 mg, 85%): 1H NMR (270 MHz, CDCl3,
(43) Herschlag, D.; Piccirilli, J . A.; Cech, T. R. Biochemistry 1991,
30, 4844. (b) Benkovic, S. J .; Schray, K. J . In The Enzymes; Boyer, P.,
Ed.; Academic Press: New York, 1971; Vol. 8, pp 201-208. (c)
Herschlan, D.; J encks, W. P. J . Am. Chem. Soc. 1989, 111, 7587.
(44) Beltz, W. R.; O’Brien, K. J . Fed. Proc. Fed. Am. Soc. Exp. Biol.
1981, 40, 1849.

Dephosphorylation of Phosphorylated Dinucleotides
J . Org. Chem., Vol. 62, No. 9, 1997 2821
reduced pressure, and then the residue was dissolved in water
(2 mL). The solution was washed with Et2O (2 mL), concen-
trated under reduced pressure, and chromatographed on
Whatman 3MM Chr papers with iPrOH-concd NH3-H2O (6:
1:3, v/v/v) to give 7 (298 A260, 70%): 1H NMR (400 MHz, D2O,
DSS) δ 3.70 (dd, J ) 3.2, 12.7 Hz, 1H), 3.76 (dd, J ) 4.1, 12.7
Hz, 1H), 4.20-4.23 (m, 2H), 4.25 (ddd, J ) 2.3, 3.2, 4.1 Hz,
1H), 4.38 (m, 1H), 4.62 (m, 1H), 4.76 (ddd, J ) 2.3, 5.0, 7.3
Hz, 1H), 4.81 (dd, J ) 5.5, 5.8 Hz, 1H), 4.93 (ddd, J ) 5.0, 7.2,
10.6 Hz, 1H), 5.83 (d, J ) 7.9 Hz, 1H), 6.02 (d, J ) 7.2 Hz,
1H), 6.10 (d, J ) 5.8 Hz, 1H), 7.77 (d, J ) 7.9 Hz, 1H), 8.23 (s,
1H), 8.46 (s, 1H); 13C NMR (100.6 MHz, D2O, dioxane) δ 64.04,
68.22, 72.99, 76.53, 76.56, 77.53, 86.82, 87.00, 89.71, 90.39,
105.01, 121.50, 142.69, 145.91, 151.95, 154.49, 155.54, 158.29,
169.89; 31P NMR (109 MHz, D2O, 85% H3PO4) δ 0.15 (brs, 1P),
46.07 (brs, 1P); MS (FAB-) calcd for C19H23O14N7P2SNa (M+
- H) 690.0397, found 690.0389; Rf 0.26 (solvent system:
iPrOH-concd NH3-H2O, 6:1:3, v/v/v).
CDCl3, TMS) δ 1.26-1.33 (m, 9H), 2.65-2.93 (m, 4H), 3.08
(q, J ) 7.3 Hz, 6H), 3.48-3.64 (m, 2H), 3.76 (s, 6H), 4.18-
4.55 (m, 9H), 5.10-5.23 (m, 1H), 5.15-5.31 (d, J ) 8.2 Hz,
1H), 5.72-6.08 (m, 3H), 6.22 (d, J ) 5.0 Hz, 1H), 6.50 (d, J )
7.3 Hz, 1H), 6.81-6.86 (m, 4H), 7.06-7.57 (m, 20H), 7.75-
7.96 (m, 7H), 8.15 (d, J ) 8.2 Hz, 1H), 11.45 (brs, 1H); 13C
NMR (67.8 MHz, CDCl3) δ 8.47, 21.31, 29.54, 45.64, 55.37,
55.10, 55.13, 62.41, 62.48, 63.02, 63.11, 72.15, 73.04, 81.19,
86.99, 87.26, 102.50, 103.22, 112.47, 113.24, 125.16, 126.88,
127.13, 127.67, 127.96, 128.09, 128.25, 128.32, 128.42, 128.68,
128.79, 128.90, 129.02, 129.63, 129.69, 130.15, 130.35, 131.23,
133.51, 134.86, 135.04, 137.72, 139.46, 143.95, 147.33, 149.63,
150.63, 158.44, 158.51, 158.62, 161.80, 163.27, 165.23, 165.32,
168.54; 31P NMR (109 MHz, CDCl3, 85% H3PO4) δ -0.32 (s,
1P), 3.66 (s, 1P). Anal. Calcd for C72H75N7O22P‚2H2O: C,
58.10; H, 5.35; N, 6.59. Found: C, 58.13; H, 5.84; N, 6.79.
Syn th esis of 2′-5′-Lin k ed 3′-P h osp h or yla ted Diu r id y-
la te U(3′-p )p U (10). The protected 2′-5′-linked 3′-phosphor-
ylated dinucleotide U(3′-p)pU (20) (53 mg, 0.028 mmol) was
dissolved in 1% trifluoroacetic acid in CH2Cl2 (2 mL). The
mixture was stirred for 15 min, and then the reaction was
quenched with CH3OH-pyridine (1:1, v/v, 1 mL). After being
concentrated in vacuo, the solution was extracted with CH2-
Cl2 (10 mL)-5% NaHCO3 (10 mL). The organic layer was
collected and dried over Na2SO4. The residue was rendered
anhydrous by repeated coevaporation with pyridine and finally
dissolved in pyridine (0.3 mL). To the solution were added
BSA (159 µL, 0.64 mmol) and DBU (4 µL, 0.084 mmol), and
the mixture was stirred for 30 min at room temperature. To
the reaction mixture was added water (0.2 mL). After being
stirred for a few minutes, the solution was extracted with H2O
(3 mL)-Et2O (3 mL). The aqueous layer was concentrated
under reduced pressure. The residue was dissolved in concd
NH3-H2O (9:1, v/v, 5 mL). After stirring for 20 h at room
temperature, the mixture was coevaporated twice with water
under reduced pressure, and then the residue was dissolved
in water (2 mL). The solution was washed with Et2O (2 mL),
evaporated under reduced pressure, and chromatographed on
Whatman 3MM Chr papers with iPrOH-concd NH3-H2O (6:
1:3, v/v/v) to give 10 (42%, 228 A260): 1H NMR (400 MHz, D2O,
DSS) δ 3.86 (d, J ) 3.4 Hz, 2H), 4.12 (m, 2H), 4.17 (dd, J )
2.4, 4.4 Hz, 1H), 4.27 (t, J ) 4.4 Hz, 1H), 4.31-4.35 (m, 2H),
4.67 (ddd, J ) 4.9, 4.9, 8.9 Hz, 1H), 4.80 (ddd, J ) 4.9, 4.9, 8.7
Hz, 1H), 5.78 (d, J ) 8.1 Hz, 1H), 5.89 (d, J ) 4.6 Hz, 1H),
5.90 (d, J ) 8.1 Hz, 1H), 6.08 (d, J ) 4.9 Hz, 1H), 7.83 (d, J )
8.1 Hz, 1H), 7.88 (d, J ) 8.1 Hz, 1H); 13C NMR (100.6 MHz,
D2O, dioxane) δ 61.70, 63.35, 65.03, 69.83, 72.88, 74.80, 77.06,
83.42, 84.76, 89.19, 89.49, 103.29, 103.48, 142.28, 143.16,
152.16, 152.47, 166.89, 166.99; 31P NMR (109 MHz, D2O, 85%
H3PO4) δ -0.32 (brs, 1P), 3.7 (brs, 1P); Rf 0.10 (solvent
system: iPrOH-concd NH3-H2O, 6:1:3, v/v/v).
Syn th esis of F u lly P r otected Din u cleotid e U(2′-5′)p U
(19a ). 2′, 3′-O,N3-Tribenzoyluridine 18 (223 mg, 0.4 mmol)
and 1H-tetrazole (70 mg, 1.0 mmol) were rendered anhydrous
by coevaporation with pyridine (1 mL × 3) and toluene (1 mL
× 2) followed by dissolution in CH3CN (4 mL). To the solution
was added 5′-O-(4,4′-dimethoxytrityl)-3′-O-(tert-butyldimeth-
ylsilyl)uridine-2′-O-(2-cyanoethoxy)-N,N-diisopropylphosphora-
midite (18) (467 mg, 0.52 mmol). After being stirred for 3.5
h, the reaction mixture was treated with tert-butyl hydro-
peroxide (500 mL, 4.0 mmol) and stirred for an additional 1
h. The solution was diluted with CH2Cl2 (15 mL), washed with
5% NaHCO3 (15 mL) and H2O (15 mL), dried over Na2SO4,
filtered, and evaporated under reduced pressure. The residue
was chromatographed on a column of silica gel with CH2Cl2-
CH3OH to give 19a (202 mg, 74%): 1H NMR (270 MHz, CDCl3,
TMS) δ -0.02 (3H, s, SiCH3 of TBDMS), 0.09-0.11 (m, 3H),
0.78 (s, 9H), 2.77, 2.90 (2t, J ) 5.8, 6.3 Hz, 2H), 3.37-3.40 (m,
1H), 3.81 (s, 3H), 4.15-4.18 (m, 2H), 4.33-4.68 (m, 5H), 4.96-
5.05 (m, 1H), 5.30-5.34 (m, 1H), 5.63-5.83 (m, 2H), 5.93-
6.02 (m, 2H), 6.15 (d, J ) 2.3 Hz, 1H), 6.42 (d, J ) 6.3 Hz,
1H), 6.87 (m, 4H), 7.24-7.63 (m, 20H), 7.76-7.97 (m, 7H), 8.10,
8.14 (2d, J ) 7.9, 8.3 Hz, 1H), 9.69, 9.82 (2s, 1H); 13C NMR
(100.6 MHz, CDCl3) δ -5.35, -4.60, 17.88, 19.43, 19.54, 25.47,
55.17, 60.54, 60.74, 62.64, 62.72, 63.22, 66.94, 66.97, 67.14,
69.24, 70.60, 70.84, 73.28, 73.53, 80.06, 80.94, 81.06, 87.26,
88.09, 102.55, 103.31, 113.172, 113.21, 116.30, 117.23, 127.26,
127.91, 128.14, 128.18, 128.27, 128.36, 128.43, 128.55, 129.99,
129.65, 129.775, 130.21, 130.42, 131.20, 133.62, 134.63, 134.72,
134.93, 139.19, 139.50, 139.96, 143.72, 149.42, 150.73, 150.80,
158.74, 161.51, 161.67, 163.13, 165.10, 165.21, 165.43, 168.25,
168.34; 31P NMR (109 MHz, CDCl3, 85% H3PO4) δ -1.28 (s,
1P), -0.73 (s, 1P). Anal. Calcd for C69H70N5O19PSi‚H2O: C,
61.37; H, 5.37; N, 5.19. Found: C, 61.22; H, 5.30; N, 5.09.
Syn th esis of P r otected 2′-5′ Lin k ed 3′-P h osp h or y-
la ted Diu r id yla te U(3′-p )p U (20). The fully protected 2′-
5′-linked diuridylate U(2′-5′)pU (19a ) (247 mg, 0.18 mmol) was
treated with Et3N-pyridine (1:3, v/v, 8 mL) for 20 h and then
evaporated under reduced pressure. The residue was dissolved
in pyridine (3.6 mL) and to the solution was added Et3N‚3HF
(202 µL, 1.2 mmol). After being further stirred for 72 h, the
mixture was diluted with CH2Cl2 (15 mL), and the usual
extraction was performed with 1 M triethylammonium hydro-
gen carbonate buffer (15 mL). The organic layer was concen-
trated in vacuo. After addition of 1H-tetrazole (378 mg, 5.4
mmol) to the residue, the mixture was rendered anhydrous
by repeated coevaporation with pyridine under reduced pres-
sure and finally dissolved in CH3CN (2 mL). To the solution
was added bis(2-cyanoethoxy)-N,N-diisopropylphosphoramid-
ite (497 mg, 1.8 mmol). After being stirred for 1 h, the mixture
was treated with tert-butyl hydroperoxide (406 µL, 3.6 mmol)
for 1 h. Then, the solution was diluted with CH2Cl2 (15 mL),
and extraction was performed with CH2Cl2-1 M triethyl-
ammonium hydrogen carbonate buffer. The organic layer was
concentrated, dried over Na2SO4, and chromatographed on
silica gel with CH2Cl2-CH3OH containing 0.5% pyridine.
Finally, extraction was performed with CH2Cl2-1 M triethyl-
ammonium hydrogen carbonate buffer, and the organic layer
was collected to give 20 (35%, 95 mg): 1H NMR (270 MHz,
Ma ter ia ls. The used nucleosides were purchased from
Yamasa. Preparation of U(2′-p)pU (1), U(2′-ps)pU (2), U(2′-
pp-biotin)pU (4), U(2′-ps-bimane)pU (5), thymidine 3′-phos-
phate (8), and thymidine 3′-thiophosphate (9) was carried out
according to our previous paper.11,12 All compounds were used
after checking the purity by HPLC.
Kin etic Mea su r em en ts. A substrate (1.0 A260) was dis-
solved in a 0.1 M AcONH4 solution (0.3 mL, pH 7.0) either
containing 10 mM metal chloride or not. The initial concen-
tration of the substrate was 1.5 × 10-4 mol dm-3, and the
reaction temperature was maintained at 90 °C. The reaction
was followed by analyzing the compositions of samples with-
drawn at appropriate intervals by HPLC. First-order rate
constants (k) for the dephosphorylation of U(2′-p)pU (1), U(2′-
ps)pU (2), thymidine 3′-thiophosphate (9), and U(3′-p)pU (10)
in Table 2 were obtained by estimating the time-dependent
product distributions. The method of least-squares was ap-
plied to fit experimental data using eq 1. The c0 value is the
initial concentration of starting materials, and c is the
concentration of the starting material which remained at time
t:
ln(c0/c) ) kt
(1)
In the case of the dephosphorylation (k1) of thymidine 3′-

2822 J . Org. Chem., Vol. 62, No. 9, 1997
Tsuruoka et al.
phosphate (8), since the appreciable hydrolysis of the glycosidic
bond occurred concurrently, the first-order rate constants were
obtained by eq 2:
thiophosphoryluridylyl(3′-5′)uridine; d(UpU), 2′-deoxyuridyl-
yl(3′-5′)-2′-deoxyuridine.
Ack n ow led gm en t. This work was supported by the
Toray Scientific Foundation, J apan Society for the
promotion of Science for Young Scientists (H. Tsuruoka),
and a Grant-in-Aid for Scientific Research on Priority
Areas from the Ministry of Education, Science, Sports
and Culture, J apan. We express hearty thanks to Prof.
H. Lo¨nnberg for his kindness in sending many reprints
including galley proofs.
c ) k1[c0 exp(-k2t) -exp{ln c0 -(k1 + k3)t}]/(k1 + k3 - k2)
(2)
In eq 2, c0 refers to initial concentration of Tp (8), and c is
concentration of thymidine at time t. Rate constants k2 and
k3 are for glycosidic hydrolysis of thymidine and thymidine
3′-phosphate, respectively. The value of k2 was determined
from other experiments in which 0.1 M AcONH4 (pH 7.0) buffer
in the presence and absence of metal chlorides as additives
was used at 90 °C (none, 0.678 × 10-6 s-1; MgCl2, 0.725 ×
10-6 s-1; MnCl2, 0.747 × 10-6 s-1; ZnCl2, 0.681 × 10-6 s-1
LaCl3, 0.758 × 10-6 s-1; YCl3, 0.753 × 10-6 s-1).
;
Su p p or tin g In for m a tion Ava ila ble: Copies of 1H, 13C,
and 31P NMR data for 6, 7, 10, 19a , 20, and 23a ; 1H-1H COSY
1
Abbreviations: U(2′-p)pU, 2′-phosphoryluridylyl(3′-5′)uri-
dine; U(2′-ps)pU, 2′-thiophosphoryluridylyl(3′-5′)uridine; UpU,
uridylyl(3′-5′)uridine; U(2′-p)p, uridine 2′,3′-bisphosphate;
U(3′-p)pU, 2′-phosphoryluridylyl(3′-5′)uridine; U(2′-p)pA, 2′-
phosphoryluridylyl(3′-5′)adenosine; U(2′-ps)pU, 2′-thiophos-
phoryluridylyl(3′-5′)adenosine; Tp, thymidine 3′-phosphate;
Tps, thymidine 3′-thiophosphate; U(2′-pU)pU, uridylyl(2′-5′)-
(3′-5′)diuridine; U(2′-pp-biotin)pU, 2′-biotinyldiphosphoryl-
uridylyl(3′-5′)uridine; U(2′-ps-bimane)pU, 2′-(S-bimanyl)-
NMR data for 6, 10, 21a , 20, and 23a ; H-13C COSY NMR
data for 6; FAB mass data for 6 and 7; detailed HPLC profiles
of hydrolysis of 1, 2, and 10 (27 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
J O970021K