Hydrolytic reactions of guanosyl-(3′,3′)-uridine and guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine)

Article abstract of DOI:10.1021/ja047568i

Hydrolytic reactions of guanosyl-(3′,3′)-uridine and guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) have been followed by RP HPLC over a wide pH range at 363.2 K in order to elucidate the role of the 2′-hydroxyl group as a hydrogen-bond donor upon departure of the 3′-uridine moiety. Under neutral and basic conditions, guanosyl-(3′,3′)-uridine undergoes hydroxide ion-catalyzed cleavage (first order in [OH^-]) of the P-O3′ bonds, giving uridine and guanosine 2′,3′-cyclic monophosphates, which are subsequently hydrolyzed to a mixture of 2′- and 3′-monophosphates. This bond rupture is 23 times as fast as the corresponding cleavage of the P-O3′ bond of guanosyl-(3′,3′)-(2,5′-di-O-methyluridine) to yield 2′,5′-O-dimethyluridine and guanosine 2′,3′-cyclic phosphate. Under acidic conditions, where the reactivity differences are smaller, depurination and isomerization compete with the cleavage. The effect of Zn²⁺ on the cleavage of the P-O3′ bonds of guanosyl-(3′,3′)-uridine is modest: about 6-fold acceleration was observed at [Zn²⁺] = 5 mmol L^-1 and pH 5.6. With guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) the rate-acceleration effect is greater: a 37-fold acceleration was observed. The mechanisms of the partial reactions, in particular the effects of the 2′-hydroxyl group on the departure of the 3′-linked nucleoside, are discussed.

Full text of DOI:10.1021/ja047568i

Published on Web 08/14/2004
Hydrolytic Reactions of Guanosyl-(3′,3′)-uridine and
Guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine)
Anu Kiviniemi, Tuomas L o¨ nnberg, and Mikko Ora*
Contribution from the Department of Chemistry, UniVersity of Turku, FIN-20014 Turku, Finland
Received April 27, 2004; E-mail: mikora@utu.fi
Abstract: Hydrolytic reactions of guanosyl-(3′,3′)-uridine and guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) have
been followed by RP HPLC over a wide pH range at 363.2 K in order to elucidate the role of the 2′-
hydroxyl group as a hydrogen-bond donor upon departure of the 3′-uridine moiety. Under neutral and basic
-
conditions, guanosyl-(3′,3′)-uridine undergoes hydroxide ion-catalyzed cleavage (first order in [OH ]) of
the P-O3′ bonds, giving uridine and guanosine 2′,3′-cyclic monophosphates, which are subsequently
hydrolyzed to a mixture of 2′- and 3′-monophosphates. This bond rupture is 23 times as fast as the
corresponding cleavage of the P-O3′ bond of guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) to yield 2′,5′-O-
dimethyluridine and guanosine 2′,3′-cyclic phosphate. Under acidic conditions, where the reactivity
differences are smaller, depurination and isomerization compete with the cleavage. The effect of Zn2 on
the cleavage of the P-O3′ bonds of guanosyl-(3′,3′)-uridine is modest: about 6-fold acceleration was
+
2
+
-1
observed at [Zn ] ) 5 mmol L and pH 5.6. With guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) the rate-
acceleration effect is greater: a 37-fold acceleration was observed. The mechanisms of the partial reactions,
in particular the effects of the 2′-hydroxyl group on the departure of the 3′-linked nucleoside, are discussed.
6
2,3,7
Introduction
ingly faster than the P-O5′ cleavage. It has been suggested
that the much faster cleavage of the P-O3′ bond could be
attributed to a proton transfer from the neighboring 2′-hydroxyl
group to the departing 3′-oxygen atom concerted with the bond
cleavage. Such rate accelerations have also been reported for
While small molecule models have been extensively used to
study the underlying principles of the cleavage of RNA by the
so-called small ribozymes utilizing an intramolecular mecha-
1
nism, the corresponding studies mimicking the intermolecular
8
the cleavage of several nonnucleosidic phosphate esters,
transesterification process of large ribozymes are scarce. Such
a mechanism has only been found to operate in the methanolysis
9
10
carboxylic esters, and acetal derivatives. To quantify the
acceleration caused by intramolecular hydrogen bonding in the
present case, we now report on a comparative kinetic analysis
of the cleavage of two diribonucleoside 3′,3′-monophosphates,
viz. guanosyl-(3′,3′)-uridine (1a) and guanosyl-(3′,3′)-(2′,5′-di-
O-methyluridine) (1b). While these compounds are cleaved by
an intramolecular attack of the 2′-hydroxy group on the
phosphorus atom, they still allow assessment of the significance
of transition-state stabilization by intramolecular hydrogen
bonding within the leaving group. The rate-limiting step of the
reaction is the departure of the 3′-linked nucleoside, either
uridine or 2′-O-methyluridine, from the phosphorane intermedi-
ate, and hence, the rate-accelerating effect of the 2′-hydroxyl
group on the cleavage (and possibly formation) of this inter-
2
of ribonucleoside 2′/3′-dimethyl phosphates and hydrolysis of
3
H-phosphonodiesters in organic solvents. We previously re-
ported a kinetic study on the hydrolytic reactions of a sugar
4
O-alkylated trinucleoside 3′,3′,5′-monophosphate, the monoan-
ionic pentacoordinated phosphorane intermediate/transition state
of which resembles the intermediate formed in the ribozyme
5
reaction of group I introns. The results obtained suggested that
the nucleophilic attack of hydroxide ion on the phosphorus atom
of the phosphotriester results in only a 3 times faster departure
of the 3′-hydroxy-esterified nucleoside compared to its 5′-
hydroxy-esterified counterpart. In other words, nucleoside 3′-
oxyanion appears to be only a three times better leaving group
than the 5′-oxyanion, and hence, the difference in the inherent
leaving group property is insufficient to explain the fact that
with group I introns the P-O3′ bond cleavage is overwhelm-
(
6) Cech, T. R.; Zaugh, A. J.; Grabowski, P. J. Cell 1981, 27, 487-496.
7) (a) Yoshida, A.; Shan, S.; Herschlag, D.; Piccirilli, J. A. Chem. Biol. 2000,
(
7
, 85-96. (b) Bakalova, S.; Siebrand, W.; Fernandez-Ramos, A.; Sme-
darchina, Z.; Petkov, D. D. J. Phys. Chem. B 2002, 106, 1476-1480. (c)
Tzokov, S. B.; Devedjiev, I. T.; Bratovanova, E. K.; Petkov, D. D. Angew.
Chem. 1994, 106, 2401-2402.
(
1) Oivanen, M.; Kuusela, S.; L o¨ nnberg, H. Chem. ReV. 1998, 98, 961-990.
2) Roussev, C. D.; Ivanova, G. D.; Bratovanova, N. G.; Petkov, D. D. J. Am.
Chem. Soc. 1999, 121, 11267-11272.
(8) (a) Forconi, M.; Williams, N. H. Angew. Chem., Int. Ed. 2002, 41, 849-
852. (b) Bender, M. L.; Lawlor, J. M. J. Am. Chem. Soc. 1963, 85, 3010-
3017. (c) Bromilow, R. H.; Kirby, A. J. J. Chem. Soc., Perkin Trans. 2
1997, 149-155. (d) Bromilow, R. H.; Khan, S. A.; Kirby, A. J. J. Chem.
Soc. (B) 1971, 1091-1097. (e) Brown, D. M.; Hall, G. E.; Higson, H. M.
J. Chem. Soc. 1958, 1360-1366.
(9) (a) Bruice, T. C.; Fife, T. H. J. Am. Chem. Soc. 1962, 82, 1973-1979. (b)
Henbest, H. B.; Lovell, B. J. J. Chem. Soc. 1957, 1965-1969.
(10) (a) Capon, B. Chem. ReV. 1969, 69, 407-498. (b) Kirby, A. J. Acc. Chem.
Res. 1997, 30, 290-296.
(
(
3) Tzokov, S. B.; Momtcheva, R. T.; Vassilev, N. G.; Kaneti J.; Petkov, D.
D. J. Am. Chem. Soc. 1999, 121, 5103-5107.
4) L o¨ nnberg, T.; Mikkola, S. J. Org. Chem. 2004, 69, 802-810.
5) (a) Piccirilli, J. A.; Vyle, J. S.; Caruthers, M. H.; Cech, T. R. Nature 1993,
(
(
3
61, 85-88. (b) Weinstein, L. B.; Jones, B. C.; Cosstick, R.; Cech, T. R.
Nature 1997, 388, 805-808. (c) Schan, S.; Yoshida, A.; Sun, S.; Piccirilli,
J. A.; Herschlag, D. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12299-12304.
11040
9
J. AM. CHEM. SOC. 2004, 126, 11040-11045
10.1021/ja047568i CCC: $27.50 © 2004 American Chemical Society

Hydrolytic Reactions
A R T I C L E S
mediate can be determined. Additionally, the effect of the action
of a catalytically active metal ion, Zn , on the cleavage rate
has been investigated. Mechanisms of hydrolytic reactions are
8a; 7b/8b; reactions E and F) and are, hence, accumulated as
intermediates only at pH 2-3. The depurination is not detected
at pH > 4 and the isomerization at pH > 6. In other words, at
pH 4-6, the cleavage reactions (A,B) compete with the
isomerization (D), and at pH > 6, they are the only reactions
observed. 3′,3′-GpMe2U (1b) reacts analogously to 3′,3′-GpU,
except that now only one cleavage reaction (reaction G), i.e.,
the one respective to reaction B in Scheme 1, may take place,
and isomerization yields only 2′,3′-GpMe2U (2b; reaction H;
Scheme 2). Examples of time-dependent product distributions
are given as Supporting Information.
2+
discussed and compared to those obtained earlier with 3′,5′-
UpU (10) containing a native 3′,5′-linkage.1
1,12
In particular,
the reactions taking place in basic and neutral solutions are of
interest.
pH-Rate Profiles. Figure 1 shows the pH-rate profiles for
Figure 1. pH-rate profile for the cleavage (kcl) of 3′,3′-GpU (1a) (9) and
guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) (1b) (0) at 363.2 K. The solid
line shows the corresponding curve for 3′,5′-UpU (10). The ionic strength
-
1
of the solutions was adjusted to 0.1 mol L with sodium chloride. Under
conditions where the isomerization competes with the cleavage, kcl represents
the disappearance of the isomeric GpUs. The curves are calculated by least-
squares fitting and indicate the theoretical dependence of the observed rate
+
constants on H concentration.
the cleavage of 3′,3′-GpU (1a, reactions A and B) and 3′,3′-
GpMe2U (1b; reaction G). The profiles consist of three
regions: a hydronium-ion-catalyzed reaction at pH < 4, a nearly
pH-independent reaction from pH 4 to 5, and a hydroxide-
ion-catalyzed reaction at pH > 5. In addition, a slight curv-
ature (leveling off) toward a constant value is observed on
going to very basic conditions. The hydroxide-ion-catalyzed
Results and Discussion
-
cleavage (first order in [OH ]) of 1a takes place 260 times as
Product Distributions and Reaction Pathways. The cleav-
age of guanosyl-(3′,3′)-uridine (3′,3′-GpU; 1a) and guanosyl-
1
1,13
fast as that of 3′,5′-UpU (10).
More importantly, 3′,3′-GpU
is cleaved 46 times as fast as 3′,3′-GpMe2U. Since there are
two scissile PO3′ bonds in 3′,3′-GpU which react as rapidly
and only one in 3′,3′-GpMe2U, one may conclude that methyl-
ation of the 2′- and 5′-hydroxyl groups of the uridine moiety
retards the cleavage of an individual PO3′ bond by a factor of
(3′,3′)-(2′,5′-di-O-methyluridine) (3′,3′-GpMe2U; 1b) was stud-
ied over a wide pH range at 363.2 K by determining the time-
dependent product composition of the aliquots withdrawn at
appropriate intervals from the reaction solution by RP HPLC.
The products were characterized by either spiking with authentic
samples or mass spectrometric analysis (HPLC/ESI-MS). Scheme
2
3.
As with 3′,5′-UpU, the isomerization is hydronium-ion
1
summarizes the reactions observed to take place with 1a.
catalyzed and approximately as fast as the cleavage at pH < 4
and nearly pH-independent under less acidic conditions (Figure
Accordingly, at pH < 4, 3′,3′-GpU is cleaved to uridine 2′,3′-
cyclic phosphate (2′,3′-cUMP; 6a; reaction A) and guanosine
2
4
). The rate of pH-independent isomerization of 3′,3′-GpU is
.5-fold compared to that of 3′,3′-GpMe2U, which is, in turn,
2′,3′-cyclic phosphate (2′,3′-cGMP; 6b; reacion B), depurinated
by loss of guanine (9; reaction C), and isomerized to a mixture
of 2′,3′-GpU (2a), 3′,2′-GpU (4) and 2′,2′-GpU (5; reaction D).
Under these conditions the cyclic phosphates (6a,b) are quite
rapidly hydrolyzed to a mixture of 2′- and 3′-phosphates (7a/
isomerized as fast as 3′,5′-UpU (10). Bearing in mind that there
are two 2′-hydroxy functions in 3′,3′-GpU, the actual rate
difference is only 2.3. At pH < 4, the differences between the
isomerization rates are even smaller.
(
11) J a¨ rvinen, P.; Oivanen, M.; L o¨ nnberg, H. J. Org. Chem. 1991, 56, 5396-
401.
12) Kuusela, S.; L o¨ nnberg, H. J. Phys. Org. Chem. 1993, 6, 347-356.
5
(13) Rayner, B.; Reese, C. B.; Ubasawa, A. J. Chem. Soc., Chem. Commun.
1980, 972-973.
(
J. AM. CHEM. SOC.
9
VOL. 126, NO. 35, 2004 11041

A R T I C L E S
Kiviniemi et al.
Scheme 1
-
1
Scheme 2
metal-ion concentration (1-10 mmol L ) and pH (5.1-5.6).
In the pH range studied, the cleavage is first order in both
2
+
-1
hydroxide-ion concentration ([Zn ] ) 5 mmol L ) and metal-
2
+
-1
ion concentration ([Zn ] ) 1-10 mmol L at pH 5.6 (Figure
). The same reactions as in the absence of the metal ions take
palce, viz. cleavage and isomerization, but only the cleavage is
3
2
+
promoted by Zn .
Mechanism of the Cleavage of P-O3′ Bond under Neutral
and Basic Conditions: Intramolecular Hydrogen Bonding.
As mentioned above, the hydroxide-ion-catalyzed cleavage (first
-
order in [OH ]) of each of the PO3’ bonds in 3′,3′-GpU 1a
proceeds 23 times as fast as the PO3’ bond cleavage in 3′,3′-
GpMe2U. One may ask what is the origin of this rate accelera-
1
4
tion. The basic mechanism suggested earlier for the phos-
phodiester cleavage of RNA in all likelihood also applies to
the reactions 3′,3′-GpU and 3′,3′-GpMe2U. Accordingly,11
a
rapid initial deprotonation of the 2′-hydroxyl group is followed
by the rate-limiting attack of the resulting oxyanion on the
phosphorus atom, giving a marginally stable dianionic phos-
phorane intermediate from which the 3′-linked nucleoside
departs “in-line” with the nucleophilic attack. The considerably
faster cleavage of 3′,3′-GpU compared to 3′,3′-GpMe2U suggests
the 2′-hydroxyl group exerts an additional rate-accelerating
effect. This group may be expected to either stabilize the
departing nucleoside 3′-oxyanion by intramolecular hydrogen
2
+
Zn -Promoted Cleavage. The rate enhancements caused
2
+
by catalytically active metal ions, such as Zn , are also of
interest in the present study. For this purpose, decomposition
of 3′,3′-GpU and 3′,3′-GpMe2U was studied as a function of
(
14) Brown, D. M.; Magrath, D. I.; Neilson, A. H.; Todd, A. R. Nature 1956,
177, 1124.
11042 J. AM. CHEM. SOC.
9
VOL. 126, NO. 35, 2004

Hydrolytic Reactions
A R T I C L E S
Scheme 3
Figure 2. pH-rate profile for the isomerization (kis) of 3′,3′-GpU (1a)
(9) and guanosyl-(3′,3′)-(2′,5′-di-O-methyluridine) (1b) (0) at 363.2 K. The
solid line shows the corresponding curve for 3′,5′-UpU (10). The ionic
strength of the solutions was adjusted to 0.1 mol L with sodium chloride.
a 3-5-fold rate acceleration could be attributed to leaving-group
-1
stabilization by hydrogen bonding, taking into account that the
1
7
transition state is very late and is, hence, stabilized as
efficiently as the initial product, the 3′-oxyanion. Accordingly,
the observed 23-fold acceleration can hardly by entirely
explained by stabilization of the leaving group by intramolecular
hydrogen bonding.
As mentioned above, another possible source of rate ac-
celeration is stabilization of the phosphorane intermediate: 2′-
OH donates a hydrogen bond to one of the phosphorane
2
,3
oxyanions (Scheme 3). The second pKa of a phosphorane
intermediate has been estimated to be higher than that of the
1
8
2
′-hydroxyl group. Accordingly, hydrogen bonding of the 2′-
hydroxy function of the leaving group to one of the nonbridging
phosphorane oxygens upon formation of a dianionic phospho-
rane appears feasible, and it may well result in an additional
3-5-fold acceleration of the cleavage. For comparison, methyl-
ation of the neighboring 2′-hydroxyl groups retards the pH-
Figure 3. Effect of metal-ion concentration on the first-order rate constants
of cleavage (kcl) of 3′,3′-GpU (1a; 9) and guanosyl-(3′,3′)-(2′,5′-di-O-
methyluridine) (1b; 0) at pH 5.60 and 363.2 K. The pH was adjusted with
HEPES buffer, and the ionic strength was maintained at 0.1 mol L with
sodium nitrate.
independent isomerization only by a factor of 2.3. This reaction,
1,19
however, proceeds via a phosphorane monoanion, and hence,
stabilization by a similar intramolecular hydrogen bonding is
less probable, since the first pKa value of the phosphorane is
-
1
18
lower than that of the 2′-hydroxyl group. It has been previously
reported that downfield shift of the 2′-hydroxyl proton in the
bonding or stabilize the dianionic phosphorane intermediate by
hydrogen bonding to a phosphorane oxyanion.2,3,7,15
1
H NMR spectrum (in CDCl3) indicates the existence of weak
The leaving group may well be stabilized by a hydrogen-
bond formation between the departing 3′-oxyanion and the
neighboring 2′-hydroxyl group. The influence of a vicinal cis-
hydroxyl group on the acidity of a sugar OH group is, however,
relatively modest. It has been shown1 that the pKa values of
aldofuranosides having cis- or trans-2,3-diol systems differ only
by about 0.5 units. Additionally, removal of the 3′-hydroxyl
group from adenosine has been reported to increase the pKa of
the 2′-hydroxyl group only by 0.54 units,1 and part of this
difference undoubtedly results from weaker electron withdrawal
by hydrogen compared to a hydroxyl group. For example, with
adenosine and ara-adenosine, the pKa difference is not larger
than 0.23 units1 The most reliable estimation for the effect
of the 2′-O-methylation is offered by the ∆pKa value between
the 3′-hydroxyl group of adenosine and 2′-O-methyladenosine,
which has been reported to be 0.74 units.1 Accordingly, only
intramolecular hydrogen bonding even in the ground state of
2
ribonucleoside 3′-dimethyl phosphates.
2
+
Mechanism of the Zn -Promoted Cleavage of P-O3′
Bond. As mentioned above, the cleavage of 3′,3′-GpU and 3′,3′-
6a
2
+
GpMe2U is first order in the concentration of Zn (Figure 3),
indicating that only one metal ion is involved on going to the
transition state. The first-order dependence of the rate on
hydroxide-ion concentration, in turn, refers to deprotonation of
the predominant ionic form (monoanion) of 3′,3′-GpU or metal
aquo ion prior to cleavage of the P-O3′ bond (Scheme 4).
Accordingly, a dianionic phosphorane may be expected to be
6b
2+
obtained. The Zn -promoted cleavage of 3′,3′-GpMe2U through
6b.
this species is approximately as fast as the corresponding
1
2
reaction of 3′,5′-UpU: the second-order rate constants for the
2
+
-1
hydroxide-ion-catalyzed cleavage at [Zn ] ) 5 mmol L are
6c
(
17) Kosonen, M.; Yousefi-Salakdeh, E.; Str o¨ mberg, R.; L o¨ nnberg, H. J. Chem.
(
15) Herschlag, D.; Ekstein, F.; Cech, T. R. Biochemistry 1993, 32, 8312-
321.
16) (a) Vesala, A.; K a¨ ppi, R.; L o¨ nnberg, H. Carbohydr. Res. 1983, 119, 25-
0. (b) Velikyan, I.; Acharya, S.; Trifonova, A.; F o¨ ldesi, A.; Chatto-
Soc., Perkin Trans. 2 1997, 2661-2666.
8
(18) (a) Lopez, X.; Schaefer, M.; Dejaegere, A.; Karplus, M. J. Am. Chem. Soc.
2002, 124, 5010-5018. (b) Davies, J. E.; Doltsinis, N. L.; Kirby, A. J.;
Roussev, C. D.; Sprik, M. J. Am. Chem. Soc. 2002, 124, 6594-6599.
(19) Mikkola, S.; Kosonen, M.; L o¨ nnberg, H. Curr. Org. Chem. 2002, 6, 523-
538.
(
3
padhyaya, J. J. Am. Chem. Soc. 2001, 123, 2893-2894. (c) Str o¨ mberg, R.;
A° str o¨ m, H.; Limen, E. Personal communication.
J. AM. CHEM. SOC.
9
VOL. 126, NO. 35, 2004 11043

A R T I C L E S
Kiviniemi et al.
K) were referred to internal TMS and 31P NMR shifts (162 MHz, 300
K) to external orthophosphoric acid. The mass spectra were acquired
using a Perkin-Elmer Sciex API 365 triple-quadrupole LC/MS/MS
spectrometer.
Scheme 4
Materials. The preparation of guanosyl-(3′,3)-(2′,5′-di-O-methyl-
4
2
uridine) has been described previously. N -(p-Isopropylphenoxyacetyl)-
′-O-(4,4′-dimethoxytrityl)-2′-O-(tert-butyldimeth ylsilyl)guanosine 3′-
5
2
(2-cyanoethyl-N,N-diisopropylphosphoramidite) and N -acetyl-5′-O-
4,4′-dimethoxytrityl)-2′-O-(triisopropylsilyloxymethyl)guanosine 3′-
(
(2-cyanoethyl-N,N-diisopropylphosphoramidite) were commercial products
of Glen Research. Uridine monophosphates (6a, 7a, and 8a), guanosine
monophosphates (6b, 7b, and 8b), guanine (9), guanosine (3b), and
uridine (3a), all used as reference materials, were commercial products
of Sigma.
2
Guanosyl-(3′,3′)-uridine (1a). N -(p-Isopropylphenoxyacetyl)-5′-O-
(4,4′-dimethoxytrityl)-2′-O-(tert-butyldimethylsilyl)guanosine 3′-(2-cya-
35 and 142 L mol s-1, respectively (T ) 363.2 K and I )
-
1
noethyl-N,N-diisopropylphosphoramidite) and 2′,5′- bis-O-(tert-bu-
tyldimethylsilyl)uridine were dissolved in a solution of 0.45 M tetrazole
1
0
4
-
1
.1 mol L with sodium nitrate). This is expected on the basis
(
0.33 mmol) in dry acetonitrile (1.5 mL). After 2 h of stirring, the
phosphite ester formed was oxidized with 0.1 M I in THF:H O:lutidine
4:2:1). The crude product was isolated by a CH Cl /aq. NaHSO work
20
2+
of the mechanism proposed earlier for the Zn -promoted
cleavage of ribonucleoside 3′-phosphodiesters. The rate-limiting
step is breakdown of the phosphorane intermediate, more
precisely cleavage of the P-O5′ bond concerted with an
intracomplex proton transfer from the aquo ligand of the
phosphorane-bound zinc ion. The âlg value of the reaction has
been shown to be low (-0.32), and hence, the reaction is not
markedly susceptible to the basicity of the leaving group. The
2
2
(
2
2
3
up and purified on a silica gel column eluted with a mixture of
dichloromethane and methanol (90:10%, v/v). The dimethoxytrityl
group was removed with 80% aqueous acetic acid solution (7 h). The
reaction mixture was evaporated to dryness, and the residue was purified
on a silica gel column eluted with a mixture of dichloromethane and
methanol (90:10%, v/v). The detritylated product was dissolved in
saturated methanolic ammonia. After being stirred for 20 h, the solution
was evaporated to dryness and purified on a silica gel column eluted
with a mixture of dichloromethane and methanol (70:30%, v/v). The
3
2
′-linked uridine is cleaved 2.0 times as rapidly as the 3′-linked
′,5′-di-O-methyluridine, the second-order rate constant obtained
-
1 -1
for 1a being 280 L mol s . This modest rate enhancement
may tentatively be attributed to the ability of the 2′-hydroxyl
group to serve, in addition to the phosphorane-bound Zn aquo
ion, as an intracomplex general acid, protonating the departing
-
1
purified product was dissolved in 1 mol L solution of tetrabutylam-
monium fluoride (0.265 g, 1.02 mmol) in tetrahydrofuran (1 mL), and
the solution was stirred for 20 h at room temperature. The mixture
was evaporated to dryness and purified by reversed phase chromatog-
raphy on a Lobar RP-18 column (37 × 440 mm, 40-63 mm), elut-
ing with a mixture of water and acetonitrile (92:8%, v/v). Finally,
2+
3
′-oxygen atom concerted with the bond cleavage.
Mechanisms of the Acid-Catalyzed Cleavage and Isomer-
ization. As shown by the pH-rate profiles in Figures 1 and 2,
the reactivity difference between 3′,3′-GpU, 3′,3′-GpMe2U, and
+
the product was passed through a Na -form Dowex 50-W (100-
3
1
200 mesh) cation exchange column. P NMR (δ
P
) (162 MHz, D
2
O):
1
2
.13. H NMR (δ ) (400 MHz, D O): 7.85 (s, 1H), 7.73 (d, 1H, J )
H 2
3′,5′-UpU is modest in acidic solutions. Under these conditions,
8
.12 Hz), 5.79 (d, 1H, J ) 4.92 Hz), 5.77 (d, 1H, J ) 5.56 Hz), 5.72
the reactions proceed via a neutral or monocationic phosphorane
intermediate and the leaving group departs as an alcohol instead
+
(d, 1H, J ) 8.12 Hz), 4.18-4.72 (m, 6H), 3.69-3.81 (m, 4H). ESI -
+
-
11
MS: m/z 588.5 [M + H] . HRMS (FAB) M calcd 588.1091, obsd
of alkoxide ion. Accordingly, stabilization of the phosphorane
intermediate or leaving group by intramolecular hydrogen
bonding does not play a role. Consistent with the observed
similarity of the reaction rates, the âlg value of the acid-catalyzed
cleavage and isomerization of ribonucleoside 3′-phosphodiesters
has been reported to be only slightly negative, -0.12 and -0.18,
respectively.17 The pH-independent cleavage, observed over a
narrow pH range of 4-5, proceeds via rapid initial formation
of a phosphorane monoanion that undergoes rate-limiting
breakdown by concerted proton transfer from the phosphorane
hydroxyl ligand to the departing oxygen.1 This means that
stabilization of neither the phosphorane intermediate nor the
leaving oxygen by intramolecular hydrogen bonding with the
5
88.1087.
Kinetic Measurements. The reactions were carried out in sealed
tubes immersed in a thermostated water bath (363.2 ( 0.1 K). The
hydronium-ion concentration of the reaction solutions was adjusted with
hydrogen chloride, sodium hydroxide, and formate, acetate, (N-[2-
hydroxyethyl]piperazine-N,-[2-ethanesulfonic acid]) (HEPES), and gly-
cine buffers. Low buffer concentration was used (30 60 mmol L ).
The initial substrate concentration was ca. 0.1 mmol L-1. The
composition of the samples withdrawn at appropriate intervals was
analyzed on a Hypersil ODS 5 column (4 × 250 mm, 5 mm) using
mixtures of acetonitrile and an acetic acid/sodium acetate buffer (0.045/
21
-
-1
9
-1
-1
0.015 mol L ) containing 0.1 mol L ammonium chloride as an eluent.
Good separation of the product mixture of 1a was obtained when a 25
min isocratic elution with buffer was followed by a linear gradient (1
min) up to 5.0% acetonitrile. After this isocratic elution with a 5.0%
content of acetonitrile (v/v) was continued. The observed retention times
2
′-hydroxy group is not important. As seen from Figure 1, the
pH-independent cleavage is much less susceptible to the 2′-O-
methylation than the hydroxide-ion-catalyzed cleavage.
R
(t /min) for the hydrolytic products of 1a on RP HPLC (flow rate was
-
1
1
6
mL min ) were 36.0 (5), 23.0 (3b), 15.0 (6b), 8.1 (8b, 7b, and 3a),
.7 (8a), 6.2 (9), 5.4 (7a), 38.0-37.0 (2a and 4), and 4.9 (6a). Uridine
Experimental Section
Methods. The NMR spectra were recorded on a Bruker AM 200 or
(3a), 3′-GMP (7b), and 2′-GMP (8b) were separated by isocratic elution
1
JEOL 400 spectrometer. The H NMR chemical shifts (400 MHz, 300
(
21) Serjeant, E. P.; Dempsey, B. Ionization Constants of Organic Acids in
Aqueous Solution; IUPAC Chemical Data Series No. 23; Pergamon:
Oxford, 1979.
(
20) Mikkola, S.; Stenman, E.; Nurmi, K.; Yousefi-Salakdeh, E.; Str o¨ mberg,
R.; L o¨ nnberg, H. J. Chem Soc., Perkin Trans. 2 1999, 1619-1625.
11044 J. AM. CHEM. SOC.
9
VOL. 126, NO. 35, 2004

Hydrolytic Reactions
A R T I C L E S
Table 1. Rate Constants for the Cleavage (Route A, k1; and
Route B, k2) of 3′,3′-GpU (1a) at 363.2 K
t. 2a, 2b, 4, and 5 are the concentrations of the isomers of 3′,3′-GpU
or guanosyl-(3′,3)-(2′,5′-di-O-methyluridine) at time t.
pH
9.1
8.0
7.1
6.0
5.6
5.0
[2a + 4 + 5]t
kis )
kdi
(3)
-
4
4
-1
-1
k1/× 10
k2/× 10
s
s
48.0
39.9
7.51
7.09
1.18
1.16
0.172
0.172
0.082
0.082
0.026
0.023
[
1a] - [1a]
0 t
-
[
2b]t
kis )
kdi
(4)
[
1b] - [1b]
0
t
with 0.1 mol L- ammonium acetate. The observed retention times(t
1
R
/
Under more acidic conditions (pH < 4), where depurination competes
with cleavage and isomerization, the first-order rate constants (kdp) for
the depurination of isomeric GpUs (route C in Scheme 1 or route I in
Scheme 2) were calculated using eq 5 by bisecting the rate constant
min) were 12.2 (6a), 14.0 (8b), and 5.9 (7b). In the case of 1b, sam-
ples were analyzed on a Hypersil ODS column (4 × 125 mm,
5
µm). With 1b a 5 min isocratic elution with buffer was followed by
a linear gradient (5 min) up to 9% acetonitrile. The observed reten-
tion times (t /min) for the hydrolytic products of 1b on RP HPLC
(kdec) for the decomposition of the isomeric GpUs to the rate constants
R
of parallel first-order reactions on the basis of the product distribution
at the early stages of the reaction, i.e., under conditions where the
formation of guanine from the cleavage products may be neglected.
1
(
(
flow rate was 0.5 mL min- ) were 13.1 (3c), 12.2 (2b), 6.6 (3b), 4.3
6b), 2.5 (7b), 2.6 (8b), and 1.9 (9). The observed retention times for
starting materials, 1a and 1b, were 42.5 and 15.6, respectively. To
identify the reaction products, the mass spectra of the hydrolysis
products were recorded (LC/MS). A mixture of acetonitrile and aqueous
[
Guanine]
sum concentration of GpU isomers (1a, 2a, 4, and 5 in the case of
′,3′-GpU and 2a and 2b in the case of guanosyl-(3′,3)-(2′,5′-di-O-
t t
and [GpUs] are the concentration of guanine (9) and the
3
-
1
ammonium acetate (5 mmol L ) was used as an eluent. In addition,
the products formed were identified by spiking with authentic reference
compounds.
methyluridine)), respectively, at time t. [GpU]
0
denotes the initial
concentration of the starting material 1a or 1b.
Calculation of the Rate Constants. The pseudo-first-order rate
constants for the disappearance of 1a and 1b (kdi) were obtained by
applying the integrated first-order rate equation to the time-dependent
diminution of the concentration of the starting material. At pH > 6,
where cleavage of the starting material is the only reaction detected,
the rate constant (kcl) for cleavage of 1a was bisected to the rate
[Guanine]t
kdp
)
kdec
(5)
[
GpU] - [GpUs]
0 t
Equation 6 was then applied to obtain the rate constants (kcl) for
cleavage of isomeric GpUs (routes A and B in Scheme 1 and route G
in Scheme 2).
1 2
constants of the two parallel first-order reactions (k , route A; and k ,
kcl ) kdec - kdp
(6)
route B in Scheme 1) by eq 1 and (2) on the basis of the product
distribution at the early stages of the reaction, i.e., under conditions
where the decomposition of the cyclic phosphates may be neglected
Conclusions
In summary, comparative kinetic measurements with guano-
syl-(3′,3′)-uridine (1a) and its 2′,5′-di-O-methyluridine analogue
(see Table 1).
(
1b) have shown that the 2′-hydroxyl group of the uridine
[
Guo] + [2′,3′-cUMP]
t
t
moiety, as an intramolecular hydrogen-bond donor, accelerates
the cleavage of dinucleoside 3′,3′-monophosphates via a dian-
ionic phosphorane by a factor 23. In all likelihood, stabilization
of both the dianionic phosphorane intermediate and the departing
nucleoside 3′-oxyanion by hydrogen bonding contributes to the
observed rate enhancement. These results, together with the
previous finding that a 3′-linked nucleoside inherently is cleaved
k1 )
kdi (1)
kdi (2)
[
Guo] + [2′,3′-cUMP] + [Urd] + [2′,3′-cGMP]
t
t
t
[
Urd] + [2′,3′-cGMP]
t
t
k2 )
[
Urd] + [2′,3′-cGMP] + [Guo] + [2′,3′-cUMP]
t t t t
At pH 4-6, where isomerization competes with cleavage, the first-
3
times as fast as its 5′-counterpart, largely explain the well-
order rate constants (kcl) for the cleavage of isomeric GpUs (routes A
and B in Scheme 1 and route G in Scheme 2) were obtained by applying
the integrated first-order rate equation to the time-dependent sum
concentration of the GpU isomers. The first-order rate constants of the
isomerization of the starting material, kis (route D in Scheme 1 and
route H in Scheme 2), were calculated by eq 3 or 4 by bisecting the
rate constant of the disappearance of the starting material (kdi) to the
rate constants of parallel first-order reactions on the basis of the product
distribution at the early stages of the reaction, i.e., under conditions
where the reverse reaction of the 3′ to 2′ migration may be neglected.
known regioselectivity governing the breakdown of the phos-
phorane intermediate of group I introns.
Acknowledgment. We thank Prof. Harri L o¨ nnberg for fruitful
discussions.
Supporting Information Available: Time-dependent product
distributions for the hydrolysis of 3′,3′-GpU and 3′,3′-GpMe U.
2
This material is available free of charge via the Internet at http://
pubs.acs.org. See any current masthead page for ordering
information and Web access instructions.
[1a]
0
and [1b]
0
are the initial concentrations of 3′,3′-GpU and guanosyl-
and [1b] denote the concentra-
(
3′,3)-(2′,5′-di-O-methyluridine). [1a]
t
t
tions of 3′,3′GpU and guanosyl-(3′,3)-(2′,5′-di-O-methyluridine) at time
JA047568I
J. AM. CHEM. SOC.
9
VOL. 126, NO. 35, 2004 11045