Hydrolysis and intramolecular transesterification of ribonucleoside 3′-phosphotriesters: The effect of alkyl groups on the general and specific acid-base-catalyzed reactions of 5′-O-pivaloyluridin-3′-yl dialkyl phosphates

Article abstract of DOI:10.1039/a707095k

Diisopropyl, diethyl, bis(2-methoxyethyl) and isopropyl 2-methoxyethyl esters of 5′-pivaloyl-2′-(tetrahydropyran-2-yl)uridin-3′-yl phosphate have been prepared. The 2′-protecting group has been removed under acidic conditions, and the isomerization of the resulting ribonucleoside 3′-phosphotriester to its 2′-counterpart and the cleavage of the isomeric mixture to 2′- and 3′-phosphodiesters and a 2′,3′-cyclic phosphate has been followed by reverse phase HPLC in aqueous hydrogen chloride and several buffer solutions over a wide acidity range from H₀ - 1.5 to pH 8. The β_1g values of the buffer-independent partial reactions, and the β_1g and Bronsted α and β values of the buffer catalyzed reactions have been determined. The mechanisms of various partial reactions are discussed on the basis of the structural effects observed.

Full text of DOI:10.1039/a707095k

View Article Online / Journal Homepage / Table of Contents for this issue
Hydrolysis and intramolecular transesterification of ribonucleoside
3Ј-phosphotriesters: the effect of alkyl groups on the general and
specific acid–base-catalyzed reactions of 5Ј-O-pivaloyluridin-3Ј-yl
dialkyl phosphates
Markus Kosonen,*,† Kristo Hakala and Harri Lönnberg
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
Diisopropyl, diethyl, bis(2-methoxyethyl) and isopropyl 2-methoxyethyl esters of 5Ј-pivaloyl-2Ј-(tetra-
hydropyran-2-yl)uridin-3Ј-yl phosphate have been prepared. The 2Ј-protecting group has been removed
under acidic conditions, and the isomerization of the resulting ribonucleoside 3Ј-phosphotriester to its
2Ј-counterpart and the cleavage of the isomeric mixture to 2Ј- and 3Ј-phosphodiesters and a 2Ј,3Ј-cyclic
phosphate has been followed by reverse phase HPLC in aqueous hydrogen chloride and several buffer
solutions over a wide acidity range from H0 Ϫ 1.5 to pH 8. The âlg values of the buffer-independent partial
reactions, and the âlg and Brønsted á and â values of the buffer catalyzed reactions have been determined.
The mechanisms of various partial reactions are discussed on the basis of the structural effects observed.
monoanionic phosphorane intermediate as oxyanions much
Introduction
more readily than the exocyclic alkoxy group. The exocyclic
bond fission thus requires general acid catalysis at pH 3 to 7, i.e.
under conditions where the endocyclic cleavage is uncatalyzed.
The present paper is an extension of our triester approach. The
acidity of the esterified alcohol(s) has now been varied, and the
βlg values for various buffer-independent and buffer-dependent
partial reactions have been determined to elucidate the transi-
tion state structures of reactions proceeding by an intra-
molecular nucleophilic attack on neutral or monocationic
phosphoester.
The kinetics and mechanisms of the phosphodiester bonds
of RNA have received increasing interest during the past
decade.1–3 As far as catalysis by Brønsted acids and bases is
concerned, the motive of these studies ranges from improved
protecting group strategies for oligoribonucleotide synthesis4,5
to the development of artificial protolytic cleaving agents6–16
and better insight into the action of ribonucleases.1,2,17 One of
the major problems encountered in mechanistic studies of ribo-
nucleoside 3Ј-phosphodiesters is the reliable assignment of the
reactive ionic form. In particular, several kinetically equivalent
mechanisms differing only in the position of a proton in the
transition state may be written for the reactions predominat-
ing under neutral conditions.2,18–21 We have previously22,23
attempted to solve some of the resulting ambiguities by using
a ribonucleoside 3Ј-dimethyl phosphate 1 as a mimetic of the
neutral ionic form of ribonucleoside 3Ј-phosphodiester 2.
Results
Preparation of the model compounds
The symmetric 3Ј-phosphotriesters of 5Ј-O-pivaloyl-2Ј-O-
(tetrahydropyran-2-yl)uridine 3a–c were prepared as described
previously for the corresponding dimethyl ester 3d.23 Accord-
ingly, the more polar diastereomer of 5Ј-O-pivaloyl-2Ј-O-
(tetrahydropyran-2-yl)uridine was phosphorylated with phos-
phoryl tris(1,2,4-triazolide),25 and the remaining two triazole
groups were displaced with 2-methoxyethanol, ethanol or
propan-2-ol to obtain 3a, 3b and 3c, respectively. The asym-
metric triester, 3e, was prepared by phosphorylating the same
starting nucleoside with 2-chlorophenyl phosphoyl bis(1,2,4-
triazolide), alcoholyzing the product in propan-2-ol, and dis-
placing the 2-chlorophenoxy group with 2-methoxyethoxide
ion in a mixture of 2-methoxyethanol and 1,4-dioxane. All
the compounds 3a–c,e were purified by adsorption chrom-
atography, and when needed additionally by RP chrom-
atography. The identity of 3a–c,e was verified by elemental
analysis, and by 1H and 31P NMR and FAB mass spectroscopy.
O
P
OH
OMe
O
OH
O
O
P
OH
OMe
OR
1
2
These studies have, among other things, suggested that a
monoanionic phosphorane intermediate is obtained by the
attack of 2Ј-oxyanion on neutral phosphodiester rather than
via the predominant ionic form, i.e. by the attack of 2Ј-hydroxy
function on the phosphodiester monoanion. The recent quan-
tum chemical calculations,24 according to which the attack of
an anionic nucleophile on protonated (neutral) phosphoester
may be even faster than expected on the basis of the reactivity
of its triester mimetic, lend additional support to the reasoning
behind this conclusion: replacing the rapidly exchangeable pro-
ton of a neutral diester with an alkyl group leads to under-
estimation rather than overestimation of the reactivity of this
particular ionic form. Furthermore, the previous23 triester
studies suggest that the sugar hydroxy groups depart from a
Kinetic measurements
All reactions were carried out at 298.2 K and their progress was
followed by analyzing the content of the aliquots withdrawn at
suitable intervals by RP HPLC. When the reactions were fol-
lowed in buffer solutions at pH > 2, 2Ј-O-tetrahydropyran-2-yl
group was first removed with 0.1 mol dmϪ3 aqueous hydrogen
chloride to give 4a–c,e, and the desired buffer system was then
created by adding an appropriate amount of the buffer base.
Under more acidic conditions, hydrolysis of the 2Ј-protecting
group constituted the first step of the reaction sequence fol-
† E-Mail: makoso@utu.fi
J. Chem. Soc., Perkin Trans. 2, 1998
663

View Article Online
lowed, and equations of consecutive first-order reactions were
applied to kinetic calculations.
It has been shown previously22,23 that 5Ј-protected uridin-3Ј-
yl dimethyl phosphates undergo in aqueous solution two paral-
lel reactions: isomerization to the 2Ј-dimethyl phosphate (5d;
Reaction A in Scheme 1) and cleavage to a mixture of 2Ј- and
3a R1 = R2 = CH2CH2OMe
PivO
U
b R1 = R2 = Et
c R1 = R2 = Pri
d R1 = R2 = Me
e R1 = Pri,
O
O
P
OThp
R2 = CH2CH2OMe
R1O
O
R2O
3a–e
A
Fig. 1 pH–rate profile for the mutual buffer-independent isomeriza-
PivO
U
PivO
O
U
k1
tion of 5Ј-O-pivaloyluridin-3Ј-yl and -2Ј-yl dialkyl phosphates at 298.2
K. The ionic strength was adjusted to 1.0 mol dmϪ3 with NaCl, except at
pH < 1.5 where aqueous HCl was used to adjust the hydronium ion
concentration. Notation: 4a/5a (᭝), 4b/5b (ᮀ), 4c/5c (᭺), 4e/5e (᭛).
O
k-1
O
P
OH
OH
R1O
O
P
R1O
O
O
R2O
R2O
B
kcl
4a–e
5a–e
PivO
U
O
Rx = R1 or R2
O
O
P
ORx
O
PivO
U
PivO
U
PivO
U
O
O
O
P
8
OH
O
O
P
RxO
OH
O
O
–O
O
–O
P
RxO
O
Fig. 2 pH–rate profiles for the buffer-independent cleavage of 5Ј-O-
pivaloyluridin-2Ј-yl and -3Ј-yl dialkyl phosphates at 298.2 K. The ionic
strength was adjusted to 1.0 mol dmϪ3 with NaCl, except at pH < 1.5
where aqueous HCl was used to adjust the hydronium ion concentra-
tion. Notation: 4a/5a (᭡), 4b/5b (᭿), 4c/5c (᭹), 4e/5e (᭜).
O–
O
6a–d
7a–d
Scheme 1
indicated in Table 1, and extrapolating the observed pseudo-
first-order rate constant to zero buffer concentration by linear
3Ј-methyl phosphates (6d, 7d) and a 2Ј,3Ј-cyclic phosphate (8;
Reaction B in Scheme 1). The latter reaction presumably pro-
ceeds via a 2Ј,3Ј-cyclic triester which, however, is too unstable
to be detected. In all likelihood, the symmetric triesters 4a–c
investigated in the present study react similarly. With the
asymmetric ester, 4e, the situation is more complicated, since in
principle either of the alkoxy groups, PriO or MeOCH2CH2O,
may be initially displaced. In fact, all possible diesters 6a, 6c,
7a, 7c and 8 were formed in aqueous hydrogen chloride and
carboxylic acid buffers, the ratio of [6a ϩ 7a]/[6c ϩ 7c] being
approximately 1:4. In more basic amine buffers, only the
isopropyl diesters (6c, 7c) were formed.
o
o
regression. The numerical values of kis and kcl are listed in
Table 1. While both of these rate constants are markedly
increased with the increasing electronegativity of the alkyl
groups, the shape of the pH–rate profiles remains almost
unchanged, being similar to that reported previously23 for the
dimethyl ester 4d. The buffer-independent isomerization and
cleavage of all the compounds are first-order in hydronium ion
concentration at pH 2 to 0, and show a modest negative devi-
ation from the linear dependence of 1g (ko/sϪ1) on H0 in concen-
trated solutions of hydrogen chloride. The latter downward
curvature possibly results from nearly quantitative protonation
of the phosphoryl oxygen, which alters the reaction order in
hydronium ion activity from one to zero. If this interpretation is
correct, the pKa value of the protonated triester is about Ϫ0.7.
The cleavage remains pH-independent from pH 2 to 7, and then
becomes hydroxide ion catalyzed, whereas the isomerization is
Fig. 1 shows the pH–rate profiles for the buffer-independent
isomerization of 4a–c,e to their 2Ј-counterparts (5a–c,e) and
Fig. 2 those for the buffer-independent cleavage of the 2Ј/3Ј-
isomeric mixture of the same compounds to a mixture of
diesters. The concentration ratio of the 2Ј- and 3Ј-triesters
approached unity with time, indicating that the migration of
the dialkyl phosphate group between the 2Ј- and 3Ј-hydroxy
functions is as fast in both directions (k1 ≈ kϪ1). The buffer-
o
hydroxide ion catalyzed at pH > 3. Accordingly, kcl may be
expressed by eqn. (1), where the partial rate and equilibrium
ϩ
kclo = kclH/(1 ϩ Ka/aHϩ) ϩ kclw ϩ kclOHKw/aH
(1)
o
independent rate constants, kiso = k1o ϩ kϪ1 for the isomeriz-
ation and kclo for the cleavage, were obtained by carrying out the
kinetic measurements at four different buffer concentrations
constants (kclH, kclw, kclOH, Ka) are those defined in Scheme 2,
664
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
Table 1 Rate constants for the cleavage and isomerization of 5Ј-O-pivaloyluridin-3Ј-yl dialkyl phosphates (4a–c,e) in carboxylic acid and amine
buffers at 298.2 K (I = 1.0 mol dmϪ3 with sodium chloride)
kclbf/10Ϫ6 dm3
kisbf/10Ϫ6 dm3
Compound
Buffer acid
ra
molϪ1 sϪ1 b
kclo/10Ϫ6 sϪ1 c
molϪ1 sϪ1 d
kiso/10Ϫ6 sϪ1 e
4a
Cyanoacetic acid
Chloroacetic acid
Formic acid
3:1
1:3
3:1
1:3
3:1
1:1
1:3
1:1
1:3
4:1
1:1
4:1
13.7 1.5
9.8 0.7
9.5 0.5
13.4 0.2
28.5 0.1
8.2 0.5
4.23 0.21
4.46 0.14
3.10 0.07
3.8 0.3
61
177
4
6
40.2 1.2
104.1 1.8
90.5 1.1
3.2 0.3
5.0 1.3
4.4 2.4
12.3 0.1
Acetic acid
Propionic acid
Triethanolammonium ion
145
268
4
8
19.2 1.4
70 50
4400 400
32
5
Diethanolammonium ion
Cyanoacetic acid
Chloroacetic acid
Formic acid
90 30
4b
3:1
1:3
3:1
1:3
3:1
1:3
3:1
1:3
3:1
1:3
4:1
1:1
4:1
5.13 0.12
2.45 0.09
2.9 0.4
1.9 0.3
3.96 0.10
7.7 0.5
2.15 0.04
0.94 0.03
1.02 0.12
0.50 0.07
0.58 0.03
0.65 0.15
0.53 0.12
0.4 0.3
0.52 0.05
0.53 0.15
0.677 0.022
1.96 0.12
3.4 0.5
3.4 1.1
1.8 0.4
6.6 1.6
4.6 1.9
79 12
1.44 0.16
1.9 0.4
1.58 0.11
6.0 0.5
6.9 0.7
29
4
Acetic acid
8.5 0.4
Ϫ1.0 0.4
66.1 1.1
27.1 0.9
10.8 0.2
32.7 0.5
5.7 0.2
13.3 1.3
690 90
Propionic acid
Ϫ1
2
89
6
Triethanolammonium ion
Diethanolammonium ion
Cyanoacetic acid
Formic acid
6
8
4c
4e
3:1
1:3
3:1
1:3
3:1
1:3
1:1
0.64 0.04
0.25 0.03
0.26 0.03
0.347 0.011
0.118 0.011
0.040 0.009
0.55 0.11
0.34 0.03
0.39 0.09
0.9 0.5
0.70 0.16
9.3 1.9
0.27 0.03
0.142 0.008
0.18 0.03
1.05 0.15
1.50 0.05
12.2 0.6
Acetic acid
0.41 0.04
0.74 0.20
0.1 0.3
0.080 0.002
Triethanolammonium ion
Cyanoacetic acid
Formic acid
3:1
1:3
3:1
1:3
3:1
1:3
4:1
1:1
4:1
4.1 0.7
1.8 0.3
2.51 0.12
5.92 0.18
5.3 0.4
15.9 0.2
0.93 0.07
1.83 0.21
0.86 0.10
0.71 0.03
0.41 0.05
0.62 0.11
0.38 0.07
1.587 0.007
4.8 0.2
5.9 0.6
4.15 0.12
5.5 0.8
1.54 0.21
1.24 0.04
3.5 0.2
8
5
9
6
1
22
3
Acetic acid
30.0 1.8
280 40
Ϫ1
Triethanolammonium ion
Diethanolammonium ion
5
2
6.41 0.14
20.8 1.3
a Buffer ratio: r = [HA]/[AϪ] and [BHϩ]/[B] for carboxylic acids and amine buffers, respectively. b Second-order rate constant for the buffer catalyzed
cleavage. The first-order rate constants were determined at [HA] ϩ [AϪ] = 0.050, 0.10, 0.30 and 0.50 mol dmϪ3, [BHϩ] ϩ [B] = 0.050, 0.075,
0.100, 0.125 mol dmϪ3 c The observed first-order rate constant for the cleavage at buffer concentration zero. d The second-order rate constant
.
for the isomerization. For experimental conditions see footnote b. e The observed first-order rate constant for the isomerization at buffer
concentration zero.
Table 2a Partial rate constants for the buffer-independent cleavage of 4a–c,e at 298.2 K (I = 1.0 mol dmϪ3 with NaCl)a
kclH/Ka/10Ϫ4
kclOH/dm3
Compound
Ka/mol dmϪ3
kclH/10Ϫ4 sϪ1
dmϪ3 molϪ1 sϪ1
kclw/10Ϫ7 sϪ1
molϪ1 sϪ1
4a
4b
4c
4e
5.0 1.8
5.0 2.1
3.9 1.2
6.0 1.6
15.5 1.8
6.5 2.1
0.74 0.17
5.4 1.2
2.0b
37.3 2.2
5.07 0.28
~0.4
14.6 1.6
1.1b
0.76 0.09
0.0234 0.0004
2.52 0.32
0.19b
0.90b,c
5.1 0.5d
c
a For the rate constants, see Scheme 2. b Estimated accuracy ca. 5%. kclH/Ka for the cleavage of the 2-methoxyethoxy group is 0.72 × 10Ϫ4 dmϪ3
w
molϪ1
s
Ϫ1, and that for the cleavage of the isopropoxy group is 0.18 × 10Ϫ4 dmϪ3 molϪ1 sϪ1
.
d kcl for the cleavage of the 2-methoxyethoxy group is
4.1 × 10Ϫ7 sϪ1, and that for the cleavage of the isopropoxy group is 1.0 × 10Ϫ7 sϪ1
.
and Kw is the ionic product of water under the experimental
conditions. A similar eqn. (2) was used to fit the values of kis ,
Table 1 also records the second-order rate constants for the
buffer catalyzed isomerization (kisbf) and cleavage (kclbf) of 4a–c,
e. The buffer catalyzed cleavage clearly predominates over the
buffer independent one at pH > 3. In 0.1 mol dmϪ3 acetic acid–
sodium acetate buffer ([AcOH]/[AcOϪ] = 1:3), for example,
87% of the cleavage of 4b/5b is buffer catalyzed. In striking
contrast, buffer catalyzed isomerization is detected only in the
most acidic buffers, i.e. in cyanoacetic, chloroacetic and formic
o
ϩ
kiso = kisH/(1 ϩ Ka/aHϩ) ϩ kisw ϩ kisOHKw/aH
(2)
although in this case the kisw term is not necessarily significant.
The values obtained by least-squares fitting for the partial con-
stants are given in Table 2.
J. Chem. Soc., Perkin Trans. 2, 1998
665

View Article Online
Table 2b Partial rate constants for the buffer-independent isomeriz-
ation of 4a–c,e at 298.2 K (I = 1.0 mol dmϪ3 with NaCl)a
kisH/Ka/10Ϫ4
kisOH/103 dm3
Compound
dmϪ3 molϪ1 sϪ1
kisw/10Ϫ7 sϪ1
molϪ1 sϪ1
4a
4b
4c
4e
2.32 0.12
0.39b
55 11
6.2 2.5
0.8 0.3
1.6 3.0
1490 80
241 29
6.0 0.9
132 16
0.064b
0.63b
a For the rate constants, see Scheme 2. b Estimated accuracy ca. 5%.
Table 3 Second-order rate constants for the general acid and general
base catalyzed cleavage at 298.2 K (I = 1.0 mol dmϪ3 with NaCl)
HA
AϪ
kcl
/
kcl /
10Ϫ6 dmϪ3
molϪ1 sϪ1
10Ϫ6 dmϪ3
molϪ1 sϪ1
Compound
Buffer acid
Fig. 3 Logarithmic second-order rate constants for the hydronium ion
4a
Cyanoacetic acid
Chloroacetic acid
Formic acid
Acetic acid
Propionic acid
Triethanolammonium ion
Diethanolammonium ion
Cyanoacetic acid
Chloroacetic acid
Formic acid
Acetic acid
Propionic acid
Triethanolammonium ion
Diethanolammonium ion
Cyanoacetic acid
Triethanolammonium ion
Cyanoacetic acid
Formic acid
Acetic acid
Triethanolammonium ion
Diethanolammonium ion
15.7
7.71
7.85
15.3
122
290
357
catalyzed isomerization of symmetric 5Ј-O-pivaloyluridin-3Ј-yl and -2Ј-
yl dialkyl phosphates (4a–d/5a–d) and their cleavage to a mixture of
diesters at 298.2 K plotted against the pKa value of the esterified alco-
hol.31 The filled circles refer to cleavage and the open circles to isomeriz-
ation. The values for 4d/5d are taken from Ref. 23. The slopes of the
lines are cleavage Ϫ0.51 0.13, isomerization Ϫ0.65 0.04.
96
22 200
1.11
1.32
9.56
36.4
43.4
27.5
3500
0.063
1.49
0.59
7.63
21.1
7
4b
6.47
3.44
2.10
Discussion
Hydronium ion catalyzed cleavage and isomerization
As seen in Figs. 1 and 2, the isomerization of 5Ј-O-pivaloyl-
uridin-3Ј-yl dialkyl phosphates (4a–e) to their 2Ј-isomers (5a–e;
Reaction A in Scheme 1) and their cleavage to a mixture of
diesters (Reaction B) are both acid catalyzed at pH < 2. The
logarithmic rate constants of these reactions are plotted against
the pKa value of the esterified alcohol in Fig. 3. Although the
acidity range of the alcohols is rather narrow and steric effects
undoubtedly play a role in addition to polar effects, it is clear
that both reactions are accelerated by the increasing electro-
negativity of the alkyl groups, the effect being more marked on
isomerization than on cleavage. The βlg value for the cleavage
is Ϫ0.51 0.13, and the β value for the isomerization
Ϫ0.65 0.04. Accordingly, the susceptibility to the electro-
negativity of the alkyl group is considerably greater than that
reported for the corresponding reactions of 3Ј-phosphodiesters:
β1g = Ϫ0.12 0.05 for the cleavage and β = Ϫ0.18 0.02 for the
isomerization.26
4c
4e
0.819
5.33
0.81
0.075
3200
H
kcl
H
kis
O
+
OH
OR1
HO
O
+
HO
P
OR1
HO
P
The hydronium ion catalyzed reactions of both 3Ј-diesters
and 3Ј-triesters have previously been suggested to proceed by a
similar mechanism. With diesters, the 2Ј-hydroxy function
attacks on the monocationic phosphodiester obtained by
preequilibrium protonation of both of the non-bridging phos-
phoryl oxygens (Scheme 3).3,27,28 The resulting monocationic
phosphorane intermediate then undergoes a kinetically invis-
ible pseudorotation and protolytic rearrangement via the neu-
tral ionic form. This process may bring either protonated 2ЈO,
3ЈO or OR to an apical position and hence lead to breakdown
of the intermediate back to the starting material, to the iso-
merization product (Reaction A), or to the cleavage product
(Reaction B), respectively. The pseudorotation cannot be the
rate-limiting step of the isomerization as far as the reaction
proceeding by initial formation of a monocationic phos-
phorane is concerned. This intermediate undergoes rapid
thermodynamically favoured deprotonation to a neutral phos-
phorane. Accordingly, if the isomerization take places by rate-
limiting pseudorotation, the product distribution would be pH-
dependent: on increasing the hydronium ion concentration, the
cleavage would become favoured over the isomerization. As
seen from the pH–rate profiles (Figs. 1 and 2), this is not the
case. With triester, the extra alkyl group is assumed to take the
role of one of the rapidly exchangeable protons of a monocati-
onic diester.3,22,23 Accordingly, the structural effects ought to be
OR2
OR2
-H+ Ka
+H+
diesters
kisw+kisOH[OH-]
O
OH
OR1
HO
O
O
P
OR1
O
P
OR2
OR2
kclw+kclOH[OH-]
Scheme 2
acid buffers, and even then its contribution to the observed rate
constant of isomerization is rather small.
The second-order rate constants obtained for the general acid
and general base catalyzed isomerization (kis and kisAϪ) and
HA
cleavage (kclHA and kclAϪ) by the breakdown of kisbf and kclbf via
eqn. (3) are listed in Table 3.
kibf ([HA] ϩ [AϪ]) = kiHA [HA] ϩ kiAϪ [AϪ]; i = cl or is (3)
666
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
O
OH
HO
O
+
HO
+
OR2
P
OR1
HO
P
R1 = alkyl
R2 = H, alkyl
OR2
OR1
+
+
P
O
OH
OR1
HO
HO
O
OR2
P
HO
OR2
OR1
Fig. 4 Logarithmic second-order rate constants for the hydroxide ion
catalyzed isomerization of symmetric 5Ј-O-pivaloyluridin-3Ј-yl and
-2Ј-yl dialkyl phosphates (4a–d/5a–d) and their cleavage to a mixture
of diesters at 298.2 K plotted against the pKa value of the esterified
alcohol.31 The filled circles refer to cleavage (y axis to the left) and the
open circles to isomerization (y axis to the right). The values for 4d/5d
are taken from ref. 23. The slopes of the lines are: cleavage Ϫ1.26
0.07, isomerization Ϫ1.10 0.16.
e
a
O
O
OR2
e
a
O
O
OR1
P
P
HO
HO
OR1
OR2
corresponding 2-ethoxyethyl and isopropyl phosphodiesters is
4.6.26
a/e
e
a/e
O
O
OR2
Hydroxide ion catalyzed cleavage and isomerization
P
HO
e
The isomerization of 4a–e becomes hydroxide ion catalyzed at
pH > 3 and the cleavage at pH > 7. Both reactions are markedly
susceptible to the acidity of the esterified alcohol, the βlg for
the cleavage being Ϫ1.26 0.07 and the β for isomerization
Ϫ1.10 0.16 (Fig. 4). It has been suggested that the first-order
dependence of rate on hydroxide ion concentration results from
preequilibrium deprotonation of the 2Ј-hydroxy function.22,23
The resulting 2Ј-oxyanion then attacks on neutral phospho-
triester centre giving a monoanionic phosphorane, which may
decompose in two alternative manners (Scheme 4). Either
HOR1
+
a
O
O
+
HO
P
OR2
Scheme 3
rather similar with triesters and diester. The extra alkyl group
on triester may, however, be expected to increase the susceptibil-
ity of the triester reaction to the acidity of the esterified alco-
hol. Increasing electronegativity of the alkyl group(s) lowers the
electron density at phosphorus and hence facilitates the attack
of the 2Ј-hydroxy function. Since symmetrical triesters (4a–d)
bear two alkyl ligands the electronegativity of which is varied,
and diesters only one, the influence is more marked with tri-
esters. Evidently the effect on preequilibrium protonation is
less important, since the inductive effect is transmitted to the
phosphoryl oxygen through an additional PO bond. For the
following reasons, the polar effects on the breakdown of the
phosphorane intermediate may also be expected to be relatively
small and comparable with triesters and diesters. As far
as isomerization is concerned, the leaving group remains un-
changed (3Ј-OH), and the nondeparting alkoxy groups have
been shown to exert only a modest effect on the hydrolysis of
phosphotriesters.29 With the cleavage reaction, the increasing
electronegativity of the alkyl group weakens the P᎐OR bond,
but simultaneously retards the protonation of the leaving
oxygen. These two effects at least partially cancel each other.
The increasing electronegativity of the alkyl group(s), however,
slightly biases the product distribution toward the isomeriz-
ation products, the effect being more marked with triesters than
with diesters. If only one of alkyl groups of a 3Ј-triester is
replaced with a better leaving group, the rate-acceleration
appears equal to that observed with the corresponding diesters.
As seen from Table 2, the isopropyl 2-methoxyethyl 3Ј-
phosphotriester 4e is cleaved 3.8 times as fast as its 3Ј-
diisopropyl counterpart 4c, while the reactivity ratio of the
–O
O
O
O–
OR1
O
P
OR2
O
P
OR1
OR2
e
e
e
a
a
O
–O
O
O
O
P
P
OR2
–O
OR1
a/e
e
a/e
OR1
OR2
a/e
a/e
O
O
P
O
OR2
Scheme 4
pseudorotation brings the 3Ј-oxygen to an apical position and
the 3Ј-oxyanion departs, or the alkoxy group leaves as alkoxide
ion. The former reaction is several orders of magnitude faster
than the latter, consistent with the results of quantum chemical
calculations30 which suggest that the endocyclic cleavage of a
monoanionic phosphorane intermediate is a more facile process
than the exocyclic cleavage.
J. Chem. Soc., Perkin Trans. 2, 1998
667

View Article Online
Fig. 6 Brønsted plots for the general base catalyzed cleavage of 5Ј-O-
Fig. 5 Logarithmic first-order rate constants for the pH-independent
cleavage of symmetric 5Ј-O-pivaloyluridin-3Ј-yl and -2Ј-yl dialkyl
phosphates (4a–d/5a–d) to a mixture of diesters at 298.2 K plotted
against the pKa value of the esterified alcohol.31 The value for 4d/5d is
taken from ref. 23. The slope of the line is: cleavage Ϫ0.94 0.13.
pivaloyluridin-3Ј-yl and -2Ј-yl dialkyl phosphates to a mixture of
diesters at 298 K. The β values are: 0.71 0.08 (4a/5a, ᭢), 0.72 0.05
(4b/5b, ᭿), 0.73 0.11 (4e/5e, ᭜). The line without experimental points
refer to 4d/5d and is taken from ref. 23. The buffer acids employed are in
the order of increasing pKa: cyanoacetic, chloroacetic, formic, acetic
and propionic acid.
Since the isomerization is much faster than the cleavage, the
breakdown of the intermediate is undoubtedly the rate-limiting
step of the cleavage reaction. The highly negative βlg value
(Ϫ1.26) is consistent with this mechanism, and it may largely be
attributed to a unimolecular cleavage of the P᎐OR bond via a
transition state where the alkoxide ion character of the leaving
group is rather well developed. In fact, the value is almost equal
to that (Ϫ1.28) reported26 for the hydroxide ion catalyzed cleav-
age of ribonucleoside 3Ј-phosphodiester, which in all likelihood
takes place via a dianionic phosphorane-like transition state or
marginally stable intermediate.1,3 One should, however, bear in
mind that with symmetric triesters the nondeparting alkyl
group may exert an additional effect on both the preequilibrium
concentration of the phosphorane intermediate and the rate of
its breakdown, and these contributions are included in the
observed βlg value. When only one of the isopropyl ligands of
the symmetrical diisopropyl triester 4c is replaced with a 2-
methoxyethyl group 4e, the rate acceleration is only 110-fold,
corresponding a βlg value of the order of Ϫ0.9.
The isomerization takes place via the same phosphorane
intermediate as the cleavage, but now the breakdown of the
intermediate alone is not rate-limiting. The process is kinetically
symmetrical, which means that the transition states for the for-
mation and breakdown of the intermediate on the reaction path
from the 3Ј-triester to its 2Ј-counterpart must be at comparable
energy levels. The pseudorotation may, at least in principle, be
rate-limiting. As far as symmetric triesters are concerned, the
pseudorotation simply consists of an interchange of two identi-
cal alkoxy groups, one in the apical and the other in the equa-
torial positions, and it hence hardly contributes to the β value.
The most probable explanation for the negative β value (Ϫ1.10)
is that, as with the hydronium ion catalyzed isomerization,
electronegative alkyl groups facilitate the formation of the
phosphorane intermediate by diminishing the electron density
at phosphorus. The susceptibility to this effect seems, however,
to be larger than with the hydronium ion catalyzed process.
Tentatively one may assume that the attack of negatively
charged 2Ј-oxyanion on neutral phosphate is more sensitive
to the electron density at phosphorus than the attack of the
neutral 2Ј-hydroxy group on an inherently electron deficient
phosphate monocation.
takes place via the same monoanionic phosphorane inter-
mediate as the hydroxide ion catalyzed isomerization and cleav-
age (Scheme 4). As discussed above, the alkoxide ion is a much
worse leaving group than the 2Ј- and 3Ј-oxyanions. Accord-
ingly, cleavage of the exocyclic P᎐OR bond of the monoanionic
phosphorane alone is rate-limiting and requires acid catalysis
under conditions where the 2Ј- and 3Ј-oxygens may depart as
oxyanions. As seen from Fig. 5, the βlg value for this reaction
(Ϫ0.94 0.13) is less negative than that of the hydroxide ion
catalyzed cleavage (Ϫ1.26), but more negative than that of the
hydronium ion catalyzed cleavage (Ϫ0.51). The relative magni-
tude of these values suggest that the bond rupture is rather
advanced in the transition state and protonation of the depart-
ing alkoxide ion takes place concertedly with the bond rupture.
In other words, hydronium ion acts as a general acid.
Buffer catalyzed cleavage
The cleavage of uridine 3Ј-phosphotriesters 4a–e is susceptible
to buffer catalysis in the pH range 2–7, i.e. under conditions
where the buffer-independent cleavage is pH-independent. As
seen from Table 3, the buffer catalysis is almost entirely general
base catalysis, except in the most acidic buffers. In other
words, general acid catalysis is only observed at pH < 3, i.e.
under conditions where the buffer-independent reaction
becomes hydronium ion catalyzed. The rate-limiting step of the
buffer-independent cleavage is, as discussed above, breakdown
of the monoanionic phosphorane intermediate. The most likely
mechanistic interpretation for the observed general base
catalysis is thus sequential specific base–general acid catalysis:
rapid initial deprotonation of the 2Ј-hydroxy function and sub-
sequent attack of the resulting 2Ј-oxyanion on phosphorus give
a monoanionic phosphorane, which undergoes general acid
catalyzed cleavage of one of the exocyclic alkoxy functions. The
mechanism is hence essentially the same as that proposed above
for the buffer-independent cleavage at pH 3–7. The β value for
the observed general base catalysis by caboxylate ions is
approximately 0.7 with all the triesters studies (Fig. 6), which
means that the α value for the general acid catalyzed step of the
sequential specific base–general acid catalysis is 0.3. The βlg
value is in turn about Ϫ1.0. Accordingly, the rupture of the
P᎐OR bond seems to be considerably more advanced in the
transition state than the proton transfer from the buffer acid to
the leaving group.
pH-independent cleavage
The cleavage of 3Ј-phosphotriesters 4a–e is pH-independent
over a wide pH range, 3–7, i.e. under conditions where the
isomerization is hydroxide ion catalyzed and much faster than
the cleavage. We have suggested previously22,23 that this reaction
As mentioned above, the cleavage reaction is also susceptible
to general acid catalysis, but only when the pKa value of the
buffer acid is <4. The α value is approximately 0.8, and the βlg
668
J. Chem. Soc., Perkin Trans. 2, 1998

View Article Online
value Ϫ0.6, i.e. almost equal to that of the hydronium ion
catalyzed cleavage (Ϫ0.51). In these markedly acidic buffers,
neither the formation nor the breakdown of the phosphorane
intermediate alone is rate-limiting, which makes it difficult to
draw firm mechanistic conclusions. The fact that the isomeriza-
tion is also buffer catalyzed in the most acidic buffer and this
reaction exhibits a βlg value comparable to that of cleavage,
suggests that the formation of the phosphorane intermediate is
buffer catalyzed. A tentative mechanistic interpretation is a
sequential specific acid–general base catalysis. The phosphoryl
oxygen is protonated in a rapid initial step, and the buffer base
facilitates the nucleophilic attack of the 2Ј-hydroxy group on
the phosphorus by abstracting the proton in the transition
state. Departure of the alkoxy group may again be catalyzed
by buffer acid, but this general catalysis is now kinetically
invisible since the buffer acid formed in the first step serves as a
catalyst in the second step. The α value of 0.8 suggests a β
value of 0.2 for the general base catalyzed partial reaction. In
other words, the transition state is rather early: neither the
proton transfer nor the PO bond formation (βlg Ϫ0.6) is very
advanced.
(LiChrospher 100 RP-18, 250 × 10 mm, 5 µm, 67% aq meth-
anol, v/v) gave chromatographically and NMR spectro-
scopically homogeneous product in 13% yield. δH(500 MHz;
CDCl3; J values in Hz throughout) 8.43 (1H, s, H3), 7.41 (1H,
d, H6, JH5,H6 8.1), 5.68 (1H, dd, H5), 6.11 (1H, d, H1Ј, JH1Ј,H2Ј
6.54), 4.90 (1H, m, H3Ј), 4.80 (1H, m, thp-2), 4.51 (1H, dd, H4Ј,
JH3Ј,H4Ј 6.41, JH4Ј,H5Ј 3.81), 4.37 (1H, dd, H5Ј, JH5Ј,H5Љ 12.6), 4.26
(1H, dd, H5Љ, JH4Ј,H5Љ 2.59), 4.22 (2H, m, CH2OP), 3.61 (3H, m,
CH2OMe and thp-6a), 3.44 (1H, m, thp-6b), 3.38 (3H, s,
CH3O), 1.4–1.8 (6H, m, thp), 1.25 (9H, s, Piv). δP(CDCl3)
Ϫ1.62 from phosphoric acid. MS/FAB 631 (M ϩ Na), 647
(M ϩ K), 525 (M Ϫ thp ϩ 2H). Elemental analysis: Found: C,
49.3; H, 6.6; N, 4.4. C25H41N2O13P requires: C, 49.3; H, 6.8; N,
4.6%.
5Ј-O-Pivaloyl-2Ј-O-(tetrahydropyran-2-yl)uridin-3Ј-yl diethyl
phosphate 3b
Compound 3b was obtained as described for 3a using ethanol
instead of 2-methoxyethanol. The crude product was purified
by silica gel chromatography, using hexane–acetone (3:2, v/v)
as eluent, yield 42%. δH(400 MHz, CDCl3) 7.91 (1H, m, H3),
5.73 (1H, dd, H5, JH5,H6 8.15), 7.40 (1H, d, H6), 6.08 (1H, d,
H1Ј, JH1Ј,H2Ј 6.12), 4.88 (1H, m, H3Ј), 4.78 (1H, m, thp-2), 4.43
(2H, m, H2Ј and H4Ј), 4.34 (1H, dd, H5Ј, JH5Ј,H5Љ 12.68, JH5Ј,H4Ј
3.92), 4.28 (1H, dd, H5Љ, JH5Љ,H4Ј 3.16), 4.16 (2H, m, Et-CH2),
3.63 (1H, m, thp-6a), 3.44 (1H, m, thp-6b), 1.4–1.8 (6H, m,
thp), 1.36 (3H, t, Et-CH3, JCH2,CH3 7.08), 1.24 (9H, s, Piv).
δP(CDCl3) Ϫ0.97 from phosphoric acid. MS/FAB (M ϩ Na),
587 (M ϩ K), 465 (100%, M Ϫ thp ϩ 2H). Elemental analysis:
Found: C, 50.7; H, 6.5; N, 5.1. C23H37N2O11P requires: C, 50.4;
H, 6.8; N, 5.1%.
Many of the β1g values discussed above are markedly
negative. Apart from partial bond breaking, this may be gener-
ally indicative of increased electron density on the bridging tri-
ester oxygen on going from the initial state to the phosphorane
intermediate.
Buffer catalyzed isomerization
In striking contrast to the cleavage reaction, the isomerization
is not markedly susceptible to general base catalysis (inter-
preted above as sequential specific base–general acid catalysis)
at pH > 3. Our previous results suggested a modest general
base catalysis also for the isomerization.23 However, the more
extensive data of the present study are in this respect more
controversial. The observed rate constants of isomerization are
in some cases buffer-dependent, but not always. Even in cases
where an apparent buffer catalysis is observed, the buffer-
dependent proportion of the observed rate constant is always
less than 50%. A possible explanation for these ambiguities
may be that with isomerization neither the formation nor the
breakdown of the phosphorane intermediate alone is rate-
limiting. In principle, the attack of the 2Ј-hydroxy function
may be general base catalyzed, and hence the departure of the
3Ј-oxygen general acid catalyzed (but kinetically invisible).
However, if this kind of a reaction pathway exists, it is always
of minor importance compared to the hydroxide ion catalyzed
reaction.
5Ј-O-Pivaloyl-2Ј-O-(tetrahydropyran-2-yl)uridin-3Ј-yl diiso-
propyl phosphate 3c
Compound 3c was obtained as described for 3a, using propan-
2-ol instead of 2-methoxyethanol. In the adsorption chrom-
atography a mixture of dichloromethane and ethanol (19:1,
v/v), and in RP-chromatography, a mixture of methanol and
water (68:32) was used as the eluent. The yield was 6%. δH(400
MHz, CDCl3) 7.89 (1H, s, H3), 7.39 (1H, d, H6, JH5,H6 8.10),
5.72 (1H, dd, H5), 6.07 (1H, d, H1Ј, JH1Ј,H2Ј 6.3), 4.86 (1H, m,
H3Ј), 4.78 (1H, m, thp-2), 4.69 (1H, m, Pri-CH, JCH,CH3 4.8),
4.63 (1H, m, Pri-CH, JCH,CH3 5.2), 4.46 (1H, dd, H4Ј), 4.42 (1H,
m, H2Ј), 4.34 (1H, dd, H5Ј, JH4Ј,H5Ј 3.20, JH5Ј,H5Љ 9.60), 4.29 (1H,
dd, H5Љ, JH4Ј,H5Љ 2.4), 3.63 (1H, m, thp-6a), 3.44 (1H, m, thp-6b),
1.4–1.8 (6H, m, thp), 1.35 (6H, m, Pri-CH3), 1.24 (9H, s, Piv).
δP(CDCl3) Ϫ2.93 from phosphoric acid. MS/FAB 599
(M ϩ Na), 615 (M ϩ K), 493 (100%, M Ϫ thp ϩ 2H). Elem-
ental analysis: Found: C, 51.3; H, 6.6; N, 4.8. C25H41N2O11P
requires: C, 52.1; H, 7.2; N, 4.9%.
Experimental
5Ј-O-Pivaloyl-2Ј-O-(tetrahydropyran-2-yl)uridin-3Ј-yl bis-
(2-methoxyethyl) phosphate 3a
5Ј-O-Pivaloyl-2Ј-O-(tetrahydropyran-2-yl)uridin-3Ј-yl isopropyl
methoxyethyl phosphate 3e
1,2,4-Triazole (265 mg, 3.82 mmol) and triethylamine (532
mm3, 3.83 mmol) were dissolved in dry acetonitrile (6.5 cm3).
Freshly distilled phosphoryl trichloride (117 mm3, 1.29 mmol)
was added, and the mixture was stirred for 30 min at room
temp. The precipitated triethylammonium chloride was
removed by filtration and the filtrate was immediately added
to predried 5Ј-O-pivaloyl-2Ј-O-(tetrahydropyran-2-yl)uridine23
(more polar diastereomer, 350 mg, 0.85 mmol). After 1 h, excess
of 2-methoxyethanol (2.5 cm3, 43 mmol) was added, and the
reaction mixture was left to stand 3.5 h at room temp. The
reaction mixture was evaporated to dryness, the residue was
dissolved in dichloromethane (30 cm3) and washed with an
aqueous phosphate buffer (30 cm3, KH2PO4–Na2HPO4 0.1/0.1
mol dmϪ3). The organic phase was dried with sodium sulfate
and evaporated to dryness. The crude product was first purified
by adsorption chromatography, using Silica gel 60 as an
adsorbent and a mixture of dichloromethane and ethanol (9:1,
v/v) as an eluent. Further purification by RP-chromatography
5Ј-O-Pivaloyl-2Ј-O-(tetrahydropyran-2-yl)uridine23 was con-
verted to its 2-chlorophenoxy isopropyl 3Ј-phosphate as
described for 3a, using 2-chlorophenyl phosphoryl-bis(1,2,4-
triazolide) (obtained from 2-chlorophenyl phosphorodichlori-
date and 1,2,4-triazole) as a phosphorylating agent, and
propan-2-ol in the subsequent alcoholysis. The reaction time
was 4 h in the phosphorylation and 40 h in the alcoholysis. The
product was purified by silica gel chromatography, using a mix-
ture of dichloromethane and ethanol (19:1, v/v) as eluent. A
5:2 mixture of the two phosphorus diastereomers was obtained
in 51% yield. δH(400 MHz, CDCl3, mixture of RP and SP
isomers) 7.48 (1H, m, Ar-3), 7.41 (1H, m, Ar-5), 7.40 (1H, d,
H6), 7.25 (1H, m, Ar-6), 7.14 (1H, m, Ar-4), 6.1 (1H, t, H1Ј),
5.73 (1H, dd, H5, JH5,H6 8.05), 4.99 and 5.05 (1H, m, H3Ј), 4.85
and 4.92 (1H, m, Pri-CH), 4.70 and 4.78 (1H, thp-2), 4.19 and
4.55 (1H, m, H4Ј), 4.45 (1H, m, H2Ј), 4.28–4.40 (2H, m, H5Ј
and H5Љ), 3.60 (1H, m, thp-6a), 3.42 (1H, m, thp-6b), 1.3–1.8
J. Chem. Soc., Perkin Trans. 2, 1998
669

View Article Online
Table 4 Retention times and chromatographic conditions for the
separation of 3–c,e and its reaction products on Hypersil RP-18
columnsa
References
1 D. M. Perreault and E. V. Anslyn, Angew. Chem., Int. Ed. Engl.,
1997, 36, 432.
2 R. Breslow, S. D. Dong, Y. Webb and R. Xu, J. Am. Chem. Soc.,
1996, 118, 6588.
3 M. Oivanen, S. Kuusela and H. Lönnberg, Chem. Rev., 1998, in
press.
4 S. Kuusela and H. Lönnberg, J. Chem. Soc., Perkin Trans. 2, 1994,
2109.
5 M. A. Morgan, S. A. Kazakov and S. Hecht, Nucleic Acids Res.,
1995, 23, 2949.
6 M. A. Podyminogin, V. V. Vlassov, R. Giege, Nucleic Acids Res.,
1993, 21, 5950.
7 J. Smith, K. Ariga and E. V. Anslyn, J. Am. Chem. Soc., 1993, 115,
362.
8 K. Shinozuka, K. Shimizu, Y. Nakashima and H. Sawai, Bioorg.
Med. Chem. Lett., 1994, 4, 1979.
9 V. V. Vlassov, G. Zuber, B. Felden, J.-P. Behr and R. Giege, Nucleic
Acids Res., 1995, 23, 3161.
Acetonitrile
(%)
Compound
tR(3)/min
tR(4)/min
tR(5)/min
a
b
c
e
30
35
40
35
21.6
19.6
20.3
21.0
6.8
6.3
6.8
5.7
5.8
6.2
7.1
6.8b
7.1b
a Eluent: formic acid–sodium formate buffer (each 0.05 mol dmϪ3),
containing 0.1 mol dmϪ3 ammonium chloride. The acetonitrile content
of the eluent is indicated in the table. Flow rate 1 cm3 minϪ1
.
The absorptions were measured at λ = 260 nm. b The RP and SP
diastereomers.
(m, thp), 1.35–1.45 (m, Pri-CH3), 1.21 and 1.23 (s, Piv).
δP(CDCl3) ϩ17.2 and Ϫ7.7 from phosphoric acid. MS/FAB 667
(M[35Cl] ϩ Na), 669 (M[37Cl] ϩ Na, 561 (M[35Cl] Ϫ thp ϩ 2H),
563 (M[37Cl] Ϫ thp ϩ 2H). Elemental analysis: Found: C, 51.7;
H, 5.8; N, 4.4. C28H38ClN2O11P requires: C, 52.1; H, 5.9; N,
4.3%.
10 V. Jubian, A. Veronese, R. P. Dixon and A. D. Hamilton, Angew.
Chem., Int. Ed. Engl., 1995, 34, 1237.
11 A. Lorente, J. F. Espinosa, M. Fernandez-Saiz, J.-M. Lehn, W. D.
Wilson and Y. Y. Zhong, Tetrahedron Lett., 1996, 37, 4417.
12 M. Endo, K. Hirata, T. Ihara, S. Sueda, M. Takagi and
M. Komiyama, J. Am. Chem. Soc., 1996, 118, 5478.
13 K. Kurz and M. W. Göbel, Helv. Chim. Acta, 1996, 79, 1967.
14 M. A. Reynolds, T. A. Beck, P. B. Say, D. A. Schwartz, B. P. Dwyer,
W. J. Daily, M. M. Vaghefi, M. D. Metzler, R. E. Klem and
L. J. Arnold Jr., Nucleic Acids Res., 1996, 24, 760.
15 M. Endo, Y. Azuma, Y. Saga, A. Kuzuya, G. Kawai and
M. Komiyama, J. Org. Chem., 1997, 62, 846.
16 M. Komiyama and K. Yoshinari, J. Org. Chem., 1997, 62, 2155.
17 R. Breslow and C. Schmuck, J. Am. Chem. Soc., 1996, 118, 6601.
18 D. A. Usher, D. I. Richardson Jr. and D. G. Oakenfull, J. Am. Chem.
Soc., 1970, 92, 4699.
19 K. N. Dalby, A. J. Kirby and F. Hollfelder, J. Chem. Soc., Perkin
Trans. 2, 1993, 1269.
20 C. L. Perrin, J. Org. Chem., 1995, 60, 1239.
21 A. Nunez and O. Nunez, J. Org. Chem., 1996, 61, 8386.
22 M. Kosonen, M. Oivanen and H. Lönnberg, J. Org. Chem., 1994, 59,
3704.
23 M. Kosonen and H. Lönnberg, J. Chem. Soc., Perkin Trans. 2, 1995,
1203.
24 J. Florian and A. Warshel, J. Am. Chem. Soc., 1997, 119, 5473.
25 A. Kraszewski and J. Stawinski, Tetrahedron Lett., 1980, 21, 2935.
26 M. Kosonen, E. Yousefi-Salakdeh, R. Strömberg and H. Lönnberg,
J. Chem. Soc., Perkin Trans. 2, 1997, 2661.
27 M. Oivanen, R. Schell, W. Pfleiderer and H. Lönnberg, J. Org.
Chem., 1991, 56, 3623.
28 P. Järvinen, M. Oivanen and H. Lönnberg, J. Org. Chem., 1991, 56,
5396.
The 2-chlorophenyl, isopropyl 3Ј-phosphate obtained (153
mg, 0.237 mmol) was dissolved in 1,4-dioxane (23 cm3), and a
solution of sodium 2-methoxyethoxide in 2-methoxyethanol
(1.05 mmol in 7 cm3) was added. The mixture was stirred for 15
min at room temp., neutralized with acetic acid and evaporated
to dryness. The residue was dissolved in dichloromethane (28
cm3), washed with aqueous phosphate buffer (40 cm3) and dried
with sodium sulfate. The product was first purified by silica gel
chromatography (dichloromethane–ethanol, 9:1, v/v) and then
by RP-chromatography (LiChrospher 100 RP-18, 250 × 10
mm, 5 µm, 65% aq methanol). A mixture of the RP and SP
diastereomers of 3e was obtained in 29% yield. δH(400 MHz,
CDCl3) 8.08 (1H, s, H3), 7.41 (1H, d, H6, JH6,H5 8.29), 5.73 (1H,
dd, H5), 6.09 (1H, t, H1Ј), 4.88 (1H, m, H3Ј), 4.80 (1H, m,
thp-2), 4.72 and 4.67 (1H, m, Pri-CH, JCH,CH3 6.58), 4.48 and
4.50 (1H, 2dd, H4Ј), 4.42 (1H, m, H2Ј), 4.36 and 4.35 (1H, 2dd,
H5Ј, JH4Ј,H5Ј 3.67, JH5Ј,H5Љ 12.44), 4.28 and 4.27 (1H, 2dd, H5Љ,
JH4Ј,H5Љ 2.93), 4.16 (2H, m, CH2O-P), 3.60 (3H, m, thp-6a and
CH2OMe), 3.49 (1H, m, thp-6b), 3.38 (3H, s, OCH3), 1.4–1.8
(6H, m, thp), 1.37 and 1.36 (6H, 2d, Pri-CH3), 1.24 (9H, s,
Piv). δP(CDCl3) Ϫ1.81 from phosphoric acid. MS/FAB 615
(M ϩ Na), 509 (100%, M Ϫ thp ϩ 2H). Elemental analysis:
Found: C, 50.9; H, 6.9; N, 5.1. C25H41N2O12P requires: C, 50.7;
H, 7.0; N, 4.7%.
29 R. H. Bromilow, S. A. Khan and A. J. Kirby, J. Chem. Soc., Perkin
Trans. 2, 1972, 911.
30 M. Uebayasi, T. Uchimaru, T. Koguma, S. Sawata, T. Shimayama
and K. Taira, J. Org. Chem., 1994, 59, 7414.
31 E. P. Serjeant and B. Dempsey, Ionization Constants of Organic
Acids in Aqueous Solution, IUPAC Chemical Data Series 23,
Pergamon Press, Oxford, 1979.
Kinetic measurements
The kinetic measurements were carried out and the aliquots
analyzed by RP HPLC (Hypersil ODS, 250 × 4 mm, 5 µm)
as described previously23 for the dimethyl ester, 4c. The reten-
tion times and chromatographic conditions are indicated in
Table 4. The rate constants were calculated as described
previously.23
Paper 7/07095K
Received 1st October 1997
Accepted 3rd November 1997
670
J. Chem. Soc., Perkin Trans. 2, 1998