Metal-Ion-Promoted Cleavage, Isomerization, and Desulfurization of the Diastereomeric Phosphoromonothioate Analogues of Uridylyl(3′,5′)uridine

Article abstract of DOI:10.1021/jo972112n

Metal-ion-promoted hydrolytic reactions of the S_P and R_P diastereomers of the phosphoromonothioate analogues of uridylyl(3′,5′)uridine [3′,5′-Up(s)U] and their cleavage products, diastereomeric uridine 2′,3′-cyclic phosphates [2′,3′-cUMPS], were followed by HPLC as a function of pH (4.7-5.6) and metal ion concentration (1-10 mmol L^-1). With 3′,5′-Up(s)U, three reactions compete: (i) cleavage to 2′,3′-cUMPS, (ii) isomerization to 2′,5′-Up(s)U, and (iii) desulfurization to an equilibrium mixture of 2′,5′- and 3′,5′-UpU. Of these, the cleavage to 2′,3′-cUMPS is markedly accelerated by Zn²⁺, Cd²⁺, and Gd³⁺, the rate enhancements observed with the Sp isomer at [M^z+] = 5 mmol L^-1 and pH 5.6 (T = 363.2 K) being 410-, 3600-, and 2000-fold, respectively. The effect of Mn²⁺ and Mg²⁺ on the cleavage rate is, in turn, modest (6- and 1.7-fold acceleration, respectively). The rate-accelerations are almost equal with the S_P and R_P diastereomers. The metal-ion-promoted reaction is first-order in both the hydroxide and metal ion concentration, and it proceeds by inversion of configuration at phosphorus, consistent with an in-line displacement mechanism. The isomerization and desulfurization are much less susceptible to metal ion catalysis: 6.4- and 7.7-fold accelerations were observed with Zn²⁺, respectively. Gd³⁺ does not promote these reactions at all. The isomerization proceeds by retention of configuration at phosphorus, consistent with formation of a pentacoordinated thiophosphorane intermediate having the entering 2′-hydroxy group apical and the leaving 3′-hydroxy equatorial, and subsequent pseudorotation posing the leaving group apical. The hydrolysis of 2′,3′-cUMPS is accelerated by metal ions slightly more efficiently than the cleavage of 3′,5′-Up(s)U to 2′,3′-cUMPS. In striking contrast to the reactions of 3′,5′-Up(s)U, the hydrolytic desulfurization to 2′,3′-cUMP is accelerated as efficiently as its endocyclic hydrolysis to uridine 2′and 3′-phosphoromonothioates [2′- and 3′-UMPS]. Somewhat unexpectedly, the latter compounds were observed to undergo metal-ion-promoted cyclization/desulfurization to 2′,3′-cUMP. The hydrolysis of 2′- or 3′-UMPS to uridine was, in turn, observed to be retarded by metal ions. The mechanisms of the partial reactions are discussed.

Full text of DOI:10.1021/jo972112n

J . Org. Chem. 1998, 63, 2939-2947
2939
Meta l-Ion -P r om oted Clea va ge, Isom er iza tion , a n d Desu lfu r iza tion
of th e Dia ster eom er ic P h osp h or om on oth ioa te An a logu es of
Ur id ylyl(3′,5′)u r id in e
Mikko Ora, Markku Peltoma¨ki, Mikko Oivanen,* and Harri Lo¨nnberg
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
Received November 18, 1997
Metal-ion-promoted hydrolytic reactions of the SP and RP diastereomers of the phosphoromonothioate
analogues of uridylyl(3′,5′)uridine [3′,5′-Up(s)U] and their cleavage products, diastereomeric uridine
2′,3′-cyclic phosphates [2′,3′-cUMPS], were followed by HPLC as a function of pH (4.7-5.6) and
metal ion concentration (1-10 mmol L-1). With 3′,5′-Up(s)U, three reactions compete: (i) cleavage
to 2′,3′-cUMPS, (ii) isomerization to 2′,5′-Up(s)U, and (iii) desulfurization to an equilibrium mixture
of 2′,5′- and 3′,5′-UpU. Of these, the cleavage to 2′,3′-cUMPS is markedly accelerated by Zn2+
,
Cd2+, and Gd3+, the rate enhancements observed with the SP isomer at [Mz+] ) 5 mmol L-1 and pH
5.6 (T ) 363.2 K) being 410-, 3600-, and 2000-fold, respectively. The effect of Mn2+ and Mg2+ on
the cleavage rate is, in turn, modest (6- and 1.7-fold acceleration, respectively). The rate-
accelerations are almost equal with the SP and RP diastereomers. The metal-ion-promoted reaction
is first-order in both the hydroxide and metal ion concentration, and it proceeds by inversion of
configuration at phosphorus, consistent with an in-line displacement mechanism. The isomerization
and desulfurization are much less susceptible to metal ion catalysis: 6.4- and 7.7-fold accelerations
were observed with Zn2+, respectively. Gd3+ does not promote these reactions at all. The
isomerization proceeds by retention of configuration at phosphorus, consistent with formation of a
pentacoordinated thiophosphorane intermediate having the entering 2′-hydroxy group apical and
the leaving 3′-hydroxy equatorial, and subsequent pseudorotation posing the leaving group apical.
The hydrolysis of 2′,3′-cUMPS is accelerated by metal ions slightly more efficiently than the cleavage
of 3′,5′-Up(s)U to 2′,3′-cUMPS. In striking contrast to the reactions of 3′,5′-Up(s)U, the hydrolytic
desulfurization to 2′,3′-cUMP is accelerated as efficiently as its endocyclic hydrolysis to uridine 2′-
and 3′-phosphoromonothioates [2′- and 3′-UMPS]. Somewhat unexpectedly, the latter compounds
were observed to undergo metal-ion-promoted cyclization/desulfurization to 2′,3′-cUMP. The
hydrolysis of 2′- or 3′-UMPS to uridine was, in turn, observed to be retarded by metal ions. The
mechanisms of the partial reactions are discussed.
In tr od u ction
reactions.11-13 A phosphorothioate linkage, having either
an RP or SP configuration, has been inserted in a selected
position of the sugar-phosphate backbone of either the
substrate chain or the ribozyme itself, and the resulting
kinetic thio effect has been utilized to pinpoint the
binding site of the metal ion7-9,14-17 and to elucidate the
stereochemistry of the ribozyme-catalyzed cleavage.15,16
Phosphorothiolate linkages, having either the 5′-bridg-
ing18,19 or 3′-bridging20,21 oxygen replaced with sulfur,
have been more recently used for related purposes.
Ribonucleoside phosphoromonothioates, containing a
chiral S-bonded phosphorus atom, were introduced in
RNA chemistry in late 1960s, and since then they have
been extensively used as mechanistic probes in elucida-
tion of the action of both protein enzymes1,2 and
ribozymes.3-10 One of the central subjects has been the
role of divalent metal ions in ribozyme-catalyzed
* To whom correspondence should be addressed. Tel.: +358 2
3336787. Fax: +358 2 3336700. E-mail: mikko.oivanen@utu.fi.
(1) (a) Eckstein, F. Annu. Rev. Biochem. 1985, 54, 367. (b) Frey, P.
Adv. Enzymol. Relat. Areas Mol. Biochem. 1989, 62, 119.
(2) Lee, J . W.; Cox, M. M. Biochemistry 1990, 29, 7666. (b) Dahnke,
T.; J iang, R. T.; Tsai, M. D. J . Am. Chem. Soc. 1991, 113, 9388. (c)
Vollmer, S. H.; Walner, M. B.; Tarbell, K. V.; Colman, R. F. J . Biol.
Chem. 1994, 269, 8082. (d) Koteiche, H. A.; Narasimhan, C.; Runquist,
J . A.; Miziorko, H. M. Biochemistry 1995, 34, 15068. (e) Hollfelder,
F.; Herschlag, D. Biochemistry 1995, 34, 12255. (f) Herschlag, D. J .
Am. Chem. Soc. 1994, 116, 11631.
(3) Koizumi, M.; Ewai, S.; Ohtsuka, E. FEBS Lett. 1988, 239, 285.
(4) Slim, G.; Gait, M. J . Nucleic Acids Res. 1991, 19, 1183.
(5) Maschhoff, K. L.; Padgett, R. A. Nucleic Acids Res. 1993, 23,
5456.
(6) Padgett, R. A.; Podar, M.; Boulanger, S. C.; Perlman, P. S. Science
1994, 266, 1685.
(11) Pyle, A. M. Science 1993, 261, 709.
(12) Piccirilli, J . A.; Vyle, J . S.; Caruthers, M. H.; Cech, T. R. Nature
1993, 361, 85.
(13) Pyle, A. M. In Metal Ions in Biological Systems; Sigel, A., Sigel,
H., Eds.; Marcel Dekker: New York, 1996; Vol. 32, Chapter 14.
(14) Dahm, S. C.; Uhlenbeck, O. C. Biochemistry 1991, 30, 9464.
(15) Koizumi, M.; Ohtsuka, E. Biochemistry 1991, 30, 5145.
(16) van Tol, H.; Buzayan, J . M.; Feldstein, P. A.; Eckstein, F.;
Bruening, G. Nucleic Acids Res. 1990, 18, 1971.
(17) Dahm, S. C.; Derrick, W. B.; Uhlenbeck, O. C. Biochemistry
1993, 32, 13040.
(18) (a) Kuimelis, R. G.; McLaughlin, L. W. Nucleic Acids Res. 1995,
23, 4753. (b) Kuimelis, R. G.; McLaughlin, L. W. Bioorg. Med. Chem.
1997, 5, 1051.
(19) Thomson, J . B.; Patel, B. K.; J imenez, V.; Eckart, K.; Eckstein,
F. J . Org. Chem. 1996, 61, 6273.
(7) Chowrira, B. M.; Burke, J . M. Biochemistry 1991, 30, 8518.
(8) Christian, E. C.; Yarus, M. Biochemistry 1993, 32, 4475.
(9) Zhou, D. M.; Kumar, P. K. R.; Zhang, L. H.; Taira, K. J . Am.
Chem. Soc. 1996, 118, 8969.
(10) Herschlag, D.; Piccirilli, J . A.; Cech, T. R. Biochemistry 1991,
30, 4844.
(20) Weinstein, L. B.; Earnshaw, D. J .; Cosstick, R.; Cech, T. R. J .
Am. Chem. Soc. 1996, 118, 10341.
(21) Sontheimer, E. J .; Sun, S.; Piccirilli, J . A. Nature 1997, 388,
801. (b) Weinstein, L. B.; J ones, B. C. N. M.; Cosstick, R.; Cech, T. R.
Nature 1997, 388, 805.
S0022-3263(97)02112-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/14/1998

2940 J . Org. Chem., Vol. 63, No. 9, 1998
Ora et al.
Sch em e 1
This kind of use of phosphorothioates as mechanistic
probes is largely based on detailed knowledge of the
effects of thiosubstitution on the underlying chemical
reactivity of phosphoesters. The metal-ion-dependent
reactivity of diastereomeric phosphorothioates is, for
example, of considerable interest since metal ions are
known to exhibit markedly different binding properties
to oxygen and sulfur.22,23 For this purpose, we now report
on the catalytic effects of some metal ions on the
nonenzymatic cleavage, isomerization, and desulfuriza-
tion of diastereomeric dinucleoside 3′,5′-phosphoromono-
thioates (1a ,b) and the corresponding reactions of dia-
stereomeric nucleoside 2′,3′-cyclic phosphoromonothioates
(3a ,b). To the best of our knowledge, no systematic
kinetic study on these subjects has been published. The
results are compared to those obtained earlier24,25 with
the same compounds in the absence of metal ions.
Moreover, the catalytic effects of the metal ions are
compared to those reported26,27 for the cleavage of their
oxygen counterparts, uridylyl(3′,5′)uridine (3′,5′-UpU; 4)
and uridine 2′,3′-cyclic phosphate (2′,3′-cUMP; 6), and
correlated with their affinities23 to a nucleoside 5′-
thiophosphate.
Resu lts
P r od u ct Distr ibu tion s. The decomposition of the SP
(1a ) and RP (1b) diastereomers of uridylyl(3′,5′)uridine
phosphoromonothioate [3′,5′-Up(s)U] was followed in the
presence of Mg2+, Mn2+, Zn2+, Cd2+, and Gd3+ (1-10 mmol
L-1) at pH 4.7-5.6 (T ) 363.2 K). The composition of
the aliquots withdrawn at appropriate intervals from the
reaction solution was determined by RP HPLC. The
signals were assigned by spiking with authentic samples.
The same three reactions observed previously24 in the
absence of metal ions took place, viz. (i) cleavage to the
SP or RP diastereomer of uridine 2′,3′-cyclic phospho-
romonothioate [2′,3′-cUMPS; 3a ,b; route A in Scheme 1],
(ii) desulfurization to a mixture of 3′,5′- and 2′,5′-UpU
(4, 5; route B), and (iii) isomerization to either the SP or
RP diastereomer of 2′,5′-Up(s)U (2a ,b; route C). 2′,3′-
cUMPS (3a ,b) underwent, also consistent with previous25
observations in the absence of metal ions, concurrent
hydrolysis to a mixture of uridine 2′- and 3′-phospho-
romonothioates [2′- and 3′-UMPS; 7, 8; route D in Scheme
2] and desulfurization to 2′,3′-cUMP (6; route E). 2′,3′-
cUMP was also formed by cyclization/desulfurization of
2′- and 3′-UMPS (route F). This reaction has not been
observed to take place in the absence of metal ions at
pH > 2.28 The reaction sequence was completed by
hydrolysis of 2′,3′-cUMP to a mixture of uridine 2′- and
3′-phosphates (2′- and 3′-UMP; 9, 10) and their dephos-
phorylation to uridine. 2′- and 3′-UMPS were addition-
ally hydrolyzed to uridine without intermediary accu-
mulation of 2′/3′-UMP (route G), but this reaction was
retarded rather than accelerated by metal ions.
Sch em e 2
(22) Pecoraro, V. L.: Hermes, J . D.; Cleland, W. W. Biochemistry
1984, 23, 5262.
(23) Sigel, R. K. O.; Song, B.; Sigel, H. J . Am. Chem. Soc. 1997, 119,
744.
(24) Oivanen, M.; Ora, M.; Almer, H.; Stro¨mberg, R.; Lo¨nnberg, H.
J . Org. Chem. 1995, 60, 5620.
(25) Ora, M.; Oivanen, M.; Lo¨nnberg, H. J . Org. Chem. 1996, 61,
3951.
(26) Kuusela, S.; Lo¨nnberg, H. J . Phys. Org. Chem. 1992, 5, 803.
(27) Kuusela, S.; Lo¨nnberg, H. J . Phys. Org. Chem. 1993, 6, 347.
(28) Ora, M.; Oivanen, M.; Lo¨nnberg, H. J . Chem. Soc., Perkin
Trans. 2 1996, 771.
Ster eoch em istr y a n d Kin etics of th e Meta l-Ion -
P r om oted Rea ction s of 3′,5′-Up (s)U (1a ,b). Table 1

Phosphoromonothioate Analogues of Uridylyl(3′,5′)uridine
J . Org. Chem., Vol. 63, No. 9, 1998 2941
Ta ble 1. F ir st-Or d er Ra te Con sta n ts for th e Clea va ge
(k1), Desu lfu r iza tion (k2), a n d Isom er iza tion (k3) of (SP )-
3′,5′-Up (s)U (1a ), (RP )-3′,5′-Up (s)U (1b), a n d 3′,5′-Up U (4) a t
363.2 Ka
Sch em e 3
[Mz+]/
k1/
k2/
k3/
compd Mz+ mmol L-1 pH
10-6 s-1
10-6 s-1 10-6 s-1
1a
Zn2+
1.0
2.0
5.0
10.0
20.0b
40.0c
100.0d
5.0
5.0
5.0
1.0
2.0
5.0
10.0
5.0
5.0
1.0
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.38
5.08
4.68
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.2
15
38
1.6
2.5
6.3
0.22
0.32
0.77
2.0
4.6
7.4
47
20
220
260
340
23
9.2
3.6
63
170
38
3.5
2.8
5.7
0.70
0.57
0.42
0.12
0.11
0.12
0.12
0.11
1.4
Mn2+
0.24
0.31
0.55
1.1
0.16
190
150
340
930
0.093
11
190
0.19
0.79
0.071
89
110
870
0.12
7.22
0.650
0.104
0.93
0.90
1.6
1.2
0.72
Mg2+
Gd3+
Cd2+
5.0
10.0
none
Zn2+
0.82
2.7
17
2.5
3.2
2.3
0.12
0.31
2.0
0.18
0.21
0.19
1b
1.0
10.0
1.0
10.0
5.0
5.0
1.0
10.0
Mn2+
Mg2+
Gd3+
Cd2+
none
Zn2+
Mn2+
Mg2+
Gd3+
Cd2+
none
2.2
0.20
4e
5.0
5.0
5.0
5.0
5.0
0.810
0.760
0.650
5.60 >1000
5.60
5.60
2.00
0.052
0.650
a
The pH was adjusted with HEPES or MES buffer and the ionic
strength (0.1 mol L-1) with sodium nitrate. The ionic strength
b
F igu r e 1. Effect of metal ion concentration on the first-order
rate constants of cleavage (k1), desulfurization (k2), and
isomerization (k3) of (SP)-3′,5′-Up(s)U (1a ) at 363.2 K. The pH
was adjusted to 5.60 with HEPES buffer and the ionic strength
to 0.1 mol L-1 with sodium nitrate. Notation: Cd2+-catalyzed
cleavage (b); Mn2+-catalyzed cleavage (2), and Zn2+-catalyzed
cleavage (9), desulfurization (0), and isomerization (1).
was 0.16, c 0.24, and 0.3 mol L-1
.
e Data taken from ref 27.
d
records the rate constants obtained for the cleavage
(route A in Scheme 1), desulfurization (B), and isomer-
ization (C) of (SP)- and (RP)-3′,5′-Up(s)U (1a ,b) at various
pH and metal ion concentrations. As seen, the cleavage
to 2′,3′-cUMPS (A) is hardly detectably accelerated by
Mg2+, and the catalytic effect of Mn2+ also is modest: a
6-fold acceleration was observed with the SP diastereomer
Some stereoselectivity, however, occurs; the Zn2+-pro-
moted cleavage of the RP diastereomer is two to three
times as fast as that of the SP form.
(1a ) at [Mn2+] ) 5 mmol L-1 at pH 5.6. In contrast, Zn2+
,
Cd2+, and Gd3+ markedly accelerate the cleavage, the rate
accelerations obtained with 1a being 410-, 3600-, and
2000-fold, respectively ([Mz+] ) 5 mmol L-1, pH 5.6). It
is worth noting that the rate-accelerating effects of the
soft metal ions, Zn2+ and Cd2+, are greater than those
observed with 3′,5′-UpU, whereas Gd3+, a typical hard
Lewis acid, accelerates the cleavage of 3′,5′-Up(s)U
considerably less efficiently than that of 3′,5′-UpU.27
The metal-ion-promoted cleavage of 3′,5′-Up(s)U pro-
ceeds by inversion at phosphorus, as depicted in Scheme
3. The reaction is first-order in metal ion concentration
(Figure 1) and also in hydroxide ion concentration (Figure
2). At high metal ion concentrations, the cleavage rate,
however, levels off to a constant value (Table 1), presum-
ably owing to saturation of the starting material with
the metal ion. With Zn2+ this saturation takes place in
the concentration range 10-50 mmol L-1, suggesting the
log(K/mol L-1) value of the 1:1 complex to be 1.5. The
SP and RP diastereomers react at a comparable rate.
Compared to the cleavage reaction, the desulfurization
of 3′,5′-Up(s)U is much less susceptible to the metal ion
catalysis. Only an 8-fold acceleration was observed in 5
mmol L-1 solution of Zn2+. Mg2+ and Gd3+ exhibited no
catalytic effect and Mn2+ a barely noticeable rate-ac-
celerating effect (Table 1). Cd2+ accelerates the desulfu-
rization very efficiently, but this acceleration appears to
be of exceptional origin. The desulfurization of both 1a
and 1b showed a marked positive deviation from the first-
order kinetics, suggesting the reaction to be subject to
efficient product catalysis. Consistent with this assump-
tion, desulfurization immediately took place when Na2S
(0.1 mmol L-1) was added to the reaction mixture. With
Zn2+, no deviation from the first-order kinetics was
observed during the first two half-lives. Both diastere-
omers are desulfurized approximately as fast. The
reaction is first-order in the concentration of metal ion
(Figure 1) but clearly less than first-order (0.7) in that
of hydroxide ion (Figure 2).

2942 J . Org. Chem., Vol. 63, No. 9, 1998
Ora et al.
Ta ble 2. F ir st-Or d er Ra te Con sta n ts for th e Hyd r olysis
(k5′′) a n d Desu lfu r iza tion (k5′) of (SP )-2′,3′-cUMP S (3a ),
(RP )-2′,3′-cUMP S (3b), a n d 2′,3′-cUMP (6) a t 363.2 Ka
compd Mz+ [Mz+]/mmol L-1 pH k5′′/10-6 s-1 k5′/10-6 s-1
3a
Zn2+
1.0
2.0
5.0
10.0
20.0b
40.0c
100.0d
5.0
5.0
5.0
1.0
2.0
5.0
10.0
5.0
5.0
1.0
2.0
5.0
10.0
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.38
5.08
4.68
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
5.60
110
270
80
240
510
1020
2700
4700
5200
6100
600
260
77
2.0
2.6
5.9
10.4
1.1
1900
830
1600
3900
9040
560
1900
1140
1500
240
100
46
Mn2+
0.93
1.3
3.5
7.3
Mg2+
Gd3+
Cd2+
0.37
80
350
810
1900
2700
0.23
130
270
520
1300
F igu r e 2. Effect of pH on the cleavage (9), desulfurization
(0), and isomerization (1) of (SP)-3′,5′-Up(s)U (1a ) in the
presence of Zn2+ (0.005 mol L-1) at 363.2 K. The pH was
adjusted with HEPES and MES buffers and the ionic strength
(0.1 mol L-1) with sodium nitrate.
none
Zn2+
0.64
3b
1.0
2.0
5.0
10.0
1.0
10.0
5.0
1.0
2.0
5.0
10.0
170
340
1300
2500
e
7.7
f
1100
Sch em e 4
Mn2+
7.7
Mg2+
Cd2+
230
497
2100
2600
1990
5200
10300
none
Zn2+
Mn2+
Mg2+
Cd2+
none
0.32
33.9
16.4
1.88
23.8
1.02
1.3
6g
5.0
5.0
5.0
5.0
a
The pH was adjusted with HEPES or MES buffer and the ionic
strength (0.1 mol L-1) with sodium nitrate. The ionic strength
b
0.16, c0.24, and d0.3 mol L-1
.
e k5 ) 1.9 × 10-6 s-1
.
f k5 ) 1.5 ×
10-6 s-1
.
Data taken from ref 26.
g
Zn2+ moderately promotes the isomerization (route C)
of 3′,5′-Up(s)U, the rate-acceleration at [Zn2+] ) 5 mmol
L-1 at pH 5.6 being 6-fold. In contrast, Mg2+, Mn2+, and
Gd3+ showed no such acceleration. With Cd2+, the
isomerization could not be studied, owing to the rapid
phosphorylation to uridine (G), cyclization/desulfurization
(F) to 2′,3′-cUMP (6), and this reaction was observed to
exhibit a slight positive deviation from first-order kinet-
ics. Accordingly, accumulation of neither of the inter-
mediates 6 or 7/8 obeys any simple rate law, even though
degradation of 3a ,b strictly follows first-order kinetics.
That is why the rate constants for the parallel reactions
D and E could only be estimated by bisecting the first-
order rate constant for disappearance of 2′,3′-cUMPS
(3a ,b) with the aid of the concentration ratio of 2′,3′-
cUMP (6) and 2′/3′-UMPS (7 + 8) at very early stages of
reaction. The results obtained are summarized in Table
2. Both of the parallel reactions (D, E) are strongly
promoted by Zn2+ and Cd2+, the hydrolysis slightly
predominating over the desulfurization. Usually, 60-
80% of the total disappearance of 2′,3′-cUMPS proceeds
via 2′- and 3′-UMPS.
desulfurization of the starting material. The Zn2+
-
promoted reaction is nearly pH-independent (Figure 2)
and first-order in the metal ion concentration (Figure 1).
The effect on the isomerization rate is almost equal with
both diastereomers (1a ,b). For comparison, Zn2+ pro-
motes neither the interconversion of 2′,5′- and 3′,5′-UpU27
nor the isomerization of the phosphodiester bonds of
poly(U).29 In striking contrast to the cleavage reaction,
the isomerization of 1a and 1b proceeds by retention of
configuration at phosphorus: 1a gives only 2a and 1b
gives 2b (Scheme 4).
Kin etics of th e Meta l-Ion -P r om oted Rea ction s of
2′,3′-cUMP S (3a ,b). Metal ions accelerate both the
hydrolysis (route D in Scheme 2) and desulfurization (E)
of 2′,3′-cUMPS (3a ,b). Determination of the rate con-
stants of these reactions was severely complicated by the
fact that the products of the hydrolysis reaction (D), i.e.,
2′- and 3′-UMPS (7, 8), undergo, in addition to dethio-
The rate-accelerating effects of various metal ions on
the hydrolysis of 2′,3′-cUMPS are somewhat larger than
their effects on the cleavage of 3′,5′-Up(s)U (Table 2). The
RP-diastereomer (3b) reacts slightly faster than the SP-
isomer (3a ), the difference in rate being usually less than
20%. Interestingly, the desulfurization of 2′,3′-cUMPS
is accelerated by metal ions much more efficiently than
the corresponding reaction of 3′,5′-Up(s)U.
(29) Kuusela, S.; Lo¨nnberg, H. J . Chem. Soc., Perkin Trans. 2 1994,
2301.

Phosphoromonothioate Analogues of Uridylyl(3′,5′)uridine
J . Org. Chem., Vol. 63, No. 9, 1998 2943
Ta ble 3. F ir st-Or d er Ra te Con sta n ts for th e Cycliza tion
(k6) a n d Deth iofosfor yla tion (k7) of 2′-UMP S (7) a t 363.2
Ka
compd
Mz+
[Mz+]/mmol L-1
k6/10-4 s-1
k7/10-4 s-1
7
Zn2+
1.0
2.0
2.7
2.4
3.4
1.5
5.0
10.0
1.0
5.9
9.5
5.6
0.94
0.15
Cd2+
2.0
5.9
5.0
10.0
1.0
10.0
5.0
11.0
7.1
0.94
0.92
0.28
Mn2+
3.5
1.3
3.0
7.0
Mg2+
none
a
The pH 5.60 was adjusted with HEPES buffer and the ionic
strength (0.1 mol L-1) with sodium nitrate.
tion (F). As mentioned above, the latter reaction does
not strictly obey first-order kinetics but is apparently
subject to product catalysis. Hence, only approximate
values for the rate constants of reactions F and G,
referring to early stages of the disappearance of 7 and 8,
may be reported. These data are collected in Table 3.
F igu r e 3. Effect of metal ion concentration on the first-order
rate constants of phosphoester hydrolysis (solid notation; k5′′)
and desulfurization (open notation; k5′) of (SP)-2′,3′-cUMPS (3a )
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-1
with sodium nitrate. Notation: Cd2+ (b, O), Zn2+ (9, 0) and
Mn2+ (2, 4).
Discu ssion
Mech a n ism of th e Meta l-Ion -P r om oted Rea ction s
of 3′,5′-Up (s)U. In the pH range studied (pH 4.7-5.6),
all the competing reactions of 3′,5′-Up(s)U are nearly pH-
independent in the absence of metal ions, the desulfur-
ization (B) predominating over cleavage (A) and isomer-
ization (C).24 It has been suggested24,30 that the reactions
take place via a common monoanionic thiophosphorane
intermediate obtained by the attack of the deprotonated
2′-hydroxy function on the neutral thiophosphate diester,
i.e., via a minor but more reactive tautomer. All the
metal ions studied accelerate the cleavage reaction (A),
but the catalytic effects differ considerably from those
observed previously for the cleavage of 3′,5′-UpU.27 While
Mg2+, Mn2+, and Gd3+ promote the cleavage of 3′,5′-
Up(s)U less efficiently than that of 3′,5′-UpU, the opposite
is true with Zn2+ and Cd2+
.
Furthermore, Zn2+ (and
possibly Cd2+) also promotes the desulfurization (B) and
isomerization (C) of 3′,5′-Up(s)U (though less efficiently
than the cleavage), in striking contrast to Gd3+. These
differences in the action most likely result from different
binding behavior of the metal ions. The studies of Sigel
et al.23 on adenosine 5′-phosphoromonothioate suggest
that soft Lewis acids, such as Zn2+ and Cd2+, prefer
coordination to the sulfur ligand, while hard Lewis acids,
such as Mn2+ and Mg2+ (and in all likelihood also Gd3+),
are predominantly bound to oxygen. As seen from Figure
5, the catalytic effect of various metal ions on the cleavage
of 3′,5′-Up(s)U roughly correlates with the stability of
their complex with adenosine 5′-phosphoromonothioate.
The mechanism of the action of the catalytically most
active metal ions, Zn2+ and Cd2+, that bind to sulfur
instead of oxygen is of special interest of the present
study. At low metal ion concentrations, the cleavage (A),
desulfurization (B), and isomerization (C) are all first
order in the concentration of metal ion (Figure 1),
indicating that only one metal ion is involved. The
cleavage reaction is also first-order in hydroxide ion
F igu r e 4. Effect of pH on the hydrolysis (9) and desulfur-
ization (0) of (SP)-2′,3′-cUMPS (3a ) in the presence of Zn2+
(0.005 mol L-1) at 363.2 K. The pH was adjusted with HEPES
and MES buffers and the ionic strength (0.1 mol L-1) with
sodium nitrate.
The Mn2+-, Zn2+-, and Cd2+-promoted hydrolysis (D)
and desulfurization (E) of 2′,3′-cUMPS are both first-
order in the metal ion concentration (Figure 3) and, at
least in the case of Zn2+, also first-order in the hydroxide
ion concentration (Figure 4).
It has been shown previously28 that 2′- and 3′-UMPS
undergo under neutral and slightly acidic conditions
hydrolytic dethiophosphorylation to uridine (G). The
divalent metal ions studied clearly retarded this reaction,
though exact kinetic data cannot be obtained due to the
concurrent metal-ion-promoted cyclization/desulfuriza-
(30) Ora, M.; Oivanen, M.; Lo¨nnberg, H. J . Org. Chem. 1997, 62,
3246.

2944 J . Org. Chem., Vol. 63, No. 9, 1998
Ora et al.
that the cleavage reaction represents a borderline case
between the concerted (dianionic transition state) and
stepwise (dianionic intermediate) mechanism.
Evidently, the pseudorotation barrier for I2-Mz+ is
high, since both of the negatively charged nonbridging
ligands (-O- of -S-) tend to avoid an apical position. It
has been shown previously24 that the hydroxide-ion-
catalyzed cleavage of 3′,5′-Up(s)U, proceeding via a
dianionic thiophosphorane, is not accompanied by either
desulfurization or isomerization. These reactions take
place only via a monoanionic thiophosphorane.24 By
analogy, protonation of I2-Mz+ to IH- might be expected
to be required to enable pseudorotation and, hence,
isomerization. This protonation may be kinetically vis-
ible, since the reaction would be first order in metal ion
concentration but pH-independent. As seen from Figure
2, the isomerization rate is only slightly increased with
the increasing pH, and the reaction proceeds by complete
retention of configuration at phosphorus (Scheme 4),
consistent with the assumed pseudorotation.
F igu r e 5. Correlation between the first-order rate constants
for the cleavage (k1) of (SP)-3′,5′-Up(s)U (1a ) in the presence
of metal ions ([M2+] ) 5 mmol L-1) and the stability constants
KM
of the complexes the metal ions form with the
M(AMPS),calc
Zn2+ also promotes the desulfurization of 3′,5′-Up(s)U.
The negatively charged sulfur ligand may be expected
thiophosphate moiety of adenosine 5′-O-thiomonophosphate.
Values for the stability constants were taken from ref 23.
to prefer an equatorial position upon formation of I2-Mz+
.
concentration. Accordingly, one proton is removed either
from the predominant ionic form (i.e., monoanion) of 3′,5′-
Up(s)U or from the metal aquo ion on going to the
transition state, and hence, a dianionic thiophosphorane
(I2-Mz+) may be expected to be obtained (Scheme 5). The
sulfur-bound metal ion increases the electrophilicity of
the phosphorus atom, and most likely it also participates
in the deprotonation of the attacking hydroxy group
either by direct coordination to the 2′-oxygen, or, as
indicated in Scheme 5, by offering an intracomplex base
catalyst, the hydroxo ligand. The dianionic thiophos-
phorane (I2-Mz+) may in principle be regarded either as
an intermediate or a transition state. Cleavage of the
P-O5′ bond then leads to 2′,3′-cUMPS. Since the reac-
tion proceeds by inversion of configuration at phosphorus
(Scheme 3), the attack of the 2′-oxyanion must take place
“in-line” to the scissile P-O5′ bond: the cleavage prod-
ucts are formed without pseudorotation. Dianionic phos-
phoranes are usually considered to be so unstable that
they are rather regarded as transition states than
intermediates,31 although opposite opinions have also
been presented.32 Quantum chemical calculations33 sug-
gest that metal ions do not stabilize dianionic oxyphos-
phoranes in a same manner as a proton does. The fact
that isomerization and desulfurization are also promoted
by Zn2+, however, suggests that the dianionic thiophos-
phorane, I2-Mz+, may in this case have a definite lifetime.
In other words, it could be regarded as a marginally
stable intermediate that predominantly decomposes by
an in-line mechanism without pseudorotation but may
also be stabilized by protonation to a monoanionic species
that pseudorotates. In this respect, the situation differs
from that observed with 3′,5′-UpU: only cleavage of the
latter, not isomerization, is promoted by metal ions.27
Accordingly, the sulfur coordinated metal ion appeares
to stabilize the dianionic thiophosphorane. We conclude
As discussed above, the considerably faster cleavage
reaction proceeds by an “in-line” mechanism, and the
apical positions may thus be assumed to be predomi-
nantly occupied by the attacking 2′-oxygen and the 5′-
linked nucleoside. Accordingly, a pseudorotation should
precede the cleavage of the PS bond. The reaction could
hence be expected to exhibit a pH-dependence similar to
that of the isomerization. However, the desulfurization
rate is increased with the increasing pH much more
markedly than the isomerization rate, the apparent
reaction order being 0.7 (Figure 2). This kind of deviation
from both zero- and first-order dependence might be
accounted for by competing mechanisms, the relative
proportions of which are varied with pH (and possibly
metal ion concentration). Possibly, the reaction partially
proceeds by direct “in-line” displacement of the metal-
bound sulfur ligand. In other words, metal ion binding
to sulfur would enable the sulfur ligand to adopt initially
an apical position in the thiophosphorane, though at a
low level compared to the 5′-linked nucleoside. Depar-
ture of the sulfur ligand then leads to formation of an
unstable cyclic triester that is decomposed to a mixture
of 2′,5′- and 3′,5′-UpU.
It is worth noting that Gd3+, which accelerates the
cleavage of 3′,5′-Up(s)U almost as efficiently as Zn2+, does
not promote the isomerization or desulfurization. Gd3+
as a hard Lewis acid may be expected to prefer coordina-
tion to oxygen. Possibly, metal ion binding to oxygen
does not stabilize the dianionic thiophosporane (I2-Mz+
)
sufficiently to allow pseudorotation, and it does not
enable the sulfur ligand to adopt an apical position upon
.
formation of I2-Mz+ In other words, I2-Gd 3+ would
rather be a transition state than an intermediate, and it
is hence decomposed only to the cleavage products. As
mentioned above, metal ions do not accelerate the
isomerization of 3′,5′-UpU.27 Binding of the metal ion
to the sulfur ligand appears to be a prerequisite for the
metal ion-promoted isomerization (and desulfurization).
The results of the present study show a completely
different dependence of the cleavage rate on the nature
of the metal ion than those reported19 for the metal-ion-
promoted cleavage of a 5′-thiolate linkage in a dinucleo-
side phosphorothiolate, uridylyl-(3′,5′)-(5′-deoxy-5′-thio-
(31) Perreault, D. M.; Anslyn, E. V. Angew. Chem., Int. Ed. Engl.
36, 432.
(32) (a) Taira, K.; Uchimaru, T.; Storer, J . W.; Yliniemela¨, A.;
Uebayasi, M.; Tanabe, K. J . Org. Chem. 1993, 58, 3009. (b) Yliniemela¨,
A.; Uchimaru, T.; Tanabe, K.; Taira, K. J . Am. Chem. Soc. 1993, 115,
3032.
(33) Uebayasi, M.; Uchimaru, T.; Koguma, T. Sawata, S.; Shi-
mayama, T.; Taira, K. J . Org. Chem. 1994, 59, 7414.

Phosphoromonothioate Analogues of Uridylyl(3′,5′)uridine
Sch em e 5
J . Org. Chem., Vol. 63, No. 9, 1998 2945
uridine) (12). With the latter compounds, Mg2+, Zn2+
,
the cleavage by a factor of 10, 71, 370, and 3400,
respectively.20 Accordingly, one may ask why a P-S5′
bond in a dimeric and polymeric substrate behaves so
differently. A tentative explanation may be that with
oligomeric substrates, but not with dimers, the bridging
sulfur participates in the preequilibrium binding of the
metal ion. Additional phosphodiester bonds in a poly-
meric substrate may result in a multidentate binding
that further strengthens the interaction with the bridging
sulfur. For comparison, it has been suggested that
chimeric ribo/2′-deoxyribo oligonucleotides form macro-
chelates with metal ions, and this enhanced metal ion
binding is reflected as accelerated cleavage. In particu-
lar, 3′-terminal monophosphate groups have been shown
to influence in this manner.34,35
and Cd2+ at the concentration of 10 mmol L-1 facilitated
the cleavage by factors of 2, 13, and 8, respectively, i.e.,
even less than the cleavage of 3′,5′-UpU.27 With these
phosphorothiolates, formation of the pentacoordinated
intermediate alone may be expected to be rate-limiting,
since an alkylthio group is known to be a several orders
of magnitude better leaving group than an alkoxy group.
Hence, the observed effects represent the influence on
the formation of a pentacoordinated species that is then
decomposed in a kinetically invisible manner. The metal
ions are coordinated to the nonbridging oxygen and
increase the electrophilicity of phosphorus and, in all
likelihood, also the nucleophilicity of the attacking 2′-
hydroxy group either by direct coordination to the 2′-
oxygen or by deprotonating the 2′-hydroxy function with
the aid of their hydroxo ligand (intracomplex general base
catalysis). Any stabilizing interaction with the leaving
group would not be reflected to kinetics. By contrast,
with 3′,5′-Up(s)U, soft Lewis acids prefer coordination to
sulfur. This increases the stability of the metal ion/
substrate complex and, hence, the concentration of the
reactive dianionic thiophosphorane, I2-Mz+‚. Accord-
ingly, a significant thio effect on the metal-ion-promoted
Mech a n ism s for th e Rea ction s of 2′,3′-cUMP S.
Zn2+ and Cd2+ promote the hydrolysis of 2′,3′-cUMPS
even more markedly than the cleavage of 3′,5′-Up(s)U
and also much more effectively than the hydrolysis of
2′,3′-cUMP (Table 2). Most probably, these metal ions
bind to the sulfur ligand in a rapid preequilibrium step.
The hydrolysis is first order in both the metal and
hydroxide ion concentration, suggesting that the reaction
proceeds via a dianionic thiophosphorane, analogously
to the cleavage of 3′,5′-Up(s)U. As discussed previously
for 2′,3′-cUMP,26 the hydroxo ligand of the phosphate-
bound metal ion may serve either as an intracomplex
nucleophile or as an intracomplex general base catalyst
that deprotonates the attacking water molecule (Scheme
6).
The striking difference between the metal-ion-pro-
moted reactions of 3′,5′-Up(s)U and 2′,3′-cUMPS is that
with the latter the desulfurization is enhanced as ef-
fectively as the diester cleavage. The desulfurization is,
similarly to hydrolysis, first-order in both the metal and
hydroxide ion concentration. According to the classical
guidelines of Westheimer36 on pseudorotating phospho-
rane intermediates, the apical positions of the thiophos-
phorane intermediate should initially be occupied by the
attacking hydroxide ion and one of the sugar oxygen
cleavage is observed with Zn2+ and Cd2+
.
In striking contrast to the results obtained with 12,19
the catalytic effects of various metal ions on the cleavage
of a P-S5′ linkage in a chimeric oligonucleotide, d(ApCpG-
pGpTpCpT)p-rCps-d(ApCpGpApGpC)18 (for ps linkage,
see 12), closely resemble those observed for the cleavage
(34) Butzow, J . J .; Eichhorn, G. L. Biochemistry 1971, 10, 2019.
(35) Kuusela, S.; Azhayev, A.; Guzaev, A.; Lo¨nnberg, H. J . Chem.
Soc., Perkin Trans. 2 1995, 1197.
of 3′,5′-Up(s)U in the present study. Mg2+, Mn2+, Zn2+
and Cd2+ at a concentration of 5 mmol L-1 accelerated
,
(36) Westheimer, F. Acc. Chem. Res. 1968, 1, 70.

2946 J . Org. Chem., Vol. 63, No. 9, 1998
Ora et al.
Ta ble 4. Obser ved Reten tion Tim es for th e Hyd r olytic
P r od u cts of 1a a n d 1b on RP HP LCa
Sch em e 6
compd
tRb/min
compd
tRc/min
3a
3b
6
12.0
35.5e
7.6
13.0
11.2
8.9
1a
2a
1b
2b
3b
4
53.0
12.8
21.0e
10.3e
7.9e
7,8d
9
10
11
11.5
7.6
10.6
5
a
On a Hypersil ODS 5 column (250 × 4 mm, 5 (µm). Formic
acid/sodium formate buffer (0.045/0.015 mol L-1) containing
tetramethylammonium chloride 0.1 mol L-1 and an appropriate
amount of acetonitrile were used as an eluent. Flow rate was 1
b
mL min-1
.
No acetonitrile added. c With 3.5% (v/v) acetonitrile
d
in the buffer. 7 and 8 may be separated from each other by using
0.1 mol L-1 ammonium acetate as an eluent. e Not detected to
accumulate when 1a was used as the starting material.
atoms. If it is accepted that the ligands may depart from
the intermediate only via an apical position, the dianionic
thiophosphorane must pseudorotate to enable desulfur-
ization. This pseudorotation of the dianionic thiophos-
phorane appears to compete with the “in-line” displace-
ments of the 2′- and 3′-oxyanions much more efficiently
than the pseudorotation of the corresponding intermedi-
ate derived from 3′,5′-Up(s)U competes with the “in-line”
displacement of the 5′-linked nucleoside. Possibly, the
metal aquo ligand really serves as an intracomplex
general acid catalyst in the latter reaction.
Rea ction s of 2′- a n d 3′-UMP S (7, 8). The metal-ion-
promoted cyclization/desulfurization of 2′- and 3′-UMPS
to 2′,3′-CUMP (route F in Scheme 2) is also of interest,
since this reaction does not take place in the absence of
metal ions. One may tentatively assume that the metal
ion undergoes a rapid initial binding to the sulfur ligand
and, hence, increases the electrophilicity of the phospho-
rus atom of the monoanionic thiophosphate. Attack of
the neighboring hydroxy group then results in an “in-
line” displacement of the metal-bound sulfur ligand.
Whether the nucleophilicity of the attacking hydroxy
group is increased by the metal ion, as suggested in
Scheme 5 for the cleavage and desulfurization of 3′,5′-
Up(s)U, remains unanswered. The dethiophosphoryla-
tion (route G in Scheme 2) that competes with the
cyclization is not metal-ion-promoted, but the reaction
is retarded by increasing concentration of the metal ion.
As discussed previously,28 this reaction most likely
proceeds by a dissociative mechanism leading to forma-
tion of a metathiophosphate ion. Evidently, binding of
a metal ion to the sulfur ligand stabilizes more markedly
the starting material than the developing metathiophos-
phate ion.
bath to quench the reactions. Composition of the samples was
analyzed by RP HPLC (UV detection at 260 nm), usually with
a
combination of two isocratic elutions. The separation
methods and observed retention times are reported in Table
4. The signal areas of all monomeric compounds were as-
sumed to be proportional to concentrations, since the base
moiety of all the compounds was the same (1-substituted
uracil). With the dinucleoside monophosphates and their
thioate analogues, the molar absorptivities of the two base
moieties were assumed to be additive.
The hydronium ion concentration of the reaction solutions
was adjusted with HEPES [4-(2-hydroxyethyl)-1-pipera-
zineethanesulfonic acid] or MES [N-morpholinoethanesulfonic
acid] buffer. The tendency of these buffer acids to complex
with the metal ions studied is known to be low.38 The pH
values of the buffers were measured at 298.2 K and extrapo-
lated to 363.2 K by using the known temperature dependence
of the pKa values of the buffer acids.38 The pH values remained
constant within (0.2 units during each kinetic run. The metal
ions were added as nitrate salts, and the ionic strength was
adjusted with NaNO3.
Ca lcu la tion of th e Ra te Con sta n ts. The first-order rate
constants (kdi) for the decomposition of 3′,5′-Up(s)U (1a or 1b)
were calculated by applying the integrated first-order rate
equation to the diminution of the peak area of the starting
material. Good correlation was obtained in each run for more
than 2 half-lives. The concentration of the isomerization
products (2a or 2b, respectively) remained low, and their
subsequent breakdown was always much faster than isomer-
ization back to the starting material.
Rate constants for the diester hydrolysis (k1) of 1a or b via
2′,3′-cUMPS (3a or 3b) were calculated by the rate equation
of parallel and consecutive first-order reactions (eq 1). Here,
[2′,3′-cUMPS]t
[3′,5′-Up(s)U]0
k
dit
5
)
(e-k - e-k t
)
(1)
k5 - kdi
[2′,3′-cUMPS]t stands for the concentration of the appropriate
diastereomer of the cyclic 2′,3′-thiophosphate at moment t and
k5 for the first-order rate constant of its breakdown (values
for k5 were determined independently, see below). [3′,5′-
Up(s)U]0 is the initial concentration of the starting material.
Rate constants k1 for the Cd2+-promoted hydrolysis of 3′,5′-
Up(s)U were calculated by the method of initial rates on the
basis of the accumulation of 2′,3′-cUMPS during the first 5%
of the reaction.
The first-order rate constants for desulfurization (k2) of 1a ,b
were calculated by eq 2, where [UpU]t stands for the concen-
tration of the equilibrium mixture of 2′,5′- and 3′,5′-UpU (4,
5) at moment t, and k4 is the first-order rate constant for their
hydrolysis.27
Exp er im en ta l Section
Ma ter ia ls. (SP)- and (RP)-3′,5′-Up(s)U (1a ,b),37 (SP)- and
(RP)-2′,3-cUMPS (3a ,b),24 and 2′- and 3′-UMPS (7, 8)24 were
obtained by the procedures described previously. 3′,5′- and
2′,5′-UpU (4, 5), uridine, and its monophosphates used as
reference materials were commercial products of Sigma. The
metal salts and buffer acids used were of analytical grade.
Kin etic Mea su r em en ts. The hydrolytic reactions were
carried out in sealed tubes immersed in a thermostated water
bath at 363.2 K ((0.1 K). The initial substrate concentration
was ca. 10-4 mol L-1. Aliquots withdrawn from the reaction
mixture at appropriate time intervals were cooled in an ice
(37) Almer, H.; Stawinski, J .; Stro¨mberg, R.; Thelin, M. J . Org.
Chem. 1992, 57, 6163.
(38) Good, N. E.; Winget, G. D.; Winter, W.; Connolly, T. N.; Izava
S.; Singh, R. M. M. Biochemistry 1966, 5, 467.

Phosphoromonothioate Analogues of Uridylyl(3′,5′)uridine
J . Org. Chem., Vol. 63, No. 9, 1998 2947
order rate constants of desulfurization (k5′) and hydrolysis (k5′′)
on the basis of the product distribution at the very early stages
of the reaction: concentration ratio [6]:([7] + [8]) extrapolated
to reaction time zero is equal with the ratio of the rate
constants of formation of these products. The first-order rate
constants for the cyclization/desulfurization (k6) and dethio-
phosphorylation (k7) of 7 were obtained analogously. Since the
total disappearance (k6 + k7) of the starting material showed
a slight positive deviation from the first-order kinetics, the
individual rate constants referring to various moments t were
extrapolated to time zero. The values obtained were then
bisected to the contributions of k6 and k7 on the basis of the
product distribution during the early stages of the reaction.
[UpU]t
k2
dit
(e-k - e-k t
)
(2)
4
)
k4 - kdi
[3′,5′-Up(s)U]0
The first-order rate constants, k3, for isomerization of 1a to
1b were calculated on the basis of the product distribution at
the very early stages of the reaction (eq 3), when the contribu-
tions of the slow reverse reaction and the subsequent reactions
of 2’,5’-Up(s)U may be neglected.
[2′,5′-Up(s)U]t
k3 )
kdi
(3)
[3′,5′-Up(s)U]0 - [3′,5′-Up(s)U]t
The first-order rate constants, k5, for the disappearance of
3a or 3b were calculated by the integrated first-order rate
equation on the basis of the diminution of the starting material
peak area. The rate constants k5 were bisected to the first-
Ack n ow led gm en t. The financial support of the
Academy of Finland is gratefully acknowledged.
J O972112N