Mimics of small ribozymes utilizing a supramolecular scaffold

Article abstract of DOI:10.1039/c2dt10193a

For elucidating the mechanism of the general acid/base catalysis of the hydrolysis of RNA phosphodiester bonds, a number of cleaving agents having two cyclen moieties tethered to a 1,3,5-triazine core have been prepared and their ability to bind and cleave uridylyl-3′,5′-uridine (UpU) studied over a wide pH range. Around neutral pH, the cleaving agents form a highly stable ternary complex with UpU and Znⁱⁱ through coordination of the uracil N3 and the cyclen nitrogen atoms to the Znⁱⁱ ions. Under conditions where the triazine core exists in the deprotonated neutral form, hydrolysis of UpU, but not of adenylyl-3′,5′-adenosine (ApA), is accelerated by approximately two orders of magnitude in the presence of the cleaving agents, suggesting general base rather than metal ion catalysis. The probable mechanism of the observed catalysis and implications to understanding the general acid/base-catalyzed phosphodiester hydrolysis by ribozymes are discussed. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2dt10193a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 3328
www.rsc.org/dalton
PAPER
Mimics of small ribozymes utilizing a supramolecular scaffold†
Tuomas A. Lönnberg,*a Mia Helkearo,a Attila Jancsób and Tamás Gajda*b
Received 3rd February 2011, Accepted 15th December 2011
DOI: 10.1039/c2dt10193a
For elucidating the mechanism of the general acid/base catalysis of the hydrolysis of RNA phosphodiester
bonds, a number of cleaving agents having two cyclen moieties tethered to a 1,3,5-triazine core have been
prepared and their ability to bind and cleave uridylyl-3′,5′-uridine (UpU) studied over a wide pH range.
Around neutral pH, the cleaving agents form a highly stable ternary complex with UpU and ZnII through
coordination of the uracil N3 and the cyclen nitrogen atoms to the ZnII ions. Under conditions where the
triazine core exists in the deprotonated neutral form, hydrolysis of UpU, but not of adenylyl-3′,5′-
adenosine (ApA), is accelerated by approximately two orders of magnitude in the presence of the cleaving
agents, suggesting general base rather than metal ion catalysis. The probable mechanism of the observed
catalysis and implications to understanding the general acid/base-catalyzed phosphodiester hydrolysis by
ribozymes are discussed.
Introduction
An increasing number of cases are reported, where nucleobases
of a ribozyme function as general acid/base catalysts for the clea-
vage of phosphodiester bonds.1 Either, the nucleophilicity of the
attacking 2′-hydroxy group is enhanced by concomitant proton
transfer to a nucleobase, and/or the departure of the 5′-oxygen is
facilitated by proton donation from a nucleobase (Scheme 1). To
learn more about the mechanistic details of these processes,
general acid/base catalysis of RNA cleavage has been exten-
sively studied with simple buffer acids and bases, such as imida-
zole, morpholine or acetic acid. Unfortunately, the observed
catalysis in such systems is so weak that high buffer concen-
trations have to be employed, making elimination of medium
effects difficult. The cleavage of UpU in 1.0 mol L−1 imidazole
buffer, for example, is only three times as rapid as at buffer con-
centration zero.2 The results are, hence, somewhat ambiguous and
open to various interpretations.3 More specifically, whether the
reaction is catalyzed by both the basic and the acidic buffer com-
ponents or only the former, as well as the extent of proton transfer
in the transition state still remain matters of some controversy.
A possible way to learn more about the role of general acid/
base catalysis in the RNA cleavage is to bring the local concen-
tration of the general acid/base catalyst to a sufficient level
Scheme 1 General acid/base-catalyzed cleavage of RNA phosphodi-
ester bonds.
without causing undesired solvent effects. This can be achieved
by anchoring the catalyst close to the scissile phosphodiester
linkage. Unfortunately, the well-known Watson–Crick base-
pairing between two complementary oligonucleotides cannot be
used for the purpose. Owing to the stacking of contiguous base-
pairs, such double helix formation forces the 5′-leaving group
into an equatorial position within the phosphorane intermediate,
effectively preventing cleavage.4
In this paper we apply another approach for bringing a cataly-
tic group close to the phosphodiester linkage of uridylyl-3′,5′-
uridine (UpU). The underlying idea is to anchor the catalyst to
the uracil bases with the aid of a scaffold that is sufficiently
flexible to allow the departing 5′-linked nucleoside to adopt the
apical position required for departure from the phosphorane
intermediate. The anchoring is based on the strong binding of
ZnII chelates of small azacrowns to the deprotonated N3 atom of
a uracil base, first reported by Kimura.5 In this type of binding,
the ZnII ion coordinates directly to the N3 atom, while two of the
amino functions of the azacrown donate hydrogen bonds to the
oxo substituents of the uracil (or thymine) base (Fig. 1). It is
worth pointing out that the resulting ternary complex is more
aDepartment of Chemistry, University of Turku, FIN-20014 Turku,
Finland. E-mail: tuanlo@utu.fi; Fax: +358 2 333 6700;
Tel: +358 2 333 6777
bDepartment of Inorganic and Analytical Chemistry, University of
Szeged, 6721 Szeged Dóm tér 7, Szeged, Hungary. E-mail: tamas.gajda@
chem.u-szeged.hu; Fax: +36 62 544 340; Tel: +36 62 544 435
†Electronic supplementary information (ESI) available: Synthetic details
1
and H and 13C NMR spectra for compounds 5, 6, 7, 2h, 2i, 2n, 2o and
8, 31P NMR spectrum for compound 8, high resolution mass spectra and
HPLC chromatograms of compounds 2h, 2i, 2n and 2o and an example
illustrating the fitting of the potentiometric data. See DOI: 10.1039/
c2dt10193a
3328 | Dalton Trans., 2012, 41, 3328–3338
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 1 Complexes of UpU with bi- and trinuclear ZnII chelates of 1,5,9-trizazcyclododecane (1a and 1b) and cyclen (2h, 2i, 2n and 2o).
stable than any of the possible binary complexes. The soundness
of this approach has been previously demonstrated with related
di- and trinuclear azacrown derivatives (1a, 1b), which have
been shown to form a highly stable complex with UpU (Fig. 1).
The binuclear bis(azacrown) compound (1b) lacks catalytic func-
tionality since both of its ZnII-chelates are engaged in binding to
the uracil bases, but still allows the uncatalyzed cleavage of UpU
to proceed at an unaltered rate. The third ZnII:azacrown unit of
1a operates as a catalytic center, promoting the hydrolysis of
UpU 100-fold compared to the corresponding monomeric com-
pound.6 In other words, even though the bi- and trinuclear ZnII:
azacrown compounds bind tightly with UpU, they still allow the
cleavage of its phosphodiester bond to take place.
In the new scaffolds (2h, 2i, 2n and 2o) employed in the
present work, the two UpU anchoring ZnII:cyclen chelates are
linked to a common 1,3,5-triazine core instead of the benzene
core applied in 1b. The former has potential to function as a
“shuttle” mediating proton transfer from the attacking 2′-hydroxy
function to a non-bridging oxygen of the phosphorane intermedi-
ate and, finally, to the leaving group. A similar mechanism has
been proposed for catalysis by the hammerhead and hairpin ribo-
zymes, with a guanine base transferring a proton from the attack-
ing 2′-oxygen to the pro-RP non-bridging phosphoryl oxygen.7 It
should be noted that the ZnII chelate of cyclen is a poorer catalyst
for phosphodiester cleavage than the ZnII chelate of the pre-
viously used 1,5,9-triazacyclododecane.8 It is, hence, expected
that the cyclen moieties are only engaged in binding with the
uracil bases and are catalytically inactive (Fig. 1). Four different
cleaving agents were prepared, differing only in the substituent
at position 6 of the triazine ring. In compounds 2h, 2n and 2o,
the role of this substituent is simply to adjust the pKa value of
the triazine core, whereas the histamine sidearm of 2i has the
potential to function as a general acid/base catalyst on its own.
The impact of the pKa of the triazine on the catalytic efficiency
of the cleaving agent should provide valuable information on the
mechanism of the general acid/base catalysed cleavage of RNA
phosphodiester bonds. A strong dependence of the rate of clea-
vage on the basicity of the catalyst would suggest that the
reaction is best understood in terms of simple general base cata-
lysis whereas a pKa-independent catalysis would be more con-
sistent with a mechanism where both the basic and the acidic
component contribute.
Results and discussion
Preparation of the cleaving agents
Synthesis of the 1,3,5-triazine-based cleaving agents 2i, 2n and
2o is presented in Scheme 2. First, a 2-aminoethyl sidearm was
attached to the free secondary amino function of N4,N7,N10-tri-
tert-butoxycarbonylcyclen (3)9 by ring-opening of N-4-nitroben-
zenesulfonylaziridine (4), prepared from ethanolamine as pre-
viously described for several other aminoalcohols.10 Subsequent
removal of the 4-nitrobenzenesulfonyl group afforded Boc-pro-
tected aminoethyl-functionalized cyclen (6) which was used to
displace two of the chloro substituents of cyanuric chloride,
giving the protected non-substituted scaffold (7). Finally, the last
chloride was displaced by histamine, methylamine or methanol,
the Boc protections were removed and the products were con-
verted to free amines (2i, 2n and 2o, respectively) by passing
them through an OH−-form strong anion exchange resin. In the
synthesis of 2h, 2,4-dichlorotriazine was used instead of cyanu-
ric chloride and the product thus obtained was deprotected and
passed through an OH−-form strong anion exchange resin to
yield 2h as the free amine (Scheme 3). The identity and purity of
1
2i, 2n, 2o and 2h was verified by ESI-HRMS, H and 13C NMR
spectroscopy and RP HPLC. For the experimental details and
the spectroscopic data, see the supporting information†.
Preparation of 2′-O-methyluridylyl-3′,5′-uridine
For a nonreactive UpU mimic to be used as a target when deter-
mining the stability of the complexes, 2′-O-methyluridylyl-
3′,5′-uridine (8) was prepared by tetrazole-promoted coupling
of commercially available 5′-O-(4,4′-dimethoxytrityl)-2′-O-
methyluridine-3′-(2-cyanoethyl)-N,N-diisopropylphosphoramidite
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3328–3338 | 3329

View Article Online
Scheme 2 Preparation of the cleaving agents 2i, 2n and 2o. Reagents and conditions: (a) 4-nitrobenzenesulfonyl chloride, pyridine, CH2Cl2, (b)
KOH, H2O, CH2Cl2, (c) 4, MeCN, (d) thioglycolic acid, DBU, DMF, (e) cyanuric chloride, DIPEA, MeCN, (f) histamine, DIPEA, MeCN, (g)
methylamine, EtOH, (h) NaOMe, MeOH, (i) TFA, CH2Cl2, MeOH.
Scheme 3 Preparation of the cleaving agent 2h. Reagents and conditions: (a) 2,4-dichlorotriazine, DIPEA, MeCN, (b) TFA, CH2Cl2, MeOH.
Scheme 4 Preparation of 2′-O-methyluridylyl-3′,5′-uridine (8, mUpU). Reagents and conditions: (a) 2′,3′-O-isopropylideneuridine, 1H-tetrazole,
MeCN, (b) I2, 2,6-lutidine, THF, H2O, (c) NH3, MeOH, (d) HCl, H2O, (e) Dowex 50Wx8 (Na+), H2O.
and 2′,3′-O-isopropylideneuridine (Scheme 4). Removal of
the cyanoethyl, dimethoxytrityl and isopropylidene pro-
tections and subsequent treatment with Na+-form strong cation
exchange resin afforded 8 as the sodium salt. For the ex-
perimental details and characterization, see the supporting
information.
3330 | Dalton Trans., 2012, 41, 3328–3338
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 1 Formation constants (logβpqrs) and some derived equilibrium
data of the ZnII complexes of 8 (mUpU) and 2n (L) (I = 0.1 M NaCl,
T = 298 K)a
Determination of the stability of the complexes by
potentiometric and spectrophotometric titrations
pqrsb (species)
logβpqrs
pqrsb (species)
logβpqrs
In order to have thermodynamic data on the interaction between
ZnII and the 1,3,5-triazine-based ligands, as well as between
UpU and the binuclear ZnII complexes, we have studied the pro-
tonation of 8 (mUpU) and 2n (L), the complex formation in the
ZnII:L, ZnII:mUpU binary, and in the ZnII:mUpU-L ternary
systems. The comparison of the forward and backward titration
curves revealed slow kinetics of complex formation processes in
the ZnII-L systems between pH 3.5–5. To minimize the errors
due to the slow kinetics, considerably increased equilibration
time has been applied in this pH range (see Experimental
section). The determined formation constants and derived data
for the relevant equilibria are listed in Table 1. The species distri-
bution curves of the ZnII:L:mUpU systems are depicted in
Fig. 2.
The protonation constants of mUpU (8) are almost identical
with those of UpU.11 Since the interaction of ZnII with mUpU is
rather weak, slight hydrolysis of ZnII was observed above pH 8.5
in the ZnII:mUpU system. Therefore only the data below pH 8.5
were considered during our evaluation, and the hydroxo com-
plexes of ZnII were taken into account.12 Under these conditions
the formation of two binary complexes Zn(HmUpU) and
Zn(HmUpU)(OH) was detected. In Zn(HmUpU) the metal ion
is very likely coordinated to one of the uracil-N3 nitrogen atoms.
Its deprotonation (pK = 8.08) is probably related to the for-
mation of a mixed hydroxo species.
1101 (Zn(HmUpU)
1001(ZnHmUpU(OH)) 5.04 0.09
1310 (ZnH3L)
13.12 0.09 1–110 (ZnL(OH)
2110 (Zn2HL)
37.75 0.05 2010 (Zn2L)
32.94 0.05 2–110 (Zn2L(OH))
24.73 0.06 2–210 (Zn2L(OH)2) 9.38 0.06
15.78 0.06 1011(Zn2L(mUpU)) 35.70 0.03
5.24 0.06
31.76 0.05
26.16 0.05
18.19 0.06
1210 (ZnH2L)
1110 (ZnH2L(OH)c)
1010 (ZnHL(OH)c)
Derived data for relevant equilibria
log K1 (Zn2++ HmUpU = Zn(HmUpU)) = 3.80
log K2 (Zn(HmUpU) + OH− = Zn(HmUpU)(OH)) = 5.67
log K3 (Zn2+ + H3L = ZnH3L) = 14.10d
log K4 (Zn2+ + H2L = ZnH2L) = 13.58d
log K5 (Zn2+ + HL = ZnHL) = 13.96d
log K6 (Zn2+ + L = ZnL) = 15.78
log K7 (ZnH2L + OH− = ZnH2L(OH)) = 5.54
pK8 (ZnH2L(OH) = ZnHL(OH) + H+) = 8.95
pK9 (ZnHL(OH) = ZnL(OH) + H+) = 10.54
log K10 (ZnL + Zn2+ = Zn2L) = 10.38
log K11 (ZnL(OH) + Zn2+ = Zn2L(OH)) = 12.95
log K12 (Zn2L + OH− = Zn2L(OH)) = 5.78
log K13 (Zn2L(OH) + OH− = Zn2L(OH)2) = 4.94
logK14 (Zn2L + mUpU = Zn2L(mUpU)) = 9.54
logK15 (Zn2L + H2mUpU = Zn2L(mUpU) + 2H+) = −9.11 (−7.94 at 90 °C)
a The protonation constants of mUpU and L are logβ0101 = 9.74 0.02,
logβ0201 = 18.65 0.02, and logβ0110 = 11.25 0.04, logβ0210 = 21.53
0.04, logβ0310 = 30.55 0.04, logβ0410 = 38.70 0.04, logβ0510 = 42.99
0.04, respectively. b p, q, r and s are the stoichiometric coefficients for ZnII,
H+, L (2n) and mUpU (8), respectively. c Main species, other protonation
isomers may also exist. d The following values were used for pK of non-zinc-
bounded cyclen ring: logβ0210/2 = 10.77 and (logβ0410 − logβ0210)/2 = 8.59
Over the studied pH range, five protonation steps of L (2n)
can be observed (Table 1). At pH 7 both cyclen rings are doubly
protonated. The first two and the second two protonation con-
stants show the expected statistical differences (∼0.9), and their
average (10.77 and 8.59) agrees well with the protonation con-
stants of N-alkyl mono-substituted cyclens.13 The fifth protona-
tion is related to the 1,3,5-triazine ring.
between the binding constants of the first and second metal ion
to L (logK6 and logK10 in Table 1, ΔlogK = 5.4) may indicate
the weak binding of the second cyclen ring to the metal ion in
ZnL, too. Such ‘earmuff’-like coordination is highly stable for
some bis(1,4,7-triaza-cyclononane) derivatives.17 Although the
relatively rigid 2,4-substituted triazine spacer would prefer the
formation of similar structure in the present case, this is rather
surprising considering the tetradentate nature of cyclen ring.
In the absence of aminoalkyl pendant arms, the metal ions in
the dinuclear Zn2L species are stronger Lewis acids than those
reported in ref. 16, therefore the deprotonation of metal bound
water molecules take place at ca 2 units lower pH. The Δlog K =
log K12–log K13 = 0.84 value corresponds to the statistical differ-
ence, and thus indicates no interaction between the two metal
centers in the dinuclear monohydroxo species (Zn2L(OH)).
In equimolar solutions of ZnII and L, considerable amounts of
binuclear complexes are present (Fig. 2A). This is the conse-
quence of the two independent and identical binding sites in L,
since under such conditions statistical considerations predict
nearly equal amounts of ligand bound in mono- and binuclear
complexes.
At first sight, the speciation diagrams in Fig. 2 show a compli-
cated picture, however, the composition of the formed species
are, in fact, rather obvious. The presence of two azacrowns allow
the formation of mono- and dinuclear complexes, and due the
acid–base properties of the non-bonded nitrogens and coordi-
nated water molecules, both may have several different protona-
tion states (an example illustrating the excellence of the fits of
the potentiometric data is included†). In the protonated mono-
nuclear ZnII complexes the metal ion is coordinated to one of the
macrocyclic units, while the other one, and the 1,3,5-triazine
ring in ZnH3L, is protonated. This is supported by the almost
identical log K values related to ZnII binding to the tri-, di- and
mono-protonated ligand (logK3, logK4 and logK5 in Table 1).
Although Kimura and his coworkers have reported a number
of equilibrium studies on Zn2-bis(cyclen) complexes related to
the equilibria Zn2L + X = Zn2LX (where X = OH−, 1-
methylthymine, TpT etc.),5,14,15 to our knowledge only a single
paper reports complete solution equilibrium study on the for-
mation of binary ZnII complexes of bis-cyclen derivatives.16
However, the ligands reported in ref. 16 contained 1,4-phenylene
spacers, and the macrocyclic units possessed aminoalkyl pendant
arms, which may also coordinate to the metal ions. As a conse-
quence, latter ligands provided two separated binding sites for
the metal ions. In contrast, the relatively large difference
In the ZnII:L:mUpU system our potentiometric study indicated
the formation of a single very stable ternary complex (Zn2L
(mUpU)), which is the sole species over a wide pH-range
(Fig. 2C). mUpU fits into the scaffold created by the Zn2L
complex almost ideally, since the value of logK14 (= 9.54 for the
reaction Zn2L + mUpU = Zn2L(mUpU)) is nearly twice as high
as the respective constant for the process Zn:cyclen + uridine =
Zn:cyclen(uridine), logK = 5.2.14
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3328–3338 | 3331

View Article Online
The Zn2L-mUpU interaction has also been studied by UV-
spectrophotometric titrations both at 25 °C and 90 °C (Fig. 3
and 4). The decrease in the UV absorbance of mUpU at 260 nm
upon addition of the Zn2L complex at 25 °C and pH 8.8 results
from the binding of the dinuclear species to the base moieties of
mUpU (Fig. 3). The spectrophotometric data depicted in Fig. 3
have been quantitatively evaluated in two different ways. Taking
into account the formation constants determined by potentio-
metry for the binary complexes, the spectrophotometric data
allowed the calculation of the formation constant of the ternary
species logβ2011 = 35.37
0.06, which agrees well with the
value determined by potentiometric measurements (see Table 1).
Furthermore, we also calculated an apparent stability constant for
the Zn2HxL + HxmUpU = Zn2L(mUpU) reaction (logKapp
=
5.98(3)), which can be directly compared to the value deter-
mined at 90 °C (see later). The measured and calculated absor-
bances at 260 nm are shown in Fig. 4 (curve a).
In order to have quantitative data on the stability of the ternary
Zn2L(mUpU) complex at the temperature used for the kinetic
measurements, we performed UV-spectrophotometric titrations
between pH* = 5–7 at 90 °C. The UV-titrations made at several
pH* allowed us to follow the Zn2L + H2mUpU = Zn2L(mUpU)
+ 2H+ reaction (Fig. 4, curve b–d). At pH* = 5 (Fig. 4, curve b)
the formation of the ternary species is negligible, therefore curve
b reflects the increasing amount of the binary Zn2L species
(which also absorbs at 262 nm at this temperature). On the other
hand, at pH* 7 (curve d) the majority of the added Zn2L
complex transforms into the ternary species until a 1 : 1 ratio of
Zn2L and mUpU is reached. These data indicate the formation
of the Zn2L(mUpU) ternary complex also at 90 °C. The equili-
brium constant determined for the above reaction at 90 °C is
logK15 = −7.94 0.06. The corresponding constant at 25 °C is
logK15 = −9.11, which indicates that with increasing temperature
the formation of the ternary complex shifts to a lower pH range.
At pH* = 7 we also calculated the apparent stability constants
Fig. 2 Species distribution curves for the ZnII complexes in the ZnII/L/
mUpU = 1/1/0 (A), 2/1/0 (B) and 2/1/1 (C) systems (T = 25 °C, I
(NaCl) = 0.1 mol L−1, [L] = 1 mmol L−1). 1110* stands for the sum of
the protonation isomers ZnHL and ZnH2L(OH), while 1010* stands for
ZnL and ZnHL(OH).
for the Zn2HxL + HxmUpU = Zn2L(mUpU) reaction (logKapp
=
5.10 0.08). The latter value shows tight binding of Zn2L to
mUpU even at this higher temperature.
Fig. 3 UV-absorption spectrum of mUpU (55 μmol L−1 aqueous solution) titrated with a solution that contained ZnII and L in a 2 : 1 ratio ([L]tot
1.55 mmol L−1) at pH 8.9 (T = 25 °C, I(NaCl) = 0.1 mol L−1).
=
3332 | Dalton Trans., 2012, 41, 3328–3338
This journal is © The Royal Society of Chemistry 2012

View Article Online
of uridine 2′- and 3′-monophosphates. Only traces of the initially
formed uridine 2′,3′-cyclic monophosphate are detected, owing
to its rapid hydrolysis to the 2′- and 3′-monophosphates. Below
pH* 8, mutual isomerization of 3′,5′- and 2′,5′-UpU is also
observed.
For the experiments carried out in the presence of an excess of
the cleaving agent, pseudo first-order rate constants for the disap-
pearance of UpU were obtained by applying the integrated first-
order rate law to the time-dependent diminution of the relative
peak area of UpU. When the reaction rate was studied as a func-
tion of UpU concentration at a constant limiting concentration of
the cleaving agent, initial rates were obtained as slopes of the
plots of the combined peak area of all the monomeric products
(2′,3′-cUMP, 3′-UMP, 2′-UMP and uridine) as a function of reac-
tion time.
Fig. 4 UV spectrophotometric titration (260 nm) of mUpU (55 μmol
L−1) with a solution that contained ZnII and L in a 2 : 1 ratio ([L]tot
=
1.55 mmol L−1) at pH = 8.9, T = 25 °C (a, ■), pH* = 5.0, T = 90 °C
(b, ○), pH* = 6.0, T = 90 °C (c, ◇) and pH* = 7.0, T = 90 °C (d, □)
(I(NaCl) = 0.1 mol L−1).
Dependence of rate on cleaving agent concentration
To verify the catalytic activity of the binuclear ZnII:cyclen com-
pounds synthesized, cleavage of UpU was followed at various
concentrations (0–80 μmol L−1) of the methylamino derivative
2n (T = 90 °C; pH* = 6.57). Cleavage of ApA was followed
under the same conditions for comparison. The dependence of
the observed pseudo first-order rate constant for the cleavage of
UpU and ApA as a function of [2n] is presented in Fig. 5.
Cleavage of UpU follows simple first-order kinetics and
attains a rate acceleration of nearly two orders of magnitude at
sufficiently high concentrations of the dizinc complex of 2n. The
observed first-order rate constant for the reaction may be
expressed by eqn (1).
Reaction pathways and product distribution
Hydrolytic reactions of UpU were followed at 90 °C over a pH*
range of 3.52–10.21 in the presence of a 12-fold excess of the
cleaving agent and at pH* 7.42 as a function of the concentration
of the cleaving agent or UpU by analyzing the composition of
the aliquots withdrawn from the reaction mixture at appropriate
time intervals by RP HPLC. In all cases, 2 eq. of Zn(NO3)2 rela-
tive to the cleaving agent were used. The reactions were started
by mixing a small amount of the UpU stock solution with the
reaction buffer prethermostated to 90 °C to give the desired UpU
concentration and quenched by cooling the aliquots to 0 °C. The
products were characterized by spiking with authentic samples.
Over the entire pH* range studied and regardless of whether
the cleaving agent is present in excess or not, the same product
distribution is observed (Scheme 5). Disappearance of UpU is
accompanied by formation of uridine and comparable amounts
½2nꢀ
kobs ¼ kuncat þ kcat
ð1Þ
Kd þ ½2nꢀ
kuncat and kcat are the first-order rate constants for the spon-
taneous and 2n-catalyzed hydrolysis, respectively, and Kd is the
Scheme 5 Hydrolytic reaction pathways of UpU.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3328–3338 | 3333

View Article Online
Fig. 6 Initial rate for the cleavage of UpU by 2n at 90 °C, pH* = 6.39,
Fig. 5 Pseudo first-order rate constants for the cleavage of UpU (■)
and ApA (○) as a function of the concentration of the cleaving agent 2n
at 90 °C, pH* = 6.57, I(NaClO4) = 0.10 mol L−1, [UpU] = [ApA] =
I(NaClO4) = 0.10 mol L−1, [UpU] + [2n] = 50 μmol L−1
.
5.0 μmol L−1
.
shape of a downward-opening parabola with a maximum at x
(UpU) = x(2n) = 0.5. As expected and also borne out by the
titration studies with 2n and 8, one molecule of UpU binds to
one molecule of 2n to form the reactive complex.
dissociation constant for the complex between UpU and
2n:2ZnII. The observed dissociation constant is logKd = −4.2
0.2, in reasonable agreement with the corresponding value
obtained by spectrophotometric titrations at pH* 7, taking into
account the difference in pH* (logKd = −logKapp = −5.1). In
contrast to the cleavage of UpU, cleavage of ApA is hardly
accelerated by 2n, indicating that binding of the nucleobases to
the ZnII:cyclen moieties of the cleaving agent is essential for cat-
alytic activity. In principle, one might argue that only one of the
azacrown chelates is engaged in UpU binding, the other one
serving as an intracomplex catalyst. The potentiometric studies,
however, show that the complex having both of the cyclen moi-
eties bound to uracil is overwhelmingly the most stable species.
This finding is also consistent with the earlier observations of
Kimura, according to which binding of 1,4-bis[(1,4,7,10-tetra-
azacyclododecan-1-yl)methyl]benzene to TpT is much stronger
than to 1-methylthymine.15 Furthermore, it has been previously
shown that the related binuclear ZnII:azacrown complex 1b:2ZnII
does not catalyze the cleavage of UpU, even though its
ZnII:1,5,9-triazacyclododecane moieties are known to be better
catalysts for phosphodiester cleavage than the ZnII:cyclen moi-
eties of 2n:2ZnII.6 It seems, hence, extremely unlikely that the
observed rate acceleration by 2n would be due to metal ion cata-
lysis. In all likelihood, the same conclusion holds also for the
other cleaving agents 2h, 2i and 2o.
pH-rate profiles
The pH-rate profiles for the disappearance of UpU in the pres-
ence (60 μmol L−1) or absence18 of the cleaving agents are pre-
sented in Fig. 7. In the absence of any cleaving agents, cleavage
of UpU becomes first-order in [OH−] at pH* 6 and remains so
until quantitative deprotonation of the attacking 2′-hydroxy func-
tion is reached around pH* 12.5.18 In contrast, in the presence of
the cleaving agents, the reaction becomes roughly second-order
Stoichiometry of the reactive complex
The fact that the 2n-catalyzed cleavage of UpU follows simple
first-order kinetics suggests 1 : 1 stoichiometry for the reactive
UpU:2n complex. For more compelling evidence, a Job’s plot
was constructed by measuring the initial rate of the cleavage of
UpU at pH* 6.39 as a function of the mole fraction of 2n,
keeping the total concentration of UpU and 2n constant: [UpU]
+ [2n] = 50 μmol L−1. The plot, presented in Fig. 6, has the
Fig. 7 pH-rate profiles for the cleavage of UpU in the absence (---)
and in the presence of the cleaving agents 2 h (□), 2i (▽), 2n (●) and
2o (▲) and the isomerization of UpU in the absence (⋯) and in the
presence of 2n (○) at 90 °C, I(NaClO4) = 0.10 mol L−1, [2] = 0/
60 μmol L−1
.
3334 | Dalton Trans., 2012, 41, 3328–3338
This journal is © The Royal Society of Chemistry 2012

View Article Online
in [OH−] at pH* 4, pH-independent at pH* 6 and finally first-
order in [OH−] at pH* 8. No data for the hydrolysis promoted by
the cleaving agents 2h, 2i, 2n and 2o could be obtained at pH*
< 4 owing to the instability of the UpU:2ZnII:2 ternary complex
under these conditions. Under the experimental conditions, i.e.
between pH* 3.52 and 10.21, the observed rate constant for the
cleavage of UpU, kcolbs, may be expressed by eqn (2).
no rate acceleration by the cleaving agents is expected. In fact,
between pH 3.5 and 4 a modest acceleration is observed but it
seems that on going to more acidic solutions this difference in
the rates of catalyzed and spontaneous reactions disappears. In
other words, under acidic conditions there is no contribution to
the cleavage of UpU by 2h, 2i, 2n or 2o but the reaction pro-
ceeds as previously described18 even when the cleaving agents
are present in a considerable excess.
k3½Zn2:2:UpUꢀmaxKa1Ka2
k4KW
½Hþꢀ
cl
obs
k
¼ k1½Hþꢀ þ k2 þ
þ
ð2Þ
2
½Hþꢀ þ Ka1½Hþꢀ þ Ka1Ka2
Cleaving agent catalyzed hydrolysis of UpU
Ka1 and Ka2 are the acidity constants of the two uracil bases
(deprotonation of N3 upon complexation with ZnII:cyclen),
[Zn2:2:UpU]max is the maximum concentration of the reactive
complex (0 or 5.0 μmol L−1), k2 is the first-order rate constant
for the pH-independent cleavage of UpU and k1, k3 and k4 are
the second-order rate constants for the hydronium ion, cleaving
agent and hydroxide ion catalyzed cleavage of UpU, respect-
ively. Assuming identical binding of mUpU and UpU to the
dizinc complex of 2n, the equilibrium data obtained at 90 °C by
spectrophotometric titrations allows one to calculate the concen-
tration of the reactive ternary species at any pH*, providing a
more accurate treatment of the kinetic data in the case of this
ligand. Thus eqn (2) is simplified to eqn (3).
Between pH 5 and 6, the cleavage of UpU is second-order in
[OH−] in the presence of the cleaving agents 2h, 2n, 2o, and 2i
and nearly pH-independent in the absence thereof. Over this
relatively narrow pH range, two pH-dependent equilibria prevail,
viz. coordination of the N3 atoms of the uracil bases to the ZnII
ions of the binuclear ZnII complexes of 2h, 2i, 2n and 2o with
concomitant deprotonation. For the coordination equilibria of
2h, 2n and 2o, the average pKa value obtained from eqn (2) is
(pKa1 + pKa2)/2 = 6.1, in excellent agreement with the spectro-
photometric titration curve obtained at 90 °C for the complex of
2′-O-methyluridylyl-3′,5′-uridine (8) and 2n:2ZnII (pK = 6.2).
With the histamine derivative 2i, eqn (2) gives pKa1 = 6.5 and
pKa2 = 4.9. The considerable difference between the two equili-
brium constants in the case of 2i is in all likelihood attributable
to the coordination ability of the imidazole ring. In other words,
in the dinuclear complex 2i:2ZnII the imidazole ring is probably
coordinated to one of the ZnII ion and this 5N-coordinated zinc
has reduced binding ability to the N3 nitrogens of UpU as com-
pared to the other, 4N-coordinated one. Therefore the ZnII pro-
moted deprotonation of UpU takes place in two separate steps.
Under conditions where the cleaving agent promoted hydroly-
sis of UpU prevails, the triazine core of the cleaving agents is in
the deprotonated neutral ionic form,19 suggesting that the cleav-
ing agents function as general base catalysts. The possibility of
general acid catalysis cannot be ruled out, however, due to the
instability of the UpU:2ZnII:2 ternary complexes under con-
ditions where the triazine core of the cleaving agents would be
protonated. Mutual isomerization of 2′,5′- and 3′,5′-UpU is
neither facilitated nor retarded by any of the cleaving agents.
The second-order rate constants for the cleaving agent cata-
lyzed hydrolysis of UpU, k3, are nearly identical for 2h, 2i, 2n
and 2o. In other words, the catalytic activity of the cleaving
agents is largely independent on the substitution at the 6-position
of the triazine ring, even when a potential general acid/base
k4KW
½Hþꢀ
cl
obs
k
¼ k1½Hþꢀ þ k2 þ k3 ½Zn2:2n:UpUꢀ þ
ð3Þ
The rate and equilibrium constants obtained by nonlinear
least-squares fitting of the experimental data to equations (2) and
(3) are presented in Table 2. With 2n, the rate constants obtained
by using equations (2) and (3) are in good agreement, supporting
the validity of our description of kinetic data. The deprotonation
of the two N3 nitrogens is strongly cooperative, and the respect-
ive dissociation constants cannot be separated based on the
kinetic data (or even by the potentiomeric measurements at
25 °C), except in the case of 2i. Therefore, for the other cleaving
agents, only the sum pKa1 + pKa2 has been tabulated.
Hydronium ion catalyzed hydrolysis of UpU
At pH < 4, cleavage of UpU is first-order in [H3O+] both in the
presence and in the absence of the cleaving agents 2h, 2i, 2n
and 2o. Under these conditions, the uracil bases remain proto-
nated and uncomplexed with the ZnII:cyclen moieties and, hence,
Table 2 Rate constants for the partial reactions of cleavage and isomerization of UpU in the presence of the cleaving agents 2h, 2i, 2n and 2o
(T = 90 °C, I(NaClO)4 = 0.10 mol L−1
)
2ha
2ia
2na
2nb
2oa
isom. with 2n
k1/10−4 mol−1 L s−1
k2/10−7 s−1
10 10
14
0.2 0.7
3.1 0.7
6
9
5
11
0.5 0.6
4.2 0.6
6
14
0.1 0.4
6
3
12.16 0.2
—
—
5
90 30
1
4
4
1
1
3
0.9 0.5
4.1 0.7
4
12.2 0.1
—
—
12
—
—
—
—
—
1
k3/mol−1 L s−1
k4/10−2 mol−1 L s−1
pKa1 + pKa2
pKa1
1
1
4
1
1
4
2
12.2 0.4
—
—
11.4 0.7
6.5 0.2
4.9 0.5
—
—
—
pKa2
a Calculated using eqn (2). b Calculated using eqn (3).
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3328–3338 | 3335

View Article Online
catalyst is attached (the histamine function of 2i). In all likeli-
hood, the substituent at the 6-position is oriented away from the
scissile phosphodiester bond in the reactive complex which, in
turn, suggests that the N3 atom is oriented towards it (Fig. 8).
For 4,6-diaminotriazines bearing a hydrogen, a methylamino
or a methoxy group at position 2, pKa values of 3.96, 5.28 and
3.54, respectively, have been determined at 25 °C.19 The pKa
value of the triazine ring of 2n, on the other hand, is 4.29, an
order of magnitude lower than the corresponding value of N2-
methyl-2,4,6-triaminotriazine. The difference is probably attribu-
table to the electron-withdrawing effect of the cyclen rings and,
in all likelihood, is essentially the same for 2h and 2o. In other
words, the difference in basicity of nearly 2 logarithmic units is
not reflected in the catalytic activity of the cleaving agents. The
observed general base catalysis is, hence, probably attributable to
sequential specific base and general acid catalysis, with the
attacking 2′-OH being deprotonated and a monoanionic phos-
phorane intermediate (I in Scheme 6A) formed in a rapid pre-
equilibrium step and proton transfer from this intermediate to the
departing oxygen taking place concerted with rate-limiting P–O
bond fission.20 In this kind of mechanism, the effects of the basi-
city of the general acid/base catalyst are opposite in the two
steps: higher basicity shifts the pre-equilibrium to favor the reac-
tive deprotonated species but at the same time retards proton
transfer at the rate-limiting step. Apparently, in the present case
these two effects entirely cancel each other out, indicating that
the catalyst-mediated proton transfer to the leaving group must
be nearly complete at the rate-limiting step. Finally, it should be
pointed out that direct general base catalysis, i.e. abstraction of
the proton from the 2′-OH by the cleaving agent at the rate-limit-
ing step, should result in strong dependence of the reaction rate
on the pKa of the triazine ring.
Together with the exocyclic amino functions at C2 and C4,
N3 of the triazine ring forms a conjugated network of hydrogen
bond donors and acceptors that, upon tautomerization, may
abstract and release a proton at any of the three nitrogen atoms.
More specifically, this network could mediate proton transfer
from the attacking 2′-OH to the phosphorane intermediate and,
ultimately, to the departing 5′-oxygen, closely mimicking the
proposed action of the catalytic guanine moiety in hammerhead
and hairpin ribozymes (Scheme 6A). One of the amino groups
of the 2,6-diaminotriazine core could deprotonate the attacking
2′-OH in a pre-equilibrium step and the other one protonate the
departing 5′-oxyanion in the rate-limiting step. In addition, the
protonated N3 of the triazine ring may interact with the nonbrid-
ging oxygen atoms by hydrogen bonding. In the initial state, the
proposed hydrogen bond between the protonated triazine N3 and
Fig. 8 Two possible conformations of the UpU:cleaving agent
complex, with the substituent X oriented away from (left) or towards
(right) the phosphodiester bond.
Scheme 6 Proposed mechanisms for (A) cleavage of UpU by 2h, 2i, 2n and 2o and (B) spontaneous cleavage of guanylyl-(3′,3′)-(2′-amino-2′-
deoxyuridine) (9).
3336 | Dalton Trans., 2012, 41, 3328–3338
This journal is © The Royal Society of Chemistry 2012

View Article Online
the nonbridging oxygen atoms of the phosphodiester linkage is
probably a weak one because the monoanionic phosphodiester is
only weakly basic. Upon the attack of the 2′-oxyanion on the
phosphorus, however, the basicity of the nonbridging oxygen
atoms increases, strengthening the hydrogen bond. In the phos-
phorane intermediate, the proton actually resides on one of the
nonbridging oxygen atoms. Upon breakdown of this intermedi-
ate, basicity of the nonbridging oxygen atoms decreases again
and negative charge builds up on the departing 5′-oxygen. The
2,6-diaminotriazine core of the cleaving agents may respond to
these changes by accepting a proton from a nonbridging oxygen
at N3 and donating another from an exocyclic amino function to
the developing 5′-oxyanion.
Although in the present case the catalytic effect of a cleaving
agent having the triazine ring protonated could not be studied,
previously obtained similar results with simpler model systems
tend to support the idea of the deprotonated triazine as a general
base catalyst. For example, the cleavage of guanylyl-(3′,3′)-(2′-
amino-2′-deoxyuridine) (9) exhibits a highly similar pH-rate
profile but in that case the reaction is first-order in [OH−]
between pH 4 and 6.21 Because the system is not complicated by
coordination equilibria between the nucleobases and cleaving
agents this first-order dependence in [OH−] could be directly
attributed to deprotonation of the 2′-amino group. In other
words, the 2′-amino function in its deprotonated form is a
general base catalyst, promoting the cleavage of 9. Also in this
case, the observed general base catalysis was interpreted in terms
of sequential specific base and general acid catalysis, as dis-
cussed above (Scheme 6B).20
in a pre-equilibrium step and proton transfer to the leaving group
is largely complete in the rate-limiting step.
Experimental
UV spectrophotometric titrations
The pH of 55 μmol L−1 solutions of mUpU (8) in 50 mmol L−1
buffers (MES, HEPES or CHES) were adjusted to the desired
value (pH = 8.9 at 25 °C, pH* = 5–7 at 90 °C. In the latter case
the pH of the solutions was measured at 298 K, and extrapolated
to 363 K using the known pKa values of the buffer acids at
363 K (throughout the paper the extrapolated pH is designated
as pH*).23 2.5 mL of these solutions were placed in a quartz cell
having an optical path of 1.0 cm. The temperature of the cell
housing block was adjusted to 25.0 0.1 °C (or 90 1 °C) and
the UV spectra were measured in the wavelength range
200–300 nm by a Unicam Helios α spectrophotometer. A more
concentrated solution of the ligand 2n (1.55 mmol L−1) and ZnII
(3.1 mmol L−1) was added portion-wise and the UV-spectrum
was measured after each addition. The spectrum of the buffer
was subtracted and the dilution of the solution was taken into
account before treatment of the data. Equilubrium data from the
spectrophotometric titrations were calculated using the
PSEQUAD computer program.24
pH-metric measurements
The protonation and coordination equilibria were investigated by
potentiometric titrations in aqueous solution (I = 0.1 M NaCl and
T = 298.0 0.1 K) under argon atmosphere using an automatic
titration set including a PC controlled Dosimat 665 (Metrohm)
autoburette and an Orion 710A precision digital pH-meter. The
Metrohm Micro pH glass electrode (125 mm) was calibrated25
via the modified Nernst equation ((4)).
Hydroxide ion catalyzed hydrolysis of UpU
At pH 8, hydrolysis of UpU in the presence of the cleaving
agents 2h, 2i, 2n and 2o becomes first-order in [OH−]. The
observed base catalysis in all likelihood refers to rate-limiting
departure of the uridine 5′-oxyanion from the marginally stable
dianionic phosphorane intermediate formed by attack of the 2′-
oxyanion on the phosphorus atom.22 The second-order rate con-
stant of the reaction is essentially independent on the presence of
the cleaving agents, suggesting that complex formation between
UpU and the cleaving agent as such has only a minor impact on
the hydrolysis of UpU and the rate acceleration observed at
lower pH is due to the ability of the 2,6-diaminotriazine moiety
to facilitate proton transfer from the attacking nucleophile to the
phosphorane intermediate and eventually to the departing
oxyanion.
JOHKw
½Hþꢀ
E ¼ E0 þ K log ½Hþꢀ þ JH½Hþꢀ þ
ð4Þ
JH and JOH are fitting parameters in acidic and alkaline media
for the correction of experimental errors, mainly due to the
liquid junction and to the alkaline and acidic errors of the glass
electrode; Kw = 10−13.75 M2 (10−12.4 M2 at 90 °C) is the autopro-
tolysis constant of water.26 The parameters were calculated by
the non-linear least squares method. The complex formation was
characterized by the following general equilibrium process eqn
((5) and (6)).
βM
H
q
L A
r s
p
pM þ qH þ rL þ sA
M H L A
s
ð5Þ
ð6Þ
ƒƒƒƒƒƒƒƒƒƒ!
½MpHqLrAsꢀ
s
p
q
r
Conclusions
βM H L A
s
¼
The cleaving agents presented in this paper form highly stable
complexes with UpU through coordination of ZnII by the cyclen
moieties of the cleaving agent and the N3 atoms of the uracil
bases. The 1,3,5-triazine core linking the ZnII:cyclen anchors, in
turn, functions as a general acid/base catalyst, shuttling a proton
from the attacking nucleophile to the leaving group. The cataly-
tic effect is independent on the basicity of the triazine ring,
suggesting that deprotonation of the attacking 2′-OH takes place
p
q
r
p
q
r
½Mꢀ ½Hꢀ ½Lꢀ ½Aꢀ
M denotes the metal ion, H the hydrogen ions, L the nonpro-
tonated ligand 2n, while A stands for mUpU (8). Charges are
omitted for simplicity, but can be easily calculated taking into
account the composition of the fully protonated ligands (H5L5+
and H2A). The corresponding formation constants (βMpHqLrAs
βpqrs) were calculated using the PSEQUAD computer program.
≡
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 3328–3338 | 3337

View Article Online
The protonation constants were determined from 3 indepen-
dent titrations (60–80 data points per titration). The complex for-
mation constants were evaluated from 5–10 independent
titrations (60–100 data points per titration), depending on the
complexity of the system. The metal-to-ligand ratios varied
between 1 : 4–1 : 1 and 2 : 1–1 : 2 in the binary ZnII–mUpU and
ZnII–2n systems, respectively, with the ZnII concentration in the
range of 0.7–1.6 mmol L−1. In the ZnII–2n–mUpU ternary
system the concentration ratios were 2 : 1 : 1, 1.5 : 2 : 1
and 1 : 1 : 1, with the ZnII concentration in the range of
Acknowledgements
Financial support from the Academy of Finland, the Hungarian
Scientific Research Found (OTKA K101541) and TÁMOP
4.2.1/B-09/1/KONV-2010-0005 is gratefully acknowledged.
A. J. wishes to thank the support of the János Bolyai research
fellowship.
0.7–1.6 mmol L−1
.
Notes and references
The comparison of the forward and backward titration curves
revealed slow kinetics of complex formation processes in the
ZnII–2n systems between pH 3.5–5. The difference between the
two titration curves decreased with increasing equilibration time,
and with increasing ligand excess (over Zn(II)). To minimize the
errors due to the slow kinetics, considerably increased equili-
bration time has been applied between pH 3.5–5 (the maximal
waiting time between two base addition was 30 min), which
resulted in 4–8 h titration time, depending on the metal-to-ligand
ratio. Under such conditions the difference between the forward
and backward titrations became negligible, especially at ligand
excess
1 S. A. Strobel, Nat. Chem. Biol., 2005, 1, 5; P. C. Bevilacqua and
R. Yajima, Curr. Opin. Chem. Biol., 2006, 10, 455.
2 E. Anslyn and R. Breslow, J. Am. Chem. Soc., 1989, 111, 4473.
3 D. M. Perrault and E. V. Anslyn, Angew. Chem., Int. Ed. Engl., 1997, 36,
432; M. Oivanen, S. Kuusela and H. Lönnberg, Chem. Rev., 1998, 98,
961; R. Breslow, S. D. Dong, Y. Webb and R. Xu, J. Am. Chem. Soc.,
1996, 118, 6588; A. J. Kirby and R. E. Marriott, J. Chem. Soc., Perkin
Trans. 2, 2002, 422.
4 D. A. Usher and A. H. McHale, Proc. Natl. Acad. Sci. USA, 1976, 73,
1149.
5 S. Aoki and E. Kimura, Chem. Rev., 2004, 104, 769; E. Kimura,
T. Shiota, M. Koike, M. Shiro and M. Kodama, J. Am. Chem. Soc., 1990,
112, 5805.
6 Q. Wang and H. Lönnberg, J. Am. Chem. Soc., 2006, 128, 10716.
7 P. B. Rupert and A. R. Ferré-D’Amaré, Nature, 2001, 410, 780;
M. Martick and W. G. Scott, Cell, 2006, 126, 309; J. Han and J.
M. Burke, Biochemistry, 2005, 44, 7864.
8 S. Kuusela and H. Lönnberg, J. Phys. Org. Chem., 1993, 6, 347;
S. Kuusela and H. Lönnberg, J. Phys. Org. Chem., 1992, 5, 803.
9 E. Kimura, S. Aoki, T. Koike and M. Shiro, J. Am. Chem. Soc., 1997,
119, 3068.
10 J. Farràs, X. Ginesta, P. W. Sutton, J. Taltavull, F. Egeler, P. Romea,
F. Urpí and J. Villarasa, Tetrahedron, 2001, 57, 7665.
11 N. Ogasawara and Y. Inoue, J. Am. Chem. Soc., 1976, 98, 7048.
12 C. F. Baes and R. E. Mesmer, The hydrolysis of Cations, John Wiley &
Sons, New York, 1976.
13 P. Davies and K. Wainwright, Inorg. Chim. Acta, 1999, 294, 103.
14 M. Shionoya, E. Kimura and M. Shiro, J. Am. Chem. Soc., 1993, 115,
6730.
15 E. Kimura, M. Kikuchi, H. Kitamura and T. Koike, Chem. Eur. J., 1999,
5, 3113.
16 C. Anda, A. Bencini, E. Berni, S. Ciattina, F. Chuburu, A. Danesi,
C. Giorgi, H. Handel, M. Le Baccon, P. Paoletti, R. Tripier, V. Turcy and
B. Valtancoli, Eur. J. Inorg. Chem., 2005, 2044.
17 R. Haidar, M. Ipek, B. DasGupta, M. Yousaf and L. J. Zompa, Inorg.
Chem., 1997, 36, 3125.
18 P. Järvinen, M. Oivanen and H. Lönnberg, J. Org. Chem., 1991, 56,
5396.
19 T. Tashiro, J. Heterocycl. Chem., 2002, 39, 615.
20 M. Kosonen and H. Lönnberg, J. Chem. Soc., Perkin Trans. 2, 1995,
1203.
21 M. Ora, H. Linjalahti and H. Lönnberg, J. Am. Chem. Soc., 2005, 127,
1826.
22 H. Lönnberg, R. Strömberg and A. Williams, Org. Biomol. Chem., 2004,
2, 2165.
23 H. S. Harned and B. B. Owen, The Physical Chemistry of Electrolytic
Solutions, Reinhold Pub., New York, 1958.
Kinetic measurements
Reactions were carried out in sealed tubes immersed in a thermo-
stated water bath, the temperature of which was adjusted to
90 °C within 0.1 °C. The hydronium ion concentration of the
reaction solutions was adjusted with formate, acetate, MES and
HEPES buffers and sodium hydroxide and checked with a pH
meter. The ionic strength of the solutions was adjusted to
0.10 mol L−1 with NaClO4. The initial substrate concentration in
the kinetic runs was typically 5 μmol L−1. The composition of
the samples withdrawn at appropriate time intervals was ana-
lyzed by HPLC on a Hypersil-Keystone Aquasil C18 column (4
× 150 mm, 5 μm) using 0.06 mol L−1 acetate buffer and MeOH
as an eluent. For the first 10 min, only buffer was used, after
which the amount of MeOH was increased linearly from to 30%
during 5 min and kept there for another 5 min. The observed
retention times (tR, min) for the hydrolytic products of UpU
were as follows: 16.0 (3′,5′-UpU), 15.3 (2′,5′-UpU), 6.5 (Urd),
5.2 (2′-UMP), 4.6 (3′-UMP), 3.8 (2′,3′-cUMP). The products
were characterized by spiking with authentic samples. For the
experiments carried out in the presence of an excess of the cleav-
ing agent, pseudo first-order rate constants for the disappearance
of UpU were obtained by applying the integrated first-order rate
law to the time-dependent diminution of the relative peak area of
UpU. When the reaction rate was studied as a function of UpU
concentration at a constant limiting concentration of the cleaving
agent, initial rates were obtained as slopes of the plots of the
combined peak area of all the monomeric products (2′,3′-cUMP,
3′-UMP, 2′-UMP and uridine) as a function of reaction time.
24 L. Zékány, I. Nagypál and G. Peintler, PSEQUAD for chemical equili-
bria, Technical Software Distributors, Baltimore, 1991
25 F. J. C. Rosotti, H. Rosotti, in The determination of stability constants,
McGraw-Hill Book Co., New York, 1962, pp. 149.
26 E. Högfeldt, in Stability Constants of Metal-Ion Complexes, Part
A. Inorganic Ligands, Pergamon, New York, 1982, pp. 32.
3338 | Dalton Trans., 2012, 41, 3328–3338
This journal is © The Royal Society of Chemistry 2012