Phosphodiester Cleavage of Ribonucleoside Monophosphates and Polyribonucleotides by Homo- and Heterodinuclear Metal Complexes of a Cyclohexane-Based Polyamino-Polyol Ligand

Article abstract of DOI:10.1002/chem.200305149

The ability of the dinuclear complexes of tdci [1,3,5-trideoxy-1,3,5-tris(dimethylamino)-cis-inositol] to promote the cleavage of the phosphodiester bonds of nucleoside 2′,3′-cyclic monophosphates, dinucleoside monophosphates and polyribonucleotides has been studied. The homodinuclear copper(II) and zinc(II) complexes efficiently promote the hydrolysis of cyclic nucleotides. The second-order rate constant (k₂≈0.44M^-1S^-1) estimated for the cleavage of 2′,3′-cAMP induced by dinuclear copper(II) complexes is about 107 times greater than that for the hydroxide-ion-catalysed reaction. The complex selectively cleaves the 2′O-P bond of 2′,3′-cUMP and forms the 3′-product in 91% yield. An equimolar mixture of copper(II), zinc(II) and tdci proved to be more efficient than either of the binary systems: a 7-20-fold rate enhancement was observed for the cleavage of 2′,3′-cNMP substrates. The half-life for the hydrolysis of 2′,3′-cAMP decreased from 300 days to five minutes at 25°C when the concentration of each of the three components was 2.5mM. In contrast to the copper(II) or zinc(II) complexes of tdci, the heterodinuclear species promoted the hydrolysis of several dinucleoside monophosphates. For two ApA isomers, cleavage of the 3′,5′-bond was about 6.5 times faster than cleavage of the 2′,5′-bond. On the basis of the kinetic data, a trifunctional mechanism is suggested for the heterodinuclear-complex-promoted cleavage of the phosphodiester bond. Double Lewis acid activation occurs when the metal ions bind to the phosphate oxygen atoms. In particular, a metal-bound hydroxide ion serves as a general base or a nucleophilic catalyst, and, presumably, a zinc(II)-bound aqua ligand behaves as a general acid and facilitates the departure of the leaving alkoxide group. The effect of the complexes on the hydrolysis of poly(U), poly(A) and type III native RNA was also investigated, and, for the first time, kinetic data on the cleavage of the phosphodiester bonds of polyribonucleotides by a dinuclear complex was obtained.

Full text of DOI:10.1002/chem.200305149

FULL PAPER
Phosphodiester Cleavage of Ribonucleoside Monophosphates and
Polyribonucleotides by Homo- and Heterodinuclear Metal Complexes of a
Cyclohexane-Based Polyamino Polyol Ligand
Attila Jancso¬,*[a] Satu Mikkola,[a] Harri Lo»nnberg,[a] Kaspar Hegetschweiler,[b] and
Tama¬s Gajda[c]
Abstract: The ability of the dinuclear
complexes of tdci [1,3,5-trideoxy-1,3,5-
tris(dimethylamino)-cis-inositol] to pro-
mote the cleavage of the phosphodiester
bonds of nucleoside 2',3'-cyclic mono-
phosphates, dinucleoside monophos-
phates and polyribonucleotides has been
studied. The homodinuclear copper(ii)
and zinc(ii) complexes efficiently pro-
mote the hydrolysis of cyclic nucleo-
tides. The second-order rate constant
(k2 ꢀ 0.44 mÀ1 sÀ1) estimated for the
cleavage of 2',3'-cAMP induced by di-
nuclear copper(ii) complexes is about
107 times greater than that for the
hydroxide-ion-catalysed reaction. The
of the binary systems: a 7 20-fold rate
enhancement was observed for the
cleavage of 2',3'-cNMP substrates. The
half-life for the hydrolysis of 2',3'-cAMP
decreased from 300 days to five minutes
at 258C when the concentration of each
of the three components was 2.5mm. In
contrast to the copper(ii) or zinc(ii)
complexes of tdci, the heterodinuclear
species promoted the hydrolysis of sev-
eral dinucleoside monophosphates. For
two ApA isomers, cleavage of the 3',5'-
bond was about 6.5 times faster than
cleavage of the 2',5'-bond. On the basis
of the kinetic data, a trifunctional mech-
anism is suggested for the heterodinu-
clear-complex-promoted cleavage of the
phosphodiester bond. Double Lewis
acid activation occurs when the metal
ions bind to the phosphate oxygen
atoms. In particular, a metal-bound
hydroxide ion serves as a general base
or a nucleophilic catalyst, and, presum-
ably, a zinc(ii)-bound aqua ligand be-
haves as a general acid and facilitates the
departure of the leaving alkoxide group.
The effect of the complexes on the
hydrolysis of poly(U), poly(A) and
type III native RNA was also investigat-
ed, and, for the first time, kinetic data on
the cleavage of the phosphodiester
bonds of polyribonucleotides by a dinu-
clear complex was obtained.
À
complex selectively cleaves the 2'O P
Keywords:
heterometallic com-
bond of 2',3'-cUMP and forms the 3'-
product in 91% yield. An equimolar
mixture of copper(ii), zinc(ii) and tdci
proved to be more efficient than either
plexes ¥ kinetics ¥ phosphodiester
cleavage ¥ phosphoesterase models
¥ RNA
such processes are catalysed by phophoesterase enzymes,
which contain, in many cases, two or three metal ions in their
active sites.[2] The use of dinuclear (or trinuclear) metal
complexes of low molecular weight ligands as mimics for these
centres has been the focus of bioinorganic/bioorganic re-
search for more than a decade.[3] Numerous homodinuclear
species have been proposed as structural or functional models
of the bimetallic centres of phosphoesterases,[4, 5] and in many
cases, their solid phase and/or solution structure has been
determined.[5] A spacer group or a bridging functional group
(e.g., alkoxide, phenoxide, phthalazine, etc.) that separates
the metal-binding sites in the molecule, is the key structural
motif of binucleating ligands. It has been observed that proper
positioning of the metal ions and a certain flexibility around
the metal centres are both required for efficient acceleration
of the hydrolytic processes. The activity of the dinuclear
species is also highly dependent upon the metal metal
distance.[6, 7] In a recent report, in which the dinuclear zinc(ii)
complex efficiently promoted the hydrolytic cleavage of
Introduction
The cleavage of biomolecular phosphodiester bonds (e.g.,
RNA, DNA, ATP, etc.) plays a crucial role in various
fundamental biochemical processes.[1] In living organisms,
¬
[a] Dr. A. Jancso, Dr. S. Mikkola, Prof. H. Lo»nnberg
Department of Chemistry, University of Turku
Vatselankatu 2, 20014 Turku (Finland)
Fax : (‡358)2-333-6700
E-mail: attila.jancso@utu.fi
jancso@chem.u-szeged.hu
[b] Prof. K. Hegetschweiler
Anorganische Chemie
»
»
Universitat des Saarlandes Saarbrucken
Postfach 15 11 50, 66041 (Germany)
[c] Dr. T. Gajda
Department of Inorganic and Analytical Chemistry University of Szeged
P.O. Box 440, 6701 Szeged (Hungary)
Supporting information for this article is available on the WWW under
http://www.chemeurj.org or from the author.
5404
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200305149
Chem. Eur. J. 2003, 9, 5404 5415

5404 5415
plasmid DNA, the precise distance between the metal centres
was provided by the pitch of the tailored heptapeptide
helix.[7]. However, questions that concern the optimal struc-
tural properties of a ligand, including the ideal donor sets of a
metal binding site, still have to be answered.
Copper(ii) and/or zinc(ii) complexes of several triamine
ligands, amongst them cis-1,3,5-triaminocyclohexane and its
congener 1,3,5-trideoxy-1,3,5-triamino-cis-inositol, have been
found to promote the hydrolysis of both activated[18a,c] [2,4-
dinitrophenyldiethyl phosphate (DNDEP)] and unactivated
(DNA) phosphoesters.[18b,c] In a recent publication, we re-
ported a kinetic and equilibrium study, as well as the structure
determination in solution, of the copper(ii) complexes of a
polyamino polyol ligand tdci (1,3,5-trideoxy-1,3,5-tris(di-
methylamino)-cis-inositol).[19] Due to the unique steric ori-
entation of the three amino and three hydroxyl groups, tdci
can bind up to three metal ions in aqueous solutions. These
studies indicated that a complicated equilibrium exists for the
complexes, since differently protonated mono-, di- and
trinuclear species are formed depending on the pH and
metal-to-ligand ratio. The dinuclear complex Cu2LHÀ3
showed outstanding catalytic activity towards the activated
phosphoester BNPP, the hydrolysis obeying Michaelis
Menten kinetics. As mentioned above, catalysis for the
cleavage of activated esters and unactivated nucleoside
phosphoesters may be different. Consequently, a species that
is able to cleave nitrophenyl phosphates is not necessarily a
good catalyst for the cleavage of unactivated esters,[8g] and
vice versa.[20]In the present paper, we report on the catalytic
activity of dinuclear copper(ii) complexes of tdci (1) for
the cleavage of two cyclic nucleoside monophosphates
(2',3'-cUMP and 2',3'-cAMP; 2) and a dinucleoside mono-
phosphate (∫3',5'-UpU, 3a). For comparison, kinetic studies
on the binary tdci zinc(ii) system have also been per-
The advantage of having two metal ions in close proximity
has been clearly proven by comparing the activity of
bimetallic systems to their mononuclear counterparts.[8]
Double Lewis acid activation of the bound phosphoester
substrates provided by the two metal ions has been re-
garded as an important factor in reaction rate enhance-
ment.[4c e, 8a, 9, 10] Nevertheless, a general base or a nucleophilic
catalyst is substantial for the cleavage of phosphoester bonds,
as was shown by the comparison of the catalytic activity of
successively formed alkoxide-bridged dinuclear zinc(ii) com-
plexes of an asymmetric ligand.[10c] The central hydroxy group
and the zinc(ii)-bound water molecule were deprotonated in
well-defined steps; this allowed the authors to unambiguously
identify the hydroxo/mixed-ligand dinuclear complex as the
catalytically active species. In general, a metal-bound hy-
droxide can assist cleavage either by acting as a general
base,[4d,e, 8f,g, 10] or as a direct nucleophilic catalyst.[4d, 6a, 7, 8c] This
has also been suggested to be the case with the hydrolytic
reactions of nucleoside phosphoesters promoted by uncom-
plexed metal aqua ions.[11] While the hydrolytic transforma-
tions of activated esters [e.g., (4-nitrophenyl)phosphate (NPP),
bis(4-nitrophenyl)phosphate (BNPP), 2,4-dinitrophenylethyl
phosphate (DNEP) and 2-hydroxypropyl-4-nitrophenyl phos-
phate (HPNP)] do not require stabilisation of the leaving
group, stabilisation of the transition state of the reaction by
the dinuclear core can strongly enhance the catalytic activi-
ty.[8f] On the other hand, the departure of an alkoxide leaving
group from an unactivated nucleoside phosphoester is most
likely a catalysed process. It has been suggested that metal-
bound water molecules in aqua ions or in mononuclear com-
plexes, metal aqua ions or serve as general acid catalysts that
enhance the cleavage of the dinucleoside monophosphate
5'O P bond.[11a] Indeed, a kinetic study conducted on a series
À
of uridine 3'-alkyl and aryl phosphates proved that this was
the case.[12] It has also been proposed that dinuclear copper(ii)
complexes catalyse the release of the leaving group.[8e, 10b]
The spectacular catalytic activity of dinuclear (or trinu-
clear) complexes is undoubtedly based on the cooperative
behaviour of the metal centres during their catalytic cycles.
Therefore, the use of two metal ions that have different
characteristics (e.g., size, Lewis acidity, favoured coordination
geometry, etc.) may be even more advantageous. A number of
hydrolytic metalloenzymes clearly display metal-ion cooper-
ation.[2a] Amongst them, purple-acid phosphatases (PAP)
have been extensively studied, and the crystal structures of
several different types of PAPs have been determined (e.g. red
kidney bean (BBPAP),[13a] and rat (TRAP)[13b]). In turn, the
above results have instigated model studies with heterodinu-
clear FeIII and ZnII,[10c, 14] FeIII and MnIII,[15] or FeIII and NiII[16]
complexes aimed at mimicking the structure or function of
native enzymes. Other groups have demonstrated that
dinucleoside monophosphate cleavage is enhanced when the
reaction is mediated by heterodinuclear[17a,b] or heterotrinu-
clear[17c] complexes of different ligands.
formed. As an extension of the studies conducted on
nucleotide monomers and dimers, the hydrolysis of homo-
polynucleotides [poly(U) and poly(A)] and native type III
RNA was also investigated. To the best of our knowledge,
kinetic data on the cleavage of polyribonucleotides by
dinuclear metal complexes have not yet been published.
Chem. Eur. J. 2003, 9, 5404 5415
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5405

¬
A. Jancso et al.
FULL PAPER
Besides binary systems that contain the ligand and one type
of metal ion, solutions of tdci and two different metal ions,
copper(ii) and zinc(ii) (ternary system), have also been
investigated. The effect of the ternary system on the
hydrolysis of four cyclic ribonucleoside monophosphates 2,
several dinucleoside mono- and diphosphates, 3a,b and 4,
respectively, as well as polyribonucleotides, has been studied
in detail. For a comparison the results of kinetic measure-
ments for the activated ester BNPP 5 are also reported. The
rates of the reactions were followed as a function of pH,
metal-to-ligand ratio, copper(ii)-to-zinc(ii) ratio and catalyst
concentration. The aim of replacing one of the copper(ii) ions
by zinc(ii) was to provide a more flexible coordination
environment with similar Lewis acid properties. The combi-
nation of the two metal ions may facilitate the binding of the
catalyst to the substrate, generate an attacking nucleophilic
reactant and promote departure of the leaving group. Hence,
markedly enhanced cleavage of the phosphoester bond is
expected.
Figure 1. Visible spectra measured for the tdci CuII (A) and tdci CuII
ZnII (B) systems in which the metal-to-ligand ratio for A is 2.98:1 0.35:1,
while the ZnII/CuII ratio is 0:2 2:1 for B. (A, [tdci] ˆ 1.35  10À3 7.49 Â
10À3 m, [CuII] ˆ 4.02  10À3 2.65  10À3 m; B, 2[tdci] ˆ [CuII] ‡ [ZnII] ˆ
4.00 Â 10À3 m).
consistent with transformations of the trinuclear complexes
into di- and mononuclear species. Figure 2A shows the species
distribution (% CuII) as a function of the copper(ii)/tdci ratio
at pH 8.6.[19] Matrix rank analysis (MRA), according to
Peintler et al.,[21] was performed on the spectral data, and
the results are summarised in Table 1. Standard deviations at
the corresponding wavelength (rows) and copper(ii)/tdci ratio
values (columns) are also given. In general, a diagonal
element can be considered to deviate from zero when it is
significantly larger than the reproducibility of the spectra (i.e.,
the accuracy of the equipment), and if it is at least twice as
large than the standard deviation. However, experimental
accuracy and the starting values used for the error propaga-
tion calculations remarkably influence the number of ele-
ments detected, especially when the minor components are
concerned. Calculation of the residual absorbance curves
(RAC) was used to verify the existence of ambiguous species,
as these curves represent features of the spectra that cannot
be explained by the presence of the (n À 1)th species. The
simple MRA calculation clearly showed the presence of four
linearly independent complexes. This was consistent with the
species distribution studies, and was also supported by the
wavelength-dependent RAC curves (Figure S1A in the Sup-
porting Information). The rows and columns of the non-zero
diagonal elements are generally very informative with respect
to the absorbing species. Unfortunately, in the present case,
the individual spectra are relatively similar, which signifi-
cantly reduces the information that can be obtained from such
Results and Discussion
Spectrophotometric measurements and matrix rank analysis
(MRA) of the absorbance matrices: The composition and
stability of dinuclear copper(ii) complexes of tdci 1 are well
established,[19] but data on the system that contains both a
copper(ii) and a zinc(ii) ion are scarce. The dinuclear
copprer(II) complexes of tdci have been characterised by
means of EPR measurements.[19] Partial replacement of
copper(ii) ions by zinc(ii) ions in an equimolar tdci CuII
solution at pH 11 resulted in a significant increase in the
intensity of the EPR signal relative to that obtained for the
dinuclear copper(ii) complexes. This indicated the presence of
a mixed dinuclear complex that contained one copper(ii) and
one zinc(ii) ion at pH 11. pH-Metric studies at lower pH would
not show conclusively that a heterodinuclear species had been
formed, due to the complexity of the species distribution
equilibria. To elucidate the possible formation of a hetero-
dinuclear complex, spectrophotometric measurements on the
tdci copper(ii) system, in the absence (Figure 1A) and
presence (Figure 1B) of zinc(ii), were performed. A continu-
ous blue shift of the d d transition band from 722 to 654 nm
was observed on decreasing the copper(ii)-to-tdci ratio; this is
5406
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5404 5415

Phosphodiester Cleavage
5404 5415
ratio is greater than about 1.00 1.25 (lmax ˆ 671 nm), the
shape of the spectrum remains unchanged. If addition of the
zinc(ii) tdci solution did not have any influence on the
copper(ii) tdci complex speciation, a simple dilution effect,
that is, a decrease in intensity without other spectral changes,
should be observed. This is clearly not the case. Indeed,
spectral changes are observed that cannot be explained by a
shift of equilibrium from tri- and dinuclear complexes to
mononuclear species. Such processes would result in the
occurence of free zinc(ii) and consequently a zinc(ii) hydrox-
ide precipitate would be formed. As it did not happen, a
plausible explanation for the blue shift is that the homodinu-
clear species are transformed to give heterodinuclear com-
plexes. MRAwas performed on the latter spectral data set and
on the matrices of both experiments. In drawing conclusions
on the basis of the data evaluated, it should be kept in mind
that: 1) if heterodinuclear complexes are only formed to a
small extent, they may only have a slight effect on the spectra
and 2) if the presence of zinc(ii) causes only a slight
disturbance in the coordination environment of the copper(ii)
ion(s) in the heterodinuclear species, the number of non-zero
diagonal elements may not increase when the two sets of
experiments are evaluated together. The simple MRA of the
spectra recorded for the solution of zinc(ii) and copper(ii) in
different ratios shows the existence of at least two, but not
more than three, linearly independent absorbing species
(Table 1). The first diagonal element, which was chosen from
the spectrum that did not con-
Figure 2. CuII (A) and ligand (B) distribution at pH 8.60 for the tdci CuII
binary system as a function of the metal-to-ligand ratio ([tdci] ˆ 0.002m,
stability constants from reference [19] were used for the calculations).
tain zinc(ii), corresponds to the
Table 1. Diagonal elements from MRA calculations[a] performed on the spectral data matrices of the tdci CuII
(pH ˆ 8.63), tdci CuII ZnII (pH ˆ 8.67) and tdci CuII/tdci CuII ZnII systems (evaluation range (l): 450
900 nm).
same complex as the first ele-
ment of the MRA of the binary
system. On the other hand, the
second diagonal element, which
was chosen from the spectrum
that contained the two metal
ions in an equimolar ratio,
probably corresponds to a het-
erodinuclear complex. The
wavelength-dependent residual
spectra suggest that three ab-
sorbing species exist in solution
(Figure S1B in the Supporting
Information). In fact, the fourth
residual curve (number of in-
dependent absorbing species
(NIAS) ˆ 3) depicted a non-
random shape in certain wave-
length ranges, but these values
Diagonal element !
1
2
3
4
5
6
tdci CuII
0.2016
0.0020
0.0400
0.0023
À 0.0187
0.0038
629
0.0164
0.0058
À 0.0044
0.0041
897
À 0.0035
0.0073
892
s[b]
l [nm][c]
695
1.76
553
0.35
450
2.50
À 0.0036
0.0051
755
R[d]
2.98
0.70
tdci CuII ZnII
0.2022
0.0020
0.0143
0.0030
À 0.0068
0.0037
450
0.0029
0.0078
864
À 0.0023
s[b]
0.0067
l [nm][c]
R [d]
709
642843
0.00
0.2022
0.0020
1.00
0.0418
0.0023
0.25
2.00
0.12
0.37
À 0.0048
0.0075
658
tdci CuII/tdci CuII ZnII
À 0.0205
0.0029
450
0.0177
0.0041
647
0.0075
0.0042
874
s[b]
l [nm][c]
R(s)[d]
709
0[f]
558
0.35[e]
2.98[e]
1.25[f]
1.53[e]
2.74[e]
[a] The elements considered to be non-zero and their standard deviations are marked with italics. [b] Standard
deviation of the given element. [c] Wavelength of the chosen element. [d] CuII/tdci (tdci CuII system) or ZnII/CuII
(tdci CuII ZnII system) ratio belonging to the column from which the element was chosen. [e] tdci CuII system.
[f] tdci CuII ZnII system.
data. However, the columns of the original matrix, from which
the first three diagonal elements are chosen, are characteristic
for the three main complexes present in solution at pH 8.6.
In a parallel experiment, a solution of tdci and zinc(ii) in a
1:2ratio, was gradually added to a similar solution of tdci and
copper(ii) at pH 8.6 (Figure 1B). In this way, the concentration
of the ligand and the total metal-ion/ligand ratio remained
constant, while the copper(ii) ions were gradually replaced
with zinc(ii). As seen in Figure 1B, a blue shift similar to the
one observed for the tdci copper(ii) binary system occurred,
but the change was not as great. After the zinc(ii)-to-copper(ii)
are within the experimental error of the measurements. Then
the two data sets were evaluated together, and, on this basis,
four or five absorbing species were suggested (Table 1).
Calculated residual curves indicated the presence of a fifth
absorbing species (Figure S1C in the Supporting Informa-
tion), as the residuals obtained on the assumption that four
species (NIAS ˆ 4) exist, contained non-random variations
larger than the experimental error of the data. Although the
similarities of the individual spectra do not allow a detailed
analysis, the increased number of non-zero diagonal elements,
as well as all the other data available about the ternary system
Chem. Eur. J. 2003, 9, 5404 5415
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5407

¬
A. Jancso et al.
FULL PAPER
(EPR, kinetics–see below), strongly suggests the presence of
heterodinuclear complexes.
Hydrolysis of nucleoside 2',3'-cyclic monophosphates in the
tdci copper(ii) system: According to formerly published
potentiometric results on the binary system,[19] several differ-
ently protonated mono-, di- and trinuclear species exist in a
complicated equilibrium that is dependent both on pH and
the metal-to-ligand ratio. The main species formed between
pH 6 and 10 are depicted in Scheme 1. Detailed kinetic
Figure 4. Pseudo-first-order rate constants and regioselectivity for the
hydrolysis of 2',3'-cAMP at pH 8.45 as a function of the CuII tdci ratio
(Tˆ 258C, [tdci] ˆ 1.85  10À3 m, [2',3'-cAMP] ꢁ 1  10À4 m, Selectivity: [2'-
AMP]/[3'-AMP]).
curve occurred when there was a twofold excess of metal to
ligand. Remarkable catalytic activity was also observed at the
equimolar ligand-to-metal ion ratio[19]; this can be attributed
to the presence of the dinuclear species even in the equimolar
solution (see Figure 2). At higher metal ion to ligand ratios,
trinuclear complexes are formed, and these become the
predominant species at [CuII]:[tdci] >2.3 (Figure 2B). These
trinuclear complexes also enhance the hydrolytic process, but
their catalytic activity is lower than that of the dinuclear
species; the observed rate constants decrease as the metal-
ion-to-ligand ratio increases (Figure 4).
Scheme 1. The main species present in the tdci CuII binary system
according to reference [19] for the pH range marked.
measurements for the hydrolysis of BNPP (5) indicated
outstanding activity of the dinuclear Cu2LHÀ3 species. On the
basis of these results, kinetic experiments for two ribonucleo-
side 2',3'-cyclic monophosphates (2',3'-cUMP, 2',3'-cAMP; 2)
have been performed. Pseudo-first-order rate constants for
the hydrolysis of the two substrates in the presence of a
twofold metal excess were determined as a function of pH
(Figure 3). The observed bell-shaped pH rate profiles and
Although the shapes of the kobs versus pH plots are similar
for both 2',3'-cUMP and 2',3'-cAMP, the reactions of 2',3'-
cAMP are seven times faster than those of 2',3'-cUMP
(Figure 3). Nevertheless, the active complexes significantly
enhance the reaction rates for both compounds. In the
absence of metal complexes, hydroxide-ion catalysis contrib-
utes to the observed rate constant, kuncat, but hydronium-ion
c]
catalysis is negligible {kuncat ˆ kH O ‡ (kOH [OHÀ])}.[22a At
2
258C, data is available for the uncatalysed reaction of 2',3'-
cUMP[22a,d] (kuncat(pH ˆ 8.57) ꢁ 1.3  10À8 sÀ1), but on the basis
of the data obtained at different temperatures,[22b d] a slightly
higher value may be estimated for 2',3'-cAMP at the 258C
(kuncat(pH ˆ 8.57) ꢁ 1.9  10À8 sÀ1). The maximum observed
kobs values at 2.5mm ligand concentration represent approx-
imately a 5000- and 25000-fold rate enhancement for the
hydrolysis of 2',3'-cUMP and 2',3'-cAMP, respectively. The
proportion of active Cu2LHÀ3 species present can be calcu-
lated from the stability constants,[19] and at the pH of the
maximum kobs values (Table 2) was calculated to be 1.1mm.
The second-order rate constants that can be estimated from
Figure 3. pH Rate profiles for 2',3'-cAMP and 2',3'-cUMP hydrolysis in
the tdci CuII 1:2system ( Tˆ 258C, [2',3'-cAMP] and [2',3'-cUMP] ꢁ 1 Â
10À4 m, 2[tdci] ˆ [CuII] ˆ 5.0  10À3 m).
the data at pH 8.57 (k2,cUMP ꢁ 0.061mÀ1 sÀ1, k2,cAMP
ꢁ
0.44 mÀ1 sÀ1)[22c,d] are approximately 23 and 107 times higher
than the rate constants of the hydroxide-ion-catalysed reac-
tions of 2',3'-cUMP and 2',3'-cAMP, respectively.
the pH maxima of the curves (pHmax ꢁ 8.5) are very similar to
those found earlier for the activated esters.[19] Since the higher
total concentrations used in the present study have only a
slight effect on the speciation the observed kinetic activity can
also be attributed to the above mentioned Cu2LHÀ3 complex.
To further support this assignment, the hydrolysis of 2',3'-
cAMP was also studied as a function of the ligand-to-metal
ratio at pH 8.5 (Figure 4). The maximum of the resultant
À
Protein nucleases selectively cleave the 2'O P bond of
cyclic nucleoside monophosphates, but base-catalysed reac-
tions produce comparable amounts of two possible products,
the 3'-O-phosphorylated compound generally being slightly
favoured.[22c, 23] Only a few metal complexes that have been
studied as nuclease mimics are able to hydrolyse nucleoside
5408
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5404 5415

Phosphodiester Cleavage
5404 5415
Table 2. Observed rate constants[a] and product distribution for the
hydrolysis of nucleoside 2',3'-cyclic monophosphates at Tˆ 258C ([sub-
strate] ˆ 5 10  10À5 m).
hydroxide is involved in the nucleophilic attack of the
phosphorous, either as a general base or as a nucleophile. A
metal-bound alkoxide group from the ligand could be the
nucleophilic reactant, but this would result in the phosphor-
ylation of the ligand and subsequent formation of a ligand
nucleotide adduct: this was not observed by high-perform-
ance liquid chromatography (HPLC). The second-order rate
constants estimated for the dinuclear-complex-promoted
cleavage of the cyclic nucleotides are much higher than those
for the hydroxide-ion-catalysed reactions, in spite of the fact
that the nucleophilicity of a metal-bound hydroxide ion is
undoubtedly lower than that of a free hydroxide ion. These
findings suggest that direct intracomplex nucleophilic attack
on the phosphorous atom by a hydroxo ligand takes place
during the cleavage of the cyclic phosphodiesters. Further-
more, the reaction is probably enhanced by double Lewis acid
activation induced by the phosphate-coordinated metal
centres. The proposed mechanism depicted in Scheme 2
explains the experimental observations, apart from the
regioselectivity.
System
Substrate
kobs Â105 [sÀ1
]
Selectivity[b]
tdci CuII 1:2 [c]
2',3'-cUMP
2',3'-cAMP
2',3'-cUMP
2',3'-cAMP
2',3'-cUMP
2',3'-cAMP
2',3'-cCMP
2',3'-cGMP
6.6 Æ 0.1
48 Æ 1
0.10 Æ 0.01
0.19 Æ 0.01
0.51 Æ 0.01
0.58 Æ 0.02
0.66 Æ 0.03
0.33 Æ 0.03
1.10 Æ 0.02
Æ 0.01
tdci ZnII 1:2 [d]
2.3 Æ 0.1
5.2 Æ 0.1
47 Æ 1
tdci CuII ZnII 1:1:1[e]
230 Æ 5
62 Æ 1
87 Æ 20.47
[a] The errors given refer to the standard deviaton of the kobs values
determined from the slope of the curves obtained after applying the
integrated first-order rate law for the data points of each kinetic run. The
values are the average of duplicate measurements with a reproducibility
ꢂ15%. [b] Ratio 2'-NMP/3'-NMP. [c] [tdci] ˆ 2.5  10À3 m, pH ˆ 8.57.
[d] [tdci] ˆ 2.5  10À3 m, pH ˆ 8.26, [e] [tdci] ˆ 2.5  10À3 m, pH ˆ 8.75.
cyclic monophosphates regioselectively.[8e, 24] The dashed
curve in Figure 4 represents the ratio of the two products,
2'-AMP and 3'-AMP, as a function of the metal-to-ligand ratio
for the tdci copper(ii) system at pH 8.50. The reaction
Hydrolysis of the cyclic nucleotides was found to be
significantly different than hydrolysis of the dinucleoside
monophosphate 3',5'-UpU; cleavage of the former com-
pounds being much more efficient. Hydrolytic cleavage of
dinucleoside monophosphates is a complicated process, and,
according to the results discussed above, catalysis by the
homodinuclear copper(ii) complex of tdci is not enough to
efficiently cleave the phosphodiester bond of 3',5'-UpU.
À
À
exhibits a clear preference for 2'O P over 3'O P bond
cleavage (see Table 2). Product distribution varies only
slightly up to a CuII:tdci ratio of about two, and suggests that
the observed activity and regioselectivity can be attributed
mostly to one species (Cu2LHÀ3).[19] Further increases of the
metal-ion excess results in a decrease in the selectivity
(Figure 4). Although the observed selectivity is lower than
À
for the dinuclear complex, cleavage of the 2'O P bond is still
preferred. The difference in selectivity between the cAMP
cleavage catalysed by di- and trinuclear complexes may arise
from differences in the coordination of the substrate to the
catalysts.
À
Cleavage of the 2'O P bond is even more favoured for 2',3'-
cUMP than for 2',3'-cAMP (see Table 2), as the mole fraction
of 2'-UMP is only 0.091. In the past, interactions (hydrogen
bonding or stacking) between the base moiety of the substrate
and the complexed ligand were believed to account for the
observed selectivities,[8e] because there did not appear to be
any clear correlation between the regioselectivity and the
base sizes and the known metal-ion affinities[25] for the bases.
Hydrogen-bond formation, which would promote a certain
orientation of the substrate in the substrate catalyst adduct,
is also possible in the present system; however, there is no
proof to support this suggestion. Such interactions may also
play a role in the observed base selectivity (see Figure 3 and
Table 2), although this may originate from metal base
interactions (see below).
Hydrolysis of 3',5'-UpU (3a) was also studied for the tdci
copper(ii) binary system. The reaction in the presence of
2.5mm tdci and 5mm CuII at pH 8.57 and 358C turned out to
be very slow, since less than 5% of the substrate was cleaved
in 15 days (kobs ꢂ 4 Â 10À8 sÀ1). At a higher pH (9.5), the
reaction was somewhat faster, but the observed rate constant
(kobs ꢁ 2 Â 10À7 sÀ1) was only about four times greater than
that observed for the uncatalysed cleavage.[26]
Scheme 2. Proposed mechanism for the hydrolysis of 2',3'-cNMP promot-
ed by the Cu2LHÀ3 complex.
Kinetic studies with the tdci zinc(ii) system: In light of the
solution equilibrium properties and catalytic behaviour of the
tdci copper(ii) system, kinetic studies of the tdci zinc(ii)
system were mainly focused on solutions that contained tdci
and zinc(ii) in a 1:2ratio. Hydrolysis of 2 ',3'-cUMP 2 was
monitored as a function of pH. A bell-shaped pH rate profile
was obtained with a maximum at around pH 8.25. At a ligand
concentration of 2.5mm, the maximum first-order rate con-
stant is about 2600 times greater than for the uncatalysed
reaction. The rate acceleration is slightly higher (ca. 4600) for
2',3'-cAMP 2 under the same conditions (see Table 2).
Cleavage of 3',5'-UpU 3a was also studied, but similarly to
the copper(ii) complexes, only a modest rate acceleration was
observed.
The shape of the pH rate profiles and the pH at the
maximum rate acceleration suggests that a metal-bound
Hydrolysis of nucleoside 2',3'-cyclic monophosphates in the
tdci copper(ii) zinc(ii) ternary system: Since efficient hy-
Chem. Eur. J. 2003, 9, 5404 5415
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5409

¬
A. Jancso et al.
FULL PAPER
drolysis was observed in the binary system when the
copper(ii)-to-tdci ratio was 2:1, the first experiments were
performed in solutions that contained the ligand, copper(ii)
and zinc(ii) ions in a ratio of 1:1:1. Several salts were tested to
adjust the ionic strength, but they caused precipitate forma-
tion at about pH 8. Therefore, the constant ionic strength
(0.05m) was adjusted by the buffers (HEPES, CHES). All the
four natural ribonucleoside 2',3'-cyclic monophosphates 2
were studied, and the catalytic properties of the ternary
system were investigated in detail with 2',3'-cUMP as the
substrate. The rate constants and product distributions are
summarised in Table 2.
The pH dependence of the kobs values for the hydrolysis of
2',3'-cUMP is depicted in Figure 5. The rate profile is sym-
metrically bell-shaped and has a maximum around pH 8.7.
The rate constants are approximately seven and 20 times
Figure 6. Pseudo-first-order rate constants at pH 8.75 for the hydrolysis of
2',3'-cUMP as a function of the CuII total metal-ion ratio (Tˆ 258C, 2 Â
[tdci] ˆ [CuII]‡[ZnII] ˆ 5.00  10À3 m, [cUMP] ꢁ 1  10À4 m).
agreement with those determined for the binary systems. The
maximum activity displayed at a CuII/ZnII ratio of 1:1 suggests
that a heterodinuclear complex is responsible for the cleavage
of the cyclic phosphodiester. As trimetallic species were
detected in the copper(ii) tdci binary system, the hydrolysis
rate for 2',3'-cUMP was also determined as a function of the
metal-to-ligand ratio, Figure S2A and Figure S2B in the
Supporting Information). The kinetic maximum was found at
a ratio of 2:1 ([CuII] ˆ [ZnII] ˆ [tdci] ˆ 2.5 mm), though nota-
ble cleavage was also observed at a 3:1 ratio, especially for the
solution that contained a twofold excess of zinc(ii) over tdci
and copper(ii) (Figure S2A in the Supporting Information). A
trinuclear species must be responsible for this activity, since
the presence of dinuclear complexes would also require the
presence of free metal ions, which in turn would result in the
precipitation of a metal hydroxide under the experimental
conditions applied. Kinetic measurements for solutions that
contained a threefold excess of the metal ions were also
obtained as a function of pH. The highest kobs values occurred
around pH 8.5 when the CuII/ZnII ratio was 1:2and between
pH 7.5 8.5 for a 2:1 CuII/ZnII ratio (data not shown). The
notable shift in pH relative to the maximum pH observed in
the equimolar solution of the three components also indicates
that trinuclear complexes possess catalytic activity.
Cleavage of 2',3'-cUMP to give phosphomonoesters was
also monitored as a function of the total concentration of the
ligand and the two metal ions. The concentration of the three
components was varied from 7.8 Â 10À5 2 .4Â 10À3 m, while
the tdci, copper(ii) and zinc(ii) ratio was kept at 1:1:1. As
depicted in Figure 7, the observed rate constants are linearly
related to the concentration of the ligand and metal ions, and
the intercept occurs close to the origin. Moreover, a straight
line with a slope of 1.12is obtained when a logarithmic plot is
used; this means that the reaction is first-order with respect to
the complex concentration. Since the mole fraction of the
active complex in the ternary system is not known, only a
lower limit for the second-order rate constant may be
estimated based on the assumption that the active complex
is the only species present. In this manner, a value of
0.19 mÀ1 sÀ1 was obtained; this is more than 70 times greater
than the second-order rate constant for the hydroxide-ion-
catalysed reaction. This again suggests a mulitfunctional
Figure 5. pH rate profile for the hydrolysis of 2',3'-cUMP in the tdci
CuII ZnII 1:1:1 system (Tˆ 258C, [2',3'-cUMP] ꢁ 1  10À4 m, [tdci] ˆ
[CuII] ˆ [ZnII] ˆ 2.5  10À3 m).
larger than the values determined for the copper(ii) tdci and
zinc(ii)-tdci binary systems, respectively (see Table 2). This
indicates that the presence of two different transition-metal
ions has a favourable effect on catalysis and that the increased
activity very likely results from the formation of heterodinu-
clear complexes. As described above, the existence of such
species is further supported by spectrophotometric and EPR
experimental data.
Several other investigations of the catalytic properties of
the heterodinuclear complexes were performed in order to
determine the ideal metal-to-ligand and copper(ii)-to-zinc(ii)
ratio for the cleavage of 2',3'-cUMP. Figure 6 shows the result
of an experiment that applied Job×s method at the pH
maximum of the pH rate profile. The concentration of the
ligand was kept at 2.5mm, while the concentration of the two
metal ions was varied between 0.0 5.0mm, maintaining a
constant 2:1 total metal-to-ligand ratio (and a constant 5.0 mm
metal ion concentration); meanwhile the ratio of the two
metal ions themselves was varied from 0:2to 2:0. The pseudo-
first-order rate constants as a function of [CuII]:([CuII] ‡
[ZnII]) ratios afforded a nearly symmetric bell-shaped curve
that has a maximum when the components are equimolar. The
end points of the curve represent the kinetic activity of the
homodinuclear complexes at the same pH and are in a good
5410
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5404 5415

Phosphodiester Cleavage
5404 5415
Table 3. Observed rate constants[a] for the hydrolysis of dinucleoside
mono- and diphosphates (Tˆ 358C) and polyribonucleotides (Tˆ 608C) in
the tdci CuII ZnII 1:1:1 system at pH ˆ 9.0: [substrate] ꢁ 1  10À4 m,
[tdci] ˆ 2.5  10À3
m
(dinucleoside mono- and diphosphates) and 1.0 Â
10À3 m (polynucleotides).
[b]
Substrate
kobs Â106 [sÀ1
]
kobs,uncat Â106 [sÀ1
]
3',5'-ApA
2',5'-ApA
3',5'-UpU
2',5'-UpU
3',5'-UpA
3',5'-ApU
3',5'-ApU-3'p
15 Æ 1
0.0053
0.016
0.017
0.025
0.018
2.3 Æ 0.1
4.1 Æ 0.1
5.0 Æ 0.1
11 Æ 1
6.6 Æ 0.1
0.0056
[c]
0.56 Æ 0.020.0042
[d]
poly(U)
poly(A)
RNA
2.1 Æ 1
12 Æ 1
8 [e]
0.57
0.18
Figure 7. Pseudo-first-order rate constant at pH 8.75 for the hydrolysis of
2',3'-cUMP as a function of the total concentration of tdci, CuII, and ZnII
(Tˆ 258C, tdci:CuII:ZnII ˆ 1:1:1, [2',3'-cUMP] ꢁ 1  10À4 m. The value de-
termined from the slope of the curve refers to the lower limit of the second-
order rate constant).
[d]
[a] The errors given refer to the standard deviaton of the kobs values
determined from the slope of the curves obtained after applying the
integrated first-order rate law for the data points of each kinetic run.
[b] Extrapolated from the data in reference [26]. [c] Calculated on the basis
of reference [28]. [d] The rate of the uncatalysed reactions were assumed to
be equal to that of 3',5'-UpU or 3',5'-ApA[31] and kuncat was extrapolated
from the data given in reference [26]. [e] Due to precipitate formation data
has to be considered as rough estimation.
mechanism in which the (double) Lewis acid activation most
likely plays a significant role.
Cleavage of diribonucleotides and oligoribonucleotides: The
hydrolysis of six different dinucleoside monophosphates
(3a,b) and one dinucleoside diphosphate (4) was investigated
in equimolar solutions of tdci, copper(ii) and zinc(ii). Since the
reaction rates were slower[26] than those observed for the
cyclic nucleotides, the experiments were performed at a
slightly higher temperature (358C). A full pH rate profile for
the metal-complex-dependent hydrolysis of 3',5'-UpU was
measured (Figure 8). The shape of the curve is distorted
monophosphates are rather modest. As with the cyclic
phosphodiesters, the presence of an adenine base in the
substrate results in faster cleavage. Surprisingly, the cleavage
rate of 2',5'-ApA is only 15% of that determined for the 3',5'-
ApA isomer. As a result, it can be speculated that the base
moiety somehow participates in binding the catalyst, although
the nature of the interaction (direct coordination, water-
mediated coordination, or hydrogen bonding) is not clear.
This interaction appears to place the catalyst in a position in
which it is better able to cleave the 3',5'-bond than the 2',5'-
bond. Direct coordination of the N3 uracil nitrogen atoms of
3',5'-UpU to the copper(ii) centres of a trinuclear complex was
recently assumed to be a key factor in the observed substrate
specifity.[27]
The
3'-phosphorylated
dinucleoside
diphosphate
(3',5'ApU-3'p, 4) is considerably less susceptible to cleavage
by the metal complexes discussed above. The rate increase for
the catalysed system over the uncatalysed system is only 11%
of that observed for 3',5'-ApU (see Table 3).[28] Evidently, the
active complex is unable to bind the substrate through two
adjacent phosphate groups or this type of binding mode is not
advantageous for cleavage. Exclusive coordination to the
dianionic 3'-phosphate is believed to predominate; this in turn
retards the cleavage of the 3',5'-phosphodiester bond. In this
respect, the above cleavage differs from the metal aqua ion
promoted cleavage of di- and oligoribonucleotides that have a
3'-terminal monophosphate group[29] because the 3'-phos-
phate group markedly accelerates hydrolysis even when the
scissile phosphodiester bonds are situated several nucleoside
units apart.[29e,f]
Figure 8. pH rate profile for the hydrolysis of 3',5'-UpU in the tdci
CuII ZnII 1:1:1 system (Tˆ 358C, [3',5'-UpU] ꢁ 8  10À5 m, [tdci] ˆ
[CuII] ˆ [ZnII] ˆ 2.5  10À3 m).
relative to the symmetric profile obtained for 2',3'-cUMP at
258C. Instead of a sharp maximum at pH 8.7, a relatively wide
plateau in the pH range of 9 9.5 is observed; this suggests a
speciation change, and/or that hydrolysis is promoted by at
least one more species beside those already considered. The
observed rate constant at about pH 9 is approximately 50- and
10-fold higher than the values obtained at the same pH for the
copper(ii) tdci and zinc(ii) tdci binary systems, respectively.
Hydrolysis of five other dinucleoside monophosphates and a
dinucleoside diphosphate was also studied at this pH, and the
kobs values obtained are presented in Table 3. The differences
in the rate constants for the cleavage of various dinucleoside
The metal-ion-induced hydrolysis of polyribonucleotides
(poly(U) and native RNA) and a number of different
oligonucleotides has previously been studied.[29 31] Metal ions
and some of their tri- and tetraazamacrocyclic chelates have
been shown to enhance the cleavage of poly(U) more
efficiently than the cleavage of UpU.[30a, 32] To our knowledge,
Chem. Eur. J. 2003, 9, 5404 5415
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5411

¬
A. Jancso et al.
FULL PAPER
data on the cleavage of oligoribonucleotides by dinuclear
metal complexes is not available. Studies that involved
poly(U), poly(A), or type III RNA in tdci copper(ii) zinc(ii)
ternary systems were carried out at 608C in order to reduce
the influence of base-stacking interactions within the polymer
chains.[33] The data obtained are presented in Table 3. Notable
activity was found, especially with poly(A); this is consistent
with the base selectivity observed for mono- and dinucleo-
tides. The rate constants determined indicate approximately a
66- and 4-fold rate increase ([tdci] ˆ 1mm) relative to the
metal-complex-independent reaction of poly(A) and poly(U),
respectively. Taking into account the different conditions
(temperature, pKw, concentration of the catalyst), these data
indicate that the cleavage of polynucleotides is significantly
slower than the hydrolysis of dinucleoside monophosphates.
In agreement with the low catalytic activity observed for the
ternary system towards the cleavage of 3',5'-ApU-3'p, the
presence of neighbouring phosphodiester groups does not
increase the efficiency of the catalyst, as with metal aqua
ions.[30a] Furthermore, the 3'-terminal cyclic phosphate groups
that are rapidly hydrolysed to give dianionic monophosphates
exert a rate-retarding effect similar to that observed for 3',5'-
ApU-3'p. Bridging coordination of the active complex to two
neighbouring phosphate diester moieties presumably does not
predominate or cannot increase the rate of hydrolysis. Kinetic
studies conducted on native RNA resulted in the formation of
a precipitate. Though first-order kinetics were obeyed, the
reliability of the data is questionable. Measurements were
also taken for the tdci copper(ii) (1:2) binary system
([tdci] ˆ 1.0 mm). The observed rate constants are only slightly
higher than the extrapolated values determined for the
uncatalysed reactions (poly(U): kobs ꢁ 8.0 Â 10À7 sÀ1, poly(A):
kobs ꢁ 1.3 Â 10À6 sÀ1).
lower than that observed for the cyclic nucleotides (pHmax
ꢁ
8.70), but is almost identical to the one observed for the
dinuclear copper(ii) complex when either unactivated (2',3'-
cUMP, 2',3'-cAMP) or activated (BNPP, NPP, DNEP) esters
were used. In contrast to the results obtained for the
unactivated esters under similar conditions, the pseudo-first-
order rate constant for hydrolysis of BNPP, at the pH
maximum, is only about one quarter of the kobs value
measured for the solution of tdci and copper(ii). The depend-
ence of the reaction rate on the total metal-to-ligand ratio
(Figure S3 in the Supporting Information) indicates that the
dinuclear species plays an essential role. The shape of the
curve is again similar to that observed for the binary system.[19]
Pseudo-first-order rate constants were found to be propor-
tional to the total concentration of the complex mixture (using
equimolar amounts of the ligand and the metal ions; Figure S4
in the Supporting Information); this indicates that the
reaction is first-order with respect to the concentration of
the active species. In the absence of information about the
concentration of the active complex in solution, only a lower
limit for the second-order rate constant could be calculated
(k2 ˆ 7.6  10À2 mÀ1 sÀ1) from the slope of the linear correla-
tion. This value is less than 10% of the second-order rate
constant determined for the Cu2LHÀ3 complex, but 3800-fold
higher than the second-order rate constant calculated for the
hydroxide-ion-catalysed reaction.[34] It can be concluded that
the ternary system is less efficient than the copper(ii) complex
in promoting the cleavage of BNPP, which has a good leaving
group. Furthermore, the homodinuclear Cu2LHÀ3 complex
may contribute to the observed rate acceleration and, hence,
an even larger difference between the catalytic activity of the
heterodinuclear species and Cu2LHÀ3, in favour of the latter
complex, is possible.
Mechanism of the heterodinuclear-complex-promoted hydro-
lytic cleavage: On the basis of the results obtained for the four
different types of substrates, some general conclusions can be
drawn with regard to the mechanism of the cleavage of the
phosphodiester bond promoted by the heterodinuclear tdci
complex. As summarised in a recent review,[35] dinuclear
metal complexes can promote the cleavage of phosphodiester
bonds by participating at several different stages of the
process. As described for the binary systems of tdci, the pH for
maximum catalytic activity and the relative magnitude of the
rate enhancement observed for cyclic phosphodiesters and
diribonucleoside monophosphates suggests that a metal-
bound hydroxide ion serves either as a direct nucleophile,
which attacks the phosphorous atom (cyclic nucleotides), or
as a general base, which deprotonates the attacking 2'-
hydroxy group (dinucleotides, polynucleotides). Moreover,
the magnitude of the second-order rate constants, relative to
the corresponding values of the hydroxide-ion-catalysed
reactions, strongly suggests that Lewis acid activation by the
metal phosphate coordination contributes to the rate accel-
eration. The metal-to-ligand ratio dependent experiments
indicate that the second metal ion is essential for hydrolysis.
However, questions arise as to the role the second metal ion
plays. Although cooperation between the metal ions in
dinuclear complexes has recently been disputed,[36] many
Hydrolysis of BNPP: To compare the results obtained for the
cleavage of BNPP 5 catalysed by the dinuclear copper(ii)
complex, studies were performed on the tdci copper(ii)
zinc(ii) ternary system. The pH rate profiles for equimolar
solutions of the components at two different complex
concentrations (Figure 9) are bell-shaped and exhibit a
maximum around pH 8.5. The pH maximum is somewhat
Figure 9. pH rate profiles for the hydrolysis of BNPP in the tdci CuII
ZnII 1:1:1 system (Tˆ 258C, [BNPP] ˆ 5  10À4 m; A: [tdci] ˆ 2.5  10À4 m,
B: [tdci] ˆ 7.1  10À4 m).
5412
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5404 5415

Phosphodiester Cleavage
5404 5415
findings on the beneficial role of a dinuclear core in which two
metal centres can perform a concerted action on the
promote the hydrolysis of dinucleoside monophosphates.
Interestingly, 3',5'-ApA is cleaved 6.5 times faster than the
2',5'-isomer. This, together with the observed preference for
the cleavage of substrates that contain adenine, suggests that
the base is involved in the catalyst substrate interaction.
When a dinucleoside diphosphate (3',5'-ApU-3'p) is used as a
substrate, a significantly lower catalytic activity is observed.
On the basis of all the kinetic data, the catalytic activity of
the ternary system can be attributed to a heterodinuclear
complex, and the presence of this species is also supported by
the spectrophotometric measurements and MRA of the
spectral data. The mechanism proposed for the heterodinu-
clear complex catalysed phosphodiester bond cleavage is
trifunctional: 1) (double) Lewis-acid activation of the sub-
strate by coordination to the phosphate group, 2) general base
or nucleophilic catalysis (depending on the substrate) of a
metal-bound hydroxide ion and 3) general acid catalysis of a
zinc(ii)-bound water molecule that promotes the departure of
the leaving alkoxide group.
For the first time, kinetic data for the cleavage of
polyribonucleotides by dinuclear complexes have been ob-
tained. In contrast to metal aqua ions, the heterodinuclear
complexes of tdci exert a smaller rate increase for the
cleavage of 3',5'ApU-3'p and polynucleotides in comparison
to the cleavage of dinucleoside monophosphates. In summary,
the copper(ii) zinc(ii) tdci ternary system provides a new
example of cooperative behaviour of bimetallic centres in
hydrolytic reactions and also provides evidence for the
beneficial influence of the presence of two different metal
ions in the complex.
e, 8a, 9, 10, 19]
phosphate (double Lewis acidactivation) exist.[4c
In the present case, double Lewis-acid activation may occur,
but the remarkable differences between the catalytic activity
of the hetero- and homodinuclear complexes suggests that the
zinc(ii) ion, in the mixed dinuclear complex, assists the
departure of the alkoxide leaving group. This is further
supported by the findings that in the hydrolysis of the
activated ester BNPP, the catalytic efficiency of the ternary
system is significantly lower than that found for the cop-
per(ii) tdci binary system. Taking into account that the pKa of
a water molecule bound to a copper(ii) ion is generally lower
than that of a zinc(ii)-bound water molecule, the latter may act
as a general acid catalyst in the hydrolytic process promoted
by a heterodinuclear tdci complex. Double Lewis acid
activation, induced by the copper(ii) ions of the dinuclear
copper(ii) species, plays a crucial role in the cleavage of BNPP
(and cyclic nucleotides), but is not capable of enhancing the
cleavage of the phosphodiester bonds of di- and polynucleo-
tides that contain poor leaving groups. The mechanism of the
heterodinuclear complex catalysed reaction is presumably
trifunctional because the catalyst promotes three steps of the
process (Scheme 3): 1) (double) Lewis acid activation of the
Experimental Section
Materials: Copper(ii) and zinc(ii) perchlorate or nitrate (Fluka) solutions
were standardised complexometrically. HEPES, CHES, 2',3'-cUMP, 2',3'-
cAMP, 2',3'-cCMP, 2',3'-cGMP, 3',5'-UpU, 2',5'-UpU, 3',5'-ApA, 2',5'-ApA,
3',5'-UpA, 3',5'-ApU, 3',5'-ApU-3'p, polyuridylic acid, polyadenylic acid,
ribonucleic acid type III from baker×s yeast, and BNPP (all Sigma
products) were used without further purification. Phosphodiesterase I
(lyophilised powder) was purchased from USB, and alkaline phosphatase
(a concentrated solution) was obtained from Boehringer, Mannheim.
Ligand tdci (tdci ¥ 3HCl ¥ 2H2O) was prepared as described earlier.[37]
Scheme 3. Proposed mechanism for the phosphodiester bond cleavage
catalysed by the heterodinuclear species.
substrate by coordination to the phosphate group, 2) general
base or nucleophile catalysis by a metal-bound hydroxide ion
and 3) general acid catalysis by a zinc(ii)-bound water
molecule for the departure of the leaving alkoxide group.
UV-Visible spectroscopic measurements and evaluation of the absorbance
matrices by MRA: UV-Visible spectra were measured on a Hewlett
Packard 8453 diode array spectrophotometer. Analyses of the absorbance
matrices obtained for the tdci copper(ii) binary system and tdci cop-
per(ii) zinc(ii) ternary system were performed by a recently described
algorithm based on the application of MRA.[21] The simple MRA method
and the calculation of the wavelength-dependent residual absorbance
curves (RAC) were also used to evaluate the NIAS. The initial standard
deviation of the actual matrix element applied for the computation of error
propagation was set to 0.002absorbance units; this was in accordance with
the reproducibility of the spectra.
Conclusion
The homodinuclear copper(ii) and zinc(ii) complexes of tdci
promote the cleavage of cyclic ribonucleoside 2',3'-mono-
phosphates by several orders of magnitude. The active
copper(ii) complex shows remarkable base selectivity in
favour of substrates that contain an adenine base and a high
regioselectivity that favours formation of the 3'-phosphory-
lated product. For the system that contains copper(ii) and
zinc(ii) in an equimolar ratio with respect to tdci, the reaction
rate increases 7 20 times to that observed for the homodi-
nuclear complexes. The ternary system, in contrast to the
homodinuclear complexes, was also found to effectively
Kinetic studies for the hydrolysis of nucleoside 2',3'-cyclic monophosphates
and transesterification of dinucleoside monophosphates: The pH of the
reaction solutions was adjusted with HEPES (0.03m), CHES (0.03m), and
by adding an appropriate amount of sodium hydroxide stock solution. The
ionic strength was adjusted to 0.1m with NaClO4 or NaNO3. After
measuring the pH of the reaction solutions at the temperature of the kinetic
measurements, the Eppendorf tubes (1.0 mL of reaction solution) were
immersed in a water bath, the temperature of which was maintained at
25.0 Æ 0.18C (cNMP) or 35.0 Æ 0.18C (NpN). The reactions were initiated
by adding 30 40 mL of the stock solution of the substrate. The initial
Chem. Eur. J. 2003, 9, 5404 5415
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5413

¬
A. Jancso et al.
FULL PAPER
substrate concentration was 0.05 0.1mm. The ligand (tdci) and metal-ion
concentrations varied from 0.0 6.0mm. In a typical experiment, the
concentration of tdci was 2.5mm, while that of the metal ion was 5.0mm.
Aliquots (7 10) were withdrawn at appropriate intervals from the reaction
solutions. The reactions were quenched by mixing a 0.1 mL aliquot with an
equal amount of the eluent used in the HPLC analyses. The samples were
kept in a freezer until analysis.
respect to all the nucleoside units of the starting material (i.e., the total
number of phosphodiester bonds).
Hydrolysis of bis(4-nitrophenyl)phosphate (BNPP): The samples were
prepared as described above and kept at 25.0 Æ 0.18C prior to analysis. The
reactions were initiated by injecting 100 mL of a 0.0103m solution of BNPP
(solid substrate dissolved in a 50 w/w% ethanol/water mixture) into an
efficiently stirred, pre-thermostated sample solution. The reactions were
followed by the increase of the absorption band for the p-nitrophenoxide
ion at 400 nm (e ˆ 18900 mÀ1 cmÀ1, pKa ˆ 6.98). The initial slope method
(ꢂ4% conversion) was used to determine the pseudo-first-order rate
constants. The increase in absorbance at 400 nm was followed immediately
after injection of the substrate. The reported data are the average of
duplicate measurements with a reproducibility of better than 10%. The
initial concentration of BNPP was 0.5mm. In a typical experiment, the
concentration of tdci, CuII and ZnII was 0.25, 0.25 and 0.25mm, respectively.
The aliquots were analysed by RP-HPLC (Perkin Elmer) by using a
Hypersil ODS RP-18 column (250 Â 4 mm, 5 mm particle size). The flow
rate was 1 mLminÀ1. The composition of the eluents was as follows: A)
0.06 moldmÀ3 acetate buffer, pH ˆ 4.3, I ˆ 0.1 moldmÀ3 (NH4Cl); and B)
eluent Awith 10% acetonitrile. The elution programs (proportion of buffer
A and duration time) used for the different substrates were: 2',3'-cUMP, 0
10 min, 100% A; 2',3'-cAMP, 0 12min, 65% A; 2 ',3'-cCMP, 0 10 min,
100% A; 2',3'-cGMP, 0 12min, 85% A; 3 ',5'-UpU and 2',5'-UpU, 0
5 min, 100% A; 5 17 min, 60% A; 3',5'-ApA and 2',5'-ApA, 0 8 min,
65% A; 8 18 min, 25% A; 3',5'-UpA, 0 8 min, 100% A; 8 17 min,
30% A; 3',5'-ApU, 0 9 min, 65% A; 9 19 min, 40% A; 3',5'-ApU-3'p,
Acknowledgements
0
3 min, 90 65% A (linear gradient); 3 16 min, 65% A; 16 18 min,
30% A. The products (nucleosides, nucleoside 2',3'-cyclic monophos-
phates, nucleoside 2'-monophosphates, and nucleoside 3'-monophos-
phates) were identified by spiking with authentic samples. The UV
detection wavelength was 260 nm. The reactions were followed for about
three half-lives. In all cases, the disappearance of the starting material
obeyed pseudo-first-order kinetics. The pseudo-first-order rate constants
This research has been supported by a Marie Curie Fellowship from the
European Community program ™Improving the Human Research Poten-
tial and the Socio-Economic Knowledge Base∫ (contract number ™HPMF-
CT-2002-01860∫), the Hungarian Research Foundation (OTKA T037385)
and the Academy of Finland.
(kobs
) for the cleavage of the starting materials were calculated by
application of the integrated first-order rate law. For nucleoside 2',3'-
cyclic-monophosphates, the reported data are the average of duplicate
measurements with a reproducibility ꢂ15%.
[1] D. E. Metzler, Biochemistry: The Chemical Reactions of Living Cells,
Academic Press, New York, 1977, Chapters 7, 8, and 11.
»
[2] a) W. N. Lipscomb, N. Strater, Chem. Rev. 1996, 96, 2375 2433;
Kinetic measurements of polynucleotides and native RNA: The temper-
ature and pH of the reaction mixtures were adjusted as described above
(Tˆ 60.0 Æ 0.28C). Aliquots from the solutions were taken periodically, the
reactions were quenched by the addition of an excess of EDTA solution
with respect to the metal ions (pH ˆ 9.1), and the samples were cooled in an
ice bath. The uncleaved polynucleotides were digested by treatment with
phosphodiesterase I, as described earlier.[30, 38] Accordingly, phosphodies-
terase I in a TRIS:HCl buffer, which contained NaCl and MgCl2 (pH ˆ
9.1), was added to the quenched aliquots. The final concentrations of TRIS,
NaCl and MgCl2 were 0.05m, 0.1m, and 0.01m, respectively. To ensure that
the enzymatic digestion was complete, the aliquots were left to stand for
20 24 h, at which time they were filtered through a syringe-driven
Hydrophobic Fluoropore (PTFE) membrane filter unit (20 mm, Millipore),
which removed the proteins. Prior to HPLC analysis, the samples were
neutralised with acetic acid. The eluent buffers employed for separation of
the homopolymer product mixtures were those described above for the
mono- and dinucleotide substrates. The elution programs applied were:
poly(U), 0 12min, 100% A; poly(A), 0 1 min, 80% A, 1 6 min, 80%
65% A (gradient); 6 16 min, 65% A. For separation of the RNA cleavage
products, buffers with 0.3m NH4Cl were used to improve the separation.
The elution program was: 0 7 min, 100% A, 7 22 min, 100% 35% A
(gradient); 22 30 min, 35% A.
b) D. E. Wilcox, Chem. Rev. 1996, 96, 2435 2458; c) G. C. Dismukes,
Chem. Rev. 1996, 96, 2909 2926.
[3] For review articles, see: a) J. Chin, Curr. Opin. Cell Biol. 1997, 1, 514
»
521; b) R. Kramer, T. Gajda in Perspectives in Bioinorganic Chemistry,
Vol. 4 (Eds.: R. W. Hay, J. R. Dilwoth, K. Nolan), JAI Press, Stamford,
1999, pp. 207 240.
[4] a) M. Yashiro, A. Ishikubo, M. Komiyama, J. Biochem. 1996, 120,
1067 1069; b) K. G. Ragunathan, H.-J. Schneider, Angew. Chem.
1996, 108, 1314 1316; Angew. Chem. Int. Ed. Engl. 1996, 35, 1219
1221; c) M. Yashiro, A. Ishikubo, M. Komiyama, Chem. Commun.
1997, 83 84; d) P. Molenveld, J. F. J. Engbersen, H. Kooijman, A. L.
Spek, D. N. Reinhoudt, J. Am. Chem. Soc. 1998, 120, 6726 6737; e) P.
Molenveld, J. F. J. Engbersen, D. N. Reinhoudt, J. Org. Chem. 1999,
64, 6337 6341.
[5] a) J. S. Seo, N.-D. Sung, R. C. Hynes, J. Chin, Inorg. Chem. 1996, 35,
7472 7473; b) T. Koike, M. Inoue, E. Kimura, M. Shiro, J. Am. Chem.
Soc. 1996, 118, 3091 3099; c) S. J. Brudenell, L. Spiccia, D. C. R.
Hockless, E. R. T. Tiekink, J. Chem. Soc. Dalton Trans. 1999, 1475
1481; d) C. He, S. J. Lippard, J. Am. Chem. Soc. 2000, 122, 184 185;
e) A. M. Barros, S. J. Lippard, Inorg. Chem. 2001, 40, 1060 1064.
[6] a) W. H. Chapman, R. Breslow, J. Am. Chem. Soc. 1995, 117, 5462
5469; b) S.-I. Kawara, T. Uchimaru, Eur. J. Inorg. Chem. 2001, 2437
2442.
[7] C. Sissi, P. Rossi, F. Felluga, F. Formaggio, M. Palumbo, P. Tecilla, C.
Toniolo, P. Scrimin, J. Am. Chem. Soc. 2001, 123, 3169 3170.
[8] a) M. Wall, R. C. Hynes, J. Chin, Angew. Chem. 1993, 105, 1696 1697;
Angew. Chem. Int. Ed. Engl. 1993, 32, 1633 1635; b) M. Yashiro, A.
Ishikubo, M. Komiyama, J. Chem. Soc. Chem. Commun. 1995, 1793
1794; c) M. J. Young, J. Chin, J. Am. Chem. Soc. 1995, 117, 10577
10578; d) K. P. McCue, D. A. Voss, Jr., C. Marks, J. R. Morrow, J.
Chem. Soc. Dalton Trans. 1998, 2961 2963; e) S. Liu, Z. Luo, A. D.
Hamilton, Angew. Chem. 1997, 109, 2794 2796; Angew. Chem. Int.
Ed. Engl. 1997, 36, 2678 2680; f) O. Iranzo, A. Y. Kovalevsky, J. R.
Morrow, J. P. Richard, J. Am. Chem. Soc. 2003, 125, 1988 1993; g) K.
Worm, F. Chu, K. Matsumoto, M. D. Best, V. Lynch, E. V. Anslyn,
Chem. Eur. J. 2003, 9, 741 747.
The method applied for the determination of the rate constants has
previously been described in detail.[30a] The reactions were followed for
10 20% transformation of the starting materials, and the aliquots
withdrawn (6 7) were treated with phosphodiesterase I to cleave all the
À
3'O P bonds of the starting material and oligomeric cleavage products. As
a result, the 5'-terminal nucleosides formed upon the chemical cleavage of
the RNA chain were detected as nucleosides, and the intrachain nucleo-
sides were detected as 5'-phosphates. Small amounts of the 2',3'-cyclic
monophosphates and the hydrolysis products (2'- and 3'-monophosphates)
that are formed upon chemical cleavage of the 3'-terminal phosphodiester
bond were also released. The amount of these products was added to the
nucleoside yields in order to obtain the total amount of phosphodiester
bonds cleaved chemically. Aliquots withdrawn from the reaction solutions
were also treated with a mixture of phosphodiesterase I and alkaline
phosphatase in order to determine the amount of 3'-terminal nucleoside
and 2',3'-cyclic phosphates formed upon chain cleavage.[30b] However, only
nucleosides were detected after the enzymatic digestion; this indicates that
chemical cleavage of the 3'-terminal cyclic phosphate groups was rapid. The
rate constants were calculated by determining the mole fraction of 5'-
terminal nucleosides (i.e., the number of cleaved bonds) at time t with
[9] N. H. Williams, J. Chin, Chem. Commun. 1996, 131 132.
¬
[10] a) T. Gajda, R. Kra»mer, A. Jancso, Eur. J. Inorg. Chem. 2000, 1635
¬
»
1644; b) T. Gajda, A. Jancso, S. Mikkola, H. Lonnberg, H. Sirges, J.
Chem. Soc. Dalton Trans. 2002, 1757 1763; c) S. Albedyhl, D.
¬
Schnieders, A. Jancso, T. Gajda, B. Krebs, Eur. J. Inorg. Chem. 2002,
1400 1409.
5414
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2003, 9, 5404 5415

Phosphodiester Cleavage
5404 5415
»
[11] a) S. Kuusela, H. Lonnberg, J. Phys. Org. Chem. 1993, 6, 347 356;
[24] a) B. Linkletter, J. Chin, Angew. Chem. 1995, 107, 529 531; Angew.
Chem. Int. Ed. Engl. 1995, 34, 472 474; b) M. Wall, B. Linkletter, D.
Williams, A.-M. Lebuis, R. C. Hynes, J. Chin, J. Am. Chem. Soc. 1999,
121, 4710 4711.
»
b) S. Kuusela, H. Lonnberg, Met. Ions Biol. Syst. 1996, 32, 272 300.
[12] S. Mikkola, E. Stenman, K. Nurmi, E. Yousefi-Salakdeh, R. Stro»m-
»
berg, H. Lonnberg, J. Chem. Soc. Perkin Trans. 2 1999, 1619 1625.
»
»
[13] a) N. Strater, T. Klabunde, P. Tucker, H. Witzel, B. Krebs, Science
[25] H. Lonnberg in Biocoordination Chemistry: Coordination Equilibria
1995, 268, 1489 1492; b) J. Uppenberg, F. Lindqvist, C. Svensson, B.
Ek-Rylander, G. Andersson, J. Mol. Biol. 1999, 290, 201 211.
[14] a) S. Albedyhl, M. T. Averbuch-Pouchot, C. Belle, B. Krebs, J. L.
Pierre, E. Saint-Aman, S. Torelli, Eur. J. Inorg. Chem. 2001, 1457
1464; b) M. Lanznaster, A. Neves, A. J. Bortoluzzi, B. Szpoganicz, E.
Schwingel, Inorg. Chem. 2002, 41, 5641 5643; c) C. Belle, I. Gautier-
Luneau, L. Karmazin, J.-L. Pierre, S. Albedyhl, B. Krebs, M. Bonin,
Eur. J. Inorg. Chem. 2002, 3087 3090.
[15] P. Karsten, A. Neves, A. J. Bortoluzzi, M. Lanznaster, V. Drago, Inorg.
Chem. 2002, 41, 4624 4626.
[16] L.-H. Yin, P. Cheng, S.-P. Yan, X.-Q. Fu, J. Li, D.-Z. Liao, Z.-H. Liang,
J. Chem. Soc. Dalton Trans. 2001, 1398 1400.
in Biologically Active Systems (Ed.: K. Burger), Ellis Horwood,
Chichester, England, 1990, Chapter VII, pp. 284 346.
»
»
[26] P. Jarvinen, M. Oivanen, H. Lonnberg, J. Org. Chem. 1991, 56, 5396
5401.
[27] M. Komiyama, S. Kina, K. Matsumura, J. Sumaoka, S. Tobey, V. M.
Lynch, E. Anslyn, J. Am. Chem. Soc. 2002, 124, 13731 13736.
[28] T. Koike, Y. Inoue, Chem. Lett. 1972, 569 570.
[29] a) J. J. Butzow, G. L. Eichorn, Biochemistry 1971, 10, 2019 2027;
b) H. Ikenaga, Y. Inoue, Biochemistry 1974, 13, 577 582; c) V. M.
Shelton, J. R. Morrow, Inorg. Chem. 1991, 30, 4295 4299; d) J. R.
Morrow, V. M. Shelton, New J. Chem. 1994, 18, 371 375; e) S.
»
Kuusela, A. Azhayev, A. Guzaev, H. Lonnberg, J. Chem. Soc. Perkin
»
[17] a) J. Kamitani, R. Kawahara, M. Yashiro, M. Komiyama, Chem. Lett.
1998, 1047 1048; b) M. Irisawa, M. Komiyama, J. Biochem. 1995, 117,
465 466; c) P. Molenveld, J. F. J. Engbersen, D. N. Reinhoudt, Angew.
Chem. 1999, 111, 3387 3390; Angew. Chem. Int. Ed. 1999, 38, 3189
3192.
Trans. 2 1995, 1197 1202; f) S. Kuusela, A. Guzaev, H. Lonnberg, J.
Chem. Soc. Perkin Trans. 2 1996, 1895 1899.
»
[30] a) S. Kuusela, H. Lonnberg, J. Chem. Soc. Perkin Trans. 2 1994, 2301
»
2306; b) S. Kuusela, H. Lonnberg, Nucleosides Nucleotides 1996, 15,
1669 1678.
[18] a) T. Itoh, H. Hisada, Y. Usui, Y. Fujii, Inorg. Chim. Acta 1998, 283,
51 60; b) T. Itoh, H. Hisada, T. Sumiyama, M. Hosono, Y. Usui, Y.
Fujii, Chem. Commun. 1997, 677 678; c) C. Sissi, F. Mancini, M.
Palumbo, P. Scrimin, P. Tecilla, U. Tonellato, Nucleosides Nucleotides
Nucleic Acids 2000, 19, 1265 1271.
[31] S. Kuusela, H. Lo»nnberg, J. Chem. Soc. Perkin Trans. 2 1994, 2109
2113.
»
[32] S. Kuusela, H. Lonnberg, Curr. Top. Solution Chem. 1997, 2, 29 47.
»
[33] S. Mikkola, U. Kaukinen, H. Lonnberg, Cell Biochem. Biophys. 2001,
34, 95 119.
»
» »
[19] T. Gajda, Y. Dupre, I. Torok, J. Harmer, A. Schweiger, J. Sandler, D.
Kuppert, K. Hegetschweiler, Inorg. Chem. 2001, 40, 4918 4927.
[20] J. K. Bashkin, L. A. Jenkins, J. Chem. Soc. Dalton Trans. 1993, 3631
3632.
[34] F. M. Menger, L. H. Gan, E. Johnson, D. H. Durst, J. Am. Chem. Soc.
1987, 109, 2800 2803.
[35] N. H. Williams, B. Takasaki, M. Wall, J. Chin, Acc. Chem. Res. 1999,
32, 485 493.
¬
¬
[21] G. Peintler, I. Nagypal, A. Jancso, I. R. Epstein, K. Kustin, J. Phys.
[36] M. Leivers, R. Breslow, Bioorg. Chem. 2001, 29, 345 356.
[37] K. Hegetschweiler, I. Erni, W. Schneider, H. Schmalle, Helv. Chim.
Acta 1990, 73, 97 105.
[38] R. Corcoran, M. Labelle, A. W. Czarnik, R. Breslow, Anal. Biochem.
1985, 144, 563 568.
Chem. A 1997, 101, 8013 8020.
[22] a) M. R. Eftink, R. L. Biltonen, Biochemistry 1983, 22, 5134 5140;
b) M. Oivanen, H. Lo»nnberg, J. Org. Chem. 1989, 54, 2556 2560;
c) M. Komiyama, J. Am. Chem. Soc. 1989, 111, 3046 3050; d) H. I.
Abrash, C.-C. S. Cheung, J. C. Davis, Biochemistry 1967, 6, 1298 1303.
»
[23] S. N. Mikhailov, M. Oivanen, P. Oksman, H. Lonnberg, J. Org. Chem.
1992, 57, 4122 4126.
Received: May 16, 2003 [F5149]
Chem. Eur. J. 2003, 9, 5404 5415
www.chemeurj.org
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5415