There is no universal mechanism for the cleavage of RNA model compounds in the presence of metal ion catalysts

Article abstract of DOI:10.1039/c3ob41554f

The transesterification of uridine 3′-phosphodiesters with a wide range of leaving group alcohols has been studied in the presence of monometallic and bimetallic complexes. The catalysis of isomerization of the phosphodiester bond was studied with a nucleoside 3′-phosphonate as a substrate. The results obtained are consistent with a step-wise mechanism, where metal ions are able to enhance both the nucleophilic attack and the departure of the leaving group. The mechanism of the catalysis depends on the acidity of the catalyst and of the leaving group alcohol: a change from general base catalysis to general acid catalysis is proposed. Catalysis of the isomerization requires efficient stabilization of the phosphorane by strong interactions with the catalyst. Catalytic strategies utilised by bimetallic complexes are also briefly discussed.

Full text of DOI:10.1039/c3ob41554f

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
There is no universal mechanism for the cleavage of
RNA model compounds in the presence of metal ion
catalysts†
Cite this: Org. Biomol. Chem., 2013, 11,
8324
Heidi Korhonen, Timo Koivusalo, Suvi Toivola and Satu Mikkola*
The transesterification of uridine 3’-phosphodiesters with a wide range of leaving group alcohols has
been studied in the presence of monometallic and bimetallic complexes. The catalysis of isomerization
of the phosphodiester bond was studied with a nucleoside 3’-phosphonate as a substrate. The results
obtained are consistent with a step-wise mechanism, where metal ions are able to enhance both the
nucleophilic attack and the departure of the leaving group. The mechanism of the catalysis depends on
the acidity of the catalyst and of the leaving group alcohol: a change from general base catalysis to
general acid catalysis is proposed. Catalysis of the isomerization requires efficient stabilization of the
phosphorane by strong interactions with the catalyst. Catalytic strategies utilised by bimetallic complexes
are also briefly discussed.
Received 29th July 2013,
Accepted 30th September 2013
DOI: 10.1039/c3ob41554f
www.rsc.org/obc
results, most probably from slow pseudorotation, since accord-
ing to Westheimer’s rules for pseudorotating oxyphosphorane
Introduction
Metal ion promoted cleavage of phosphodiesters has been exten- intermediates,9 placing a negatively charged oxyanion in an
sively studied over the last three decades and a number of review apical position, which is inevitable with a dianionic phosphor-
articles covering different aspects of metal ion catalysis have ane, is energetically unfavourable.
been published recently.1–7 The two main objectives of such
Studies with small molecular weight substrates, such as bis-
studies are the need to understand the role of metal ions in the (p-nitrophenyl)phosphate (1), 1-(2-hydroxypropyl)-p-nitrophenyl-
catalysis by metal ion dependent ribozymes or enzymes and phosphate (2) and nucleoside 3′-aryl (3) and alkyl esters (4),
the need to develop efficient catalysts for the cleavage of RNA.
have revealed the basic features of metal ion dependent reac-
In the absence of metal ions RNA is cleaved by intramole- tions of phosphodiesters. Reviews discussing these studies can
cular transesterification, where 2′-OH attacks the phosphate be found for example in ref. 3 and 4. References to more
resulting in the formation of a pentacoordinated phosphorane specific articles are found in the Results and discussion sec-
(Scheme 1).8 The fate of the phosphorane depends on tions. Zn2+ and Cu2+ complexes of nitrogen ligands have com-
the conditions: under neutral and acidic conditions monly been used as catalysts. Mononuclear complexes studied
cleavage and isomerization of a phosphodiester bond are include Zn2+ complexes with cyclic and acyclic nitrogen
observed. Under alkaline conditions cleavage is the only reac- ligands (8–13), such as azacrowns, 1,5,9-triazacyclododecane
tion. The difference results from the properties of the phos- (8), 1,4,7-triazacyclononane (9) and cyclen (10) as well as Cu2+
phorane under the experimental conditions. While neutral complexes with bipyridine (14) and terpyridine (15) and their
and monoanionic species are stable enough to pseudorotate, derivatives. The shapes of the pH-rate profiles are usually sig-
which is a prerequisite for isomerization,9 pseudorotation of a moidal or bell-shaped, depending on the system, and a cata-
dianionic species has not been observed. The exact status of lytically important deprotonation at pH close to the pKa of a
the dianionic phosphorane was unknown for long, but in 2004 metal bound aquo ligand has been observed with a wide
two independent papers were published proposing that the variety of catalyst–substrate combinations. Furthermore, metal
alkaline cleavage is a step-wise reaction with phosphorane ion catalysts with acidic aquo ligands, such as Zn2+–8 with a
being an intermediate.10,11 The absence of isomerization pKa of 7.3, are generally the best catalysts. Electrophilic cata-
lysis, nucleophilic catalysis and general acid and base catalysis
have all been proposed as the catalysis mechanisms in the
University of Turku, Department of Chemistry, FIN-20014 Turku, Finland.
cleavage of various types of substrates.
E-mail: satu.mikkola@utu.fi
The catalysis by monometallic complexes is fairly modest,
and in catalyst development a new era began with the
†Electronic supplementary information (ESI)
available. See DOI:
10.1039/c3ob41554f
8324 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 Intramolecular transesterification reactions of phosphodiester bonds of RNA.
observation that bimetallic complexes can be hundreds of is a more than 1000 times as efficient catalyst than the
times more efficient as catalysts than their monometallic corresponding monomer Zn2+–20. Second-order rate constants
counterparts. For example, the rate enhancement by the Cu2+– of 53 M−1 s−1 and 0.046 M−1 s−1 have been reported for reac-
TerPy dimer Cu2+2–16 in comparison to Cu2+–TerPy is 51-fold tions catalyzed by Zn2+2–19a16 and Zn2+–20,17 respectively. The
for the cleavage of HPNP (2)12 and 285-fold for the cleavage of rate increase can be further enhanced by using an organic
uridine 2′,3′-cyclic monophosphate (5a).13 A Zn2+–9 dimer medium: the dimer Zn2+2–21a in methanol is 1.5 × 104 times
Zn2+2–17a is a 120 times more efficient catalyst for the cleavage more efficient as a catalyst for the cleavage of 2 than Zn2+–8.18
of 2 than Zn2+–9.14 Cu2+2–18 is a more than 500 times more
In previous papers we have reported on a very efficient clea-
efficient catalyst than Cu2+–9 for the cleavage of 3′,5′-ApA (6a), vage of 3′,5′-UpU (6b)19 and a series of uridine 3′-alkyl phos-
while the difference with 2′,3′-cAMP (5b) is 287-fold.15 An even phates (4b,4d–4h)20 by the bimetallic Zn2+ complex Zn2+2–19a.
more significant rate enhancement has been observed with Interestingly this catalyst promoted also the isomerization of
(2-pyridylyl)methylamine based complexes: dimeric Zn2+2–19a phosphodiester bonds. The catalysis on isomerization was
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8325

View Article Online
Paper
Organic & Biomolecular Chemistry
modest in comparison to that for the cleavage reaction, but it similar kinetic evidence can be proposed to support different
was confirmed using a nucleoside 3′-alkylphosphonate 7a as a mechanisms. For example pH-rate profiles for the cleavage of a
substrate.19 In the absence of competing cleavage, the rate- wide variety of phosphodiester substrates are fairly similar. In
enhancement could be observed clearly. This was a first obser- the present paper we approach the question of the catalysis
vation of a sufficiently stable phosphorane intermediate along mechanism from a different angle. The catalytic activity of
the reaction route in a metal ion dependent reaction.
various mono- and bimetallic complexes in the cleavage of a
Despite extensive research, a universally accepted mechan- range of uridine 3′-aryl (3a–e) and alkyl phosphates (4a–h) has
ism for the metal ion promoted cleavage of RNA model com- been studied. By contrasting the catalytic effect of metal ion
pounds does not exist. It is generally accepted that the complexes against the nature of the background reaction that
strength of interaction between the catalyst and the substrate varies depending on the acidity of the leaving group, we have
is essential, but the role of the coordinated metal ion catalyst gathered information that complements the literature data
is not clear. Electrophilic catalysis by coordinated metal ion obtained by studying pH-dependence and solvent isotope
catalysts is widely supported,21–24 but general acid25 and effects. Furthermore, the catalytic effect of various metal ion
base26,27 catalyses by metal ion coordinated aquo and hydroxo complexes on the isomerization of nucleoside 3′-phosphonate
ligands have also been proposed. Direct coordination to the 7b was studied. In addition to providing new insights
attacking nucleophile is well established with simpler model into the mechanisms utilised by metal ion complexes in the
compounds.28–30 A difficulty in mechanistic research is that catalysis of cleavage of different phosphodiesters, the factors
8326 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
that influence the catalytic efficiency of bimetallic complexes under the experimental conditions, and the rate constants for
can be evaluated.
the background reaction for all alkyl substrates studied in the
present work were calculated using parameters reported by
Kosonen et al.31 The procedures are explained in the ESI.†
According to such an analysis, the reaction of the trichloro-
ethyl ester (4h; UpEtCl3) at pH 6.5 and 90 °C takes place
almost entirely via the base-catalysed route, whereas the predo-
Results
Analysis of the uncatalysed reaction
As the first step of the analysis of the kinetic data, the nature minant reaction of the isopropyl ester (UpiPr; 4b) is spon-
of the background reaction under the experimental conditions taneous cleavage. Information on the nature of the
was identified. According to results of Kosonen et al.,31 spon- background reaction is included in Table 1 recording the cata-
taneous and base-catalysed reactions (Scheme 2a and 2b, lytic activity of mononuclear metal ion complexes for the clea-
respectively) are the predominant cleavage routes under vage of uridine 3′-alkylphosphates.
neutral conditions. The poorer the leaving group is, the higher
Mechanistically the two background reactions are fairly
is the proportion of spontaneous cleavage under the given con- different. The spontaneous cleavage is a two-step process with
ditions, as is shown by the shape of pH-rate profiles reported a monoanionic phosphorane intermediate (Scheme 2a). The
for 4c, 4e, 4g and 4h.31 The proportions of these reactions reaction is believed to involve a concerted proton transfer from
Scheme 2 Predominant mechanisms for the cleavage of phosphodiester bonds under neutral conditions in the absence of metal ions. a. Spontaneous cleavage,
b. alkaline cleavage. From ref. 31.
Table 1 Catalysis of the cleavage of uridine 3’-alkyl phosphates 4 by mononuclear Zn2+ and Cu2+ complexes as measured by relative rate constants krel = kobs/k0
UpEtCl3 (4h)
UpEtCl2 (4g)
UpEtF2 (4f)
UpEtOEt (4e)
UpMe (4d)
UpEt (4c)
UpiPr (4b)
UpnPe (4a)
pKa (LG)a
12.2
12.9
13.0
14.8
15.5
15.8
17.1
17.3
k0/s−1 pH 6.6 at 90°b
Background reactionb,c
krel 10 mM Zn2+–8
krel 2 mM Zn2+–8
krel 2 mM Zn2+–9
krel 2 mM Zn2+–9
krel 2 mM Zn2+–10
krel 2 mM Zn2+–11
krel 2 mM Zn2+–12
krel 2 mM Zn2+–13
krel 10 mM Cu2+–14
krel 10 mM Cu2+–15
k0/s−1 pH 5.6 at 90°b
Background reactionc
krel 10 mM Zn2+–8
7.3 × 10−5
1.2 × 10−5
9.5 × 10−6
7%/93%
1.3 × 10−7
37%/63%
1873
410
2.9 × 10−8
62%/37%
1.8 × 10−8
65%/35%
3000
2.1 × 10−9
90%/10%
809
1.6 × 10−9
89%/6%
3%/97%
29
6
1.5
1
3
1.4
1
6%/94%
146
26
3.3
2.4
5.7
1.6
1.6
28
(2116)d
(181)d
52
81
119
20
4.5
8.5
7.5
1429
4333
2.0 × 10−9
96%/1%
448
140
110
42
26
40
1.3
1.3
1.3
291
2164
5.9 × 10−8
84%/14%
989
594
44
37
2.8
28
2.4
3.3
1.8
1.1
11
61
1.6
15
16
80
1502
1900
171
2048
13 000
9.0 × 10−6
22%/78%
18
1.9 × 10−6
1.5 × 10−6
41%/58%
2.2 × 10−8
88%/7%
1.3 × 10−8
94%/5%
870
2.0 × 10−9
69%/1%
39%/41%
50
25
5
e
krel 10 mM Znaq
krel 2 mM Znaq
12
1.5
1231
1524
8
a From ref. 33, except for that for 4a that was determined kinetically in 1 M NaOH at 25 °C using the known βLG value of −1.28 from ref. 31. b Rate
constant for the cleavage in the absence of a metal ion catalyst calculated using the data in ref. 31 as explained in the ESI. c Percentage of
spontaneous and alkaline reactions in the background reaction calculated using data reported in ref. 31. d Catalysis of the cleavage of UpMe (4d)
was exceptionally high and the values should be regarded with caution. e Rate constants have been reported in ref. 25.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8327

View Article Online
Paper
Organic & Biomolecular Chemistry
the phosphate to the leaving group, and it is characterized by a
moderately negative βLG of −0.59.31 The alkaline cleavage
involves a nucleophilic attack by a 2′-oxyanion, the formation
of a dianionic phosphorane, and the departure of the leaving
group as an alkoxy anion (Scheme 2b). Consistent with a nega-
tive charge on the leaving group oxygen, the alkaline cleavage
is characterized by a highly negative βLG of −1.28.31 In other
words, the analysis of the background reaction shows that
while the better alkyl leaving groups, e.g. trichloroethoxy and
dichloroethyoxy groups, can depart as alkoxy anions at pH 6.6,
protonation of the poorer leaving groups is required to allow
the cleavage.
As mentioned in the introduction, alkaline cleavage is also
a step-wise reaction with a change in the rate-limiting step at a
pKa of 12.6. Accordingly, with UpEtCl (4h) and with aryl esters
3a–e the formation of the phosphorane is the rate-limiting
step, whereas with esters with pKa(LG) > 12.6 the departure of
the leaving group determines the rate of the cleavage reaction.
The difference in kinetic solvent isotope effects determined
previously20 for the spontaneous cleavage of trichloroethyl
(3.8) and dichloroethyl (6.0) esters at pH 6.5 is consistent with
the change. Values 4.9 and 7.2 have been reported for the alka-
line cleavage of nucleoside 3′-phenyl and -methoxyethyl esters,
respectively.32
Fig. 1 Rate constants for the cleavage of uridine 3’-alkylesters in the presence
of Zn2+ complexes of different acidity at pH 6.6 and 90 °C. Notation: squares –
uncatalysed cleavage; rate constants calculated as explained in the ESI, circles –
2 mM Zn2+–11 (pKa = 9.9), triangles – 2 mM Zn2+–9 (pKa = 9.0), diamonds –
2 mM Zn2+–8 (pKa = 7.5). The pKa values given refer to the deprotonation of a
metal aquo ligand at 25 °C. They are discussed in more detail in the Results
section. Lines are manually drawn to emphasise the trend.
resembles closely that by Zn2+ but the behavior is less pro-
aq
nounced. Results obtained with Zn2+–9 are shown in Fig. 1.
Other Zn2+ complexes tested were Zn2+–11, Zn2+–12 and Zn2+–
13 that showed a very modest catalytic effect and only with
phosphodiesters with the poorest leaving groups. Results
obtained with Zn2+–11 are included in Fig. 1 as an example of
these inefficient catalysts.
Catalysis of the cleavage of uridine 3′-alkyl esters by
monomeric Zn2+ complexes
The cleavage of a number of uridine 3′-alkyl and aryl esters
was studied in the presence of various Zn2+ complexes. Table 1
shows the results obtained with the alkyl esters by recording
the rate-enhancement observed in the presence of the cata-
lysts. Actual rate constants are given in Table S2 in the ESI.†
Examples of catalysis by different Zn2+ complexes are also
shown in Fig. 1, where the logarithmic rate constants for the
cleavage of alkyl esters in the presence of 2 mM Zn2+–8, Zn2+–9
and Zn2+–11 are shown as a function of pKa of the leaving
group. It is clear on the basis of the results shown that none of
the Zn2+ based mononuclear catalysts significantly enhance
the cleavage of the trichloroethyl ester (4h; UpEtCl3). At 2 mM
concentration only Zn2+–8 and Zn2+–10 showed any observable
effect. An experiment at a higher catalyst concentration con-
firms that Zn2+–8 does enhance the cleavage. A more signifi-
cant rate-enhancement is observed with the less reactive
substrates bearing a more basic leaving group. The largest
effect among mononuclear Zn2+-catalysts is observed with
10 mM Zn2+–8 that at pH 6.6 promotes the cleavage of the
ethyl ester UpEt (4c) by a factor of 3000. As the leaving group
becomes poorer, the catalytic effect of Zn2+–8 decreases, as is
clearly shown by the results in Fig. 1. The same trend is
observed with Zn2+–10.
Zn2+ based catalysts can, hence, be roughly divided into
three categories. The first group (group A in the following dis-
cussion) includes catalysts Zn2+–8 and Zn2+–10 which were the
only ones with any observable catalytic activity at 2 mM con-
centration for the cleavage of the trichloroethyl ester UpEtCl3.
These catalysts show the highest catalytic activity with ethoxy-
ethyl UpEtOEt, ethyl UpEt or methyl UpMe esters. As the
acidity of the leaving group further decreases, the catalytic
activity decreases. The second group of catalysts consists of
Zn2+ and Zn2+–9 (group B) that only modestly enhance the
aq
cleavage of UpEtOEt with a moderately acidic leaving group,
but which show their highest catalytic activity with the least
reactive substrates UpiPr or UpnPe. There is no drop in the cata-
lytic activity of Zn2+aq. The third group of catalysts consists of
Zn2+-complexes Zn2+–11, Zn2+–12 and Zn2+–13 (group C) which
show only a very modest catalytic activity and only with the
least reactive uridine 3′-alkyl esters. The difference between
the groups is clearly seen in Fig. 1, where the logarithmic rate
constants for the cleavage of alkyl esters in the presence of
2 mM Zn2+–8, Zn2+–9 and Zn2+–11 as representatives of group
A, B and C catalysts are shown as a function of the pKa of the
leaving group.
In contrast to the behaviour of Zn2+–8, Zn2+ aquo ions at
pH 5.6 exhibit the most efficient catalysis with the least reactive
substrates as is shown by the results in Table 1. The catalysis
The difference between the catalysts is in their acidity. pKa’s
of 7.534 and 8.028 have been reported for group A catalysts Zn2+–
8 and Zn2+–10, respectively, whereas the pKa values of group B
catalysts are slightly higher. The pKa of the Zn2+ aquo ion is
by Zn2+ on the cleavage of the reactive esters 4e–h is more
aq
modest than that by Zn2+–8, but with 4b and 4c, it is actually
the better of the two catalysts. The catalysis by Zn2+–9
8328 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
9.0.35 The information on the acidity of Zn2+–9 is slightly respectively. Accordingly, at pH 6.6, group A catalysts exist to a
ambiguous: according to Zompa34 and Kimura et al.28 the pKa large extent in their hydroxo form, while with group B catalysts
cannot be determined by potentiometric titration because of the hydroxo form is a minor component and the aquo form
solubility problems. A kinetic value of 9.2 has been proposed predominates. Group C catalysts exist almost entirely in their
on the basis of the pH-dependence of the Zn2+–9 promoted aquo form.
cleavage of HPNP (2).36 Despite a concern for solubility, this
value seems reliable. The shape of the pH-rate profile is
Monomeric Zn2+ complexes as catalysts of the cleavage of
uridine 3′-aryl esters
similar to those obtained with other catalysts, and generally
pKa values obtained from pH-dependence agree well with
values obtained by potentiometric titration. Group C catalysts
are even more basic: pKa values of 9.8 and 9.9 have been
reported for Zn2+–13 and Zn2+–11, respectively.37 The pKa of
Zn2+–12 is not known, but the similarity of its catalytic inactiv-
ity to that of Zn2+–11 suggests that the pKa is of the same
order.38
The pKa values given above refer to 25 °C, but the tempera-
ture dependence of the pKa values is most probably steep. pKa
values of 7.30 and 7.89 have been determined for the deproto-
nation of the water ligand of Zn2+–8 at 25 and 0 °C, respecti-
vely.28 Corresponding pKa values of Zn2+–10 bound water are
8.02 and 8.54. Assuming that the temperature dependence is
linear, pKa values of 5.8 and 6.7 at 90 °C can be calculated for
aquo ligands of Zn2+–8 and Zn2+–10, respectively. The assump-
tion of a linear dependence is very crude, but the large drop in
pKa values is consistent with the steep temperature depen-
dence of the water autoprotolysis constant pKW.39 The 0.3 unit
drop in the pKa value of the Zn2+–9 bound alcohol group as the
temperature increases from 25 to 40 °C is also of the same
order.30
Complexes with a fairly low pKa value exist as equilibrium
mixtures of aquo and hydroxo forms under neutral conditions.
Provided that the estimated pKa values are correct, at pH 6.6
and 90 °C Zn2+–8 exists predominantly in the hydroxo form,
whereas with Zn2+–10 the concentrations of the aquo and
hydroxo forms are approximately equally large. Assuming that
the temperature dependence of the pKa values of other Zn2+
complexes of the same type is similar, pKa values of 7.2
and 8.1 at 90 °C can be estimated for Zn2+–9 and Zn2+–11,
Table 2 presents catalytic data obtained with mononuclear
complexes and uridine 3′-aryl esters 3a–e. As can be seen, the
rate-enhancement by Zn2+
and Zn2+ complexes is very
aq
modest. Zn2+–8 is again a better catalyst than Zn2+–9 or Zn2+
aq
and experiments carried out at a higher catalyst concentration
confirm the rate-enhancing effect. The results show also that
even though the effects are modest, the rate-enhancement
increases as the acidity of the leaving group increases on going
from the phenyl ester 3a to the p-nitrophenyl compound 3e,
which is the total opposite of the situation with uridine
3′-alkyl phosphates. It is possible that the effect is actually
even larger, but the increasing electronegativity of the leaving
group may influence the observed catalytic activity in conflict-
ing ways. While the increasing acidity of the leaving group
enhances the reaction, the affinity of the metal ion towards the
phosphate may decrease. Consistent with this, phenyl phos-
phate has been shown to bind the catalyst Zn2+–8 2.5 times as
strongly as p-nitrophenylphosphate.29
Catalysis by bipyridine and terpyridine complexes of Cu2+
The bipyridine complex Cu2+–14 and the terpyridine complex
Cu2+–15 at 10 mM concentration show good catalytic activity
in the cleavage of alkyl esters, with Cu2+–15 being up to 10
times better as a catalyst than Cu2+–14 (Table 1). The catalytic
activity of Cu2+–15 is approximately the same as that of Zn2+–8
with the more reactive alkyl esters as a substrate (4e–4h; 12 <
pKa(LG) < 15). However, the catalytic activity of Cu2+–15 or
Cu2+–14 does not drop as the acidity of the leaving group
decreases. The higher catalytic activity of Cu2+–15 is inconsist-
ent with the trend observed with Zn2+ catalysts for the pKa of
Table 2 Catalysis of the cleavage of uridine 3’-aryl phosphates 3 by mononuclear Zn2+ and Cu2+ complexes as measured by relative rate constants krel = kobs/k0
UpPh (3a)
Up-pClPh (3b)
Up-oClPh (3c)
UpPhCl2 (3d)
UpPhNO2 (3e)
a
pKa
9.95
0.39 0.1
6.6
26
114
3.65 0.05
1.7
13
2.8
22
0.25 0.02
2.8
10
0.618
15
9.38
1.33 0.02
5.6
36
167
9.1 0.2
1.9
10
4.4
8.48
1.84 0.04
7.6
57
205
11.7 0.4
1.8
11
4.4
46
0.95 0.01
4.6
24
1.92
72
7.51
13.2 0.3
12
108
250
51
2.6
7.14
43.1 0.6
18
116
179
106
5.7
kuncat/10−6 s−1, pH 6.5, 25 °C
krel 2 mM Zn2+–8
krel 10 mM Cu2+–14
krel 10 mM Cu2+–15
kuncat/10−4 s−1, pH 6.5, 90 °C
krel 2 mM Zn2+–9
3
5
krel 2 mM Zn2+–8
krel 10 mM Cu2+–14
krel 10 mM Cu2+–15
31
kuncat/10−6 s−1, pH 5.9, 25 °Cb
0.68 0.02
3.5
17
1.40
32
5.06 0.04
12
51
10.4
99
18.9 0.1
33
58
krel 10 mM Zn2+
b
aq
krel 10 mM Zn2+–8
kuncat/10−5 s−1, pH 7.5, 25 °Ca
krel 10 mM Zn2+–8
a From ref. 40. b Rate constants reported in ref. 25.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8329

View Article Online
Paper
Organic & Biomolecular Chemistry
Cu2+–15 (8.2) is higher than that of Cu2+–14 (7.8).41 A similar could then be explained by changes in the concentration of
difference has been observed before and has been explained monomeric and dimeric species.
by the formation of a catalytically inactive Cu2+–14 dimer
The results obtained with aryl esters (Table 2) show also
under the experimental conditions.41 In comparison with the that the difference in catalytic activity of Cu2+–14 and Cu2+–15
100-fold difference between 10 mM Cu2+–15 and 10 mM Cu2+– decreases as the acidity of the leaving group increases. With
14 in the cleavage of 3′,5′-ApA (6a), the difference observed in the p-nitrophenyl ester 3e, Cu2+–15 and Cu2+–14 are approxi-
the present work is modest, but it most probably results from mately as good catalysts. This is consistent with the results
different conditions. The previously reported 100-fold differ- obtained with HPNP (2),12 where Cu2+–14 is actually a better
ence has been obtained by comparing maximal rate constants catalyst than Cu2+–15 at pH 6.5. This may suggest a change in
obtained at optimal pH, whereas our values refer to a fixed pH the catalytic mechanism as the nucleophilic attack becomes
value of 6.5.
Cu2+–15 is also a fairly good catalyst for the cleavage of
more clearly rate-limiting.
uridine 3′-aryl esters (3a–e), while the catalysis by Cu2+–14 is
again more modest (Table 2). Similar to the catalysis by Zn2+
complexes the rate-enhancing effect increases as the acidity of
the leaving group increases. The increase is slightly more pro-
minent with Cu2+–14, and as a result of this, the difference
between the two catalysts decreases as the acidity of the
leaving group increases. Another curious feature of the cata-
lysis by the Cu2+ complexes is the drop in the activity as the
temperature increases. This is clearly seen when the catalytic
activity of Cu2+–15 is compared to that of Zn2+–8. At 25 °C
Cu2+–15 is even a better catalyst than Zn2+–8, but at 90 °C the
situation is the opposite (note that [Zn2+–8] is lower than
[Cu2+–15].
Catalysis by bimetallic complexes
The catalysis by several dinuclear metal ion complexes was
also studied. Results obtained with alkyl and aryl esters are
presented in Tables 3 and 4, respectively. As the catalysis by
dinuclear complexes is generally more efficient than that
observed with monomeric complexes, the experiments have
been carried out at lower temperatures, where the rate con-
stants of the uncatalysed reactions are available only for the
aryl esters and UpEtCl3 (4h). Unfortunately, data obtained with
UpEtCl3 cannot be extended to other alkyl esters, since, as dis-
cussed above, the proportion of the spontaneous cleavage
increases with less reactive esters, and the temperature depen-
dence of this reaction is not known. Rate-enhancement can
hence be only estimated in most of the cases.
Similar to the situation with monomeric complexes, the
poorest catalysis is observed with UpEtCl3 (4h) as the substrate.
Rate-enhancing effects observed in this case are larger,
consistent with the known activity of bimetallic complexes.
However, none of the other complexes studied in the present
work were even nearly as efficient catalysts as Zn2+2–19a that at
25 °C and pH 6.5 promoted the cleavage of UpEtCl3 by a factor
of 240 000. Even though the rate constants for the uncatalysed
reaction of the other alkyl esters at 25 °C are not known, it can
be deduced that similar to the situation with monometallic
complexes, the catalytic activity increases as the acidity of the
Dimerisation of the catalysts offers most likely an expla-
nation also for the illogical behavior of Cu2+ complexes with
the aryl esters. While most of Cu2+–14 is dimerized under
neutral conditions,41 the Cu2+–15 dimer has been reported as
a minor species under neutral conditions at a total Cu2+–15
concentration of 1.5 mM.42 We can hence assume that under
the experimental conditions of the present study, Cu2+–14
exists mostly as a dimer, but a monomer is the predominant
form of Cu2+–15 at 90 °C. However, the Cu2+–15 concentration
at 10 mM favours the dimerization, and at 25 °C, the pro-
portion of the Cu2+–15 dimer may be more significant. The
changes in the relative catalytic activities at 25 °C and 90 °C
Table 3 Catalysis of the cleavage of uridine 3’-alkyl phosphates 4 by binuclear Zn2+ complexes and a comparison to the activity of corresponding mononuclear
complexes
UpEtCl3 (4h)
kobs/k0 (UpEtCl3)a UpEtCl2 (4g)
UpEtF2 (4f)
UpEtOEt (4e)
b
pKa
12.2
12.9
13.0
14.8
2 mM Zn2+–9 at 50 °C, pH 6.5
2 mM Zn2+–8 at 50 °C, pH 6.5
1 mM Zn2+2–17a at 50 °C, pH 6.5
k (Zn2+2–17a)/k (Zn2+–9)c
(1.06 0.1) × 10−6
(8.8 0.3) × 10−6
2
19
(2.83 0.5) × 10−7
(3.51 0.05) × 10−7 (1.5 0.1) × 10−8
(6.02 0.04) × 10−7
(5.74 0.07) × 10−6
(6.43 0.03) × 10−4 1397
607
(1.65 0.01) × 10−4 (1.96 0.06) × 10−4 (7.5 0.4) × 10−6
583
(4.25 0.05) × 10−7 (5.91 0.07) × 10−7 (4.9 0.7) × 10−8
1.61 1.74 1.02
(3.55 0.05) × 10−7 (4.41 0.04) × 10−7 (3.6 0.2) × 10−8
558
500
1 mM k (Zn2+–17b) at 50 °C, pH 6.5 (1.12 0.01) × 10−6
2
k (Zn2+–17b)/k (Zn2+–9)c
1.06
1 mM (Zn2+2–21b) at 50 °C, pH 6.5
k (Zn2+2–21b)/k (Zn2+–8)c
(1.30 0.01) × 10−6
0.15
3
0.06
0.06
1 mM Zn2+2–22 at 50 °C, pH 7.0
1 mM Zn2+2–19c at 50 °C, pH 6.5
1 mM Zn2+2–19a at 25 °C, pH 6.5e
(1.34 0.04) × 10−4 89d
(4.7 0.4) × 10−5
(5.6 0.3) × 10−5
(3.8 0.1) × 10−4
(3.67 0.06) × 10−5 (1.6 0.2) × 10−6
(3.3 0.5) × 10−4
(5.03 0.06) × 10−3 239 524
717
(5.0 0.3) × 10−5
(4.5 0.1) × 10−4
(1.40 0.03) × 10−6
(4.4 0.4) × 10−6
a Rate constant for the cleavage of UpEtCl3 (4h) in the presence of a catalyst relative to that of the uncatalysed reaction under the same
conditions. k0 values of 2.1 × 10−8 s−1 and 4.6 × 10−7 s−1 at 25 °C and 50 °C and pH 6.5 were calculated as explained in ESI. b From ref. 33. c Rate
constant for the cleavage in the presence of a bimetallic complex relative to that obtained in the presence of the corresponding monomeric
catalyst. d A rate constant of 1.5 × 10−6 for the uncatalysed reaction was calculated assuming a first-order dependence on [HO−]. e From ref. 20.
8330 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 4 Catalysis of the cleavage of uridine 3’-aryl phosphates 3 by binuclear Zn2+ complexes and a comparison to the activity of corresponding mononuclear
complexes
UpPh (3a)
Up-pClPh (3b)
Up-oClPh (3c)
UpPhCl2 (3d)
UpPhNO2 (3e)
a
pKa
9.95
9.38
8.48
7.51
7.14
k0/s−1 pH 6.6 at 25°
k/s−1, 1 mM Zn2+2–19c
krel Zn2+2–19cb
(3.92 0.09) × 10−7
(1.19 0.02) × 10−3
3036
(1.33 0.02) × 10−6
(4.6 0.1) × 10−3
3459
(1.84 0.04) × 10−6
(1.28 0.02) × 10−2
6957
(1.32 0.03) × 10−5
(4.31 0.06) × 10−5
k/s−1, 1 mM Zn2+2–17a
krel Zn2+2–17a
(4.14 0.09) × 10−4
1056
(8.4 0.2) × 10−4
632
(2.15 0.08) × 10−3
1168
(1.58 0.07) × 10−2
1197
k/s−1, 1 mM Zn2+2–22c
krel Zn2+2–22b,d
(9.0 0.1) × 10−5
230
(3.01 0.03) × 10−4
226
(6.5 0.2) × 10−4
353
(7.7 0.4) × 10−3
(1.11 0.7) × 10−2
257
583
k0/s−1 pH 6.5, 50 °Ce
k/s−1, 2 mM Zn2+–9
k/s−1, 2 mM Zn2+–8
k/s−1, 1 mM Zn2+–17b
krel Zn2+–17bb
7.5 × 10−6
2.2 × 10−5
3.0 × 10−5
1.7 × 10−4
4.7 × 10−4
(1.18 0.01) × 10−5
(1.23 0.02) × 10−4
(1.76 0.04) × 10−5
2.3
(4.17 0.08) × 10−5
(3.94 0.05) × 10−4
(5.0 0.3) × 10−5
2.3
(7.3 0.2) × 10−5
(7.4 0.3) × 10−4
(1.05 0.03) × 10−4
3.5
(7.3 0.2) × 10−4
(7.6 0.2) × 10−3
(1.30 0.06) × 10−3
(5.5 0.6) × 10−3
11.7
7.6
1.8
k Zn2+–17b/k Zn2+–9f
k/s−1 Zn2+2–21b
1.5
1.2
1.5
(1.73 0.01) × 10−5
2.3
(5.76 0.05) × 10−5
2.6
(1.06 0.08) × 10−4
3.5
(1.05 0.03) × 10−3
(2.9 0.4) × 10−3
6.2
krel Zn2+2–21bb
6.2
0.14
k Zn2+2–21b/k Zn2+–8f
0.14
0.15
0.14
a From ref. 40. b Rate constant relative to that for the uncatalysed reaction. c Determined at pH 7.0. d Rate constant for the uncatalysed cleavage
calculated assuming a first-order dependence on [HO−]. e Calculated by interpolation on the basis of rate constants determined at 25 and 90 °C.
f Rate constant for the cleavage in the presence of a bimetallic complex relative to that obtained in the presence of the corresponding monomeric
catalyst.
leaving group decreases: the βLG value for Zn2+2–19a promoted that the difference depends on the substrate. With UpEtCl3
reaction is less negative than that for the uncatalysed reaction (4h) as a substrate the difference between the catalytic
under the experimental conditions.20 Assuming that the shape activity of dimeric Zn2+2–17a and monomeric Zn2+–9 is 600-
of the log k vs. pKa plot for the uncatalysed reaction is the fold, but it seems to decrease as the acidity of the leaving
same at 25 °C and 90 °C, a nearly 106 fold rate-enhancement group decreases. For the cleavage of aryl esters, rate constants
by Zn2+2–19a could be expected for the cleavage of UpEtOEt at the same temperature are not available, but assuming that
(4e). This is consistent with our previous estimation of monomeric Zn2+–9 is equally inactive at 25 °C and 90 °C, the
approximately 106-fold rate-enhancement with 3′,5′-UpU (6b) dimeric complex Zn2+2–17a is approximately 400–500 times as
as a substrate.19 Complex Zn2+2–19c, the non-amino analogue efficient a catalyst for the cleavage of the aryl esters as Zn2+–9
of Zn2+2–19a, and complex Zn2+2–17a showed a more modest is. The difference decreases as the acidity of the leaving group
activity: at pH 6.5 and 50 °C, rate-enhancements of 717 and increases. In other words, the difference between the dimer
1400 were observed, respectively (Table 3).
and the monomer decreases as the nucleophilic attack
The catalysis on the cleavage of uridine 3′-aryl esters is becomes more clearly rate-limiting. Consistent with this, the
again fairly modest, but in this case the dependence on the difference between the dimer and the monomer in the reaction
acidity of the leaving group is not quite as obvious. The cata- of HPNP (2) has been reported to be 120-fold.14
lysis by the most efficient catalyst Zn2+2–19a could not be
A comparison of the results obtained with complex Zn2+
–
2
studied because the reactions were too fast to follow using 17a and its analogue with ligand 17b shows that in this case
HPLC even at 25 °C. The complex Zn2+2–19c, a less active ana- the formation of a bimetallic complex depends on the hydroxyl
logue, showed a fairly good catalysis (Table 4). In this case, the group on the linker: the complex with 17b without the hydroxo
catalytic activity seems to increase as the acidity of the leaving group is catalytically approximately as active as the monomer.
group increases, but no firm conclusions on a trend should be In this case the lack of enhanced catalysis is to be attributed to
drawn on the basis of just three data points. Unlike the situ- coordination chemistry of the ligand: the ligand binds only
ation with alkyl esters, the catalytic activity of the complex one Zn2+ ion.43 The results obtained in the present study show
Zn2+2–17a is poorer than that of Zn2+2–19c: an approximately also that linking two active catalysts with the hydroxyl contain-
1000-fold rate enhancement by Zn2+2–17a was observed with ing linker does not necessarily result in an efficient dinuclear
no clear dependence on the acidity of the leaving group. catalyst. A surprisingly poor performance was shown by the
The results obtained with the complex Zn2+2–22 suggest a Zn2+–8-based dimer Zn2+2–21b whose catalytic activity was
drop in catalytic activity with the most reactive arylphosphate approximately only one-tenth of that of the Zn2+–8. In contrast
substrate 3e.
to results obtained in aqueous solutions, Zn2+2–21b catalyses
Results obtained with the complex Zn2+2–17a confirm the the cleavage of 2 in methanol, but even then the catalytic
earlier reports according to which the dimer is a significantly effect is very modest in comparison to that obtained with
better catalyst than monomeric Zn2+–9.14 Our results show also Zn2+2–21a.44
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8331

View Article Online
Paper
Organic & Biomolecular Chemistry
Catalysis on the isomerization of phosphodiester bonds
catalysis on isomerisation is observed19,20 only with catalysts
with particularly strong interactions with the phosphorane,
such as Zn2+2–19a.16
Several complexes were also tested as catalysts for the isomeri-
zation of phosphonate substrates (7). As reported previously,20
the complex Zn2+2–19a promotes also the isomerization of the
phosphodiester bond. The rate enhancement with other com-
plexes was, however, very modest. At 25 °C the rate constant of
the Zn2+2–19a promoted isomerization of uridine 3′-methyl
phosphonate (7b) was 2.1 × 10−7 s−1, whereas a rate constant
of 0.15 × 10−7 s−1 was obtained for the reaction in the presence
of 1 mM Zn2+2–19b. Zn2+2–17a (1 mM) showed no detectable
activity in one month. Assuming that a 1% change in the con-
centration of the starting material could be reliably observed, a
limiting value of 4 × 10−9 s−1 can be estimated.
Another conclusion that can be drawn on the basis of the
results discussed is that there is no universal catalysis mechan-
ism covering all the substrate–catalyst combinations, but the
mechanism depends on the substrate and possibly also on the
catalyst utilized. We suggest that minimum catalysis observed
for the cleavage of UpEtCl3 (4h) is mainly of electrostatic
nature: the catalysts bind the phosphate, enhance thereby the
nucleophilic attack, stabilize the phosphorane and enhance
thereby the overall reaction. The fact that catalysis is almost
non-existent for monometallic complexes, even for the most
efficient among them, is not surprising considering the low
proportion of Zn2+-bound substrate molecules under the
experimental conditions. According to Koike and Kimura,29
the binding constant for complexation between Zn2+–8 and
phosphodiesters cannot be accurately determined, but a limit-
ing log K value of 0.5 has been proposed. Using this value, it
can be estimated that not more than 3% of the substrate mole-
cules are bound as Zn2+–8 complexes in our experiments at
room temperature. In fact, this is a crude estimate and several
factors, such as temperature, pH and the structure of the sub-
strate, influence the binding.
Zn2+2–19c promoted isomerization of 7b was studied at
50 °C, where a 3-fold rate enhancement was observed in the
presence of a 1 mM complex. Monomeric Zn2+–9 and Zn2+–8,
as well as Cu2+–14 and Cu2+–15, showed a barely observable
catalytic activity at 90 °C; 1.5-fold rate-enhancements were
obtained in the presence of a 10 mM complex. Increasing pH
had no observable effect on the isomerization in the presence
of Cu2+–15, whereas increasing complex concentration slightly
enhanced the reaction.
Discussion
Catalysis of the cleavage when the departure of the leaving
group is rate-limiting
The results discussed above can be understood by considering
a step-wise transesterification reaction, where the pentacoordi-
nated phosphorane is an intermediate even in the presence of
a metal ion catalyst. As was described above, the existence of
an intermediate in the uncatalysed reaction has been proven
by a Brønstedt plot with a breakpoint at a pKa of 12.6, where
the pKa of the leaving group equals that of the attacking
nucleophile.10,11 At this point the energy profile for the clea-
vage is symmetrical, and as the energy minimum is probably
very shallow, a nearly concerted reaction can be proposed.
According to our results, this is also approximately the same
point where the poorest catalytic activity was observed. With
any catalyst studied, the lowest rate-enhancement was
observed with UpEtCl3 (4h)‡. However, as the acidity of the
leaving group either increases or decreases, the energy profile
becomes increasingly asymmetrical. At the same time the
observed catalytic activity increases, as we have experimentally
shown. Accordingly, metal ion catalysts can enhance both
steps of the cleavage reaction. The lack of catalysis on the iso-
merization reaction in the presence of most metal ion com-
plexes most probably results from the instability of the
phosphorane. Similar to alkaline cleavage, a dianionic phos-
phorane is formed. Apparently, in most of the cases, the inter-
actions with the catalyst are not strong enough for a metal ion
to act as a substitute for a proton neutralizing the charge. The
phosphorane is, hence, too unstable to pseudorotate, and
The pKa values show that the complexes exist as equilibrium
mixtures of hydroxo and aquo forms, and the proportion of
these forms is different for different complexes. General acid
and base catalysis is often proposed for the metal ion pro-
moted reactions of RNA model compounds: a hydroxo ligand
can act as a general base that deprotonates the attacking
nucleophile (Scheme 3a), whereas an aquo ligand can facilitate
the reaction by protonating the leaving group oxygen
(Scheme 3b). A sequential general base–general acid catalysis
(Scheme 3c) is also possible in the presence of complexes with
both hydroxo and aquo forms present under experimental con-
ditions. As mentioned in the Results section, three different
catalyst groups were identified on the basis of experimental
data and they are characterized by different acidity and cata-
lytic properties. Group A complexes are acidic with a high pro-
portion of the hydroxo form. They are good catalysts as long as
protonation of the leaving group is not essential, but the cata-
lytic activity decreases abruptly for substrates with more basic
leaving groups as is shown with Zn2+–8 in Fig. 1. This behavior
is easily attributed to the reaction route involving general base
catalysis only (Scheme 3a). Similarly, the very modest catalysis
by complexes belonging to group C, only observed with the
most basic leaving groups, can be explained by the reaction
route involving only general acid catalysis (Scheme 3b). Appar-
ently, only the most basic leaving groups are able to abstract a
proton from these inefficient catalysts. Group B catalysts show
a fair and consistently increasing catalytic activity as the
leaving group becomes poorer. As both aquo and hydroxo
forms are available (though the aquo form is predominant),
‡It is to be noted that the pKa of 12.6 indicated in ref. 10 and 11 refers to kine-
tics at 25 °C, whereas our experiments with monometallic complexes were
carried out at 90 °C, where the pKa’s can be expected to be lower.
8332 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 3 a. General base catalysis by a coordinated metal complex in the hydroxo form on the nucleophilic attack of 2’-OH on the phosphate. b. Possible reaction
routes involving general acid catalysis by a coordinated metal complex in the aquo form for the departure of the leaving group. c. Sequential general base–general
acid catalysis by a coordinated metal complex.
Scheme 4 Catalysis mechanisms showing the metal ion complex as an electrophilic catalyst.
they can be proposed to act as bifunctional catalysts following value for spontaneous cleavage.25 Group A complexes with
the general base–general acid route in Scheme 3c. Quite poss- only hydroxo ligands, such as Zn2+–8, cannot provide the
ibly the system is a continuum with a gradual change from required proton and hence a drop in the catalytic activity is
one mechanism to another depending on the acidities of the observed.
aquo ligand and the leaving group alcohol.
General base catalysis on the nucleophilic attack is more
The general acid catalysis on the departure of the leaving controversial. Even though the catalytic activity of the com-
group is easy to confirm, for it is clearly seen in the results: for plexes on the more reactive alkyl esters (12 < pKa < 15), where
the least reactive esters (4a, b), catalysts with higher proportion protonation of the leaving group is not essential, seems to cor-
of the acidic aquo form, such as group B catalysts Zn2+ and relate with the proportion of the hydroxo form, the situation is
aq
Zn2+–9, retain their catalytic activity, whereas the catalytic by no means clear; a pKa value is also a measure of the electro-
activity of group A catalysts decreases abruptly (Fig. 1). As the philicity of a metal ion center. Electrophilic catalysis would
analysis of the uncatalysed reaction showed, these are sub- enhance the reaction by assisting the nucleophilic attack on
strates that under the experimental conditions react through a the phosphate. Even if the second step was rate limiting, the
mechanism that involves a concerted proton transfer to the reaction would be enhanced, because the equilibrium concen-
leaving group. It is hence a logical conclusion that metal ion tration of the phosphorane intermediate would increase.
complexes with an aquo ligand provide the required proton, Experimental evidence could, hence, be explained also by
assisting hence the cleavage of these unreactive esters. This is mechanisms shown in Scheme 4. The mechanism described is
consistent with the previously reported βLG value of −0.32 for essentially metal ion catalysis on the alkaline cleavage. The
Zn2+ promoted reaction that is even less negative than the fact that the catalysis decreases when the proportion of the
aq
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8333

View Article Online
Paper
Organic & Biomolecular Chemistry
alkaline cleavage increases argues, however, against such complex Zn2+2–19c.49 At a qualitative level the results obtained
a mechanism. Furthermore, a metal complex in its hydroxo with monometallic complexes suggest that bell-shaped depen-
form could be expected to be a poorer electrophilic catalyst dence is observed with more basic complexes, whereas a
than the aquo form because of the less positive net charge. sigmoidal dependence refers to catalysis by more acidic com-
The reaction system is, however, complex and these arguments plexes. In the case of Zn2+–8 promoted cleavage of poly-U, the
cannot be taken as conclusive evidence against electrophilic rate increase levels off at around pH 6.6. On the basis of the
catalysis.
discussion on temperature dependence of the pKa values, it
As has been pointed out,21 electrophilic catalysis by the seems that this is close to the pKa of the aquo ligand. Consist-
metal catalyst aquo form on the alkaline cleavage and general ent with different catalysis mechanisms, Yashiro et al.49,50
base catalysis of the cleavage of a monoanionic substrate are have compared the pH-rate profiles of the cleavage of 3′,5′-ApA
kinetically equivalent and difficult to distinguish from each (6a) to species distribution curves. According to their analysis
other. As will be discussed later, several reports supporting the the bell-shaped dependence is observed when the catalyst is
electrophilic catalysis have been published. Most of these, present in a form with both an aquo and a hydroxo ligand, as
however, concentrate on reactions where the nucleophilic with Zn2+2–23a and Zn2+2–23b, whereas a sigmoidal depen-
attack is the rate limiting step. We believe that at least in cases dence refers to the presence of a species with only hydroxo
where the departure of the leaving group is rate limiting, ligands, such as Zn2+2–19c.
general base catalysis is more feasible as a part of the catalytic
system than electrophilic catalysis. This suggestion is based on
results obtained with Zn2+–8 catalyzed cleavage of phosphodi-
ester bonds in short oligonucleotides. The results show clearly
that Zn2+–8 binds only one phosphate group, but this phos-
phate is not necessarily the scissile phosphodiester bond.45 As
evidence of this, phosphodiester bonds in 3′-phosphorylated
oligonucleotides are cleaved more rapidly than oligonucleo-
tides with no such good coordination site.
On the basis of the observations discussed above, we
propose, hence, a general acid–base catalyzed reaction system
Catalysis on the cleavage when the nucleophilic attack is rate
limiting
for the cleavage of nucleoside 3′-alkyl esters where the depar- The results obtained with nucleoside 3′-aryl esters show that
ture of the leaving group is rate determining (Scheme 3a–c). In the catalysis by monometallic complexes is very modest, but it
principle, both general acid and base catalysis are possible, seems to increase, as the acidity of the leaving group increases.
but the preference depends on the acidity of metal aquo This can again be explained by a step-wise mechanism, where
ligand and on the acidity of the leaving group. It would be the catalysis increases as the difference between the energy
logical to assume that there is no clear breakpoint, but a barriers of the individual steps increases. As we proposed a
gradual change from general base catalysis to general acid cata- general acid–base catalysis (Scheme 3a–c) for the reactions of
lysis through systems where both are involved. This suggestion substrates with a poor leaving group, a general base catalysis
is consistent with pH-rate profiles of metal ion promoted (Scheme 3a) would be a logical choice here, but the situation
cleavage of RNA substrates (pKa of 14.4 has been determined is not straightforward. In addition to general base catalysis
for 3′,5′-UpU (6b) by kinetic experiments under alkaline con- (Scheme 5b),26,27 specific base catalysis of the cleavage of
ditions20). Both bell-shaped and sigmoidal pH-rate profiles the metal bound substrate (Scheme 5a)21,22 and nucleophilic
have been observed. The cleavage of a dinucleoside dipho- catalysis (Scheme 5c)30 have all been proposed for metal ion
sphate by Zn2+–9,46 the cleavage of oligonucleotides by Cu2+– promoted cleavage of RNA models. Sigmoidal pH-rate pro-
1547,48 and the cleavage of dinucleoside monophosphates by files12,21,26,30,52,53 and a correlation between a pH-rate profile
bimetallic complexes Zn2+2–23a49 and Zn2+2–23b50 are charac- and a species distribution curve26,27 have been attributed to a
terized by bell-shaped pH-dependence. Sigmoidal pH-rate kinetically significant deprotonation of a metal bound water
profiles have been observed with the cleavage of poly-U in the ligand. For a wider perspective, studies with different sub-
presence of Zn2+–8,51 the cleavage of UpU by bimetallic strates are discussed below together with our own experi-
complex Zn2+2–17a21 and the cleavage of ApA (6a) by bimetallic mental results.
Scheme 5 a. Electrophilic + specific base catalysis. b. General base catalysis. c. Nucleophilic catalysis.
8334 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Catalysis on the cleavage of non-nucleosidic substrates with an of
aryl leaving group
2
in the presence Zn2+–9, Zn2+–8 and Zn2+–10,
respectively.14,53
Results obtained with Cu2+–15 and Cu2+–14 as catalysts
On the basis of their extensive research, Morrow and coworkers support also the suggestion that the reaction of 2 could involve
have considered two kinetically equivalent mechanisms, intra- nucleophilic catalysis by the coordinated metal ion. These
complex general base catalysis by a metal hydroxo ligand catalysts show a very different preference for substrates. As was
(Scheme 5b) and specific base catalyzed cleavage of a co- discussed earlier, Cu2+–15 is a better catalyst for the cleavage
ordinated substrate (Scheme 5a), and have ended up favouring of uridine 3′-alkyl esters (4) and dinucleoside monophosphate
the latter.21 This mechanism is supported by the absence of a 6b than Cu2+–14 is,40 but with cyclic monophosphate 5b13 and
significant solvent isotope effect in the Zn2+2–17a promoted 212 the situation has been reported to be the opposite. While
cleavage of UpPhNO2 (3e)22 and the inhibition of the catalysis the dimerisation of Cu2+–15 can explain the inactivity observed
by metal ion complexes observed upon addition of phosphate with nucleoside 3′-alkyl phosphates, the fact that with simpler
anions.53,54 According to the results obtained, the strongest systems Cu2+–14 actually is a better catalyst suggests that the
inhibition of the HPNP (2) cleavage is observed at a lower pH catalysis mechanisms required are different. Results reported
where the catalyst is present in the protonated form, suggesting in the literature show that Cu2+–14 is an active catalyst in reac-
that this is the species involved in the catalysis.
tions where the nucleophile is a Cu2+ bound hydroxo ligand,
In contrast, Bonfa et al.30 have interpreted the sigmoidal such as in the cleavage of 1, where Cu2+–15 is inactive as a
pH-dependence and the absence of a significant solvent catalyst.58 Cu2+–14 is also a far better catalyst than Cu2+–15
isotope effect in the cleavage of 2 as evidence for nucleophilic for cleavage of dinucleoside oligophosphates, where an intra-
catalysis, i.e. direct interaction of the attacking nucleophile molecular nucleophilic attack is not possible.59 This indirect
and the metal ion catalyst (Scheme 5c). A large drop in the evidence suggests that in cases where Cu2+–14 is a better
solvent isotope effect was observed in the presence of the catalyst than Cu2+–15, a nucleophilic catalysis should be
metal ion catalyst. While the cleavage of 2 in the absence of a considered.
catalyst was characterized by a solvent isotope effect of 4.01
reflecting the difference in equilibrium deprotonation of the
attacking nucleophile, the value observed in the presence of a
Catalysis on the cleavage of nucleosidic substrates with an aryl
leaving group
mononuclear Zn2+ complex was 1.43. The difference suggests While nucleophilic catalysis (Scheme 5c) may be a feasible cata-
that the catalyst has a more active role than just that of a Lewis lysis with 2, with nucleoside based phosphodiesters it most
acid. However, according to Morrow and coworkers55 heavy probably is of minor importance. Theoretically thinking, the
atom isotope effects are more in favour of a reaction where flexible structure of 2 might be able to allow the simultaneous
there is no interaction between the nucleophilic OH and the interaction to the phosphate and to the attacking nucleophile.
metal ion catalyst.
With nucleoside based phosphodiesters this could be expected
Nucleophilic catalysis for the cleavage of 2 receives further to be more difficult because of the more rigidly positioned
support from the similarity of the metal ion promoted reac- nucleophile and constraints posed by the ribose ring. The
tions of 2 and BNPP (1). Even though 1 and 2 are fundamen- results obtained in the present study may be taken as experi-
tally different in that one reacts by intramolecular mental evidence for the suggestion of different mechanisms:
transesterification and the other is cleaved by hydrolysis invol- with all the phosphodiesters studied Cu2+–15 is a more
ving an intermolecular nucleophile, the basic kinetic obser- efficient catalyst than Cu2+–14 is, while the opposite is true
vations for metal ion promoted reactions are similar: the with 2 as a substrate. Even if the suggestion of nucleophilic
reactions of BNPP are generally characterized by very clearly catalysis for the cleavage of 2 is not correct, the different pre-
sigmoidal pH-rate profiles28,29,56,57 and a hydroxo species has ferences suggest different catalysis mechanisms for different
been identified as the catalytically active form. In the case of 1 substrates. The fact that the difference in the catalytic activity
the hydroxo ligand acts as the attacking nucleophile. An essen- between Cu2+–15 and Cu2+–14 decreases as the acidity of an
tial observation is that with both 1 and 2 the catalytic activity aryl leaving group decreases suggests that there may be a
of metal ion complexes depends not only on the acidity of an slight change in the preferred catalysis mechanism on going
aquo ligand, but also on the coordination geometry.29,30 The to more reactive nucleoside 3′-aryl esters.
analysis by Bonfa et al.30 shows that while logarithmic rate
As mentioned above, specific base catalyzed cleavage of the
constants obtained with complexes of different tridentate metal bound substrates (Scheme 5a) has been preferred over
ligands depend linearly on the pKa, those obtained using tetra- the kinetically equivalent mechanism, where a metal bound
dentate ligands fall below the line. The trend is more pro- hydroxo ligand acts as a general base catalyst (Scheme 5b).
nounced with 1 than with 2, but the difference between the The absence of a significant solvent isotope effect on the clea-
tridentate and tetradentate complexes can be seen with both vage of UpPhNO2 (3e) has been taken as an indication of a
129,30 and 2.30,53 Zn2+–10 (pKa 8.028) is in fact even poorer as a mechanism where the catalysis is based merely on the
catalyst for the cleavage of 2 than Zn2+–9 (pKa 9.2)36 is. increased electrophilicity of a phosphate group upon coordi-
Second-order rate constants of 2.1 × 10−3 M−1 s−1, 0.016 M−1 nation.22 However, as has been pointed out above, the solvent
s−1 and 6.3 × 10−4 M−1 s−1 have been reported for the cleavage isotope effect allows different interpretations. When the
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8335

View Article Online
Paper
Organic & Biomolecular Chemistry
reactions are carried out at the same pL, the pre-equilibrium Catalysis by bimetallic complexes
deprotonation in the alkaline cleavage is observed as a signifi-
cant difference in rate constants in H2O and D2O.32 The The catalysis by bimetallic complexes follows the same trends
results obtained in Zn2+–17a catalyzed cleavage of UpPh (3a) as those observed with monometallic complexes: The catalysis
show that a normal SIE is observed until pL 8.5; above pL 8.5 by any metal complex is the weakest with UpEtCl3, and it
the reaction in D2O is slightly faster (kH/kD = 0.8).22 Assuming increases as the asymmetry of the energy profile becomes
that the kinetically observed pKa refers to deprotonation of the more pronounced. As numerous reports have shown before,
catalyst, kH/kD of 0.8 refers to conditions where the catalyst is the catalysis by bimetallic complexes can be significantly more
in its lyoxo form both in H2O and in D2O. If the reaction efficient than that by their monometallic counterparts, and
involved a specific base catalysis on the metal bound substrate, there is no doubt that enhanced interactions play a significant
equilibrium deprotonation of the substrate HO was the only role in the enhanced catalysis. The effect has been described
proton transfer process, and the same equilibrium isotope in different terms: simultaneous interaction with two posi-
effect of approximately 5 should be observed. In contrast to tively charged metal ions increases the equilibrium concen-
this, intracomplex general base catalyzed reaction might well tration of bound substrate molecules, the double Lewis acid
be characterized even by an inverse solvent isotope effect. The activation makes the phosphate increasingly electrophilic and
general base catalysis involves a proton transfer from the the negatively charged phosphorane is more strongly stabilized
2′-OH which is characterized by a normal isotope effect, but by the enhanced interactions. Williams and his coworkers
this is compensated for by the inverse effect resulting from the have previously shown that the interactions can be further
protonation of the catalyst. This is consistent with the very enhanced by hydrogen bonding.16,17,61 In the absence of the
modest solvent isotope effects observed for the reactions of hydrogen bonding amino groups Zn2+2–19c binds less strongly,
phosphodiesters under neutral and acidic conditions, which and a less efficient catalysis is observed: Zn2+2–19a is 726
involve intramolecular proton transfer reactions.32 The normal times as efficient a catalyst for the cleavage of HPNP as Zn2+
–
2
solvent isotope effect observed at a lower pH can be explained 19c.16 The results obtained in the present work fit well with
by the concentration of the hydroxo form of the complex, this: comparing the results obtained with Zn2+2–19a at 25 °C
which is lower in D2O under conditions where the deprotona- and with Zn2+2–19c at 50 °C, a 370-fold difference in catalytic
tion is incomplete.
activity can be calculated.
The discussion above does not offer a clear-cut answer to
The efficiency of the binding is determined by the shape of
the question about the mechanism of the metal ion promoted the complex and consequently by the structure of the ligand
cleavage of nucleoside 3′-aryl phosphates. Three different and the coordination geometry of metal ions. Morrow and her
mechanisms have been proposed, and often the same experi- co-workers have shown that in the case of Zn2+2–17a, the
mental evidence has been taken to support two different linker hydroxyl group is involved in interactions with the two
mechanisms: pH-rate profiles and solvent isotope effects Zn2+ centers.14 Apparently, this interaction induces a structure
allow different interpretations. Similarly, both general base that allows a productive binding of a substrate with two metal
catalysis (Scheme 5b)60 and specific base catalysis ions and an efficient catalysis. In the absence of the linker
(Scheme 5a)23,24 as mechanisms of metal complex promoted alkoxy group, the catalytic advantage of a bimetallic system is
cleavage of 2 have been supported on the basis of theoretical lost nearly completely. In the case of ligand 17b the inactivity
calculations.
results from the fact that the ligand forms only 1 : 1 complexes
As the discussion on the reactions of nucleoside 3′-alkyl with Zn2+.43 The results obtained in the present study show
esters showed, there is most probably not a single mechanism also that linking two active catalysts with the hydroxyl contain-
covering all the substrate–catalyst combinations studied. It is, ing linker does not necessarily result in an efficient dinuclear
hence, quite possible that this is the case also with aryl esters catalyst. The catalytic activity of the Zn2+–8 -based dimer
(nucleoside derivatives and 2). Particularly, when the nature of Zn2+2–21b was only approximately one-tenth of that of the
the substrates is different, caution should be exercised when Zn2+–8. It could be tentatively suggested that intracomplex
results obtained with one type of substrate are extended to the interactions prevent the binding to the substrate. The impor-
reactions of another type of compound. The possibility of tance of the correct coordination geometry is shown for
different catalysis mechanisms utilized by different complexes example by the fact that a Cu2+ complex, Cu2+2–17a, has been
should also be considered. As mentioned above, we are reported to be nearly inactive as a catalyst for the cleavage
inclined to believe that metal ion promoted reactions of of 2.14
nucleoside 3′-aryl esters generally involve a general base cata-
While the complexes studied in the present work utilize the
lysis by a hydroxo ligand of the coordinated metal ion. While linker with a hydroxyl ligand to achieve an overall structure
the experimental evidence obtained with these substrates is favouring the productive substrate binding, this is by no
fairly limited, none of it argues explicitly against the general means the only strategy. Several ligands with a more rigid aro-
base catalyzed reaction. The mechanism of reactions of HPNP matic linker such as 18,15 or 23a49 and 23b50 have been
may be different, at least with some catalysts, and results reported to form efficient bimetallic catalysts. Rigidity in
obtained with this substrate should not be used as evidence ligand 16 is brought about by the amide linker.13 The inter-
against a given mechanism for another substrate.
actions between the catalyst and the substrate can also be
8336 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
enforced by using a less polar medium. Zn2+2–21b shows a par- different reaction mechanisms are actually probable, and
ticularly large rate-enhancing effect in methanol or ethanol, because of this, caution should be exercised when any results
whereas those with aromatic linkers seem to be less prone to are extended to another system, no matter how closely it is
the enhancing effect of alcohol solvent.44
related. The same problems are encountered in the design
The discussion above emphasises the importance of of bifunctional catalysts: the catalytic activity of a monomer
enhanced binding, but it is also clear that, similar to catalysis together with design that is successful elsewhere is not a
by mononuclear complexes, the bimetallic catalysts play a guarantee of an efficient bimetallic catalyst.
more active role when the acidity of the leaving group
increases. The difference in catalytic activity between Zn2+
–
2
19a and Zn2+2–19c, for example, decreases, as the leaving
group becomes poorer. We have previously proposed a general
acid catalysis on the basis of solvent isotope effect values
ranging from 2 to 3.20 As has been pointed out above, the
values are clearly different from those of the uncatalysed reac-
tions and hence consistent with a catalyst induced proton
transfer process, even if the interpretation is not straight-
forward at all with such a complicated system. The decrease in
the difference between the catalytic activities of the Zn2+–9
monomer and the bimetallic complex Zn2+2–17a can most
probably be attributed to different pKa values: a value of 7.852
determined for Zn2+2–17a is clearly lower than the value of
9.235 estimated for the monomer. Therefore Zn2+2–17a is likely
to be a general base catalyst, whereas Zn2+–9 with an aquo
ligand under neutral conditions may act as a general acid cata-
lyst as well.
Experimental
Synthetic procedures
Phosphodiester substrates and ligands were synthesized using
known methods. Synthesis and characterization are described
in detail in the ESI.†
Kinetic measurements
The pH of the reaction solutions was adjusted using MOPSO
[3-(N-morpholino)-2-hydroxypropanesulphonic acid]. pKa
values of MOPSO at higher temperatures were calculated using
the data found in the literature62 and NaOH was used to adjust
the buffer ratio. Reactions were carried out in Eppendorf tubes
at 25 and 50 °C and in glass tubes at 90 °C. The temperature
was controlled using a thermostated water bath. Aliquots were
withdrawn at suitable intervals and were kept in an ice bath
until the analysis using HPLC. An excess of EDTA was added to
samples to quench metal ion catalyzed reactions.
Analysis was carried out with RP HPLC using a Waters
Atlantis™ dC18 column (150 × 4.6 mm, 5 µm particle size) or a
Supelcosil™ LC-18 column (250 × 4 mm, 5 µm particle size).
Mixtures of acetic acid buffer ([AcOH] = 0.045 M, [AcONa] =
0.015 M, [NH4Cl] = 0.1 M]) and acetonitrile were used as
eluents. Isocratic elution with 5–20% acetonitrile or gradient
elution (0 → 20% acetonitrile) was used. With the methylphos-
phonate compound 7b, an isocratic elution with 0.025 M
Conclusions
Transesterification of uridine 3′-phosphodiesters in the pres-
ence of metal ion catalysts can be understood as a two-step
reaction, where phosphorane is an intermediate. Metal ion cata-
lysts can enhance both the nucleophilic attack and the depar-
ture of the leaving group. No universal catalysis mechanism
exists, but the mechanism depends on both the catalyst
and the substrate. A change from general base catalysis
(Scheme 3a) to general acid catalysis (Scheme 3b) through
bifunctional general base–general acid catalysis (Scheme 3c) is
suggested as the acidity of the metal aquo ligand and of the
leaving group alcohol decrease. Catalysis on the isomerization
depends on the strength of the binding: strong interactions
stabilize the phosphorane allowing the pseudorotation.
Pseudorotation is, however, rate-limiting, and only a modest
catalysis is observed.
triethylammonium acetate containing 0.2
M tetramethyl-
ammonium chloride was used. UV detection was done at
260 nm.
Rate constants of the cleavage were calculated by following
the decrease of the signal area of the substrate and by applying
the integrated rate-law of first order reactions. The pKa values
of the leaving groups were drawn from the literature.33,40 The
pKa of neopentanol (2,2-dimethylpropan-1-ol) was determined
by comparing the reactivity of 4a and a series of alkyl esters in
1.00 M NaOH at 25 °C, where the rate of the cleavage is
reported to be strongly dependent on the nature of the leaving
group, βLG being −1.28.31 At 1.00 M NaOH and 25 °C, 4a was
cleaved at a rate of (0.46 0.03) × 10−6 s−1 which refers to a
pKa value of 17.3.
Catalytic advantage achieved with bimetallic complexes
depends on interactions within the complex and interactions
between the catalyst and the substrate. Different strategies can
be used to induce the cooperative binding with two metal
ions. The basic catalytic mechanisms utilized by monometallic
and bimetallic complexes are the same.
The results discussed underline the difficulty of mechanis-
tic research: any change in reaction conditions may change
several parameters at the same time, and all changes should
be considered. Furthermore, the kinetic data obtained with
different substrates are often similar, but they can be inter-
preted in different ways. As the discussion above shows,
References
1 J. Weston, Chem. Rev., 2005, 105, 2151–2174.
2 T. Niittymäki and H. Lönnberg, Org. Biomol. Chem., 2006,
4, 15–25.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8337

View Article Online
Paper
Organic & Biomolecular Chemistry
3 J. R. Morrow, T. L. Amyes and J. P. Richard, Acc. Chem. Res., 29 T. Koike and E. Kimura, J. Am. Chem. Soc., 1991, 113, 8935–
2008, 41, 539–548.
8941.
4 H. Lönnberg, Org. Biomol. Chem., 2011, 9, 1687–1703.
5 R. S. Brown, Z. Lu, T. C. Liu, W. Y. Wang, D. R. Edwards
and A. A. Neverov, J. Phys. Org. Chem., 2010, 23, 1–15.
6 D. Desbouis, I. P. Troitsky, M. J. Belousoff, L. Spiccia and
B. Graham, Coord. Chem. Rev., 2012, 256, 897–937.
30 L. Bonfa, M. Gatos, F. Mancin, P. Tecilla and U. Tonellato,
Inorg. Chem., 2003, 42, 3943–3949.
31 M. Kosonen, E. Yousefi-Salakdeh, R. Strömberg and
H. Lönnberg, J. Chem. Soc., Perkin Trans. 2, 1998, 1589–
1595.
7 H. Korhonen, N. H. Williams and S. Mikkola, J. Phys. Org. 32 N. Virtanen, L. Polari, M. Välilä and S. Mikkola, J. Phys.
Chem., 2013, 26, 182–186. Org. Chem., 2005, 18, 385–397.
8 M. Oivanen, S. Kuusela and H. Lönnberg, Chem. Rev., 1998, 33 E. P. Serjeant and P. Dempsey, Ionization constants of
98, 961–990.
organic acids in aqueous solutions. IUPAC Chemical data
9 F. H. Westheimer, Acc. Chem. Res., 1968, 1, 70–78.
series 23, Pergamon Press, Oxford, 1979.
10 H. Lönnberg, R. Strömberg and A. Williams, Org. Biomol. 34 L. J. Zompa, Inorg. Chem., 1978, 17, 2531–2536.
Chem., 2004, 2, 2165–2167. 35 D. D. Perrin, J. Chem. Soc., 1962, 4500–4502.
11 M. Padovani, N. H. Williams and P. Wyman, J. Phys. Org. 36 O. Iranzo, A. Y. Kovalevsky, J. R. Morrow and J. P. Richard,
Chem., 2004, 17, 472–477. J. Am. Chem. Soc., 2003, 125, 1988–1983.
12 S. Liu and A. D. Hamilton, Bioorg. Med. Chem. Lett., 1997, 37 T. Itoh, Y. Fujii, T. Tada, Y. Yoshikawa and H. Hisada, Bull.
7, 1779–1784. Chem. Soc. Jpn., 1996, 69, 1265–1272.
13 S. Liu, Z. Luo and A. D. Hamilton, Angew. Chem., Int. Ed. 38 S. Kuusela and H. Lönnberg, J. Phys. Org. Chem., 1992, 5,
Engl., 1997, 36, 2678–2680. 803–811.
14 O. Iranzo, T. Elmer, J. P. Richard and J. R. Morrow, Inorg. 39 H. S. Harned and B. B. Owen, The physical chemistry
Chem., 2003, 42, 7737–7746.
of electrolytic solutions, Chapman and Hall Ltd, London,
15 M. J. Young and J. Chin, J. Am. Chem. Soc., 1995, 117,
10577–10578.
16 G. Feng, J. C. Mareque-Rivas, R. Torres Martin de Rosales
1958.
40 A. M. Davis, A. D. Hall and A. Williams, J. Am. Chem. Soc.,
1988, 110, 5105–5108.
and N. H. Williams, Angew. Chem., Int. Ed. Engl., 2005, 45, 41 B. Linkletter and J. Chin, Angew. Chem., Int. Ed. Engl., 1994,
7056–7059. 34, 472–474.
17 G. Feng, J. C. Mareque-Rivas and N. H. Williams, Chem. 42 L. A. Jenkins, J. K. Bashkin, J. D. Pennock, J. Florian and
Commun., 2006, 1845–1847. A. Warshel, Inorg. Chem., 1999, 38, 3215–3222.
18 A. A. Neverov, Z. Lu, C. I. Maxwell, M. F. Mohamed, 43 B. DasGupta, R. Haidar, W. Hsieh and L. J. Zompa, Inorg.
C. J. White and J. S. W. Tsang, J. Am. Chem. Soc., 2006, 128,
16398–16405.
19 H. Linjalahti, G. Feng, J. C. Mareque-Rivas, S. Mikkola and
Chim. Acta, 2000, 306, 78–86.
44 M. F. Mohamed, A. A. Neverov and R. S. Brown, Inorg.
Chem., 2009, 48, 11425–11433.
N. H. Williams, J. Am. Chem. Soc., 2008, 130, 4232– 45 S. Kuusela, A. Azhayev, A. Guzaev and H. Lönnberg,
4233. J. Chem. Soc., Perkin Trans. 2, 1995, 1197–1202.
20 H. Korhonen, S. Mikkola and N. H. Williams, Eur. J. Chem., 46 V. M. Shelton and J. R. Morrow, Inorg. Chem., 1991, 30,
2011, 18, 659–670. 4295–4299.
21 A. O’Donoghue, S. Y. Pyun, M. Yang, J. R. Morrow and 47 M. K. Stern, J. K. Bashkin and E. D. Sall, J. Am. Chem. Soc.,
J. P. Richard, J. Am. Chem. Soc., 2006, 128, 1615–1621. 1990, 112, 5357–5359.
22 M. Yang, O. Iranzo, J. P. Richard and J. R. Morrow, J. Am. 48 L. A. Jenkins Autry and J. K. Bashkin, Inorg. Chim. Acta,
Chem. Soc., 2005, 127, 1064–1065. 1997, 263, 49–52.
23 K. Selmeczi, C. Michel, A. Milet, I. Gautier-Luneau, 49 M. Yashiro, H. Kaneiwa, K. Onaka and M. Komiyama,
C. Philouze, J. Pierre, D. Schieders, A. Rompel and C. Belle,
Chem.–Eur. J., 2007, 13, 9093–9106.
24 H. Gao, Z. Ke, N. J. DeYonker, J. Wang, H. Xu, Z. Mao,
Dalton Trans., 2004, 605–610.
50 M. Yashiro and R. Kawahara, J. Biol. Inorg. Chem., 2004, 9,
914–921.
D. L. Phillips and C. Zhao, J. Am. Chem. Soc., 2011, 133, 51 S. Kuusela and H. Lönnberg, J. Chem. Soc., Perkin Trans. 2,
2904–2915. 1994, 2301–2306.
25 S. Mikkola, E. Stenman, K. Nurmi, E. Yousefi-Salakdeh, 52 M. Yang, J. R. Morrow and J. P. Richard, Bioorg. Chem.,
R. Strömberg and H. Lönnberg, J. Chem. Soc., Perkin Trans.,
1999, 1619–1625.
26 T. Gajda, R. Krämer and A. Jancso, Eur. J. Inorg. Chem.,
2000, 1635–1644.
27 X. Wu, H. Lin, J. Shao and H. Lin, J. Mol. Catal. A: Chem.,
2008, 293, 79–85.
28 E. Kimura, T. Shiota, T. Koike, M. Shiro and M. Kodama,
J. Am. Chem. Soc., 1990, 112, 5805–5811.
2003, 35, 366–374.
53 R. A. Matthews, C. S. Rossiter, J. R. Morrow and
J. P. Richard, Dalton Trans., 2007, 3804–3811.
54 M. Yang, J. P. Richard and J. R. Morrow, Bioorg. Chem.,
2007, 35, 366–374.
55 T. Humphry, S. Iyer, O. Iranzo, J. R. Morrow, J. P. Richard,
P. Paneth and A. C. Hengge, J. Am. Chem. Soc., 2008, 130,
17858–17866.
8338 | Org. Biomol. Chem., 2013, 11, 8324–8339
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
56 C. Vichard and T. A. Kaden, Inorg. Chim. Acta, 2002, 337, 59 S. Valakoski, S. Heiskanen, S. Andersson, M. Lähde and
173–180. S. Mikkola, J. Chem. Soc., Perkin Trans. 2, 2002, 604–610.
57 M. Subat, K. Woinaroschy, C. Gerstl, B. Sarkar, 60 Y. Fan and Y. Q. Gao, J. Am. Chem. Soc., 2007, 129, 905–913.
W. Kaim and B. König, Inorg. Chem., 2008, 47, 4661– 61 J. C. Mareque Rivas, R. Torres Martin de Rosales and
4668.
S. Parsons, Dalton Trans., 2003, 4385–4386.
58 J. R. Morrow and W. C. Trogler, Inorg. Chem., 1988, 27, 62 Y. C. Wu, P. A. Berezensky, D. Feng and W. F. Koch, Anal.
3387–3394.
Chem., 1993, 65, 1084–1087.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 8324–8339 | 8339