Copper(II) and Zinc(II) Complexes of Conformationally Constrained Polyazamacrocycles as Efficient Catalysts for RNA Model Substrate Cleavage in Aqueous Solution at Physiological pH

Article abstract of DOI:10.1002/chem.201601175

As part of a quest for efficient artificial catalysts of RNA phosphodiester bond cleavage, conformationally constrained mono- and bis-polyazamacrocycles in which tri- or tetraazaalkane chains link the ortho positions of a benzene ring were synthesized. The catalytic activities of mono- and dinuclear copper(II) and zinc(II) complexes of these polyazamacrocycles towards cleavage of the P?O bond in 2-hydroxypropyl-4-nitrophenylphosphate (HPNP) in aqueous solution at pH 7 have been determined. Only the complexes of the ligands incorporating three nitrogen atoms in a macrocycle proved to be capable of efficiently catalyzing HPNP transesterification. The dinuclear complexes were found to be approximately twice as efficient as their mononuclear counterparts, and exhibited Michaelis–Menten saturation kinetics with calculated rate constants of k_cat≈10^?4s^?1. By means of quantum chemical calculations (DFT/COSMO-RS), several plausible reaction coordinates were described. By correlating the calculated barriers with the experimental kinetic data, two possible reaction scenarios were revealed, with activation free energies of 20–25 kcal mol^?1.

Full text of DOI:10.1002/chem.201601175

DOI: 10.1002/chem.201601175
Full Paper
&
Cleavage Reactions
Copper(II) and Zinc(II) Complexes of Conformationally
Constrained Polyazamacrocycles as Efficient Catalysts for RNA
Model Substrate Cleavage in Aqueous Solution at Physiological
pH
Daniel Bꢀm,[a, b] Eva Svobodovꢁ, Vꢁclav Eigner, Lubomꢀr Rulꢀsek,* and Jana Hodacovꢁ*
[a]
[c]
[b]
[a]
ˇ
ˇ
Chem. Eur. J. 2016, 22, 1 – 13
1
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Abstract: As part of a quest for efficient artificial catalysts of
RNA phosphodiester bond cleavage, conformationally con-
strained mono- and bis-polyazamacrocycles in which tri- or
tetraazaalkane chains link the ortho positions of a benzene
ring were synthesized. The catalytic activities of mono- and
dinuclear copper(II) and zinc(II) complexes of these polyaza-
macrocycles towards cleavage of the PꢀO bond in 2-hydrox-
ypropyl-4-nitrophenylphosphate (HPNP) in aqueous solution
at pH 7 have been determined. Only the complexes of the li-
gands incorporating three nitrogen atoms in a macrocycle
proved to be capable of efficiently catalyzing HPNP transes-
terification. The dinuclear complexes were found to be ap-
proximately twice as efficient as their mononuclear counter-
parts, and exhibited Michaelis–Menten saturation kinetics
with calculated rate constants of kcat ꢁ10ꢀ4 sꢀ1. By means of
quantum chemical calculations (DFT/COSMO-RS), several
plausible reaction coordinates were described. By correlating
the calculated barriers with the experimental kinetic data,
two possible reaction scenarios were revealed, with activa-
tion free energies of 20–25 kcalmolꢀ1
.
Introduction
convenient experimental setup, model substrates are often
used in studies on the catalytic properties of artificial catalysts.
Among them, 2-hydroxypropyl-4-nitrophenylphosphate (HPNP)
is the most frequently used RNA model. HPNP undergoes intra-
molecular transesterification (Scheme 1) with the release of 4-
nitrophenol, the concentration of which can be conveniently
determined by UV/Vis spectrophotometry.
A phosphate diester linkage is present in many biological com-
pounds, such as RNA, DNA, and cGMP. Phosphate diesters are
extremely stable towards hydrolytic cleavage under physiologi-
cal conditions, with half-lives of approximately 100 years for
RNA[1] and up to millions of years for DNA.[2] Specific nucleases
are able to accelerate the rate of hydrolysis of the PꢀO bond
by a factor of up to 1017, which makes them some of the most
efficient enzymes encountered in nature.[3] The development
of artificial catalysts mimicking the catalytic activities of natural
enzymes has attracted considerable attention over the past
few decades.[4] From a fundamental (academic) perspective,
studies on simpler synthetic catalysts can provide us with
better insight into the mechanisms of more complex enzyme
functions. Furthermore, artificial ribonucleases can be conju-
gated with structural units capable of recognizing specific oli-
gonucleotide sequences, and thus result in conjugates with
potential applications as new biotechnological tools or thera-
peutics targeting nucleic acids.
Scheme 1. Cleavage of HPNP.
To date, many artificial catalytic systems capable of acceler-
ating cleavage of the PꢀO bond in RNA model substrates have
been reported,[6] mostly inspired by phosphodiesterases. Since
the active sites of phosphodiesterases usually contain two or
more metal ions, the catalytic abilities of numerous mono-, di-,
and trinuclear metal complexes of various synthetic ligands
have been investigated, and many of them have been shown
to efficiently catalyze the cleavage of RNA model substrates.[6]
Among them, kinetically labile complexes of zinc(II) and cop-
per(II) ions have become increasingly popular. Typically, multi-
nuclear complexes are more efficient catalysts than their mon-
onuclear counterparts. Despite enormous progress in this field,
only a few synthetic catalysts have been shown to be capable
of accelerating the cleavage of RNA models in water at physio-
logical pH. In 1993, Chin et al.[7] described the dinuclear cop-
per(II) complex of 2,6-bis[bis(2-benzimidazolylmethyl)amino-
methyl]-4-methylphenol as an efficient catalyst of HPNP cleav-
age in aqueous solution at pH 7 and 258C. The observed
pseudo first-order rate constant (k’obs =2.1ꢃ10ꢀ3 sꢀ1) was
about 52 times higher than that obtained in the presence of
the analogous mononuclear complex. Later, Hamilton et al.[8]
showed that the dinuclear copper(II) complex of a bis-terpyri-
dine ligand accelerated HPNP cleavage in water at pH 7 and
258C by a factor of about 65 (k’obs =3.86ꢃ10ꢀ4 sꢀ1) with re-
In comparison with hydrolytic DNA cleavage, RNA cleavage
is significantly faster owing to the presence of the ribose 2’-hy-
droxy group, which acts as an internal nucleophile. The cleav-
age of RNA is known to proceed in two steps: transesterifica-
tion of the phosphodiester by the 2’-hydroxy group to form
a 2’,3’-cyclic phosphodiester, followed by hydrolytic opening of
the 2’,3’-cyclic phosphodiester.[5] Because of the practical and
ˇ
[a] D. Bꢀm, Dr. E. Svobodovꢁ, Prof. J. Hodacovꢁ
Department of Organic Chemistry, Faculty of Chemical Technology
University of Chemistry and Technology
Technickꢁ 5, 166 28 Prague 6 (Czech Republic)
Fax: (+420)220-444-288
E-mail: Jana.Hodacova@vscht.cz
ˇ
[b] D. Bꢀm, Dr. L. Rulꢀsek
Institute of Organic Chemistry and Biochemistry
v.v.i. and Gilead Sciences Research Center
Academy of Sciences of the Czech Republic
ˇ
Flemingovo nꢁmestꢀ 2, 166 10 Prague 6 (Czech Republic)
E-mail: lubos@uochb.cas.cz
[c] V. Eigner
Department of Solid State Chemistry, Faculty of Chemical Technology
University of Chemistry and Technology
Technickꢁ 5, 166 28 Prague 6 (Czech Republic)
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/chem.201601175.
&
&
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
2
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
spect to its mononuclear counterpart. Similar results were ob-
tained for the cleavage of ribonucleoside 2’,3’-cyclic mono-
phosphates in water at pH 7.5.[9] Another example was pub-
lished by Morrow et al.,[10] who described a dinuclear zinc(II)
complex of a ligand consisting of two 1,4,7-triazacyclononane
rings connected by a 2-hydroxypropane-1,3-diyl spacer. This
that there is no general mechanism[17] and that most likely
there is subtle interplay between contributions from various
activation modes. This delicate balance depends on both the
substrate and the catalyst structure. Last but not least, the
nature of the solvent also plays an important role in the reac-
tion. Thus, greater catalytic turnover was observed when meth-
anol or ethanol was used as reaction medium instead of
water,[18] but it was shown[19] that other solvents also influence
the rate of RNA model cleavage. To date, most reaction mecha-
nisms for the cleavage of RNA model systems have been infer-
red from available experimental data. Only recently have quan-
tum chemical methods been exploited to provide a solid theo-
retical basis for the mechanistic descriptions of this reaction.[20]
A recent study by Brown and co-workers is particularly nota-
ble,[20c] as it provided an overview of plausible PꢀO bond-cleav-
age pathways in the decomposition of HPNP or bis(4-nitrophe-
nyl)phosphate promoted by the dinuclear zinc(II) complex of
1,3-bis(1,5,9-triazacyclododec-1-yl)propane. Nevertheless, sever-
al important questions still need to be addressed. For example,
Richard and Morrow[21] assumed that the decomposition of
HPNP promoted by the dinuclear zinc(II) complex of 1,3-
bis(1,5,9-triazacyclododec-1-yl)propan-2-ol in water was subject
to specific base catalysis, whereas Brown and co-workers[20c] as-
sumed general base catalysis in light alcohols for their above-
mentioned system. This discrepancy has a significant implica-
tion for the rate-determining step (RDS) of the reaction. Fur-
thermore, quantum chemical calculations performed on cata-
lytic systems with metal ions other than zinc(II) or in different
solvents (for example, in water) may greatly extend our under-
standing of the key factors underpinning the mechanism of
the reaction in general.
complex gave
a
second-order rate constant
kZn =25ꢃ
10ꢀ2 Lmolꢀ1 sꢀ1 for HPNP cleavage in water at pH 7.6 and
258C, corresponding to a 120-fold acceleration with respect to
the reaction catalyzed by the mononuclear zinc(II) complex of
1,4,7-triazacyclononane. A calix[4]arene-based trinuclear cop-
per(II) complex in which the copper ions were bound within
1,5,9-triazacyclododecane macrocycles was found to catalyze
HPNP cleavage in water at pH 7.0 and 258C with a rate con-
stant of k’obs =1.3ꢃ10ꢀ4 sꢀ1, corresponding to a 590-fold cata-
lytic efficiency with respect to the uncatalyzed reaction and
about a 39-fold rate increase relative to the process catalyzed
by the mononuclear copper(II) complex of 1,5,9-triazacyclodo-
decane.[11] The same catalyst was found to accelerate the cleav-
age of diribonucleoside 3’,5’-monophosphates with selectivity
for UpU and UpG. A 105-fold rate acceleration compared with
the uncatalyzed reaction was reported in water at pH 7.0 and
508C.[11] In 2005, Williams et al.[12] demonstrated that the at-
tachment of amino groups to pyridine rings of tris(2-pyridine-
methyl)amine significantly improved the catalytic activity of
the corresponding zinc(II) complex towards HPNP cleavage.
The zinc(II) complex of a ligand bearing amino groups was 20
times more effective than its non-aminated analogue. Combin-
ing 2-aminopyridin-6-yl structural motifs and a hydroxy group
in a single ligand structure resulted in a highly efficient binu-
clear zinc(II) complex that catalyzed HPNP cleavage in water
with a rate constant of kcat =1.7ꢃ10ꢀ2 sꢀ1 (pH 7.4, 258C).[13]
This complex is the best small-molecule catalyst hitherto re-
ported. Faster HPNP cleavage has been achieved with poly-
mers or supramolecular assemblies bearing multiple zinc(II)
complexes.[14] Notably, gold nanoparticles coated with a mono-
layer of zinc(II) complexes of 1,4,7-triazacyclononane-terminat-
ed long-chain thiols (so-called nanozymes) successfully cleaved
the HPNP substrate in water at pH 7.5 and 258C with rate con-
Herein, we report the synthesis of conformationally con-
strained polyazamacrocycles and the catalytic properties of
their mono- and dinuclear copper(II) and zinc(II) complexes to-
wards hydrolytic cleavage of HPNP. The experimental results
are complemented by quantum chemical calculations with sev-
eral DFT functionals in conjunction with what is presumably
the most accurate implicit solvation model (conductor-like
screening model for realistic solvation, COSMO-RS), in the
hope of obtaining quantitative agreement with the reported
kinetic data. This may give us a certain confidence in formulat-
ing an unambiguous mechanistic picture of the title reaction
in aqueous solution and in identifying key factors underpin-
ning the catalytic action of these complexes as simple models
of phosphodiesterases.
stants of up to kcat =3.7ꢃ10ꢀ2 sꢀ1 [15]
.
It has been postulated that the cleavage of RNA and its sim-
pler models proceeds through transesterification of the phos-
phodiester by the 2’-hydroxy group and subsequent opening
of the five-membered 2’,3’-cyclic phosphodiester. Several
mechanisms have been proposed[5,6,16] to explain the activating
effect of metal-ion catalysts in the cleavage of RNA models:
1) the metal ion acts as a Lewis acid interacting with the phos-
phoryl oxygen atom of the substrate and activates the bound Results and Discussion
substrate towards nucleophilic attack by the 2’-hydroxy group;
2) coordination of the metal ion to the 2’-hydroxy group can
Catalyst design
activate a nucleophile; 3) the alkoxide leaving group can be ac-
tivated by interaction of its oxygen atom with the metal ion;
4) a metal-coordinated water molecule or hydroxy group can
act as an intramolecular acid or base, respectively, in general
acid/base catalysis; or 5) electrostatic interactions between the
metal ion and the noncoordinated phosphate ester may also
provide some rate enhancement. It has been demonstrated
Many of the reported artificial polynuclear catalysts for hydro-
lytic cleavage of the phosphodiester bond are copper(II) or
zinc(II) complexes of ligands containing two or three triaza- or
tetrazamacrocycles as metal-binding sites.[4] Typically, each aza-
crown is connected to a rigid central unit at a single point via
a simple alkyl linker. Unfortunately, some of these ligands are
rather flexible, which may not be favorable for catalysis. To im-
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
3
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
prove the ligand preorganization, we proposed appending the
ends of the triaza- or tetraazaalkanediyl chain to the ortho po-
sitions of a benzene ring. The benzylic CH2 groups enforce an
out-of-plane orientation of the azacrown moiety. In the case of
two azacrown units, this may lead to two isomers in which the
macrocycles are oriented either to the same side (syn orienta-
tion) or to opposite sides (anti) of the benzene ring. The syn
orientation is represented in Figure 1.
complexes were then used in kinetic studies of the hydrolytic
cleavage of HPNP.
Synthesis of the ligands
Ligands 1 and 2 having three nitrogen atoms in the macrocy-
clic ring were prepared according to Scheme 3. This synthetic
pathway is a modification of synthesis of 1,5,9-triazacyclodode-
cane by Alder et al.[24] The synthesis started from 1,5,7-triazabi-
cyclo[4.4.0]dec-5-ene, which was doubly alkylated with bromo-
Figure 1. Representation of the syn orientation of two azacrown units at-
tached to a central benzene ring.
In comparison with structures in which the azacrown unit is
connected to the rigid benzene platform by a single bond,
these ortho-bridged ligands are conformationally less flexible.
It is expected that the syn orientation of the azacrown moieties
may provide a conformationally constrained binding site capa-
ble of accommodating two metal ions at a close-to-optimal
mutual separation (ca. 5 ꢄ, as can be inferred a priori from
simple models). Chin et al.[1] postulated that twofold Lewis acid
activation through the coordination of both phosphoryl
oxygen atoms of a phosphate diester to metal ions is possible,
with metal–metal distances in the range of 2.9–7.0 ꢄ. Accord-
ingly, Scrimin et al.[22] concluded, on the basis of results ob-
tained for oligopeptide-based ligands, that a distance of 6.3 ꢄ
between the two zinc(II) ions might be optimal for HPNP cleav-
age, whereas Meyer et al.[23] showed that the distance between
two zinc(II) ions in the catalyst should be at least 4 ꢄ. Shorter
Zn···Zn distances result in inactivation of the catalyst due to
the presence of a bridging hydroxide ion (that is, formation of
the stable but inactive m-1,2-OH complex).
Scheme 3. Synthesis of ligands 1 and 2.
methyl derivatives of benzene to yield the mono- (7) or bis-
guanidinium salt (8). Since the distance between the nitrogen
atoms in the starting bicyclic guanidine fits well with the dis-
tance between the ortho benzylic carbon atoms, polycyclic
guanidinium salt 8 was formed regioselectively. Its structure
was confirmed by X-ray crystallographic analysis (see Figure 2).
The above considerations led us to design and synthesize
bis-macrocyclic ligands 2, 4, and 6 and their mono-macrocyclic
analogues 1, 3, and 5 (Scheme 2). Their copper(II) and zinc(II)
Scheme 2. Bis-macrocyclic ligands 2, 4, and 6 and their mono-macrocyclic
analogues 1, 3, and 5.
Figure 2. Perspective ORTEP drawing of bis-guanidinium salt 8. Ellipsoids are
set at 50% probability.[50]
&
&
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
4
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
The NMR spectra of guanidinium salts 7 and 8 suggested
that these were formed almost quantitatively, although the
presence of some impurities was detected. Therefore, both
salts were purified by crystallization before the next reaction
step. Later, it was found that this purification was not necessa-
ry, and that both the alkylation and reduction steps could be
performed in one pot. Next, guanidinium salts 7 and 8 were
reduced to orthoamides 9 and 10 with NaBH4 in methanol (or
in acetonitrile/methanol in the one-pot procedure). The result-
ing orthoamides 9 and 10 were not entirely pure, but several
attempts to purify them proved unsuccessful (the fact that
they are oily liquids precluded crystallization, and attempts to
purify them by column chromatography led only to their par-
tial decomposition). Therefore, the crude orthoamides were
used in the final hydrolytic step and the impurities were suc-
cessfully removed during purification of the final products
1 and 2 by crystallization of their hydrochlorides from an aque-
ous HCl/MeOH and aqueous HCl, respectively. The structure of
ligand 2 was unambiguously confirmed by X-ray analysis of its
hexahydrobromide (Figure 3). (For experimental details of the
preparation of ligand 2 without isolation of the intermediates,
see the Experimental Section; the elementary synthetic steps
and characterization of the compounds shown in Scheme 3
can be found in the Supporting Information.)
kis(bromomethyl)benzene and subsequent hydrolytic cleavage
of the aminal core.[28]
All ligands were obtained in the form of solvates of their hy-
drochloride salts. As the number of bound solvent molecules
in these solvates differed from batch to batch, the exact con-
tent of the ligand in each sample was determined by quantita-
tive NMR measurement.
Kinetic studies
The catalysts were formed in situ by mixing stoichiometric
amounts of Cu(ClO4)2 or Zn(ClO4)2 and a ligand in aqueous
HEPES buffer. The catalytic activities of the resulting copper(II)
and zinc(II) complexes of ligands 1–6 in the cleavage of HPNP
(Scheme 1) were determined from the rates of formation of 4-
nitrophenol, monitored spectrophotometrically at 400 nm in
a thermostated UV cell at 208C.
Catalytic activity was only observed for complexes of ligands
1 and 2 in aqueous HEPES buffer, whereas initial screening re-
vealed that copper(II) and zinc(II) complexes of ligands 3–6 ex-
hibit little or no catalytic activity at pH 7 in the two media
tested: 10 mm aqueous HEPES and acetonitrile/10 mm aque-
ous HEPES (1:1), both at pH 7.0. Moreover, no catalytic activity
was observed for complexes of ligands 1 and 2 in acetonitrile/
water (1:1). This observation indicated that acetonitrile acts as
an inhibitor in the studied catalytic systems. From the prelimi-
nary studies, it could be further deduced that an efficient cata-
lyst must have no more than three strongly coordinated li-
gands in the coordination sphere of the metal ion (such as the
macrocycle nitrogen atoms or the CꢂN group of acetonitrile)
so that the metal center is accessible to a substrate and/or nu-
cleophile. Furthermore, blank experiments carried out in the
absence of ligands resulted in only negligible changes in ab-
sorption at 400 nm, which indicate that the catalytic contribu-
tion of free copper(II) or zinc(II) ions to HPNP cleavage was in-
significant (to within an error bar of the experimental data).
On the basis of the above preliminary results, detailed stud-
ies on the catalytic activities of the copper(II) and zinc(II) com-
plexes of ligands 1 and 2 in aqueous HEPES buffer were carried
out. The total concentration of metal ions was kept constant
(0.2 mm) in all experiments to obtain directly comparable rate
constants. The obtained kinetic data conformed to pseudo
first-order kinetics, and the corresponding pseudo first-order
rate constants k’obs were determined at various pH values
(6.25–7.5) by a nonlinear least-squares fitting method employ-
ing Equation (1):
Figure 3. Perspective ORTEP drawing of the cationic part of the hexahydro-
bromide of 2. Ellipsoids are set at 50% probability; anions and solvent mole-
cules are omitted for clarity.[50]
Ligand 1 has been previously synthesized by Sherry and
Chavez[25] using the classical Richman–Atkins method.[26] Tosy-
lated 1,5,9-triazanonane was alkylated with 1,2-bis(bromome-
thyl)benzene in the presence of K2CO3 in DMF yielding the to-
sylated macrocycle, from which the tosyl groups were reduc-
tively removed to afford ligand 1 in 57% overall yield. Howev-
er, this synthetic approach is not suitable for the preparation
of ligand 2, since on alkylation of tosylated 1,5,9-triazanonane
with 1,2,4,5-tetrakis(bromomethyl)benzene, the tosylated aza-
crown chains preferentially link the meta benzylic positions.[27]
Therefore, we used the modified Alder method for the prepa-
ration of ligand 2. Synthesis of the simpler ligand 1 was exam-
ined as a model preparation to test this synthetic route and to
optimize the reaction conditions.
0
obs t
½NPꢃ ¼ ½HPNPꢃ0ð1ꢀeꢀk
Þ
ð1Þ
where k’obs is the rate constant, [NP] the concentration of 4-ni-
trophenol, [HPNP]0 the initial concentration of HPNP, and t the
time. The calculated rate constants k’obs obtained at pH 7.0 are
summarized in Table 1, and the pH dependences of k’obs are
depicted in Figure 4.
Ligands 3–6 were synthesized according to previously pub-
lished procedures involving alkylation of tricyclic aminals (the
products of condensation of linear tetraamines with butane-
2,3-dione) with 1,2-bis(bromomethyl)benzene or 1,2,4,5-tetra-
For comparison, the rate constant for uncatalyzed HPNP
cleavage was estimated from literature data obtained under
similar experimental conditions. A pseudo first-order rate con-
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
5
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
7.00–7.25. The acceleration of the reaction with increasing pH
may be attributable to the higher concentration of hydroxide
ions, which can serve as a general base. Hydroxide ions can
also deprotonate the hydroxy group of the substrate, increas-
ing its nucleophilicity. On the other hand, the decrease in reac-
tivity at even higher pH is likely caused by coordination of hy-
droxide ion to the metal ion in M-1 and M2-2, which effectively
blocks the free coordination site for the substrate. In the case
of the dinuclear catalysts M2-2, hydroxide ion may also serve
as a bridge between the metal ions. Furthermore, this process
is in accordance with the observed retardation of the reaction
on addition of Clꢀ ions, which may also coordinate to the
metal ions in the catalysts or serve as bridging ligands. The
largest difference in k’obs between Cu-1 and Cu2-2 was ob-
served at pH 6.50, which indicates a synergistic effect between
the copper(II) ions in Cu2-2 at this pH. With increasing pH, dif-
ferences in k’obs for Cu-1 and Cu2-2 decreased, such that nearly
equal values were reached at pH 7.50.
Table 1. Pseudo first-order rate constants for HPNP cleavage catalyzed by
copper(II) and zinc(II) complexes of ligands 1 and 2, obtained in 10 mm
aqueous HEPES buffer at pH 7.0 and 208C.
Catalyst
k’obs [sꢀ1
]
Cu-1
Zn-1
Cu2-2
Zn2-2
(1.98ꢄ0.18)ꢃ10ꢀ5
(7.05ꢄ2.90)ꢃ10ꢀ6
(2.80ꢄ0.21)ꢃ10ꢀ5
(1.43ꢄ0.04)ꢃ10ꢀ5
The dependence of the initial rate of reaction on the con-
centration of the substrate (HPNP) was studied at pH 7.0 in
10 mm aqueous HEPES buffer with a fixed catalyst concentra-
tion (0.2 mm) at 208C. Michaelis–Menten saturation kinetic be-
havior was observed, and the Michaelis–Menten parameters
were calculated by fitting the obtained data with Equation (2):
Figure 4. The observed pH dependences of the rate constants k’obs of HPNP
cleavage catalyzed by copper(II) and zinc(II) complexes of ligands 1 and 2.
dc
dt
nmax½HPNPꢃ0
ð2Þ
n0 ¼
¼
stant of k’obs =3.33ꢃ10ꢀ8 sꢀ1 has been reported for HPNP
cleavage (substrate concentration 0.18 mm) in 10 mm aqueous
HEPES buffer at pH 7.0 and 378C,[29] and a value of k’obs =1.9ꢃ
10ꢀ8 sꢀ1 has been obtained in 20 mm aqueous HEPES buffer/
acetonitrile (1:1) with a substrate concentration of 0.15 mm at
pH 7.0 and 258C.[30] On this basis, it can be safely postulated
that the rate constant of noncatalyzed HPNP cleavage under
KM þ ½HPNPꢃ0
where v0 is the initial reaction rate, vmax the maximum reaction
rate, [HPNP]0 the initial concentration of HPNP, and KM the Mi-
chaelis constant (substrate concentration at half of the maxi-
mum rate). The results are summarized in Table 2.
our experimental conditions would be on the order of 10ꢀ8 sꢀ1
.
Table 2. Michaelis–Menten parameters for HPNP cleavage in 10 mm
aqueous HEPES buffer at pH 7.0 catalyzed by the copper(II) and zinc(II)
complexes of ligands 1 and 2 at 208C.
Thus, the k’obs values reported in Table 1 indicate that the reac-
tion was accelerated by approximately three orders of magni-
tude.
[a]
[a]
[b]
[c]
Catalyst vmax [molLꢀ1 sꢀ1
]
Km [molLꢀ1
]
Kass [Lmolꢀ1
]
kcat [sꢀ1
]
The copper(II) complexes of both ligands 1 and 2 exhibited
higher catalytic activities than the corresponding zinc(II) com-
plexes, whereas the mononuclear M-1 complexes were approx-
imately one- to two-thirds less efficient than their M2-2 coun-
terparts. Furthermore, the k’obs value of 1.98ꢃ10ꢀ5 sꢀ1 for the
Cu-1 catalyst at pH 7.0 and 208C is nearly an order of magni-
tude greater than that of the copper(II) complex of 1,5,9-triaza-
cyclododecane obtained under similar conditions (k’obs =3.3ꢃ
10ꢀ6 sꢀ1; 20 mm HEPES in water; pH 7.0; 258C).[11] The higher
catalytic activity could be attributed to the presence of the
benzene ring in close proximity to the catalytic site. The aro-
matic ring may create a more hydrophobic microenvironment,
which might be helpful in the catalytic process, or it may facili-
tate release of the leaving group (4-nitrophenolate) through
p–p stacking interactions.
Cu-1
Zn-1
Cu2-2
Zn2-2
(1.14ꢄ0.50)ꢃ10ꢀ7 0.0103ꢄ0.0058
(5.39ꢄ0.31)ꢃ10ꢀ8 0.0184ꢄ0.0054
(1.75ꢄ0.60)ꢃ10ꢀ7 0.0078ꢄ0.0040
(3.24ꢄ0.88)ꢃ10ꢀ8 0.0070ꢄ0.0023
97
54
129
143
5.68ꢃ10ꢀ4
2.69ꢃ10ꢀ4
8.73ꢃ10ꢀ4
1.62ꢃ10ꢀ4
[a] Determined by curve fitting. [b] Calculated by Km =1/Kass. [c] Calculated
by kcat =vmax/[catalyst].
Quantum chemical calculations
Distinct features of reaction mechanism(s) for HPNP cleavage
revealed by quantum chemical calculations
To obtain a full mechanistic picture of the studied reactions,
the experimental data were complemented by a comprehen-
sive computational study on the reaction mechanism of HPNP
decomposition catalyzed by the M2-2 catalysts. To this end,
a set of DFT functionals was used in combination with what is
presumably one of the most reliable and robust solvation
models, namely, the COSMO-RS method.[31] We expected to
The pH dependence of the rate constants exhibited bell-
shaped behavior with maxima at pH 7.00–7.15 and 7.25 for di-
nuclear and mononuclear complexes, respectively (Figure 4). A
rapid decrease in activity was observed when the pH was low-
ered to 6.25, and a slight decrease was also observed at pH>
&
&
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
6
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
obtain near-quantitative results that could be correlated with
the experimental data to elucidate the catalytic efficiency of
the Cu2-2 and Zn2-2 species in aqueous solution. Since the re-
action rate constants k’obs indicated faster reaction in the pres-
ence of the dinuclear complexes (catalysts), the presented the-
oretical study was carried out for these more complex (and
computationally more challenging) catalytic systems. It can be
assumed that both metal ions participate in substrate binding
and reaction. Elaborating on the work of Brown and co-work-
ers,[20c] we considered the binding mode of HPNP depicted in
Figure 5. The substrate is bound to the metal ions through the
relaxing the rest of the molecule (relaxed potential-energy
scan). The nature of the stationary points, that is, minima and
transition states (TSs), was confirmed by vibrational analyses,
as well as by minuscule displacements of the TS geometries,
which always led to the corresponding local minima, that is, re-
actants, products, or intermediates, along the reaction coordi-
nate (similar to the intrinsic reaction coordinate procedure,
IRC). Finally, the reaction pathways were also studied by two-
dimensional potential-energy surface (PES) scans with the Pꢀ
Onuc and PꢀOleaving distances as variables. As a representative
example, the PES for the studied reaction is depicted in
Figure 6 (for the HPNP·Zn2-2 complex in 5C,4C coordination
mode).
Figure 6. Two-dimensional PES scan of the studied reaction with the PꢀOnuc
and PꢀOleaving distances used as the variables. The Zn2-2 catalyst was consid-
ered and the energies were obtained by the RI-DFT(PBE)/def2-SVP method.
Although the overall features of the theoretical investigation
discussed above were very similar for both of the studied ions
(Zn, Cu), there were certain noticeable differences. Therefore,
the computational results are discussed separately for each
metal ion.
Zn2-2 complex
Careful inspection of the reaction coordinate revealed that the
two-step reaction mechanism was the likely alternative in the
gas phase (for both the ZnRC and ZnRC’ pathways). Both reac-
tion pathways involve the reactant structure ZnRC (ZnRC’), two
transition states ZnTS1 (ZnTS1’) and ZnTS2 (ZnTS2’), one inter-
mediate ZnIM (ZnIM’) between TS1 and TS2, and the product of
the reaction ZnPC (ZnPC’). The situation changed slightly when
the effect of solvation was taken into account through the
COSMO-RS model. The Gibbs free energy of ZnTS2’ decreased
and became lower than that of ZnIM’. We interpret this in
terms of the reaction starting with ZnRC’ (5C,5C coordination
mode) being a one-step process in the solvent (mechanism A;
Figure 7). On the contrary, the reaction proceeding from the
ZnRC structure (5C,4C coordination mode) remains a two-step
process in the solvent (mechanism B; Figure 8). Our observa-
tions are in accordance with the work of Brown and co-work-
ers[20c] on a similar system, for which a two-step mechanism
was proposed with the leaving group (4-nitrophenol) bound
to the ZnB ion, whereas a one-step mechanism was postulated
for the (5C,5C) pathway (corresponding to ZnRC’ in our case).
We observed that the dissociation of the leaving group was fa-
Figure 5. Binding modes of the HPNP substrate to the Zn2-2 complex. Two
systems differing in the coordination sphere of the ZnB ion were considered.
The binding modes of the substrate in the HPNP·Cu2-2 system are similar.
Zn yellow, P green, O red, N blue, C gray, H white.
phosphoryl oxygen atoms, whereas the deprotonated 2-hy-
droxy group completes the coordination sphere of the ZnA
(CuA) ion, giving rise to an overall (5C,4C) coordination mode
of M2-2 (Figure 5, MRC). Similarly, the ZnB (CuB) coordination
sphere can be considered as pentacoordinate, with a water
molecule occupying the fifth coordination site, that is, the
(5C,5C) mode depicted in Figure 5 (MRC’).
It seems reasonable to assume that the reaction starts with
intramolecular attack of the coordinated and deprotonated 2-
hydroxy group on the central phosphorus atom, followed by
cleavage of the bond between the phosphorus atom and the
oxygen atom of the leaving 4-nitrophenol moiety (SN2 reaction
mechanism). Thus, we searched for the respective transition
state (TS) by shortening the PꢀO distance (that is, the distance
between the nucleophilic ROꢀ group and phosphorus), while
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
7
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 7. Proposed mechanism A.
Figure 8. Proposed mechanism B.
cilitated by the formation of a hydrogen bond between its
oxygen atom and the hydrogen atom of the bound solvent
molecule. As shown in Figure 9, the barriers for both processes
were quite similar and one can assume that both pathways
may be operational.
reaction, spin-state flipping may have occurred as a conse-
quence of the sizeable spin–orbit coupling in copper systems,
which allows formally spin-forbidden reactions to proceed. At
the same time, the reaction profiles would be virtually un-
changed if only one spin multiplicity were to be considered
along the reaction pathway (spin conservation).
Reaction coordinate
The computed activation energies (TS barriers) and free ener-
gies are summarized in Table S4 of the Supporting Information
and graphically depicted in Figures 9 and 10. Contrary to the
situation with the zinc catalyst, the second TS for the copper(II)
catalyst was found to have lower energy than the intermediate
structure (IM), both in the gas phase and in the solvent and
for both tetra- and pentacoordinate CuB. Thus, we postulate
that the reaction proceeds by a one-step mechanism, the es-
sential features of which are identical to the mechanism dis-
cussed above for zinc(II).
Figure 9. Energy diagram for the proposed reaction mechanisms with the
Zn2-2 catalyst. All Gibbs free energies [kcalmolꢀ1] were calculated with inclu-
sion of solvation effects by using the COSMO-RS model. The presented
values were obtained by using the PBE0 functional; values for other func-
tionals are shown in Tables S4 and S5 (Supporting Information).
The experimental activation energy could be derived from
the Eyring equation [Eq. (3)]:
Cu2-2 complex
Copper(II) is a d9 ion with one unpaired electron in its valence
d shell. Therefore, two spin states must be considered for the
dinuclear HPNP·Cu2-2 system: triplet (S=1) and open-shell sin-
glet (S=0), which originate from the ferro- or antiferromagnet-
ic (super)exchange coupling of the two unpaired spins on the
respective copper(II) centers. The calculated spin-state split-
tings are listed in Table S3 of the Supporting Information. In all
cases (RC, TSs, IM, PC; with various functionals), the energy dif-
ferences between the two spin states were on the order of
200 cmꢀ1 (<0.6 kcalmolꢀ1) and thus well within the error bar
of the DFT methods used for computations of reaction ener-
getics. We decided to construct reaction coordinates from the
(electronic) ground-state energy, be it singlet or triplet. Saying
this, we implicitly assumed that within the time frame of the
Figure 10. Energy diagram for the proposed reaction mechanisms with the
Cu2-2 catalyst. All Gibbs free energies [kcalmolꢀ1] were calculated with inclu-
sion of solvation effects by using the COSMO-RS model. The presented
values were obtained by using the PBE0 functional; values for other func-
tionals are shown in Tables S4 and S5 (Supporting Information).
&
&
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
8
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
ꢀDG°
Deprotonation of the 2-hydroxy group is followed by nucle-
ophilic attack of the alcoholate on the phosphorus center, re-
sulting in a change from the original tetrahedral coordination
to trigonal-bipyramidal geometry. These two consecutive
steps, deprotonation and formation of the pentacoordinate
transition state, represent the RDS of the reaction. The depro-
tonation is presumably a fast process limited only by the rate
of diffusion and associated with a small reaction barrier (activa-
tion energy). Thus, the activation (free) energy for the RCPROT
(RC’PROT)!IM (IM’) reaction step is the sum of the energy of
deprotonation and the energy of TS1 (TS1’); see Table S4 of
the Supporting Information. These observations are in contrast
to those reported by Richard, Morrow et al. in a study[21] per-
formed on a similar system. We have found that the HPNP de-
composition process is subject to general base catalysis and
that 2-hydroxy deprotonation is part of the RDS. The same was
observed by Brown and co-workers for the process in light al-
cohols.[20c] From TS1, the product complex (PC) is formed,
either directly (one-step mechanism) or via IM and TS2 (Fig-
ures 7 and 8).
kBT
h
k0 ¼
e
RT
ð3Þ
The pseudo first-order rate constant k’ corresponds to the
whole process beginning with the formation of the HPNP·M2-2
complex from the free reactants (HPNP and M2-2), followed by
deprotonation of the HPNP hydroxy group and cleavage of the
PꢀO bond through TS1, which is presumably the RDS. There-
fore, to compare the experimental and calculated activation
(free) energies, it was necessary to calculate the Gibbs free en-
ergies of the free reactants. However, calculation of the ener-
getics of the first step, that is, formation of the RCPROT (RC’PROT
)
complex from the free reactants (catalyst and substrate), is
a computationally formidable task, in particular owing to the
nonzero molecular charges of the species involved. It is, in
principle, equivalent to computational prediction of stability
constants (association free energies of the charged species), as
[32]
ˇˇ
ˇ
recently discussed by Gutten, Besseovꢁ, and Rulꢀsek, since
the association of the RCPROT involves two huge and opposing
terms of hundreds of kilocalories per mole (gas-phase interac-
tion energies and (de)solvation energies). The computational
intricacies involved in this approach are reflected by the huge
differences in the computed Gibbs free energies for this asso-
ciative process (HPNP+Zn2-2!ZnRCPROT (ZnRC’PROT) yielded by
various respectable DFT functionals (from ꢀ21.2 kcalmolꢀ1 ob-
tained with PBE to 4.0 kcalmolꢀ1 with wB97xD). Therefore, in
agreement with the standard approach used in theoretical
modeling of enzymatic reaction mechanisms, we consider
RCPROT (RC’PROT) as the initial point on the reaction coordinate
and assign zero free energy to this reactant complex. This allows
us to compare the calculated activation (free) energy with the
experimental kcat (derived from Michaelis–Menten kinetics).
Starting from RCPROT (RC’PROT), the first step in the considered
mechanism is deprotonation of the 2-hydroxy group of the
substrate. This is facilitated by coordination of this group to
the metal ion, which presumably lowers the pKa of the secon-
dary alcohol by several units. The free energy difference be-
tween the starting RCPROT (RC’PROT) complex and RC (RC’) could
be directly determined using the pKa value of this hydroxy
group: DG=ꢀRTlnKa. An attempt was made to calculate this
pKa value by using the COSMO-RS solvation model (def2-
TZVPD-FINE parametrization). We computed pKa values of 9.2
and 13.4 for HPNP bound to the Zn2-2 and Cu2-2 complexes,
respectively. While the former value seems reasonable for Zn2-
2, we think that the latter is somewhat too high for Cu2-2. Al-
ternatively, one can exploit the fact that the known pKa values
of copper(II)- or zinc(II)-bound water or alcohol groups are usu-
ally comparable, typically in the range of 8–10.[33] This corre-
sponds to Gibbs free energies of 10.7–13.4 kcalmolꢀ1 at
293.15 K. To allow comparison of the experimental data with
the values gained by quantum chemical calculations, we opted
to use a pKa value of 9 for HPNP bound to both the Zn2-2 and
Cu2-2 complexes (12.1 kcalmolꢀ1 at 293.15 K). We consider this
to be as accurate as those obtained from the quantum chemi-
cal/COSMO-RS calculations discussed above, which may in-
volve errors of a few kilocalories per mole on an absolute (un-
calibrated) scale.
Regarding comparison of the copper(II) and zinc(II) catalysts,
both the experimental and calculated activation energies are
lower for the Cu2-2 catalyst than for the Zn2-2 catalyst. From
the experimental rate constants, a lowering of the reaction
barrier by about 1 kcalmolꢀ1 can be inferred on going from
zinc(II) to copper(II). Unfortunately, rigorous comparison of the
absolute values (calculated) is hampered by the uncertainty in
the pKa of the Cu2-2-bound HPNP. Nevertheless, the next step
along the reaction pathway, intramolecular attack of the nucle-
ophile on the central phosphorus atom, is not subject to this
uncertainty, and indeed calculations predict that the barrier for
Cu2-2 is lower by about 3 kcalmolꢀ1
.
To fully exploit the potential of electronic structure calcula-
tions, we analyzed the contributions to and origin of the reac-
tion barrier for copper(II). Two factors were considered:
1) better pre-organization of atoms in the reactant-complex
structure and 2) stabilization of the transition state. The former
can be ruled out, since the bond length between the attacking
nucleophile (oxygen atom of the HPNP hydroxy group) and
the phosphorus atom is shorter and the angle between the
nucleophilic oxygen atom, central phosphorus atom, and the
oxygen atom of the leaving group is closer to 1808 in the
zinc(II) complexes, both of which would favor the attack. This
leaves us with stabilization of the transition-state structure as
the most likely explanation of the higher reactivity of the Cu2-2
species. As can be seen in Table S5 of the Supporting Informa-
tion, the gas-phase energy differences of reactant complexes
and transition states are quite similar for both metal ions,
whereas solvation stabilizes the CuTS1 (CuTS1’) structure to
a greater degree than the ZnTS1 (ZnTS1’) structure (in compari-
son with their respective RC structures), by about 6 kcalmolꢀ1.
Finally, we briefly comment on the accuracy and robustness
of the computed values. As can be seen in Tables S4 and S5 of
the Supporting Information, the calculated reaction energetics
(except for the initial pre-association step discussed above) is
not heavily dependent on the chosen DFT functional. Nine
DFT functionals were used in this study, from which it was as-
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
9
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
certained that most of them provide the same mechanistic pic-
ture, differing by no more than a few kilocalories per mole. At
the same time, we admit that the choice of DFT functionals
used in this work was inspired by the benchmark study of Fer-
nandes and co-workers.[34] In their work, CCSD(T)/CBS//B3LYP/
6-311+ +G(2d,2p) data were employed as a reference and the
hydrolysis pathways of dimethyl phosphate were examined.
Comparing 52 DFT functionals with the reference, they con-
cluded that the H-GGA (hybrid-generalized gradient approxi-
mation) and HM-GGA (hybrid-meta-GGA) functionals, such as
mPWB1K, PBE0, TPSSh, and M06, could be recommended (with
MAD<2 kcalmolꢀ1).
vides new insight into the first key step of the reaction, that is,
deprotonation of the hydroxy group of HPNP, and resolves the
controversy whether this deprotonation occurs by specific or
general base catalysis,[20,21] by yielding some evidence in favor
of the latter. This implies that the deprotonation is part of the
RDS. We believe that this study presents a structurally distinct
class of catalysts of phosphodiester bond cleavage that may in
turn open new avenues in our quest for mimicking nature in
efficient RNA hydrolysis.
In summary, we have introduced a novel dinuclear catalyst
for phosphodiester PꢀO bond cleavage based on conforma-
tionally constrained bis-triazamacrocycles. By virtue of the free
positions on the central benzene ring, the structure of the
ligand may be further synthetically modified. Additional metal
complexes, nucleophiles, or functionalities suitable for supra-
molecular assembly may be attached to the ligand to improve
the catalytic efficiency of this class of artificial metalloenzyme
models.
Conclusions
Effective catalysis of PꢀO bond cleavage in DNA or RNA model
compounds by artificial models mimicking the function of
phosphodiesterases is still a formidable task in aqueous solu-
tion at physiological pH. To date, very few catalysts with this
ability have been reported. In this work, a series of conforma-
tionally constrained mono- and bis-benzoperazacrowns has
been reported.
Experimental Section
Only the key experimental details are presented in this section;
a full account is available in the Supporting Information.
On the basis of preliminary results of UV/Vis kinetic studies
on transesterification of the RNA model HPNP promoted by
zinc(II) and copper(II) complexes of ligands 1–6, we have
shown that only ligands 1 and 2 having triazacrown chains are
active towards catalysis of the title reaction. Further examina-
tion of the reaction revealed Michaelis–Menten saturation ki-
netic behavior with both zinc and copper complexes of ligands
1 and 2, with calculated rate constants kcat on the order of
10ꢀ4 sꢀ1. Pseudo first-order rate constants obtained by fitting
the experimental data indicated approximately three orders of
magnitude acceleration compared with the uncatalyzed reac-
tion. Furthermore, the mononuclear complex Cu-1 merits spe-
cial mention, since the k’obs value of 1.98ꢃ10ꢀ5 sꢀ1 observed at
pH 7.0 and 208C is nearly an order of magnitude greater than
that reported for the copper(II) complex of 1,5,9-triazacyclodo-
decane under similar conditions.
Synthesis of ligand 2
K2CO3 (5.5 g) was added to a solution of 1,2,4,5-tetrakis(bromome-
thyl)benzene (2.9 g, 6.45 mmol) and 1,5,7-triazabicyclo[4.4.0]dec-5-
ene (1.8 g, 12.9 mmol) in acetonitrile (140 mL) and the mixture was
heated to reflux for 2 h. The reaction was monitored by TLC (SiO2;
CH2Cl2/MeOH, 97:3). On consumption of the bromomethyl com-
pound, the reaction mixture was allowed to cool to RT, methanol
(80 mL) was added, and NaBH4 (1.38 g, 36.5 mmol) was added por-
tionwise. The reaction mixture was stirred at RT overnight. The
solids were then filtered off, washed with chloroform, and the
combined filtrate and washings were concentrated to dryness.
Water (30 mL) was added to the residue, the mixture was extracted
with chloroform (3ꢃ60 mL), and the organic phases were com-
bined. After removal of the solvent, 3m HCl (150 mL) was added to
the white solid, and the resulting solution was heated to reflux for
24 h. The mixture was concentrated in vacuo until a white solid
started to precipitate, and then it was kept at 58C for 20 h. The
white solid was collected and dried at 708C/20 Pa. Ligand 2 was
obtained as its hexahydrochloride tetrahydrate in 24% overall yield
By correlating the experimental data for the M2-2 systems
with quantum chemical calculations, we provide a fairly consis-
tent picture concerning the reaction mechanism of the cata-
lyzed PꢀO bond cleavage. It has been shown that the HPNP
substrate binds to both metal ions and takes advantage of
their cooperation. We propose two independent reaction
mechanisms, mainly differing in the coordination mode of the
MB ion. While the reaction has been shown to be a one-step
process with a single transition state when MB is pentacoordi-
nate (with a water molecule occupying the fifth coordination
site), it is likely to be a two-step process when the overall
(5C,4C) coordination mode is considered in the Zn2-2 system.
This was not observed for the copper(II) complex, in which
case the reaction mechanism was a one-step process irrespec-
tive of the coordination number of the CuB ion. Quantum
chemical calculations also indicated that the higher reaction
rate with the Cu2-2 catalyst than with the Zn2-2 catalyst can be
attributed to better solvation of the transition state and ac-
cordingly its enhanced stabilization. The present work also pro-
1
(1.1 g, 1.6 mmol). H NMR (D2O): d=7.78 (s, 2H; ArH), 4.43 (s, 8H;
ArCH2N), 3.38 (m, 16H; NCH2), 2.28 ppm (m, 8H; CH2); 13C NMR
(D2O): d=133.10, 132.96, 45.59, 43.02, 42.13, 20.14 ppm; ESI-MS:
m/z (%): 389 [Mꢀ6HCl+H]+ (100), 425 [Mꢀ5HCl+H]+ (12); HR-ESI-
MS: m/z calcd for C22H41N6 389.3393; found: 389.3392; elemental
analysis calcd (%) for C22H40N6·6HCl·4H2O: C 38.89, H 8.01, Cl
31.31, N 12.37; found: C 39.02, H 7.95, Cl 30.32, N 11.87.
Kinetic studies
Kinetic UV/Vis data were recorded with a Varian CARY 50 CONC
spectrophotometer at 208C in standard cells (10 mm optical path
length). In a typical experiment, ligand 1 (100 mL, 6 mm in water)
and Zn(ClO4)2 (100 mL, 6 mm in water) were added to a UV cell con-
taining HEPES buffer solution (2.7 mL, 10 mm in water). After equili-
bration for 5 min, HPNP (100 mL, 6 mm in water) was added and
the absorbance at l=400 nm was recorded at 1 min intervals over
&
&
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
10
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
a period of 10 h. The obtained data conformed to pseudo first-
order kinetics. The pseudo first-order rate constants k’obs were de-
termined by a nonlinear least-squares fitting method based on
Equation (1). Kinetic parameters were obtained from an average of
at least three independent measurements.
Keywords: cleavage
calculations · ligand design · macrocyclic ligands · reaction
mechanisms
reactions
·
density
functional
The dependence of the initial rate of the reaction on the concen-
tration of the substrate (HPNP) was studied at pH 7 (10 mm aque-
ous HEPES buffer) and 208C. Kinetic UV/Vis data were obtained by
the same procedure as described above. The concentration of
zinc(II) or copper(II) ions (0.2 mm) was the same in all experiments,
while the substrate concentration was varied in each experiment
from 0.2 to 4.0 mm. Michaelis–Menten saturation kinetic behavior
was observed. The Michaelis–Menten parameters were obtained by
fitting the data with Equation (2) above.
[1] N. H. Williams, B. Takasaki, M. Wall, J. Chin, Acc. Chem. Res. 1999, 32,
485–493.
[2] G. K. Schroeder, C. Lad, P. Wyman, N. H. Williams, R. Wolfenden, Proc.
Natl. Acad. Sci. USA 2006, 103, 4052–4055.
[3] a) J. Weston, Chem. Rev. 2005, 105, 2151–2174; b) W. W. Cleland, A. C.
´
Hengge, Chem. Rev. 2006, 106, 3252–3278; c) N. Mitic, S. J. Smith, A.
Neves, L. W. Guddat, L. R. Gahan, G. Schenk, Chem. Rev. 2006, 106,
3338–3363; d) W. L. Ward, K. Plakos, V. J. DeRose, Chem. Rev. 2014, 114,
4318–4342.
[4] a) J. R. Morrow, O. Iranzo, Curr. Opin. Chem. Biol. 2004, 8, 192–200;
b) A. K. Yatsimirsky, Coord. Chem. Rev. 2005, 249, 1997–2011; c) F.
Mancin, P. Tecilla, New J. Chem. 2007, 31, 800–817; d) D. Desbouis, I. P.
Troitsky, M. J. Belousoff, L. Spiccia, B. Graham, Coord. Chem. Rev. 2012,
256, 897–937; e) F. Mancin, P. Scrimin, P. Tecilla, Chem. Commun. 2012,
48, 5545–5559; f) H. Korhonen, N. H. Williams, S. Mikkola, J. Phys. Org.
Chem. 2013, 26, 182–186; g) R. Salvio, Chem. Eur. J. 2015, 21, 10960–
10971; h) M. Subat, K. Woinaroschy, C. Gerstl, B. Sarkar, W. Kaim, B.
Kçnig, Inorg. Chem. 2008, 47, 4661–4668; i) B. Gruber, E. Kataev, J.
Aschenbrenner, S. Stadlbauer, B. Kçnig, J. Am. Chem. Soc. 2011, 133,
20704–20707; and references therein.
[5] E. V. Anslyn, D. M. Perreault, Angew. Chem. Int. Ed. Engl. 1997, 36, 432–
450; Angew. Chem. 1997, 109, 470–490.
[6] a) T. Niittymꢅki, H. Lçnnberg, Org. Biomol. Chem. 2006, 4, 15–25; b) H.
Lçnnberg, Org. Biomol. Chem. 2011, 9, 1687–1703.
[7] M. Wall, R. C. Hynes, J. Chin, Angew. Chem. Int. Ed. Engl. 1993, 32, 1633–
1635; Angew. Chem. 1993, 105, 1696–1697.
[8] S. Liu, A. D. Hamilton, Bioorg. Med. Chem. Lett. 1997, 7, 1779–1784.
[9] S. Liu, Z. Luo, A. D. Hamilton, Angew. Chem. Int. Ed. Engl. 1997, 36,
1633–1635; Angew. Chem. 1997, 109, 1706–1709.
[10] O. Iranzo, T. Elmer, J. P. Richard, J. R. Morrow, Inorg. Chem. 2003, 42,
7737–7746.
[11] R. Cacciapaglia, A. Casnati, L. Mandolini, D. N. Reinhoudt, R. Salvio, A.
Sartori, R. Ungaro, J. Am. Chem. Soc. 2006, 128, 12322–12330.
[12] G. Feng, J. C. Mareque-Rivas, R. T. M. de Rosales, N. H. Williams, J. Am.
Chem. Soc. 2005, 127, 13470–13471.
[13] G. Feng, D. Natale, R. Prabaharan, J. C. Mareque-Rivas, N. H. Williams,
Angew. Chem. Int. Ed. 2006, 45, 7056–7059; Angew. Chem. 2006, 118,
7214–7217.
[14] a) J. Suh, S. H. Hong, J. Am. Chem. Soc. 1998, 120, 12545–12552; b) F.
Manea, F. B. Houillon, L. Pasquato, P. Scrimin, Angew. Chem. Int. Ed.
2004, 43, 6165–6169; Angew. Chem. 2004, 116, 6291–6295; c) F. Aveni-
er, J. B. Domingos, L. D. Van Vliet, F. Hollfelder, J. Am. Chem. Soc. 2007,
129, 7611–7619; d) F. Avenier, F. Hollfelder, Chem. Eur. J. 2009, 15,
12371–12380.
Theoretical studies
Quantum-chemical calculations were carried out with the Turbo-
mole 6.4 program.[35] Geometry optimizations were carried out at
the RI-DFT level of theory by employing the Perdew–Burke–Ernzer-
hof (PBE) functional[36] and the def2-SVP basis set.[37] Normal-mode
analyses (vibrational frequency calculations) were carried out at
the same level of theory. They were used to characterize (and con-
firm) the nature of transition states and to provide gas-phase ther-
mochemical data. Single-point energies were calculated by em-
ploying the PBE, PBE0,[38] B3LYP,[39] B3LYP-D,[40] and TPSSh[41] func-
tionals and Ahlrichs’ def2-TZVP basis set.[42] Another set of func-
tionals, presumably transition-state-barrier calibrated, included
mPW1K,[43] mPWB1K,[44] M06,[45] and wB97xD,[46] employed in con-
junction with the 6-311G+(2d,p) basis set. These single-point cal-
culations were carried out with the Gaussian 09 program.[47] Solva-
tion (free) energies of all studied species were calculated by the
COSMO-RS[31] method (conductor-like screening model for realistic
solvation) as implemented in the COSMOtherm program, by using
the BP_TZVP_C30_1201.ctd parametrization file. The geometries
were first optimized in water (e=80) by using the Becke–Perdew
(BP86) functional[39a,48] and the COSMO implicit solvation model.[49]
COSMO-RS calculations were performed according to the recom-
mended protocol, which includes RI-BP86/def-TZVP calculations
with e=1 (ideal conductor) and e=1 (vacuum) as a prerequisite
for final calculations in water. The Gibbs free energy was subse-
quently calculated as the sum of the following contributions
[Eq. (4)]:
G ¼ Eel þ Gsolv þ EZPVEꢀRT lnðqtransqrotqvibÞ þ pV
ð4Þ
[15] M. Diez-Castellnou, F. Mancin, P. Scrimin, J. Am. Chem. Soc. 2014, 136,
1158–1161.
[16] a) M. Oivanen, S. Kuusela, H. Lçnnberg, Chem. Rev. 1998, 98, 961–990;
b) T. Lçnnberg, H. Lçnnberg, Curr. Opin. Chem. Biol. 2005, 9, 665–673.
[17] H. Korhonen, T. Koivusalo, S. Toivola, S. Mikkola, Org. Biomol. Chem.
2013, 11, 8324–8339.
[18] R. S. Brown, Z.-L. Lu, C. T. Liu, W. Y. Tsang, D. R. Edwards, A. A. Neverov, J.
Phys. Org. Chem. 2010, 23, 1–15.
[19] a) L. Lain, H. Lçnnberg, T. A. Lçnnberg, Org. Biomol. Chem. 2015, 13,
3484–3492; b) I. S. Hong, J. Suh, Org. Lett. 2000, 2, 377–380.
[20] a) Y. Fan, Y. Q. Gao, J. Am. Chem. Soc. 2007, 129, 905–913; b) H. Gao, Z.
Ke, N. J. DeYonker, J. Wang, H. Xu, Z.-W. Mao, D. L. Phillips, C. Zhao, J.
Am. Chem. Soc. 2011, 133, 2904–2915; c) C. I. Maxwell, N. J. Mosey, R. S.
Brown, J. Am. Chem. Soc. 2013, 135, 17209–17222; d) R. Bonomi, G.
Saielli, P. Scrimin, F. Mancin, Supramol. Chem. 2013, 25, 665–671; e) X.
Zhang, X. Zheng, D. L. Phillips, C. Zhao, Dalton Trans. 2014, 43, 16289–
16299; f) X. Zhang, Y. Zhu, H. Gao, C. Zhao, Inorg. Chem. 2014, 53,
11903–11912.
where Eel is the single-point energy in vacuo, EZPVE the zero-point
vibrational energy, Gsolv the COSMO-RS solvation energy, and
RT(lnqtransqvibqrot) accounts for the change in entropic terms and in
the thermal correction to the enthalpy, obtained from the harmon-
ic vibrational frequencies by the ideal-gas, rigid-rotor, harmonic-os-
cillator approximation at T=298.15 K.
Acknowledgements
Financial support from specific university research (MSMT No
20/2015), the Academy of Sciences of the Czech Republic (RVO
61388963), and the Czech Science Foundation (14-31419S, 14-
03636S) is gratefully acknowledged.
[21] M.-Y. Yang, O. Iranzo, J. P. Richard, J. R. Morrow, J. Am. Chem. Soc. 2005,
127, 1064–1065.
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
11
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[22] a) P. Rossi, F. Felluga, P. Tecilla, F. Formaggio, M. Crisma, C. Toniolo, P.
Scrimin, J. Am. Chem. Soc. 1999, 121, 6948–6949; b) A. Scarso, U.
Scheffer, M. Gçbel, Q. B. Broxterman, B. Kaptein, F. Formaggio, C. Tonio-
lo, P. Scrimin, Proc. Natl. Acad. Sci. USA 2002, 99, 5144–5149.
[23] B. Bauer-Siebenlist, F. Meyer, E. Farkas, D. Vidovic, S. Dechert, Chem. Eur.
J. 2005, 11, 4349–4360.
1993, 98, 5648–5652; d) P. J. Stephens, F. J. Devlin, M. J. Frisch, C. F.
Chabalowski, J. Phys. Chem. 1994, 98, 11623–11627.
[40] T. Schwabe, S. Grimme, Acc. Chem. Res. 2008, 41, 569–579.
[41] J. M. Tao, J. P. Perdew, V. N. Staroverov, G. E. Scuseria, Phys. Rev. Lett.
2003, 91, 146401.
[42] A. Schꢅfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829–5835.
[43] B. J. Lynch, P. L. Fast, M. Harris, D. G. Truhlar, J. Phys. Chem. A 2000, 104,
4811–4815.
[24] R. W. Alder, R. W. Mowlam, D. J. Vachon, G. R. Weisman, J. Chem. Soc.
Chem. Commun. 1992, 507–508.
[25] F. Chavez, A. D. Sherry, J. Org. Chem. 1989, 54, 2990–2992.
[26] J. E. Richman, T. J. Atkins, J. Am. Chem. Soc. 1974, 96, 2268–2270.
[44] a) C. Adamo, V. Barone, J. Chem. Phys. 1998, 108, 664–675; b) Y. Zhao,
D. G. Truhlar, J. Phys. Chem. A 2004, 108, 6908–6918.
ˇ
[27] J. Hodacovꢁ, J. Zꢁvada, P. C. Junk, Collect. Czech. Chem. Commun. 1999,
[45] Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215–241.
[46] J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008, 10, 6615–
6620.
64, 1827–1832.
ˇ ˇ
´
ˇ
[28] M. Chadim, M. Budesꢀnsky, J. Hodacovꢁ, J. Zꢁvada, Collect. Czech. Chem.
Commun. 2000, 65, 99–105.
[47] Gaussian 09 (Revision E.01), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢆ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[29] R. Breslow, D. L. Huang, Proc. Natl. Acad. Sci. USA 1991, 88, 4080–4083.
[30] R. Cacciapaglia, A. Casnati, L. Mandolini, D. N. Reinhoudt, R. Salvio, A.
Sartori, R. Ungaro, J. Org. Chem. 2005, 70, 624–630.
[31] A. Klamt, F. Eckert, Fluid Phase Equilib. 2000, 172, 43–72.
ˇˇ
ˇ
[32] O. Gutten, I. Besseovꢁ, L. Rulꢀsek, J. Phys. Chem. A 2011, 115, 11394–
11402.
[33] a) Metal Sites in Proteins and Models: Phosphatases, Lewis Acids and Va-
nadium (Eds.: H. A. O. Hill, P. J. Sadler, A. J. Thompson), Springer, Berlin,
1997; b) R. Gilson, M. C. Durrant, Dalton Trans. 2009, 10223–10230.
[34] A. J. M. Ribeiro, M. J. Ramos, P. A. Fernandes, J. Chem. Theory Comput.
2010, 6, 2281–2292.
[35] R. Ahlrichs, M. Bar, M. Haser, H. Horn, C. Kolmel, Chem. Phys. Lett. 1989,
162, 165–169.
[36] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–
3868.
[37] a) F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305;
b) A. Schꢅfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829–
5835.
[48] J. P. Perdew, Phys. Rev. B 1986, 33, 8822–8824.
[49] A. Klamt, G. Schuurmann, J. Chem. Soc. Perkin Trans. 2 1993, 799–805.
[50] CCDC 1461784 (2) and 1461785 (8) contain the supplementary crystallo-
graphic data for this paper. These data are provided free of charge by
The Cambridge Crystallographic Data Centre.
[38] C. Adamo, M. Cossi, V. Barone, J. Mol. Struct. THEOCHEM 1999, 493,
145–157.
[39] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) C. Lee, W. Yang,
R. G. Parr, Phys. Rev. B 1988, 37, 785–789; c) A. D. Becke, J. Chem. Phys.
Received: March 11, 2016
Published online on && &&, 0000
&
&
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
12
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
Artificial metalloenzyme models: Cata-
lytic activities of mono- and dinuclear
copper(II) and zinc(II) complexes of con-
formationally constrained polyazama-
crocycles towards cleavage of the PꢀO
bond in the RNA model substrate 2-hy-
droxypropyl-4-nitrophenylphosphate
(HPNP) in aqueous solution at pH 7 are
reported. A reaction mechanism of the
catalyzed PꢀO bond cleavage (see
figure) is proposed on the basis of ex-
perimental results and quantum chemi-
cal calculations.
&
Cleavage Reactions
D. Bꢀm, E. Svobodovꢁ, V. Eigner,
ˇ
ˇ
L. Rulꢀsek,* J. Hodacovꢁ*
&& – &&
Copper(II) and Zinc(II) Complexes of
Conformationally Constrained
Polyazamacrocycles as Efficient
Catalysts for RNA Model Substrate
Cleavage in Aqueous Solution at
Physiological pH
Macrocyclic Ligands
The catalytic activity of mono- and dinuclear copper(II) and
zinc(II) complexes of conformationally constrained
polyazamacrocycles towards cleavage of the PꢀO bond in
the RNA model substrate 2-hydroxypropyl-4-
nitrophenylphosphate (HPNP) in aqueous solution at pH 7
are presented. For more details, see the Full Paper by L.
ˇ
ˇ
Rulꢀsek, J. Hodacovꢁ, et al. on page && ff.
Chem. Eur. J. 2016, 22, 1 – 13
www.chemeurj.org
13
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ