Switching between inner- and outer-sphere PCET mechanisms of small-molecule activation: Superoxide dismutation and oxygen/superoxide reduction reactivity deriving from the same manganese complex

Article abstract of DOI:10.1021/jacs.6b08394

Readily exchangeable water molecules are commonly found in the active sites of oxidoreductases, yet the overwhelming majority of studies on small-molecule mimics of these enzymes entirely ignores the contribution of water to the reactivity. Studies of how these enzymes can continue to function in spite of the presence of highly oxidizing species are likewise limited. The mononuclear Mn^II complex with the potentially hexadentate ligand N-(2-hydroxy-5-methylbenzyl)-N,N′,N′-tris(2-pyridinylmethyl)-1,2-ethanediamine (L^OH) was previously found to act as both a H₂O₂-responsive MRI contrast agent and a mimic of superoxide dismutase (SOD). Here, we studied this complex in aqueous solutions at different pH values in order to determine its (i) acid-base equilibria, (ii) coordination equilibria, (iii) substitution lability and operative mechanisms for water exchange, (iv) redox behavior and ability to participate in proton-coupled electron transfer (PCET) reactions, (v) SOD activity and reductive activity toward both oxygen and superoxide, and (vi) mechanism for its transformation into the binuclear Mn^II complex with ^(H)OL-L^OH and its hydroxylated derivatives. The conclusions drawn from potentiometric titrations, low-temperature mass spectrometry, temperature- and pressure-dependent ¹⁷O NMR spectroscopy, electrochemistry, stopped-flow kinetic analyses, and EPR measurements were supported by the structural characterization and quantum chemical analysis of proposed intermediate species. These comprehensive studies enabled us to determine how transiently bound water molecules impact the rate and mechanism of SOD catalysis. Metal-bound water molecules facilitate the PCET necessary for outer-sphere SOD activity. The absence of the water ligand, conversely, enables the inner-sphere reduction of both superoxide and dioxygen. The L^OH complex maintains its SOD activity in the presence of ^?OH and Mn^IV-oxo species by channeling these oxidants toward the synthesis of a functionally equivalent binuclear Mn^II species. (Figure Presented).

Full text of DOI:10.1021/jacs.6b08394

Article
pubs.acs.org/JACS
Switching between Inner- and Outer-Sphere PCET Mechanisms of
Small-Molecule Activation: Superoxide Dismutation and Oxygen/
Superoxide Reduction Reactivity Deriving from the Same Manganese
Complex
Isabell Kenkel,† Alicja Franke,† Maximilian Durr,† Achim Zahl,† Carlos Ducker-Benfer,† Jens Langer,†
̈
̈
Milos R. Filipovic,† Meng Yu,‡ Ralph Puchta,† Stephanie R. Fiedler,§ Matthew P. Shores,§
́
Christian R. Goldsmith,*,‡ and Ivana Ivanovic-Burmazovic*,†
́
́
†Department of Chemistry and Pharmacy, University Erlangen-Nuremberg, 91058 Erlangen, Germany
‡Department of Chemistry and Biochemistry, Auburn University, Auburn, Alabama 36849, United States
§Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523-1872, United States
S
* Supporting Information
ABSTRACT: Readily exchangeable water molecules are commonly found in
the active sites of oxidoreductases, yet the overwhelming majority of studies on
small-molecule mimics of these enzymes entirely ignores the contribution of
water to the reactivity. Studies of how these enzymes can continue to function in
spite of the presence of highly oxidizing species are likewise limited. The
mononuclear MnII complex with the potentially hexadentate ligand N-(2-
hydroxy-5-methylbenzyl)-N,N′,N′-tris(2-pyridinylmethyl)-1,2-ethanediamine
(LOH) was previously found to act as both a H2O2-responsive MRI contrast
agent and a mimic of superoxide dismutase (SOD). Here, we studied this
complex in aqueous solutions at different pH values in order to determine its (i)
acid−base equilibria, (ii) coordination equilibria, (iii) substitution lability and
operative mechanisms for water exchange, (iv) redox behavior and ability to
participate in proton-coupled electron transfer (PCET) reactions, (v) SOD
activity and reductive activity toward both oxygen and superoxide, and (vi)
mechanism for its transformation into the binuclear MnII complex with (H)OL−LOH and its hydroxylated derivatives. The
conclusions drawn from potentiometric titrations, low-temperature mass spectrometry, temperature- and pressure-dependent
17O NMR spectroscopy, electrochemistry, stopped-flow kinetic analyses, and EPR measurements were supported by the
structural characterization and quantum chemical analysis of proposed intermediate species. These comprehensive studies
enabled us to determine how transiently bound water molecules impact the rate and mechanism of SOD catalysis. Metal-bound
water molecules facilitate the PCET necessary for outer-sphere SOD activity. The absence of the water ligand, conversely,
enables the inner-sphere reduction of both superoxide and dioxygen. The LOH complex maintains its SOD activity in the
presence of •OH and MnIV-oxo species by channeling these oxidants toward the synthesis of a functionally equivalent binuclear
MnII species.
bulkiness of the ligands, and each of these has been
demonstrated to impact the reactivity of the metal centers.1
There is evidence that the primary ligand sphere, however, is
not the sole determinant of activity, that crucial contributions
may come from interactions with solvent molecules, and that
these interactions can be tuned via acid−base equilibria.2 Many
oxidoreductases are believed to function via proton-coupled
electron transfer (PCET) mechanisms that rely on water/
hydroxide/oxide ligands to provide or abstract the protons and
to modulate the redox potentials of the involved metal ions. An
excellent illustration of the impact of the coordinated solvent
INTRODUCTION
■
Determining the factors that control the solution behavior and
redox reactivity of metal centers is essential both for optimizing
small-molecule catalysts and for understanding the operation of
metal-containing oxidoreductases. Small-molecule models have
long been used in bioinorganic chemistry to provide insight
into how metalloenzymes function. In these models, a
polydentate ligand is commonly used to approximate the
peptide-derived coordination sphere provided by the enzyme.
Most synthetic efforts have focused on making changes to the
polydentate ligand in order to modulate the geometry,
stereochemistry, and overall charge of the metal complexes;
the identities of the donor atoms; and the rigidity/flexibility and
Received: August 14, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
molecules (H2O or OH−) on the redox properties of
biologically relevant metal centers is provided by Mn and Fe
superoxide dismutase (SOD) enzymes and their Fe- or Mn-
substituted analogues.3−6 Even though these interactions are
widely acknowledged to be important, most small-molecule
mimics do not contain coordinated water molecules and are
studied in nonaqueous solutions. The active sites of SOD
enzymes, for instance, are often approximated by hexa- and
heptacoordinate Mn complexes with polydentate amino-
pyridines and one or more monodentate ligands;2,7,8 these
monodentate ligands, usually MeCN or Cl− (neither of which is
found in the native enzyme), remain coordinated in non-
aqueous solutions.2 Although such studies nicely illustrate how
changes to the pyridylamine ligand influence the redox
properties of the bound Mn related to SOD activity, they do
not address how the coordination of water can mediate its
reactivity.2
That motivated us to study the aqueous chemistry of the
[(LOH)MnII(solv)](ClO4)2 complex (1), where LOH is the
potentially hexadentate ligand N-(2-hydroxy-5-methylbenzyl)-
N,N′,N′-tris(2-pyridinylmethyl)-1,2-ethanediamine. The ligand
is closely related to the pyridylamine ligands used in the
aforementioned studies2 and can be thought of as a close
analogue of N,N,N′,N′-tetrakis(2-pyridylmethyl)-1,2-ethanedi-
amine (tpen).9 The LOH and tpen ligands differ in that the
former contains a phenol in place of one of the pyridine
moieties; the phenol may approximate the tyrosine found in the
active sites of Mn and Fe SODs.10 Despite the lack of a
macrocyclic component or a carboxylate group, the MnII
complex with LOH was found to be stable enough in water to
serve as a H2O2-responsive contrast agent for magnetic
resonance imaging (MRI).11 The hexadenticity, coupled with
the ability of MnII to accommodate heptacoordination, allows
for the binding of a single monodentate ligand. Subsequent
work suggested that 1 may serve as a functional mimic of
SOD.12
present in the catalytic cycle responsible for superoxide
dismutation. The role of the coordination equilibrium is
reminiscent of the reactivity of the Mn SOD enzyme in that its
catalytic cycle is proposed to alternate between six- and five-
coordinate active sites.13−15 The observed mechanistic behavior
was compared to that of one of the most efficient classes of
MnSOD mimetics, i.e., pentaazamacrocyclic SOD mimetics,7,16
that also exhibit coordination equilibria between hepta- and
hexacoordinate species, and general conclusions were drawn.
EXPERIMENTAL SECTION
■
Materials and Instrumentation. Except where noted otherwise,
chemicals were purchased from Sigma-Aldrich and used as received.
Anhydrous acetonitrile (MeCN) was purchased from Acros Organics
and stored in a glovebox free of moisture and oxygen. Hydrogen
peroxide (H2O2, 50 wt %) was bought from Fisher. Dry nitrogen (N2)
and oxygen (O2) were purchased from Airgas. N-(2-Hydroxy-5-
methylbenzyl)-N,N′,N′-tris(2-pyridinylmethyl)-1,2-ethanediamine
(LOH), [(LOH)MnII(MeCN)](ClO4)2 (1), and [(HOL−LO−)-
MnII (MeCN)2](ClO4)3 (2) were prepared as previously described.11
2
Deionized Millipore water was used for all aqueous measurements. All
buffers were treated with Chelex sodium exchange resin to remove all
trace metals.
Determination of in Vitro SOD Activity via Stopped-Flow
Technique. Superoxide dismutase activity was tested by a direct
method using a stopped-flow technique described elsewhere.17
Experiments were carried out on a Biologic SFM-400 instrument,
using syringes 1, 2, and 3, equipped with an Energetiq LDLS ENQ
EQ-99-FC laser-driven light source and a J&M TIDAS diode array
detector (integration time 0.5 ms, λ = 180−724 nm). The source of
superoxide was commercially available KO2 dissolved in dry DMSO
([O2•−] ≈ 1−2 mM). Each complex was tested at four different
concentrations between 0.9 and 9 μM in aqueous solutions buffered
with HEPES or sodium phosphate to either pH 7.4 or pH 8.1. The
aqueous solution containing the MnII complex was mixed in a 9:1 ratio
with the superoxide solution in DMSO using a high-density mixer. In
each experiment, the concentration of superoxide exceeded that of the
MnII complex by at least 10-fold in order to ensure catalytic
conditions. Millipore water was used for the preparation of the buffer
solutions; the buffers were treated with Chelex 100 sodium exchange
resin for at least 12 h before use in order to remove adventitious metal
ions. The data analysis was performed using BioKine V4.66 software.
Each reported kobs value is the average of at least 10 measurements. kcat
values were determined from the slope of the kobs vs [Mn] plot.
Cryo-MS Measurements. Cryospray-ionization MS (CSI-MS)
measurements were performed on a UHR-TOF Bruker Daltonik
(Bremen, Germany) maXis plus, an ESI-quadrupole-time-of-flight
(qToF) mass spectrometer capable of resolution of at least 60 000
fwhm, which was coupled to a Bruker Daltonik Cryospray unit.
Detection was in positive-ion mode, with a source voltage of 4.5 kV.
The flow rates were 250 μL/h. Both the N2 drying gas (N2), which
aided solvent removal, and the spray gas were maintained at
temperatures of 5 °C for aqueous solutions. For MeCN solutions,
the temperature of drying gas was kept at −20 °C and the spray gas at
−40 °C. The machine was calibrated prior to each experiment via
direct infusion of the Agilent ESI-TOF low concentration tuning
mixture, which provided an m/z range of singly charged peaks up to
2700 Da in both ion modes.
In order to better understand the solution behavior and
reactivity of bioinorganic model systems in aqueous solutions,
we have conducted a more thorough study of 1. We have
assessed the water exchange dynamics of the MnII-bound water
molecule, the speciation of 1 between various six- and seven-
coordinate complexes, the redox activity as a function of pH,
and the reactivity with superoxide and dioxygen. One of our
groups previously reported that the LOH ligand in 1 undergoes
oxidative coupling upon reaction with H2O2 in either MeCN or
water to yield the binuclear complex [(HOL−LO−)-
MnII (solv)2](ClO4)3 (2).11 Here, we demonstrate that this
2
coupling reaction protects the catalytic sites from the reactive
oxygen species (ROS) and MnIV-oxo species that are produced
outside of the mechanism that is predominantly responsible for
SOD mimicry. Instead of degrading and deactivating the
mononuclear complex, these oxidants transform it into the
more stable 2, which displays SOD activity that is comparable
to two equiv of 1.
The ability of 1 to toggle between catalytic superoxide
dismutation and the reduction of oxygen to yield ROS and
MnIV-oxo species is, to the best of our knowledge, unique. The
reactivity switches between outer- (OS) and inner-sphere (IS)
PCET mechanisms of small-molecule activation. Whether the
OS or IS mechanism is operable is determined by the
coordination equilibrium between seven- and six-coordinate
[(LO−)MnII(OH2)]+ and [(LO−)MnII]+ species and the lesser
lability of the related five-coordinate [(LO−)MnIII(OH−)]+ form
Cyclic Voltammetry. Redox potentials were determined on an
Autolab instrument equipped with a PGSTAT 101 potentiostat and
the following: a Au disk working electrode (A = 0.07 cm2, Metrohm), a
Pt auxiliary electrode (Metrohm), and a Ag/AgCl reference electrode
(3 M NaCl, Metrohm, 0.222 V vs NHE at 20 °C). Solid samples of the
complex were dissolved in small amounts of MeCN, which were then
added to a larger volume of the respective aqueous buffer to yield a 1
mM solution. The ionic strength of the solution was adjusted by the
addition of LiClO4 to achieve a concentration of 0.1 M. All solutions
were degassed with N2 before measurement. All experiments were
performed under N2.
B
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Potentiometric Titration. Potentiometric titration was performed
on a Metrohm 702 SM Titrino in a jacketed, airtight glass titration cell
equipped with a pH glass electrode (Metrohm, 3 M NaCl), an inlet
and an outlet for N2, and a graduated 10 mL microburet (Metrohm).
The pH electrode was calibrated before measurement using
commercially available standard buffer solutions (pH 4.0, 7.0, and
10.0). For titration, a 1 mM solution of the complex and a carbonate-
free 0.05 M solution of NaOH were used. Sodium perchlorate was
added to adjust the ionic strength of the solutions to 0.1 M. All
solutions were degassed before use, and a nitrogen atmosphere was
maintained during the experiment. The temperature was kept constant
at 21 °C by a thermostat. Data analysis was performed using the
software TITFIT.18
CCDC 1496504 contains the supplementary crystallographic data
for [(LO−)MnII](ClO4)·1.5H2O. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif. (For X-ray crystallography of
[(LOH)MnII(OAc)](ClO4)·nH2O see SI.)
Quantum Chemical Methods. Density functional theory
calculations were performed using the combination of exchange and
correlation PW91-functionals;24−26 this treatment yielded results
comparable to those obtained with hybrid functional B3LYP27−29
calculations. The performances of these functionals are well-
documented.30,31 In all cases, we applied the def2svp basis set.32,33
For the PW91 calculations, the density fitting approach was
applied.34,35 All structures were characterized as local minima by
computation of vibrational frequencies, and the stability of the wave
function was successfully tested. The Gaussian 09 suite of programs
was used.36
Water Exchange: 17O NMR Measurements. 17O NMR spectra
were recorded on a Bruker AVANCE DRX 400WB spectrometer
equipped with a spectrospin superconducting widebore magnet
operating at a resonance frequency of 54.24 MHz at a magnetic
induction of 9.4 T. The measurements at atmospheric pressure were
performed with a commercial 5 mm Bruker broadband probe
thermostated with a Bruker B-VT 3000 variable temperature unit.
Relaxation rates were measured for both the paramagnetic solutions
and metal-free aqueous buffer solutions. The line widths at half-height
of the signal were determined by a deconvolution procedure on the
real part of the Fourier transformed spectra with a Lorentzian shape
function in the data analysis module of Bruker Topspin 1.3 software.
Pressure-dependent measurements were done with a custom-made
thermostated high-pressure probe.19 The sample was measured in a
standard 5 mm NMR tube cut to a length of 50 mm. To enable
pressure transmittance to the solution, the NMR tube was closed with
a moveable Macor piston. The advantage of this method is that
oxygen-sensitive samples can be easily placed in the NMR tube and
sealed with the Macor piston under an argon atmosphere. This
facilitates subsequent transfer of the sample to the high-pressure
probe. The pressure was applied to the high-pressure probe via a
perfluorinated hydrocarbon (hexafluoropropylene oxide, Hostinert
175, Hoechst) and measured by a VDO gauge with an accuracy of
1%. Temperature was adjusted with circulating, thermostated water
(Colora thermostat WK 16) to 0.1 K of the desired value and
monitored before each measurement with an internal Pt-resistance
thermometer with an accuracy of 0.2 K. Enriched (10%) 17O-labeled
water (D-Chem Ltd. Tel Aviv, Israel) was used for the 17O NMR
water exchange experiments. Samples were prepared by dissolving
solid [(LOH)MnII(MeCN)](ClO4)2 in MeCN and refilling the sample
with degassed buffer solution to get a 5 mM complex solution
(containing 10% MeCN). To this solution, 10%-enriched 17O-labeled
water was added, and the resultant solution was transferred to a NMR
tube. For measurements at pH 7.4, 60 mM HEPES (4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid) buffer was used since
it shows a pKa value that is almost temperature- and pressure-
independent. A detailed description of the experiments at pH 4.8 and
general data treatment can be found in the Supporting Information
(SI).
RESULTS AND DISCUSSION
■
Speciation of [(LOH)MnII(solv)](ClO4)2 (1) in Aqueous
(An)aerobic Solutions at Different pH Values: Potentio-
metric Titration, 17O NMR Water Exchange Measure-
ments, X-ray Structures, and CSI-Mass Spectrometry.
The [(LOH)MnII(MeCN)](ClO4)2 (1) complex was previously
studied in organic solvents, and the structure of a species
crystallized from MeCN solution was reported.11,12 To better
understand the behavior of 1 in aqueous solutions, we
investigated the protonation state of the LOH ligand, the
number and protonation states of coordinated water molecules,
and the overall coordination number of the MnII center as
functions of pH. Analysis of potentiometric titration (see Figure
1, Table 1) suggests the existence of four MnII species from pH
Figure 1. Species distribution of [(LOH)MnII(solv)](ClO4)2 (1) in
aqueous solution at 21 °C.
EPR and Magnetic Measurements. Methods for the EPR and
magnetic measurements are outlined in the SI.
X-ray Crystallography. The intensity data for complexes were
collected with an Agilent Technologies SuperNova, Dual, AtlasS2
diffractometer. The crystal of compound [(LO−)MnII](ClO4)·1.5H2O
was a nonmerohedral twin. The twin law, relating the two twin
2 to 10; these correspond to the blue, pink, green, and orange
curves in Figure 1. The observed acid−base, coordination, and
redox equilibria (vide infra) are summarized in Scheme 1.
Water exchange experiments were monitored by 17O NMR at
pH 4.8 and 7.4; the titration data suggest that a single species
accounts for over 80% of the MnII at these points. Additional
CSI-MS experiments were performed to better understand the
species distribution.
Since the reactivity studies of the species present in solution
at pH >6 are of higher biological relevance, we will further
focus on the speciation of 1 under these experimental
conditions. Characterization of the species at pH <6 is provided
in the SI.
20
components, was determined by CrysAlisPro as (1.0003, 0.0014,
0.0009/−0.0010, −1.0000, 0.0007/−0.9849, −0.0021, −1.0011).
Using Olex2,21 the structure was solved by direct methods (ShelXT)22
and refined with ShelXL23 using least squares minimization. The
fractional contribution of the minor twin component was refined to
0.4614(10). All non-hydrogen atoms were refined anisotropically.
Most of the hydrogen atoms have been placed on calculated positions
and were refined isotropically in a riding model. The positions of the
hydrogen atoms of the three water molecules were located by
difference Fourier synthesis and refined with an isotropic thermal
parameter 1.5 times that of the attached oxygen atom.
C
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Temperature- and pressure-dependent water exchange experi-
ments (Figure 2 and Table 2) show the exchange of one water
molecule. However, the water exchange process is one order of
magnitude faster than at pH 4.8, where the hexacoordinate
species with one aqua ligand and noncoordinated protonated
phenol group exists (Scheme 1; for details see the SI). The
mechanism changes its character from an associative to a
dissociative one, which is best demonstrated by the positive
activation volume (Table 2). These results can be explained by
the coordination of the phenolate anion and the existence of a
seven-coordinate structure in the ground state, with the singly
deprotonated LO− ligand providing six donor atoms in addition
to the one from the aqua ligand. As previously mentioned,
water exchange involving seven-coordinate MnII centers usually
proceeds through Id mechanisms, which display ΔV⧧ values
between approximately +3 and +5 cm3 mol−1.38−41 The
measured ΔV⧧ of +8.1 cm3 mol−1 for 1 at pH 7.4 is
significantly more positive (Table 2), and the fast exchange
rate of 1.3 × 108 s−1 at 298 K suggests an additional labilization
effect. Such fast water exchange on seven-coordinate MnII has
only been observed for overall negatively charged complexes
with polyanionic ligands, such as EDTA.39−41 Thus, both the
volume of activation and the rate constant indicate the possible
agency of a pure dissociative (D) mechanism, which has been
unknown for ligand exchange processes involving MnII species.
A D mechanism would imply the existence of a six-coordinate
Table 1. pMn Values, Stability Constant, and pKa Values for
the Ligand and Complex 1
a
pKL1
pKL2
4.11
10.01
3.63
3.93
6.16
6.51
5.40
a
b
pKa([(H2LOH)2+MnII]4+)
c
pKa([(HLOH)+MnII]3+)
d
pKa([(LOH)MnII]2+)
e
logK([(LOH)Mn]2+)
f
pMn (7.4)
a
Ligand pKa values related to the (de)protonation of the free ligand
(protonated pyridine and phenol groups, respectively), KL1
=
[ L O H ] [ H + ] / [ ( H L O H ) + ] , K L 2
=
[ L O − ] [ H + ] / [ L O H ] .
b
4 +
Ka ([(H2 LO H
)
2 + MnI I
]
)
=
[[(HLO H )+ MnI I
]
3 + ][H+ ]/
[[(H2LOH)2+MnII]4+], related to a protonated pyridine group in the
c
complex. Ka([(HLOH)+MnII]3+
) =
[[(LOH)MnII]2+][H+]/
[[(HLOH)+MnII]3+], related to a protonated pyridine group in the
d
complex. Ka([(LOH)MnII]2+) = [[(LO−)MnII]+][H+]/[[(LOH)-
MnII]2+], related to the deprotonation of the phenol group in the
e
f
complex. K([(LOH)MnII]2+) = [[(LOH)MnII]2+]/[Mn2+][LOH]. pMn
= −log [Mn2+]free calculated for [complex] = 1 mM, T = 21 °C, pH
7.4.
At pH values above 6, the phenol in the LOH ligand is
deprotonated (Scheme 1 and Figure 1). The pKa value
obtained for coordinated phenol is 6.16 (Table 1), which is
similar to values measured for other MnII−phenol complexes.37
Scheme 1. Species Involved in Acid−Base, Coordination, and Redox Equilibria
D
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
negatively charged polydentate ligands, the Mn−Ocarboxylate
bond distances were found to be 2.161 and 2.1809 Å for
[(H21,2-pdta)MnII(OH2)] and [(edta)MnII(OH2)]2−, respec-
tively.43,44 Complexes with EDTA derivatives have been
documented to undergo water exchange through Id mecha-
nisms.40 The greater strength of the Mn−Ophenolate bond may
necessitate a D, rather than an Id, mechanism for water
exchange for the seven-coordinate [(LO−)MnII(OH2)]+. The
strong coordination of the phenolate anion probably plays a
role in the labilization of the aqua ligand, enabling the isolation
of the non-water-bound six-coordinate intermediate predicted
by the D mechanism.
The deprotonated phenol interacts via hydrogen bonds with
a water molecule that is in close proximity to the vacant
coordination site on MnII and is part of a hydrogen bond
network that involves two additional water molecules and the
coordinated phenolate of a second complex unit (see Figure 3).
The hydrogen bond network can assist in rapid water exchange
and formation of the six-coordinate intermediate that is in fast
equilibrium with the seven-coordinate species. The six-
coordinate [(LO−)MnII]+ form has also been detected by
CSI-MS (m/z = 507.1815) as the main species in solution at
pH 7.4 (see Figure S12, SI). The EPR studies of the
[(tpen)MnII]2+ complex suggested that it is structurally
heterogeneous in frozen MeCN solution.2 Our results lead us
to believe that a seven−six-coordination equilibrium may be
operable for this system as well.
Electrochemistry. Cyclic voltammetry measurements
demonstrate that the redox behavior of 1 is highly dependent
on pH. Above pH 6 (e.g., pH 6.5 in Figure 4), one clear
oxidation (0.546 V vs Ag/AgCl) and one reduction (0.386 V vs
Ag/AgCl) signal appear. The peak separation of 160 mV is
consistent with a quasi-reversible redox process, which we
attribute to the subsequent deprotonation of a bound water
molecule upon oxidation of MnII to MnIII.2 There is a linear
correlation between E1/2 and pH (Figure S4, Table S4, SI), with
a slope that is consistent with the transfer of one proton per
electron, corresponding to the [(LO−)MnII(OH2)]+/[(LO−)-
MnIII(OH−)]+ couple. At lower pH, the MnII could not be
oxidized. The inability to deprotonate the phenol group and the
aqua ligand under these conditions precludes the stabilization
of a MnIII species. At more positive potentials (higher than 0.5
V; see Figure S5, SI), an additional oxidation signal becomes
clearly observable, which also shifts to lower potentials and
becomes more prominent by increasing the pH; at the same
time, the reduction signal almost disappears by scanning toward
cathodic potentials.
Figure 2. (a) Temperature dependence of the reduced transverse
relaxation rate (ln(1/T2r)) for the [(LO−)MnII(OH2)]+ complex at pH
7.4 (see eqs S4, S7, and S8 in the SI). (b) Pressure dependence of
ln(kex) at pH 7.4 and 298 K (see eq S6 in the SI).
intermediate with no bound water molecule. We crystallized
this species from a neutral aqueous solution (see Figure 3). The
MnII center is coordinated in a distorted octahedral environ-
ment by the phenolate oxygen and the five nitrogen donors
from the LO− ligand. The distorted geometry leaves a quite
large gap between N4, N5, and O1 for the coordination of the
aqua ligand during the water exchange process (see Figure 3
and Table 3). As in the structure obtained from MeCN,11 the
O-donor from the polydentate ligand participates in hydrogen
bond interactions with outer-sphere water molecules.
The Mn−O bond in [(LO−)MnII]+ is much shorter (2.037 Å)
than that in the crystal structure of [(LOH)MnII(MeCN)]2+,
which contains a neutral phenol group (Mn−O = 2.367 Å).11
The contraction of the Mn−O bond provides strong evidence
for the assignment of a deprotonated phenol group in the
[(LO−)MnII]+ structure. It should be mentioned that the
phenolate anion in [(LO−)MnII]+ is anticipated to coordinate
more strongly to MnII than carboxylate groups, based on the
shorter Mn−O distance in 1 compared to MnII complexes that
contain one monodentate carboxylate (average Mn−O
This additional process can be assigned to the formation of a
MnIV-oxo species. The high reactivity associated with MnIV-oxo
species45−47 provides an explanation for the disappearance of
the reduction signal and our inability to observe it even with
distance: 2.18
0.07 Å).42 In the MnII complexes of highly
Table 2. Activation Parameters for the Water Exchange on Complex 1 {six-coordinate [(LOH)MnII(OH2)]2+ and seven-
coordinate [(LO−)MnII(OH2)]+} Depending on pH
pH 4.8, [(LOH)MnII(OH2)]2+
pH 7.4, [(LO−)MnII(OH2)]+
nH2O
1
1
kex (s‑1)
1.9 × 107
26.4 0.2
−17.0 0.7
−10.9 0.4
1.3 × 108
26.6 0.5
−0.65 1.68
+8.1 0.4
298
ΔH⧧ (kJ mol‑1)
ΔS⧧ (J mol‑1 K‑1)
ΔV⧧ (cm3 mol‑1)
E
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. Molecular structure of the cation of [(LO−)MnII](ClO4)·1.5H2O (a) and two symmetry-independent [(LO−)MnII]+ units bridged by a
chain of three molecules of water (b). The displacement ellipsoids represent a probability of 50%; counterion, solvent molecules, and hydrogen
atoms are omitted for clarity.
Table 3. Selected Bond Lengths and Angles from the Structure of [(LO−)MnII](ClO4)·1.5H2O
bond
distance (Å)
bond
angle (deg)
bond
angle (deg)
N1−Mn1
N2−Mn1
N3−Mn1
N4−Mn1
N5−Mn1
O1−Mn1
O1−C27
2.311(3)
2.270(2)
2.321(2)
2.194(2)
2.251(3)
2.037(6)
1.334(4)
N1−Mn1−N2
N1−Mn1−N3
N1−Mn1−N4
N1−Mn1−N5
N1−Mn1−O1
N2−Mn1−N3
N2−Mn1−N4
N2−Mn1−N5
74.61(9)
104.26(9)
92.07(9)
171.80(9)
87.93(9)
80.66(9)
147.95(9)
97.27(9)
N2−Mn1−O1
N3−Mn1−N4
N3−Mn1−N5
N3−Mn1−O1
N4−Mn1−N5
N4−Mn1−O1
N5−Mn1−O1
91.13(8)
74.50(9)
75.11(9)
162.76(1)
95.59(9)
117.87(9)
91.07(9)
faster scan rates. The complex likely quickly reacts further prior
to the reduction scan.
on the E1/2 values,2 the oxidation of the water-bound complex 1
is much more thermodynamically favored. The differences
between [(LO−)MnII(OH2)]+ and [(tpen)MnIICl]+ also illus-
trate the importance of water coordination as a means to
facilitate PCET and thereby enable novel reactive pathways.
SOD Activity. The observed PCET redox behavior of the
complex prompted us to investigate its activity toward oxygen
species. Manganese complexes that exhibit the proper redox
potentials and coordination equilibria have been found to be
potent SOD mimetics,38,39,48−51 imitating the function of the
manganese cofactor in mitochondrial SOD that cycles between
five- and six-coordinate geometries during catalysis.15 Thus, we
followed the reaction between the [(LO−)MnII(OH2)]+/
[(LO−)MnII]+ complex and superoxide by using an established
direct stopped-flow assay in which a DMSO solution of KO2 is
rapidly mixed with an aqueous solution containing catalytic
amounts of the complex.17 The SOD activity was studied in
three different aqueous solutions: 60 mM HEPES at pH 8.1, 60
mM HEPES at pH 7.4, and 50 mM phosphate buffer at pH 7.4.
As anticipated, the reactions were rapid, and the superoxide
vanishes via a first-order process. The rate law suggests catalytic
decomposition of superoxide. Superoxide will degrade without
a catalyst, but this reaction is second-order with respect to
The species involved in the acid−base and redox equilibria
are summarized in Scheme 1. Notably, the related seven-
coordinate [(tpen)MnIIICl]2+/[(tpen)MnIICl]+ redox couple,
though studied in MeCN, exhibits a redox potential at a ca. 0.6
V more positive value (E1/2 = 1.05 V vs SCE).2 The second
anodic process attributed to the formation of a MnIV species
could be detected only at −20 °C at a 1 V more positive
potential (+1.85 V vs SCE) than that for 1. The much lower
redox potentials required for the generation of the MnIII and
MnIV species from 1 allow us to observe these oxidized species
merely by dissolving the MnII precursor in aerobic buffer
solutions at pH >6 (the reactivity toward oxygen is discussed
further below), whereas the oxidation of [(tpen)MnIICl]+ and
other related MnII complexes with pyridylamine ligands can
only be achieved by either bulk electrolysis or the addition of a
strong oxidizing agent, such as H2O2 or mCPBA.2 This
dichotomy vividly illustrates how a PCET mechanism can
render a difficult redox event into a thermodynamically
favorable process. Even though the structurally related
[(LO−)MnII(OH2)]+ and [(tpen)MnIICl]+ complexes hold the
same overall charge, which remains the predominant influence
F
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure S13, SI). More basic conditions facilitate the oxidation of
the MnII in saturated oxygen solutions; under these conditions,
the oxidation can be followed spectrophotometrically. A new
UV/vis absorption band develops at 420 nm, a wavelength that
is characteristic for MnIII-related LMCT bands (Figure S7a,
SI).2,54,55 These compounds include [(tpen)MnIIICl]2+, which
coincidentally also features a UV/vis band at 420 nm, albeit in
MeCN.2 We observed the same UV/vis spectral features upon
electrochemical generation of Mn(III) at 500 mV (vs Ag wire),
which confirmed the characterization of the product of the
reaction with oxygen (see SI material and Figure S7b).
One notable difference between MnII and MnIII is that MnIII
does not support seven-coordinate geometry.56 Heptacoordi-
nate MnIII of a capped−trigonal prism geometry was only
reported in the solid state for the aforementioned [(tpen)-
MnIIICl]2+, which is structurally analogous to 1.2 However, in
that structure, two Mn−N bonds are substantially elongated
with average bond lengths of 2.47 Å, and it is questionable
whether it retains its heptacoordinate geometry in solution.
Parallel computational studies revealed that the [(LO−)-
MnIII(OH−)]+ cation is more stable with a pentacoordinate
geometry, since two of the Mn−N distances (d = 2.81 and 2.63
Å) are inconsistent with Mn−N bonds (Figure 5a). This is
again a consequence of the strong coordination of the
phenolate and hydroxo groups; this induces additional
labilization of the Mn−N bonds in comparison to the structure
of [(tpen)MnIIICl]2+, which has only one negatively charged
donor atom, from the more weakly binding chlorido ligand.2 By
way of comparison, our calculated average Mn−O bond length
of 1.88 Å (Table S5, SI) is identical to that experimentally
observed by EXAFS measurements for a related Mn(III)
complex with phenolate and hydroxo groups in the
coordination sphere.57
Furthermore, in aerated complex solutions at neutral to
slightly basic pH, the adducts of the complex with oxygen and
its reduced species, respectively, could also be detected by CSI-
MS in addition to [(LO−)MnIII(OH−)]+ and [(LO−)MnIVO]+
(see Figures S13 and S14, SI). One of the generated species has
a peak at m/z = 539.1703, which corresponds to either
[(LO−)MnII(O2)]+ or its redox tautomer [(LO−)MnIII(O2•−)]+
(Figure S14, SI). Our DFT calculations support its assignment
as a MnIII-superoxo species [Figure 5b, Table S5 (SI)]. These
show that coordination of O2 to hexacoordinate [(LO−)MnII]+
results in a weakened O−O bond. The 1.27 Å O−O bond
distance is consistent with a metal-bound superoxide in an end-
on binding mode (1.280 Å).58 In addition, the shortening of the
Mn−phenolate bond and the significant elongation of the
bonds between the Mn and the two backbone amine nitrogens
suggest the significant MnIII character of the metal center
(Figure 5b).
Figure 4. Cyclic voltammograms of [(LO−)MnII(OH2)]+/[(LO−)-
MnIII/(OH−)]+ in aqueous solution in relation to pH. [complex] = 1
mM, [LiClO4] = 100 mM, scan rate 0.1 V/s.
superoxide. The observed rate constants exhibit a linear
dependence on the complex concentrations, and catalytic rate
constants (kcat) could be determined from the corresponding
slopes [see Figure S6 (SI) and Table 4].
Previously, a chemiluminescent assay suggested that [(LO−)-
MnII]+ was an extremely effective superoxide scavenging agent
at pH 8.12 Despite this, the observed SOD activity reported
here is in the lower range for characterized SOD mimetics,
which have kcat values that span from 106 to 109 M−1 s−1.17,52
This sort of discrepancy has precedence. The results obtained
by indirect SOD assays often disagree with those from direct
stopped-flow measurements, since side reactions interfere with
the former.17,49,53 As reported for different classes of SOD
mimetics, the nature of the buffer also impacts the catalytic
rate.17 At pH 7.4, the complex degrades O2 about twice as
•−
fast in HEPES than it does in phosphate buffer. The inhibitory
effect of phosphate has previously been observed with
pentaazamacrocyclic SOD mimetics and was believed to result
from the competitive binding of hydrogen phosphate
(HPO42−) to the manganese center.17
Reductive Activity toward Oxygen and Superoxide.
Besides SOD activity, the MnII species is also capable of
reducing oxygen. At pH >6, where the phenol group is
deprotonated, the oxidation of the manganese center to MnIII
becomes feasible in aerobic solutions. The MnIII complex can
be further stabilized by deprotonation of the bound water (see
Scheme 1), which yields the hydroxo species [(LO−)-
MnIII(OH−)]+. Above pH 6, the [(LO−)MnIII(OH−)]+ could
be detected in aerobic solutions by CSI-MS (m/z = 524.1834;
The second new feature at m/z = 540.1788 corresponds to
either the protonated superoxide (perhydroxyl radical) or
•
hydroperoxo adduct, i.e. [(LO−)MnII(HO2 )]+ or [(LO−)-
−
MnIII(HO2 )]+ (Figure S14, SI). The latter assignment is
Table 4. Catalytic Rate Constants of the Reaction of 1 and 2 with Superoxide
kcat (M‑1 s‑1)
buffer, pH
[(LO−)MnII]+/[(LO−)MnII(OH2)]+ (1)
[(HOL−LO−)MnII2(OH2)x]3+ (2)
HEPES, 8.1
HEPES, 7.4
phosphate, 7.4
6.2 0.5 × 105
3.6 0.1 × 106
1.1 0.02 × 106
1.6 0.1 × 106
5.1 0.3 × 106
2.5 0.2 × 106
G
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 5. Calculated structures (UPW91PW91/def2svp) of (a) [(LO−)MnIII(OH−)]+, (b) [(LO−)MnIII(O2•−)]+, and (c) [(LO−)MnIII(O22−)]. The
hydrogen atoms of the polydentate ligand were omitted for clarity.
more plausible, since the superoxide adduct of a MnII complex
with an analogous pentadentate ligand was proven to have
MnIII-peroxo character.59 The existence of these species
suggests that oxidation of the MnII complex to [(LO−)MnIVO]+
proceeds through oxygen binding to the six-coordinate MnII
center that triggers an inner-sphere electron transfer and
superoxide generation. This we could also confirm by DFT
calculations that demonstrate cleavage of the MnIII−superoxo
bond upon addition of a proton and release of HO2 (Figure
S8a, SI). The latter can bind to excess MnII present in solution
and undergoes an inner-sphere reduction to the corresponding
oxygen atom that is accompanied by elongation of the O−O
bond and shortening of the Mn−O bond with negatively
charged phenolate and vicinal oxygen (from the O−O moiety)
atoms, respectively (Figure S8b, SI). This at least qualitatively
explains a tendency toward O−O bond splitting and generation
of the MnIV-oxo species and hydroxyl radical upon protonation
of coordinated peroxide.
On the basis of the redox potential of 1, an outer-sphere
reduction of O2 to O2•− by the MnII form is thermodynamically
unfavorable; therefore, an inner-sphere pathway, proceeding
through the observed hexacoordinate [(LO−)MnII]+ intermedi-
ate, that binds O2 represents the only possible mechanism for
its reduction.
We experimentally confirmed the generation of reactive
oxygen species (ROS) by following of the oxidation of 2-nitro-
5-thiobenzoate (TNB) to 5,5′-dithiobis(2-nitrobenzoic acid)
(DTNB) spectrophotometrically, as well as by EPR measure-
ments (for details see the SI).
On the basis of the observed manganese and reduced oxygen
species, as well as their adducts, an overall mechanism for the
oxygen/superoxide reduction by 1 is depicted in Scheme 2 (red
hydroperoxo adduct [(LO−)MnIII(HO2 )], with subsequent
−
homolytic cleavage of the O−O bond resulting in the MnIV-oxo
species and the OH• radical (red pathway in Scheme 2). We
also probed the character of the MnII-superoxo adduct by DFT
studies. As expected (see above), binding of superoxide to
hexacoordinate [(LO−)MnII]+ results in a structure that
resembles that of a pentacoordinate end-on MnIII-peroxo
complex [Figure 5c, Table S5 (SI)] based on the Mn−N,
Mn−phenolate, O−O, and Mn−O2(unit) bond lengths.
Addition of a proton leads to protonation of the distal peroxo
H
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2. Overall Reaction Scheme, Which Links Outer Sphere, Inner Sphere, and Water Exchange Reactivity of [(LO−)MnII]+
at pH >6
Scheme 3. Oxidative Coupling of Two Molecules of [(LO−)MnII]+ Induced by [(LO−)MnIVO], OH•, or H2O2
field). Importantly, the binding of superoxide to six-coordinate
MnII and its subsequent inner-sphere reduction does not seem
to contribute to the superoxide dismutation path, since the
decay of the corresponding MnIII-hydroperoxo adduct yields a
MnIV-oxo species and OH•, rather than the MnIII required to
continue the SOD catalytic cycle (see above). Even more, we
probed coordination of superoxide to the MnIII form of 1 that
exists under our experimental conditions, i.e., pentacoordinate
[(LO−)MnIII(OH)]+, by computational methods. However,
superoxide did not show a tendency to bind to [(LO−)-
MnIII(OH−)]+ at all, suggesting that repulsions between the
negatively charged superoxide and phenolate, the steric bulk of
MnII(OH2)]+ and pentacoordinate [(LO−)MnIII(OH−)]+ (see
blue field in Scheme 2). In comparison, manganese complexes
with pentaazamacrocyclic ligands also exhibit equilibria
between seven- and six-coordinate species, yet they possess
significantly higher SOD activity, with kcat up to three orders of
magnitude bigger than 1, despite their having much more
positive redox potentials that do not favor superoxide
reduction.49−51,61 The oxidized and reduced forms of these
complexes, either six- and seven-coordinate MnII ([(L)-
MnII(OH2)] and [(L)MnII(OH2)2]) or six-coordinate MnIII
([(L)MnIII(OH2)]), possess at least one readily exchangeable
water molecule in their coordination spheres.39,61 Further, their
pentagonal bipyramidal coordination environment with the
neutral pentadentate ligand in the equatorial plane does not
hinder the access of negatively charged substrate molecules to
the metal center.39,49−51,62 The ability of the pentaazamacro-
cylic complexes to operate through an inner-sphere mechanism
60
the LO− ligand, and the lesser kinetic lability of MnIII make
[(LO−)MnIII(OH−)]+ coordinatively saturated and inert toward
substitution/addition reactions. All together, our results suggest
that the SOD catalysis displayed by 1 is predominantly
achieved through an outer-sphere mechanism that involves
coordinationally saturated species: heptacoordinate [(LO−)-
I
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 6. CSI-MS of a solution of 1 in acetonitrile mixed with mCPBA. Detection of [(LO−)MnII]+ (m/z = 507.1815) in addition to [(LO−)MnIVO]+
(m/z = 523.1766) and [(HOL−LO−)MnII ]3+ (m/z = 337.7860).
2
compensates for their less favorable redox potentials and results
in efficient catalysis.49−51
Scheme 2). In the oxidizing solutions, where generation of 2
was observed, we could also detect trace amounts of mono- and
Transformation into Binuclear Complex [(HOL−LO−)-
dihydroxylated forms {[(HOL(OH)−LO−)MnII ]3+, m/z =
2
MnII ]3+ (2) by ROS Prolongs SOD Catalysis (stability,
343.1174 and [(HOL(OH)−(HO)LO−)MnII2]3+, m/z =
348.4486; Figures 6 and S15 (SI)}, which feature one or two
further hydroxylated phenol groups of ligand, respectively. The
presence of this product is consistent with the agency of a
MnIV-oxo intermediate; such species are known to activate
aromatic C−H bonds.45 The additional OH groups render the
ligand environment around the Mn centers within the binuclear
structure even more crowded.
2
magnetic properties, and SOD activity of 2). The
competition between outer-sphere SOD catalysis and inner-
sphere reduction of oxygen and superoxide is determined by
the equilibrium between the heptacoordinate MnII complex and
the hexacoordinate MnII species that serves as the intermediate
for water exchange (see green field in Scheme 2). The reaction
with oxygen is slow, whereas the reaction with superoxide is
rapid. Nonetheless, one would expect that the OH• and MnIV-
oxo species generated by the inner-sphere pathway would
eventually result in oxidative degradation of the complex, as has
been observed in MeCN for similar MnII complexes,63 and the
concomitant loss of SOD activity.
We attempted to compare the stability of the mono- and
binuclear analogues and measure the formation constant of 2
through a competitive binding assay with tpen. The tpen ligand
had previously been found to remove the MnII from 1.11 Even
after adding four equiv of tpen to a solution of the binuclear
complex 2 in a mixture of MeCN and MeOH, there was no
evidence of its dissociation. No new resonances that could be
assigned to the dimeric HOL−LOH ligand appeared, as would be
anticipated from its displacement from the highly paramagnetic
MnII.11 The results suggest that the stability of 2, which is
obtained upon interaction between reactive oxygen species and
1, is superior to that of its mononuclear analogue 1. By way of
comparison, it was previously demonstrated that the binuclear
structures of seven-coordinate MnII SOD mimetics based on
the pentaazamacrocyclic ligands provide higher complex
stabilities than their corresponding mononuclear structures.38
In order to understand the overall mechanism behind the
catalytic superoxide decomposition initially induced by 1, we
studied the SOD activity of pure 2 under the same experimental
conditions. The obtained catalytic rate constants per
manganese center are comparable to those observed for the
mononuclear complex [see Table 4 and Figure S6 (SI)]. This
suggests that there is no significant electronic communication
between the two manganese centers within the binuclear
complex11 and that they behave independently as catalysts. A
similar phenomenon was observed for binuclear analogues of
the MnII pentaazamacrocyclic SOD mimetics.38
One of our groups previously reported that the reaction
between 1 and H2O2 results in oxidative coupling of the phenol
groups from two different equivalents of 1, yielding a stable
binuclear MnII complex 2 (see Scheme 3).11 During this
process, a transient color change from pale yellow to brown was
observed and ascribed to the temporary oxidation of the
manganese.11 Reacting 1 with the less reactive organic peroxide
mCPBA (see Figure 6, ROOH = mCPBA) instead of H2O2 in
MeCN allowed us to observe both MnIV-oxo (m/z = 523.1766)
and the binuclear species 2 ([(−OL−LOH)MnII ]3+, m/z =
2
337.7860; Figure 6) by CSI-MS. These results suggest that the
MnIV-oxo complex may be an intermediate for the formation of
the binuclear species. Since the reaction between 1 and ROOH
is not anticipated to result in generation of an organic radical
(scheme in Figure 6), the MnIV-oxo species seems to be the
oxidizing agent responsible for the oxidative coupling of the
LOH ligand. MS analysis of the reaction mixture obtained after
the catalytic decomposition of superoxide also revealed the
formation of 2. We currently speculate that both MnIV-oxo45
and OH•, each of which is capable of hydrogen atom
abstraction, contribute to the oxidative self-dimerization of
the complex through the ligand phenol groups (see red field in
J
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The previously reported crystal structure of 2 obtained from
MeCN showed that the two MnII centers retain the same
seven-coordinate geometry as the metal center in 1.11 One of
the phenols is deprotonated and the seventh coordination site
is occupied by the solvent molecule. In aqueous solution, the
water molecules coordinated to 2 might be significantly more
acidic than those in 1, as has been observed in the case of
mono- and binuclear pentaazamacrocyclic SOD mimetics.38
of the structurally related [(tpen)MnII(Cl)]+, which has the
same overall charge but lacks an inner-sphere water. The
lowering of these redox potentials also facilitates the generation
of a MnIV-oxo species, which contributes to the superoxide and
oxygen reduction ability of the complex. Similar effects would
be anticipated for the active sites of other oxido-reductase
enzymes; these are typically highly accessible to water
molecules.
This would stabilize dihydroxo [(HOL−LO−)MnII (OH−)2]+
The operation of SOD catalysis according to an IS
mechanism, cannot be provided by 1. In order for an IS
mechanism to proceed, either a vacant coordination site or a
labile ligand in the coordination sphere of both MnII and MnIII
forms is required. The MnSOD enzyme cycles through five-
coordinate MnII and MnIII trigonal bipyramidal forms that are
sterically uncongested enough to accommodate superoxide as
the sixth ligand in the equatorial plane.64 Although the MnIII
center in [(LO−)MnIII(OH−)]+ is also pentacoordinate with
three neutral nitrogen and two negatively charged oxygen
donor atoms as in the MnSOD enzyme, its distorted square
pyramidal structure and the diamine backbone of the ligand
block the access of potential sixth ligands. The axial position is
therefore inaccessible and is not capable of binding superoxide.
In addition, the strong coordination of phenolate hinders
entrance of a third negatively charged ligand into the
coordination sphere of [(LO−)MnIII(OH−)]+. In the case of
efficient seven-coordinate pentaazamacrocyclic SOD mimetics,
the equatorially positioned pentadentate ligands do not exert
any steric hindrance for the entry of additional substitu-
ents;39,49−51,62 consequently, both their oxidized and reduced
forms possess at least one labile aqua ligand, enabling IS
mechanisms. The pentadentate ligands for these complexes are
neutral; this results in more positive overall complex charges
and much higher redox potentials. The higher potentials
preclude efficient OS SOD catalysis. Nonetheless, these
compounds still exhibit remarkable SOD activity (higher than
that of 1) with catalytic rate constants comparable with those of
the enzymes (kcat ∼ 109 M−1 s−1 for MnSOD,65 1.60 × 109 M−1
s−1 for M4040152). The results here demonstrate how the
operation of an IS mechanism can assist SOD catalysis.
Different from the MnIII form of 1, its MnII hexacoordinate
form is capable of reducing oxygen and superoxide by the IS
mechanism. Importantly, reduction of oxygen to superoxide,
which is formally the reverse of what proceeds within the SOD
cycle, would not be feasible without coordination of oxygen.
Upon binding of oxygen to MnII, the redox potential of the O2/
2
species over the diaqua species. This difference is significant,
since hydroxide anions should be more difficult to displace
from the metal ions than water molecules. Attempts to
determine the pKa values, stability constants, and rate constants
for water exchange of 2 failed due to the low water solubility of
the compound.
To further assess the electronic independence of the MnII
centers in 2, we performed variable temperature magnetic
susceptibility measurements. The magnetic data for 2 are
consistent with the notion that the MnII centers are essentially
separate entities, arising from the lack of an obvious orbital
pathway for communication of spin information between the
ions (for details, see the SI).
On the basis of the collective obtained results, a proposed
mechanism that accounts for the observed interactions of 1
with superoxide and oxygen is summarized in Scheme 2. Once
the binuclear complex 2 is formed, it probably also contributes
to the SOD catalytic cycle by an outer-sphere mechanism
(Scheme 2). This would be in agreement with the fact that the
electronic nature of the manganese centers is the same in 1 and
2 and should therefore result in similar outer-sphere catalytic
rates. A preference for the outer-sphere SOD mechanism
results from the fact that (i) the MnIII forms both in 1 and 2 are
inert toward superoxide binding (see above), (ii) the binding of
negatively charged superoxide to two MnII centers in the
binuclear structure might not be sterically/electrostatically
favored, (iii) the binuclear MnII form might be more acidic than
the mononuclear one and would consequently exist in a form,
[(HOL−LO−)MnII (OH−)2]+, with a less labile exogenous ligand
2
that would hinder the IS mechanism, and (iv) the binding of
superoxide to six-coordinate mononuclear MnII leads to
reactions that generate MnIV-oxo and OH•, inducing the
formation of the binuclear complex that continues to
superoxide dismutation via an OS mechanism (Scheme 2).
CONCLUSION
■
−
The LOH complex with MnII is capable of both outer-sphere
and inner-sphere reactions with superoxide, with the coordina-
tion of water determining which pathway is followed. The
heptacoordinate complex [(LO−)MnII(OH2)] engages in super-
oxide dismutation, whereas the hexacoordinate species [(LO−)-
MnII]+ only reduces superoxide. Enzymatic systems and some
efficient SOD mimetics rely on the operation of both inner-
sphere and PCET processes for the rapid catalysis of superoxide
degradation,49−51 whereas the latter can also assist in operation
of an OS electron transfer mechanism. Although OS PCET is
provided by 1, the IS mechanism leads to superoxide reduction
side reactivity. In [(LO−)MnII(OH2)], the bound water
molecule supports PCET by lowering the MnIII/MnII and
MnIV/MnIII redox potentials enough to allow efficient OS SOD
catalysis. The importance of water coordination is nicely
illustrated by the fact that [(LO−)MnII(OH2)]+ can be oxidized
to MnIII and MnIV at potentials that are at least 0.6 and 1 V,
respectively, less positive than those required for the oxidation
O2 couple shifts to values higher than that for the
corresponding MnIII/MnII couple, thus enabling IS reduction
of O2.49,50
One substantial difference between the aforementioned
pentaazamacrocyclic SOD mimetics and 1 and its binuclear
form 2 is that the latter two complexes have more sterically
crowded metal centers, with negatively charged phenolate
groups close to the only accessible coordination site for an
entering substrate. Thus, superoxide can only bind to MnII
through the six-coordinate water-free intermediate. The
obtained MnIII-hydroperoxo intermediate undergoes homolytic
O−O cleavage instead of being protonated to release H2O2. A
byproduct of the O−O cleavage is a MnIV-oxo species that
reacts with another equivalent of Mn complex to form 2. The
low redox potential for the MnIV/MnIII couple favors such
reactivity. Outside of Mn−corrole systems,66 this is the only
known example of the involvement of a MnIV-oxo species in
SOD catalysis, albeit through the separate superoxide/oxygen
K
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
reduction path. Another key difference between the pentaaza-
macrocyclic SOD mimetics and 1 is the coordination of the
MnIII members of the main catalytic cycle. In the former, the
MnIII forms are proposed to be six-coordinate with neutral
pentadentate and aqua ligands;52 conversely, the MnIII form of
1, [(LO−)MnIII(OH−)]+, is pentacoordinate, with the bulky and
negatively charged LO− and a OH− ligand that is substitution-
ally inert relative to the aforementioned aqua ligands.
AUTHOR INFORMATION
Corresponding Authors
*crg0005@auburn.edu
*ivana.ivanovic@chemie.uni-erlangen.de
■
ORCID
Jens Langer: 0000-0002-2895-4979
Christian R. Goldsmith: 0000-0001-7293-1267
Ivana Ivanovic-Burmazovic: 0000-0002-1651-3359
́
́
The high redox potentials of the pentaazamacrocyclic SOD
mimetics inhibit the formation of high-valent MnIV or MnV
species, representing a means to defend the active site against
inadvertent oxidative damage. Study of the solution properties
and redox reactivity of 1, conversely, reveals that the compound
retains its SOD activity via its unique ability to self-scavenge the
OH• and MnIV-oxo species that are generated through inner-
sphere reactions between the complex and oxygen, superoxide,
and/or hydrogen peroxide. These side products convert the
complex into what appears to be a more stable binuclear form,
2, that maintains both the same coordination geometry around
the MnII centers and the same catalytic activity. The SOD
activity of 2 appears to operate through fundamentally the same
OS mechanism as 1. The channeling of these oxidants toward a
reaction that produces a catalytic species that is less susceptible
to side reactivity represents an alternative mechanistic strategy
to both defend a catalyst against oxidative damage and prolong
catalysis. This strategy could potentially be applied to prolong
the catalysis of other oxidation reactions, such as non-heme
iron-promoted C−H activation. Although the OS mechanism
for SOD activity of 1 resulted in slower rates of reaction (with
the magnitude of 106 M−1 s−1) than the IS mechanism in the
case of pentaazamacrocyclic SOD mimetics (with the
magnitude of 107 s−1),67 the results here show that the heavy
use of this pathway limits the formation of deleterious OH• and
MnIV-oxo species.
To summarize, we have determined important factors that
regulate SOD catalysis through both inner- and outer-sphere
mechanisms. The steric bulk and charge distribution of the
polydentate ligand regulate the operation of inner- or outer-
sphere mechanisms. Metal-bound water molecules both
facilitate the PCET necessary for OS pathways and determine
the feasibility of competing OS and IS mechanisms through
coordination equilibria. The importance of the water ligand
contrasts sharply with the traditional view of active-site water
molecules merely serving as place-holders for other reagents in
bioinorganic chemistry.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
I.K., A.F., M.D., A.Z., M.R.F, and I.I.-B. gratefully acknowledge
the support of an intramural grant from the University of
Erlangen-Nuremberg (Emerging Initiative: Medicinal Redox
Inorganic Chemistry). I.K. thanks Dr. Anne Dees for her
support with the interpretation of the potentiometric titration
data and Stephan Pflock for the graphical presentation of the
crystal structures. R.P. thanks Prof. Tim Clark for hosting this
work at the CCC and Regionales Rechenzentrum Erlangen
(RRZE) and for a generous allotment of computer time. M.P.S.
and S.F. thank the NSF (CHE-1058889 and CHE-1363274)
for support and Mr. Tarik Ozumerzifon for experimental
assistance.
REFERENCES
■
(1) Karlin, K. D.; Lippard, S. J.; Valentine, J. S.; Burrows, C. J. Acc.
Chem. Res. 2015, 48, 2659.
(2) Groni, S.; Hureau, C.; Guillot, R.; Blondin, G.; Blain, G.;
Anxolabehere-Mallart, E. Inorg. Chem. 2008, 47, 11783.
́ ̀
(3) Jackson, T. A.; Gutman, C. T.; Maliekal, J.; Miller, A.-F.; Brunold,
T. C. Inorg. Chem. 2013, 52, 3356.
(4) Miller, A.-F. Acc. Chem. Res. 2008, 41, 501.
(5) Grove, L. E.; Xie, J.; Yikilmaz, E.; Miller, A.-F.; Brunold, T. C.
Inorg. Chem. 2008, 47, 3978.
(6) Jackson, T. A.; Brunold, T. C. Acc. Chem. Res. 2004, 37, 461.
(7) Riley, D. P. Chem. Rev. 1999, 99, 2573.
(8) Riley, D. P.; Lennon, P. J.; Neumann, W. L.; Weiss, R. H. J. Am.
Chem. Soc. 1997, 119, 6522.
(9) Hureau, C.; Blanchard, S.; Nierlich, M.; Blain, G.; Rivier
Girerd, J.-J.; Anxolabehere-Mallart, E.; Blondin, G. Inorg. Chem. 2004,
43, 4415.
̀
e, E.;
́
̀
(10) Stallings, W. C.; Pattridge, K. A.; Strong, R. K.; Ludwig, M. L. J.
Biol. Chem. 1984, 259, 10695.
(11) Yu, M.; Beyers, R. J.; Gorden, J. D.; Cross, J. N.; Goldsmith, C.
R. Inorg. Chem. 2012, 51, 9153.
(12) Yu, M.; Ambrose, S. L.; Whaley, Z. L.; Fan, S.; Gorden, J. D.;
Beyers, R. J.; Schwartz, D. D.; Goldsmith, C. R. J. Am. Chem. Soc. 2014,
136, 12836.
(13) Sheng, Y.; Stich, T. A.; Barnese, K.; Gralla, E. B.; Cascio, D.;
Britt, R. D.; Cabelli, D. E.; Valentine, J. S. J. Am. Chem. Soc. 2011, 133,
20878.
(14) Sheng, Y.; Butler Gralla, E.; Schumacher, M.; Cascio, D.;
Cabelli, D. E.; Valentine, J. S. Proc. Natl. Acad. Sci. U. S. A. 2012, 109,
14314.
(15) Sheng, Y.; Abreu, I. A.; Cabelli, D. E.; Maroney, M. J.; Miller, A.-
F.; Teixeira, M.; Valentine, J. S. Chem. Rev. 2014, 114, 3854.
(16) Salvemini, D.; Wang, Z.-Q.; Zweier, J. L.; Samouilov, A.;
Macarthur, H.; Misko, T. P.; Currie, M. G.; Cuzzocrea, S.; Sikorski, J.
A.; Riley, D. P. Science 1999, 286, 304.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b08394.
Crystallographic data for [(LO−)MnII](ClO4)·1.5H2O in
CIF format (CIF)
Crystallographic data for [(LOH)MnII(OAc)](ClO4)·
nH2O in CIF format (CIF)
Additional experimental procedures; detailed description
of data treatment of water exchange measurements;
details concerning the speciation of [(LOH)MnII(solv)]-
(ClO4)2 (1) below pH 6; and additional data concerning
X-ray crystallography, electrochemistry, SOD activity,
oxygen reduction, quantum chemical methods (including
Cartesian coordinates), ROS detection by EPR and
TNB, CSI-MS, and magnetic measurements (PDF)
(17) Friedel, F. C.; Lieb, D.; Ivanovic-Burmazovic, I. J. Inorg. Biochem.
2012, 109, 26.
(18) Zuberbuhler, A. D.; Kaden, T. A. Talanta 1982, 29, 201.
̈
(19) Zahl, A.; Neubrand, A.; Aygen, S.; van Eldik, R. Rev. Sci. Instrum.
1994, 65, 882.
L
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(20) CrysAlisPro, version 1.171.38.43; Rigaku Oxford Diffraction,
Yarnton, England, 2016.
(21) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339.
(22) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Adv. 2015, 71,
3.
(48) Liu, G.-F.; Durr, K.; Puchta, R.; Heinemann, F. W.; van Eldik,
̈
R.; Ivanovic-
(49) Ivanovic-
64, 53.
(50) Ivanovic-
́
Burmazovic,
́
́
I. Dalton Trans. 2009, 6292.
Burmazovic,
́
I.; Filipovic, M. R. Adv. Inorg. Chem. 2012,
́
Burmazovic,
́
I. Adv. Inorg. Chem. 2008, 60, 59.
(51) Ivanovic-Burmazovic, I.; van Eldik, R. Dalton Trans. 2008, 5259.
(52) Aston, K.; Rath, N.; Naik, A.; Slomczynska, U.; Schall, O. F.;
Riley, D. P. Inorg. Chem. 2001, 40, 1779.
(23) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
2008, 64, 112.
(53) Liu, G.-F.; Filipovic,
́ ́
M.; Heinemann, F. W.; Ivanovic-
(24) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.;
Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B: Condens. Matter
Mater. Phys. 1992, 46, 6671.
(25) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.;
Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. Rev. B: Condens. Matter
Mater. Phys. 1993, 48, 4978.
(26) Perdew, J. P.; Burke, K.; Wang, Y. Phys. Rev. B: Condens. Matter
Mater. Phys. 1996, 54, 16533.
(27) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter
Mater. Phys. 1988, 37, 785.
(28) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J.
Burmazovic, I. Inorg. Chem. 2007, 46, 8825.
́
(54) Goldsmith, C. R.; Cole, A. P.; Stack, T. D. P. J. Am. Chem. Soc.
2005, 127, 9904.
(55) Hubin, T. J.; McCormick, J. M.; Alcock, N. W.; Busch, D. H.
Inorg. Chem. 2001, 40, 435.
(56) Ivanovic-
315.
(57) Lassalle-Kaiser, B.; Hureau, C.; Pantazis, D. A.; Pushkar, Y.;
Guillot, R.; Yachandra, V. K.; Yano, J.; Neese, F.; Anxolabehere-
Mallart, E. Energy Environ. Sci. 2010, 3, 924.
́ ́ ́
Burmazovic, I.; Jelkovic, K. Adv. Inorg. Chem. 2004, 55,
́
̀
(58) Wurtele, C.; Gaoutchenova, E.; Harms, K.; Holthausen, M. C.;
̈
Phys. Chem. 1994, 98, 11623.
Sundermeyer, J.; Schindler, S. Angew. Chem., Int. Ed. 2006, 45, 3867.
(29) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(30) Puchta, R.; Dahlenburg, L.; Clark, T. Chem. - Eur. J. 2008, 14,
8898.
(59) Groni, S.; Blain, G.; Guillot, R.; Policar, C.; Anxolabeh
Mallart, E. Inorg. Chem. 2007, 46, 1951.
(60) Lieb, D.; Zahl, A.; Shubina, T. E.; Ivanovic-Burmazovic, I. J. Am.
́ ́
Chem. Soc. 2010, 132, 7282.
́ ̀
ere-
(31) Selcu̧ ki, C.; van Eldik, R.; Clark, T. Inorg. Chem. 2004, 43, 2828.
(32) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297.
(33) Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 1057.
(61) Riley, D. P.; Schall, O. F. Adv. Inorg. Chem. 2006, 59, 233.
(62) Ivanovic-Burmazovic, I.; Hamza, M. S. A.; van Eldik, R. Inorg.
Chem. 2006, 45, 1575.
(63) Groni, S.; Dorlet, P.; Blain, G.; Bourcier, S.; Guillot, R.;
Anxolabehere-Mallart, E. Inorg. Chem. 2008, 47, 3166.
̈
́
́
(34) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R.
Chem. Phys. Lett. 1995, 240, 283.
̈
(35) Eichkorn, K.; Weigend, F.; Treutler, O.; Ahlrichs, R. Theor.
Chem. Acc. 1997, 97, 119.
́
̀
(64) Lah, M. S.; Dixon, M. M.; Pattridge, K. A.; Stallings, W. C.; Fee,
(36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.;
Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.;
Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.;
Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.;
Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A.
J. A.; Ludwig, M. L. Biochemistry 1995, 34, 1646.
(65) Miriyala, S.; Spasojevic, I.; Tovmasyan, A.; Salvemini, D.;
Vujaskovic, Z.; St. Clair, D.; Batinic-Haberle, I. Biochim. Biophys. Acta,
Mol. Basis Dis. 2012, 1822, 794.
(66) Haber, A.; Gross, Z. Chem. Commun. 2015, 51, 5812.
(67) Weiss, R. H.; Riley, D. P. In Uses of Inorganic Chemistry in
Medicine; Farrell, N., Ed.; Royal Society of Chemistry: Cambridge, UK,
1999; Chapter 7.
̈
D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J.
Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009.
(37) Gale, E. M.; Mukherjee, S.; Liu, C.; Loving, G. S.; Caravan, P.
Inorg. Chem. 2014, 53, 10748.
(38) Lieb, D.; Friedel, F. C.; Yawer, M.; Zahl, A.; Khusniyarov, M.
M.; Heinemann, F. W.; Ivanovic-
222.
(39) Dees, A.; Zahl, A.; Puchta, R.; van Eikema Hommes, N. J. R.;
Heinemann, F. W.; Ivanovic-Burmazovic, I. Inorg. Chem. 2007, 46,
2459.
(40) Drahos,
Chem. 2010, 49, 3224.
́ ́
Burmazovic, I. Inorg. Chem. 2013, 52,
́
́
̌
̌ ́
B.; Kotek, J.; Hermann, P.; Lukes, I.; Toth, E. Inorg.
(41) Maigut, J.; Meier, R.; Zahl, A.; van Eldik, R. J. Am. Chem. Soc.
2008, 130, 14556.
(42) Deeth, R. J. Inorg. Chem. 2008, 47, 6711.
(43) Trifunovic,
H.; Yamasaki, M. Transition Met. Chem. 2000, 25, 680.
́ ́ ́
S. R.; Matovic, Z. D.; Milovanovic, V.; Kawaguchi,
(44) Wang, X. F.; Gao, J.; Wang, J.; Zhang, Z. H.; Wang, Y. F.; Chen,
L. J.; Sun, W.; Zhang, X. D. J. Struct. Chem. 2008, 49, 724.
(45) Wu, X.; Seo, M. S.; Davis, K. M.; Lee, Y.-M.; Chen, J.; Cho, K.-
B.; Pushkar, Y. N.; Nam, W. J. Am. Chem. Soc. 2011, 133, 20088.
(46) Parsell, T. H.; Behan, R. K.; Green, M. T.; Hendrich, M. P.;
Borovik, A. S. J. Am. Chem. Soc. 2006, 128, 8728.
(47) Chen, J.; Yoon, H.; Lee, Y.-M.; Seo, M. S.; Sarangi, R.;
Fukuzumi, S.; Nam, W. Chem. Sci. 2015, 6, 3624.
M
DOI: 10.1021/jacs.6b08394
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX