Combining Structural with Functional Model Properties in Iron Synthetic Analogue Complexes for the Active Site in Rabbit Lipoxygenase

Article abstract of DOI:10.1021/jacs.1c04422

Iron complexes that model the structural and functional properties of the active iron site in rabbit lipoxygenase are described. The ligand sphere of the mononuclear pseudo-octahedral cis-(carboxylato)(hydroxo)iron(III) complex, which is completed by a tetraazamacrocyclic ligand, reproduces the first coordination shell of the active site in the enzyme. In addition, two corresponding iron(II) complexes are presented that differ in the coordination of a water molecule. In their structural and electronic properties, both the (hydroxo)iron(III) and the (aqua)iron(II) complex reflect well the only two essential states found in the enzymatic mechanism of peroxidation of polyunsaturated fatty acids. Furthermore, the ferric complex is shown to undergo hydrogen atom abstraction reactions with O-H and C-H bonds of suitable substrates, and the bond dissociation free energy of the coordinated water ligand of the ferrous complex is determined to be 72.4 kcal·mol-1. Theoretical investigations of the reactivity support a concerted proton-coupled electron transfer mechanism in close analogy to the initial step in the enzymatic mechanism. The propensity of the (hydroxo)iron(III) complex to undergo H atom abstraction reactions is the basis for its catalytic function in the aerobic peroxidation of 2,4,6-tri(tert-butyl)phenol and its role as a radical initiator in the reaction of dihydroanthracene with oxygen.

Full text of DOI:10.1021/jacs.1c04422

pubs.acs.org/JACS
Article
Combining Structural with Functional Model Properties in Iron
Synthetic Analogue Complexes for the Active Site in Rabbit
Lipoxygenase
Emiel Dobbelaar, Christian Rauber, Thorsten Bonck, Harald Kelm, Markus Schmitz,
Matina Eloïse de Waal Malefijt, Johannes E. M. N. Klein, and Hans-Jörg Kruger*
̈
Cite This: J. Am. Chem. Soc. 2021, 143, 13145−13155
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Iron complexes that model the structural and functional properties of
the active iron site in rabbit lipoxygenase are described. The ligand sphere of the
mononuclear pseudo-octahedral cis-(carboxylato)(hydroxo)iron(III) complex, which
is completed by a tetraazamacrocyclic ligand, reproduces the first coordination shell
of the active site in the enzyme. In addition, two corresponding iron(II) complexes
are presented that differ in the coordination of a water molecule. In their structural
and electronic properties, both the (hydroxo)iron(III) and the (aqua)iron(II)
complex reflect well the only two essential states found in the enzymatic mechanism
of peroxidation of polyunsaturated fatty acids. Furthermore, the ferric complex is
shown to undergo hydrogen atom abstraction reactions with O−H and C−H bonds of suitable substrates, and the bond dissociation
free energy of the coordinated water ligand of the ferrous complex is determined to be 72.4 kcal·mol−1. Theoretical investigations of
the reactivity support a concerted proton-coupled electron transfer mechanism in close analogy to the initial step in the enzymatic
mechanism. The propensity of the (hydroxo)iron(III) complex to undergo H atom abstraction reactions is the basis for its catalytic
function in the aerobic peroxidation of 2,4,6-tri(tert-butyl)phenol and its role as a radical initiator in the reaction of
dihydroanthracene with oxygen.
reaction cycle, the ferric hydroxide converts the pentadiene
unit into a pentadienyl radical via an overall H atom
abstraction (Scheme 1b).
The current state of research describes the elementary
reaction step of C−H bond peroxidation as a concerted
proton-coupled electron transfer (PCET).2,14−16 After rear-
rangement, the pentadienyl radical formed reacts with triplet
oxygen, yielding a peroxyl radical intermediate, which then
abstracts a hydrogen atom from the ferrous aqua unit under
regeneration of the original ferric hydroxide state.
The objective of the synthetic analogue approach (struc-
tural−functional modeling) in bioinorganic studies entails the
synthesis of small inorganic model complexes that provide a
rational basis to understanding the structural, electronic, or
mechanistic properties of the metal sites in metalloproteins and
help to identify those intrinsic properties of the enzyme
associated with the metal site and its first coordination sphere
and those related to the protein matrix.17−19 In addition to the
INTRODUCTION
■
Lipoxygenases (LOXs) are enzymes found in plants and
animals that contain a mononuclear cis-(carboxylato)-
(hydroxo)iron(III) coordination unit as the catalytically active
site for the hydroperoxidation of (Z,Z)-1,4-pentadiene
moieties with molecular oxygen.1−3 These enzymes play a
major role, for example, in regulating the metabolism of
polyunsaturated fatty acids, such as linoleate and arachidonate,
in controlling cellular differentiation and growth, and in
initiating inflammatory processes in mammals.4,5 On the basis
of the structural parameters of the enzymes in the reduced cis-
(carboxylato)(aqua)iron(II) state, the distance found between
the oxygen atoms deriving from the water ligand and the
carbonyl group of the carboxylate ligand (deriving from the C-
terminal amino acid isoleucine) indicates the existence of a
hydrogen bonding interaction between these ligands.6 While
the residual distorted octahedral coordination sphere of the
active nonheme iron(III) unit is completed by four histidine
nitrogen donors in rabbit and human 15-lipoxygenases
(rLOXs7,8 and hLOXs9), the composition of the protein-
derived ligand environment in lipoxygenases from other
sources varies by substitution of one of the histidine residues,
e.g., by an asparagine carbonyl oxygen donor.2,6,10−12 The
most widely accepted mechanism involves two states for the
iron site: a high-spin ferric hydroxide and a high-spin ferrous
aqua complex (Scheme 1a).2,13 In the initial step of the
Received: April 28, 2021
Published: August 12, 2021
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155
© 2021 American Chemical Society
13145

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Scheme 1. (a) Iron Sites of the Enzymatic States Involved in
the Catalytic Cycle; (b) Generally Accepted Radical
Mechanism for the Peroxidation of the 1,4-Pentadiene Unit
in Polyunsaturated Fatty Acids by rLOX
RESULTS AND DISCUSSION
■
Synthesis and Structural Characterization. The mono-
nuclear (benzoato)(hydroxo)iron(III) complex [Fe(L-
t
N4 Bu2)(O2CPh)(OH)]+ (1) was obtained by the reaction of
t
the bis(methanolato)iron(III) complex [Fe(L-N4 Bu2)-
(OMe)2]+ (4) with one equivalent of benzoic acid in an
acetonitrile solution containing only small amounts of water.
Complex 4, in turn, is conveniently prepared from the
45
t
dichloroiron(II) complex [Fe(L-N4 Bu2)Cl2] (5) in meth-
anol under aerobic conditions upon base addition. Complex 5
also serves as a starting material in the synthesis of the
t
respective (benzoato)iron(II) complexes [Fe(L-N4 Bu2)-
(O2CPh)(OH2)]+ (2) and [Fe(L-N4 Bu2)(O2CPh)]+ (3). In
t
alcoholic solutions, the coordinated water molecule in 2 can be
replaced by a wide variety of alcohol molecules (e.g., by
t
MeOH in [Fe(L-N4 Bu2)(O2CPh)(MeOH)]+ (6)); in con-
trast, using anhydrous acetonitrile, complex 3 can be obtained
where the benzoate unit acts as a bidentate chelating ligand.
Under air, complex 2 forms complex 1, but this reaction is not
fully quantitative and is, therefore, an impractical method of
obtaining complex 1 as an analytically pure material. Complex
cations 1, 2, and 4 were isolated with a variety of counterions.
The synthetic efforts, the structures of 1, 2, and 3, and a list of
individual analytically pure substances obtained in this study
are contained in Scheme 3; synthetic details are described in
the Supporting Information.
gains in knowledge of bioinorganic issues, the results of such a
study could reward researchers interested in developing novel
catalysts with potential information about those features, which
are essential for a reactivity similar to the enzyme’s.
Both functional13,20−22 and structural23,24 model complexes
for lipoxygenases have previously been reported; however,
none of these combine both structural and functional
properties. Probably, the most challenging part of generating
a structural model for the active site in rabbit lipoxygenase is
the preparation of a mononuclear (hydroxo)iron(III) complex
that maintains its mononuclearity in solution and does not
undergo the usually observed formation of μ-oxo-bridged
oligonuclear iron(III) species.25,26 Until recently, reports of
mononuclear (hydroxo)iron(III) complexes that had been
structurally characterized were rare. However, an increasing
number of new examples have been published over the past
few years, providing evidence of a developing interest in such
compounds.23,24,27−44 Strategies to achieve mononuclearity
include (i) sterically shielding the FeIIIOH site, (ii) increasing
the electron donating capabilities of the ancillary ligand by
incorporating negatively charged donor groups, (iii) increasing
the coordination number of the iron(III) ion, and (iv)
furnishing the ancillary ligand with hydrogen bond donors
for the interaction with the electron lone pairs of the
coordinated hydroxide ligand. On the basis of the results of
a preliminary study on iron complexes with sterically less
demanding diazapyridinophane derivatives, the tetradentate
macrocyclic N,N′-di-tert-butyl-2,11-diaza[3.3](2,6)-
Scheme 3. (a) Synthetic Reactions and (b) Structures of 1,
a
2, and 3
t
pyridinophane ligand (L-N4 Bu2, Scheme 2) is employed in
t
Scheme 2. Tetradentate Macrocyclic Ligand L-N4 Bu2
a
The cationic complexes were crystallized and isolated as [1]BPh4
(1a), [1]BPh4·MeCN (1b), [1]PF6 (1c), [2]ClO4 (2a), [2]PF6
(2b(1), P21/n, and 2b(2), C2/c), [2]CF3SO3 (2c), [3]ClO4·MeCN
(3a), [4]BPh4 (4a), and [4]PF6 (4b).
All novel compounds (1−4, 6) were structurally charac-
terized by X-ray structure analyses. Representative perspective
views of the complex cations 1, 2, and 3 are given in Figure 1,
including the applied numbering of those atoms relevant to the
subsequent discussion. Since complex cations 1 and 2 could be
structurally investigated with different counterions, the
measured lengths for equivalent bonds from these different
structure analyses span a range. Especially, due to different
hydrogen-bonding patterns of the coordinated water molecule
this investigation. Upon coordination, the large tert-butyl
substituents are located above and below the two remaining
cis-coordination sites at the metal ion with a distorted
octahedral environment, thereby preventing the formation of
μ-oxo bridged dinuclear metal complexes by steric shielding.
13146
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 1. Perspective views of the complex cations in the ferric compound 1a, and both ferrous compounds 2a and 3a with thermal ellipsoids
displaying a probability level of 50%. Hydrogen atoms are omitted for clarity with the exception of those bound to O(3).
in the individual crystal packings, the spread of bond lengths in
2 is somewhat larger than that in 1. Thus, in Table 1, the
an overall hydrogen abstraction reaction from the O−H and
C−H bonds of suitable substrates is enabled while maintaining
the mononuclearity of the ferric complex (Figure S27). In
addition, the other hydrogen atom of the coordinated water
ligand in 2 can engage in various intermolecular hydrogen
bonding interactions with either the counterion or a
neighboring complex cation as established by several crystal
structure analyses of complexes with different counterions
(Table S8).
Table 1. Selected Bond Lengths and Interatomic Distances
a
(in Å) in the Ferric Complex Cation 1 and the Ferrous
b
c
Complex Cations 2 and 3 and Differences (Δd) in the
d
Lengths of the Respective Bonds upon Oxidation of 2 to 1
Δ(d(1) −
3
2
1
d(2))
As usually observed in pseudo-octahedral complexes with
diazapyridinophane ligands, the axial Fe−Namine bonds are
considerably longer than the Fe−Npy bonds in the equatorial
coordination plane of the metal ion.45,46 The Fe−N bond
lengths to the macrocyclic ligand can be used to
unambiguously identify the spin state of the iron ion. The
relatively long Fe−N bonds in complexes 1 and 2 are
consistent with other high-spin iron complexes containing this
macrocyclic ligand. The comparison of the bond lengths in
complexes 1 and 2 reveals that all Fe−N/O bonds decrease in
length upon oxidation except for the Fe−Npy bond trans to the
hydroxide/water ligand. The reverse trend for the latter bond
is related to the stronger trans influence of the anionic
hydroxide ligand. Despite this detail, the smallest change in
bond length is observed for the Fe−Npy bonds; the already
quite elongated axial Fe−Namine bonds react to the change in
the oxidation state of the metal ion with a somewhat larger
alteration of the bond length, probably due to the fact that
these already weak axial bonds are characterized by rather soft
vibration modes.
The Fe−O bonds to the carboxylate and especially to the
water/hydroxide ligand experience the largest change in length
with 0.06−0.10 and 0.23−0.29 Å, respectively (Table 1). The
pronounced shortening of the latter Fe−O bond is caused by
the emergence of a charge on the ligand. The Fe−O bond
lengths of 1.83−1.85 and 2.08−2.12 Å agree well with those
reported for other octahedral high-spin iron(III) hydrox-
ide23,36,47 and iron(II) aqua37,48,49 complexes, validating the
assignments as a hydroxide and a water ligand in 1 and 2,
respectively. The variations in bond length and charge arising
at the two oxygen donor ligands affect the hydrogen bonding
interaction considerably; the significantly shorter interatomic
distance O(1)···O(3) as well as the larger O(1)−H(3A)−
O(3) angle indicate the presence of a substantially stronger
hydrogen bonding interaction in 2 compared to 1. Both
aspects, the rather long axial Fe−Namine bonds as well as the
Fe−O(2)
Fe−O(3)
2.08
2.02−2.05
2.08−2.12
2.35−2.39
1.95−1.96
1.83−1.85
2.29−2.34
−(0.06−0.10)
−(0.23−0.29)
−(0.01−0.10)
Fe−N(1) or
2.31,
2.40
2.14
2.11
2.24
Fe−N(3)
Fe−N(2)
Fe−N(4)
Fe···O(1)
O(1)···O(3)
2.08−2.10
2.10−2.12
3.41−3.42
2.58−2.63
2.10−2.11
2.08−2.10
3.30−3.36
+(0−0.03)
−(0−0.04)
−(0.05−0.12)
+(0.38−0.16)
2.79−2.96
a
b
From the structural analyses of 1a, 1b, and 1c. From the structural
c
analyses of 2a, 2b(1), 2b(2), and 2c. From the structural analysis of
d
3a. Despite the higher experimental accuracy, all bond lengths are
rounded to a hundredth of an Å for clarity. N(1) and N(3) are the
axial amine donors; N(2) is the pyridine nitrogen atom trans to the
hydroxide/aqua oxygen donor O(3); N(4) is the pyridine nitrogen
atom trans to the carboxylate oxygen donor O(2). O(1) refers to the
carbonyl oxygen atom of the carboxylate ligand. For more details, see
the Supporting Information.
ranges of some selected bond lengths or interatomic distances
and their changes between the ferric complex 1 and the ferrous
complex 2 are provided. Upon coordination, the tetradentate
macrocyclic ligand is folded along the Namine−Namine axis,
leaving two cis coordination sites trans to the pyridine nitrogen
donor atoms to be occupied by the two oxygen donor atoms
originating from monodentate benzoate and hydroxide or
water ligands in 1 or 2 and from the chelating benzoate ligand
in 3, respectively. The coordinated tetraazamacrocyclic ligand
is slightly twisted, resulting in an idealized C2-symmetry of the
t
[Fe(L-N4 Bu2)] fragment. In both complexes 1 and 2, the most
noteworthy structural feature is the almost planar hexagon
formed by the iron ion, the coordinated carboxylate group, the
hydroxide/water ligand, and the hydrogen bonding interaction
between carboxylate and hydroxide/water moieties (Tables S5
and S10). As a consequence of the prominent hydrogen
bonding bridge, the lone pairs of the coordinated hydroxo
ligand in 1 are oriented in such a way that a reaction path for
13147
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
increase in the strength of the hydrogen bonding interaction,
contribute to the driving force of 1 to act as a reagent for H
atom abstraction reactions. In 3, the benzoate ligand is
coordinated in a bidentate chelating fashion to a high-spin
iron(II) ion. The difference of 0.16 Å in length between both
Fe−O bonds tends to indicate a more localized charge
character within the carboxylate functional group.
of the hydroxide ligand in hydrogen bonding interactions.
Similar O−H vibrational energies have been observed for other
iron(III) hydroxide complexes.20,23 For the iron(II) aqua
complex, two distinct O−H stretching vibrations are expected.
However, only a single O−H stretching vibration between
3575 and 3338 cm−1 (2a−c) is observed (Figure S7).
Deuterium substitution experiments of 2a proved very helpful
in identifying the weaker and broader second vibration (Figure
S8). Thus, two features at 2490 and 2192 cm−1 are detected
for the deuterated sample 2d as compared to one feature at
3345 cm−1 in 2a. Using the same ratio for the changes of
vibrational energies due to the H/D substitution effect, the
second vibration is calculated to occur at 2944 cm−1 for the H-
substituted complex 2a and is, therefore, obscured by the C−H
vibrations detected in this energy range. The vibrations at 3345
and 2944 cm−1 are assigned to the individual O−H bonds
involved in intermolecular and intramolecular hydrogen
bonding interaction, respectively.
Characterization of the Complexes in Solution. The
spin state and coordination environment of the iron(III)
complex 1 are preserved in solution. Thus, the visible range of
the electronic absorption spectrum is devoid of any features
with intensities usually displayed by spin-allowed d−d
transitions, and the EPR spectrum of a frozen solution of
complex 1 in dimethylformamide (DMF) containing 0.2 mol·
L−1 tetrabutylammonium perchlorate (TBAP) displays a
rhombic signal that arises from an S = 5/2 species with real
g-values of 1.999, 1.997, and 1.995, a zero-field splitting
constant D of +2.75 cm−1, and a rhombicity E/D of 0.107
(Figures 2, S10, and S11) as defined by the spin Hamiltonian
described by eq 1 (with μB = Bohr magneton).
Magnetic and Spectroscopic Characterizations of the
Complexes in the Solid State. The magnetochemical, UV−
̈
vis−NIR, and Moßbauer spectroscopic characterizations of
solid 1 and 2 support the assignment of high-spin states to the
iron ions. Temperature-dependent measurements of the
magnetic susceptibility χ reveal χT-values of 4.61 and 3.61
cm3Kmol−1 at 298 K for the ferric and the ferrous complex,
respectively (Figure S13). Furthermore, the solids retain their
high-spin states over the entire temperature range from 2 to
300 K. In addition, only very weak absorptions are observed in
the vis−NIR spectrum of the solid ferric compound 1 (Figure
S15), which is consistent with the sextet spin state. In contrast,
the spectrum of the solid ferrous compound 2 (Figure S15)
displays two d−d bands at 736 and 1370 nm, which arise from
5
the splitting of the Eg excited state in Oh symmetry to two
5
5
states A1g and B1g (assuming idealized D4h local symmetry).
This finding is mainly attributed to the structural distortions
imposed by the coordinated macrocyclic ligand such as the
very elongated axial Fe−Namine bonds. The solid-state spectrum
of the ferrous compound 3 (Figure S15) with two d−d bands
at ∼891 and ∼1307 nm is distinct from that of 2. The decrease
in energetic separation of the two d−d transitions from 6300
to 3570 cm−1 indicates that the difference in ligand field
strength between the axial and the equatorial ligands is reduced
̈
in 3. Moßbauer spectroscopy was applied to verify both the
spin state and the oxidation state in complexes 1 and 2 (Figure
S9). The ferric complex showed an asymmetric doublet with
an isomeric shift (δIS) of 0.42 mm/s and a quadrupole splitting
(ΔEQ) of 2.25 mm/s (90 K). The ferrous complex showed a
doublet with δIS = 1.04 mm/s and ΔEQ = 2.63 mm/s (295 K).
The decrease in δIS is consistent with a reduced electron
shielding in the ferric compound as opposed to the ferrous
compound, and the values for δIS are in line with the
50
̈
Moßbauer data of other iron complexes. The unusually high
quadrupole splitting in the ferric complex is explained by the
pronounced differences in bond strength between the axial
amine ligands and the equatorial ligands imposed on the
complex by the distorting coordination properties of the
macrocyclic ligand.45 This substantial reduction in symmetry
of the ligand field results in a high electric field gradient at the
only seemingly totally symmetric high-spin d5 shell of the high-
Figure 2. X-band EPR spectrum (9.3452 GHz) of a frozen solution of
1 in DMF containing 0.2 mol·L−1 TBAP at 10 K. Effective g-values are
indicated. Some very minor paramagnetic contaminants are found at
geff = 4.84 and geff = 4.28. The latter is commonly observed for ferric
samples. More details, temperature-dependent spectra, and simu-
lations can be found in the Supporting Information.
̈
spin iron(III) ion. In our study, Mossbauer spectroscopy also
proved to be a very valuable tool for detecting small
contaminations by the iron(II) starting complex in the
attempted preparation of pure 1 by oxidation of 2 with
molecular oxygen.
Depending on the substance investigated (1a, 1b, or 1c), the
ATR-IR spectrum of 1 displays a feature between 3289 and
3392 cm−1 (Figure S7). On the basis of the shift in energy of
the vibration to 2470 cm−1 upon H/D isotope substitution, the
vibration can be unambiguously assigned to the O−H
stretching vibration (Figure S8). The extent of decrease in
energy of this vibration compared to the value of an
unperturbed free hydroxide in the gas phase (3700−3570
cm−1)51 is a measure of the binding strength of the hydroxide
ligand to the metal ion as well as of the extent of involvement
The rhombicity matches the low symmetry of the iron(III)
site. A zero-field splitting constant D of similar size is obtained
by fitting the magnetic susceptibility data of solid 1 assuming
an isotropic g-value (D = +3.23 cm−1).
Ä
É
Å
Å
Å
Å
Ñ
Ñ
Ñ
Ñ
2
S(S + 1)
2
2
T
̂
̂
̂
̂
̂
H = μ B gS + DÅSz
−
Ñ + E[Sx − Sy ]
Å
Ñ
Ñ
B
Å
3
Å
Ñ
(1)
Å
Ç
Ñ
Ö
13148
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
[FeIII(L-N4 Bu2)(OBz)(OH)]+/[FeII(L-N4 Bu2)(OBz)(OH)]0
(1/1−), a half-wave redox potential E1/2 = −0.31 V vs SCE is
calculated. A comparison of this redox potential to those of
other (hydroxo)iron(III) complexes in acetonitrile solution is
provided in Table S22. In contrast, the oxidation of a solution
of 2 in MeCN occurs irreversibly at an oxidative peak potential
Ep,ox of 678 mV vs SCE.
t
t
ESI mass spectra of solutions of 1 in MeCN show a
dominant signal at a m/z value of 546.3 (Figure S18). On the
basis of the m/z value and its isotope pattern, this signal
corresponds to the monocationic complex 1 (Figure S19),
indicating the structural preservation of this species in solution.
No signal arising from a μ-oxo bridged dinuclear complex can
be detected. However, a considerably less intense second signal
at m/z = 529.3 has been observed, which is attributed to the
monocationic ferrous complex 3. The strength of this signal
varies substantially with the experimental conditions employed
in the mass spectrometric investigation (Figure S20). Since we
performed the mass spectrometric experiments only with
analytically and spectroscopically pure compounds 1a and 1c,
we attribute the observation of 3 to the partial loss of a
hydroxyl radical from 1 under ESI-MS experimental con-
ditions.
To summarize the results presented up to this point, the
structural aspects of the active site in rLOX, namely, the
FeN4O2 immediate coordination environment and, partic-
ularly, the intramolecular hydrogen bonded cis-[FeIII(O2CR)-
(OH)]+ and cis-[FeII(O2CR)(OH2)]+ coordination units, are
well reproduced in the ferric complex 1 and the ferrous
complex 2, respectively. In addition to the structural model
qualities, the complexes display the same electronic properties
as the enzyme. In our opinion, even complex 3 could prove
itself to be an adequate model for an enzymatic state of
lipoxygenase, which has not yet been explicitly demonstrated
to exist for rLOX. In soybean LOX (sLOX), on the other hand,
spectroscopic studies and EXAFS investigations reveal the
existence of an equilibrium between a 5-coordinate and a 6-
coordinate species in solution (40/60%), thus implying a
dynamic coordination environment at the active iron(II) site
when no substrate is present.9,47,52 This observation is in line
with the generally accepted role of carboxylate ligands in
metalloproteins of providing easy access to a further open
coordination site at a metal ion by switching from the
bidentate chelating to the monodentate coordination mode.53
Such an equilibrium also exists between model complexes 2
and 3. Although the structural and electronic properties of the
enzyme site are well reproduced by the presented complexes,
the model complexes do not seem to match the redox
properties of the enzyme. A considerably higher redox
potential has been reported for the enzyme (+0.6 0.1 V vs
SHE for pH = 7−9).54 The reason behind this discrepancy
could be the slightly longer Fe−N and Fe−O bonds in the
rLOX7,8 and the extensive hydrogen bonding donor network
provided by the protein environment around the active site of
the enzyme.2
Solutions of both ferrous complexes 2 and 3 in MeCN
render almost identical electronic absorption spectra with two
absorption maxima observed at 732 and 1296 nm (Figures 3
Figure 3. Electronic absorption spectra of 1 and 2 in acetonitrile
solutions between 210 and 600 nm with a molar extinction coefficient
ranging from 0 to 60 000 L mol−1cm−1 (main figure) and between
400 and 1600 nm with a molar extinction coefficient ranging from 0
to 10 L mol−1cm−1 (inset).
and S15). The energetic separation of both maxima of 5945
cm−1 lies between those observed for the spectra of the solid
complexes. In addition, the slightly asymmetric appearance of
the absorption bands, especially that around 732 nm, points
toward a superposition of two spectra most likely arising from
cations 2 and 3, respectively. A superposition of two spectral
components with similar transitions as those observed for the
solid complexes 2 and 3 can be confirmed by Gaussian curve
analysis of the spectra of the dissolved complexes (Figure S16).
These results suggest that an equilibrium between complexes 2
and 3 exists in the MeCN solutions.
The ESI spectrum of dissolved complex 2 in MeCN reveals a
dominant signal at m/z = 529.3 corresponding to 3 (Figures
S18 and S19). No signal at m/z = 547.3 corresponding to the
monocation 2 is detected. This shows that, under the
experimental conditions of the ESI-mass spectrometric
measurement, complex 2 easily loses its coordinated water
molecule, rendering complex 3 during the transfer of cationic
species into the gas phase.
H Atom Abstractions from O−H Bonds. The functional
model character was established by investigating the reactivity
of the ferric complex 1 in abstracting H atoms from both O−H
and C−H bonds of organic substrates. The reactions studied
are displayed in Scheme 4.
Under anaerobic conditions at room temperature (RT), the
reaction of 1 and 2,4,6-tri-tert-butylphenol (TTBP) resulted in
the formation of the corresponding phenoxyl radical and the
ferrous complex 2 within minutes (reaction II). The generation
of the phenoxyl radical was unambiguously demonstrated by
EPR spectroscopy (Figure S24) and by UV−vis spectroscopy
(Figure S26). Moreover, the formation of the ferrous aqua
complex 2 was deduced from the ESI mass spectrum of the
reaction solution (Figure S21). The cationic mass spectrum
reveals a species with an identical m/z value as the ferrous
complex cation 3, which in an earlier paragraph was described
as also deriving from 2 under ESI-MS conditions. In addition,
unreacted complex 1 was detected. The immutability of the
ratio of intensities of the signals corresponding to 3 and 1
indicates that the reaction is not quantitative but instead leads
to the constitution of an equilibrium. In order to determine the
associated equilibrium constant K, a series of nine quantitative
EPR experiments was carried out with varying initial
concentrations of the reactants (Table S15, Figure S29). The
The electrochemical properties of the complexes were
investigated by cyclic voltammetry (Figure S17). At a scan
rate of 100 mV/s, the cyclic voltammogram of complex 1
displays an irreversible reduction process at a peak potential
Ep,red of −366 mV vs SCE with a considerably less intensive
first reoxidative current peak at Ep,ox of −262 mV vs SCE.
When both current peaks are attributed to the redox pair
13149
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Scheme 4. Reactions of 1 with Organic Substrates
Investigated for Establishing H Atom Abstraction from
Suitable O−H (I−III) and C−H (IV−VI) Bonds
Table 2. Comparison of O−H and N−H Bond Dissociation
Free Energies (BDFEs) of Various Iron(II) Complexes and
a
Organic Substrates
BDFE
compound
solvent
[kcal·mol−1
]
reference
TTBP
TEMPOH
MeCN
MeCN
H2O
DMSO
MeCN
74.8
66
84
80
72.4
55
55
22
20
[FeII(PyPz)(OH2)2]4+
[FeII(PY5)(OH2)]2+
[FeII(L-N4 Bu2)(O2CPh)(OH2)]+
this
t
work
57
55
b
[FeII(OMe2N4(tren))(OH2)]+
[FeII(H2bim)3]2+
MeCN
MeCN
62.4
68.4
a
PyPz = tetramethyl-2,3-pyridinoporphyrazine, PY5 = 2,6-bis(bis(2-
pyridyl)methoxymethyl)pyridine, (OMe2N4(tren)) = 3-((2-(bis(2-
aminoethyl)amino)ethyl)imino)2-methylbutan-2-olate, and H2bim =
b
2,2′-bi-imidazoline. The value in the original publication was
determined to be 64.7 kcal·mol−1 using CG,MeCN = 54.9 kcal·
mol−1.57 The value for CG,MeCN was later revised to be 52.6 kcal·
mol−1.55 Employing the new value of CG,MeCN, the BDFE(MeCN) for
[FeII(OMe2N4(tren))(OH2)]+ was recalculated. The structures of all
referenced complexes are depicted in Scheme S3.58
carbonyl oxygen atom stabilizes the second O−H bond of the
water ligand, resulting in an increase of its BDFE, and is,
therefore, considered as an essential feature for the reactivity of
the active site in the enzyme.56 On the basis of the value for
the BDFE of 2, the reaction of 1 with 2,2,6,6-tetramethylpiper-
idin-1-ol (TEMPOH, reaction I) is expected to be almost
quantitative, which could, accordingly, be verified by the
experiment. It is noteworthy that, in contrast to 2, the complex
[FeII(OMe2N4(tren))(OH2)]+ reacts as a H atom donor with
the TEMPO radical.
Under aerobic conditions, complex 1 functions as a catalyst
for the peroxidation of TTBP, rendering a mixture of 4,4′-
peroxybis(2,4,6-tri(tert-butyl)-2,5-cyclohexadien-1-one) and its
2,4-isomer (reaction III) as shown by NMR spectroscopic and
X-ray structure analytical investigations (see the Supporting
Information for more details) of the crystalline material
formed. The peroxidation of TTBP is assumed to proceed via a
radical recombination step between two phenoxyl radicals and
a triplet oxygen molecule. The catalyst 1 is regenerated under
air as the ferrous complex 2 reacts readily with molecular
oxygen (Figure S22).
H Atom Abstractions from C−H Bonds. To demon-
strate the capability of a H atom abstraction from weak C−H
bonds, anaerobic reactions of 1 with an excess of 1,4-
cyclohexadiene (CHD) or 9,10-dihydroanthracene (DHA)
were investigated at RT (reactions IV and V). Both CHD and
DHA are frequently used substitute substrates for the 1,4-
pentadiene unit of the fatty acid derivatives, which are the
physiological substrates of the enzyme. H atom abstraction
from C−H bonds by metal complexes with similar BDFE are
known to occur, albeit at an approximately 10 000 times lower
rate than that of O−H bonds.59 Thus, even in the presence of a
10-fold excess of CHD, complete (>99%) conversion of the
ferric complex to the ferrous complex at room temperature
(RT) required 23 days under anaerobic conditions (Figure 4).
The ferrous product and benzene could be identified by ESI-
MS (Figure S33) and NMR spectroscopy (Figure S35),
respectively. Analogous studies with DHA yield a similarly
slow reaction. The formation of anthracene in this reaction is
also demonstrated by NMR spectroscopy (Figure S36). Under
attainment of equilibrium conditions was ensured. Measure-
ment of the EPR signal intensities allowed the determination of
the equilibrium concentrations of the phenoxyl radical
produced in each of the nine reactions starting with different
initial concentrations of the educts. Using these values, the
mean equilibrium constant K and the corresponding mean free
reaction energy ΔG at 20 °C could be calculated to be K =
1.77 × 10−2 and ΔG = 2.35 kcal·mol−1 (estimated method
error of 0.14 kcal·mol−1), respectively. Further experimental
details are discussed in the Supporting Information. With the
help of the reported bond dissociation free energy BDFE of
TTBP in MeCN (74.8 kcal·mol−1), the BDFE associated to the
OH bond in 2 in MeCN is deduced to be 72.4 kcal·mol−1.55 A
comparison to previously reported values for iron(II)
complexes is presented in Table 2.
When one neglects the intrinsic difficulties in determining
BDFE and the resulting uncertainties in comparing the values,
the value determined in this study lies within the general range
of values (68−84 kcal·mol−1) observed for other iron(III)
complexes capable of H atom abstraction. At the lower end of
the range, there is the iron(II) complex [FeII(H2bim)3]2+
possessing a N−H bond that participates in H atom transfer
reactions. Iron(II) aqua complexes display O−H bonds with
higher BDFE. The BDFE increases with the positive charge of
the complex. Thus, the tetracationic low-spin iron(II) complex
[FeII(PyPz)(OH2)2]4+ with a trans FeN4O2 coordination
environment containing a positively charged planar tetraaza-
macrocyclic ligand exhibits the highest BDFE with 84 kcal·
mol−1. The BDFE is lowered somewhat in [FeII(PY5)-
(OH2)]2+ with a FeN5O coordination environment arising
from neutral ligands. Both monocationic complexes 2 and
[FeII(OMe2N4(tren))(OH2)]+ possess a high-spin cis-FeIIN4O2
coordination environment, but only in 2 is the aqua ligand
involved in an intramolecular hydrogen bonding interaction.
The presence of a hydrogen bonding interaction involving one
of the O−H bonds of the coordinated water ligand and the
13150
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
In our opinion, the special properties of the tetraazamacro-
t
cyclic ligand L-N4 Bu2 are also crucial to the reactivity of the
ferric complex 1. Thus, the usually highly favorable formation
of μ-oxo-bridged diferric complexes is prevented by the steric
hindrance caused by the bulky amine substituents of the ligand,
ensuring the mononuclearity of the iron(III) site. In addition,
rather long axial bonds are a trademark of the ligand, thus
transferring less electron density from the ligand to the
iron(III) ion and, thereby, raising the redox potential of the
iron(III) site to some extent. Finally, the steric properties of
the macrocyclic ligand do not prevent a potential substrate
from reaching the hydroxo oxygen atom and reacting with it.
Theoretical Calculations. In order to probe if such a
mechanistic scenario is viable, the initial C−H bond cleaving
step was studied computationally. As a computational
methodology, PW6B95-D3(BJ)/def2-TZVPP/SMD-
(MeCN)//M06-L/def2-SVP/PCM(MeCN) was selected.62
Subsequently, the transition states for C−H bond cleavage
for the model substrates CHD and DHA were computed. The
resulting barriers (ΔG‡298 K) are 24.9 and 25.0 kcal·mol−1,
respectively, consistent with a reaction that occurs slowly at
RT. With the computed reaction paths, not only was an overall
H atom abstraction step confirmed to be mechanistically
reasonable, but also, more interestingly, the exact nature of the
C−H bond cleaving step could be determined. For reactions
that follow a concerted proton-coupled electron transfer type
(cPCET, alternative acronyms are EPT and CPET) mecha-
nism,63−67 two possible pathways can be differentiated: a H
atom transfer (HAT), which is characterized by the transfer of
a genuine H atom (electron and proton are transferred as a
discrete entity) and, alternatively, a cPCET where the proton
and electron are transferred separately.68 For LOXs, a cPCET
scenario has been well established.2,14−16 Therefore, the
electron flow along the intrinsic reaction coordinates (IRCs)
of these reactions was studied using intrinsic bond orbitals
(IBOs; Figure 5),69,70 a methodology that has previously been
shown to be capable of differentiating these scenarios.71 For
the oxidation of CHD, the C−H bond is found to be cleaved
in a cPCET fashion. This can be seen in Figure 5b where the
changes of the individual IBOs associated with the α-spin (i)
and β-spin (ii) manifold of the C−H bond are shown. As the
C−H bond is cleaved homolytically, the β IBO transforms into
an Fe d-orbital and the α IBO remains with the derived organic
radical in a continuous fashion; this finding is based on plotting
the root square deviation of the partial charge changes of the
studied IBOs (Figure 5a). This establishes the fate of the C−H
bond and indicates that the electron transfer is decoupled from
the proton transfer. If the lone pairs on the oxygen of the
FeIII−OH unit are inspected in an analogous way, the β-spin
manifold reveals how a lone pair is transformed into an O−H
bond, which corresponds to the proton transfer (Figure 5b,
(iii)). Thus, it is concluded that both electron and proton
transfers are indeed very much concerted. This is analogously
observed for the oxidation of DHA (see the Supporting
Information).
Figure 4. Reaction monitoring via ESI-MS by sampling the reaction
solution on an almost daily basis (see the Supporting Information for
further details). (a) Plot of the peak intensity I of the ESI mass signal
versus the reaction time t for the reaction of 1 with CHD (initial
concentrations c0 (1c) = 2 × 10−4 mol·L−1 and c0 (CHD) = 2 × 10−3
mol·L−1). (b) Logarithmic plot of the intensity I versus reaction time t
demonstrating first order kinetic properties. Experimental details are
provided in the Supporting Information.
aerobic conditions in the presence of 20% of 1, anthraquinone
can be isolated at a yield of 87% (with respect to 1) after 1
week. Under a pure oxygen atmosphere in the presence of 1%
of 1, yields of 53% can be achieved overnight (reaction VI),
demonstrating that 1 acts as a radical initiator via H atom
abstraction.
In preliminary reactivity experiments, the consumption of
the ferric complex and the formation of the ferrous complex
were followed by ESI-MS in the anaerobic reaction of 1 with a
10-fold excess of CHD at RT. An apparent pseudo-first order
reaction kinetic was observed in these experiments (Figure 4).
However, further, more carefully planned experiments are
needed to determine reliable values for the kinetic parameters
and the dependency of the rate constant on the substrate
concentration. Nevertheless, the preliminary reactivity studies
unambiguously demonstrate C−H abstraction reactivity of 1
with CHD as well as with DHA. The reactivity of 1 appears to
be slower than that observed for [FeIII(PyPz)(OH)(OH2)2]4+
(in H2O), [FeIII(Py5)(OH)]2+ (in MeCN), and
[FeIII(H2bim)2(Hbim)]2+ (in MeCN).20,22,60 Notably, in
contrast to the observed C−H abstraction reactivity of the
complex [FeIII(Py5)(OMe)]2+ in MeCN,61 no reactivity could
t
be observed for the ferric complex [FeIII(LN4 Bu2)(OMe)2]+
(4) even with TTBP under the same experimental conditions,
providing a strong argument for the special reactivity of the cis-
(carboxylato)(hydroxo)iron(III) structural motif. There seem
to be rate-enhancing cooperative effects caused by the
hydrogen bonding interaction between the coordinated
benzoate and hydroxide ligands. At the same time, the lack
of reactivity of complex 4, where the access to the lone pairs of
the alcoholate ligands is sterically encumbered (see the
Supporting Information for further details), rules out a simple,
base-assisted oxidation mechanism of the substrate induced by
prior dissociation of the basic ligand in solution. An overall H
atom abstraction is the most likely occurring initial step in this
model complex, similar to the mechanism proposed for
lipoxygenases.
Therefore, the theoretical calculations establish that the
model complex 1 not only is a structural and functional model
of LOX but also mimics the event of C−H bond cleavage in
the same fashion as LOX at an electronic structure level, which
makes complex 1 a complete model.
13151
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
functional model with very limited structural resemblance to
the active site. The current study renders a model complex that
reproduces all the relevant structural features of the first
coordination shell of the active site in the oxidized state of
rLOX, namely, the cis-FeN4O2 coordination environment and
the cis-(carboxylato)(hydroxo)iron(III) coordination unit.
Particularly noteworthy is the synthesis of both the
corresponding cis-(carboxylato)(hydroxo)iron(III) and cis-
(carboxylato)(aqua)iron(II) complexes as they represent
models for the two enzymatic states observed in the catalytic
cycle of LOX. The electronic properties of both complexes
reflect those of the iron states in the enzyme. In addition, the
(hydroxo)iron(III) complex shows reactivity in the overall H
atom abstraction reaction with O−H and C−H bonds of
suitable substrates, albeit with a lower rate than that observed
for the enzyme. The BDFE of the O−H bond in the cis-
(carboxylato)(aqua)iron(II) unit could be determined by
quantitative EPR spectroscopic methods using the BDFE of
the well-studied 2,4,6-tri(tert-butyl)phenol as a reference point.
Theoretical calculations of the reactivity of the model complex
indicate that the mechanism is consistent with the concerted
proton-coupled electron transfer found in the enzyme.
Furthermore, the ferric complex 1 can use its capability to
abstract H atoms from weak O−H and C−H bonds in order to
act as a catalyst in the aerobic peroxidation of TTBP and as a
radical initiator in the oxidation of DHA to anthraquinone by
air.
This investigation pursued the objectives of the synthetic
analogue approach in bioinorganic studies to identify those
intrinsic properties of the enzyme associated with the metal
site and its first coordination sphere and those contributed by
the protein matrix.17 Very rarely, a synthetic analogue complex
is a good model for the structural and/or electronic properties
and, at the same time, a good model for the reactivity of an
enzyme. The successful structural−functional modeling
presented here indicates that the fundamental H atom
abstraction reactivity of rLOX is derived from the cis-
(carboxylato)(hydroxo)iron(III) coordination unit in an
FeN4O2 ligand donor environment that is supplied by the
protein matrix. Though the observation of a significantly
slower reaction hints toward further protein matrix effects not
captured by these simplified model complexes, this system
offers a unique and advantageous opportunity for more
detailed studies: specifically, how well-defined changes of
structural and electronic factors at this coordination unit will
influence its reactivity, a study that is hardly possible only with
the enzyme.
Figure 5. (a) Reaction plots of the orbital changes (e−) along the
IRCs for the C−H cleavage of CHD by complex 1. Relative changes
to the electronic energies are plotted in black circles, and changes to
the IBOs are plotted separately for the α-spin (orange squares) and β-
spin (purple triangles) manifolds of the C−H bond and the α-spin
manifold of the lone pair on O (green circles). (b) Depiction of
changes to IBOs relevant to C−H cleavage of CHD by complex 1
along the IRC. Localized orbitals are depicted for the α-spin (i) and
β-spin (ii) manifolds of the C−H bond and the α IBO corresponding
to the lone pair on O, resulting in the newly formed O−H bond (iii).
All data were obtained at the M06-L/def2-SVP/CPCM level of
theory.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c04422.
■
sı
Methods, synthetic and experimental procedures,
supplementary figures and tables displaying character-
ization and structural analyses, theoretical calculations,
and crystallographic information for 1a, 1b, 1c, 2a,
2b(1), 2b(2), 2c, 3a, 4a, 4b, and 6 (PDF)
CONCLUSION
■
In conclusion, here, we report on a structural as well as
functional model for the active site in rLOX. Previous model
studies have either succeeded in reproducing the structural and
electronic aspects of the first coordination sphere of the
oxidized active site, however, at the expense of losing reaction
functionality,23,24 or were able to provide an adequate
Accession Codes
CCDC 2077026−2077036 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting the
13152
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(5) Kuhn, H.; Banthiya, S.; van Leyen, K. Mammalian lipoxygenases
and their biological relevance. Biochim. Biophys. Acta, Mol. Cell Biol.
Lipids 2015, 1851 (4), 308−330.
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: + 44 1223 336033.
(6) Xu, S.; Mueser, T. C.; Marnett, L. J.; Funk, M. O. Crystal
structure of 12-lipoxygenase catalytic-domain-inhibitor complex
identifies a substrate-binding channel for catalysis. Structure 2012,
20 (9), 1490−1497.
(7) Gillmor, S. A.; Villasenor, A.; Fletterick, R.; Sigal, E.; Browner,
M. F. The structure of mammalian 15-lipoxygenase reveals similarity
to the lipases and determinants of substrate specificity. Nat. Struct.
Biol. 1997, 4 (12), 1003−1009.
(8) Choi, J.; Chon, J. K.; Kim, S.; Shin, W. Conformational flexibility
in mammalian 15S-lipoxygenase: Reinterpretation of the crystallo-
graphic data. Proteins: Struct., Funct., Genet. 2008, 70 (3), 1023−1032.
(9) Kuban, R. J.; Wiesner, R.; Rathman, J.; Veldink, G.; Nolting, H.;
AUTHOR INFORMATION
Corresponding Author
■
Hans-Jörg Kruger − Department of Chemistry, Technische
̈
Universität Kaiserslautern, 67663 Kaiserslautern, Germany;
orcid.org/0000-0002-8652-2154; Email: krueger@
chemie.uni-kl.de
Authors
Emiel Dobbelaar − Department of Chemistry, Technische
Universität Kaiserslautern, 67663 Kaiserslautern, Germany;
orcid.org/0000-0002-5650-9231
Solé, V. A.; Kuhn, H. The iron ligand sphere geometry of mammalian
̈
Christian Rauber − Department of Chemistry, Technische
Universität Kaiserslautern, 67663 Kaiserslautern, Germany
Thorsten Bonck − Department of Chemistry, Technische
Universität Kaiserslautern, 67663 Kaiserslautern, Germany
Harald Kelm − Department of Chemistry, Technische
Universität Kaiserslautern, 67663 Kaiserslautern, Germany
Markus Schmitz − Department of Chemistry, Technische
Universität Kaiserslautern, 67663 Kaiserslautern, Germany
Matina Eloïse de Waal Malefijt − Molecular Inorganic
Chemistry, Stratingh Institute for Chemistry, Faculty of
Science and Engineering, University of Groningen, 9747 AG
Groningen, The Netherlands
15-lipoxygenases. Biochem. J. 1998, 332 (1), 237−242.
(10) Ivanov, I.; Kuhn, H.; Heydeck, D. Structural and functional
biology of arachidonic acid 15-lipoxygenase-1 (ALOX15). Gene 2015,
573 (1), 1−32.
(11) Newcomer, M. E.; Brash, A. R. The structural basis for
specificity in lipoxygenase catalysis. Protein science: a publication of the
Protein Society 2015, 24 (3), 298−309.
(12) Navaratnam, S.; Feiters, M. C.; Al-Hakim, M.; Allen, J. C.;
Veldink, G. A.; Vliegenthart, J. Iron environment in soybean
lipoxygenase-1. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol.
1988, 956 (1), 70−76.
(13) Goldsmith, C. R.; Jonas, R. T.; Stack, T. D. P. C-H bond
activation by a ferric methoxide complex: modeling the rate-
determining step in the mechanism of lipoxygenase. J. Am. Chem.
Soc. 2002, 124 (1), 83−96.
(14) Lehnert, N.; Solomon, E. I. Density-functional investigation on
the mechanism of H-atom abstraction by lipoxygenase. JBIC, J. Biol.
Inorg. Chem. 2003, 8 (3), 294−305.
(15) Knapp, M. J.; Rickert, K.; Klinman, J. P. Temperature-
dependent isotope effects in soybean lipoxygenase-1: correlating
hydrogen tunneling with protein dynamics. J. Am. Chem. Soc. 2002,
124 (15), 3865−3874.
(16) Li, P.; Soudackov, A. V.; Hammes-Schiffer, S. Impact of
Mutations on the Binding Pocket of Soybean Lipoxygenase:
Implications for Proton-Coupled Electron Transfer. J. Phys. Chem.
Lett. 2018, 9 (22), 6444−6449.
(17) Ibers, J. A.; Holm, R. H. Modeling coordination sites in
metallobiomolecules. Science (Washington, DC, U. S.) 1980, 209
(4453), 223−235.
(18) Holm, R. H.; Solomon, E. I. Preface: biomimetic inorganic
chemistry. Chem. Rev. 2004, 104 (2), 347−348.
(19) Majumdar, A.; Sarkar, S. Bioinorganic chemistry of
molybdenum and tungsten enzymes: A structural−functional
modeling approach. Coord. Chem. Rev. 2011, 255 (9−10), 1039−
1054.
(20) Goldsmith, C. R.; Stack, T. D. P. Hydrogen atom abstraction by
a mononuclear ferric hydroxide complex: insights into the reactivity of
lipoxygenase. Inorg. Chem. 2006, 45 (15), 6048−6055.
(21) Mei, F.; Ou, C.; Wu, G.; Cao, Li; Han, F.; Meng, X.; Li, J.; Li,
D.; Liao, Z. Non-heme iron(II/III) complexes that model the
reactivity of lipoxygenase with a redox switch. Dalton transactions
2010, 39 (18), 4267−4269.
(22) Gao, H.; Groves, J. T. Fast Hydrogen Atom Abstraction by a
Hydroxo Iron(III) Porphyrazine. J. Am. Chem. Soc. 2017, 139 (11),
3938−3941.
(23) Ogo, S.; Yamahara, R.; Roach, M.; Suenobu, T.; Aki, M.;
Ogura, T.; Kitagawa, T.; Masuda, H.; Fukuzumi, S.; Watanabe, Y.
Structural and spectroscopic features of a cis (hydroxo)-Fe(III)-
(carboxylato) configuration as an active site model for lipoxygenases.
Inorg. Chem. 2002, 41 (21), 5513−5520.
(24) Ogo, S.; Wada, S.; Watanabe, Y.; Iwase, M.; Wada, A.; Harata,
M.; Jitsukawa, K.; Masuda, H.; Einaga, H. Synthesis, Structure, and
Spectroscopic Properties of [Feiii(tnpa)(OH)(PhCOO)]ClO4: A
Johannes E. M. N. Klein − Molecular Inorganic Chemistry,
Stratingh Institute for Chemistry, Faculty of Science and
Engineering, University of Groningen, 9747 AG Groningen,
The Netherlands; orcid.org/0000-0002-1290-597X
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c04422
Funding
H.-J.K. and E.D. would like to acknowledge financial support
by the University of Kaiserslautern and funding supplied by the
DFG (INST 248/266-1 FUGG).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
J.E.M.N.K. and M.E.d.W.M. would like to thank the Center for
Information Technology of the University of Groningen for
their support and for providing access to the Peregrine high-
performance computing cluster. In addition, H.-J.K. and
J.E.M.N.K. thank the COST Action CM1305 (ECOSTBio)
for financial support to attend its conferences.
DEDICATION
Dedicated to the memory of Richard H. Holm.
■
REFERENCES
■
(1) Andreou, A.; Feussner, I. Lipoxygenases - Structure and reaction
mechanism. Phytochemistry 2009, 70 (13−14), 1504−1510.
(2) Gaffney, B. J. EPR Spectroscopic Studies of Lipoxygenases.
Chem. - Asian J. 2020, 15 (1), 42−50.
(3) Schilstra, M. J.; Veldink, G. A.; Vliegenthart, J. F. The
dioxygenation rate in lipoxygenase catalysis is determined by the
amount of iron (III) lipoxygenase in solution. Biochemistry 1994, 33
(13), 3974−3979.
(4) Bryant, R. G.; Bailey, J. M.; Schewe, T.; Rapoport, S. M.
Positional Specificity of a Reticulocyte Lipoxygenase. J. Biol. Chem.
1982, 257 (257), 6050−6055.
13153
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Model Complex for an Active Form of Soybean Lipoxygenase-1.
Angew. Chem., Int. Ed. 1998, 37 (15), 2102−2104.
(25) Zaragoza, J. P. T.; Yosca, T. H.; Siegler, M. A.; Moënne-Loccoz,
P.; Green, M. T.; Goldberg, D. P. Direct Observation of Oxygen
Rebound with an Iron-Hydroxide Complex. J. Am. Chem. Soc. 2017,
139 (39), 13640−13643.
(26) Ching, W.-M.; Zhou, A.; Klein, J. E. M. N.; Fan, R.; Knizia, G.;
Cramer, C. J.; Guo, Y.; Que, L. Characterization of the Fleeting
Hydroxoiron(III) Complex of the Pentadentate TMC-py Ligand.
Inorg. Chem. 2017, 56 (18), 11129−11140.
(41) Yadav, V.; Siegler, M. A.; Goldberg, D. P. Temperature-
Dependent Reactivity of a Non-heme FeIII(OH)(SR) Complex:
Relevance to Isopenicillin N Synthase. J. Am. Chem. Soc. 2021, 143
(1), 46−52.
(42) Keshari, K.; Bera, M.; Velasco, L.; Munshi, S.; Gupta, G.;
Moonshiram, D.; Paria, S. Characterization and reactivity study of
non-heme high-valent iron−hydroxo complexes. Chemical science
2021, 12 (12), 4418−4424.
(43) MacBeth, C. E.; Golombek, A. P.; Young, V. G.; Yang, C.;
Kuczera, K.; Hendrich, M. P.; Borovik, A. S. O2 activation by
nonheme iron complexes: A monomeric Fe(III)-Oxo complex derived
from O2. Science 2000, 289 (5481), 938−941.
(44) Cook, S. A.; Ziller, J. W.; Borovik, A. S. Iron(II) complexes
supported by sulfonamido tripodal ligands: endogenous versus
exogenous substrate oxidation. Inorg. Chem. 2014, 53 (20), 11029−
11035.
(27) Mukherjee, J.; Lucas, R. L.; Zart, M. K.; Powell, D. R.; Day, V.
W.; Borovik, A. S. Synthesis, structure, and physical properties for a
series of monomeric iron(III) hydroxo complexes with varying
hydrogen-bond networks. Inorg. Chem. 2008, 47 (13), 5780−5786.
(28) Bénisvy, L.; Halut, S.; Donnadieu, B.; Tuchagues, J.-P.;
Chottard, J.-C.; Li, Y. Monomeric iron(II) hydroxo and iron(III)
dihydroxo complexes stabilized by intermolecular hydrogen bonding.
Inorg. Chem. 2006, 45 (6), 2403−2405.
(45) Kruger, H.-J. Spin transition in octahedral metal complexes
̈
containing tetraazamacrocyclic ligands. Coord. Chem. Rev. 2009, 253
(19−20), 2450−2459.
(29) Çelenligil-Çetin, R.; Paraskevopoulou, P.; Dinda, R.; Lalioti, N.;
Sanakis, Y.; Rawashdeh, A. M.; Staples, R. J.; Sinn, E.; Stavropoulos, P.
Oxidative Ligand Rearrangement Due to Incipient Aminyl Radicals in
the Oxidation of Iron(II) Species with Dioxygen. Eur. J. Inorg. Chem.
2008, 2008 (5), 673−677.
(46) Koch, W. O.; Kruger, H.-J. A Highly Reactive and Catalytically
̈
Active Model System for Intradiol-Cleaving Catechol Dioxygenases:
Structure and Reactivity of Iron(III) Catecholate Complexes ofN,N′-
Dimethyl-2,11-diaza[3.3](2,6)pyridinophane. Angew. Chem., Int. Ed.
Engl. 1996, 34 (2324), 2671−2674.
(30) Ortega-Villar, N.; Ugalde-Saldívar, V. M.; Flores-Pérez, B.;
Flores-Alamo, M.; Real, J. A.; Moreno-Esparza, R. A new N6
hexadentate ligand and a novel heptacoordinated N6O-type Fe(III)
compounds: Synthesis, characterization and structure of [Fe-
(47) Scarrow, R. C.; Trimitsis, M. G.; Buck, C. P.; Grove, G. N.;
Cowling, R. A.; Nelson, M. J. X-ray spectroscopy of the iron site in
soybean lipoxygenase-1: changes in coordination upon oxidation or
addition of methanol. Biochemistry 1994, 33 (50), 15023−15035.
(48) Becker, E.; Kirchner, K.; Mereiter, K. Hexaaqua-iron(II) bis-
fac-tribromido-tricarbonyl-ferrate(II). Acta Crystallogr., Sect. E: Struct.
Rep. Online 2009, 65 (10), i71.
(49) Yahyaoui, S.; Rekik, W.; Naïli, H.; Mhiri, T.; Bataille, T.
Synthesis, crystal structures, phase transition characterization and
thermal decomposition of a new dabcodiium hexaaquairon(II)
bis(sulfate): (C6H14N2)[Fe(H2O)6](SO4)2. J. Solid State Chem.
2007, 180 (12), 3560−3570.
̈
̈
(50) Gutlich, P.; Bill, E.; Trautwein, A. X. Mossbauer spectroscopy and
transition metal chemistry; Springer: Heidelberg, 2011.
(51) Lutz, H. D.; Eckers, W.; Haeuseler, H. OH stretching
frequencies of solid hydroxides and of free OH− ions. J. Mol. Struct.
1982, 80, 221−224.
(52) Pavlosky, M. A.; Zhang, Y.; Westre, T. E.; Gan, Q.-F.; Pavel, E.
G.; Campochiaro, C.; Hedman, B.; Hodgson, K. O.; Solomon, E.I.
Near-infrared circular dichroism and magnetic circular dichroism and
x-ray absorption spectral comparison of the non-heme ferrous active
sites of plant and mammalian 15-lipoxygenases. J. Am. Chem. Soc.
1995, 117, 4316−4327.
(53) Holm, R. H.; Kennepohl, P.; Solomon, E. I. Structural and
Functional Aspects of Metal Sites in Biology. Chem. Rev. 1996, 96,
2239−2314.
(54) Nelson, M. J. Catecholate complexes of ferric soybean
lipoxygenase 1. Biochemistry 1988, 27, 4273−4278.
−
−
(dimpyen)(OH)](A)2 (A = PF6 or ClO4 ). Inorg. Chim. Acta
2011, 375 (1), 213−219.
(31) Yeh, C. Y.; Chang, C. J.; Nocera, D. G. ″Hangman″ porphyrins
for the assembly of a model heme water channel. J. Am. Chem. Soc.
2001, 123 (7), 1513−1514.
(32) Chambers, M. B.; Groysman, S.; Villagrán, D.; Nocera, D. G.
Iron in a trigonal tris(alkoxide) ligand environment. Inorg. Chem.
2013, 52 (6), 3159−3169.
(33) Burger, J.; Klufers, P. Polyol-Metall-Komplexe. 17. Kristalline
̈
Eisen(III)-Komplexe mit zweifach deprotonierten Anhydroerythrit-
Liganden. Z. Anorg. Allg. Chem. 1996, 622 (10), 1740−1748.
(34) Hodges, K. D.; Wollmann, R. G.; Kessel, S. L.; Hendrickson, D.
N.; van Derveer, D. G.; Barefield, E. K. Preparations and properties of
nitrosyl complexes of iron tetramethylcyclam. X-ray structures of
[Fe(C14H32N4)NO](BF4)2, a S = 3/2−1/2 spin-equilibrium complex,
and [Fe(C14H32N4)(NO)(OH)](ClO4)2·CH3CN. J. Am. Chem. Soc.
1979, 101 (4), 906−917.
(35) Drummond, M. J.; Ford, C. L.; Gray, D. L.; Popescu, C. V.;
Fout, A. R. Radical Rebound Hydroxylation Versus H-Atom Transfer
in Non-Heme Iron(III)-Hydroxo Complexes: Reactivity and Struc-
tural Differentiation. J. Am. Chem. Soc. 2019, 141 (16), 6639−6650.
(36) Soo, H. S.; Komor, A. C.; Iavarone, A. T.; Chang, C. J. A
hydrogen-bond facilitated cycle for oxygen reduction by an acid- and
base-compatible iron platform. Inorg. Chem. 2009, 48 (21), 10024−
10035.
(55) Wise, C. F.; Agarwal, R. G.; Mayer, J. M. Determining Proton-
Coupled Standard Potentials and X-H Bond Dissociation Free
Energies in Nonaqueous Solvents Using Open-Circuit Potential
Measurements. J. Am. Chem. Soc. 2020, 142 (24), 10681−10691.
(56) In general, a comparison with values reported in the literature
turns out to be difficult because the respective BDFE values of the
complexes depend decisively on the method used for their
determinations. The most frequent method is the “Bordwell” analysis
based on measuring the redox potentials and pKa values of species
involved in the proton-coupled electron transfer reaction of a metal
complex, but here, experimental problems such as the frequent
observation of irreversible electrochemical responses in cyclic
voltammograms lead to less reliable data. The recently reported
method of open-circuit potential measurement by J. M. Mayer seems
to avoid these kinetic problems,55 rendering thermodynamically more
accurate data. In addition, the comparison of BDFE values obtained in
different solvent, especially when the solvents differ substantially in
(37) Matson, E. M.; Bertke, J. A.; Fout, A. R. Isolation of iron(II)
aqua and hydroxyl complexes featuring a tripodal H-bond donor and
acceptor ligand. Inorg. Chem. 2014, 53 (9), 4450−4458.
(38) Celenligil-Cetin, R.; Paraskevopoulou, P.; Dinda, R.; Staples, R.
J.; Sinn, E.; Rath, N. P.; Stavropoulos, P. Synthesis, characterization,
and reactivity of iron trisamidoamine complexes that undergo both
metal- and ligand-centered oxidative transformations. Inorg. Chem.
2008, 47 (3), 1165−1172.
(39) Yadav, V.; Gordon, J. B.; Siegler, M. A.; Goldberg, D. P.
Dioxygen-Derived Nonheme Mononuclear FeIII(OH) Complex and
Its Reactivity with Carbon Radicals. J. Am. Chem. Soc. 2019, 141 (26),
10148−10153.
(40) Yadav, V.; Rodriguez, R. J.; Siegler, M. A.; Goldberg, D. P.
Determining the Inherent Selectivity for Carbon Radical Hydrox-
ylation versus Halogenation with FeIII(OH)(X) Complexes: Rele-
vance to the Rebound Step in Non-heme Iron Halogenases. J. Am.
Chem. Soc. 2020, 142 (16), 7259−7264.
13154
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
hydrogen bonding properties, is accompanied by a high degree of
inaccuracy. Furthermore, errors in deriving constants used in the
“Bordwell” analysis as well as confusion concerning the usage of
BDFE versus BDE considerably complicate a reasonable comparison
of values.
(57) Brines, L. M.; Coggins, M. K.; Poon, P. C. Y.; Toledo, S.;
Kaminsky, W.; Kirk, M. L.; Kovacs, J. A. Water-soluble Fe(II)-H2O
complex with a weak O-H bond transfers a hydrogen atom via an
observable monomeric Fe(III)-OH. J. Am. Chem. Soc. 2015, 137 (6),
2253−2264.
(58) For a depiction of the structures, see the Supporting
Information.
(59) Mayer, J. M. Understanding hydrogen atom transfer: from
bond strengths to Marcus theory. Acc. Chem. Res. 2011, 44 (1), 36−
46.
(60) Roth, J. P.; Mayer, J. M. Hydrogen Transfer Reactivity of a
Ferric Bi-imidazoline Complex That Models the Activity of
Lipoxygenase Enzymes. Inorg. Chem. 1999, 38 (12), 2760−2761.
(61) Jonas, R. T.; Stack, T. D. P. C−H Bond Activation by a Ferric
Methoxide Complex: A Model for the Rate-Determining Step in the
Mechanism of Lipoxygenase. J. Am. Chem. Soc. 1997, 119 (36),
8566−8567.
(62) For a detailed description of the computational methods used
and the corresponding references, see the Supporting Information.
Here, we also provide a brief exploration of the reliability of the
selected computational methodology.
(63) Hammes-Schiffer, S. Proton-Coupled Electron Transfer:
Moving Together and Charging Forward. J. Am. Chem. Soc. 2015,
137 (28), 8860−8871.
(64) Weinberg, D. R.; Gagliardi, C. J.; Hull, J. F.; Murphy, C. F.;
Kent, C. A.; Westlake, B. C.; Paul, A.; Ess, D. H.; McCafferty, D. G.;
Meyer, T. J. Proton-coupled electron transfer. Chem. Rev. 2012, 112
(7), 4016−4093.
(65) Warren, J. J.; Tronic, T. A.; Mayer, J. M. Thermochemistry of
proton-coupled electron transfer reagents and its implications. Chem.
Rev. 2010, 110 (12), 6961−7001.
(66) Huynh, M. H. V.; Meyer, T. J. Proton-coupled electron transfer.
Chem. Rev. 2007, 107 (11), 5004−5064.
(67) Mayer, J. M. Proton-coupled electron transfer: a reaction
chemist’s view. Annu. Rev. Phys. Chem. 2004, 55, 363−390.
(68) Usharani, D.; Lacy, D. C.; Borovik, A. S.; Shaik, S.
Dichotomous hydrogen atom transfer vs proton-coupled electron
transfer during activation of X-H bonds (X = C, N, O) by nonheme
iron-oxo complexes of variable basicity. J. Am. Chem. Soc. 2013, 135
(45), 17090−17104.
(69) Knizia, G. Intrinsic Atomic Orbitals: An Unbiased Bridge
between Quantum Theory and Chemical Concepts. J. Chem. Theory
Comput. 2013, 9 (11), 4834−4843.
(70) Knizia, G.; Klein, J. E. M. N. Electron flow in reaction
mechanisms–revealed from first principles. Angew. Chem., Int. Ed.
2015, 54 (18), 5518−5522.
(71) Klein, J. E. M. N.; Knizia, G. cPCET versus HAT: A Direct
Theoretical Method for Distinguishing X-H Bond-Activation
Mechanisms. Angew. Chem., Int. Ed. 2018, 57 (37), 11913−11917.
13155
https://doi.org/10.1021/jacs.1c04422
J. Am. Chem. Soc. 2021, 143, 13145−13155