Series of structural and functional models for the ES (enzyme-substrate) complex of the Co(II)-containing quercetin 2,3-dioxygenase

Article abstract of DOI:10.1021/ic402695c

A series of mononuclear Co^II-flavonolate complexes [Co ^IIL^R(fla)] (L^RH = 2-{[bis(pyridin-2-ylmethyl) amino]methyl}-p/m-R-benzoic acid; R = p-OMe (1), p-Me (2), m-Br (4), and m-NO₂ (5); fla = flavonolate) were designed and synthesized as structural and functional models for the ES (enzyme-substrate) complexes to mimic the active site of the Co(II)-containing quercetin 2,3-dioxygenase (Co-2,3-QD). The metal center Co(II) ion in each complex shows a similar distorted octahedral geometry. The model complexes display high enzyme-type dioxygenation reactivity (oxidative O-heterocyclic ring opening of the coordinated substrate flavonolate) at low temperature, presumably due to the attached carboxylate group in the ligands. The reactivity exhibits a substituent group dependent order of -OMe (1) > -Me (2) > -H (3)14b > -Br (4) > -NO₂ (5), and the Hammett plot is linear (ρ = -0.78). This can be explained as the electronic nature of the substituent group in the ligands may influence the conformation and redox potential of the bound flavonolate and finally bring different reactivity. The structures, properties, and reactivity of the model complexes show some dependence on the substituent group in the supporting model ligands, and there is some relationship among them. This study is the first example of a series of structural and functional ES models of Co-2,3-QD, with focus on the effects of the electronic nature of substituted groups and the carboxylate group of the ligands to the dioxygenation reactivity, that will provide important insights into the structure-property-reactivity relationship and the catalytic role of Co-2,3-QD.

Full text of DOI:10.1021/ic402695c

Article
pubs.acs.org/IC
Series of Structural and Functional Models for the ES (Enzyme−
Substrate) Complex of the Co(II)-Containing Quercetin 2,3-
Dioxygenase
Ying-Ji Sun,* Qian-Qian Huang, and Jian-Jun Zhang
School of Chemistry, Dalian University of Technology, 2 Linggong Road, Dalian 116024, China
S
* Supporting Information
ABSTRACT: A series of mononuclear CoII−flavonolate
complexes [CoIILR(fla)] (LRH = 2-{[bis(pyridin-2-ylmethyl)-
amino]methyl}-p/m-R-benzoic acid; R = p-OMe (1), p-Me
(2), m-Br (4), and m-NO2 (5); fla = flavonolate) were
designed and synthesized as structural and functional models
for the ES (enzyme−substrate) complexes to mimic the active
site of the Co(II)-containing quercetin 2,3-dioxygenase (Co−
2,3-QD). The metal center Co(II) ion in each complex shows
a similar distorted octahedral geometry. The model complexes
display high enzyme-type dioxygenation reactivity (oxidative
O-heterocyclic ring opening of the coordinated substrate flavonolate) at low temperature, presumably due to the attached
carboxylate group in the ligands. The reactivity exhibits a substituent group dependent order of −OMe (1) > −Me (2) > −H
(3)14b > −Br (4) > −NO2 (5), and the Hammett plot is linear (ρ = −0.78). This can be explained as the electronic nature of the
substituent group in the ligands may influence the conformation and redox potential of the bound flavonolate and finally bring
different reactivity. The structures, properties, and reactivity of the model complexes show some dependence on the substituent
group in the supporting model ligands, and there is some relationship among them. This study is the first example of a series of
structural and functional ES models of Co−2,3-QD, with focus on the effects of the electronic nature of substituted groups and
the carboxylate group of the ligands to the dioxygenation reactivity, that will provide important insights into the structure−
property−reactivity relationship and the catalytic role of Co−2,3-QD.
Scheme 1
INTRODUCTION
■
Co ion plays an important role in many biological processes. It
is mainly included in coenzyme B12;1 however, there are various
kinds of Co(II)-substituted enzymes. Most of them are
nonredox zinc(II) enzymes,2 while Co(II) substitution is also
found in some oxidoreductase, such as amine oxidase (Cu),3a
catechol dioxygenase (Fe),3b acireductone dioxygenase (ARD)
(Fe and Ni),3c and quercetin 2,3-dioxygenase (2,3-QD) (Cu
and Fe).4 The iron-containing dioxygenase has been extensively
studied; however, the noniron metal, especially the Co(II)-
containing dioxygenases, received much less attention until
recently. The metal diversity among the various kinds of
dioxygenases for catalysis and the variability in metal cofactor
selectivity in the same dioxygenase are a very interesting and
poorly understood field and will be a topic of future study.
Quercetin 2,3-dioxygenase (2,3-QD) catalyzes the oxygen-
ative ring-opening reaction of the O-heterocycle of quercetin
(3′,4′,5,7-tetrahydroxyflavonol, QUE) to the corresponding
depside (phenolic carboxylic acid esters) and carbon monoxide
by activation of dioxygen (Scheme 1) .5
one water molecule, and one carboxylate group of Glu. Another
is a distorted tetrahedral or square pyramidal geometry,
depending on whether there is not or is an additionally weak
interaction of the carboxylate group of Glu.
Under anaerobic conditions, the deprotonated substrate was
coordinated to the Cu(II) center via displacement of the water
molecule to form an ES (enzyme−substrate) complex.8 Site-
directed mutagenesis experiments show that the enzyme
activity was lost by mutation of Glu73,6 indicating that the
carboxylate group of Glu73 plays an important role in the
catalytic reaction. It may act as an active site base to
deprotonate the substrate, help to stabilize the bound substrate
through a hydrogen bond, modulate the redox potential of the
metal ion, and lower the energy barriers.6,9
The mononuclear M(II) (M = Cu, Fe) active sites of 2,3-QD
from Aspergillus japonicus6 and Bacillus subtilis4a,7 have similar
structure. Each of them shows two distinct coordination
environments. One is distorted trigonal-bipyramidal geometry,
and the M(II) ion is coordinated by three histidine imidazoles,
Received: October 26, 2013
Published: March 6, 2014
© 2014 American Chemical Society
2932
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
Metal-replacement experiments of the native Fe(II)-contain-
ing 2,3-QD from B. subtilis demonstrate that the Co(II)- and
Mn(II)-substituted enzymes show 24- and 35-times increased
electronic substituent effects of the supporting model ligands
on the reactivity was reported.
To gain insights into the electronic substituent effects and
the carboxylate effects of the supporting model ligand on the
catalytic role of Co-2,3-QD, we herein designed and
synthesized a series of CoII−flavonolate complexes
[CoIILR(fla)] supported by the p/m-substituted carboxylate
ligand LR (LRH = 2-{[bis(pyridin-2-ylmethyl)amino]methyl}-
p/m-R-benzoic acid (Figure 1); R = p-OMe (1), p-Me (2), m-
catalytic activity relative to the native enzyme, respectively.4b
A
recent study of 2,3-QD from Streptomyces sp. FLA (mono-
cupin) has shown that the Ni(II)- and Co(II)-containing
enzymes exhibit the highest and next highest level of
reactivity.4c These results clearly indicate that the catalytic
activity of the enzymes are remarkably affected by the metal
ions, and the 2,3-QD enzymes have surprising variability in
metal cofactor selectivity and structural flexibility for the metal
cofactor.4,7
Generally, the amino acid residues of/around the active site
of the enzyme and their substituent groups could influence the
properties and catalytic activity of the enzyme by affecting the
coordination environment of the metal cofactor and the
conformation of the enzyme. However, detailed roles of the
electronic nature, carboxylate group, and metal ion on the
catalytic mechanism are not clear.
To date, most of the reported 2,3-QD model complexes are
Cu(II) based,10,16,17 and little attention has been paid to other
transition-metal ions,11−15 especially the Co(II) com-
plexes.12−14 Nishinaga’s group reported [CoII(salen)]/
[CoIII(salen)(OH)] (salenH2 = 1,6-bis-(2-hydroxyphenyl)-2,5-
diaza-hexa-1,5-diene) could be used as catalyst in the
dioxygenation of the substrate flavonol derivatives at room
temperature.12a,b The electronic substituent effects of the
substrate flavonol and solvent effects on the catalytic reactivity
were also discussed. They also reported a series of [CoIII(4′/
7R-fla)(5′/5R-salen)] (fla = flavonolate)12c,d 2,3-QD model
complexes; however, no detailed kinetic study was reported.
Thus far, [CoIII(fla)(salen)],12c [(6-Ph2TPA)CoII(fla)]ClO4 (6-
Ph2TPA = N,N-bis((6-phenyl-2-pyridyl)methyl)-N-((2-
pyridyl)methyl)amine),13a and [CoIILH(fla)] (3) (LHH = 2-
{[bis(pyridin-2-ylmethyl)amino]methyl}benzoic acid)14b are
the only three examples of the structurally characterized Co−
flavonolate complexes. However, in [CoIII(fla)(salen)] the
oxidation state of Co ion is different from that of Co-2,3-QD.
Kinetic study of [(6-Ph2TPA)CoII(fla)]ClO4 indicates it
displays “dioxygenase-type reactivity” upon being irradiated
by UV light.13b Our previously reported [CoIILH(fla)] (3)14b is
the only structurally characterized ES-model complex of Co-
2,3-QD with enzyme-type dioxygenation reactivity. The
observed higher dioxygenation reactivity of [CoIILH(fla)]
(3)14b sparked our interest to further study the role of the
higher potential Co(II) center in O2 activation and the C−C
bond cleavage process.
Figure 1. Structure of ligand LRH.
Br (4), and m-NO2 (5); fla = flavonolate) as structural and
functional ES models of Co−2,3-QD. Details of their
structures, spectroscopic features, redox properties, and
reactivity toward dioxygen are discussed.
EXPERIMENTAL SECTION
■
General and Physical Methods. All chemicals used in this study
except the ligands and complexes were commercial products of the
highest available purity and further purified by the standard method18
if necessary. FT-IR spectra were recorded with a Nicolet 6700
spectrophotometer. UV−vis spectra were measured using an Agilent
Technologies HP8453 diode array spectrophotometer. ESI-MS
(electrospray ionization mass spectra) measurements were performed
1
on an Agilent Technologies HP1100LC-MSD. H NMR spectra were
recorded on a Bruker 400WB. EPR spectra were obtained on a Bruker
MEX-Plus spectrometer fitted with a liquid helium cooled probe.
Spectra of the complexes (4 mM in 0.5 mL of DMF) were recorded at
−173 °C, about 20 mW microwave powers, and about 9.40 GHz. The
sample was put into a nitric acid washed quartz EPR tube and frozen
in liquid nitrogen. Cyclic voltammetry data was collected using a
CHI620b system. All CV data were obtained under N2 in DMF with a
complex concentration of 2 mM and KClO4 (0.5 M) as the supporting
electrolyte. The experimental setup consisted of a glassy carbon
working electrode, an SCE reference electrode, and a platinum wire
auxiliary electrode. All potentials are reported versus SCE. HPLC-MS
measurements were performed on a Thermo Fisher Scientific LTQ
Orbitrap XL HPLC-MS.
Most of the model ligands used in the model complexes of
2,3-QD are polyamine N-chelating ligands;10−13,15,16 the role of
active site carboxylate group of Glu has received much less
attention. Two reports11c,16 focused on the reactivity of model
complexes of 2,3-QD also reveal that dioxygenation of the
bound substrate flavonolate can be promoted by addition of
excess free carboxylate ligands, due to the change of
coordination mode of the bound substrate flavonolate from
bidentate to monodentate resulting from coordination of the
exogenous carboxylate. Thus far, five types of ligands bearing a
carboxylate group were used in 2,3-QD model complexes,14,17
but the effect of the introduced carboxylate group on the
reactivity was not observed except [MIILH(fla)].14b Besides,
several biomimetic studies about the electronic effects of the
substituted groups of the substrate flavonol on the reactivity
have been reported.10,11c,15 However, no example about the
X-ray Crystallography. All single crystals of the complexes were
mounted on a glass capillary. X-ray diffraction data were collected by a
Bruker Smart APEXII CCD single-crystal diffractometer using Mo Kα
radiation (λ = 0.71073 Å) to 2θ
of 50−55.0° at 296 K.
max
Crystallographic calculations were performed by a combination of
direct and heavy atom methods using SHELXTL 97.19 All non-
hydrogen and hydrogen atoms were refined anisotropically and
isotropically.
Kinetic Measurements. [CoIILR(fla)] Dependence. The reactions
of the complexes [CoIILR(fla)] with O2 were performed in a 10 mm
path length UV−vis cell that was held in a Unisoku thermostatted cell
holder USP-203 (a desired temperature can be fixed within 0.5 °C).
After the solution (3 mL of DMF) of [CoIILR(fla)] (0.78−1.7 × 10−4
M) was kept at a desired temperature (70 °C) under N2 for several
minutes, the N2 was replaced with O2. The time courses of the
reactions were followed by monitoring the absorption changes at a
λmax due to the π → π* transition of the coordinated flavonolate. The
2933
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
Figure 2. ORTEP representation of (a) [CoIILOMe(fla)]·CH3OH (1), (b) [CoIILBr(fla)]·CH3OH·H2O (4), and (c) [CoIILNO2(fla)]·Et2O·0.5H2O
(5). (d) Plot of the torsion angle at C(21)− C(22)− C(30)− C(31) vs Hammett constants σ. Ellipsoids are shown at the 50% probability level.
Hydrogen atoms and solvent molecules are omitted for clarity.
initial concentrations of [CoIILR(fla)] were determined using the ε
value at the λmax of each complex (see Table 3).
Reaction Product Analysis. [CoIILR(fla)] (2.0 × 10−3 M in 2 mL
of DMF) was stirred at 70 °C for 8 h under O2. After the reaction, the
mixture was concentrated by evaporation and the remaining residue
was dissolved in 1.9 mL of MeOH. o-Methylbenzoic acid (10 mM in
100 μL of MeOH solution, total 0.5 mM in 2.0 mL of solution) was
added to the above solution as an internal standard, and the reaction
products were analyzed with a Thermo Fisher Scientific LTQ Orbitrap
XL HPLC-MS with an online UV−vis detector (λ: 210 nm). A
Hypersil GOLD C18 column (Thermo Fisher Scientific 150 mm × 2.1
mm, 5 μm) was used for HPLC analysis at room temperature with a
mobile phase (MeOH and 5 mM NH4OAc) with some gradient at a
constant flow rate of 0.5 mL min−1. The yields of the reaction products
were calculated using the standard calibration curve.
O2 Dependence. In a typical experiment, [CoIILR(fla)] (1.0 × 10−4
M in 60 mL of DMF) was kept under N2 for several minutes in a
thermostatted reaction vessel connecting to a manometer to regulate
constant dioxygen pressure. The solution was then heated to the
desired temperature (60−75 0.5 °C), and the N2 was replaced with
O2. The reaction mixture was taken by syringe periodically (ca. every 2
min), and the time courses of the reactions were followed by
monitoring the absorption changes at a λmax due to the π → π*
transition of the coordinated flavonolate. O2 concentrations were
calculated from literature data20 taking into account the partial
pressure of DMF21 and assuming the validity of Dalton’s law.
Determination of the Activation Parameters. The activation
parameters for degradation of [CoIILR(fla)] were obtained from an
Eyring plot (Table 4, Supporting Information, Table S3 and Figure
S5) with [CoIILR(fla)] (1.0 × 10−4 M) at the temperature range of
55−85 °C. The experimental procedures were similar to the kinetic
study described above.
Synthesis of the Model Ligands and Complexes. The
synthesis and characterization details of the model ligands LRH (R =
p-OMe, p-Me, m-Br, and m-NO2) are shown in the Supporting
Information.
General Procedure for Synthesis of [CoIILR(fla)]. Under a nitrogen
atmosphere, a dry CH3OH (1.0 mL) solution of LRH (0.05 mmol)
2934
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
Table 1. Summary of X-ray Data Collection and Refinement
[CoIILOMe(fla)]·CH3OH (1)
[CoIILBr(fla)]·CH3OH·H2O (4)
[CoIILNO2(fla)]·Et2O·0·5H2O (5)
empirical formula
C37H33CoN3O7
690.59
C36H32BrCoN3O7
757.49
C39H37CoN4O8.50
756.66
Mr
cryst syst
space group
a/Å
orthorhombic
Pbca
monoclinic
P2(1)/c
8.9469(6)
17.8871(9)
23.1484(14)
90
monoclinic
P2(1)/c
8.8402(4)
18.0551(8)
23.3936(14)
90
9.6655(10)
22.650(2)
30.634(3)
90
b/Å
c/Å
α/deg
β/deg
γ/deg
V/Å3
Z
90
99.073(4)
90
97.638(3)
90
90
6706.5(12)
8
3658.2(4)
4
3700.7(3)
4
d
calcd/g cm−3
1.368
1.375
1.358
T/K
296(2)
prism
296(2)
296(2)
cryst habit
prism
prism
color
wine red
0.20 × 0.15 × 0.10
0.566
wine red
0.30 × 0.25 × 0.20
1.611
wine red
0.25 × 0.18 × 0.12
0.523
cryst size (mm3)
μ/mm−1
2θmax/deg
55.00
50.00
52.00
completeness to θ (%)
no. of reflns collected
no. of independent reflns
Rint
99.7
99.7
99.7
43 075
13 154
14 550
7682
6429
7247
0.0606
0.0304
0.0392
variable params
433
432
454
a
b
R1 /wR2
0.0568/0.1714
1.043
0.0668/0.2297
1.026
0.0540/0.1527
1.006
goodness-of-fit on F2
−3
Δρmax/min/e Å
0.973/−0.433
1.901/−0.911
0.432/−0.436
a
b
R1 = ∑(||Fo| − |Fc||)/∑|Fo|. wR2 = [∑w (Fo − Fc2)2/∑w(Fo )2]1/2, where w = 1/[σ2(Fo ) + (aP)2 + bP].
2
2
2
Table 2. Selected Bond Distances (Angstroms) and Bond Angles (degrees) for the Complexes
[CoIILOMe(fla)]·CH3OH (1)
[CoIILBr(fla)]·CH3OH·H2O (4)
[CoIILNO2(fla)]·Et2O·0·5H2O (5)
Co(1)−O(1)
Co(1)−O(3)
Co(1)−O(4)
Co(1)−N(1)
Co(1)−N(2)
Co(1)−N(3)
O(3)−C(21)
O(4)−C(27)
2.035(2)
1.997(2)
2.157(2)
2.194(2)
2.131(3)
2.163(3)
1.310(4)
1.250(4)
1.378(4)
99.92(9)
87.79(10)
90.59(9)
86.30(9)
163.19(10)
168.69(10)
96.94(9)
94.73(10)
80.15(9)
104.59(9)
172.87(10)
86.63(10)
79.47(9)
75.53(10)
100.15(10)
−0.914(12)
2.050(3)
2.012(3)
2.058(2)
2.007(2)
2.150(2)
2.219(3)
2.111(3)
2.132(3)
1.301(4)
1.249(4)
1.388(4)
101.46(9)
88.92(9)
88.78(9)
87.31(10)
163.56(9)
168.86(9)
96.10(10)
94.11(9)
79.68(9)
105.16(9)
173.68(10)
88.69(9)
79.85(10)
76.17(9)
96.35(10)
−11.030(7)
2.138(3)
2.212(4)
2.104(4)
2.151(4)
1.308(5)
1.242(5)
C(21)−C(22)
1.392(6)
O(1)−Co(1)−O(3)
O(1)−Co(1)−O(4)
O(1)−Co(1)−N(1)
O(1)−Co(1)−N(2)
O(1)−Co(1)−N(3)
O(3)−Co(1)−N(1)
O(3)−Co(1)−N(2)
O(3)−Co(1)−N(3)
O(3)−Co(1)−O(4)
O(4)−Co(1)−N(1)
O(4)−Co(1)−N(2)
O(4)−Co(1)−N(3)
N(1)−Co(1)−N(2)
N(1)−Co(1)−N(3)
N(2)−Co(1)−N(3)
torsion angle at C(21)−C(22)−C(30)−C(31)
100.40(13)
90.25(13)
89.53(13)
86.86(14)
164.87(14)
169.33(14)
96.75(14)
94.11(13)
79.74(12)
104.29(14)
175.00(14)
88.16(14)
79.79(15)
76.31(13)
95.69(15)
−9.876(7)
was added dropwise to a dry CH3OH solution (1.0 mL) of Co(OAc)2·
4H2O (12.8 mg, 0.05 mmol) at room temperature. After stirring for 30
min, a CH3OH/CH2Cl2 (1:1) solution containing flavonol (11.9 mg,
0.05 mmol) and Me4NOH·5H2O (9.05 mg, 0.05 mmol) was added
2935
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
[CoIILNO2(fla)]·Et2O·0.5H2O (5) with electron-withdrawing
groups crystallize in the monoclinic system and P2(1)/c space
group.
dropwise to the above solution; then the mixture was stirred for 1 day
under N2. [CoIILR(fla)] was isolated as a reddish-yellow powder by
filtration. Wine red single crystals of [CoIILOMe(fla)] (1), [CoIILBr(¯a)]
(4), and [CoIILNO2(fla)] (5) suitable for X-ray crystallographic analysis
were obtained by slow diffusion of ether into the dichloromethane
solution of the complexes at −20 °C.
The Co−O(1) (benzoate of LR) bond distance is 2.035(2) Å
for 1, 2.050(3) Å for 4, and 2.058(2) Å for 5, which is close to
that of [CoIILH(fla)]·CH3OH (3) (2.0165(14) Å).14b The
bond lengths of Co−O(3) and Co−O(4) are 1.997(2) and
2.157(2) Å for 1, 2.012(3) and 2.138(3) Å for 4, and 2.007(2)
and 2.150(2) Å for 5, respectively, which are similar to those of
[CoIILH(fla)]·CH3OH (3) (1.9993(14) and 2.1797(14) Å)14b
and [CoII(6-Ph2TPA)(fla)]ClO4 (1.956(2) and 2.172(2) Å).13a
The difference [Δd(Co−O)] between Co−O(3) and Co−
O(4) is 0.16 Å for 1, 0.13 Å for 4, and 0.14 Å for 5, which are
also close to those of [CoIILH(fla)]·CH3OH (3) (0.18 Å)14b
and [CoII(6-Ph2TPA)(fla)]ClO4 (0.22 Å).13a The average Co−
N bond distance is 2.16 Å for both 1 and 4 and 2.15 Å for 5,
which are similar to those of [CoIILH(fla)]·CH3OH (3) (2.16
Å)14b and [CoII(6-Ph2TPA)(fla)]ClO4 (2.22 Å).13a Besides,
bond valence calculation22 (2.04 for 1, 2.05 for both 4 and 5)
also proves the metal center is the +2 state, which is in good
agreement with ESI-MS, EPR, and magnetic moment results
described below.
[CoIILOMe(fla)]·CH3OH (1). Yield: 23.7 mg, 69%. ESI-MS: m/z (pos.)
= 659.1 ([CoIILOMe(fla)]H+) (main peak), 681.3 ([CoIILOMe(¯a)]-
Na+). Anal. Calcd for C37H33CoN3O7 (690.59): C, 64.35; H, 4.82; N,
6.08. Found: C, 64.46; H, 4.93; N, 6.19. FT-IR (solid sample, KBr,
cm−1): 3485 (m), 1593 (s), 1558 (s), 1485 (m), 1412 (m), 1388 (w),
1217 (s), 1101 (m), 760 (m). FT-IR (solution sample, in ethanol,
cm−1): 3343 (m), 1614 (s), 1549 (s), 1487 (m), 1416 (m), 1381 (w),
1217 (s), 754 (m).
[CoIILMe(fla)]·CH3OH (2). Yield: 19.3 mg, 57%. ESI-MS: m/z (pos.)
= 643.3 ([CoIILMe(fla)]H+) (main peak), 665.3 ([CoIILMe(fla)]Na+).
Anal. Calcd for C37H33CoN3O6 (674.61): C, 65.87; H, 4.93; N, 6.23.
Found: C, 65.73; H, 5.02; N, 6.36. FT-IR (solid sample, KBr, cm−1):
3480 (m), 1610 (s), 1550 (s), 1490 (m), 1420 (s), 1360 (w), 1220 (s),
907 (m), 755 (s). FT-IR (solution sample, in ethanol, cm−1): 3346
(m), 1590 (s), 1557 (s), 1486 (m), 1409 (s), 1353 (w), 1214 (s), 906
(w), 757 (s).
[CoIILBr(fla)]·CH3OH·H2O (4). Yield: 29.3 mg, 77%. ESI-MS: m/z
(pos.) = 707.1 ([CoIILBr(fla)]H+), 729.1 ([CoIILBr(fla)]Na+) (main
peak). Anal. Calcd for C36H32BrCoN3O7 (757.49): C, 57.08; H, 4.26;
N, 5.55. Found: C, 57.21; H, 4.36; N, 5.69. FT-IR (solid sample, KBr,
cm−1): 3440 (m), 1610 (s), 1560 (s), 1480 (m), 1410 (s), 1350 (m),
1220 (s), 1150 (w), 906 (w), 754 (s). FT-IR (solution sample, in
ethanol, cm−1): 3357 (m), 1603 (s), 1553 (s), 1482 (s), 1410 (s),
1352 (m), 1215 (s), 1148 (w), 752 (s).
It should be noted that there are some structural and
conformational changes in the bound substrate flavonolate. The
distances of C(21)−O(3) (3-hydroxylate) (1.310(4) Å for 1,
1.308(5) Å for 4, and 1.301(4) Å for 5) are slightly contracted
relative to that of free flaH (1.357(3) Å).23 In contrast, the
distances of C(27)O(4) (4-carbonyl) (1.250(4) Å for 1,
1.242(5) Å for 4, and 1.249(4) Å for 5) and C(21)C(22)
(1.378(4) Å for 1, 1.392(6) Å for 4, and 1.388(4) Å for 5) are
slightly elongated relative to that of free flavonol (1.232(3) and
1.363(4) Å).23 However, all are close to the corresponding
distances in [CoIILH(fla)]·CH3OH (3) (1.313(2), 1.259(2),
and 1.376(2) Å).14b It should be noted that the elongated C
C bond may be helpful for its cleavage during the
dioxygenation reaction described below (Scheme 3). Interest-
ingly, the B ring of the bound flavonolate in each complex is
bent out from the plane defined by the rest of the flavonol
molecule. The torsion angle of C(21)−C(22)−C(30)−C(31)
is −0.914(12)° for 1, 6.275(44)° for 3,14b −9.876(7)° for 4,
and −11.030(7)° for 5, indicating that the C(22) atom
(corresponding to C(2) of the enzymatic substrate, Scheme 1)
in each complex has some sp3 character by pyramidalization, as
also observed in the ES adduct of the CuII-containing 2,3-QD
from A. japonicus.8 The plot of the torsion angle of flavonolate
vs Hammett constant σ is linear (R = 0.95) (Figure 2d),
indicating that the torsion angle of flavonolate is affected by the
electronic nature of the substituent group in the supporting
model ligand. Such a structural and conformational change
could stabilize the substrate radical formed during the reaction
with O2 and influence the electron distribution of the C(22)
atom through the “electron conduit” conferred by the benzoate
group, Co(II) ion, and the conjugated double bonds O(4)
C(27)−C(21)C(22) (corresponding to O(carbonyl)
C(4)−C(3)C(2) of the enzymatic substrate, Scheme 1;
Scheme 2 in pink).
[CoIILNO2(fla)]·Et2O·0.5H2O (5). Yield: 26.2 mg, 70%. ESI-MS: m/z
(pos.)
=
674.0 ([CoIILNO2(fla)]H+) (main peak), 696.0
([CoIILNO2(fla)]Na+). Anal. Calcd for C39H37CoN4O8.50 (756.66): C,
61.91; H, 4.93; N, 7.40. Found: C, 62.04; H, 5.01; N, 7.54. FT-IR
(solid sample, KBr, cm−1): 3480 (m), 1614 (w), 1551 (s), 1487 (m),
1418 (s), 1321 (m), 1217 (s), 908 (w), 754 (s), 669 (s). FT-IR
(solution sample, in ethanol, cm−1): 3334 (m), 1604 (w), 1558 (s),
1485 (m), 1414 (s), 1346 (s), 1215 (w), 1155 (w), 759 (s).
RESULTS AND DISCUSSION
■
Synthesis and Structural Characterization. Four new
model ligands bearing carboxylate group LRH (R = p-OMe, p-
Me, m-Br, and m-NO2) (Figure 1) were designed and
synthesized, and their corresponding cobalt(II) complexes
[CoIILR(fla)] (R = p-OMe (1), p-Me (2), m-Br (4), and m-
NO2 (5)) were synthesized by mixing 1 equiv of Co(OAc)2·
4H2O, ligand LRH, Me4NOH·5H2O, and flavonol in MeOH/
CH2Cl2 under N2. All complexes are relatively stable under air
in the solid state but react with O2 in solution (see below).
The single-crystal X-ray structures of the complexes
[CoIILOMe(fla)]·CH3OH (1), [CoIILBr(fla)]·CH3OH·H2O (4),
and [CoIILNO2(fla)]·Et2O·0.5H2O (5) are shown in Figure 2a,
2b, and 2c, respectively. Crystallographic data of the complexes
are summarized in Table 1. Selected bond distances and angles
are listed in Table 2. The structures of the complexes are similar
to each other and also similar to that of the nonsubstituted
analogue [CoIILH(fla)]·CH3OH (3),14b all bearing a distorted
octahedral metal center, which is coordinated by two oxygen
atoms from flavonolate (O(3), 3-hydroxylate; O(4), 4-carbon-
yl), one carboxylate oxygen O(1), and three nitrogen atoms
from LR. Interestingly, their crystal systems and space groups
vary with the substituent group of the ligands. [CoIILOMe(fla)]·
CH3OH (1) with an electron-donating group crystallizes in the
orthorhombic system and Pbca space group, which is similar to
that of the nonsubstituted analogue [CoIILH(fla)]·CH3OH
(3).14b In contrast, [CoIILBr(fla)]·CH3OH·H2O (4) and
Spectroscopic and Redox Properties of the Com-
plexes. Infrared Spectroscopy. In order to gain insights into
the solution structure and binding mode of the carboxylate
group in LR of the complexes, FT-IR spectra of both the
ethanol solution and the solid sample of the complexes were
recorded (Table 3, Supporting Information Table S1 and
Figure S1 for 2). The solution spectrum of each complex is
2936
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
(ε = 9.8 × 103 M−1 cm−1) for 1, 419 nm (ε = 1.1 × 104 M−1
cm−1) for 2, 415 nm (ε = 1.1 × 104 M−1 cm−1) for 4, and 413
Scheme 2. Products of the Reactions of Model Complexes
with O2
Figure 3. (a) UV−vis spectra of complexes [CoIILR(fla)] (∼ 0.1 mM
except 3 (7.2 × 10−2 mM) in DMF). (b) Plot of λmax of the complexes
vs Hammett constant σ.
similar to that of the corresponding solid sample, and spectra of
all complexes are similar to each other. Each spectrum shows a
CO stretching vibration ν(CO) of the coordinated
carbonyl group of flavonolate at ∼1560 cm−1, which is 40−
50 cm−1 red shifted relative to free flavonol (1602 cm−1),
resembling other flavonolate complexes reported so far.10−17 In
both states, each complex shows the asymmetric νas(COO−)
and symmetric νs(COO−) stretching frequencies of the
carboxylate group of LR at ∼1600 and ∼1410 cm−1,
respectively. They are also red shifted as compared with
those of free ligands LRH (∼1700 and ∼1600 cm−1,
respectively). The difference between them (Δν = νas(COO−)
− νs(COO−)) is in the range of 181−200 cm−1, rendering a
monodentate carboxylate binding mode in all complexes in
both states,24,25 being in good agreement with the X-ray
structures shown in Figure 2. It can be concluded that each
complex keeps its mononuclear structure in solution, which is
consistent with other spectroscopic results such as ESI-MS,
EPR, and solution magnetic moment described below.
nm (ε = 1.1 × 104 M−1 cm−1) for 5 (Figure 3a and Table 3).
λmax values match well with those of [CoIILH(fla)] (3) (417
nm),14b [CoII(fla)(L1)] (L1 = N-propanoate-N,N-bis(2-
pyridylmethyl)amine) (423 nm),14a and [(6-Ph2TPA)-
CoII(fla)]ClO4 (422 nm);13a all are blue shifted (∼40 nm)
with respect to those of free flavonolate (458 nm for
Me4Nfla13a and 465 nm for Kfla27). A similar phenomenon is
also observed in the enzymatic4,6−9 and other synthetic model
systems.10−17
Furthermore, the λmax values of the complexes are in the
order of [CoIILOMe(fla)] (1) > [CoIILMe(fla)] (2) >
[CoIILH(fla)] (3)14b > [CoIILBr(fla)] (4) > [CoIILNO2(fla)]
(5), and the plot of the λmax vs Hammett constant σ is linear (R
= 0.95) (Figure 3b). The largest λmax was observed in 1, bearing
the strongest electron-donating group (OMe), presumably due
to the best planarity and conjugation in flavonolate molecule of
1, as confirmed by its smallest torsion angle. These results
indicate that the λmax of the coordinated flavonolate is also
UV−vis Spectroscopy. In DMF, an intense absorption band
due to the π → π* transition of the coordinated flavonolate26 of
the complexes [CoIILR(fla)] appears at about 420 nm (422 nm
Table 3. Summary of Spectroscopic and CV Data for the Complexes
[CoIILOMe(fla)] (1)
[CoIILMe(fla)] (2)
[CoIILBr(fla)] (4)
1560
[CoIILNO2(fla)] (5)
1551
FT-IR
ν(CO)/cm−1
νas(CO2)/cm−1
νs(CO2)/cm−1
Δν/cm−1
1558
1550
1593
1412
181
1610
1420
190
1610
1410
200
1614
1418
196
UV−vis, λ/nm (ε/M−1 cm−1
)
277 (16 081)
277 (16 998)
277 (24 477)
269 (14 846)
422 (9778)
419 (10 510)
415 (10 660)
413 (10 510)
498 (71)
543 (152)
510 (182)
540 (102)
595 (37)
605 (78)
570 (86)
630 (65)
Dq/cm−1
B/cm−1
920
879
934
845
939
783
835
809
ESI-MS m/z
659.1 ([CoIILOMe(fla)]H+)
681.3 ([CoIILOMe(fla)]Na+)
4.76
643.3 ([CoIILMe(fla)]H+)
665.3 ([CoIILMe(fla)]Na+)
4.82
707.1 ([CoIILBr(fla)]H+)
729.1 ([CoIILBr(fla)]Na+)
5.01
674.1 ([CoIILNO2(fla)]H+)
696.0 ([CoIILNO2(fla)]Na+)
4.39
μeff/μB
CV/V
CoIII/II
Epa
−0.132
−0.206
−0.169
0.88
−0.123
−0.181
−0.152
1.06
−0.047
−0.133
−0.090
0.79
Epc
−0.080
E1/2
ipc/ipa
Epa
fla−/fla•
+0.315
+0.225
+0.270
0.82
+0.334
+0.259
+0.297
0.72
+0.409
+0.341
+0.375
0.51
+0.588
+0.231
+0.410
0.48
Epc
E1/2
ipc/ipa
2937
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
Figure 4. Cyclic voltammograms of [CoIILR(fla)] in DMF at room temperature. (a) [CoIILBr(fla)] (4) under slightly aerobic condition. (b) Plot of
Epc of the CoIII/II vs Hammett constant σ. (c) Plot of E1/2 of fla−/fla• vs Hammett constant σ.
affected by the electronic nature of the substituent group in the
ligands via the “electron conduit” mentioned above.
contribution of the orbital moment of the Co(II) ion. The
values are close to that of the nonsubstituted analogue
[CoIILH(fla)]·CH3OH (3) (4.13 μB)14b and also fall in the
range of the reported values (4.3−5.2 μB) for mononuclear
Co(II) complexes. These results indicate that all model
complexes remain in their oxidation state and mononuclear
structures in solution, which is consistent with the X-ray
structure, solution FT-IR, and ESI-MS results described above.
Cyclic Voltammetry. Cyclic voltammetry of [CoIILR(fla)]
was performed in DMF at room temperature. The redox
potentials (versus SCE) for all of the observed waves are listed
in Table 3. For each complex there is a quasi-reversible redox
couple at E1/2 = −0.169 V (ΔEp = 74 mV, ipc/ipa = 0.88) for
[CoIILOMe(fla)] (1), −0.152 V (ΔEp = 58 mV, ipc/ipa = 1.06)
The complexes also exhibit two weak absorption bands at
about 16 950 and 20 000 cm−1, respectively, which can be
assigned to the spin-allowed d−d transition bands of CoII (3d7)
in the distorted octahedral ligand field, ν2 (4T1g(P) ← 4T1g(F))
and ν3 (4A2g
←
4T1g(F)), respectively. The ligand-field
constants Dq (845 − 934 cm−1) and Racah parameter B
(783−939 cm−1) are calculated28 (Table 3) using the observed
bands ν2 and ν3. Both results are smaller than those of
[Co(H2O)6]2+ (970 cm−1) and free CoII ion (1030 cm−1).
ESI-MS Spectroscopy. The solution structures of the
complexes were also examined by ESI-MS. Each complex
shows two peak clusters that can be assigned to [CoIILR(¯a)]H+
(m/z (pos.) = 659.1 for 1, 643.3 for 2, 707.1 for 4, and 674.1
for 5) and [CoIILR(fla)]Na+ (m/z (pos.) = 681.3 for 1, 665.3
for 2, 729.1 for 4, and 696.0 for 5); however, no carboxylate- or
flavonolate-bridged dimer peak was observed. The m/z value
and isotope distribution pattern of each peak cluster match well
with the calculated value, indicating that each complex keeps its
oxidation state and mononuclear structure in solution, which is
in line with the X-ray structure and other spectroscopic results
such as solution FT-IR, EPR, and solution magnetic moment
results described below.
for [CoIILMe(fla)] (2), and −0.090 V (ΔEp = 86 mV, ipc/ipa
=
0.79) for [CoIILBr(fla)] (4) (Figure 4a), which can be assigned
to one-electron oxidation from CoII to CoIII. The E1/2 values are
comparable with the nonsubstituted analogue [CoIILH(fla)] (3)
(E1/2 = −0.062 V, ΔEp = 201 mV, ipc/ipa = 0.94).14b For
[CoIILNO2(fla)] (5), we only observed a reduction wave at
−0.080 V but did not observe any oxidation wave even under
O2. The Epc values of CoIII/II of the complexes show some
differences in the range from −0.206 to −0.080 V (over a range
of 126 mV) and are in the order of [CoIILOMe(fla)] (1) <
Magnetic Properties. The X-band EPR spectra of the
complexes (4.0 mM in 0.5 mL DMF) were examined under N2
at 100 K (Supporting Information Figure S2). Each complex
exhibits a very weak signal at about g = 5.8 and 2.2 (Supporting
Information Figure S2-A for 1), which is akin to that of the
high-spin distorted octahedral complex [CoIILH(fla)] (3) (g =
5.60 and 2.25),14b those of the native high-spin CoII-containing
enzyme and its ES adduct from Streptomyces sp. FLA (g = 6.0, A
= 95 G; g = 3.8 and 2.3),4c and that of the CoII-containing
enzyme from B. subtilis (g = 6.5).4b However, the signal of the
complexes is too weak at liquid nitrogen temperature to be
analyzed in detail.
[CoIILMe(fla)] (2) < [CoIILH(fla)] (3) (Epc = −0.162 V)14b
<
[CoIILBr(fla)] (4) < [CoIILNO2(fla)] (5), and the polt of Epc vs
Hammett constant σ is linear (R = 0.98) (Figure 4b). These
results indicate that the redox potentials of the CoII ion are
affected by the electronic nature of the substituent group of the
ligands, namely, the electron-donating group could increase the
electron density around the cobalt(II) center and make it more
easily oxidized.
When a DMF solution of [CoIILR(fla)] contacted slowly with
air, a new quasi-reversible redox couple appeared at E1/2
=
+0.270 V (ΔEp = 90 mV, ipc/ipa = 0.82) for [CoIILOMe(fla)] (1),
+0.297 V (ΔEp = 75 mV, ipc/ipa = 0.72) for [CoIILMe(fla)] (2),
+0.375 V (ΔEp = 68 mV, ipc/ipa = 0.51) for [CoIILBr(fla)] (4)
(Figure 4a), and +0.410 V (ΔEp = 357 mV, ipc/ipa = 0.48) for
[CoIILNO2(fla)] (5). The redox couple observed in the slow
oxygenation reaction could be tentatively assigned to one-
electron oxidation from fla− to fla•,14b which are comparable
1H NMR spectra of the complexes were examined in CDCl3
at ambient temperature (Supporting Information Figure S3 for
2). In each spectrum of the complex, we observed several
paramagnetically shifted weak and broad resonances over a
range from −2 to 10 ppm, which are similar to those of the
mononuclear complexes [CoIILH(fla)] (3)14b and [CoII(6-
Ph2TPA)(fla)]ClO4.13a
The solution magnetic susceptibility of the complexes was
also measured by NMR using Frei−Bernstein’s method.29 The
solution magnetic moment of each complex was calculated to
be 4.76 μB for 1, 4.82 μB for 2, 5.01 μB for 4, and 4.39 μB for 5
(Table 3). The values are larger than the calculated value using
the spin-only equation of the mononuclear high-spin CoII
center (3.87 μB). The difference can be explained by the
with the nonsubstituted analogue [CoIILH(fla)] (3) (E1/2
=
+0.306 V, ΔEp = 180 mV, ipc/ipa = 0.94).14b The E1/2 values of
the fla−/fla• redox couples show some differences in the range
from +0.270 to +0.410 V (over a range of 140 mV) and are in
the order of [CoIILOMe(fla)] (1) < [CoIILMe(fla)] (2) <
[CoIILH(fla)] (3)14b < [CoIILBr(fla)] (4) < [CoIILNO2(fla)] (5),
and the polt of E1/2 of fla−/fla• vs Hammett constant σ is linear
(R = 0.99) (Figure 4c). Again, the results can be explained by
the electronic nature of the substituent group of the ligand. The
2938
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
stronger the electron-donating group in the ligand, the more
electron rich the C(22) atom (via the “electron conduit”),
making the bound flavonolate more easily oxidized.
initial concentrations of both [CoIILR(fla)] (Figure 5b) and O2
(Figure 5c), so the rate law could be determined as
−d[CoIILR(fla)]/dt = k[CoIILR(fla)][O2]. The second-order
reaction rate constants k were thus determined as 9.60−49.4 ×
10−2 M−1 s−1 at 70 °C (ΔH‡ = 72−86 kJ mol−1, ΔS‡ = −15 to
−43 J mol−1 K−1) (Table 4, Supporting Information Table S3
and Figure S5). To date, two examples of Co(II)−flavonolate
complexes bearing dioxygenation reactivity, [CoII(fla)(L1)]
(2.09 × 10−2 M−1 s−1 at 80 °C, not structurally character-
ized)14a and [CoIILH(fla)] (3) (34.0 × 10−2 M−1 s−1 at 70
°C),14b were reported. Other reported ES model complexes
exhibit lower reactivity and require higher temperature (65−
130 °C).10−16,17b,c Compared with them, our model complexes
show higher reactivity at lower temperature (55−85 °C), which
can be attributed to the existing carboxylate group in the ligand.
In order to get insights into the carboxylate effects of the
model ligand, the noncarboxylate group analogue [CoIIL0(fla)-
(CH3O)] (L0 = benzyl-bis-pyridin-2-ylmethyl-amine) sup-
ported by the 3N chelating ligand was synthesized and
characterized (Supporting Information). The band of the π
→ π* transition of the coordinated flavonolate appears at 416
nm (ε = 11 614 M−1 cm−1) (Supporting Information Figure
S5a), which is a little blue shifted compared with that of the o-
carboxylate group substituted analogue [CoIILH(fla)] (3)14b
(417 nm). The reactivity of [CoIIL0(fla)(CH3O)] toward
dioxygen was also examined (Supporting Information Table
S3). The plots of the initial reaction rate vs the initial
concentrations of both [CoIIL0(fla)(CH3O)] and O2 are linear
(Supporting Information Figure S5b and S5c). Thus, the rate
law also can be described as −d[CoIIL0(fla)(CH3O)]/dt =
k[CoIIL0(fla)(CH3O)][O2]. The second-order reaction rate
constant k is 7.20 × 10−2 M−1 s−1 at 70 °C (ΔH‡ = 90.14 kJ
mol−1, ΔS‡ = −5.12 J mol−1 K−1) (Table 4 and Supporting
Information Figure S6), much smaller (about 5 times) than that
of the o-carboxylate group substituted analogue [CoIILH(fla)]
(3).14b These results clearly indicate that the presence of
carboxylate group in the ligand can enhance the reactivity via
lowering the redox potential of the coordinated flavonolate by
electron donation.
Degradation of the Complexes (Enzyme-Type Dioxyge-
nation Reactivity). Reactions of the model complexes
[CoIILR(fla)] with O2 in DMF at 70 °C result in two C−C
bond cleavage of the O-heterocycle to give the ring-opening
product o-benzoylsalicylic acid (HObs) (m/z (neg.: 241 (M −
H)− and 301 (M + OAc)−) (6.2−79%) (Supporting
Information Figure S4) and CO14b as the primary products.
HObs was then hydrolyzed with a small amount of water in the
solvent and amidated by the solvent DMF to give salicylic acid
(m/z (neg.) 137 (M − H)−) (13 − 76%), benzoic acid (m/z
(neg.) 121 (M − H)−) (19−89%) and 2-hydroxy-N,N-
dimethylbenzamide (m/z (pos.) 166 (M + H)+) (8.3−48%)
as the final products (Scheme 2, characterized by HPLC-MS
and HPLC, results seen in Supporting Information Table S2).
Conversions of the complexes are over 95%.
The dioxygenation reaction of the complexes was followed
by monitoring the decrease of the absorbance of the
coordinated flavonolate (π → π* transition) at the correspond-
ing λmax (Figure 5a for 1, Figure 3a, Table 3). Experimental
conditions and the observed initial rates are summarized in
Supporting Information Table S3. The initial reaction rate of
each complex exhibits a linear relationship with respect to the
Although the structures of the ES model complexes
[CoIILR(fla)] are similar, the dioxygenation reactivity of
[CoIILR(fla)] shows some differences and is in the order of
[CoIILOMe(fla)] (1) > [CoIILMe(fla)] (2) > [CoIILH(fla)] (3)14b
> [CoIILBr(fla)] (4) > [CoIILNO2(fla)] (5) (Figure 5), and the
Hammett plot is linear (ρ = −0.78) (Figure 5d). This order is
reverse of the order of the torsion angle of flavonolate and the
E1/2 of fla−/fla• described above, and the plot of k vs each of
them (Figure 5e (R = 0.99) and 5f (R = 0.99), respectively) is
linear. All these correlations provide evidence that the structure,
property, and dioxygenation reactivity of the model complexes
were influenced by the electronic nature of the substituent
group in the ligand via the “electron conduit”, and there is some
relationship among them, providing important insights into the
structure−property−reactivity relationship.
In fact, when the CH2Cl2/MeOH solutions of [CoIILR(fla)]
were exposed to air at room temperature, we initially observed
a [CoIIILR(fla)]+ peak cluster in ESI-MS spectra (Table 5) and
the π → π* transition band of flavonolate (∼ 0.1 mM)
disappeared after 1 day, indicating that our model complexes
are also active in CH2Cl2/MeOH. Such phenomenon inspires
us to check the solvent effects on the dioxygenation reactivity of
the complexes. Thus, the same dioxygenation reaction of
[CoIILOMe(fla)] but in CH2Cl2/MeOH (0.1 mM in 50 mL of
Figure 5. Dioxygenation of the [CoIILR(fla)] complexes at 70 °C
under O2. (a) Spectral change observed upon introduction of O2 gas
into a DMF solution of [CoIILOMe(fla)] (1) (1.0 × 10−4 M). (Inset)
Time course of the absorption changes of [CoIILOMe(fla)] (1) at 422
nm. (b) Plot of −d[CoIILR(fla)]/dt vs [CoIILR(fla)]0. (c) Plot of
−d[CoIILR(fla)]/dt vs [O2]0. (d) Hammett plot of the dioxygenation
of [CoIILR(fla)]. (e) Plot of k vs torsion angle of flavonolate. (f) Plot
of k vs E1/2 of fla−/fla•.
2939
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
Table 4. Kinetic Data (at 70 °C) and Activation Parameters of the Complexes
complexes
102k (M−1 s−1
)
ΔH‡ (kJ mol−1
)
ΔS‡ (J mol−1 K−1
)
Ea (kJ mol−1
)
[CoIILOMe(fla)] (1)
[CoIILMe(fla)] (2)
[CoIILBr(fla)] (4)
[CoIILNO2(fla)] (5)
[CoIIL0(fla)]
49.4 0.24
40.5 0.17
13.7 0.10
9.60 0.04
7.20 0.03
71.96
72.77
84.69
85.93
90.14
−41.87
−42.69
−16.19
−15.02
−5.120
74.75
75.58
87.53
88.85
93.08
Table 5. ESI-MS Results of the Complexes When Exposed to Air and after Reaction with O2
[CoIILOMe(fla)] (1)
[CoIILMe(fla)] (2)
[CoIILBr(fla)] (4)
[CoIILNO2(fla)] (5)
ESI-MS (m/z (pos.))
exposed to air
after reaction
658.3 ([CoIIILOMe(fla)]+)
542.2 [CoIIILOMe(ben)]+
642.3 ([CoIIILMe(fla)]+)
542.2 [CoIIILMe(sal)]+
706.1 ([CoIIILBr(fla)]+)
590.0 [CoIIILBr(ben)]+
673.1 ([CoIIILNO2(fla)]+)
573.1 [CoIIILNO2(sal)]+
1:1 CH2Cl2/MeOH, reflux for 14 h) and DMSO (1 mM in 10
mL of DMSO, at 130 °C for 48 h) was performed, and their
products were characterized by HPLC-MS and HPLC. In both
reactions the same enzyme-type dioxygenation ring-opening
products as that in DMF were observed (HObs (m/z (neg.)
241 (M − H)−) (29.6% and 28.5%), salicylic acid (m/z (neg.)
137 (M − H)−) (32.6% and 18.5%), benzoic acid (m/z (neg.)
121 (M − H)−) (34.1% and 17.4%), shown in Scheme 2,
Supporting Information Table S2 and Figure S7). Besides,
some unoxidized substrate flavonol (m/z (neg.) 237 (M)−;
33.1% and 48.5%, respectively) was also detected, and the total
conversions were 64% and 47%, respectively. The kinetic result
shows that the dioxygenation reaction of [CoIILOMe(fla)] in
DMSO is much slower (k about 3.0 × 10−2 M−1 s−1 at 100 °C,
Supporting Information Figure S8) than that in DMF (k 49.4 ×
10−2 M−1 s−1 at 70 °C). These results clearly indicate that the
same but much slower dioxygenation reaction occurs both in
CH2Cl2/MeOH and in DMSO. Thus, CH2Cl2/MeOH and
DMSO are not suitable solvents for the kinetic study.
Figure 6. Spectral change of [CoIILOMe(fla)] and [CoIIILOMe(¯a)]-
(OAc) (1.0 mM in DMF) in the presence of excess NBT under O2
(blue and red) and N2 (black) at room temperature.
(benzoate) or sal (salicylate), one of the oxidation products,
Table 5) and the weak CoII EPR signal disappeared
(Supporting Information Figure S2-B). These results indicate
that a fast radical−radical reaction exists between the generated
[CoIIILR(fla•)]+ (B) and O2•−, which leads to the dioxygenated
enzyme-type ring-opening free organic products (Scheme 2)
and coordinated products complex [CoIIILR(prod)]+ with
concomitant release of CO.
Dioxygenation Reaction Mechanism. Compared with
the free ion, the CoII ion in the CoII−flavonolate complex is
more easily oxidized to CoIII, which can be explained by the
electron donation contribution of the coordinated flavonolate
and ligand. In fact, when the [CoIILR(fla)] solution was exposed
to air, we initially observed a peak cluster [CoIIILR(fla)]+ (Table
5), indicating that the CoII ion was oxidized by O2 to CoIII to
form the CoIII−flavonolate species [CoIIILR(fla)]+ (A) at the
On the basis of the structure, spectroscopic, redox properties,
kinetic, and products analysis results, we proposed the
dioxygenation mechanism (Scheme 3) as follows. First,
[CoIILR(fla)] may react with one O2 molecule to produce
•−
[CoIIILR(fla)]+ (A) and O2 quickly. The following step is a
•−
first stage of the reaction, and the O2 was reduced to O2
rate-determining one-electron transfer from the bound fla− to
another O2 molecule to form [CoIIILR(fla•)]+ (B) and another
O2•−. The final step is a fast radical−radical reaction, forming
the dioxygenated products and releasing CO. The mechanisms
of the complexes are similar each other and also similar to that
of the nonsubstituted analogue [CoIILH(fla)] (3).14b
(superoxide radical), as proved by the reaction with nitroblue
tetrazolium (NBT2+). Significantly, when excess NBT2+ was
added to the O2-saturated [CoIILR(fla)] solution, a new
absorption band arose at 530 nm (Figure 6, blue line for 1).
It can be assigned to the monoformazan (MF+) due to
reduction of NBT2+ by free O2 generated in the reaction of
•−
[CoIILR(fla)] with O2.30
Although the oxidation state change of the metal center CoII
ion is not observed by EPR in the native Co-2,3-QD, there still
exists the possibility that one-electron transfer from the
substrate flavonol via the CoII ion to dioxygen is triggered by
dioxygen binding with a “transient change of the oxidation state
of the metal center”.4c Unfortunately, such an oxidation state
change could not be detected by the conventional experimental
method. In comparison, our mechanism is similar to that of
Co−2,3-QD except the oxidation state change in the first step.
Then [CoIIILR(fla)]+ (A) is proposed to react with another
O2 molecule to form a radical species [CoIIILR(fla•)]+ (B) and
•−
another O2 (Scheme 3, step 2). In order to prove this
postulation, we synthesized the CoIII−flavonolate analogue
[CoIIILOMe(fla)](OAc) independently, checked its reaction with
dioxygen (2.0 × 10−3 M in 2 mL of DMF, at 70 °C for 8h), and
observed O2•− by NBT2+ (Figure 6 red line) as well as the same
enzyme-type ring-opening products (salicylic acid 31.5%,
benzoic acid 16.1%, N,N-dimethylbenzamide 21.0%, and
HObs 62.3%, conversion 94%). Thus, step 2 in Scheme 3
was proved.
CONCLUSIONS
■
In summary, we designed and synthesized a series of CoII−
flavonolate complexes [CoIILR(fla)] (R = p-OMe (1), p-Me
(2), m-Br (4), and m-NO2 (5)) as structural and functional ES
After the dioxygenation reaction of [CoIILR(fla)], we
observed a peak cluster of [CoIIILR(prod)]+ (prod = ben
2940
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
Scheme 3. Proposed Dioxygenation Reaction Mechanism of Model Complexes [CoIILR(fla)]
(2) (a) Moratal, J. M.; Martínez-Ferrer, M. J.; Donaire, A.; Aznar, L. J.
Inorg. Biochem. 1992, 45, 65−71. (b) Ciancaglini, P.; Pizauro, J. M.;
Leone, F. A. J. Inorg. Biochem. 1995, 60, 155−162. (c) Ben-Meir, D.;
Spungin, A.; Ashkenazi, R.; Blumberg, S. Eur. J. Biochem. FEBS 1993,
212, 107−112. (d) Luchinat, C.; Monnanni, R.; Roelens, S.; Vallee, B.
L.; Auld, D. S. J. Inorg. Biochem. 1988, 32, 1−6.
models for the active site of Co−2,3-QD. Their structures,
spectroscopic futures, redox properties, and reactivity toward
dioxygen were investigated in detail. The reaction of each
model complex with O2 shows high enzyme-type dioxygenation
reactivity (oxidative O-heterocyclic ring opening of the bound
substrate flavonolate) at low temperature and first-order
dependence versus the initial concentrations of both complex
and O2. The reactivity is remarkably promoted by the
carboxylate group in the supporting model ligand by electron
donation. Besides, it also exhibits a substituent group
(3) (a) Mills, S. A.; Klinman, J. P. J. Am. Chem. Soc. 2000, 122, 9897−
9904. (b) Fielding, A. J.; Kovaleva, E. G.; Farquhar, E. R.; Lipscomb, J.
D.; Que, L., Jr. J. Biol. Inorg. Chem. 2011, 16, 341−355. (c) Dai, Y.;
Wensink, P. C.; Abeles, R. H. J. Biol. Chem. 1999, 274, 1193−1195.
(4) (a) Gopal, B.; Madan, L. L.; Betz, S. F.; Kossiakoff, A. A.
Biochemistry 2005, 44, 193−201. (b) Schaab, M. R.; Barney, B. M.;
Francisco, W. A. Biochemistry 2006, 45, 1009−1016. (c) Merkens, H.;
Kappl, R.; Jakob, R. P.; Schmid, F. X.; Fetzner, S. Biochemistry 2008,
47, 12185−12196.
dependent order of −OMe (1) > −Me (2) > −H (3)14b
>
−Br (4) > −NO2 (5), and the Hammett plot is linear (ρ =
−0.78). The structures, properties, and reactivity of the model
complexes are all influenced by the electronic nature of the
substituent group in the ligand via the benzoate, Co(II) ion,
and O(4)C(27)−C(21)C(22) “electron conduit”. A
stronger electron-donating group in the ligand may induce a
smaller torsion angle between the B and the C ring, larger λmax
of the π → π* transition, lower redox potential of flavonolate,
and finally higher reactivity of the model complexes. As the first
example of a series of structural and functional ES models of
Co-2,3-QD, our study focuses on both the electronic
substituent and the carboxylate effects of the supporting
model ligands on the enzyme-type dioxygenation reactivity and
will provide important insights into the structure−property−
reactivity relationship, the electronic and carboxylate effects on
the enzymatic reactivity, the role of metal ion, and the catalytic
role of Co−2,3-QD.
(5) Wollenweber, E. Flavonoids: Advances in Research; Harborne, J. B.,
Mabry, T. J., Eds.; Chapman & Hall: London, U.K., 1982.
(6) Fusetti, F.; Schroter, K. H.; Steiner, R. A.; van Noort, P. I.;
̈
Pijning, T.; Rozeboom, H. J.; Kalk, K. H.; Egmond, M. R.; Dijkstra, B.
W. Structure 2002, 10, 259−268.
(7) (a) Laura, B.; Shirley, A. F.; Victoria, J. J.; Stephen, B. FEBS Lett.
2004, 557, 45−48. (b) Barney, B. M.; Schaab, M. R.; LoBrutto, R.;
Francisco, W. A. Protein Expr. Purif. 2004, 35, 131−141.
(8) Steiner, R. A.; Kalk, K. H.; Dijkstra, B. W. Proc. Natl. Acad. Sci.
2002, 99, 16625−16630.
(9) (a) Steiner, R. A.; Kooter, I. M.; Dijkstra, B. W. Biochemistry
2002, 41, 7955−7962. (b) Antonczak, S.; Fiorucci, S.; Golebiowski, J.;
Cabrol-Bass, D. Phys. Chem. Chem. Phys. 2009, 11, 1491−1501.
(10) (a) Kaizer, J.; Balogh-Hergovich, E.; Czaun, M.; Casy, T.;
Speier, G. Coord. Chem. Rev. 2006, 222, 2222−2233. (b) Pap, J. S.;
Kaizer, J.; Speier, G. Coord. Chem. Rev. 2010, 254, 781−793. (c) Kaizer,
J.; Pap, J. S; Speier, G. Copper Dioxygenases. In Copper-Oxygen
Chemistry; Karlin, K. D., Shinobu, I., Eds.; John Wiley & Sons, Inc.:
ASSOCIATED CONTENT
* Supporting Information
■
S
́
Hoboken, NJ, 2011; pp 23−52. (d) Balogh-Hergovich, E.; Kaizer, J.;
Ligands synthesis, reaction products analysis data, kinetic data,
FT-IR, EPR, 1H NMR, LC-MS spectra, Eyring plot, and
crystallographic information (CIF) CCDC 957283 for 1,
957282 for 4, and CCDC 957281 for 5. This material is
available free of charge via the Internet at http://pubs.acs.org.
Speier, G.; Argay, G.; Parkanyi, L. J. Chem. Soc., Dalton Trans. 1999,
́
́
́
3847−3854. (e) Balogh-Hergovich, E.; Kaizer, J.; Pap, J.; Speier, G.;
Huttner, G.; Zsolnai, L. Eur. J. Inorg. Chem. 2002, 2287−2295.
(f) Balogh-Hergovich, E.; Kaizer, J.; Speier, G.; Fulop, V.; Par
́
́
kan
́
yi, L.
̈
̈
Inorg. Chem. 1999, 38, 3787−3795.
(11) (a) Kaizer, J.; Barath, G.; Pap, J. S.; Speier, G. Chem. Commun.
2007, 5235−5237. (b) Kaizer, J.; Pap, J. S.; Speier, G. On Biomimetics;
Pramatarova, L. D., Ed.; Intech Open Access Publisher: New York,
AUTHOR INFORMATION
Corresponding Author
*E-mail: yingjis@dlut.edu.cn.
■
2011; pp 29−42. (c) Barat
́ ́ ́
h, G.; Kaizer, J.; Speier, G.; Parkanyi, L.;
Kuzmann, E.; Vertes, A. Chem. Commun. 2009, 3630−3632.
́
Notes
(12) (a) Nishinaga, A.; Tojo, T.; Matsuura, T. J. Chem. Soc., Chem.
Commun. 1974, 896−897. (b) Nishinaga, A.; Numada, N.; Maruyama,
K. Tetrahedron Lett. 1989, 30, 2257−2258. (c) Hiller, W.; Nishinaga,
A.; Rieker, A. Z. Naturforsch., B: Chem. Sci. 1992, 47, 1185−1188.
(d) Nishinaga, A.; Kuwashige, T.; Tsutsui, T.; Mashino, T.; Maruyama,
K. J. Chem. Soc., Dalton Trans. 1994, 805−810.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the financial support of the National
Natural Science Foundation of China (Nos. 20641002,
20771020, 20811140328) and the Fundamental Research
Funds for the Central Universities (DUT12YQ04).
(13) (a) Grubel, K.; Rudzka, K.; Arif, A. M.; Klotz, K. L.; Halfen, J.
A.; Berreau, L. M. Inorg. Chem. 2010, 49, 82−96. (b) Grubel, K.;
Marts, A.; Greer, S. M.; Tierney, D. L.; Allpress, C. J.; Anderson, S. N.;
Laughlin, B. J.; Smith, R. C.; Arif, A. M.; Berreau, L. M. Eur. J. Inorg.
Chem. 2012, 4750−4757. (c) Grubel, K.; Laughlin, B. J.; Maltais, T. R.;
Smith, R. C.; Arif, A. M.; Berreau, L. M. Chem. Commun. 2011, 47,
10431−10433.
REFERENCES
■
(1) Roth, J. R.; Lawrence, J. G.; Bobik, T. A. Annu. Rev. Microbiol.
1996, 50, 137−181.
2941
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942

Inorganic Chemistry
Article
(14) (a) Matuz, A.; Giorgi, M.; Speier, G.; Kaizer, J. Polyhedron 2013,
63, 41−49. (b) Sun, Y.-J.; Huang, Q.-Q.; Tano, T.; Itoh, S. Inorg.
Chem. 2013, 52, 10936−10948.
(15) Barhacs, L.; Kaizer, J.; Speier, G. J. Mol. Catal. A. 2001, 172,
117−125.
́
(16) Balogh-Hergovich, E.; Kaizer, J.; Speier, G. J. Mol. Catal. A.
2003, 206, 83−87.
(17) (a) Aramice, Y. S.; Malkhasian, F. M. E.; Nikolovski, B.; Menon,
A.; Kucera, B. E.; Chavez, F. A. Inorg. Chem. 2007, 46, 2950−2952.
́ ́
(b) Kaizer, J.; Goger, S.; Speier, G.; Reglier, M.; Giorgi, M. Inorg.
Chem. Commun. 2006, 9, 251−254. (c) Volkman, J.; Nicholas, K. M.
Tetrahedron 2012, 68, 3368−3376.
(18) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth-Heinemann: Oxford, U.K., 1996.
(19) (a) Program for the Solution of Crystal Structures; University of
Gottingen: Gottingen, Germany, 1997. (b) Program for the Refinement
of Crystal Structures; University of Gottingen: Gottingen, Germany,
1997.
(20) Kruis, A. Landolt-Bornstein; Springer-Verlag: Berlin, 1976; Board
̈
4, Teil 4, p 269.
(21) Ram, G.; Sharaf, A. R. J. Ind. Chem. Soc. 1968, 45, 13−16.
(22) Brydon, P. Philos. Mag. Lett. 1999, 79, 383−387.
(23) Etter, M. C.; Urbanczyk-Lipkowska, Z.; Baer, S.; Barbara, P. F. J.
Mol. Struct. 1986, 144, 155−167.
(24) Christou, G.; Perlepes, S. P.; Libby, E.; Folting, K.; Huffman, J.
C.; Webb, R. J.; Hendrickson, D. N. Inorg. Chem. 1990, 29, 3657−
3666.
(25) Chadjistamatis, I.; Terzis, A.; Raptopoulou, C. P.; Perlepes, S. P.
Inorg. Chem. Commun. 2003, 6, 1365−1371.
(26) Jurd, L.; Geissman, T. A. J. Org. Chem. 1956, 21, 1395−1401.
́
(27) (a) Barhacs, L.; Kaizer, J.; Speier, G. J. Org. Chem. 2000, 65,
́
3449−3452. (b) Pap, J. S.; Matuz, A.; Barath, G.; Kripli, B.; Giorgi, M.;
Speier, G.; Kaizer, J. J. Inorg. Biochem. 2012, 108, 15−21.
(28) Lever, A. B. P. Inorgnic Electronic Spectroscopy, 2nd ed.; Elsevier:
Amsterdam, 1984.
(29) (a) Freit, K.; Bernstein, H. J. J. Chem. Phys. 1962, 37, 1891−
1892. (b) Kuchel, P. W.; Chapman, B. E.; Bubb, W. A.; Hansen, P. E.;
Dur, C. J. Concepts Magn. Reson. 2003, A 18, 56−71.
(30) Blelski, B. H. J.; Shiue, G. G.; Bajuk, S. J. Phys. Chem. 1980, 84,
830−833.
2942
dx.doi.org/10.1021/ic402695c | Inorg. Chem. 2014, 53, 2932−2942