Nonphotochemical base-catalyzed hydroxylation of 2,6-dichloroquinone by H2O2 occurs by a radical mechanism

Article abstract of DOI:10.1021/jp208536x

Kinetic and structural studies have shown that peroxidases are capable of the oxidation of 2,4,6-trichlorophenol (2,4,6-TCP) to 2,6-dichloro-1,4- benzoquinone (2,6-DCQ). Further reactions of 2,6-DCQ in the presence of H ₂O₂ and OH^- yield 2,6-dichloro-3-hydroxy-1,4- benzoquinone (2,6-DCQOH). The reactions of 2,6-DCQ have been monitored spectroscopically [UV-visible and electron spin resonance (ESR)] and chromatographically. The hydroxylation product, 2,6-DCQOH, has been observed by UV-visible and characterized structurally by ¹H and ¹³C NMR spectroscopy. The results are consistent with a nonphotochemical base-catalyzed oxidation of 2,6-DCQ at pH > 7. Because H₂O ₂ is present in peroxidase reaction mixtures, there is also a potential role for the hydrogen peroxide anion (HOO^-). However, in agreement with previous work, we observe that the nonphotochemical epoxidation by H₂O₂ at pH < 7 is immeasurably slow. Both room-temperature ESR and rapid-freeze-quench ESR methods were used to establish that the dominant nonphotochemical mechanism involves formation of a semiquinone radical (base -catalyzed pathway), rather than epoxidation (direct attack by H₂O₂ at low pH). Analysis of the kinetics using an Arrhenius model permits determination of the activation energy of hydroxylation (E_a = 36 kJ/mol), which is significantly lower than the activation energy of the peroxidase-catalyzed oxidation of 2,4,6-TCP (E_a = 56 kJ/mol). However, the reaction is second order in both 2,6-DCQ and OH ^- so that its rate becomes significant above 25 °C due to the increased rate of formation of 2,6-DCQ that feeds the second-order process. The peroxidase used in this study is the dehaloperoxidase-hemoglobin (DHP A) from Amphitrite ornata, which is used to study the effect of a catalyst on the reactions. The control experiments and precedents in studies of other peroxidases lead to the conclusion that hydroxylation will be observed following any process that leads to the formation of the 2,6-DCQ at pH > 7, regardless of the catalyst used in the 2,4,6-TCP oxidation reaction.

Full text of DOI:10.1021/jp208536x

ARTICLE
pubs.acs.org/JPCB
Nonphotochemical Base-Catalyzed Hydroxylation of
,6-Dichloroquinone by H O Occurs by a Radical Mechanism
2
2
2
,†
†
§
§
§
Stefan Franzen,* Koroush Sasan, Bradley E. Sturgeon, Blake J. Lyon, Benjamin J. Battenburg,
‡
†
†
Hanna Gracz, Rania Dumariah, and Reza Ghiladi
†
‡
Department of Chemistry, and Department of Molecular and Structural Biochemistry, North Carolina State University, Raleigh,
North Carolina 27695, United States
§
Department of Chemistry, Monmouth College, Monmouth, Illinois 61462, United States
S Supporting Information
b
ABSTRACT:
Kinetic and structural studies have shown that peroxidases are capable of the oxidation of 2,4,6-trichlorophenol (2,4,6-TCP) to 2,
ꢀ
6
3
-dichloro-1,4-benzoquinone (2,6-DCQ). Further reactions of 2,6-DCQ in the presence of H O and OH yield 2,6-dichloro-
2
2
-hydroxy-1,4-benzoquinone (2,6-DCQOH). The reactions of 2,6-DCQ have been monitored spectroscopically [UVꢀvisible and
electron spin resonance (ESR)] and chromatographically. The hydroxylation product, 2,6-DCQOH, has been observed by
1
13
UVꢀvisible and characterized structurally by H and C NMR spectroscopy. The results are consistent with a nonphotochemical
base-catalyzed oxidation of 2,6-DCQ at pH > 7. Because H O is present in peroxidase reaction mixtures, there is also a potential role
2
2
ꢀ
for the hydrogen peroxide anion (HOO ). However, in agreement with previous work, we observe that the nonphotochemical
epoxidation by H O at pH < 7 is immeasurably slow. Both room-temperature ESR and rapid-freeze-quench ESR methods were
2
2
used to establish that the dominant nonphotochemical mechanism involves formation of a semiquinone radical (base -catalyzed
pathway), rather than epoxidation (direct attack by H O at low pH). Analysis of the kinetics using an Arrhenius model permits
2
2
determination of the activation energy of hydroxylation (E = 36 kJ/mol), which is significantly lower than the activation energy of
a
the peroxidase-catalyzed oxidation of 2,4,6-TCP (E = 56 kJ/mol). However, the reaction is second order in both 2,6-DCQ and
a
ꢀ
OH so that its rate becomes significant above 25 °C due to the increased rate of formation of 2,6-DCQ that feeds the second-order
process. The peroxidase used in this study is the dehaloperoxidase-hemoglobin (DHP A) from Amphitrite ornata, which is used to
study the effect of a catalyst on the reactions. The control experiments and precedents in studies of other peroxidases lead to the
conclusion that hydroxylation will be observed following any process that leads to the formation of the 2,6-DCQ at pH > 7,
regardless of the catalyst used in the 2,4,6-TCP oxidation reaction.
’
INTRODUCTION
Although the native substrate of DHP A is believed to be 2,4,
-tribromophenol (2,4,6-TBP), most kinetic studies have been
conducted using the more soluble2,4,6-TCP. The reactivity of
these two 2,4,6-TXPs is similar. Trichlorophenols are of interest
because they are important targets for bioremediation efforts.
6
Several peroxidases have been observed to catalyze the oxida-
3ꢀ6
tion of 2,4,6-trihalophenols (2,4,6-TXPs) to the corresponding
1
ꢀ4
2
,6-dihalo-1,4-benzoquinones (2,6-DXQs).
The haloqui-
nones, 2,6-DXQs, are quite reactive and can undergo further
reactions, which may be both spontaneous and catalyzed in
living organisms. We have studied the oxidation of 2,4,
2
,4,6-TCP is of particular importance because it is ranked number
6
-trichlorophenol (2,4,6-TCP) in the dehaloperoxidase-hemo-
Received: September 5, 2011
Revised: November 12, 2011
Published: December 07, 2011
globins, DHP A andDHPB, from Amphitrite ornata, which serve as
a model system for the study of the fate of the product 2,6-DCQ.
r 2011 American Chemical Society
1666
dx.doi.org/10.1021/jp208536x
|J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
1 among the top pollutants registered by the United State
ARTICLE
2
’ EXPERIMENTAL SECTION
Environmental Protection Agency (Appendix A to part 40 of the
CFR 423 of EPA regulation).
Sample Preparation. His-tagged DHP A was overexpressed
and purified as previously described. The concentration of the
3
Although much of the chemistry described in this study is
1
7
protein was determined using the Soret band at 406 nm with a
general to peroxidases and even heme proteins in general, we
discovered the reactions as part of our research into the structure
and function of DHP A. It is therefore useful to briefly introduce
the current state of knowledge of DHP A to put the further
reactivity of the enzymatic product, 2,6-DCQ, in context. Re-
sonance Raman and NMR studies have shown that 2,4,6-TXPs
ꢀ
1
ꢀ1 24
molar absorptivity of 116 400 M cm
purity was that the A406/A280 ratio was greater than 4. Purified
XHis-DHP A was oxidized in the presence of K [Fe(CN) ].
.
The criterion for
6
3
6
Size exclusion chromatography (Sephadex G-25) was used to
separate excess K [Fe(CN) ], and the freshly oxidized protein
3
6
8
ꢀ14
was further purified by ion-exchange chromatography (CM-52)
prior to each experiment. For the kinetic assays, the elution buffer
bind to an external site on DHP, while X-ray crystallography,
1
5,16
17,18
19
NMR,
-halophenols (4-XPs) bind in a well-defined hydrophobic cavity
in the distal pocket above the heme. We have recently shown that
FTIR,
and Raman have demonstrated that
was 150 mM KH PO buffer at pH 7.0. 2,4,6-Trichlorophenol
4
2
4
from Acros Organics was dissolved in 150 mM KH PO , pH 7
2
4
12
buffer with a final concentration of 3 mM and stored at 4 °C until
4
-XPs are not substrates, as originally proposed, but rather are
inhibitors of DHP. On the basis of these studies, it is clear that
the mechanism of DHP A and B is similar to the established
1
4
use. The H O solution was diluted from 30% reagent grade
2 2
H O solution (Fisher Scientific) with 150 mM KH PO , pH 7
2
2
2
4
2
0
buffer to a stock concentration of 17.6 mM. The H O solution
peroxidase mechanism in that the substrate is oxidized at
2 2
was prepared freshly before use and stored at 4 °C during the
course of a series of experiments.
Kinetic Assays. The kinetic assays were conducted in a 1 cm
path length quartz cuvette with a total volume of 1.5 mL. The
the surface of the enzyme where it can readily diffuse into
2
1,22
solution.
Thus, it has been suggested that 2,4,6-TXP is
oxidized by two one-electron steps rather than a two-electron
6
,23ꢀ27
concerted process.
More recently, we have also pro-
final concentration of ferric DHP A in the cuvette was [E] ≈
T
posed that TXP is being oxidized by a single 1-electron
•
2 μM, and the 2,4,6-TCP concentration was 300 μM for the kinetic
experiments. Spectra were obtained using a photodiode array
UVꢀvisible spectrometer (Agilent 8453) equipped with a Peltier
temperature controller using benchtop mixing of the reagents.
To reach thermal equilibrium, DHP A and 2,4,6-TCP (150 mM
KH PO , pH 7 buffer) were allowed to incubate for 5 min in the
peroxidase oxidation of TCP to form TCP , followed by
enzyme-independent radical reactions leading to the 2-electron
2
3
oxidized product.
In the course of enzymatic studies of DHP A, it was observed
that a slower secondary process occurred, which consumed the
product 2,6-DCQ. Above 20 °C, the DHP A-catalyzed oxidation
2
4
thermostatted cuvette. The H O solution concentration was 75
2
2
of 2,4,6-TCP shows a maximum in 2,6-DCQ production (λ
=
max
times that of DHP A. H O was added to the cuvette within 1 s of
2
2
2
73 nm), which then decreases with time as 2,6-DCQ is
24
initiation of data collection.
transformed into one or more secondary products that have
not yet been characterized. We noted a similar behavior in the
4
The data were measured over the wavelength range from 200
to 700 nm with a time sampling rate of 3.1 s. The 2,6-DCQ
absorption band (λmax = 273 nm) was monitored as the primary
product, with 2,6-DCQOH (524 nm) as the secondary product.
The data were analyzed by IgorPro5.0 by fitting the initial rate,
which comprised the first 30 s of data collection. The analysis of
the MichaelisꢀMenten kinetics of DHP A has been presented
mutant Y38F, which has been studied in detail in the context of
4
,25,26,28
radical pathways in the DHP A and B.
tions are similar to the previously reported hydroxylation of
These observa-
29
quinone products observed using a variety of iron catalysts, in
3
0,31
7,32
peroxidases,
not required for the oxidation of 2,4,6-TCP as shown by the
and in heme proteins, more generally. Iron is
24
elsewhere.
3
3
oxidation reactions in Ralstonia eutropha. Previous reports have
HPLC/UVꢀVis Combined Data Collection. Samples for com-
bined HPLC and UVꢀvisible spectroscopic data collection were
prepared by dissolving 100 μM 2,6-DCQ in 1 mL of ethanol and
ꢀ
ꢀ
shown that both OH and HOO may play a role in the
formation of the semiquinone, which then further reacts to
form hydroxylated quinones and other products.
3
0,34
Under-
then adding 100 mM potassium phosphate (KP ) buffer to a total
i
standing secondary product formation in DHP-catalyzed oxi-
dations is important both in a biological context and in
bioremediation applications. For example, it is believed that
volume of 100 mL, both with and without 200 μM H O .
2
2
A sample for HPLC analysis was transferred immediately to
the auto sampler vial at the same time the sample was placed in
the UVꢀvisible spectrometer. When handled in this manner, the
sample was injected onto a C-18 HPLC column ∼2 min after
preparation. HPLC conditions: mobile phase A = acetonitrile,
mobile phase B = 0.1% TFA in pure water and a flow rate of
35,36
DHP A acts to oxidize 2,4,6-TBP in the coelom of A. ornata,
so the further reactivity of 2,6-DCQ determines the fate of these
37,38
molecules in benthic ecosystems.
Moreover, the secondary
reactions of 2,4,6-TBP may have an impact on the cellular meta-
3
9,40
bolism in A. ornata.
Haloquinone degradation pathways may
1
mL/min. A linear gradient from 0 to 30% B was applied for
involve direct oxidation, for example, epoxidation or radical path-
ways. Herein, we have applied kinetic, structural, and computa-
tional methods to quantitatively understand the secondary reac-
tions, with a specific emphasis on the use of electron spin resonance
1
0 min, followed by a linear gradient from 30 to 100% B for an
additional 10 min. UV absorbance measurement was at 273 nm
(λmax for 2,6-DCQ).
Room-TemperatureESRSpectroscopy. X-band (∼9.8 GHz)
ESR spectroscopy was performed using a Bruker EMX spectro-
meter (Billerica, MA) equipped with a high sensitivity cavity. The
samples were prepared by dissolving 0.5 mM 2,6-DCQ with
0.5 mM H O at pH 7.4 (100 mM KP buffer) introduced into the
(
ESR) to determine the role of radicals in the reaction pathway.
The use of both room-temperature and rapid-freeze-quench
RFQ ) ESR spectroscopic methods provides unique electron
(
structural and kinetic information that elucidates the reaction
pathways of 2,6-DCQ with greater accuracy than was previously
possible.
2
2
i
41
cavity using the Bruker Aqua-X sample holder. ESR simula-
tions were performed using the WINSIM program. Simulations
42
1
667
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
Figure 1. UVꢀvisible spectrophotometric measurement of the DHP-catalyzed oxidation of 2,4,6-TCP. Initial concentrations were [DHP A] ≈ 2 μM,
buffer (pH 7). (A) 2,4,6-TCP (312 nm) was converted
[
2,4,6-TCP] ≈ 300 μM, [H
2
O
2
] ≈ 150 μM, and the assay was conducted at 25 °C in 150 mM KP
i
to the oxidation product, 2,6-DCQ (273 nm). (B) The formation of 2,6-DCQOH (524 nm). The proposed reaction corresponding to the change each of
the three absorption bands is shown beside the spectra.
shown use the best-fit parameters as determined by the WINSIM
program.
Preparation of Reaction Mixtures by Rapid-Freeze-Quench
NMR Spectroscopy. All of the pulsed field NMR experiments
were carried out on a Bruker AVANCE 500 MHz spectrometer
(1996) with an Oxford narrow bore magnet (1989), HP XW
4200 host workstation, Topspin software version 1.3. A 5 mm ID
(
RFQ) Methods. DCQ and H O solutions were freshly prepared.
2 2
109
31
Prior to each experiment, a stock solution of 20 mM 2,6-DCQ was
prepared in ethanol under minimal light conditions. The concen-
tration of H O stock solution was determined by measuring its
1H/BB ( Agꢀ P) triple-axis gradient probe (ID500-5EB,
Nalorac Cryogenic Corp.) was used for all measurements. The
13
2
2
NMR probe was tuned to the1 C frequency of 125.75 MHz in
the 500 MHz spectrometer ( H frequency was 500.128 MHz).
The products of hydroxylation of 2,6-DCQ were studied using a
ꢀ1
ꢀ1
absorbance at 240 nm (ε240 = 43.6 M cm ). The stock so-
lutions were then diluted to the appropriate premixing concentra-
tion in 100 mM potassium phosphate (0.333 mM H O , variable
1
1
combination of homonuclear 2D COSY ( Hꢀ H) and hetero-
2
2
1
13
pH) or ethanol (1.0 mM 2,6-DCQ) and immediately loaded into
the stopped-flow/freeze quench apparatus. Rapid-freeze-quench
experiments were performed with a BioLogic SFM 400 freeze-
quench apparatus by mixing a 0.333 mM H O solution (initial
nuclear Hꢀ C correlated NMR methods. Both 2D hetero-
nuclear single quantum coherence (HSQC) and 2D hetero-
nuclear multiple bond correlation (HMBC) experiments were
applied to aid in the assignments. The HSQC experiment mea-
2
2
1
13
concentration) in 100 mM potassium phosphate (pH 5, 7.4, and 9)
with 1.0 mM 2,6-DCQ solution in ethanol at 20 °C using a 3:1
mixing ratio. Reaction times were as follows: pH 5 and pH 7.4, 100
ms, 1 s, and 10 s; pH 9, 20 ms, 1 s, 10 s. For comparative purposes,
solutionswerealso preparedby hand mixingand werequenched at
sures Hꢀ C one bond coupling, and the HMBC experiment
1
13
measures Hꢀ C long-range couplings. Because of the low
solubility of 2,6-DCQ in H O, the NMR samples were pre-
2
pared by dissolving 100 mM 2,6-DCQ in ∼0.6 mL of DMSO
followed by the transfer of the solution to a 5 mm NMR tube
for analysis. Subsequently, ∼0.2 mL of H O was added to make
16 s (pH 7.4) and 13 s (pH 9). A standard 4 mm outside diameter
2 2
a solution with a final concentration of 300 mM H O and
quartz ESR tube was connected to a Teflon funnel, and both the tube
and the funnel were completely immersed in an isopentane bath at
2
2
1
00 mM 2,6-DCQ.
DFT Calculations. The generalized gradient approximation of
ꢀ
110 °C. The reactions were quenched by spraying the mixtures
43
PerdewꢀBurkeꢀErnzerhof density functional theory was em-
into the cold isopentane, and the frozen material so obtained was
packed at the bottom of the quartz tube using a packing rod equip-
ped with a Teflon plunger. Samples were then transferred to a liquid
nitrogen storage dewar until they were analyzed.
ployed in the ground-state calculations as implemented by the
3,44ꢀ47
electronic structure package DMol
using a double numeric
basis set with one polarization function. Optimized geometries
were obtained using the conjugate gradient method constrained
X-Band ESR Spectroscopy for RFQ Samples. ESR spectra
were recorded with an X-band (9 GHz) Varian E-9 ESR spectro-
meter (Varian, El Palo, CA). Standard 3 mm ꢁ 4 mm quartz ESR
tubes were used. The temperature of the samples was kept at
ꢀ
6
to an energy difference of 10 hartrees. Ground-state calcula-
tions implemented both the thermal treatment of electron
48
occupancy with an electronic temperature of 0.02 Ha and the
3
COSMO module employed by DMol (conductor-like screening
1
90 K for the duration of the data acquisition with a constant
49
model), ε = 78.4.
flow of nitrogen gas cooled with liquid nitrogen. The flow of the
cooled nitrogen gas was adjusted using a Varian temperature
controller. The typical spectrometer settings were as follows:
field sweep, 100 G; scan rate, 3.33 G/s; modulation frequency,
’
RESULTS
1
2
00 kHz; modulation amplitude, 2.0 G; and microwave power,
mW. The exact resonant frequency of each ESR experiment
Observation of Spontaneous Reactivity of 2,6-DCQ. The
H O /DHP-catalyzed major oxidation product of 2,4,6-TCP is
2
2
was measured by an EIP-78 (PhaseMatrix, San Jose, CA) in-line
microwave frequency counter and is indicated in the figure
captions. Typically, 150ꢀ200 individual scans were averaged to
achieve sufficient signal-to-noise.
2,6-DCQ. Figure 1A shows that the substrate, 2,4,6-TCP
(312 nm), was catalytically converted to the product, 2,6-DCQ
(273 nm), in the presence of DHP A, following the addition of
H O . However, in these kinetic data, there was also an increase
2
2
1
668
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
hydroxylation of 2,6-DCQ in the absence of enzyme. Figure S3
shows that the rate of hydroxylation of 2,6-DCQ was essentially
the same in the absence of DHP A as in its presence. Evidence for
the spontaneous formation of 2,6-DCQOH is shown in Figure 3,
which shows a kinetic trace in the absence of DHP A, but in the
presence of 150 μM H O at pH 7.
2
2
Determination of the Structure of the Hydroxylation
Product of 2,6-DCQ. We investigated the hypothesis that 2,
6
-DCQ is hydroxylated to 2,6-DCQOH by H O in aqueous so-
2 2
1 13 1 13
lution using 2D Hꢀ C HSQC and 2D Hꢀ C HMBC NMR
spectroscopic experiments. The aqueous solubility of 2,6-DCQ is
too low for NMR studies in D O. Consequently, DMSO-d was
2
6
selected as the solvent. The HMBC experimental data shown in
Figure 4A indicated that there are four correlation peaks between
protons and carbon atoms in the 2,6-DCQ solution prior to
treatment with H O . These data lead to the conclusion that
there are four distinguishable carbons in 2,6-DCQ as expected
Figure 2. Kinetic trace of 2,4,6-TCP (black trace, 312 nm), 2,6-DCQ
2 2
(
blue trace, 273 nm), and 2,6-DCQOH (red trace, 524 nm) during the
oxidation of 2,4,6-TCP with H catalyzed by DHP at 20 °C. Initial
concentrations were [DHP A] ≈ 2 μM, [2,4,6-TCP] ≈ 300 μM,
] ≈ 150 μM, and the assay was conducted at 20 °C in 150 mM
buffer (pH 7). The red trace is baseline subtracted for clarity and
2
O
2
(Figure 4A). After the addition of ∼150 μM H O to the 2,
2
2
6-DCQ solution, six cross-peaks were observed in 2D HMBC
spectra, indicating the formation of six distinguishable carbons
2 2
[H O
KP
i
(
Figure 4B). The six peaks can be assigned to the six carbons of
represents ΔA at 524 nm.
2
1
,6-DCQOH: δ 177 (C dO), δ 168 (C dO), δ 152 (C ꢀOH), δ
1
4
2
22 (C ꢀCl), δ 142 (C ꢀCl), and δ 116 (C ꢀH) ppm. There is
3 5 6
13
excellent agreement between the predicted C chemical shifts
53
for 2,6-DCQOH and our experimental NMR data (see Table S1).
Figure S2 of the Supporting Information shows 2D HSQC data,
which exhibit a change in correlation between the hydrogen
atoms and carbon atoms 3 and 5 after addition of H O .
2
2
Hydroxylation of 2,6-DCQ Is a Temperature-Dependent
Reaction. The activation energy of the rate of oxidation of 2,4,
6
-TCP by H O was determined by the measurement of the
2 2
24
temperature dependence of the rate constant. Figure 5 shows
that the rate of spontaneous hydroxylation of 2,6-DCQ to 2,
6
-DCQOH increases with temperature as well. The effect of this
spontaneous hydroxylation on the concentration of the 2,6-DCQ
product becomes evident above 25 °C at pH 7 for w.t. DHP A.
Figure 6 provides an explanation for this observation. The spon-
taneous rate of formation of 2,6-DCQOH increases with tem-
perature more rapidly than does 2,6-DCQ formation. The kinetics
of the 2,6-DCQ product has a biexponential form above 20 °C,
which results from an increase in ΔA due to 2,6-DCQ formation
followed bya decreaseinΔA dueto itsconversionto2,6-DCQOH
observed for the 273 nm data in Figure 6. This biexponential
behavior and the fit to the data are presented in Figure S4 of the
Supporting Information. The model function, P (t) = A [1 ꢀ
Figure 3. UVꢀvisible spectroscopic measurement of the hydroxylation
of 2,6-DCQ (nonenzymatic). The absorbance of 2,6-DCQOH has a
maximum at 524 nm. Initial concentrations were [2,6-DCQ] ≈ 150 μM,
[
H
2
O
2
] ≈ 150 μM, and the assay was conducted at 25 °C in 100 mM
i
KP buffer (pH 7).
in absorbance at 524 nm that has not been investigated in pre-
vious studies of DHP (Figure 1B). The observed 524 nm band
grows more slowly than the product peak (273 nm) at 20 °C.
The 524 nm band observed in Figure 1B has been pre-
viously assigned to a hydroxylated quinone. Specifically, 2,6-DCQ
has been observed to convert to 2,6-dichloro-3-hydroxy-1,
1
1
exp(ꢀk t)] ꢀ A exp(ꢀk t), has a rapid process, k , that leads to
1
2
2
1
34,50ꢀ52
2,6-DCQ formation, and a slower hydroxylation process, k , with
2
4
-benzoquinone (2,6-DCQOH) in the presence of H O
2 2
ꢀ2 ꢀ1
ꢀ3 ꢀ1
values k = 1.8 ꢁ 10
s
and k = 3.3 ꢁ 10 s , respectively.
1
2
in a photochemical, temperature-, and pH-dependent process. A fit
These values are comparable to the rate constants obtained in-
to the data in Figure 2 using the method of initial rates gives rates
ꢀ7
ꢀ7
ꢀ1
dependently from a fit to substrate decay at 312 nm, S(t) = A
3
of 9.1 ꢁ 10 and 1.2 ꢁ 10 M s for the formation of 2,
-DCQ and 2,6-DCQOH, respectively. The analysis is shown in
Figure S1 in the Supporting Information. According to previous
exp(ꢀk t), and growth of the secondary product, P , at 524 nm,
1
2
6
P (t) = A (1 ꢀ exp(ꢀk t)), as shown in Figures S4 and S5 in the
2
4
2
SupportingInformation and are confirmed bythe methodofinitial
rates in Figure S1 (see analysis in the Supporting Information).
Thus, the present study confirms that the hydroxylation of 2,
5
0
reports of similar secondary oxidations in peroxidases, we
hypothesize that the observed process consists of the hydroxyla-
tion of 2,6-DCQ to 2,6-DCQOH. The extinction coefficient of
6
-DCQ is about 8 times slower than its catalytic formation under the
standard conditions used in enzymatic assays ([DHP A] ≈ 2.4 μM,
H O ] < 1.5 mM, [2,4,6-TCP] < 1.5 mM).The fact that the
ꢀ
1
ꢀ1
2
,6-DCQOH is 2530 M cm at 524 nm, which is significantly
smaller than the extinction coefficient of 2,6-DCQ (13 200
[
2
2
ꢀ1
ꢀ1
50
M
cm at 276 nm). We have observed recently that 2,
formation of 2,6-DCQOH is ∼8 times slower than 2,6-DCQ
formation validates the assumption made in the determination
of activation parameters for 2,6-DCQ formation in a previous
6
-DCQ appears stable below 10 °C, but further reactions take place
24
at temperatures above 10 °C. The absorption band of the sec-
ondary product, 2,6-DCQOH (524 nm), was used to study the
24
study. The exponential model was necessary because of the
1
669
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
0
Figure 4. HMBC data obtained on 2,6-DCQ and 2,6-DCQOH in DMSO-d
6
. (A) The correlations between the identical H atoms (on C3 and C3 ) and
all C atoms of 2,6-DCQ are shown. (B) Data obtained on 2,6-DCQOH indicate six different C atoms that are correlated with the single H atom on C-6.
Figure 6. Kinetic trace of 2,4,6-TCP (black trace, 312 nm), 2,6-DCQ
Figure 5. Kinetic traces of 2,6-DCQOH formed during the reaction of
(
blue trace, 273 nm), and 2,6-DCQOH (red trace, 524 nm) during the
2
2 2
,4,6-TCP with H O as catalyzed by DHP A at 10, 15, 20, 25, and
oxidation of 2,4,6-TCP with H
2 2
O as catalyzed by DHP A. Initial
3
0 °C. Initial concentrations were [DHP A] ≈ 2 μM, [2,4,6-TCP] ≈
concentrations were [DHP A] ≈ 2 μM, [2,4,6-TCP] ≈ 300 μM,
300 μM, [H O ] ≈ 150 μM, and the assay was conducted in 150 mM
2
2
[
H O ] ≈ 150 μM, and the assay was conducted at 25 °C in 150 mM
2
2
i
KP buffer at pH 7.
KP buffer (pH 7). The red trace is baseline subtracted for clarity and
i
represents ΔA at 524 nm.
biphasic appearance of the data in Figure 6, which was analyzed in
Figures S4 and S5. However, the conclusion that the rate constant
for formation of 2,6-DCQOH is approximately 8 times less than
that for 2,6-DCQ under typical conditions for enzymatic conver-
sion was also verified using the short time approximation, that is,
with a linear fit to the data in the first 15 s of the kinetic trace.
Linear fits were used for the temperature dependence of the
formation of 2,6-DCQOH presented in Figure 5. The fit shown in
Figure S7 to the linearized form of the Arrhenius rate constant, ln
k = ꢀE /RT + ln A, gives an activation energy of E = 36.3 kJ/mol.
range from 4 to 10. The studies showed that changing pH can
cause a significant change in the rate of TCP oxidation. DHP
shows the highest activity between pH 6.5 and 7.5. To under-
stand the effect of pH on secondary product formation, we
studied the degradation of 2,6-DCQ at pH 6.2, 7.4, and 9.0 in the
presence and absence of H O , following the 2,6-DCQ peak by
4
2
2
both UVꢀvisible spectroscopy and HPLC. On the basis of the
data shown in Figure 7, 2,6-DCQ reacts completely with H O in
2
2
a
a
∼2 min at pH 9.0, but 2,6-DCQ reacts more slowly with H O at
2 2
The details of the analysis are given in Figure S6 and Table S2 of
the Supporting Information. It is possible that other products are
formed above 30 °C given the observation that the 2,6-DCQOH
absorption band at 524 nm begins to level off at the higher tem-
peratures as shown in Figure 5.
lower pH. The single exponential time constant is ∼10 min at pH
7.4, but less than <1 min at pH 9.0.
Mechanism of Hydroxylation of 2,6-DCQ. To determine
whether there is a radical mechanism for the hydroxylation of 2,
6-DCQ, a solution of 2,6-DCQ/H O was subjected to room-
2 2
Effect of pH on the Hydroxylation of 2,6-DCQ. Most
enzymes are sensitive to pH and have specific ranges of activity.
A change in pH can protonate or deprotonate an amino acid side
temperature ESR analysis (Figure 8). When a sample containing
2,6-DCQ/H O at pH 7.4 was aspirated into the ESR spectrometer,
semistable radical signals (a = 2.38 G and a = 2.23 G) were
observed immediately (Figure 8A) and decayed within ∼3 min.
A radical was not observed at pH 6 or 9.0. Scheme 1 provides a
2
2
H
H
5
4ꢀ58
group, thereby changing the reactivity of the enzyme.
Oxi-
dation of TCP as catalyzed by DHP has been studied in the pH
1
670
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
Figure 7. HPLC measurements of the nonenzymatic reaction of 2,6-DCQ with H O at (A) pH 6.2, (B) 7.4, and (C) 9.0 (25 °C). Initial concentrations
2
2
were [2,6-DCQ] ≈ 100 μM, [H
2
O
2
] ≈ 200 μM.
(
SQ2, species 2, Figure 8B). We can also consider a termination
reaction shown in Scheme 3 that removes 2,6-dichlorohydroqui-
none dianion (HQ1) by reaction with H O (Table 1). It is also
2
2
3
0
possible that 2,6-dichlorohydroquinone reacts with O2. SQ1
dismutation (2SQ1 f HQ1 + 2,6-DCQ) shown in the bottom
row of Scheme 2 is a dominant reaction in solution, at least
initially, because 2,6-DCQ has a higher concentration than that
of 2,6-DCQOH. If the concentration of 2,6-DCQOH surpasses
that of 2,6-DCQ, then SQ2 dismutation (2SQ2 f HQ2 + 2,
6
-DCQOH) would be the radical termination step (reaction not
shown in Scheme 2). Scheme 3 indicates the likely fate of HQ1
in a reaction with H O , although there may be competing
2
2
radical reactions to form peroxyl or hydroxyl radicals that are not
considered here. An analogous reaction exists for HQ2 (i.e.,
ꢀ
HQ2 + H O f 2,6-DCQOH + 2OH ), with similar energetics
2
2
o
of ΔG ≈ ꢀ26.4 kJ/mol. At pH 9.0, base catalysis leads to an
acceleration in the formation of HQ2, which makes it likely that
the entire reaction proceeds too rapidly for detection using the
room temperature ESR. This observation is consistent with the
dramatic increase in rate constant observed by HPLC and would
explain why no radicals are observed by room-temperature ESR
at pH 9.
RFQ ESR spectra were obtained at pH 5, 7, and 9 to further
test the nonphotochemical mechanism of 2,6-DCQ hydroxyla-
tion. No signal was observed at pH 5 in either the presence or the
Figure 8. ESR spectrum of semiquinone detected when 2,6-DCQ was
mixed with H at pH 7.4. (A, solid line) The ESR spectrum detected
when 0.5 mM 2,6-DCQ was mixed with 0.5 mM H O and rapidly
2 2
O
2
2
transferred (∼1 min after mixing) to the ESR spectrometer. ESR
conditions: 9.80 GHz microwave frequency, 10 mW microwave power,
0.2 Gpp modulation amplitude, and a single scan of 81 s/conversion time
1
64 ms. (A, dashed line) Simulation of (A) using two radical species.
H
absence of H O at any quench time examined (up to 10 s),
Species 1 (81%), 2 spin 1/2 with a = 2.38 G; and species 2 (19%),
1
amplitude). (C) Simulation of species 1 alone.
2 2
H
spin 1/2 with a = 2.23 G. (B) Simulation of species 2 alone (4ꢁ
consistent with the results obtained from the room-temperature
ESR experiment. At pH 7, the reaction of 2,6-DCQ with H O
2
2
led to the observation of a radical signal at 100 ms that is
best described as an anisotropic singlet with a geff = 2.0064
(Figure 9A). The signal exhibited no time-dependent changes in
the shape of the spectral features between the 100 ms, 1 s, and 10
s data (only subtle changes in signal intensity were noted, which
were attributed to the differences in packing the ESR tube under
RFQ conditions), suggesting that the chemical nature of the
radical remained unchanged over the entire time range examined.
In contrast to the room-temperature ESR experiment, the
reaction of 2,6-DCQ with H O also yielded an observable radical
possible mechanism for the formation of hydroxylated secondary
ꢀ
products from 2,6-DCQ in the presence of H O , OH , and
2
2
ꢀ
OOH , corresponding to low, intermediate, and high pH, re-
spectively. At low pH, 2,6-DCQOH has been observed to form by
electrophilic addition by H O via either an epoxide or a cis diol
2
2
intermediate by a photochemical mechanism. Thus, at low pH,
,6-DCQOH does not result in the formation of radical inter-
2
mediates. In addition, 3-dihydroxy-6-chloroquinone (CQ-DiOH)
isformed asa sideproductbyHCl elimination. Atintermediate pH
2
2
ꢀ
of 7.4, the reaction is catalyzed by OH . 2,6-Dichloro-3-hydroxy-
at pH 9, with a g = 2.0067 (Figure 9B). Although not quantified,
eff
hydroquinone dianion (HQ2) is formed by nucleophilic attack
the signal intensity at pH 9 was qualitatively greater than that
observed at pH 7.4. Again, no time-dependent changes in the
shape of this radical signal were noted between the 20 ms, 1 s, and
10 s data. Overall, the shape and position of the radical signals
obtained at pH 7.4 and 9 were virtually indistinguishable, strongly
suggesting that the chemical identity of the radical was the same
regardless of pH. However, the rise of the signal at pH 9 is
ꢀ
ꢀ
ꢀ
by OH with further base catalysis by O H or OH to form
2
radical intermediates. At pH 7.4, radical semiquinones SQ1 and
SQ2 are formed by reaction of 2,6-DCQ and HQ2. The ESR data
indicate that the primary semiquinone is a symmetric radical with
H
two equivalent protons (a = 2.38) (SQ1, species 1, Figure 8C),
H
whereas the secondary semiquinone has a single proton (a = 2.23)
1
671
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
a
Scheme 1. Mechanism of Hydroxylation of 2,6-DCQ at Low, Intermediate, and High pH
a
5,6
The low pH mechanism has been shown to be a viable photochemical pathway, but has negligible rate under nonphotochemical conditions. The
intermediate and high pH reactivities are observed at pH 7.4 and 9, respectively.
Scheme 2. Radical Intermediates in the Formation of
Hydroxylated Chloroquinones
Scheme 3. Recycling Step for the 2,6-DCHQ Formed during
a
2
the Radical Chemistry Steps Shown in Scheme To Reform
a
2
,6-DCQ
a
2
Because O is present in the reaction mixture, we include the reduc-
tion of O to H O , which is a known pathway for oxidation of
2
2
2
hydroquinones.
arise from the absence of an explicit solvent in the calculation.
There is also an error of ∼4 kJ/mol for the method. The value of
the calculations is to identify high energy intermediates, which have
energies far outside the error of the method. The base -atalyzed path-
a
SQ1 and SQ2 are formed from the reaction of 2,6-DCQ and HQ2 at
pH 7.4, which we have called intermediate pH in Scheme 1. Two SQ1
radicals can also disproportionate to form 2,6-DCQ and HQ1. The net
effect of the radical pathway is to form 2,6-DCQOH as the thermo-
dynamic product.
ꢀ
way involves the reaction of OH with 2,6-DCQ to form the
ꢀ
hydroxyl anion in the 3 position (2,6-DCQ-3-O ) anion with a
o
free energy change of ΔG = 41.0 kJ/mol. Because this high
indicative of an increased rate consistent with the hypothesis of
base catalysis.
energy species is formed on the pathway, and not directly ob-
served, the free energy change for its formation is the barrier for
DFT Calculations. We have employed DFT calculations to
determine the plausibility of the reaction pathways in Scheme 1
by calculating the change in Gibbs free energy for each step.
Ground-state calculations implemented both the thermal treat-
the reaction and should be comparable to the activation energy, which
ꢀ
was measured as E = 36 kJ/mol. Subsequently, 2,6-DCQ-3-O
a
ꢀ
ꢀ
anion react with either OH or O H to form HQ2 with
2
o
ΔG = ꢀ5.4 and ꢀ1.2 kJ/mol, respectively (Table 1). The DFT
59
ment of electron occupancy extrapolated to 0 K, and the
COSMO module (conductor-like screening model) employed
calculation corresponding to the low pH reactivity shows that
H O attack on 2,6-DCQ leads to formation of 2,6-DCQ
2
2
3
,49,60
by DMol.
Table 1 divides the possible reactions into base-
epoxide, with ΔG°
= +8.4 kJ/mol. The epoxide can
react further with H O to yield 3,5-dichloro-2,3-dihydroxy-1,
DCQ Epox
catalyzed reactions at intermediate and high pH (>7.0) and
H O epoxidation at low pH (<7.0). The overall free energy
2
2
2
4-benzoquinone (2,6-DCQ diOH) with a free energy of
change of the hydroxylation of 2,6-DCQ by H O is calculated to
be ΔG = +8.0 kJ/mol. This positive free energy change may
ΔG°
= +43.9 kJ/mol, which is thus a high energy inter-
mediate along the low pH reaction pathway. There are two
2
2
DCQ DiOH
o
1
672
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
a
Table 1. Calculated Free Energy for Each Step in the Hydroxylation of 2,6-DCQ
o
o
reaction
G (kJ/mol)
ΔG (kJ/mol)
Nucleophilic Attack (Base-Catalyzed Pathway)
ꢀ135.1 f ꢀ94.1
ꢀ
ꢀ
2
2
2
,6-DCQ + OH f 2,6-DCQ-3-O
+41.0
ꢀ5.4
ꢀ1.2
+1.2
ꢀ
ꢀ
,6-DCQ-3-O + OH f HQ2 + H
2
O
ꢀ136.8 f ꢀ142.3
ꢀ
ꢀ
,6-DCQ-3-O + O
2
H
f HQ2 + H
2
O
2
ꢀ151.5 f ꢀ152.7
HQ2 + 2,6-DCQ f SQ1 + SQ2
ꢀ186.6 f ꢀ185.4
SQ2 + 2,6-DCQ f SQ1 + 2,6-DCQOH
ꢀ186.1 f ꢀ186.1
0.0
ꢀ
HQ1 + H
2
O
2
f 2,6-DCQ + 2OH
ꢀ151.0 f ꢀ177.8
ꢀ26.4
+8.0
2
,6-DCQ + H
2
O
2
f 2,6-DCQOH + H
2
O
ꢀ150.9 f ꢀ143.0
2 2
Formation of Epoxide and Cis Diol by H O Addition
2
2
2
2
,6-DCQ + H
2
O
2
f 2,6-DCQ epoxide + H
O f 2,6-DCQ-diOH
,6-DCQ epoxide f 2,6-DCQOH
,6-DCQ epoxide + H O f CQ-diOH + HCl
2
O
ꢀ151.4 f ꢀ143.1
ꢀ143.1 f ꢀ99.1
ꢀ95.2 f ꢀ94.8
ꢀ143.1 f ꢀ138.9
+8.4
+43.9
ꢀ0.4
+4.2
,6-DCQ epoxide + H
2
2
a
2 2
The base-catalyzed pathway is the nonphotochemical pathway observed in this study. The H O addition to form epoxide is primarily a photochemical
pathway.
2 2
Figure 9. ESR spectra of the reaction mixture of H O with 2,6-DCQ at pH 7.4 (A) and 9.0 (B). Rapid-freeze-quench samples (100 ms, 1 s, and 10 s)
were prepared from the reaction of H O (final concentration of 0.25 mM) with 2,6-DCQ (final concentration of 0.25 mM) at 20 °C and rapidly frozen
2
2
in an isopentane slurry. The samples quenched at 16 and 13 s (A and B) were prepared using manual mixing. Spectra were recorded at 190 K using the
spectrometer settings described in the Experimental Section. The frequency of the experiments was 9.0698 and 9.0699 GHz in panels A and B,
respectively.
possible outcomes for the 2,6-DCQ diOH intermediate. The
minor pathway consists of the elimination of HCl to form CQ-
that the formation of 2,6-DCQ is the primary product of the
catalytic oxidation of 2,4,6-TCP catalyzed by DHP A and B and
24
DiOH with a net free energy change of ΔG°
= +4.2 kJ/mol.
cosubstrate H O by a one-electron mechanism. Evidence sug-
CQ diOH
2 2
The major pathway is the elimination of H O to form 2,
gests that both the catalytic transformation leading to 2,6-DCQ
formation and the ensuing spontaneous hydroxylation reaction
occur by a radical mechanism. Radical formation is thus a per-
vasive feature of catalytic oxidation of 2,4,6-TCP and similar
halogenated phenols, which likely is related to the toxicity of
these molecules. While phenoxyl radicals have been discussed in
a number of studies, the central role played by quinone radicals,
revealed by this study, is likely to have a significant consequence
in A. ornata, as well as for any development of DHP A and B for
biotechnology applications.
2
6
-DCQOH with ΔG°2
= ꢀ0.4 kJ/mol. These calcula-
,6‑DCQOH
tions were carried out in a vacuum. It is anticipated that inclusion
of solvent would make the free energy for product formation
more favorable. The free energy considerations suggest that 2,
6
-DCQOH would be the favored product at low pH, although
the accuracy of the method is not greater than 4 kJ/mol so that it
is possible that the dihydroxy compound, 2,6-DCQ diOH, is
formed as well. The numerical values for each intermediate and
product are presented in Table 1.
Radical chemistry is important for all aspects of oxidation
of 2,4,6-TCP and related phenolic compounds by peroxidases.
’
DISCUSSION
3
Following the cloning and expression of DHP A, we pointed out
1
2
The present study confirms that 2,6-DCQOH is the major
that if the substrate binds internally, as originally proposed,
product of the nonphotochemical base-catalyzed oxidation of
then a one-electron oxidation mechanism would require a radi-
2
,6-DCQ. This observation builds on previous work showing
cal product, a phenoxyl radical, to diffuse through the protein.
1
673
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
Given the high reactivity of phenoxyl radicals, there would be a sig-
nificant possibility of radical reactions as this radical would pass
by tyrosines 28, 34, and 38 on its migration toward the solvent.
Consistent with this concern, we have shown that tyrosine
radicals are observed with nearly 100% yield in DHP A when
form 2,6-DCQOH (Arrhenius plot in Table S2 and Figure S3).
This activation energy is consistent with DFT calculations that
show that the initial hydroxylation step has a free energy of
41 kJ/mol (Table 1), which would suggest that attack by
hydroxide is the rate-limiting step in the formation of 2,
6-DCQOH at pH 7.4 and hence explains the activation barrier.
However, the relative activation energies of the two processes,
formation of 2,6-DCQ and 2,6-DCQOH, do not explain why the
hydroxylation is observed first above 20 °C. The empirical fact
that the rate of formation of 2,6-DCQOH increases more rapidly
than the catalyzed rate of 2,6-DCQ formation as a function of
temperature shown in Figure 6 can be modeled as a pseudofirst-
order process as described in the Supporting Information. The
justification for a pseudofirst-order treatment is that concentra-
25,26
substrate is not present.
We have recently confirmed that
2
,4,6-trichlorophenoxyl radicals can be directly observed follow-
23
ing oxidation of 2,4,6-TCP by DHP. To resolve the paradox of
radical reactivity at an internal site, we suggested that a two-
electron oxidation could be a reasonable mechanism to explain
an internal substrate binding site. However, the hypothesis that
the internal site was a substrate binding site was disproven by
inhibition studies combined with resonance Raman spectros-
copy that showed opposing effects of inhibitor 4-BP and sub-
12
14
ꢀ
strate 2,4,6-TCP on the distal histidine (H55) conformation.
tions of enzyme (DHP A), H O , and OH are essentially con-
2
2
Furthermore, NMR data clearly showed different patterns of
interaction for inhibitors, such as 4-BP, and substrates, such as
stant on the time scale of the reaction. This simplifying assump-
tion can explain the temperature dependence of formation of
2,6-DCQ and subsequent formation of 2,6-DCQOH. The
∼36 kJ/mol activation energy for the intrinsic rate constant for
the hydroxylation reaction means that it decreases by a factor ∼2
between 15 and 30 °C. The temperature dependence of the for-
mation of 2,6-DCQ is greater, and it changes by ∼3.5 over this
range. Thus, even though the temperature dependence of the
hydroxylation reaction is smaller than that of the formation, the
time scale for that process shifts from ∼100 s at 30 °C to ∼300 s
at 15 °C for the observable onset of 2,6-DCQOH formation.
Because our kinetic analyses were conducted for 300 s, we did not
observe a significant decay of 2,6-DCQ at temperatures below
20 °C. The kinetic model calculated from the rate equations
based on a finite difference approach presented in Figures S8ꢀ
S10 explains the appearance of the data at both 273 and 524 nm.
At low pH, the oxidation of 2,6-DCQ may occur by H O
2
,4,6-TCP, which suggested internal and external modes of
15,16
binding, respectively.
radical (2,4,6-TCPO ) dismutation is rapid, such that it can only
be observed by flow-ESR. The observed product in spectro-
We have also shown that phenoxyl
•
23
photometric assays is 2,6-DCQ, which has a maximum absorbance
2
4,26
at 273 nm.
Thus, we currently understand the mechanism of
substrate oxidation to involve external binding of 2,4,6-TCP
followed by one electron oxidation to form the 2,4,6-TCPO•
radical followed by radical dismutation. The oxidized 2,4,6-TCP
ꢀ
cation rearranges to form 2,6-DCQ and Cl .
Our study finds significant nonphotochemical reactivity of 2,
-DCQ via a radical mechanism that has not been previously
6
described. We have correlated the decline in product, 2,6-DCQ,
concentration at 273 nm with the appearance of 2,6-DCQOH
ꢀ1
ꢀ1
(
λmax = 524 nm, ε = 2350 M cm ) (Figures S4 and S5) using
2
2
studies to show both temperature and pH-dependence. Although
we have relied on DHP A for the mechanistic comparisons that
involve enzymatic catalysis, it is believed that the spontaneous
hydroxylation chemistry has general validity for all peroxidases,
and indeed for any process that produces 2,6-DCQ. While the
observed spectral changes are consistent with hydroxylation of
epoxidation of the quinone (Scheme 1). The epoxide could be
ring opened to form either the trans or the cis diols in aqueous
solution. There are then two possible mechanisms leading to
distinct products. First, the elimination of HCl and formation of
CQ-DiOH is possible. The second mechanism, involving the
elimination of H O leading to the formation of 2,6-DCQOH, is
2
2
,6-DCQ, the hypothesis of a secondary product requires a
the favored reaction based on the DFT calculations. The photo-
chemistry of 2,6-DCQ appears to be more complex than was
previously thought based on the observation of radicals in the
RFQ-ESR and reactivity even in the absence of H O as shown in
1
13
structural proof. We have used 2D Hꢀ C HMBC NMR ex-
periments to establish the identity of the secondary product, 2,
6
-DCQOH. We were compelled to use DMSO as a solvent for
2
2
these studies due to the low solubility of 2,6-DCQ in aqueous
solution. The addition of H O to 2,6-DCQ in DMSO provides
firm evidence that 2,6-DCQOH is the major species formed by
the spontaneous oxidation of 2,6-DCQ.
Figure S13. We note that the epoxidation mechanism does not
occur via a radical mechanism and would be ESR-silent. The
UVꢀvis data in Figure S13 may contain a contribution from this
pathway. The ESR data in Figures S11 and S12 suggest that
photochemistry can affect the radical pathway at higher pH.
These facts are mentioned in this report for the sake of complete-
ness, although we make no claim to have conducted a compre-
hensive study of the photochemical pathway. The photochemical
pathway is worth including in our discussion because there clearly
is relevance to the biological degradation of 2,4,6-TBP and 2,4,
6-TCP. The nonphotochemical yield of hydroxylation product
falls nearly to zero by pH 5, which strongly suggests that base
catalysis is required for nonphotochemical reactivity in the pre-
sence of H O .
2
2
The temperature dependence of the reaction confirms the
kinetic model for the formation of the hydroxylated product.
Specifically, it explains why 2,6-DCQOH is observed in signifi-
cant yield only when the temperature exceeds 20 °C. The non-
photochemical hydroxylation of 2,6-DCQ has a measurable rate
above 10 °C. The kinetics of the 2,6-DCQ product formation and
decay by the nonphotochemical hydroxylation reaction have a
biexponential form above 25 °C. The fit to an Arrhenius model
associated with the MichaelisꢀMenten kinetic scheme gave an
activation energy ofE ≈ 56 kJ/mol for substrate oxidation to form
a
2
2
2
4
2
,6-DCQ. This activation barrier was hypothesized to be pri-
This study validates the hypothesis that a base-catalyzed
radical-mediated process is responsible for the nonphotochem-
ical oxidation of 2,6-DCQ. The base-catalyzed reaction may also
marily due to the requirement for substrate diffusion to the active
site onthe surfaceofDHP A, which was therate-limiting processin
24
ꢀ
ꢀ
the complex peroxidase reaction scheme. The data presented
involve reactivity by HOO at high pH. Attack by OH yields
HQ2 and radical intermediates, SQ1 and SQ2, that can form
2,6-DCQOH and regenerate 2,6-DCQ as shown in Scheme 3.
here show that 2,6-DCQOH has a significantly lower activation
energy of E ≈ 36 kJ/mol for the hydroxylation of 2,6-DCQ to
a
1
674
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
The data obtained at pH 7.4 showing the transient concentration
of two radicals are readily explained by the mechanism shown in
Schemes 2 and 3. Scheme 1 provides a plausible mechanism for
the base-catalyzed hydroxylation of 2,6-DCQ in aqueous solu-
tion at pH 7.4. 2,6-DCQ can undergo base-catalyzed nucleophilic
addition leading to HQ2. Product HQ2 is a hydroquinone
dianion, which can be involved in a cross-oxidation reaction with
at physiological temperatures and pH, which suggests that 2,
6-DCQOH is an important intermediate in the degradation of 2,4,
6-TCP in benthicecosystemswhere A. ornata isfound, aswellasin
anybioremediationprocess designedtouseDHPAor B. Thestudy
focused on the nonphotochemical process, which occurs at phy-
siological pH. However, we have also shown that this set of oxi-
dation reactions is even more complex and that a complete study
would have to include temperature, pH, H O concentration, and
2
,6-DCQ to yield the observed semiquinone radicals, SQ1 and
2
2
SQ2 (Scheme 2). Semiquinone SQ2 can react with 2,6-DCQ to
yield SQ1 and the hydroxylated product, 2,6-DCQ-3-OH
light flux as variables that would be studied by flow and rapid-freeze-
quench ESR as key methods to elucidate the pathways.
(
Scheme 2). On the basis of the mechanism in Scheme 2, which
leads to HQ2 and subsequent formation of semiquinone radicals,
OH is thought to dominate at pH 7.4, even though the [OH ]
and [HOO ] concentrations are nearly equal at this pH. We
’
ASSOCIATED CONTENT
ꢀ
ꢀ
ꢀ
S
Supporting Information. Table S1: Predicted and experi-
b
ꢀ
ꢀ7
note that [OH ] = 2.5 ꢁ 10 M at pH 7.4 as compared to
mental chemical shifts. Figure S1: Calculation of initial rates
for formation of 2,6-DCQ and 2,6-DCQOH. Figure S2: HSQC
data for 2,6-DCQ and 2,6-DCQOH. Figure S3: Single wave-
length (524 nm) kinetic trace of 2,6-DCQOH. Figure S4: The
biexponential kinetics of 2,6-DCQ formation. Figure S5: The
exponential growth of the secondary product, 2,6-DCQOH, at
524 nm. Figure S6: Study of the hydroxylation of 2,6-DCQ at
different temperatures. Table S2: Rate constants for hydroxyla-
tion of 2,6-DCQ at different temperatures. Figure S7: Arrhenius
plot of the data fit to the model equation. Figure S8: Calculated
model for kinetics at 30 °C. Figure S10: Calculated model for
kinetics at 15 °C. Figure S19: Comparison of kinetics calculated
at 30 and 15 °C. Figure S10: ESR spectra of 2,6-DCQ with added
H O at pH 7.4. Figure S11: ESR spectra of 2,6-DCQ with added
ꢀ
ꢀ7
[
HOO ] = 3.0 ꢁ 10 M for [H O ] = 500 μM. The latter is
2
2
based on the pK of 11.6 for H O , a reaction pH of 7.4, and the
a 2 2
ꢀ
HendersonꢀHasselbach equation rearranged to give [HOO ] =
pHꢀpK
a
[H O ]10
. On the basis of RFQ-ESR, we have also found
2
2
that there is an additional base-catalyzed photochemical reac-
tion, which may yield a dihydroxylated product, 2,6-dichloro-
3
,5-dihydroxyquinone, shown in Scheme S1 of the Supporting
Information. The RFQ-ESR data showing the radicals produced
under those conditions are given in Figures S11 and S12.
Finally, we consider the high pH mechanism and the observa-
tion that no radical signal was observed in room-temperature
benchtop mixing experiments at pH 9.0, and yet radicals were
ꢀ
detected under RFQ-ESR conditions. At pH 9.0, [HOO ] =
1
2
2
ꢀ
6
ꢀ
ꢀ5
.2 ꢁ 10 M and [OH ] = 10 M, which would also suggest
H O at pH 9. Figure S12: UVꢀvis spectra of the reaction of 2,
2
2
formation of HQ2. However, the HPLC data show that the
reaction is an order of magnitude more rapid at this pH. The fact
that no radical was observed in the ESR spectrum obtained in
benchtop mixing ESR experiments at room temperature suggests
two possibilities. First, it is possible that the reaction is so much
faster that no radicals are observed under rapid mixing condi-
tions. The time constant for formation of 2,6-DCQOH is less
than 1 min at pH 9.0, whereas it is ∼10 min at pH 7.4. A second
6-DCQ with and without H O . Scheme S1: Possible model for a
2 2
second hydroxylation. This material is available free of charge via
the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
Tel.: (919) 515-8915. E-mail: stefan_franzen@ncsu.edu.
*
ꢀ
possibility is a direct nucleophilic attack by HOO at pH 9. As
shown in Scheme 1, this mechanism could potentially generate
the 2,6-DCQOH product without forming radicals. Because
radicals are observed by RFQ-ESR at pH 9, we favor the first
possibility and include the second one for completeness. The
conclusions reached in this study show how the flow and rapid-
freeze quench ESR data complement each other. The assignment
of the radicals would have been difficult using the RFQ-ESR
technique. However, the kinetic trajectory of radical formation
and reaction is not discernible from flow ESR alone.
’
ACKNOWLEDGMENT
R.G. and S.F. gratefully acknowledge support by Army Re-
search Office Grant 57861-LS. We gratefully acknowledge the
receipt of a PUC-19 plasmid containing the eukaryotic DHP A
gene from Drs. Lincoln and Dawson of the University of South
Carolina.
’
REFERENCES
(
1) Ferrari, R. P.; Laurenti, E.; Trotta, F. JBIC, J. Biol. Inorg. Chem.
999, 4, 232–237.
2) Laurenti, E.; Ghibaudi, E.; Ardissone, S.; Ferrari, R. P. J. Inorg.
1
’
CONCLUSION
(
For both applied and fundamental reasons, it is valuable to
Biochem. 2003, 95, 171–176.
understand the mechanism of oxidation of 2,4,6-TCP to 2,
-DCQ by DHP. Chlorophenols, which have been produced in-
dustrially on a large scale, constitute a significant class of environ-
(3) Belyea, J.; Gilvey, L. B.; Davis, M. F.; Godek, M.; Sit, T. L.;
Lommel, S. A.; Franzen, S. Biochemistry 2005, 44, 15637–15644.
6
(4) Franzen, S.; Gilvey, L. B.; Belyea, J. L. Biochim. Biophys. Acta,
6
1ꢀ63
Proteins Proteomics 2007, 1774, 121–130.
(5) Franzen, S.; Belyea, J.; Gilvey, L. B.; Davis, M. F.; Chaudhary,
C. E.; Sit, T. L.; Lommel, S. A. Biochemistry 2006, 45, 9085–9094.
mental pollutants.
Specifically, 2,4,6-TCP has been used
64
extensively as wood preservatives and pesticides. 2,4,6-TCP is
also useful as a model substrate in DHP A and B, because it has
close structural similarity to 2,4,6-TBP, the native substrate of
DHP, and is more soluble. The present study demonstrates that
(
6) Osborne, R. L.; Coggins, M. K.; Raner, G. M.; Walla, M.;
Dawson, J. H. Biochemistry 2009, 48, 4231–4238.
7) Osborne, R. L.; Coggins, M. K.; Walla, M.; Dawson, J. H.
(
2
,6-DCQOH is a spontaneous oxidation product of 2,6-DCQ
Biochemistry 2007, 46, 9823–9829.
(8) Chen, Z.; de Serrano, V.; Betts, L.; Franzen, S. Acta Crystallogr.,
Sect. D: Biol. Crystallogr. 2009, D65, 34–40.
formed during the catalytic oxidation of 2,4,6-TCP by DHP. This
work shows that the secondary hydroxylation reaction is significant
1
675
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676

The Journal of Physical Chemistry B
ARTICLE
(
9) de Serrano, V.; Chen, Z.; Davis, M. F.; Franzen, S. Acta Crystal-
logr., Sect. D: Biol. Crystallogr. 2007, D63, 1094–1101.
10) de Serrano, V.; D’Antonio, J.; Franzen, S.; Ghiladi, R. A. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66, 529–538.
11) de Serrano, V. S.; Davis, M. F.; Gaff, J. F.; Zhang, Q.; Chen, Z.;
(43) Perdew, J. P.; Burke, K.; Wang, Y. Phys. Rev. B: Condens. Matter
1996, 54, 16533–16539.
(44) Andzelm, J.; King-Smith, R. D.; Fitzgerald, G. Chem. Phys. Lett.
2001, 335, 321–326.
(
(
(45) Baker, J.; Kessi, A.; Delley, B. J. Chem. Phys. 1996, 105, 192–212.
(46) Delley, B. J. Chem. Phys. 1990, 92, 508–517.
(47) Delley, B. J. Chem. Phys. 2000, 113, 7756–7764.
(48) Weinert, M.; Davenport, J. W. Phys. Rev. B 1992, 45, 13709.
(49) Andzelm, J.; Kolmel, C.; Klamt, A. J. Chem. Phys. 1995, 103,
9312–9320.
(50) Lente, G.; Espenson, J. H. New J. Chem. 2004, 28, 847–852.
(51) Sorokin, A.; DeSuzzoniDezard, S.; Poullain, D.; Noel, J. P.;
Meunier, B. J. Am. Chem. Soc. 1996, 118, 7410–7411.
(52) Lente, G.; Espenson, J. H. Chem. Commun. 2003, 1162–1163.
(53) (ACD/Laboratories). CAS Registry Number: 89465-84-9;
Software V9.07 ed.; Advanced Chemistry Development.
(54) Cook, P. F.; Cleland, W. W. Biochemistry 1981, 20, 1797–1805.
(55) Zaks, A.; Klibanov, A. M. Proc. Natl. Acad. Sci. U.S.A. 1985,
82, 3192–3196.
D’Antonio, E. L.; Bowden, E. F.; Rose, R.; Franzen, S. Acta Crystallogr.,
Sect. D: Biol. Crystallogr. 2010, 66, 770–782.
(
12) Lebioda, L.; LaCount, M. W.; Zhang, E.; Chen, Y. P.; Han, K.;
Whitton, M. M.; Lincoln, D. E.; Woodin, S. A. Nature (London) 1999,
01, 445.
13) LaCount, M. W.; Zhang, E.; Chen, Y. P.; Han, K.; Whitton,
M. M.; Lincoln, D. E.; Woodin, S. A.; Lebioda, L. J. Biol. Chem. 2000,
75, 18712–18716.
14) Thompson, M. K.; Davis, M. F.; deSorrano, V.; Nicoletti,
F. P.; Howes, B. D.; Smulevich, G.; Franzen, S. Biophys. J. 2010, 99,
4
(
2
(
1
586–1595.
15) Davis, M. F.; Bobay, B. G.; Franzen, S. Biochemistry 2010, 49,
199–1206.
16) Davis, M. F.; Gracz, H.; Vendeix, F. A. P.; de Serrano, V.;
Somasundaram, A.; Decatur, S. M.; Franzen, S. Biochemistry 2009,
8, 2164–2172.
17) Nienhaus, K.; Deng, P.; Belyea, J.; Franzen, S.; Nienhaus, G. U.
J. Phys. Chem. B 2006, 110, 13264–13276.
18) Nienhaus, K.; Nickel, E.; Davis, M. F.; Franzen, S.; Nienhaus,
G. U. Biochemistry 2008, 47, 12985–12994.
19) Nicoletti, F. P.; Thompson, M. K.; Howes, B. D.; Franzen, S.;
Smulevich, G. Biochemistry 2010, 49, 1903–1912.
(
1
(
(56) Vitello, L. B.; Huang, M.; Erman, J. E. Biochemistry 1990, 29,
4283–4288.
(57) Thomas, P. G.; Russell, A. J.; Fersht, A. R. Nature 1985, 318,
4
(
375–376.
(
(58) Conroy, C. W.; Tyma, P.; Daum, P. H.; Erman, J. E. Biochim.
Biophys. Acta, Protein Struct. 1978, 537, 62–69.
(59) Mermin, D. Phys. Rev. A 1965, 137, 1441–1443.
(60) Delley, B. Mol. Simul. 2006, 32, 117–123.
(61) Reddy, G. V. B.; Gelpke, M. D. S.; Gold, M. H. J. Bacteriol. 1998,
180, 5159–5164.
(
(20) Dunford, H. B. Xenobiotica 1995, 25, 725–733.
(
21) Ator, M. A.; David, S. K.; Demontellano, P. R. O. J. Biol. Chem.
1
987, 262, 14954–14960.
(62) Duran; Deschler; Precigou; Goulas Appl. Microbiol. Biotechnol.
(
22) Raven, E. L. Subcell. Biochem. 2000, 35, 317–349.
2002, 59, 284–288.
(
23) Sturgeon, B. E.; Battenburg, B. J.; Lyon, B. J.; Franzen, S. Chem.
(63) Bumpus, J. A.; Aust, S. D. BioEssays 1987, 6, 166–170.
(64) Hiratsuka, N.; Oyadomari, M.; Shinohara, H.; Tanaka, H.;
Wariishi, H. Biochem. Eng. J. 2005, 23, 241–246.
Res. Toxicol., ASAP.
24) Ma, H.; Thompson, M. K.; Gaff, J. F.; Franzen, S. J. Phys. Chem.
B 2010, 114, 13823–13829.
25) Thompson, M. K.; Franzen, S.; Ghiladi, R. A.; Reeder, B. J.;
Svistunenko, D. A. J. Am. Chem. Soc. 2010, 132, 17501–17510.
26) Feducia, J.; Dumarieh, R.; Gilvey, L. B. G.; Smirnova, T.;
Franzen, S.; Ghiladi, R. A. Biochemistry 2009, 48, 995–1005.
(
(
(
(
(
27) D’Antonio, J.; Ghiladi, R. A. Biochemistry 2011, 50, 5999–6011.
28) D’Antonio, J.; D’Antonio, E. L.; Thompson, M. K.; Bowden,
E. F.; Franzen, S.; Smirnova, T.; Ghiladi, R. A. Biochemistry 2010,
49, 6600–6616.
(29) Lente, G.; Espenson, J. H. Green Chem. 2005, 7, 28–34.
(30) Song, Y.; Buettner, G. R. Free Radical Biol. Med. 2010, 49,
9
5
1
19–962.
(31) Wong, S. K.; Wan, J. K. S.; Sytnyk, W. Can. J. Chem. 1972,
0, 3052–&.
(32) Sorokin, A.; Meunier, B.; Seris, J.-L. Science 1995, 268,
163–1166.
(
(
33) Xun, L.; Webster, C. M. J. Biol. Chem. 2004, 279, 6696–6700.
34) Song, Y.; Wagner, B. A.; Lehmler, H.-J.; Buettner, G. R. Chem.
Res. Toxicol. 2008, 21, 1359–1367.
35) Fielman, K. T.; Woodin, S. A.; Walla, M. D.; Lincoln, D. E. Mar.
Ecol.: Prog. Ser. 1999, 181, 1–12.
36) Lincoln, D. E.; Fielman, K. T.; Marinelli, R. L.; Woodin, S. A.
Biochem. Syst. Ecol. 2005, 33, 559–570.
37) Woodin, S. A.; Lindsay, S. M.; Lincoln, D. E. Mar. Ecol.: Prog.
Ser. 1997, 157, 303–306.
38) Woodin, S. A.; Walla, M. D.; Lincoln, D. E. J. Exp. Mar. Biol.
Ecol. 1987, 107, 209–217.
39) Han, K.; Woodin, S. A.; Lincoln, D. E.; Fielman, K. T.; Ely, B.
Mar. Biotechnol. 2001, 3, 287–292.
40) Chen, Y. P.; Woodin, S. A.; Lincoln, D. E.; Lovell, C. R. J. Biol.
Chem. 1996, 271, 4609–4612.
41) Gunther, M. R.; Sturgeon, B. E.; Mason, R. P. Free Radical Biol.
Med. 2000, 28, 709–719.
42) Duling, D. R. J. Magn. Reson., Ser. B 1994, 104, 105–110.
(
(
(
(
(
(
(
(
1
676
dx.doi.org/10.1021/jp208536x |J. Phys. Chem. B 2012, 116, 1666–1676