Mechanisms of N-demethylations catalyzed by high-valent species of heme enzymes: Novel use of isotope effects and direct observation of intermediates [2]

Full text of DOI:10.1021/ja981357u

10762
J. Am. Chem. Soc. 1998, 120, 10762-10763
Scheme 1
Mechanisms of N-Demethylations Catalyzed by
High-Valent Species of Heme Enzymes: Novel Use
of Isotope Effects and Direct Observation of
Intermediates
Yoshio Goto,† Yoshihito Watanabe,*,†,‡ Shunichi Fukuzumi,§
Jeffrey P. Jones,| and Joseph P. Dinnocenzo
We have examined the N-demethylation mechanisms through
Department of Structural Molecular Science
The Graduate UniVersity for AdVanced Studies
Myodaiji, Okazaki, 444-8585 Japan
novel isotope effect studies carried out by direct observation of
the reduction of the high-valent species responsible for catalysis.
Reactions of compound I of HRP and of OdFeIVTMP porphyrin
π-cation radical (which serves as a functional model for the active
species of cytochrome P450) with a series of para-substituted
dimethylanilines (DMAs) have been investigated (TMP )
5,10,15,20-tetramesitylporphine dianion). The dependence of the
rate constants on the one-electron oxidation potentials of DMAs
and the comparison of the kinetic and product isotope effects have
allowed us to clarify the mechanism of N-demethylation.
Upon the mixing of HRP compound I and DMAs under
stopped-flow conditions in a buffer solution (pH 7.0) at 273K,15
rapid formation of compound II (k1) and the following relatively
slow conversion (k2) to the resting state were observed. The same
reactions were also carried out with deuterated compounds
(DMAs(-CD3)2)16 to determine k1D and k2D. The k1 and k2 values
and the KIEs (k1H/k1D and k2H/k2D) are listed in Table 1 together
with the one-electron oxidation potentials (E0) of DMAs.17 Both
the k1 and k2 values increase with decrease in the E0 value. A
linear correlation between log k1 and E0 is shown in Figure 1a.18
No kinetic isotope effects (KIEs) are observed for either k1 or k2.
The dependence of k1 and k2 on E0 and the absence of KIEs clearly
indicate that the rate-determining steps are the electron transfers
from DMA to HRP compound I and compound II, respectively.
While transient formation of compound I of P450 has been
reported, because it is formed in a mixture of other species,19 it
has not been fully characterized. Thus, we have employed
OdFeIVTMP•+ (1)20 as a model complex for the proposed high-
valent catalytic intermediate of P450 in the kinetic measurement
of N-demethylation of DMAs since 1 is able to mimic most P450-
catalyzed oxidations. The reactions of 1 with DMAs in CH2Cl2
at 223 K were monitored by UV-vis spectral changes of 1 and
found to afford FeIIITMP without observation of any intermedi-
ates.21 The rate constants (k3) and kinetic isotope effects (k3H/
k3D) are summarized in Table 1. In addition, product (intramo-
lecular) isotope effects (kH/kD)22 are observed in the reactions of
1 with DMAs(-CD3, -CH3)16 as listed in Table 1.
Institute for Molecular Science
Myodaiji, Okazaki 444-8585 Japan
Department of Applied Chemistry
Osaka UniVersity, Suita, 565 Japan
Departments of Pharmacology and Chemistry
UniVersity of Rochester, Rochester, New York 14642
ReceiVed April 21, 1998
The molecular mechanisms of amine N-demethylation by heme
enzymes including peroxidases and cytochrome P450 have been
studied for over three decades.1-14 While the overall reaction
mechanism consists of R-hydroxylamine formation followed by
hydrolysis to afford N-demethylated products and formaldehyde,
the mechanism of R-hydroxylamine formation is still controver-
sial. Large intramolecular isotope effects observed for the
N-demethylation of N,N-dimethylaniline (DMA)3,5,6,13,14 have been
interpreted to be due either to direct hydrogen abstraction or to
proton transfer from the aminium radical, which is the one-
electron oxidation product of the amine. In the case of horseradish
peroxidase (HRP)-catalyzed oxidation of several amines, the
corresponding aminium radicals have been detected by EPR to
support the involvement of the electron-transfer process before
the R-hydroxylation (Scheme 1).8,9 On the other hand, the
mechanism of the N-demethylation catalyzed by P450 is still
under debate.3,5,13 It has recently been disclosed that the
magnitude of the intramolecular isotope effects on P450-catalyzed
N-demethylations of substituted DMAs determined by product
analysis are nearly identical to those on H-atom abstraction by a
tert-butoxyl radical (tBuO•).13,14 This result suggests the involve-
ment of H-atom abstraction in the P450 reaction. However, direct
observation of the oxidation process is crucial for elucidation of
the detailed mechanism.
† The Graduate University for Advanced Studies.
‡ Institute for Molecular Science.
Figure 1b shows a linear correlation between log k3 and E0
similar to the correlation observed for log k1 and E0 (Figure 1a).
§ Osaka University.
| Department of Pharmacology, University of Rochester.
Department of Chemistry, University of Rochester.
(1) Ortiz de Montellano, P. R. Cytochrome P450. Structure, Mechanism,
and Biochemistry, 2nd ed.; Plenum Publishing Corporation: New York, 1995.
(2) Guengerich, F. P.; Macdonald, T. L. Acc. Chem. Res. 1984, 17, 9-16.
(3) Miwa, G. T.; Walsh, J. S.; Kedderis, G. L.; Hollenberg, P. F. J. Biol.
Chem. 1983, 258, 14445-14449.
(15) Kinetic experiments involving HRP compounds I and II were
performed with a Hi-Tech SF-43 stopped-flow instrument. Compound I was
prepared by mixing with a stoichiometric amount of H2O2 followed by mixing
with 10-100 equiv of DMA at 273 K in 50 mM phosphate buffer (pH 7.0)
with double-mixing mode (delay time, 10 ms; final concentration of HRP,
2.5 µM). The UV-vis spectral changes showed rapid conversion of compound
I to compound II (Supporting Information). The rate of compound II formation
(k1) was determined by fitting the change in absorbance at 411 nm with a
least-squares procedure; the fit was linearly dependent on the concentration
of the substrates. Compound II was prepared by addition of stoichiometric
amounts of H2O2 to a solution of ferric HRP followed by reduction with
stoichiometric amounts of potassium ferricyanide. The reaction rate (k2) of
compound II and DMA was also determined from the absorbance change at
403 nm by a similar method.
(4) Miwa, G. T.; Garland, W. A.; Hodshon, B. J.; Lu, A. Y. H.; Northrop,
D. B. J. Biol. Chem. 1980, 255, 6049-6054.
(5) Guengerich, F. P.; Yun, C.-H.; Macdonald, T. L. J. Biol. Chem. 1996,
271, 27321-27329.
(6) Okazaki, O.; Guengerich, F. P. J. Biol. Chem. 1993, 268, 1546-1552.
(7) Macdonald, T. L.; Gutheim, W. G.; Martin, R. B.; Guengerich, F. P.
Biochemistry 1989, 28, 2071-2077.
(8) Griffin, B. W.; Ting, P. L. Biochemistry 1978, 17, 2206-2211.
(9) Van der Zee, J.; Duling, D. R.; Mason, R. P.; Eling, T. E. J. Biol.
Chem. 1989, 264, 19828-19836.
(16) Dinnocenzo, J. P.; Karki, S. B.; Jones, J. P. J. Am. Chem. Soc. 1993,
115, 7111-7116.
(10) Kedderis, G. L.; Hollenberg, P. F. J. Biol. Chem. 1983, 258, 8129-
8138.
(17) Oxidation potentials of DMAs were determined by cyclic voltammetric
measurements (BAS 100B/W) in CH2Cl2 containing 2 mM DMA derivatives
and 50 mM tetrabutylammonium hexafluorophosphate at 223 K with a voltage
sweep rate of 100-200 mV/s.
(11) Kedderis, G. L.; Koop, D. R.; Hollenberg, P. F. J. Biol. Chem. 1980,
255, 10174-10182.
(12) Lindsey Smith, J. R.; Mortimer, D. N. J. Chem. Soc., Perkin Trans.
2 1986, 1743-1749.
(18) The linear correlation between log k2 and E0 was also observed.
(19) Egawa, T.; Shimada, H.; Ishimura, Y. Biochem. Biophys. Res.
Commun. 1994, 201, 1464-1469.
(13) Manchester, J. I.; Dinnocenzo, J. P.; Higgins, L.; Jones, J. P. J. Am.
Chem. Soc. 1997, 119, 5069-5070.
(14) Karki, S. B.; Dinnocenzo, J. P.; Jones, J. P.; Korzekwa, K. R. J. Am.
Chem. Soc. 1995, 117, 3657-3664.
(20) Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1988, 110, 8443-
8452.
10.1021/ja981357u CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/01/1998

Communications to the Editor
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10763
Table 1. Bimolecular Rate Constants and the Isotope Effects of the Reactions of HRP Compounds I, II and OdFeIVTMP•+ with a Series of
Para-substituted DMAs of which Oxidation Potentials Are in the Second Column from the Left
HRP system
k1H/k1D
k2b [M-1 s-1
TMPFe system
k3b [M-1 s-1
k3H/k3D
para-substituent
of DMA
E0 [V vs Fc]
k1b [M-1 s-1
]
]
k2H/k2D
]
kH/kD
OMe
Me
H
Cl
Br
CN
NO2
0.14
0.33
0.53a
0.49
0.47
0.82
0.92
(9.5 ( 0.6) × 107 1.0 ( 0.1 (4.6 ( 0.1) × 106 1.0 ( 0.1 (5.6 ( 0.1) × 107 1.3 ( 0.1 3.9 ( 0.1
(3.1 ( 0.1) × 107 1.0 ( 0.1 (2.1 ( 0.1) × 105 1.0 ( 0.1 (2.7 ( 0.1) × 105 1.5 ( 0.1 3.8 ( 0.1
(6.0 ( 0.3) × 105 1.1 ( 0.1 (1.6 ( 0.1) × 104 1.1 ( 0.1 (1.7 ( 0.1) × 104 1.9 ( 0.2 2.8 ( 0.1
(1.6 ( 0.1) × 106 1.0 ( 0.1 (3.2 ( 0.2) × 104 1.0 ( 0.1 (3.4 ( 0.1) × 103 1.8 ( 0.1 2.6 ( 0.1
(1.8 ( 0.1) × 106 1.0 ( 0.1 (3.4 ( 0.2) × 104 1.0 ( 0.1 (6.6 ( 0.1) × 103 2.0 ( 0.1 3.4 ( 0.1
(8.7 ( 0.8) × 102 1.0 ( 0.2 ndc
nd
nd
(9.0 ( 0.4) × 10
(2.2 ( 0.1) × 10
5.5 ( 0.4 6.9 ( 0.1
5.9 ( 0.4 6.2 ( 0.9
nd
nd
nd
a Peak potential was used because no reversible wave was detected on cyclic voltammetry. b Determined by the experiments with DMA derivatives
of which methyl groups are not deuterated. c Not determined.
Scheme 2
k3H/k3D ) (kH/kD){(kD/kH) + (k-et/kH)}/{1 + (k-et/kH)} (1)
When electron transfer is the primary rate-determining step
(k-et/kH , 1), eq 1 reduces to k3H/k3D ) 1. On the other hand,
when k-et/kH . 1, eq 1 reduces to k3H/k3D ) kH/kD. Thus, the
difference between kH/kD and k3H/k3D should decrease with an
increase in k-et/kH. The k-et/kH value can be obtained from the
k3H/k3D and kH/kD values using eq 2, which is derived from eq 1.
Figure 1. Dependence of kinetic values on the oxidation potential of
DMAs (E0). The slopes of each line are shown in parentheses. (a)
(diamond): bimolecular rate constant of HRP compound I with DMAs,
k1. (b) (circle): bimolecular rate constant of OdFeIVTMP•+ with DMAs,
k3. (c) (triangle): k-et/kH determined by eq 2.
k-et/kH ) {(k3H/k3D) - 1}/{(kH/kD) - (k3H/k3D)} (2)
The k-et value for the back electron transfer from OdFeIVTMP
to DMA•+ is expected to increase with an increase in E0 , the
one-electron reduction potential of DMA•+. On the other hand,
the H-atom transfer rate constant (kH) may be less sensitive to E0
than the back electron-transfer rate constant (k-et).23 Thus, the
k-et/kH value may increase with an increase in the E0 value. In
fact, such a correlation between k-et/kH and E0 is demonstrated
in Figure 1c.
In conclusion, our results are in agreement with the currently
accepted mechanism in which the N-demethylation of DMAs by
HRP compounds I and II both proceed via a rate-determining
electron-transfer step. Our results suggest that demethylation by
OdFeIVTMP•+ also proceeds via electron transfer; however,
H-atom transfer from DMAs•+ to OdFeIVTMP competes with
back electron transfer, and this competition is responsible for the
observed deuterium isotope effects. The difference between the
kinetic and product isotope effects depends on the ratio of the
rate constant of the back electron transfer to that of the hydrogen
transfer.
The parallel relationship between these two plots indicates that
electron transfer from the DMAs is similar for HRP and the model
complex. However, the KIEs are observed only for the reactions
of OdFeIVTMP•+. The k3H/k3D value increases with an increase
in the E0 value to reach a value of 5.9 ( 0.4 in the case of the
p-nitro derivative. More importantly, there is a significant
difference between the kinetic (k3H/k3D) and product (kH/kD) isotope
effects. Coupled with the fact that no compound II is seen as an
intermediate, these isotope effects suggest that there is a significant
reverse electron transfer, so that any compound II-aminium
radical pair formed can partition either toward products (kH) or
to initial reactants (k-et) as shown in Scheme 2. The redox
potential of the DMA is likely to have a significant effect on
k-et. Using a steady-state assumption (the compound II-radical
pair is “constant” and at a low concentration), one can derive an
expression (eq 1) that relates the KIEs (k3H/k3D) and the product
isotope effects (kH/kD). The product isotope effects reflect the
H-atom transfer step from DMA•+ to OdFeIVTMP.
(21) Kinetic experiments on the iron-TMP complex were performed with
a Hi-Tec SF-43 sequential-mixing stopped-flow instrument. In a typical run,
a CH2Cl2 solution of FeIIITMP(OH) (40 µM) was mixed with an equal volume
of CH2Cl2 containing 4 equiv of mCPBA and 1 equiv of 3-chlorobenzoic
acid at 223 K. Formation of 1 was complete in 3 min and confirmed by its
UV-vis spectrum. DMA (10-100 equiv) in CH2Cl2 was then mixed with 1
(final concentration of 1, 10 µM), a very rapid isosbestic change of 1 to FeIII-
TMP was observed (Supporting Information). The reaction rate constant (k3)
was determined from analysis of the change in absorbance at 413 nm.
(22) KIEs by product analysis were determined by GC-MS analyses of
single turnover reactions. A 1.0 mM solution of FeIIITMP(OH) was added to
a CH2Cl2 solution containing 1.2 equiv of mCPBA at 223 K. After confirmation
of OdFeIVTMP+• formation, a CH2Cl2 solution of 10 equiv of a DMA(-CH3,
-CD3) derivative was added. After the solution color changed to brown,
trifluoroacetic anhydride was added, and the N-methylaniline derivative was
detected as a trifluoroacetylamide form. The kH/kD value was determined with
GC/SIM from the M+ peak area ratio of trideuteriomethyl- and methyl-
trifluoroacetylaniline (ref 16).
Acknowledgment. This work was partly supported by JSPS-NSF Joint
Program, Photo-Induced Charge Transfer, and also supported by a Grant-
in-Aid for Priority Area, Molecular Biometallics to Y.W. from the
Ministry of Education, Science, Sport, and Culture in Japan.
Supporting Information Available: UV-vis spectral changes for
the reactions of DMAs with both HRP compound I and OdFeIVTMP•+
,
derivation of eq 1 (4 pages, print/PDF). See any current masthead page
for ordering information and Web access instructions.
JA981357U
(23) The energetics of hydrogen transfer consist of those for transfer of a
proton and for an electron. The substitution effects on the proton- and electron-
transfer processes may be largely canceled out since the acidity of DMA•+
increases but the donor ability of DMA(-H)• decreases with an increase E0.