Formation of a quinoneimine intermediate of 4-Fluoro- N -methylaniline by FMO1: Carbon oxidation plus defluorination

Article abstract of DOI:10.1021/tx1000688

Here, we report on the mechanism by which flavin-containing monooxygenase 1 (FMO1) mediates the formation of a reactive intermediate of 4-fluoro-N-methylaniline. FMO1 catalyzed a carbon oxidation reaction coupled with defluorination that led to the formation of 4-N-methylaminophenol, which was a reaction first reported by Boersma et al. (Boersma et al. (1993) Drug Metab. Dispos. 21, 218 -230). We propose that a labile 1-fluoro-4-(methylimino) cyclohexa-2,5-dienol intermediate was formed leading to an electrophilic quinoneimine intermediate. The identification of N-acetylcysteine adducts by LC-MS/MS and NMR further supports the formation of a quinoneimine intermediate. Incubations containing stable labeled oxygen (H₂¹⁸O or ¹⁸O₂) and ab initio calculations were performed to support the proposed reaction mechanism.

Full text of DOI:10.1021/tx1000688

Chem. Res. Toxicol. 2010, 23, 861–863
861
Formation of a Quinoneimine Intermediate of
4-Fluoro-N-methylaniline by FMO1: Carbon Oxidation Plus
Defluorination
James P. Driscoll,* Ignacio Aliagas,‡ Jennifer J. Harris, Jason S. Halladay,
Sheerin Khatib-Shahidi, Alan Deese,† Nathaniel Segraves,† and S. Cyrus Khojasteh-Bakht
Departments of Drug Metabolism and Pharmacokinetics, Small Molecule Analytical Chemistry, and Computational
Chemistry & Cheminformatics, Genentech, Inc., 1 DNA Way, South San Francisco, California 94080
ReceiVed February 22, 2010
Here, we report on the mechanism by which flavin-containing monooxygenase 1 (FMO1) mediates the
formation of a reactive intermediate of 4-fluoro-N-methylaniline. FMO1 catalyzed a carbon oxidation reaction
coupled with defluorination that led to the formation of 4-N-methylaminophenol, which was a reaction first
reported by Boersma et al. (Boersma et al. (1993) Drug Metab. Dispos. 21, 218-230). We propose that a
labile 1-fluoro-4-(methylimino)cyclohexa-2,5-dienol intermediate was formed leading to an electrophilic
quinoneimine intermediate. The identification of N-acetylcysteine adducts by LC-MS/MS and NMR further
supports the formation of a quinoneimine intermediate. Incubations containing stable labeled oxygen
(H218O or 18O2) and ab initio calculations were performed to support the proposed reaction mechanism.
4-Hydroxylanilines and analogues are known to convert to
quinoneimine intermediates, either chemically or enzymatically
(1), that can lead to organ toxicity. For example, amodiaquine
undergoes bioactivation to an electrophilic quinoneimine me-
tabolite that results in oxidative stress or conjugation to cysteinyl
sulfydryl groups on proteins that elicits toxicity by either
cytotoxic or immunological mechanisms (2-5). In the clinical
setting, the use of amodiaquine has been limited due to severe
adverse reactions, including hepatotoxicity (6-10). Replacing
the para hydroxyl group in the case of amodiaquine with a
fluorine atom diminishes such toxicity (2, 11). The p-fluoroa-
niline analogue of amodiaquine blocks the oxidation to the
reactive quinoneimine that results in the depletion of intracellular
glutathione and covalent binding to proteins (11, 12). The
rationales are that the p-fluoro substitution decreased the electron
density in the benzene ring and that the fluoro group could block
the metabolically labile site at the para position (13). There
are several reports where the para substituted fluorine in aniline
containing compounds are oxidized to form a phenol, and it is
known that these reactions involve cytochrome P450 or per-
oxidases (14, 15). Boersma et al. (16) demonstrated that in the
case of 4-fluoro-N-methylaniline (1), FMO also catalyzes this
reaction to form 4-(methylamino)phenol (2) in purified rat liver
FMO. The focus of this work was to understand the underlying
mechanism of the oxidative defluorination of 1 by FMO.
Incubation of 1 (1 mM) with rat liver microsomes (0.5 mg/
mL in 0.5 mL of 100 mM potassium phosphate buffer) and
NADPH (1 mM) led to the formation of 2 in an NADPH- and
time-dependent manner. In the presence of an inactivator of
cytochrome P450 (preincubation with rat liver microsomes,
NADPH, and 1 mM 1-aminobenzotriazole (ABT) for 15 min)
or FMO inhibitor (coincubation with 0.1 mM methimazole),
only methimazole inhibited the formation of 2, which indicated
that FMO was the major enzyme system responsible for the
formation of this metabolite in rat liver microsomes. The
identification of 2 as 4-(methylamino)phenol was confirmed by
comparing the LC retention time (3.4 min), the product ion
fragmentation pattern (m/z 53, 80, 109), and high resolution mass
spectra ([MH]+ ) 124.07535 within 3 ppm of the exact mass
of C7H10NO+) to the authentic standard.
Furthermore, incubations with recombinant human FMO
(rFMO) enzymes were performed. Of the three isoforms tested
(rFMO1, 3, and 5), only rFMO1 was responsible for the
formation of 2. In incubations with rFMO1 (0.5 mg/mL),
NADPH, and 1, high resolution LC-MS was used to identify
major metabolites, which included 4-(methylamino)phenol,
4-aminophenol, N-(4-fluorophenyl)-N-methylhydroxylamine, N-
(4-fluorophenyl)hydroxylamine, and 4-fluoroaniline. Compound
2 was the major metabolite on the basis of LC-MS peak area,
and its formation rate was calculated to be 1.34 nmol/min/mg
(20.1 µM in 30 min) using a standard curve for its quantification.
Although Boersma et al. used purified rat liver FMO (16) and
we used recombinant human FMO1, the same metabolites were
identified and presented in Figure 1.
To further understand the mechanism leading to the oxidative
defluorination of 1 mediated by FMO1, we investigated the
source of the oxygen atom that was incorporated in 2 by using
either stable labeled H218O or 18O2 in the incubation. In the
reaction mixtures containing 18O2, rFMO1, NADPH, and 1, the
incorporation of 18O in 2 was observed. The incorporation of
18O was calculated to be 91% when compared to the amount of
18O incorporated into the N-hydroxylation of 1 in the same
incubation, which was a known FMO catalyzed reaction (16).
The incorporation of 18O into 2 was confirmed by high resolution
mass spectra which had an exact mass of m/z 126.07950 (within
4 ppm of C7H10N18O+). In the reaction mixtures containing
H218O, rFMO1, NADPH, and 1, no 18O labeled 2 was detected.
Additional incubations were performed to ensure that the
oxidative defluorination was not formed due to a side reaction
or a nonenzymatic process such as release of superoxide and
hydrogen peroxide which FMO has been shown to produce
(17). To assess this, incubations (1 (1 mM), rFMO1 (0.5 mg/
mL), and NADPH (1 mM)) were performed in the presence of
superoxide dismutase (500 units) or catalase (2000 units),
* To whom correspondence should be addressed. Tel: 650-225-8370.
Fax: 650-467-3487. E-mail: Driscoll.James@gene.com.
† Department of Small Molecule Analytical Chemistry.
‡ Department of Computational Chemistry & Cheminformatics.
10.1021/tx1000688  2010 American Chemical Society
Published on Web 04/06/2010

862 Chem. Res. Toxicol., Vol. 23, No. 5, 2010
Driscoll et al.
quinoneimine (4). The quinoneimine was then reduced to
generate the final product 2. This reaction mechanism, depicted
in Scheme 1, is consistent with reported FMO reactions and is
reasonable for nucleophilic moieties.
To test our hypothesis for the formation of the reactive
intermediate 4, incubations with rFMO1, NADPH (1 mM), and
1 (1 mM) were carried out in the presence of a nucleophilic
thiol, N-acetylcysteine (NAC; 5 mM). If 4 was formed and has
a reasonable half-life, it should form the NAC conjugates 5 and
6 (similar to NAPQI in the case of acetaminophen (20)). In
fact, we detected two NAC conjugates that correspond to 5 and
6 within 2 ppm by high resolution mass spectrometry (m/z
285.08987). The formation of these conjugates was further
confirmed by high resolution product ion scans and was within
4 ppm of the expected mass accuracy. The same product ions
were detected for 5 and 6, and the high resolution mass spectra
are found in Supporting Information. When glutathione (GSH)
was used as the trapping agent instead of NAC, only trace
amounts of GSH conjugates were detected. This was consistent
with a previous finding that GSH cannot access the active site
of FMO (21). Therefore, we proposed that this may be the
reason why GSH does not react with 4. Perhaps this is a
protective feature of the active site of FMO from futile cycling
of GSH oxidation. Further investigation is needed to prove this
hypothesis. The trace amounts that were found could be due to
GSH reacting with 2 after it was released from the active site.
Figure 1. Composite extracted ion chromatograms of the metabolites
of 1 incubated with rFMO after 30 min. (a) (+) NADPH and (+) NAC,
(b) (+) NADPH and (-)NAC, and (c) (+) NADPH, (+) NAC, and
superoxide dismutase (500 units). The metabolites include 4-aminophe-
nol (m/z 110.06004, RT ) 2.2), 4-(methylamino)phenol (m/z 124.07569,
RT ) 3.4), N-(4-fluorophenyl)hydroxylamine (m/z 128.05062, RT )
4.3), 4-fluoroaniline (m/z 112.05570, RT ) 4.8), N-(4-fluorophenyl)-
N-methylhydroxylamine (m/z 142.06627, RT ) 6.8), and N-acetylcys-
teine conjugates (m/z 285.09035, RT ) 5.9 and 6.3).
To characterize the regiochemistry of the conjugates, synthetic
2 (1 mM in 3 mL of 100 mM potassium phosphate buffer) was
prepared by chemical reaction of NAC (5 mM) under basic
conditions (pH 8) to facilitate the formation of 4 for 24 h (22).
The samples were analyzed by LC-MS, and peaks containing
NAC adducts at m/z 285 were fraction collected. The resulting
fractions were dried under vacuum, dissolved in DMSO-d6 (D,
99.8%, Cambridge Isotope) containing 0.05% V/V TMS as an
internal chemical shift reference standard, and transferred to 3
mm NMR tubes (Norell, S-3-600-SC-7), purged with nitrogen,
and sealed. A combination of many NMR methods (NOE and
HMBC) were performed to confirm the regiochemistry of the
two NAC conjugates using a Bruker Avance 3, 600 MHz
spectrometer equipped with a 5 mm, TCI, Z gradient CryoProbe
(see Supporting Information). Interestingly, when GSH was used
instead of NAC under the same conditions, two GSH adducts
were detected by mass spectrometry, which further supports our
hypothesis that GSH does not access the active site of FMO1
to interact with 4.
Scheme 1
Various computational studies on the behavior of peroxy
groups, including the FAD-OOH group, have been published
to confirm the proposed viability or mechanism of reactions.
These computational studies have been carried out using
semiempirical, ab initio, and density functional theory regarding
the reactivity, transition state geometry, and activation barrier
of FAD reacting with various substrates (23). To further support
our mechanism in Scheme 1, an ab initio study of the frontier
molecular orbital analysis (24, 25) was performed on p-fluoro-
N-methylaniline. The mechanism of nucleophilic substitution
is consistent with the frontier molecular orbital analysis, which
showed the fluorine having a strong inductive effect in extracting
electron density from the aromatic ring and the directly attached
ipso carbon bearing the largest positive calculated atomic
electrostatic potential (ESP) charge of all the carbons in the
phenyl ring. The frontier molecular orbital analysis showed that
the carbon with the highest HOMO (highest occupied molecular
orbital) coefficient (Figure 2) being consistent with the para-
carbon initiating the nucleophilic attack. These calculations were
respectively. Under these conditions, the formation of 2 was
not inhibited, which further supports the fact that the oxidative
defluorination was an enzymatic process.
The FMO-mediated reaction to form 4-(methylamino)phenol
is notable because it includes carbon oxidation and defluorina-
tion as well as the formation of a reactive quinoneimine
intermediate. FMO enzymes are well documented as catalyzing
the oxidation of soft nucleophilic heteroatoms such as sulfur
and nitrogen (18). In the catalytic mechanism of the enzyme,
incorporation of an oxygen atom into the substrate originates
from molecular oxygen (O2) that exists as a peroxyflavin group
(FAD-OOH) located in the active site of the enzyme (19). We
proposed that the lone pair of electrons from the aniline nitrogen
could delocalize and facilitate the para position carbon to attack
the distal oxygen on the FAD-OOH (Scheme 1). From this, a
1-fluoro-4-(methylimino)cyclohexa-2,5-dienol intermediate (3)
would be formed that could then eliminate HF to form a reactive

Rapid Report
Chem. Res. Toxicol., Vol. 23, No. 5, 2010 863
(3) Clarke, J. B., Maggs, J. L., Kitteringham, N. R., and Park, B. K. (1990)
Immunogenicity of amodiaquine in the rat. Int. Arch. Allergy Appl.
Immunol. 91, 335–342.
(4) Harrison, A. C., Kitteringham, N. R., Clarke, J. B., and Park, B. K.
(1992) The mechanism of bioactivation and antigen formation of
amodiaquine in the rat. Biochem. Pharmacol. 43, 1421–1430.
(5) Jewell, H., Maggs, J. L., Harrison, A. C., O’Neill, P. M., Ruscoe,
J. E., and Park, B. K. (1995) Role of hepatic metabolism in the
bioactivation and detoxication of amodiaquine. Xenobiotica 25, 199–
217.
(6) Neftel, K. A., Woodtly, W., Schmid, M., Frick, P. G., and Fehr, J.
(1986) Amodiaquine induced agranulocytosis and liver damage. Br.
Med. J. (Clin. Res. Ed.) 292, 721–723.
(7) Hatton, C. S., Peto, T. E., Bunch, C., Pasvol, G., Russell, S. J., Singer,
C. R., Edwards, G., and Winstanley, P. (1986) Frequency of severe
neutropenia associated with amodiaquine prophylaxis against malaria.
Lancet 1, 411–414.
(8) Larrey, D., Castot, A., Pessayre, D., Merigot, P., Machayekhy, J. P.,
Feldmann, G., Lenoir, A., Rueff, B., and Benhamou, J. P. (1986)
Amodiaquine-induced hepatitis. A report of seven cases. Ann. Intern.
Med. 104, 801–803.
(9) Phillips-Howard, P. A., and West, L. J. (1990) Serious adverse drug
reactions to pyrimethamine-sulphadoxine, pyrimethamine-dapsone and
to amodiaquine in Britain. J. R. Soc. Med. 83, 82–85.
(10) Raymond, J. M., Dumas, F., Baldit, C., Couzigou, P., Beraud, C., and
Amouretti, M. (1989) Fatal acute hepatitis due to amodiaquine. J. Clin.
Gastroenterol. 11, 602–603.
(11) O’Neill, P. M., Harrison, A. C., Storr, R. C., Hawley, S. R., Ward,
S. A., and Park, B. K. (1994) The effect of fluorine substitution on
the metabolism and antimalarial activity of amodiaquine. J. Med.
Chem. 37, 1362–1370.
Figure 2. Molecular representations HOMO orbitals and coefficients.
Calculations were performed at the HF/6-31+G** level using the
Gaussian 03 program. (a) Representation of the HOMO orbitals. The
tightly packed orbital around the fluorine atom indicates the lesser
conjugation of that atom to the ring, while the nitrogen shows strong
conjugation with the π-ring system. The blue and red color scheme
shows the molecular orbital wave function sign. (b) Representation of
the calculated HOMO orbital coefficients. The carbon next to the
fluorine bears the largest coefficient on the phenyl ring consistent with
it initiating the nucleophilic attack.
(12) Tingle, M. D., Jewell, H., Maggs, J. L., O’Neill, P. M., and Park,
B. K. (1995) The bioactivation of amodiaquine by human poly-
morphonuclear leucocytes in vitro: chemical mechanisms and the
effects of fluorine substitution. Biochem. Pharmacol. 50, 1113–
1119.
(13) Bohm, H. J., Banner, D., Bendels, S., Kansy, M., Kuhn, B., Muller,
K., Obst-Sander, U., and Stahl, M. (2004) Fluorine in medicinal
chemistry. ChemBioChem 5, 637–643.
in line with what O’Hagan (26) discusses in his review of the
carbon-fluorine bond. He states that the three lone pairs of
electrons on fluorine are held tightly and that the electrons in
the C-F bond are reluctant to get involved in any resonance
structures due to its high electronegativity.
The electron donation from the nitrogen to the ring has an
effect on lowering the HOMO and lowest unoccupied molecular
orbital (LUMO) energy gap, given that the simple fluoro phenyl
HOMO-LUMO energy gap was larger (0.42 eV vs 0.35 eV
for the p-fluoroaniline). It was also noteworthy that the
HOMO-LUMO gap for the p-fluoro-N-methylanaline was
larger (0.28 eV vs 0.35 eV) than that of p-fluoroaniline. This
was consistent with our observation that no carbon oxidation
was observed from an incubation of p-fluoroaniline and rFMO1.
In summary, we demonstrated an FMO-based carbon oxida-
tion plus defluorination reaction. Consistent with FMO reactions,
one oxygen atom was incorporated from O2. We propose that
the lone pair of electrons from the nitrogen facilitates the
nucleophilic attack by the C-1 carbon to the distal oxygen from
FAD-OOH followed by the loss of HF and the formation of 3.
The formation of the quinoneimine intermediate was confirmed
by trapping it with NAC and the detection of 5 and 6. To our
knowledge, this is the first time FMO has been shown to cause
the formation of a quinoneimine intermediate.
(14) Li, X., Kamenecka, T. M., and Cameron, M. D. (2009) Bioactivation of
the epidermal growth factor receptor inhibitor gefitinib: implications for
pulmonary and hepatic toxicities. Chem. Res. Toxicol. 22, 1736–1742.
(15) Chen, H., Shockcor, J., Chen, W., Espina, R., Gan, L. S., and Mutlib,
A. E. (2002) Delineating novel metabolic pathways of DPC 963, a
non-nucleoside reverse transcriptase inhibitor, in rats. Characterization
of glutathione conjugates of postulated oxirene and benzoquinone
imine intermediates by LC/MS and LC/NMR. Chem. Res. Toxicol.
15, 388–399.
(16) Boersma, M. G., Cnubben, N. H., van Berkel, W. J., Blom, M.,
Vervoort, J., and Rietjens, I. M. (1993) Role of cytochromes P-450
and flavin-containing monooxygenase in the biotransformation of
4-fluoro-N-methylaniline. Drug Metab. Dispos. 21, 218–230.
(17) Patton, S. E., Rosen, G. M., and Rauckman, E. J. (1980) Superoxide
production by purified hamster hepatic nuclei. Mol. Pharmacol. 18,
588–593.
(18) Cashman, J. R. (2005) Some distinctions between flavin-containing
and cytochrome P450 monooxygenases. Biochem. Biophys. Res.
Commun. 338, 599–604.
(19) Ziegler, D. M. (1993) Recent studies on the structure and function of
multisubstrate flavin-containing monooxygenases. Annu. ReV. Phar-
macol. Toxicol. 33, 179–199.
(20) Chen, W., Shockcor, J. P., Tonge, R., Hunter, A., Gartner, C., and
Nelson, S. D. (1999) Protein and nonprotein cysteinyl thiol modifica-
tion by N-acetyl-p-benzoquinone imine via a novel ipso adduct.
Biochemistry 38, 8159–8166.
Supporting Information Available: NMR data and NAC
adduct spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
(21) Poulsen, L. L., Ed. (1991) Chemistry and Biochemistry of FlaVoen-
zymes, Vol. 3, CRC Press, Boca Raton, FL.
(22) Dykens, J. A., Sullivan, S. G., and Stern, A. (1987) Oxidative reactivity
of the tryptophan metabolites 3-hydroxyanthranilate, cinnabarinate,
quinolinate and picolinate. Biochem. Pharmacol. 36, 211–217.
(23) Bach, R. D., Ed. (2006) General and Theoretical Aspects of the
Peroxide Group, Vol. 2, John Wiley & Sons, Ltd., New York.
(24) Herndon, W. C. (1972) Theory of cycloaddition reactions. Chem. ReV.
72, 157–179.
(25) Houk, K. N. (1975) The frontier molecular orbital theory of cycload-
dition reactions. Acc. Chem. Res. 8, 361–369.
(26) O’Hagan, D. (2008) Understanding organofluorine chemistry. An
introduction to the C-F bond. Chem. Soc. ReV. 37, 308–319.
References
(1) Kalgutkar, A. S., Gardner, I., Obach, R. S., Shaffer, C. L., Callegari,
E., Henne, K. R., Mutlib, A. E., Dalvie, D. K., Lee, J. S., Nakai, Y.,
O’Donnell, J. P., Boer, J., and Harriman, S. P. (2005) A comprehensive
listing of bioactivation pathways of organic functional groups. Curr.
Drug Metab. 6, 161–225.
(2) Maggs, J. L., Tingle, M. D., Kitteringham, N. R., and Park, B. K.
(1988) Drug-protein conjugates-XIV. Mechanisms of formation of
protein-arylating intermediates from amodiaquine, a myelotoxin and
hepatotoxin in man. Biochem. Pharmacol. 37, 303–311.
TX1000688