Cytochrome P450-catalyzed oxidation of N-benzyl-N-cyclopropylamine generates both cyclopropanone hydrate and 3-hydroxypropionaldehyde via hydrogen abstraction, not single electron transfer

Article abstract of DOI:10.1021/ja054938+

The suicide substrate activity of N-benzyl-N-cyclopropylamine (1) and N-benzyl-N-(1′-methylcyclopropyl)amine (2) toward cytochrome P450 and other enzymes has been explained by a mechanism involving single electron transfer (SET) oxidation, followed by rin

Full text of DOI:10.1021/ja054938+

Published on Web 02/15/2006
Cytochrome P450-Catalyzed Oxidation of
N-Benzyl-N-cyclopropylamine Generates Both
Cyclopropanone Hydrate and 3-Hydroxypropionaldehyde via
Hydrogen Abstraction, Not Single Electron Transfer
Matthew A. Cerny and Robert P. Hanzlik*
Contribution from the Department of Medicinal Chemistry, UniVersity of Kansas,
Lawrence, Kansas 66045
Received July 22, 2005; E-mail: rhanzlik@ku.edu
Abstract: The suicide substrate activity of N-benzyl-N-cyclopropylamine (1) and N-benzyl-N-(1′-methyl-
cyclopropyl)amine (2) toward cytochrome P450 and other enzymes has been explained by a mechanism
involving single electron transfer (SET) oxidation, followed by ring-opening of the aminium radical cation
(protonated aminyl radical) and reaction with the P450 active site. Although the SET oxidation of
N-cyclopropyl-N-methylaniline (3) by horseradish peroxidase leads exclusively to ring-opened (non-
cyclopropyl) products, P450 oxidation of 3 leads to formation of cyclopropanone hydrate and no ring-
opened products, and 3 does not inactivate P450. To help reconcile these discrepant behaviors we have
determined the complete metabolic fate of 1 with P450 in vitro. 3-Hydroxypropionaldehyde (3HP), the
presumptive “signature metabolite” for SET oxidation of a cyclopropylamine, was observed for the first
time in 57% yield, along with cyclopropanone hydrate (34%), cyclopropylamine (9%), benzaldehyde (6%),
benzyl alcohol (12%), and benzaldoxime (19%). Unexpectedly, N-benzyl-N-cyclopropyl-N-methylamine (4)
was found not to inactivate P450 and not to give rise to 3HP as a metabolite without first undergoing
oxidative N-demethylation to 1. These and other observations argue against a role for SET mechanisms
in the P450 oxidation of cyclopropylamines. We suggest that a conventional hydrogen abstraction/hydroxyl
recombination mechanism (or its equivalent as a one-step “insertion” mechanism) at C-H bonds in 1-4
leads to nonrearranged carbinolamine intermediates and thereby to “ordinary” N-dealkylation products
including cyclopropanone hydrate. Alternatively, hydrogen abstraction at the N-H bond of secondary
cyclopropylamines 1 gives a neutral aminyl radical which could undergo rapid ring-opening leading either
to enzyme inactivation or 3HP formation.
Introduction
Chart 1. Structures of Compounds 1-11
The cyclopropylamine substructure is found in several
biologically active natural products and is increasingly found
within synthetic drug and drug candidate molecules. Since the
initial reports that N-benzyl-N-cyclopropylamine (1, Chart 1)
1
is a suicide substrate for cytochrome P450 enzymes, as well
as for monoamine oxidase,2 cyclopropylamines have been
widely applied as mechanistic probes of these and other
oxidative enzymes.3
-12
The observation that compound 2, the
(
1) Hanzlik, R. P.; Kishore, V.; Tullman, R. J. Med. Chem. 1979, 22, 760-
7
61.
(
2) Silverman, R. B.; Hoffman, S. J. J. Am. Chem. Soc. 1980, 102, 884-886.
(
3) Shah, M. A.; Trackman, P. C.; Gallop, P. M.; Kagan, H. M. J. Biol. Chem.
1′-methyl derivative of 1, inactivates both P4501
3,14
and
1
993, 268, 11580-11585.
(
4) Kuttab, S.; Shang, J.; Castagnoli, N., Jr. Bioorg. Med. Chem. 2001, 9, 1685-
689.
5) Bhakta, M. N.; Hollenberg, P. F.; Wimalasena, K. J. Am. Chem. Soc. 2005,
15
monoamine oxidase (MAO) nearly as effectively as 1 ruled
1
(
1
27, 1376-1377.
(10) Mitchell, D. J.; Nikolic, D.; Jang, M.-H.; van Breeman, R. B.; Hille, R.;
Silverman, R. B. Biochemistry 2001, 40, 8523-8530.
(
6) Bhakta, M. N.; Wimalasena, K. J. Am. Chem. Soc. 2002, 124, 1844-1845.
7) Cooper, A. B.; Wright, J. J.; Ganguly, A. K.; Desai, J.; Loebenberg, D.;
Parmegiani, R.; Feingold, D. S.; Sud, I. J. J. Chem. Soc., Chem. Commun.
(
(11) Vint e´ m, A. P. B.; Price, N. T.; Silverman, R. B.; Ramsay, R. R. Biochem.
Biophys. Res. Commun. 2005, 13, 3487-3495.
1
989, 898-900.
(12) Totah, R. A.; Hanzlik, R. P. J. Am. Chem. Soc. 2002, 124, 10000-10001.
(13) Hanzlik, R. P.; Tullman, R. H. J. Am. Chem. Soc. 1982, 104, 2048-2050.
(14) Macdonald, T. L.; Zirvi, K.; Burka, L. T.; Peyman, P.; Guengerich, F. P.
J. Am. Chem. Soc. 1982, 104, 2050-2052.
(
8) Angelastro, M. R.; Laughlin, M. E.; Schatzman, G. L.; Bey, P.; Blohm, T.
R. Biochem. Biophys. Res. Commun. 1989, 162, 1571-1577.
(
9) Njar, V. C. O.; D u¨ erkop, J.; Hartman, R. W. J. Enzyme Inhib. 1995, 10,
4
7-56.
(15) Silverman, R. B.; Yamanski, R. B. Biochemistry 1984, 23, 1322-1332.
3346
9
J. AM. CHEM. SOC. 2006, 128, 3346-3354
10.1021/ja054938+ CCC: $33.50 © 2006 American Chemical Society

P450 Oxidation of Cyclopropylamines
A R T I C L E S
Scheme 1. Hydrogen Atom Transfer (HAT) Mechanism
Scheme 3. SET versus P450 Oxidation of Probe 3
Marcus analysis of similar data suggests that the active oxygen
2
4,25
species of P450, presumed
to be the ferryl or iron-oxene
3+
complex FeO , has a very much higher reduction potential than
compound I of horseradish peroxidase, a heme enzyme well-
known to perform single-electron oxidations of its substrates.
While these observations appear to make a persuasive
argument in favor of a role for SET reactions in the P450-
catalyzed oxidations of amines (and perhaps other substrates),
Scheme 2. Single Electron Transfer (SET) Mechanism
2
6
more careful consideration indicates that alternate explanations
are possible, even probable, in some of these cases. Evidence
against the role of SET reactions in the P450-catlalyzed
N-dealkylations comes from studies of the regioselectivity of
N-dealkylation of N-ethyl-N-fluoroethylanilines Ph(Et)NCH2-
2
7
CHnF3-n. With HRP, loss of the fluoroethyl group increases
with the degree of fluorine substitution and the acidity of the R
hydrogens, indicating an important role for a C-deprotonation
reaction following the initial SET oxidation at nitrogen. In
contrast the opposite trend is observed with P450; loss of the
fluoroethyl group decreases from 68% to only 8% across the
series from -CH2CH2F to CH2CF3.
Clearly, the role of SET mechanisms in P450-catalyzed
N-dealkylations in general, and more specifically their corollary
role in the suicide substrate activity of cyclopropylamines, is
far from settled. Until recently, one very important type of
evidence relevant to this latter question had been entirely
missing, namely, the unambiguous identification of the fate of
the three cyclopropyl carbons lost from the nitrogen center.
out the involvement of N-cyclopropylidene Schiff base inter-
mediates in the inactivation process (Scheme 1) and focused
attention on the single electron transfer (SET) mechanism shown
in Scheme 2. Cyclopropyl aminium ions16 are known to undergo
rapid unimolecular ring-opening rearrangement to distonic
iminium ion-radical species.1 The latter have been suggested
to account for the enzyme-inactivating ability of cyclopropyl-
amines versus other types of amines in this regard (Scheme 2).
In addition to the suicide substrate activity of 1 and especially
7,18
2
8,29
Shaffer et al. first reported
that the oxidation of N-methyl-
N-cyclopropylaniline (3) by horseradish peroxidase (HRP), a
well-known SET oxidant, leads exclusively to fragmentation
of the cyclopropane ring. The initial nitrogen-centered cation
radical fragments to a distonic open-chain form which then
recyclizes via radical addition to the phenyl ring; further
oxidation generates the ultimate product, N-methylquinolinium,
quantitatively (Scheme 3). In sharp contrast, the P450-catalyzed
oxidation of 3 was shown to lead exclusively to ring-intact
metabolites including cyclopropanone hydrate, N-cyclopropy-
laniline, and p-hydroxy-3, with no signs of either acrolein or
2
toward cytochrome P450, several other observations have also
been taken as support for an SET mechanism for the P450-
catalyzed oxidation of amines.19 For example, P450 oxidation
of 4-alkyl-1,4-dihydropyridine derivatives leads to the extrusion
of alkyl radicals that can be trapped by the P450 heme (leading
2
0
to suicide inactivation) or by spin-traps in solution. An SET
mechanism was also invoked to explain the uniformly low
kinetic deuterium isotope effects observed for P450-catalyzed
N-dealkylation of amines and anilines.21 In addition, the P450-
inactivating potential of cyclopropylamines and other hetero-
atom-substituted cyclopropanes correlates to their one-electron
oxidation potentials.22 The P450-catalyzed N-demethylation of
p-substituted N,N-dimethylanilines shows a Hammett F value
of -0.6, indicating the development of partial positive charge
in the transition state for carbinolamine formation.2 Finally,
3
-hydroxypropionaldehyde (3HP) and no signs of P450 inac-
30
tivation by 3, as would be predicted by Scheme 2.
In view of the disparate behavior of cyclopropyl probe 3 with
SET versus P450 enzymes, we undertook a detailed exploration
of the interaction of 1 with P450 enzymes, both as a substrate
and as an enzyme inactivator. For this work we used the original
source of P450 enzymes in which the suicide substrate activity
3
(
23) Burka, L. T.; Guengerich, F. P.; Willard, R. J.; Macdonald, T. L. J. Am.
(
2
16) The neutral species R N‚ (with three electrons on nitrogen) is an aminyl
Chem. Soc. 1985, 107, 2549-2551.
radical, while its protonated form (a nitrogen radical cation) is an aminium
ion.
(24) Macdonald, T. L.; Gutheim, W. G.; Martin, R. B.; Guengerich, F. P.
Biochemistry 1989, 28, 2071-2077.
(
(
17) Qin, X.-Z.; Williams, F. J. Am. Chem. Soc. 1987, 109, 595-597.
(25) Dowers, T. S.; Rock, D. A.; Rock, D. A.; Jones, J. P. J. Am. Chem. Soc.
2004, 126, 8868-8869.
18) Bouchoux, G.; Alcaraz, C.; Dutuit, O.; Nguyen, M. T. J. Am. Chem. Soc.
1
998, 120, 152-160.
(26) Karki, S. B.; Dinnocenzo, J. P. Xenobiotica 1995, 25, 711-724.
(27) Bhakta, M.; Hollenberg, P. F.; Wimalasena, K. Chem. Commun. 2005, 265-
267.
(19) Guengerich, F. P.; Yun, C. H.; Macdonald, T. L. J. Biol. Chem. 1996,
71, 27321-27239.
(20) Augusto, O.; Beilan, H. S.; Ortiz de Montellano, P. R. J. Biol. Chem. 1982,
57, 11288-11295.
(21) Miwa, G. T.; Walsh, J. S.; Kedderis, G. L.; Hollenberg, P. F. J. Biol. Chem.
(
2
(28) Shaffer, C. L.; Morton, M. D.; Hanzlik, R. P. J. Am. Chem. Soc. 2001,
123, 8502-8508.
2
(29) Shaffer, C. L.; Morton, M. D.; Hanzlik, R. P. J. Am. Chem. Soc. 2001,
123, 349-350.
1
983, 258, 14445-14449.
22) Guengerich, F. P.; Willard, R. J.; Shea, J. P.; Richards, L. E.; McDonald,
(30) Shaffer, C. L.; Harriman, S.; Koen, Y. M.; Hanzlik, R. P. J. Am. Chem.
Soc. 2002, 124, 8268-8274.
T. L. J. Am. Chem. Soc. 1984, 106, 6446-6447.
J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006 3347

A R T I C L E S
Cerny and Hanzlik
of 1 was discovered, namely, liver microsomes from phenobar-
bital-treated rats (PB-microsomes), as well as several expressed,
purified, and reconstituted P450 isoforms. Based on this work
we recently reported that a significant portion of the inactivation
of P450 enzymes during metabolism of 1 occurs via formation
of a robust, spectroscopically distinct metabolic intermediate
complex between R-nitrosotoluene (a metabolite of 1 but not
of benzylamine) and the heme iron of P450.31 In this Article
we describe the complete metabolic fate of 1 in PB-microsomes.
Significantly, we report for the first time that 3HP, long
anticipated as the “signature” metabolite for SET oxidation of
a cyclopropylamine by cytochrome P450 (Scheme 2), is indeed
a major metabolite of 1. Surprisingly, however, tertiary amine
4
, the N-methyl derivative of 1, is not an inactivtor of P450
until it is N-demethylated to form 1. These and other observa-
tions are discussed in terms of SET versus HAT mechanisms
of stable metabolite formation as well as P450 inactivation by
cyclopropylamines.
Figure 1. Kinetics of oxidation of 1 by PB-microsomes.
(imine) carbon of the authentic DNP derivative (152.6 ppm in
MeOH-d4); this peak was absent in extracts from control
incubations. Thus, it is clear that 3HP was formed as a
microsomal metabolite of 1; as described below, 3HP proved
to be a major metabolite of 1.
Results
Metabolite Identification. Since previous work in our
laboratory28-30
13
In parallel with the above experiments, a microsomal incuba-
demonstrated the power and convenience of
C
1
3
13
tion of [7- C]-1 was carried out and analyzed by C NMR.
NMR for identification of metabolites of 3 in quenched
After incubation, acidification to pH 1 and extraction with ether
incubation mixtures, we adopted a similar approach for the
1
3
13
to remove acidic and neutral metabolites, the C NMR of the
aqueous incubation mixture (Figure 3S in the Supporting
Information) showed peaks corresponding to the nonenriched
carbon atoms of unmetabolized substrate at 3.1, 29.6, and
analysis of metabolites of 1. Thus, comparison of the C NMR
1
3
spectrum of a complete microsomal incubation of [1′- C]-1
Figure 1S in the Supporting Information) to that of a blank
(
incubation in which only substrate was omitted (Figure 2S in
the Supporting Information) readily revealed a major peak from
the C-1′ carbon of unmetabolized substrate (29.5 ppm) plus
smaller peaks for the nonenriched carbons (i.e., the aromatic
carbons at 128.6-130.6 ppm, the benzylic carbon at 51.5 ppm,
and C-2′ and C-3′ of the cyclopropyl group which appear as a
doublet (due to the 13C-enrichment at C-1′; J ) 10 Hz) at 3.00
1
28.7-130.4 ppm, as well as a large peak for the benzylic
carbon of the substrate (51.7 ppm). The only metabolite peak
detected in the aqueous concentrate was benzylamine, identified
by its chemical shift (43.2 ppm). Analysis of the ether extract
1
3
by C NMR revealed three significant peaks at 65.5, 149.1,
and 192.2 ppm that were not present in blank incubations (data
not shown). The peaks at 65.5 and 192.2 ppm were consistent
with benzyl alcohol and benzaldehyde, respectively, but the
identity of the peak at 149.1 ppm was not immediately apparent.
Hence, the ether extract was submitted to analysis by GC/MS,
which yielded three peaks having retention times of 5.8, 6.9,
and 9.4 min associated with m/z values of 107, 109, and 122,
respectively. The injection of authentic standards of unlabeled
benzaldehyde and benzyl alcohol resulted in peaks with retention
times of 5.8 and 6.9 min and m/z values of 106 and 108,
ppm). Closer examination revealed five additional peaks at 22.8,
7
9.4, 88.1, 88.7, and 206.9 ppm that were not present in the
blank and were not associated with the substrate. Two peaks at
2.7 and 79.4 ppm were identified as arising from cyclopro-
2
pylamine and cyclopropanone hydrate, respectively, by their
chemical shifts and by spiking the incubation sample with
authentic standards and rerecording the spectrum (not shown).
Two peaks at 88.1 and 206.9 ppm, which were more intense
than those at 22.7 or 79.4 ppm, were initially puzzling. However,
the carbonyl carbon of 3HP and its corresponding gem-diol
hydrate have been reported to have chemical shifts of 206.3
1
3
respectively (reflecting the absence of C-enrichment in the
standards). The apparent molecular weight of the 9.4 min peak,
1
3
32
together with its large C chemical shift, led us to suspect that
it might be benzaldoxime, and comparison with an authentic
standard confirmed this identification. The appearance of
benzaldoxime among the metabolites of 1 was unexpected and
and 88.6 ppm, respectively. To confirm the formation of 3HP
as a metabolite of 1 the incubation mixture was treated with
2
9
2
,4-dinitrophenylhydrazine (DNPH) reagent, extracted, and the
extract was analyzed by HPLC, 3C NMR, and LC-MS
1
31
led us to investigate separately the mechanism of its formation
and the role of its nitroso tautomer in the inactivation of P450
by 1.
comparison to an authentic standard. HPLC analysis of incuba-
tion extracts showed a peak at 11.5 min, the tR of the standard,
that increased with incubation time. LC-MS analysis of material
collected from this HPLC peak from incubations of 1 and [1′-
Quantitation of Metabolites. Incubation of 1 with PB-
microsomes resulted in a rapid initial rate of consumption that
decreased gradually but did not stop, even after 30 min when
the rate had slowed to e10% of the initial rate (Figure 1). A
replot of log[1] versus time (not shown) is strongly convex
downward, reflecting substantial enzyme inactivation concurrent
with substrate depletion. Despite the enzyme inactivation taking
place during turnover of 1, approximately 20% of the substrate
(initially 1 mM) is consumed over 60 min of incubation.
1
3
C]-1 showed apparent molecular ions of m/z 255 and 256,
respectively, and fragmentation patterns identical to that of the
13
authentic standard (m/z 255). Finally, the C NMR of the extract
13
showed a single C-enriched peak corresponding to the carbonyl
(
31) Cerny, M. A.; Hanzlik, R. P. Arch. Biochem. Biophys. 2005, 436, 265-
75.
32) Vollenweider, S.; Grassi, G.; Puhan, Z. J. Agric. Food Chem. 2003, 51,
287-3293.
2
(
3
3348 J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006

P450 Oxidation of Cyclopropylamines
A R T I C L E S
Scheme 4. Metabolites of 1 Formed by PB-Microsomes
Figure 2. Formaldehyde production from microsomal incubation of 4 (500
µM) at 33 °C in the absence (curve A, closed squares) and presence (curve
B, open triangles) of 250 µM 1. The effect of preincubating microsomes
plus NADPH in the absence and presence of 1 (250 µM) for 10 min prior
to adding substrate 4 is shown by curves C (open circles) and D (closed
circles), respectively. For each data point a 500 µL aliquot of the incubation
mixture was removed and analyzed for formaldehyde formation as described
in the Methods.
To quantitate the cyclopropane-derived metabolites identified
by C NMR experiments the microsomal metabolism of [1′-
C]-1 (initial concentration 1.0 mM) was monitored by
13
1
4
reversed-phase HPLC with radiochemical detection. All non-
substrate-associated radioactivity was found to elute in the
solvent front, indicating the highly hydrophilic nature of the
metabolites, i.e., cyclopropylamine, cyclopropanone hydrate, and
continuously due to enzyme inactivation (see above), the
3
HP. Integration of the solvent front 14C peak area showed these
disappearance of 4 followed strictly linear (zero-order) kinetics
1
3
metabolites to be present at a total concentration of 199 µM
after 60 min, which agrees with the 20% consumption of
substrate mentioned above. Because HPLC failed to resolve the
solvent front constituents, quantitation of the individual me-
tabolites necessitated prior derivatization to enable separation
throughout the 60 min incubation. C NMR studies showed
that compound 1, cyclopropanone hydrate, and N-methylcyclo-
1
3
propylamine were the only C-enriched metabolites of 4. The
gem-diol hydrate of formaldehyde was also clearly detectable
1
3
at 85.0 ppm, even though this carbon was not C-enriched.
Thus, the principal route of microsomal metabolism of 4 is
N-dealkylation, with N-demethylation to form 1 greatly pre-
dominating.
(
see Methods). Thus, cyclopropylamine, analyzed as its 4-ni-
33
trobenzamide derivative, accounted for 9% of total metabolites,
while 3HP, analyzed as its DNP derivative, accounted for 57%
of total metabolites; cyclopropanone hydrate was estimated by
subtraction to account for 34% of total metabolites. Similar
Since 1 is an inactivator of microsomal P450, yet 4 is oxidized
to 1 with linear (i.e., zero-order) kinetics, we also examined
the interaction of these compounds with P450 by direct
photometric titration and by coincubation under several condi-
tions. Interestingly, microsomal metabolism of 4 in the absence
and presence of 1 showed that the presence of 1 had no effect
of N-demethylation of 4 (Figure 2, curves A and B). However,
preincubation of PB-microsomes with 1 alone, followed by
addition of 4, led to a considerable decrease in N-demethylation
activity toward 4 compared to a control in which the preincu-
bation was performed in the absence of 1 (Figure 2, curves D
and C, respectively). These results suggest that 4 is indeed not
an inactivator of P450 but first must be converted to 1 before
P450 inactivation occurs and that the presence of substrate 4
actually protects P450 from inactivation by 1. To determine if
14
studies using [7- C]-1 showed that after 60 min of incubation,
benzylamine was also a major (63%) metabolite while benzyl
alcohol, benzaldehyde, and benzaldoxime accounted for 12%,
6
%, and 19% of total metabolites, respectively). Thus the overall
metabolism of 1 is as shown in Scheme 4.
Comparative Metabolism of 1 and Its N-Methyl Derivative
4). The observed formation of substantial amounts of the ring-
(
opened metabolite 3HP was consistent with the possible
involvement of an SET mechanism in the microsomal metabo-
lism of 1 (viz., Scheme 2). To explore this possibility further,
we submitted the tertiary cyclopropylamine 4 (1 mM) to
metabolism by PB-microsomes. Because tertiary amines are
much more susceptible than secondary or primary amines to
oxidation by electron-abstracting agents,3 we anticipated that
4,35
4
had a higher affinity for P450 than 1 we examined their
4
should be at least as susceptible to SET oxidation as 1 and
interaction with P450 directly by spectrophotometric titration.
Compound 4 interacted very strongly with P450 in PB-
microsomes and showed a classical type I (substrate-like)
binding spectrum and an apparent dissociation constant (Ks) of
that it would therefore also give a good yield of 3HP metabolite.
At an initial concentration of 1 mM, compound 4 was an
excellent substrate for metabolism by microsomal P450, being
5
0% oxidized after 60 min, even with microsomes from
8
.3 µM (data not shown). Such difference spectra arise from
noninduced rats. Surprisingly, however, microsomal metabolism
ligand-induced displacement of an iron-coordinated water
molecule and are quite distinct from the type II spectral change
which occurs with nitrogenous ligands that coordinate directly
to the heme iron. In contrast, compound 1 had little if any
observable effect on the P450 difference spectrum at concentra-
tions up to 1.4 mM. Although this result cannot distinguish
whether 1 binds only very weakly to ferric P450, or binds
without displacing the iron-coordinated water in the axial
position above the heme, the former interpretation would be
13
of 4 and [1′- C]-4 generated only traces of 3HP as determined,
13
respectively, by DNPH trapping and HPLC and by C NMR.
Moreover, in contrast to incubations with 1 which slow
(
33) Clark, C. R.; Teague, J. D.; Wells, M. W.; Ellis, J. H. Anal. Chem. 1977,
4
9, 912-915.
(
(
34) Wang, F.; Sayre, L. M. J. Am. Chem. Soc. 1992, 114, 248-255.
35) Rosenblatt, D. H.; Davis, G. T.; Hull, L. A.; Forberg, G. D. J. Org. Chem.
1
968, 33, 1649-1651. Hull, L. A.; Davis, G. T.; Rosenblatt, D. H.;
Williams, H. K. R.; Weglein, R. C. J. Am. Chem. Soc. 1967, 89, 1163-
170.
1
J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006 3349

A R T I C L E S
Cerny and Hanzlik
Table 1. Formation of 3HP from Incubation of Cyclopropylamines
Scheme 5. Possible Role of N-Hydroxylation in Formation of 3HP
incubation results in complete loss of 5 and generation of [1′-
1
3
13
C]-3HP as identified by DNPH trapping and C NMR. Thus
it is conceivable that a portion of the 3HP observed in actual
microsomal incubations of 1 could have arisen via acid-catalyzed
decomposition of 5 formed by the FMO enzyme but only if 5
were able to accumulate to a significant extent (up to 57% of
substrate metabolized) during the microsomal incubations of
1. However, such extensive accumulation seems unlikely for
two reasons. First, the FMO enzyme in rat liver microsomes
efficiently oxidizes secondary hydroxylamines to nitrones,38 and
second, hydroxylamine 5 reacts rapidly with both FMO and
P450 in PB-microsomes. Thus, it is likely that most if not all
of the 3HP observed is formed as a true metabolite of 1 and
not as an artifact from the chemical decomposition of 5 during
workup of incubations.
a
Incubations were conducted at 37 °C with 0.5 mg of PB-microsomal
protein, 0.5 µmol of substrate, and 0.5 µmol of NADPH in a total volume
of 0.5 mL. Values represent nmol of 3HP formed/20 min.
b
consistent with its inability to interfere with the metabolism of
, as shown by curves A and B in Figure 2.
Formation of 3HP from Secondary and Tertiary Cyclo-
3
1
4
propylamines and Secondary Hydroxylamine 5. In view of
the significant difference in metabolism of 1 versus 4, we
examined several other pairs of secondary and tertiary cyclo-
propylamines to determine if other secondary cyclopropylamines
were also more readily converted to 3HP than their tertiary
N-methylated derivatives. As indicated in Table 1, for three of
the four pairs tested, production of 3HP occurred to a much
greater extent with the secondary cyclopropylamine than with
its tertiary N-methyl analogue. This suggests that formation of
HP from these compounds may actually require the N-H bond
of a secondary amine. In this context it is interesting to note
that of the several other cyclopropylamine-based P450 inacti-
vators described in the literature, all are secondary amines.
Discussion
A central part of the concept behind mechanism-based or
suicide substrate enzyme inactivators is that they resemble
normal substrates, bind to the active site of an enzyme, and
undergo early steps of the catalytic cycle as if they were a
normal substrate. However, having unique chemical features
not found in ordinary substrates, suicide substrates can, at least
part of the time, exit the normal catalytic cycle prematurely in
a way that leads to covalent inactivation of the enzyme. Thus,
the study of suicide substrates can potentially illuminate at least
part of the catalytic mechanism followed by normal substrates
whose turnover does not lead to enzyme inactivation. A case
in point is the P450-catalyzed N-dealkylation of tertiary amines,
which has been conclusively shown to occur via the interme-
3
7-9,36
Furthermore, whereas oxidation of 4-alkyl-1,4-dihydropyridine
Hantzsch esters by P450 enzymes leads to alkyl radical extrusion
and enzyme inactivation, their N-alkyl derivatives are also
3
7
inactive in this regard until first undergoing N-deaklylation.
diacy of a carbinolamine which decomposes solvolytically to a
The significance of these observations is that, collectively, they
cast doubt on the role of SET mechanisms in the formation of
stable metabolites from 1 and possibly in the mechanism of
suicide inactivation of P450 enzymes by 1.
The generally more extensive production of 3HP from
secondary cyclopropylamines versus their tertiary N-methyl
derivatives, as well as our previous demonstration of the
involvement of the microsomal flavin-containing monooxyge-
nase (FMO) in the N-oxidation of 1,31 led us to consider that
production of 3HP might also occur via a secondary cyclopro-
pylhydroxylamine intermediate such as 5 as shown in Scheme
39-41
carbonyl compound and a dealkylated amine.
Carbinola-
mine formation had been assumed to occur by a reaction
analogous to aliphatic hydroxylation, but with the discovery that
both 1 and 2 are suicide substrates for P450, the SET mechanism
was proposed as a mechanism that could potentially account
for both the N-dealkylation of normal amines and the suicide
13,14
substrate activity unique to cyclopropylamines.
Initially, the SET mechanism also appeared attractive as a
way to rationalize the enigmatically small kinetic deuterium
isotope effects observed for the N-dealkylation of low-E1/2
21
amines (1.0 < kH/kD < 1.8), which contrast the relatively large
1
3
13
5
. Using C NMR, we demonstrated that, whereas [1′- C]-5
is stable in incubation buffer for >4 h, addition of 15% zinc
sulfate solution to simulate the workup of a microsomal
isotope effects observed for N-dealkylation of high-E1/2 amides,
(38) Cashman, J. R.; Yang, Z. C.; H o¨ gberg, T. Chem. Res. Toxicol. 1990, 3,
4
28-432.
(
39) Rose, J.; Castagnoli, N., Jr. Med. Res. ReV. 1983, 3, 73-88.
(
36) Castagnoli, N., Jr.; Rimoldi, J. M.; Bloomquist, J.; Castagnoli, K. P. Chem.
Res. Toxicol. 1997, 10, 924-940.
37) Lee, J. S.; Jacobsen, N. E.; Ortiz de Montellano, P. R. Biochemistry 1988,
(40) Shea, J. P.; Valentine, G. L.; Nelson, S. D. Biochem. Biophys. Res. Commun.
1982, 109, 231-235.
(41) Kedderis, G. L.; Dwyer, L. A.; Rickert, D. E.; Hollenberg, P. F. Mol.
Pharmacol. 1983, 23, 758-760.
(
2
7, 7703-7710.
3350 J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006

P450 Oxidation of Cyclopropylamines
A R T I C L E S
Scheme 6. Proposed Mechanisms for Interaction of 1 or 2 with
Cytochrome P450
sistent with an SET mechanism. In fact, the formation of 3HP
uniquely from secondary cyclopropylamines is entirely consis-
tent with a variation of the HAT mechanism first proposed by
26
Karki and Dinnocenzo, as shown in Scheme 6B. In this scheme
the enzymatic oxidant is depicted as the ferryl or iron-oxene
3
+ 24,25
derivative of P450 (FedO ),
but the N-HAT mechanism
shown is not inconsistent with a radical-like Fe(III)-OO‚
precursor to the ferryl species being involved as the H-
abstracting agent.
In conclusion, we believe the many results presented above
argue persuasively in favor of the “traditional” HAT-type
mechanism for most if not all P450-catalyzed N-dealkylation
reactions, including those of cyclopropylamines. Other important
questions remain, however. For example, at what point in the
turnover cycle of P450 with 1 or 2 does the mechanism branch
to suicide inactivation; is there any role for SET in amine
metabolism by P450 enzymes; what isoforms of P450 in PB-
microsomes are inactivated by 1; in addition to metabolic
O-dealkylation of ethers, and C-hydroxylation of aliphatic
3
1
systems (4 < kH/kD < 10).42-45 However, isotope effect profiling
intermediate complex formation, how do 1 and 2 inactivate
P450? Ongoing studies in our laboratory are directed toward
answering these important questions.
studies of P450-catalyzed N-dealkylations and related HAT-
type model reactions have been interpreted to favor the classical
4
6
HAT (i.e., abstraction/recombination) mechanism, and a
moderately large (ca. 3.1) kinetic deuterium isotope effect more
consistent with a HAT mechanism has recently been reported
Methods
Glucose-6-phosphate, glucose-6-phosphate dehydrogenase, and NAD-
PH were purchased from Sigma (St. Louis, MO). Compounds 6-11
were available from earlier work in our laboratory. All other chemicals
and reagents were of reagent grade or higher purity and were purchased
from commercial suppliers. Liver microsomes from phenobarbital-
6
for P450-catalyzed N-decyclopropylation reactions. The current
consensus is that kinetic isotope effects do not provide support
for SET involvement in N-dealkylation reactions.6
,26,27
More recently the use of product studies to investigate the
role of SET mechanisms in the oxidative metabolism of
cyclopropylamines has been enabled by the development of
facile synthetic methods for isotopic labeling of cyclopropy-
3
1,48,49
treated rats were prepared, stored, and handled as described.
P450
48
ligand binding spectra were recorded as described by Chaurasia et al.
Other general laboratory instrumentation and procedures were as
described. Unless otherwise specified, the buffer used for all mi-
31
2
8,47
lamines
and by the increased sensitivity of modern NMR
crosomal incubations was 0.1 M potassium phosphate, pH 7.5. Zinc
sulfate solution (15% w/v) was used to precipitate proteins, but unlike
many literature descriptions of this method we added barium hydroxide
only when formaldehyde was the metabolite being measured. The
spectrometers. In agreement with expectations for an SET
reaction mechanism, oxidation of the low-E1/2 substrate N-
cyclopropyl-N-methylaniline (3) by horseradish peroxidase was
found to generate exclusively cyclopropyl ring-opened prod-
13
14
13
14
syntheses of [1′- C]-1, [1′- C]-1, [7- C]-1, [7- C]-1, and 5 have been
described previously.31 2,4-Dinitrophenylhydrazine derivatives of ac-
rolein and 3-hydroxypropionaldehyde were prepared as described by
Shaffer et al.29
2
8,29
ucts.
In contrast, oxidation of 3 by cytochrome P450
enzymes gave the ring-intact metabolite cyclopropanone hydrate
as the only metabolite from the cyclopropyl moiety; moreover
13
N-Benzyl-N-methyl-[1′- C]-cyclopropylamine‚HCl (4‚HCl). To
a 30 mL culture tube (2.5 cm × 10 cm) was added N-benzyl-N-
0
3
, unlike 1, does not inactivate P450.3
Because of the dichotomous behavior of 3 toward HRP versus
13
methylamine‚HCl (0.79 g, 5.00 mmol), sodium [ C]-formate (0.28 g,
P450, and because HRP does not oxidize 1, we undertook to
elucidate the complete P450 metabolic profile of 1. As reported
above, in addition to all of the anticipated normal N-dealkylation
products (Scheme 6A), 3HP has for the first time been observed
as a metabolite, indeed a major (57%) metabolite, derived from
the cyclopropyl moiety of 1. 3HP had long been anticipated as
the “signature” metabolite for an SET mechanism for cyclo-
propylamine oxidation (Scheme 2). However, the fact that
tertiary cyclopropylamines, which on the basis of oxidation
potential should be more susceptible to SET oxidation than their
secondary analogues,43 apparently yield 3HP only after under-
going metabolism to a secondary cyclopropylamine is incon-
4
.00 mmol), and 10 mL of acetonitrile. The tube was capped with a
Teflon-lined screw cap, and the contents were stirred at 82 °C for 48
h. The reaction was cooled to room temperature and diluted with ethyl
acetate (15 mL), and the resulting solution rinsed with 1 M NaOH (15
mL) and 1 M HCl (15 mL). The ethyl acetate solution was then dried
over magnesium sulfate, filtered, and the solvent was removed in vacuo
yielding N-benzyl-N-methyl-[ C]-formamide as a yellow liquid (0.55
g, 74% yield). H NMR (500 MHz, CDCl
Hz), 4.46 (d, 2H, J ) 3.83 Hz), 7.20-7.55 (m, 5H), 8.22 (dd, 1H, J )
13
1
3
) δ 2.82 (d, 3H, J ) 3.51
6
4
1
2.81, 192.87 Hz). 13C NMR (125.8 MHz, CDCl
8.56, 54.30, 128.16, 128.41, 128.87, 128.99, 129.44, 129.65, 136.40,
3
) δ 30.24, 34.87,
13
13
36.66, 163.53 ( C-enriched carbon), 163.69 ( C-enriched carbon).
].
+
GC/MS: t 14.5 min, m/z 150 [M
R
13
N-benzyl-N-methyl-[ C]-formamide (0.55 g, 3.68 mmol) was dis-
solved in 8 mL of dry tetrahydrofuran in a three-neck round-bottom
flask and stirred vigorously under nitrogen. Titanium isopropoxide (1.35
mL, 4.60 mmol) was added, followed by 3.68 mL of a 3.00 M solution
(
42) Guengerich, F. P.; Peterson, L. A.; B o¨ cker, R. H. J. Biol. Chem. 1987,
2
63, 8176-8183.
(43) Hall, L. R.; Hanzlik, R. P. J. Biol. Chem. 1990, 265, 12349-12355.
(44) Hanzlik, R. P.; Ling, J. K.-H. J. Am. Chem. Soc. 1993, 115, 9363-9370.
(45) Iyer, K. R.; Jones, J. P.; Darbyshire, J. F.; Trager, W. F. Biochemistry 1997,
3
6, 7136-7143.
(
46) Manchester, J. I.; Dinnocenzo, J. P.; Higgins, L. A.; Jones, J. P. J. Am.
Chem. Soc. 1997, 119, 5069-5070.
(48) Chaurasia, C. S.; Alterman, M. A.; Lu, P.; Hanzlik, R. P. Arch. Biochem.
Biophys. 1995, 317, 161-169.
(49) Narasimhan, N.; Weller, P. E.; Buben, J. A.; Wiley, R. A.; Hanzlik, R. P.
Xenobiotica 1988, 18, 491-499.
(
47) Chaplinski, V.; de Meijere, A. Angew. Chem., Int. Ed. Engl. 1996, 35,
4
13-414.
J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006 3351

A R T I C L E S
Cerny and Hanzlik
of ethylmagnesium bromide in ether (11.05 mmol). Upon addition of
the Grignard reagent, gas evolved and a yellow color formed. The
mixture was heated for 15 min at 45 °C, whereupon the reaction mixture
became black. The reaction mixture was then cooled to room temper-
ature, stirred for 12 h, and quenched by addition of 15 mL of saturated
aqueous ammonium chloride solution. After adding 20 mL of 1 M
NaOH the resulting mixture was extracted with ether (3 × 25 mL).
The extracts were pooled, dried over magnesium sulfate, filtered, and
concentrated in vacuo giving a yellow liquid (0.41 g). This material
was then applied to a silica gel column and eluted with 2.5% ether in
pentane to obtain 0.21 g of N-benzyl-N-methyl-[1′- C]-cyclopropy-
lamine as a clear colorless liquid. This liquid was then dissolved in 10
mL ether and bubbled with HCl gas resulting in the formation of a
white precipitate. The white precipitate that formed was collected by
suction filtration and dissolved in a minimum volume of hot absolute
ethanol, cooled to room temperature, and the product was precipitated
(m, 1H), 6.35 (s, 1H), 7.90-7.92 (m, 2H), 8.26-8.29 (m, 2H). GC/
+
MS: t
R
16.3 min, m/z 206 [M ].
N,N-Diethyl-4-nitrobenzamide. N,N-Diethyl-4-nitrobenzamide, used
as an internal standard for HPLC, was prepared and purified as
described for N-cyclopropyl-4-nitrobenzamide giving the desired
1
product as white needles (43% yield). Mp 64-66 °C. H NMR (400
MHz, CDCl ) δ 1.11-1.14 (m, 3H), 1.26-1.31 (m, 3H), 3.21 (m, 2H),
3
3.57 (m, 2H), 7.55 (d, 2H, J ) 8.77 Hz), 8.28 (d, 2H, J ) 8.74 Hz).
+
GC/MS: t 18.4 min, m/z 222 [M ].
R
N-Cyclopropyl-N-methyl-N-cyclohexylmethylamine‚HCl (7‚HCl).
In a 16 mm × 100 mm culture tube was combined the hydrochloride
salt of N-cyclopropyl-N-(cyclohexylmethyl)amine (0.47 g, 2.50 mmol),
1
3
5
mL of methanol, formaldehyde (12.7 M solution, 0.22 mL), and acetic
acid (0.57 mL, 10 mmol) along with 3 Å molecular sieves. After stirring
for 30 min, sodium cyanoborohydride (0.24 g, 3.75 mmol) was added
and the tube was capped with a Teflon-lined screwcap. The reaction
was allowed to stir overnight. The reaction was then quenched by
addition of water (10 mL), and the resulting solution was extracted
with ether (15 mL). The aqueous portion was then made alkaline by
addition of 10 mL of 1 M NaOH and extracted with ether (3 × 15
mL). The extracts were pooled, dried over magnesium sulfate, and
concentrated under reduced pressure. The colorless liquid obtained was
dissolved in 5 mL of ether and bubbled with HCl gas resulting in the
by addition of ether giving 0.15 g of a white crystalline solid (20%).
1
H NMR (400 MHz, D O) δ 0.80 (m, 2H), 0.91 (m, 2H), 2.81 (dm,
2
1
3
1
H, J ) 186.18 Hz), 2.94 (m, 2H) 7.50 (m, 5H). C NMR (100.6
1
3
MHz, D
1
2
O) δ 4.65, 39.28 ( C-enriched carbon), 42.60, 61.31, 129.27,
+
29.44, 130.38, 131.67. GC/MS (free base): t
R
11.2 min, m/z 161 [M ].
1
3
13
[
1′- C]-N-Benzyl-N-cyclopropylhydroxylamine ([1′- C]-5). In a
1
3
three-neck round-bottom flask was combined [1′- C]-1 (56 mg, 0.38
mmol), sodium tungstate dihydrate (5 mg, 0.015 mmol), and 3 mL of
water. Hydrogen peroxide solution (30%, 0.10 mL) was added over
1
formation of a white precipitate (0.054 g, 11%). H NMR (400 MHz,
D O) δ 0.95-1.29 (m, 9H), 1.69 (m, 5H), 1.90 (m, 1H), 2.79 (m, 1H),
2
2.92 (s, 5H), 3.15 (s, 5H). 13C NMR (100.6 MHz, D O) δ 25.20, 25.70,
30 min with constant cooling on ice. The reaction was allowed to warm
2
to room temperature and stirred an additional 3 h. The reaction was
halted by addition of 6 mg of sodium hydrogen sulfite and 48 mg of
sodium chloride. The resulting solution was extracted with methylene
chloride (4 × 10 mL), and the extracts were pooled, dried over
magnesium sulfate, and concentrated in vacuo giving 0.048 g of the
crude nitrone product as determined by TLC. Without purification the
crude nitrone was dissolved in 2 mL of dry tetrahydrofuran and added
dropwise over 30 min to a stirred ice-cold slurry of lithium aluminum
hydride (0.045 g, 1.19 mmol) in 2 mL of tetrahydrofuran under an
atmosphere of nitrogen with constant cooling. After addition of the
nitrone solution was complete, the reaction was left to stir at 0 °C for
30.56, 32.95, 40.62, 42.23, 64.64.
N-Cyclopropyl-N-methyl-1-methyl-2-phenylethylamine‚HCl (11‚
HCl). In a 16 mm × 100 mm culture tube the hydrochloride salt of
N-cyclopropyl-1-methyl-2-phenylethylamine (10) (0.53 g, 2.50 mmol),
methanol (4 mL), formaldehyde (12.7 M solution, 0.22 mL, 2.79 mmol),
and acetic acid (0.57 mL, 10 mmol) were combined over 3 Å molecular
sieves. After stirring for 30 min, sodium cyanoborohydride (0.24 g,
3
.75 mmol) was added, and the tube was capped with a Teflon-lined
screwcap. The reaction was allowed to stir overnight, water was then
added, and the resulting solution was extracted with ether (15 mL).
The aqueous portion was then made alkaline by addition of 10 mL of
3
h. At this point TLC showed that no starting material remained. The
1
M NaOH and extracted with ether (3 × 15 mL). The extracts were
reaction was quenched by the slow addition of 3 mL of water with
continual cooling. The resulting mixture was extracted with ether (4
pooled, dried over magnesium sulfate, and concentrated under reduced
pressure. The colorless liquid obtained was dissolved in 5 mL of ether
×
10 mL), and the extracts were pooled, dried over magnesium sulfate,
and bubbled with HCl gas resulting in formation of a white precipitate
and concentrated giving 0.025 g of a yellow liquid. After silica gel
chromatography using 3:1 pentane/ether as solvent, 0.015 g of the
desired product was obtained as an amorphous white solid (24% yield
1
(
0.149 g, 26%). H NMR (400 MHz, D
2
O) δ 0.87-1.07 (m, 4H), 1.26
(t, 3H, J ) 5.71 Hz), 2.75-2.97 (m, 5H), 3.25 (dd, 1H, J ) 4.07,
13
1
3.17 Hz), 3.75-3.84 (m, 1H), 7.30-7.59 (m, 5H). C NMR (100.6
MHz, D O) δ 3.59, 3.80, 5.50, 5.77, 12.34, 13.76, 36.51, 36.57, 37.04,
7.55, 37.65, 38.13, 64.16, 64.45, 127.59, 127.67, 129.27, 129.30,
1
3
1
from [1′- C]-1). H NMR (400 MHz, MeOH-d ) δ 0.57 (m, 4H), 2.30
4
2
13
(dm, 1H, J ) 171.77 Hz), 3.96 (s, 2H), 7.21-7.35 (m, 5H). C NMR
3
1
1
3
(
6
100.6 MHz, MeOH-d
3.32, 128.18, 129.13, 130.81.
N-Cyclopropyl-4-nitrobenzamide. Cyclopropylamine (5.00 mmol),
4
) δ 6.86, 7.02, 41.65 ( C-enriched carbon),
29.72, 136.48, 136.72.
Incubation and Metabolite Analysis by 13C NMR. Concentrated
PB-microsomes (729 µL, 30 mg of protein, 101 nmol of P450) were
triethylamine (10.00 mmol), and 5 mL of dry tetrahydrofuran were
combined in a 50 mL culture tube. A 1 M solution of 4-nitrobenzoyl
chloride (10 mL, 10.00 mmol) in dry tetrahydrofuran was added
dropwise with stirring. The tube was capped with a Teflon-lined
screwcap and heated at 65 °C for 4 h. The reaction was quenched by
addition of water (10 mL), and the organic portion was separated and
concentrated in vacuo. The resulting material was then dissolved in 40
mL of chloroform and rinsed with saturated aqueous sodium carbonate
diluted with 8.9 mL of buffer and warmed to 37 °C. To this mixture
was added 40 µL of either buffer or a solution of [1′-13C]-1 or [7-13C]-
1 or [1′-13C]-4 in buffer (0.5 M, 20 µmol) and 53 µL of an NADPH
regeneration system (0.13 units/µL glucose-6-phosphate dehydrogenase
and 1 M glucose-6-phosphate in buffer). This mixture was vortexed
thoroughly and preincubated for 3 min at 37 °C. The incubation was
started by addition of 200 µL of a 50 mM solution of NADPH in buffer
(10 µmol) bringing the total volume to 10 mL. The incubation was
conducted for 90 min at 37 °C in an oscillating water shaker bath. The
incubation was quenched by addition of 2 mL of 15% zinc sulfate
solution (which dropped the pH to around 4), the precipitate was
sedimented by centrifugation, and the resulting supernatant was
removed. For the [7-13C]-1 incubations the supernatant was further
acidified to pH 1 with 1 mL of 1 M HCl and extracted with ether (3
× 5 mL). The extracts were set aside for later analysis (see below).
After extraction the supernatant was concentrated with gentle heating
under vacuum to a volume of approximately 500 µL. This liquid was
(
2 × 30 mL) and H
2
O (30 mL). The resulting solution was dried over
magnesium sulfate, filtered, and the solvent was removed in vacuo to
give N-cyclopropyl-4-nitrobenzamide. The crude product was dissolved
in hot benzene, and hexane was added until a cloudy solution resulted.
Upon standing in a refrigerator, the desired product formed as white
needles (51% yield). Mp 175 °C (Lit. 178-179 °C). H NMR (400
MHz, CDCl ) δ 0.62-0.69 (m, 2H), 0.90-0.95 (m, 2H), 2.90-2.97
5
0 1
3
(50) Horrom, B. W.; Lynes, T. E. J. Med. Chem. 1963, 19, 528-532.
3352 J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006

P450 Oxidation of Cyclopropylamines
A R T I C L E S
then transferred to a 5 mm NMR tube along with 50 µL of D
samples were then analyzed by C NMR. After acquisition of the initial
2
O. These
was removed and combined with 10 µL of a hexanes-extracted DNPH
solution (0.15 M) and allowed to stand at room temperature for 30
13
28
1
3
13
C NMR spectra for the PB-microsomal incubation of [1′- C]-1,
cyclopropylamine (5.9 µL of 7.61 M solution in H O, 45 µmol) and
cyclopropanone hydrate (32.2 µL, 1.4 M in H
O, 45 µmol)3 were
added, and the sample was reanalyzed by C NMR.
Analysis of Ether Extracts for Non-Amine Metabolites of [7- C]-
min. The resulting mixture was extracted with ethyl acetate (2 × 500
µL), and the combined extracts were dried in vacuo. The resulting
orange residue was dissolved in 200 µL of acetonitrile, and the solution
of DNPH-trapped material (20 µL) was analyzed by HPLC using a
Vydac C18 column (5µ, 4.6 mm × 250 mm) with UV monitoring at
350 nm eluting at 1 mL/min with a gradient of solvent B (acetonitrile)
2
0,51
2
1
3
13
1
3
1
by C NMR and GC/MS. A blank incubation and an incubation
1
3
containing [7- C]-1 were conducted in parallel as described above.
The ether extracts of the quenched incubation supernatants were dried
over magnesium sulfate and filtered. To avoid loss of volatile
metabolites, the ether extracts were concentrated to <250 µL using a
Vigreux distillation column. To this liquid was then added 1 mL of
deuterated chloroform, and the sample transferred to a 5 mm NMR
tube and analyzed by 1 C NMR. After acquisition of the NMR spectrum,
an aliquot of the NMR solution was removed and without dilution
analyzed by GC/EIMS (DB5 column) using the following oven
program: 50 °C initial temperature increasing to 150 °C at 10 °C/min,
then 30 °C/min to 250 °C, then hold at 250 °C for 10 min. To limit the
loss of volatile compounds the injection port temperature was main-
tained at 150 °C. To confirm the retention times and mass spectral
fragmentation patterns, authentic standards were injected using the same
oven program.
2
in solvent A (MeOH/H O, 1:1 v/v) as follows: 0-13 min, 9% B; 13-
18 min, 9-30% B; 18-21 min, 30% B; 21-23 min, 30-100% B;
23-26 min, 100% B; 26-27 min, 100-9% B; 27-35 min, hold at
9% B. Formation of 3HP was quantified using a calibration curve
prepared from a standard solution of 3,3-diethoxypropanol (which
generates 3HP, but not acrolein, upon hydrolysis under the above
conditions).
3
Quantitation of Cyclopropylamine Formation. Incubation aliquots
(200 µL) were removed at 0, 15, 30, and 60 min and combined with
45 µL of 15% zinc sulfate solution and 20 µL of 1 mM N,N-diethyl-
4-nitrobenzamide (internal standard). After brief centrifugation the
supernatant was removed and combined with 180 µL of 5 M potassium
hydroxide and 400 µL of freshly prepared 100 mM 4-nitrobenzoyl
chloride in tetrahydrofuran and allowed to stand at room temperature
for 30 min. The resulting mixture was extracted with ethyl acetate (2
× 250 µL), the extracts were concentrated in vacuo, and the resulting
material was dissolved in 100 µL of acetonitrile. HPLC analysis was
performed by injection of 20 µL of this solution onto a Vydac C18
column (5µ, 4.6 mm × 250 mm) monitored by UV at 270 nm and
eluted at a flow rate of 1 mL/min using a gradient of solvent B
1
4
14
Incubation of 1, [1′- C]-1, and [7- C]-1 for Metabolite Quan-
titation. These incubations were carried out essentially as described
above, except that the incubations were scaled down to 1.5-2.0 mL
and the microsomal protein concentration was decreased to 2 mg/mL.
Aliquots (200 µL) were removed at various times for analysis as
described below.
(acetonitrile) in solvent A (70:30 H O/acetonitrile) as follows: 0-8
2
Incubations for amine analysis were quenched by addition to a
mixture of 45 µL of 15% zinc sulfate solution and 20 µL of 2.75 mM
min, 0% B; 8-15 min, 0-55% B; 15-20 min, hold at 55% B; 20-23
min, 55-0% B. Formation of cyclopropylamine was quantified using
a calibration curve prepared from a standard solution of cyclopropy-
lamine.
4-methoxybenzylamine (internal standard). The precipitate was removed
by centrifugation for 10 min. The supernatant was removed, extracted
with pentane/ether (9:1), and residual pentane/ether was removed by
briefly blowing a stream of nitrogen over the solution. The quenched
incubation samples were then analyzed by HPLC (100 µL injection)
using an isocratic solvent system consisting of 15% acetonitrile in 100
Effect of 1 on the N-Demethylation of 4. To examine the direct
effect of coincubating 1 with 4, a suspension of PB-microsomes (2 mg
of protein/mL, 4436 µL) was added to each of two 25 mL Erlenmeyer
flasks. To each of these flasks was also added 95 µL of a 25 mM
solution of 4 (2.38 µmol) in buffer and 29 µL of an NADPH
regenerating system containing glucose-6-phosphate dehydrogenase
(0.13 units/mL) and glucose-6-phosphate (1.0 M) in buffer. One flask
then received 95 µL of a solution of 1 (12.5 mM, 1.19 µmol) in buffer
while the other received 95 µL of buffer alone. Prior to initiating the
reaction by addition of NADPH (see below), a 490 µL “zero-time”
aliquot was removed from each flask and added to a 1.7 mL Eppendorf
tube containing 250 µL each of 15% zinc sulfate solution and saturated
aqueous barium hydroxide solution. Each tube was then vortexed, 10
µL buffer was added, and the mixture was held for later analysis. The
two incubation tubes were then warmed in an oscillating water bath at
33 °C for 3 min prior to initiating reaction by the addition of 85 µL of
4
mM NaClO (acidified to pH 2.25 using perchloric acid) at a 1 mL/
5
2
min flow rate on a Kromasil C4 column (5 µm, 4.6 mm × 150 mm).
The column effluent was passed through a UV detector (210 nm) in
series with a Ramona radioactivity flow detector with a solid scintillant
cell. The amounts of benzylamine formed and 1 remaining were
quantified based on radiochemical peak integration or using a calibration
curve prepared from standard solutions of nonradiolabeled benzylamine
and 1.
For analysis of non-amine metabolites incubations were quenched
by addition of 15% zinc sulfate solution (45 µL) and acetophenone
(20 µL, 1.25 mM in acetonitrile) as an internal standard. The quenched
incubation samples were then analyzed by HPLC (100 µL injection)
using an isocratic solvent system consisting of 27.5:72.5 acetonitrile/
ammonium formate solution (50 mM, adjusted to pH 3.6) at a flow
rate of 1 mL/min on a Vydac C18 column (5µ, 4.6 mm × 250 mm).
Column effluent was passed through a UV detector (254 nm) in series
with a Ramona radioactivity flow detector with a solid scintillant cell.
Non-amine metabolite formation was quantified based on radiochemical
peak area integration.
2 4
an NADPH solution (50 mM in 0.1 M KH PO buffer). At 0.5, 1, 3,
5, 10, 15, 30, and 45 min after the addition, a 500 µL aliquot was
removed from each flask and added to separate 1.7 mL Eppendorf tubes
containing 250 µL of 15% zinc sulfate solution. The tubes were
vortexed, a saturated solution of barium hydroxide was added (250
µL), and the tubes were vortexed again. After brief centrifugation, 800
µL of the supernatant was removed, combined with 400 µL of Nash
reagent,53 and heated at 60 °C for 10 min. The absorbance of each
incubation sample was then measured at 415 nm versus a similarly
treated buffer (blank) solution. Formaldehyde formation was quantified
using a calibration curve prepared from standard solutions.
DNPH Trapping and Quantitation of 3HP. Aliquots (200 µL) of
an incubation mixture of 1 with PB-microsomes (2 mg of protein/mL)
were removed at 0, 15, 30, and 60 min and placed into a mixture of
1
5% zinc sulfate solution (45 µL) and internal standard (20 µL × 2
mM acetaldehyde or glycolaldehyde in acetonitrile). Precipitated protein
To examine the effect of preincubation of 1 on the N-demethylation
of 4, PB-microsomes were diluted with buffer to a protein concentration
of 2 mg/mL and kept on ice until use. Diluted microsomes (3502 µL)
were added to each of two 25 mL Erlenmeyer flasks along with 23 µL
was removed by centrifugation for 10 min. The supernatant (200 µL)
(
51) Wiseman, J. S.; Nichols, J. S.; Kolpak, M. X. J. Biol. Chem. 1982, 257,
6
328-6332.
(52) Roberts, J. M.; Diaz, A. R.; Fortin, D. T.; Friedle, J. M.; Piper, S. D. Anal.
Chem. 2002, 74, 4927-4932.
(53) Nash, T. Biochem. J. 1953, 55, 416-421.
J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006 3353

A R T I C L E S
Cerny and Hanzlik
of an NADPH regenerating system as described above. To one of the
two flasks was added a solution 1 in buffer (75 µL, 12.5 mM, 0.94
µmol), and to the other was added 75 µL of buffer alone. To each
flask was then added 75 µL of an NADPH solution (50 mM in buffer),
and the resulting mixtures were incubated in the absence of 4 in an
oscillating water bath at 33 °C for 10 min. Prior to addition of 4, a 490
µL aliquot was removed from each flask and added to a 1.7 mL
Eppendorf tube containing 250 µL each of 15% zinc sulfate and a
saturated aqueous barium hydroxide solution and 10 µL of buffer to
generate a zero-time sample. After 10 min of preincubation time, a
solution of 4 (65 µL, 25 mM in buffer, 1.63 µmol) was added to each
flask. This was followed by further incubation in the oscillating water
bath at 33 °C. At 1, 3, 5, 10, 20, and 25 min after the addition of 4, a
ethylene glycol (250 µmol, internal standard) in water. The mixture
was thoroughly vortexed, and a 550 µL aliquot was removed, added
to a 5 mm NMR tube along with 50 µL of D O, and analyzed by C
2
NMR after 30 min and again after 4 h. To the remaining solution was
added 1 mL of a 15% solution of zinc sulfate which resulted in the
precipitation of a white solid. Precipitate was sedimented by centrifuga-
tion, and the supernatant was removed and concentrated with gentle
heating under vacuum to a volume of approximately 500 µL. This liquid
1
3
2
was then transferred to a 5 mm NMR tube along with 50 µL of D O
and analyzed by 13C NMR.
Acknowledgment. We thank Sara Neuenswander and Dr.
David Vander Velde for assistance with NMR experiments and
Pfizer Global Research and Development for financial support
of this work. We also appreciate helpful suggestions made by
several anonymous reviewers.
5
00 µL aliquot was removed from each flask in parallel and added to
a 1.7 mL Eppendorf tube containing 250 µL of 15% zinc sulfate
solution; after vortexing to mix thoroughly, 250 µL of saturated aqueous
barium hydroxide solution was added and the tubes were vortexed again.
Samples were processed and analyzed for formaldehyde as described
above.
Supporting Information Available: 13C NMR spectra. This
material is available free of charge via the Internet at
http://pubs.acs.org.
1
3
Mock Incubation of [1′- C]-5. To a 16 mm × 100 mm culture
tube was added 5394 µL of buffer, 56 µL of a 100 mM solution of
1
3
[
1′- C]-5 (5.6 µmol) in methanol, and 100 µL of a 2.5 M solution of
JA054938+
3354 J. AM. CHEM. SOC.
9
VOL. 128, NO. 10, 2006