3-Pyrrolines are mechanism-based inactivators of the quinone-dependent amine oxidases but only substrates of the flavin-dependent amine oxidases

Article abstract of DOI:10.1021/ja0205434

We previously reported that 3-pyrroline and 3-phenyl-3-pyrroline effect a time-dependent inactivation of the copper-containing quinone-dependent amine oxidase from bovine plasma (BPAO) (Lee et al. J. Am. Chem. Soc. 1996, 118, 7241-7242). Quinone cofactor model studies suggested a mechanism involving stoichiometric turnover to a stable pyrrolylated cofactor. Full details of the model studies are now reported along with data on the inhibition of BPAO by a family of 3-aryl-3-pyrrolines (aryl = substituted phenyl, 1-naphthyl, 2-naphthyl), with the 4-methoxy-3-nitrophenyl analogue being the most potent. At the same time, the parent 3-phenyl analogue is a pure substrate for the flavin-dependent mitochondrial monoamine oxidase B from bovine liver. Spectroscopic studies (including resonance Raman) on BPAO inactivated by the 4-methoxy-3-nitrophenyl analogue are consistent with covalent derivatization of the 2,4,5-trihydroxyphenylalanine quinone (TPQ) cofactor. The distinction of a class of compounds acting as an inactivator of one amine oxidase family and a pure substrate of another amine oxidase family represents a unique lead to the development of selective inhibitors of the mammalian copper-containing amine oxidases.

Full text of DOI:10.1021/ja0205434

Published on Web 09/21/2002
3-Pyrrolines Are Mechanism-Based Inactivators of the
Quinone-Dependent Amine Oxidases but Only Substrates of
the Flavin-Dependent Amine Oxidases
Younghee Lee,† Ke-Qing Ling,† Xingliang Lu,‡ Richard B. Silverman,‡
E. M. Shepard,§ D. M. Dooley,§ and Lawrence M. Sayre*,†
Contribution from the Department of Chemistry, Case Western ReserVe UniVersity,
CleVeland, Ohio 44106, Departments of Chemistry and of Biochemistry, Molecular Biology, and
Cell Biology, Northwestern UniVersity, EVanston, Illinois 60208-3113, and Department of
Chemistry and Biochemistry, Montana State UniVersity, Bozeman, Montana 59717
Received April 15, 2002
Abstract: We previously reported that 3-pyrroline and 3-phenyl-3-pyrroline effect a time-dependent
inactivation of the copper-containing quinone-dependent amine oxidase from bovine plasma (BPAO) (Lee
et al. J. Am. Chem. Soc. 1996, 118, 7241-7242). Quinone cofactor model studies suggested a mechanism
involving stoichiometric turnover to a stable pyrrolylated cofactor. Full details of the model studies are now
reported along with data on the inhibition of BPAO by a family of 3-aryl-3-pyrrolines (aryl ) substituted
phenyl, 1-naphthyl, 2-naphthyl), with the 4-methoxy-3-nitrophenyl analogue being the most potent. At the
same time, the parent 3-phenyl analogue is a pure substrate for the flavin-dependent mitochondrial
monoamine oxidase B from bovine liver. Spectroscopic studies (including resonance Raman) on BPAO
inactivated by the 4-methoxy-3-nitrophenyl analogue are consistent with covalent derivatization of the 2,4,5-
trihydroxyphenylalanine quinone (TPQ) cofactor. The distinction of a class of compounds acting as an
inactivator of one amine oxidase family and a pure substrate of another amine oxidase family represents
a unique lead to the development of selective inhibitors of the mammalian copper-containing amine oxidases.
Scheme 1
Introduction
The quinone-dependent copper amine oxidases mediate a
pyridoxal phosphate (PLP)-like transamination mechanism for
conversion of primary amines to aldehydes.1 Of the several
mammalian enzymes in this enzyme class, plasma amine oxidase
(PAO) appears to be responsible for regulating the levels of
blood-borne biogenic amines and for metabolizing exogenous
amines, though the exact roles of this and other tissue-specific
copper amine oxidases are only partially understood.
Substrate oxidation is initiated by Schiff base formation
between primary amine and the 2,4,5-trihydroxyphenylalanine
quinone (TPQ) cofactor derived posttranslationally from an
active-site tyrosine residue present in each monomer of these
dimeric enzymes.2 The resulting “substrate Schiff base” tau-
tomerizes to a “product Schiff base”, which in turn hydrolyzes
to aldehyde product and a reductively aminated cofactor. The
latter is oxidatively recycled to the starting quinone at the
expense of reducing O2 to H2O2 (Scheme 1). This mechanism
has been reproduced by use of simple models for the TPQ
cofactor.3,4
Mechanism-based irreversible inactivators5 are superior en-
zyme inhibitors because they result in long-acting inhibition in
vivo (recovery of activity requires new protein synthesis) and
because they offer a second level of selectivity: a given
mechanism-based inhibitor may bind reversibly to and thus
* Corresponding author: Phone: (216) 368-3704. Fax: (216) 368-3006.
E-mail: lms3@po.cwru.edu.
† Case Western Reserve University.
‡ Northwestern University.
§ Montana State University.
(3) (a) Wang, F.; Bae, J. Y.; Jacobson, A. R.; Lee, Y.; Sayre, L. M. J. Org.
Chem. 1994, 59, 2409-2417. (b) Lee, Y.; Sayre, L. M. J. Am. Chem. Soc.
1995, 117, 3096-3105. (c) Lee, Y.; Sayre, L. M. J. Am. Chem. Soc. 1995,
117, 11823-11828.
(4) Mure, M.; Klinman, J. P. J. Am. Chem. Soc. 1995, 117, 8707-8718.
(5) Silverman, R. B. Mechanism-Based Enzyme InactiVation: Chemistry and
Enzymology; CRC Press: Boca Raton, FL, 1988; pp 3-30.
(1) (a) Knowles, P. F.; Dooley, D. M. Metal Ions Biol. Sys. 1994, 30, 361-
403. (b) Anthony, C. Biochem. J. 1996, 320, 697-711. (c) Klinman, J. P.
Chem. ReV. 1996, 96, 2541-2561.
(2) (a) Tanizawa, K. J. Biochem. (Tokyo) 1995, 118, 671-678. (b) Dooley,
D. M. J. Biol. Inorg. Chem. 1999, 4, 1-11. (c) Klinman, J. P. J. Biol.
Chem. 1996, 271, 27189-27192.
9
10.1021/ja0205434 CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 12135-12143
12135

A R T I C L E S
Lee et al.
Scheme 2
competitively inhibit several enzymes but will have a long-acting
effect only on the enzyme that suffers inactivation. Several
mechanism-based inactivators for the copper amine oxidase
family have been reported over the years. In most cases, the
overall approach is to take advantage of the change in
hybridization of the amino group-bearing carbon from sp3 to
sp2 that occurs following transamination, in the same manner
that Abeles and Maycock6 pioneered for the development of
amino acid-based inactivators of PLP enzymes. Thus, turnover
of â-haloamines or â,γ-unsaturated amines results in R-haloal-
dehydes or R,â-unsaturated aldehydes, respectively, that may
bind covalently to active-site nucleophiles, thereby affording
irreversible inhibition. However, other enzymes that carry out
the same overall stoichiometry of RCH2NH2 f RCHdO, such
as the flavin-dependent monoamine oxidase (MAO) associated
with the outer mitochondrial membrane, would potentially also
be irreversibly inactivated by such agents. Examples of com-
pounds that inhibit both the quinone- and flavin-dependent
amine oxidases have been reported.7
members of the copper amine oxidase family indicated that
simple secondary amines (e.g., N-methylbenzylamine) are not
substrates. According to the transamination mechanism (Scheme
1), processing of a secondary amine (Scheme 2) would lead to
an iminium substrate Schiff base and then an iminium product
Schiff base that, upon hydrolysis, would not afford the normal
reductively aminated cofactor; one alkyl arm of the secondary
amine would still be attached. It is probable that the enzyme
evolved in a manner to deliberately sterically exclude secondary
amine substrates, so that it would not have to deal with oxidation
of the N-alkyl reductively aminated cofactor. We postulate that
steric exclusion, rather than the issue of forming quaternary
Schiff base intermediates, is the main reason the copper amine
oxidases do not normally metabolize secondary amines.
It was thus of great interest to find that, despite detection of
no O2 uptake when BPAO was incubated with up to 10 mM
pyrrolidine or 3-pyrroline, preincubation with 3-pyrroline
resulted in a concentration- and time-dependent irreversible
pseudo-first-order loss of BPAO activity.12 In this regard, it is
clear that irreversible inactivation serves as a much more
sensitive test of enzymatic processing than does substrate
activity. Kitz and Wilson analysis of the first-order inactivation
rate constants for 3-pyrroline at 30 °C, pH 7.2, yielded values
Substantial evidence has been accrued suggesting that MAO-A
and -B oxidize their amine substrates by a mechanism initiated
by electron transfer and involving radical intermediates.8
Although other mechanisms for MAO have also been con-
sidered,9-11 it is clear that these flavin-dependent reactions do
not utilize an imine shift (transamination) mechanism. Thus,
amines capable of effecting a transamination-specific irrevers-
ible modification should potentially inactivate only the quinone-
dependent copper amine oxidase family. Such compounds would
be superior probes for evaluating the physiological functions
of these enzymes and may have therapeutic potential as selective
inhibitors.
We previously reported that 3-pyrrolines inactivate bovine
PAO (BPAO) and that model studies with a TPQ mimic suggest
that inactivation reflects irreversible pyrrolylation of the enzyme
quinone cofactor through a transamination-dependent mecha-
nism.12 This finding was of special interest in that it represented
the first observation of processing of a secondary amine by this
enzyme class. Although 3-pyrroline itself was a weak inactivator
of BPAO, its 3-phenyl analogue exhibited improved inhibitory
properties, a finding in line with this enzyme preferentially
metabolizing arylalkylamines rather than simple alkylamines.
In this paper, we divulge full details of the inhibitory action of
a variety of 3-aryl-3-pyrrolines on BPAO, as well as of model
studies directed at elucidating their mechanism of action. In
addition, we show that the parent 3-phenyl analogue is an
excellent pure substrate (no inactivation) of MAO-B purified
from beef liver.
of Ki ) 50 mM and kinact ) 0.3 min-1 12
The rather high Ki
.
reflects the expected very weak binding for this simple aliphatic
amine. Furthermore, loss of the characteristic 480 nm absorption
of the TPQ anion, and the absence of an increase at A440
normally observed upon derivatization of the quinone cofactor
with phenylhydrazine, suggested complete loss of the cofactor
quinone character.
Reaction of Pyrrolidine and 3-Pyrroline with TPQ Models.
The reactivity of TPQ with pyrrolidine and 3-pyrroline was
explored with TPQ models, as monitored by NMR spectroscopy
in CD3CN. After 2 days, the NMR tube reaction of quinone 1a
with pyrrolidine (3 equiv) was seen to generate (pyrrolidino)-
resorcinol 6a (Scheme 3) in 40% yield, based on integration,
along with other unidentified compounds. The appearance of
6a is understood in terms of generation of quinoniminium
intermediate 3a, and its reduction by a product of redox turnover,
either 2a arising from addition-elimination (Scheme 3, path
A) or (alkylamino)resorcinol 5a arising from transamination via
4a (Scheme 3, path B).
Results and Discussion
Irreversible Inactivation of Bovine Plasma Amine Oxidase
by the Secondary Amine 3-Pyrroline. Early studies on several
(6) Abeles, R. H.; Maycock, A. L. Acc. Chem. Res. 1976, 9, 313-319.
(7) (a) Lyles, G. A.; Fitzpatrick, C. M. S. J. Pharm. Pharmacol. 1984, 37,
329-335. (b) Yu, P. H.; Zuo, D.-M. Biochem. Pharmacol. 1992, 43, 307-
312. (c) Riceberg, L. L.; Simon, M.; Van Vunakis, H.; Abeles, R. H.
Biochem. Pharmacol. 1975, 24, 119-125.
(8) Silverman, R. B. Acc. Chem. Res. 1995, 28, 335-342.
(9) Miller, J. R.; Edmondson, D. E. Biochemistry 1999, 38, 13670-13683.
(10) Kim, J.; Bogdan, M. A.; Mariano, P. S. J. Am. Chem. Soc. 1993, 115,
10591-10595.
Operation of the transamination path B would mandate
generation of 4a/5a. When the reaction mixture of 1a and
pyrrolidine was acid-quenched (with HCl), evaporated, and
analyzed by 1H NMR in D2O, signals corresponding to 4a were
(11) Anderson, A. H.; Kuttab, S.; Castagnoli, N., Jr. Biochemistry 1996, 35,
5-3340.
(12) Lee, Y.; Huang, H.; Sayre, L. M. J. Am. Chem. Soc. 1996, 118, 7241-
7242.
1
observed, verified by comparison to the H NMR spectrum of
9
12136 J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002

Interactions of 3-Pyrrolines with Amine Oxidases
A R T I C L E S
Scheme 3
Scheme 4
the authentic compound independently synthesized by acid
deprotection of 8a (eq 1). However, 4a/5a could also arise via
path A if 4-aminobutanal, generated from 1-pyrroline, condensed
with 1a or 3a to give 7a, which was then reduced to 5a by 2a.
The reaction of pyrrolidine was repeated with the tert-butyl
TPQ model 1b to aid in identification of reaction products
besides 6. In addition to the pyrrolidinoresorcinol 6b, a product
9b was isolated (eq 2), formally representing condensation
(electrophilic aromatic substitution) of 6b with 1-pyrroline. Since
an aryl ring “recognition element” to 3-pyrroline would improve
binding and thus inhibitory efficiency. A series of 3-aryl-3-
pyrrolines 14a-f was synthesized from N-(carboethoxy)-3-
pyrrolidone by the Grignard/deprotection/dehydration sequence
shown in Scheme 5.13,14
Scheme 5
1-pyrroline can arise either from the addition-elimination
mechanism (Scheme 3, path A) or from hydrolysis of 7 formed
from oxidation of 5 (Scheme 3, path B), the discovery of 9b
does not help distinguish between operation of the two mech-
anisms considered in Scheme 3.
In contrast to pyrrolidine, reaction of 3-pyrroline with 1b in
CD3CN was found to give 11b quantitatively (Scheme 4) and
not a trace of triol 2b, pyrrole, or (3-pyrrolino)resorcinol
analogous to 6. The lack of side products is understood in terms
of the irreversible deprotonation of 10b giving 11b, thereby
averting generation of a reducing species analogous to 5a. This
result highlights the superb utility of 3-pyrroline as a mechanistic
probe for a transamination mechanism. In addition, this observa-
tion suggests the dominance of transamination over addition-
elimination also for pyrrolidine (Scheme 3), because partitioning
of the initial carbinolamine intermediate between elimination
and dehydration should hardly be affected by the presence or
absence of the remote CdC in 3-pyrroline.
To permit spectroscopic studies on interaction of the inhibitor
with BPAO, a nitro-containing analogue was also sought.
Attempts to prepare 3-(4-nitrophenyl)-3-pyrroline by direct
nitration of 14a or its N-acetyl analogue, under a variety of
nitration conditions, all caused aromatization or destruction of
the pyrroline ring. This failure raised concern that the desired
3-(4-nitrophenyl)-3-pyrroline might be unstable with respect to
internal oxidation-reduction to give 3-(4-nitrosophenyl)pyrrole.
(13) Wu, Y.; Gould, W.; Lobeck, W. G., Jr.; Roth, H. R.; Feldkemp, R. F. J.
Med. Chem. 1962, 5, 752-769.
(14) Gould, W. A.; Lish, P. M.; Wu, Y.-H.; Roth, H. R.; Lobeck, W. G., Jr.;
Berdahl, J. M.; Feldkemp, R. F. J. Med. Chem. 1964, 7, 60-67.
Synthesis of 3-Aryl-3-pyrrolines. Since the preferred sub-
strates of BPAO are arylalkylamines, it was surmised that adding
9
J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002 12137

A R T I C L E S
Lee et al.
Table 1. Inactivation of BPAO by 3-Aryl-3-pyrrolines (14a-g)
inactivator
concentration (mM)
t1/2a (min)
14a
14b
14c
14d
14e
14f
0.4
0.4
7
10
31
45
13
6
0.4
0.1b
0.4
0.2
14g
0.015
3
a Incubation time required for 50% inactivation. b Limited solubility of
14d precluded its evaluation at higher concentration.
Thus it was decided alternatively to prepare 3-(4-methoxy-3-
nitrophenyl)-3-pyrroline (14g), presuming that the nitration of
4-methoxy analogue 14b could be accomplished smoothly and
that, with the nitro group being meta rather than para, the
compound would be stable with respect to internal redox.
Unfortunately, even very mild nitration of 14b or its N-acetyl
derivative again resulted in complete loss of the pyrroline ring,
indicating that nitration would have to be effected prior to
generation of the 3-pyrroline CdC. In this regard, nitration of
alcohol 13b in acetic anhydride with copper nitrate at room
temperature successfully afforded the acetylated nitro product
15; the latter was then heated at reflux with aqueous hydro-
chloric acid to give the target molecule 14g (eq 3).
It was considered that the synthetic route could be shortened
by one step if the carbethoxy precursor 12b could be nitrated,
and the carbamate could be hydrolyzed in the final acid
dehydration step. Although nitration of 12b proceeded smoothly,
conditions could not be found for acid hydrolysis of the ethyl
carbamate that preserved the CdC presumed to form quickly.
In addition, attempted basic hydrolysis of the nitrated carbamate
resulted in nucleophilic aromatic substitution (HO replacing
CH3O). Thus, the pathway summarized in eq 3 appears to be
the best route to 14g. Finally, although the successful nitration
of alcohol 13b suggested that we might have been able to
prepare the originally sought 4-nitrophenyl analogue via nitration
of 13a, this attempt resulted in dehydration to the exclusion of
nitration of the benzene ring.
Evaluation of 3-Aryl-3-pyrrolines as Inhibitors of Bovine
Plasma Amine Oxidase. As was the case for the parent
3-phenyl analogue 14a,12 all the 3-aryl analogues were more
potent time-dependent inhibitors of BPAO than 3-pyrroline
itself. Table 1 lists inactivation half-life data for single con-
centrations of the 3-aryl-3-pyrrolines. The presence of electron-
donating 4-methoxy (14b) and 4-(dimethylamino) (14c) sub-
stituents and the sterically bulky 1-naphthyl substituent (14e)
resulted in less potent inactivators than the parent 3-phenyl
analogue 14a. On the other hand, the 2-naphthyl analogue 14f
and especially the 3-nitro-4-methoxyphenyl analogue 14g were
more potent than 14a. The latter two compounds and the parent
14a all displayed a pseudo-first-order loss of BPAO activity
[Figure 1(top) and Supporting Information]. Kitz and Wilson
replots of the inactivation rate data [Figure 1 (bottom) and
Figure 1. Time-dependent inhibition (top panel) of BPAO by 3-(4-
methoxy-3-nitrophenyl)-3-pyrroline (14g) and Kitz and Wilson replot of
the data (bottom panel).
Supporting Information] in all three cases came too close to
the 0-0 origin to justify an attempt to estimate the intercepts.
We suspect that the apparent lack of saturation5,15 reflects a
more efficient turnover (leading to inactivation) at the expense
of a binding limited by the normal steric exclusion of secondary
amines. For comparative purposes, inactivation by 14a, 14f, and
14g can be quantified in terms of second-order rate constants,
230, 528, and 6970 M-1 min-1, respectively, which can be
compared to kinact/Ki ) 6 M-1 min-1 for 3-pyrroline itself.12
The high potency of 14g clearly reflects the electron-withdraw-
ing nitro group. For all three cases, enzyme activity was not
recovered by gel filtration, indicative of irreversible modifica-
tion, and the absence of an increase in A440 upon treating the
inactivated enzyme with excess phenylhydrazine is suggestive
of irreversible modification of the quinone cofactor. The
inactivation kinetics of the other analogues were not fully
characterized, but they most likely follow the same irreversible
behavior as do 14a, 14f, and 14g.
Additional evidence for modification of the quinone cofactor
is provided by the nitroblue tetrazolium (NBT) redox cycling
(15) Saturation may become apparent at low temperature.
9
12138 J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002

Interactions of 3-Pyrrolines with Amine Oxidases
A R T I C L E S
assay16 carried out on a sample of purified BPAO that was
inhibited by 14a to the extent of 98.4% and then denatured.
Compared to the color intensity of the control sample as
representing 100% redox activity, the inhibited sample had no
detectable redox cycling activity. On the basis of the ac-
cumulated enzymologic data (including the observed lack of
O2 uptake) and the model study results, the most likely
explanation for the observed behavior of the 3-pyrrolines is that
they induce a stoichiometric covalent modification of the
quinone cofactor in a single turnover. To determine whether a
native enzyme structure is required for the modification of the
quinone cofactor to occur, the NBT assay was conducted on a
sample of enzyme treated with the inhibitor after denaturation.
In this case, the NBT staining was not totally eliminated, as it
was when the incubation with inhibitor occurred prior to
denaturation (data not shown). The finding that there was still
about 60% reduction of the NBT staining observed for the
control enzyme sample indicates that the denatured enzyme is
still capable of reacting with 14a, though not to the extent that
occurs for the native enzyme. We interpret this difference to
indicate the extent to which cofactor derivatization by 14a is
facilitated by the normal catalytic assistance of the active site.
Additional support for compounds 14 satisfying the criteria
of mechanism-based inactivation is that substrate benzylamine
effected a concentration-dependent (but incomplete) protection
against inhibition by 14f (see Experimental Section) in the
presence of catalase used to prevent H2O2-dependent inactivation
by the benzylamine.17
Figure 2. Spectra of 14g (9.60 × 10-4 M) and 16b (1.70 × 10-3 M).
Spectroscopic Analysis of BPAO Inactivated by 3-(4-
Methoxy-3-nitrophenyl)pyrroline (14g). The results of the
model studies suggested that the mechanism of inactivation of
BPAO by 3-pyrroline and its 3-aryl analogues involves con-
densation with the TPQ cofactor and transamination to give,
following loss of proton, the corresponding pyrrolylated cofac-
tor.12 The availability of chromophoric inhibitor 14g suggested
that direct evidence for the nature of enzyme inactivation might
be obtained by comparing the spectral properties of the 14g-
inactivated enzyme with those of small molecule adducts
generated by reaction of 14g with TPQ models. Thus, 14g was
condensed with quinone cofactor models 1a and 1b in buffered
aqueous CH3CN, yielding the corresponding model cofactor
pyrrole derivatives 16a and 16b (eq 4). Absorption spectra of
Figure 3. Spectrum of native BPAO (16.95 µM TPQ), spectrum of BPAO
inactivated by 1 equiv of 14g at 30 °C for 3.5 h with subsequent gel filtration
and dilution to 17 µM of 14g-BPAO monomer complex, and difference
spectrum.18
has λmax at 371 nm with ꢀ ) 570 M-1 cm-1. From these data,
one would expect that derivatization of the native cofactor to a
pyrrole derived from reaction with 14g should result in an
inactivated enzyme spectrum with a loss of the native 480 nm
quinone chromophore and a gain of absorbance near 370 nm.
In Figure 3 are shown spectra for native BPAO and for a
sample of enzyme inactivated by 14g. The inactivated enzyme
spectrum exhibits the expected loss at 480 nm and a shoulder
near 370 nm, clearly consistent with the proposed enzymatic
oxidation of 14g to give the respective TPQ-pyrrole derivative.
Though discerning the shoulder at 370 nm is limited by the
strong background absorbance in this region, the estimated
difference spectrum,18 also shown in Figure 3, shows the
shoulder at 371 nm more prominently. Furthermore, the
increased absorbance in the spectrum of the inactivated enzyme
between 300 and 350 nm, as compared to native BPAO, is
attributed to a strong absorbance of the TPQ-14g derivative
(18) Calculation of the difference spectrum required precise knowledge of the
concentration of both native and inactivated enzyme cofactor. Whereas this
could be calculated in the former case from the known extinction coefficient
of enzyme-bound TPQ, this was estimated for the BPAO-14g TPQ
derivative by assuming that the shoulder at 370 nm represented the proposed
pyrrole and that the displacement of the shoulder over the background rise
in absorbance in this region could be converted by use of the same extinction
coefficient as exhibited by model 16b at 370 nm. The concentration
estimated in this way agreed reasonably well with that based on dilution
of the sample obtained following gel filtration.
14g (0.96 mM, HCl salt) and 16b (1.7 mM) in methanol are
displayed in Figure 2. Compound 14g has λmax at 340 nm (355
in H2O) with ꢀ ) 1960 M-1 cm-1 (2160 in H2O), while 16b
(16) Fluckiger, R.; Paz, M. A.; Gallop, P. M. Methods Enzymol. 1995, 258,
14-19.
(17) Lee, Y.; Shepard, E.; Smith, J.; Dooley, D. M.; Sayre, L. M. Biochemistry
2001, 40, 822-829.
9
J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002 12139

A R T I C L E S
Lee et al.
the range 1350-1358 cm-1 in the model spectra shown is
presumably due to different solvent interactions. The data
obtained for model 16b in 63% CD3OD and 37% D2O (not
shown) depict a shift from 1358 cm-1 (in 100% deuterated
methanol) to 1355 cm-1. Additionally, a shift from 1344 to 1350
cm-1 was observed in the rR spectrum of a denatured prepara-
tion of BPAO inactivated by 14g (data not shown; see
Experimental Section). Except for this shift, which we believe
to be solely due to interactions with solvent, the rR spectrum
for the denatured sample was identical to that shown in Figure
4 for the nondenatured enzyme. We can tentatively assign the
1344-1358 cm-1 frequency (inactivated enzyme and models)
to the nitro aromatic stretch (aryl-NO2) originating from 14g,
as stretching modes for such groups range from ∼1330 to 1360
cm-1 20
This assignment is also in agreement with the 1357
.
cm-1 mode seen in the Raman spectrum for 14g itself in H2O
(data not shown) and with our previous rR analysis of BPAO
inactivated by 4-nitrobenzylamine, where a 1344 cm-1 mode
was observed.17
Three or more coupled CdC stretches have a characteristic
vibration between 1540 and 1620 cm-1 20
, so it is possible that
the 1565 and 1620 cm-1 modes are due to the aryl conjugation
present in the adduct. Since there are modes at 1570 and 1620
cm-1 in the Raman spectrum of 14g (data not shown), it is
plausible that the frequencies at 1565 and 1620 cm-1 are due
to the coupled deformation of ring modes in the TPQ-14g
derivative. It should be noted that the apparent shift of the mode
at 1565-1567 cm-1 observed for model 16b to 1556 cm-1 for
16a in 75% CD3OD and 25% D2O is not a solvent shift and
must reflect subtle differences between the two quinone models,
since this mode in the rR spectrum of 16b in 63% CD3OD and
Figure 4. Resonance Raman spectra of 14g-derived pyrrole model
compounds (approximately 5 mM) and of BPAO inactivated by 14g (∼200
µM BPAO dimer). P* denotes protein mode.
37% D2O still occurs at 1565 cm-1
.
chromophore in this region, as seen in the spectrum of model
In the rR spectrum of the BPAO-methylamine adduct, an
intense mode at 1616 cm-1 was assigned to the NdC stretch
of the product Schiff base.21 The absence of a stretching
frequency at 1616 cm-1 in the spectrum of the inactivated
enzyme is consistent with resonance delocalization in a pyrro-
lylated cofactor, rather than any derivative containing a localized
imine. Overall, the similarities in the spectra of models 16a/
16b and the inactivated BPAO-14g derivative indicate that the
pyrrole structure seen to form in the model reaction is a good
candidate for the structure of the adduct formed enzymatically.
16b (Figure 2).
Additional evidence to support enzymatic transformation of
14g to the TPQ-pyrrole derivative was obtained through
comparison of resonance Raman (rR) data on the inactivated
enzyme and respective 14g-derived models 16a and 16b (Figure
4). In native BPAO, the region from 1450 to 1690 cm-1 is
dominated by TPQ ring stretches, with the C5dO stretching
mode at 1678 cm-1 and a protein mode at 1450 cm-1 19
the observation that the amide I vibration is between 1650 and
1660 cm-1 20 we tentatively assign the 1652 cm-1 stretch as a
. Given
,
Interaction of 3-Phenyl-3-pyrroline (14a) with Bovine
Liver Mitochondrial Monoamine Oxidase B. The reaction of
14a with MAO-B was followed spectrophotometrically at 30
°C. Over a period of 1 h, there was observed a monotonic
conversion of the spectrum of 14a (λmax 251 nm) to one
exhibiting λmax values at 225 and 277 nm with isosbestic points
at 236 and 264 nm (difference spectrum is shown in Figure 5),
at which point no further spectral change was seen. The final
spectrum was identical with that of independently prepared
3-phenylpyrrole (18), indicating that 14a is a substrate for clean
dehydrogenation by MAO-B, presumably first giving isopyrrole
17, which rapidly tautomerizes to 18 (eq 5). In addition,
protein mode, in agreement with previous reports.21,22 Analysis
of the rR data for the inactivated enzyme reveals the presence
of the 1450 and 1652 cm-1 modes and the absence of the C5d
O stretch. The latter suggests that the C5dO bond has been
altered upon modification by 14g.
Comparison of the rR spectra of the inactivated enzyme and
respective model compounds reveals similar vibrational patterns,
with the 1565 and 1620 cm-1 modes matching exactly. We
propose that the 1344 cm-1 mode in the inactivated enzyme
spectrum relates to the 1350-1358 cm-1 modes seen in the
model rR data in solution. The variation of the latter mode over
(19) Nakamura, N.; Matsuzaki, R.; Choi, Y. H.; Tanizawa, K.; Sanders-Loehr,
J. J. Biol. Chem. 1996, 271, 4718-4724.
(20) Lin-Vien, D.; Colthup, N.; Fately, W.; Grasselli, J. The Handbook of
Infrared and Raman Characteristic Frequencies of Organic Molecules;
Academic Press: Boston, 1991.
(21) Nakamura, N.; Moe¨nne-Loccoz, P.; Tanizawa, K.; Mure, M.; Suzuki, S.;
Klinman, J. P.; Sanders-Loehr, J. Biochemistry 1997, 36, 11479-11486.
(22) Lord, R. C.; Yu, N. T. J. Mol. Biol. 1970, 50, 509-524.
9
12140 J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002

Interactions of 3-Pyrrolines with Amine Oxidases
A R T I C L E S
Conclusions. Our studies have shown that the inactivation
of bovine plasma amine oxidase by 3-pyrrolines generalizes to
3-aryl-3-pyrrolines with differing phenyl ring substituents. The
fact that the 4-methoxy-3-nitrophenyl analogue is one of the
most potent BPAO inactivators known (15 µM effects 50%
inactivation in 3 min at 30 °C), is remarkable, considering that
secondary amines were previously thought to be nonsubstrates
for this enzyme. It appears that inactivation efficiency is greatly
improved by electron-withdrawing groups that enhance amine
CR deprotonation, as has been seen for the substrate activity of
substituted benzylamines.25 Model studies, supported by spec-
troscopic studies on the inactivated enzyme, suggest a mecha-
nism involving a unique conversion of the quinone cofactor to
a reduced pyrrolylated form that cannot be reoxidized to the
starting quinone. There is no evidence for competing productive
turnover, and indeed, cofactor pyrrolylation would correspond
to mechanism-based inactivation with a partition ratio of 0. In
contrast, for bovine liver mitochondrial monoamine oxidase B,
the parent 3-phenyl analogue 14a is an excellent substrate for
dehydrogenation to 3-phenylpyrrole. The action of a specific
class of compounds as pure inactivators of one enzyme class
and pure substrates for another enzyme class is extraordinary,
permitting us to refer to the 3-pyrrolines as transamination-
specific mechanism-based inactivators.
The finding that the 3-pyrrolines are selective inhibitors of
the quinone-dependent amine oxidases does not preclude the
ability of appropriately substituted 3-pyrrolines to exert selective
inactivation of particular enzymes within this family. Indeed,
preliminary studies on 14a against a series of purified amine
oxidases indicate that whereas 1 equiv (based on active sites)
of 14a effected ∼50% loss of activity of equine plasma amine
oxidase (EPAO) and Pichia pastoris lysyl oxidase (PPLO) in
addition to BPAO, 40 equiv was needed to effect ∼80%
inhibition of Arthrobacter globiformis phenethylamine oxidase
(AGAO), and 40 equiv effected only 13-16% inhibition of pea
seedling amine oxidase (PSAO) and human kidney amine
oxidase (HKAO).26 Thus, in addition to selective inhibition of
the quinone-dependent amine oxidases on the basis of mecha-
nism, a further level of discrimination can be achieved by fine-
tuning preferential recognition by the enzyme substrate binding
cavity. Overall, the development of highly selective inhibitors
should be possible, and such inhibitors should aid in unraveling
the only partly understood physiological functions of the various
mammalian quinone-dependent amine oxidases, as well as
serving as potential drug candidates.
Figure 5. Time course of MAO-B-mediated oxidation of 3-phenyl-3-
pyrroline (14a). The spectrophotometer was blanked on a solution of 14a
(100 µM) in 50 mM sodium phosphate, pH 7.2, at 30 °C. Then 25 nmol of
MAO-B was added and the spectral changes were followed over time. The
50 min spectrum overlapped completely with the 30 min spectrum.
following incubation of 14a with MAO-B for 1 h at room
temperature, reverse-phase HPLC analysis of the evaporated
ether extract of the reaction mixture indicated complete conver-
sion of starting 14a to a single product with identical retention
time to that seen for authentic 18. Although mass spectrometry
was also confirmatory, exhibiting the expected strong molecular
ion at m/z 143, it was found that starting 14a also exhibited a
m/z 143 mass spectrum, owing to apparently facile dehydro-
genation in the mass spectrometer.
Steady-state MAO-B substrate kinetics (25 °C) for 14a, where
the ABTS-horseradish peroxidase coupled assay was used to
monitor generation of the stoichiometric byproduct H2O2,
afforded Michaelis parameters Km ) 252 µM and kcat ) 13.9
min-1, indicating that 14a is an excellent substrate for MAO-
B. Tight binding of 14a was also apparent by its evaluation as
a reversible competitive inhibitor of the MAO-B-mediated
oxidation of benzylamine. The Dixon plot of the data (30 °C)
revealed a Ki of 35 µM for 14a. When the benzylamine oxidase
activity of MAO-B was followed over time (25 °C) in the
presence of various concentrations of 14a, concentration-
dependent inhibition was observed during the first few hours,
but no decrease in enzyme activity was observed at 20 h relative
to the control incubation containing no 14a. Thus, 14a appears
to be a pure substrate for MAO-B with no irreversible inhibitory
properties.
Experimental Section
The substrate activity of 3-phenyl-3-pyrroline (14a) for
MAO-B is consistent with the recent report that the tertiary
N-methyl derivatives of 3-pyrroline, 3-phenyl-3-pyrroline, and
ring-substituted isoindolines are also substrates for MAO-B
(presumably undergoing conversion to the corresponding pyr-
roles) whereas N-methylpyrrolidine is not.23 Since the latter
compound lacks an allylamine fragment, it appears that MAO-B
substrate behavior for both the secondary and tertiary five-
membered cyclic amines depends on the presence of allylic
unsaturation, just as in the case where MAO-B substrate activity
is seen for 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)
but not for 1-methyl-4-phenylpiperidine.24
General Methods. NMR spectra were obtained at 300 MHz (13C
NMR at 75 MHz), with chemical shifts referenced to the solvent peak.
In the 13C NMR spectra, attached proton test (APT) designations are
given as (+) for CH2 and C and (-) for CH3 and CH. High-resolution
mass spectra (HRMS), electron impact or fast atom bombardment
(FAB), were obtained at 20-40 eV on a Kratos MS-25A instrument.
UV-vis spectra were obtained on either Perkin-Elmer Lambda 3
(chemical studies) or Lambda 10 (kinetics studies) spectrophotometers
fitted with a jacketed (temperature-controlled) cell compartment.
(24) Heikkila, R. E.; Manzino, L.; Cabbat, F. S.; Duvoisin, R. C. J. Neurochem.
1985, 45, 1049-1054.
(25) Hartmann, C.; Klinman, J. P. Biochemistry 1991, 30, 4605-4611.
(26) Purification of these enzymes and conduct of their activity assays is as
described: Shepard, E. M.; Smith, J.; Elmore, B. O.; Kuchar, J. A.; Sayre,
L. M.; Dooley, D. M. Eur. J. Biochem. 2002, 269, 3645-3658.
(23) Wang, Y.-X.; Mabic, S.; Castagnoli, N., Jr. Bioorg. Med. Chem. 1998, 6,
143-149.
9
J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002 12141

A R T I C L E S
Lee et al.
Absorption spectra were taken on a Hewlett-Packard 8453 photodiode
array spectrophotometer. Raman spectra were collected on a Spex
Triplemate 1877 (1800 groove setting), with a liquid-N2 cooled Spex
Spectrum One CCD detector and Coherent Ar ion laser. Doubly distilled
water was used for all experiments. All solvents, reagents, and organic
fine chemicals were the most pure available from commercial sources.
Pyrrolidine was freshly distilled before use. Bovine plasma amine
oxidase for kinetic studies (BPAO, 100 units/g of protein) and Sephadex
G-25 were purchased from Sigma. The concentration of BPAO active
monomer was estimated from the rate of benzylamine oxidation (1 unit
oxidizes 1.0 µmol of benzylamine to benzaldehyde per minute at 25
°C), using an activity of 0.48 unit/mg of protein for the pure monomer
of molecular weight 85 00027 and ∆ꢀ250 ) 12 800 M-1 cm-1 for
benzaldehyde [corresponding to an A250 of 12.8 min-1 (unit of activity)-1
for 1 mL volume]. BPAO used for spectroscopic studies was purified
as described previously.27 Bovine mitochondrial monoamine oxidase
B was isolated and assayed as described previously28 and stored as a
concentrated solution (15-25 mg/mL) in 50 mM sodium phosphate
buffer, pH 7.2, at 4 °C. O2 uptake was monitored by a Yellow Springs
Instruments 5300 biological oxygen meter. All evaporations were
conducted at reduced pressure in a rotary evaporator.
and mass spectral data on the mixture and then subtracting out signals
due to 6b. 9b: 1H NMR (CD3CN) δ 1.32 (s, 9H), 1.56 (m, 1H), 1.87
(m, 2H), 1.92 (m, 4H), 2.23 (m, 1H), 2.88 (m, 4H), 3.05 (m, 2H), 4.62
(dd, J ) 9.60 and 7.11 Hz, 1H), 6.96 (s, 1H); 13C NMR (CD3CN) δ
25.1(2C), 26.1, 30.1(3C), 33.2, 35.1, 45.8, 54.7(2C), 57.6, 112.2, 119.8,
126.8, 128.2, 149.8, 156.7; HRMS calcd for C18H28N2O2 m/z (rel
intensity) 304.2152, found 304.2100 (16.8%).
Reaction of TPQ Model 1b with 3-Pyrroline. To a solution of 1b
(18 mg, 0.10 mmol) in degassed CD3CN (0.6 mL) was added
3-pyrroline (11.6 µL, 0.15 mmol) via syringe. The1H NMR spectrum
was recorded periodically and no further change was seen after 2 h, at
which point the only species present was 4-(pyrrol-1-yl)-6-tert-
butylresorcinol (11b): 1H NMR (CD3CN) δ 1.34 (s, 9H), 4.83 (br s,
OH, 2H), 6.18 (t, J ) 2.14 Hz, 2H), 6.43 (s, 1H), 6.85 (t, J ) 2.14 Hz,
2H), 6.98 (s, 1H); 13C NMR (CD3CN) δ 29.88 (3C), 34.61, 105.61,
109.13 (2C), 122.95 (2C), 125.49, 128.43, 128.84, 150.25, 155.87;
HRMS calcd for C14H17NO2 m/z (rel intensity) 231.1260, found
231.1260 (88.7%).
General Procedure for the Synthesis of Ethyl 3-Aryl-3-hydroxy-
1-pyrrolidinecarboxylates 12. To a solution of ethyl-3-oxo-1-pyrro-
lidinecarboxylate13 (25 mmol) in anhydrous ether (150 mL) was added
dropwise a solution of the substituted aryl Grignard reagent (28 mmol)
(prepared from either bromobenzene, 4-bromoanisole, 4-bromobiphenyl,
1-bromonaphthalene, or 2-bromonaphthalene) in anhydrous ether (100
mL) at room temperature. After the addition was completed, the reaction
mixture was heated at reflux for 10 min and then cooled to room
temperature and poured into 3 M aqueous ammonium chloride (100
mL). The ether layer was separated and dried over sodium sulfate to
afford 12 after evaporation of solvent. In the case of 12c, it was found
necessary to replace the use of diethyl ether with THF as both Grignard
(from 4-bromo-N,N-dimethylaniline) and reaction solvent. The products
were purified by crystallization (white solids) or flash silica gel
chromatography (oils), with isolated yields ranging from 60% to 80%.
NMR analysis indicated that compounds 12 exist as mixtures of syn
and anti carbamate isomers.
Reaction of TPQ Model 1a with Pyrrolidine. Pyrrolidine (7.5 µL,
0.09 mmol) was added to a 5 mm NMR tube containing 2-hydroxy-
5-(2-pivalamidoethyl)-1,4-benzoquinone (1a, 7.5 mg, 0.03 mmol) in
degassed CD3CN (0.6 mL). Integration of the 1H NMR spectrum
recorded after 2 days at room temperature indicated generation of 6a
in 40% yield (by comparison to the authentic compound synthesized
by “redox cycling” reaction3a of triol 2a with pyrrolidine), along with
other unidentified compounds. The reaction mixture was quenched by
pouring it into 0.1 N aqueous HCl (1 mL), and the resulting solution
was evaporated under high vacuum to give a product mixture, which
was dissolved in D2O. In addition to the major product (HCl salt of
6a), signals corresponding to the iminium salt 4a (identified as discussed
1
below) were seen in the H NMR spectrum (∼3% of the amount of
6a).
General Procedure for the Synthesis of 3-Aryl-3-pyrrolidinols
13. A mixture of 10 mmol of the ethyl 3-aryl-3-hydroxy-1-pyrroli-
dinecarboxylate 12, n-propanol (5 mL), H2O (5 mL), and KOH (3 g)
was heated at reflux for 20 h.14 After cooling, the alcoholic layer was
separated and the aqueous layer was extracted with ether. The combined
organic layer was dried over sodium sulfate and concentrated to give
the corresponding 3-aryl-3-pyrrolidinols 13 as white solids, which were
used in the next step without purification (for yields in excess of 80%)
or recrystallized (for yields 60-80%).
General Procedure for the Synthesis of 3-Aryl-3-pyrroline
Hydrochlorides 14a-f. A solution of 3-aryl-3-pyrrolidinol 13 (7 mmol)
in 12 N hydrochloric acid (15 mL) was heated at reflux for 1 h. The
crude residue obtained upon evaporation of solvent was recrystallized
from ethanol or, in the case of 14b, from ethanol-ether.
Time-Dependent Inactivation of BPAO by 3-Aryl-3-pyrrolines.
Solutions (0.9 mL) of 3-aryl-3-pyrrolines (final concentration ranging
from 0.05 to 1 mM for 14a-f and from 1 to 20 µM for 14g) in 100
mM potassium phosphate buffer, pH 7.2, were mixed with 0.1 mL of
a 20 µM solution of BPAO, and the mixture was incubated at 30 °C.
Aliquots (0.1 mL) were periodically withdrawn and diluted with 1.0
mL of benzylamine (10 mM in 50 mM sodium phosphate buffer, pH
7.2) in a 1 cm cuvette (1.5 mL). The rate of oxidation of benzylamine
to benzaldehyde was measured by recording the increase in absorbance
at 250 nm and compared to the rate of benzylamine oxidation in a
companion control solution of enzyme lacking the inhibitor. Since the
Kitz and Wilson plots (inactivation t1/2 vs [inhibitor]-1) went through
the 0-0 origin within experimental error, the inactivation kinetics for
14a were repeated at 5 °C, but the Kitz and Wilson plot still came too
Independent Generation of Iminium Salt 4a. To a 5 mm NMR
tube containing triol 2a (25.3 mg, 0.1 mmol) in CD3CN (0.6 mL) was
added 4-aminobutyraldehyde diethylacetal (90%, 19 µL, 0.1 mmol) via
syringe. The 1H NMR spectrum after overnight standing showed clean
“redox cycling” conversion3a of the starting materials to 4-(4,4-
diethoxybutanamino)-6-(2-pivalamidoethyl)resorcinol (8a): 1H NMR
(CD3CN) δ 1.11 (s, 9H), 1.12 (t, J ) 6.99 Hz, 6H), 1.60 (m, 4H), 2.63
(t, J ) 6.80 Hz, 2H), 3.01 (m, 2H), 3.24 (m, 2H), 3.39-3.65 (4H),
4.46 (m, 1H), 6.36 (s, 2H), 6.76 (br t, J ) 4.38 Hz, 1H); 13C NMR
(CD3CN) δ 15.7 (2C), 25.5, 27.8 (3C), 30.6, 32.2, 39.1, 41.5, 45.6,
61.9 (2C), 103.6, 104.0, 115.4, 117.3, 131.4, 145.0, 147.5, 180.1. The
solution of 8a was poured into 0.1 N aqueous HCl (6 mL), and the
resulting solution was concentrated to dryness under high vacuum to
give iminium salt 4a as a solid: 1H NMR (D2O) δ 1.03 (s, 9H), 2,42
(m, 2H), 2.71 (t, J ) 6.38 Hz, 2H), 3.38-3.42 (4H), 4.58 (m, 1H),
6.54 (s, 1H), 7.21 (s, 1H), 9.02 (br s, 1H); 13C NMR (D2O) δ 18.9,
26.5 (3C), 28.4, 36.4, 38.3, 39.2, 61.7, 103.4, 117.3, 118.5, 125.8, 149.4,
157.4, 180.6, 182.0.
Reaction of TPQ Model 1b with Pyrrolidine. To a solution of
2-hydroxy-5-tert-butyl-1,4-benzoquinone (1b, 5.4 mg, 0.03 mmol) in
degassed CD3CN (0.5 mL) was added pyrrolidine (7.5 µL, 0.09 mmol)
via syringe. After 3 days, all volatiles were evacuated, and the remaining
residue was redissolved in degassed CD3CN (0.5 mL) for recording of
spectral data. Integration of the 1H NMR spectrum indicated a mixture
of 6b and another major product, 2-(pyrrolidin-2-yl)-4-(pyrrolidin-1-
yl)-6-tert-butylresorcinol (9b), in a ratio of 3:4, along with traces of
other side products. Separation of the air-sensitive product mixture was
impractical; thus, identification of 9b was made by acquisition of NMR
close to the 0-0 origin to permit a valid estimate of Ki and kinact
.
Irreversibility of BPAO Inactivation by 3-Pyrroline and 3-Aryl-
3-pyrrolines. The irreversibility of the inhibition for 3-pyrroline,
(27) Janes, S. M.; Klinman, J. P. Biochemistry 1991, 30, 4599-4605.
(28) Banik, G. M.; Silverman, R. B. J. Am. Chem. Soc. 1990, 112, 4499-4507.
9
12142 J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002

Interactions of 3-Pyrrolines with Amine Oxidases
A R T I C L E S
3-phenyl-3-pyrroline, and 3-(2-naphthyl)-3-pyrroline was verified by
applying enzyme preparations that were 50-90% inhibited to a
Sephadex G-25 column (1 × 10 cm) to separate noncovalently bound
small molecules. Remeasurement of the enzyme activity and correcting
for the volume change indicated no detectable regain of enzyme activity
in up to 16 h. Furthermore, incubation of 90% inhibited enzyme (by
3-pyrroline) with phenylhydrazine following Sephadex G-25 chroma-
tography failed to reveal the normal 450 nm complex chromophore.
Nitroblue Tetrazolium Redox Cycling Assay on BPAO Inacti-
vated by 14a. Purified BPAO was incubated with 3.3 mM 14a for 2
h at 30 °C, which achieved >98% inactivation. SDS-PAGE analysis
of the inactivated, control, and phenylhydrazine-inactivated enzyme
incubations were performed by standard methods on a polyacrylamide
slab gel (6% acrylamide with 0.16% bisacrylamide). Inhibited and
control enzyme samples (40 µL) were mixed with 20 µL of standard
denaturing buffer (10% SDS with 0.5 M 2-mercaptoethanol) and the
mixture was heated at 100 °C for 6 min before application in duplicate
to two halves of the gel. After the gel was run (at 0.02 amp constant
current, voltage near 200 V), it was cut in half and the two halves
were stained with either Coomassie blue (0.25% Coomassie in 50%
methanol and 7% acetic acid) or 0.24 mM nitroblue tetrazolium in 2
M potassium glycinate, pH 10, for 120 min (BPAO) in the dark.16 No
alteration of the electrophoretic properties of the enzyme or the intensity
of Coomassie staining was observed.
Substrate Activity of 3-Phenyl-3-pyrroline (14a) for MAO-B. For
spectrophotometric studies, MAO-B (10 µM) was added to a solution
of 14a (0.1 mM) in sodium phosphate buffer, pH 7.2 (100 mM), in a
1 mL cuvette, and the reaction was monitored in the range of 200-
400 nm at 30 °C for 2 h. Steady-state rates of oxidation of 14a were
determined at 25 °C by the horseradish peroxidase-coupled assay for
the determination of released H2O2 with 2,2′-azinobis(3-ethylbenzthi-
azoline)-6-sulfonic acid (ABTS) as chromogen, according to a modi-
fication of the published method.29 In brief, a calibration curve from
serial dilutions of a standard H2O2 solution was first constructed, with
monitoring of the ABTS oxidation product at 414 nm. Then, 5 µL
aliquots of incubations of various concentrations of 14a (0.1-1.2 mM)
with MAO-B (2 µM) at 25 °C were removed at standard periods of
time and diluted with 0.5 mL of a solution of horseradish peroxidase
(10 units/mL) and ABTS (0.2 mM). The amounts of H2O2 formed at
the various times were then read off the calibration plot, and these
values were plotted vs time to obtain the velocity of H2O2 production.
Steady-state kinetic parameters (Km and Vmax) were obtained graphically
from a plot of (velocity)-1 vs [14a]-1. The value for Vmax was converted
to kcat by dividing by [enzyme].
Inhibition of MAO-B-Mediated Metabolism of Benzylamine by
14a. The initial rates of oxidation of various concentrations of
benzylamine (0-2 mM) by MAO-B (10 µM) in the presence of various
concentrations of 14a (0-80 µM) were determined by spectrophoto-
metric monitoring of the production of benzaldehyde at 250 nm in 100
mM Tris buffer, pH 9.0, at 30 °C. A Dixon plot of the data (velocity)-1
vs [14a] was consistent with competitive inhibition, for which the
reversible inhibition constant Ki could be calculated. To address the
possibility of irreversible MAO-B inhibition by 14a, the enzyme (100
µM) was incubated with various concentrations of 14a (0-6 mM) in
100 mM phosphate buffer, pH 7.2, at 25 °C. At various time intervals
up to 20 h, aliquots (20 µL) of the primary incubation mixture were
removed and diluted with 0.5 mL of a solution of benzylamine (1 mM)
in 100 mM Tris buffer, pH 9.0, at 30 °C, and the rate of production of
benzaldehyde was determined spectrophotometrically.
Substrate Protection against BPAO Inactivation by 3-(2-Naph-
thyl)-3-pyrroline (14f). Incubation of BPAO (1.7 µM) with 14f (0.1
mM) in 100 mM phosphate buffer, pH 7.2, for 20 min at 30 °C resulted
in 35% remaining activity after Sephadex G-25 gel filtration. When
the inactivation experiment described above was performed in the
presence of varying amounts of benzylamine (4, 10, 20, and 30 mM)
and catalase (670 units/mL), increased levels of remaining activity
(42%, 46%, 56%, and 64%, respectively) were measured after Sephadex
G-25 gel filtration of the corresponding incubation solution.
Spectroscopic Characterization of Pyrrole Models 16 and BPAO
Modified by 14g. A solution of highly purified BPAO (∼25 µM dimer,
∼50 µM TPQ) was incubated with 1 or 4 equiv of 14g based on TPQ
content at 30 °C for 3.5 or 1 h, respectively. The protein sample was
assayed and found to have no remaining activity. The sample was
subsequently run over a Sephadex G-25M PD-10 column in 0.1 M
phosphate buffer, pH 7.2, and then concentrated to ∼400 µM TPQ,
reassayed, and found to contain no activity. Optical spectra of
appropriate dilutions of native and inactivated enzyme were recorded
(Figure 3). Raman spectra on the undiluted samples (Figure 4) were
collected at room temperature after 413.1 nm excitation. The collection
times were 5 min (at 30 mW) for model 16b in 100% acetonitrile, 8
min (at 70 mW) for model 16a in 75% deuterated methanol and 25%
D2O, 15 min (at 30 mW) for model 16b in 100% deuterated methanol,
and 60 min (at 50 mW) for BPAO inactivated by 14g. Model 16b in
100% chloroform was only collected for 0.5 min (at 20 mW) due to
the formation of a precipitate. Raman spectra were also collected for
model 16b in 63% CD3OD and 37% D2O (8 min at 75 mW), 14g in
H2O (30 min at 20 mW), and a denatured sample of BPAO inactivated
by 14g (13 min at 28 mW). Preparation of the latter sample involved
adding 55.5 µM 14g to a solution of BPAO (52.6 µM in TPQ). Time
point aliquots were removed and assayed for activity. At 3.5 h, 1.6%
activity remained and the sample was subsequently run over a Sephadex
column as described above. The concentrated sample (∼400 µM TPQ)
was reassayed and found to contain 2.1% activity. The sample was
then thoroughly mixed with 0.45% SDS and 0.1 mM dithiothreitol
before being placed into a thermocycler programmed to heat the sample
to 55 °C for 6 min with subsequent cooling to 25 °C. Raman data
were then immediately collected.
Acknowledgment. We are grateful to the National Institutes
of Health for support of this research through Grants GM 48812
(L.M.S.), GM32634 (R.B.S.), and GM 27659 (D.M.D.). We
thank Doreen Brown and Elizabeth Wilkinson for help in
obtaining resonance Raman spectra, Brooke Hale for inhibitory
studies on 14a against PSAO, AGAO, EPAO, and PPLO, and
Heather Heggem for inhibitory studies on 14a and 14g against
HKAO. We also thank Dr. Gang Sun for help with the NBT
assays.
Supporting Information Available: Preparation and charac-
terization of 6a-b; characterization data on 12a-f, 13a-f, and
14a-f; preparation and characterization of 15, 14g, 16a-b, and
18; activity vs time semilog plots and Kitz and Wilson plots
for inactivation of BPAO by 14a and 14f; Lineweaver-Burk
plot for substrate activity of 14a for MAO-B; Dixon plot for
inhibition of MAO-B-mediated oxidation of benzylamine by
14a; activity vs time plot for MAO-B-mediated oxidation of
benzylamine in the presence of various concentrations of 14a
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA0205434
(29) Szutowicz, A.; Kobes, R. D.; Orsulak, P. J. Anal. Biochem. 1984, 138,
86-94.
9
J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002 12143