Highly potent 3-pyrroline mechanism-based inhibitors of bovine plasma amine oxidase and mass spectrometric confirmation of cofactor derivatization

Article abstract of DOI:10.1016/j.bmc.2006.11.025

Despite the quinone-dependent copper amine oxidases being described as having the ability to metabolize unbranched primary amines to the corresponding aldehydes, we previously showed that the secondary amines 3-pyrrolines are metabolized as mechanism-base

Full text of DOI:10.1016/j.bmc.2006.11.025

Bioorganic & Medicinal Chemistry 15 (2007) 1868–1877
Highly potent 3-pyrroline mechanism-based inhibitors of bovine
plasma amine oxidase and mass spectrometric confirmation
of cofactor derivatization
Yuming Zhang, Chongzhao Ran, Guangyin Zhou and Lawrence M. Sayre*
Department of Chemistry, Case Western Reserve University, Cleveland, OH 44106, USA
Received 28 July 2006; revised 5 November 2006; accepted 13 November 2006
Available online 16 November 2006
Abstract—Despite the quinone-dependent copper amine oxidases being described as having the ability to metabolize unbranched
primary amines to the corresponding aldehydes, we previously showed that the secondary amines 3-pyrrolines are metabolized as
mechanism-based inactivators of bovine plasma amine oxidase (BPAO), and that the 3-(3-nitro-4-methoxyphenyl)-substituted ana-
log was a particularly potent and efficient inactivator. We now show that additional 3-aryl-3-pyrrolines containing highly electron-
withdrawing aryl groups (pyridyl, quinolyl, isoquinolyl, and pentafluorophenyl) are some of the most potent inactivators of BPAO
reported to date. We also provide mass spectroscopic confirmation of the proposed mechanism of inhibition involving pyrrolyla-
tion of the active-site cofactor, through identification by MALDI-TOF and LC–ESI-MS/MS of the (3-arylpyrrol-1-yl)resorcinol
derivatives of the cofactor-containing thermolytic peptides.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
by bovine plasma amine oxidase (BPAO) as mecha-
nism-based inactivators.9,10 We proposed, with the sup-
port of optical spectroscopic and model studies, a
transamination of the cofactor-derived iminium substrate
Schiff base and subsequent irreversible aromatization,
resulting in a stable pyrroloresorcinol cofactor derivative
(Scheme 1) that had no tendency to be reoxidized.9,10 We
found that 3-aryl-3-pyrrolines are particularly efficient
irreversible inactivators of BPAO, that electron-with-
drawing aryl substituents increase inactivation potency,
The copper-containing amine oxidases are typically
defined in terms of their ability to convert unbranched
primary amines to the corresponding aldehydes, through
the utilization of an active-site 2,4,5-trihydroxyphenyl-
alanine quinone (TPQ)-dependent transamination
mechanism.1,2 Interest in the selective inhibition of these
enzymes has increased in recent years with the realiza-
tion of their important physiological roles in plants,3,4
and particularly with the recognition that the human en-
zymes in this class, known as semicarbazide-sensitive
amine oxidases, appear to be involved in a spectrum
of metabolic and signaling pathways.5–8
and
that
the
parent
compound
3-phenyl-
3-pyrroline is a pure substrate for the flavin-dependent
mitochondrial monoamine oxidase B from bovine liver.10
On the basis that 3-(4-methoxy-3-nitrophenyl)-3-pyrro-
line exhibited an IC50 of 15 lM at 3 min of incubation,10
we envisioned that additional analogs bearing electron-
withdrawing aryl substituents might exhibit even more
pronounced inactivation potency. We now report on
several new 3-aryl-3-pyrrolines, where the 3-aryl group
is pentafluorophenyl, pyridyl, quinolyl, or isoquinolyl,
some of which are among the most potent inhibitors
of BPAO that have been reported to date. This is
remarkable considering that these compounds are sec-
ondary amines. We also report mass spectroscopic stud-
ies for two analogs that confirm modification of the
active site TPQ cofactor in the inactivated enzyme.
Despite the historical characterization of TPQ-dependent
enzymes as being capable of metabolizing only primary
amines, a factor that distinguishes them from their fla-
vin-dependent cousins, we recently showed that 3-pyrro-
lines, though being secondary amines, are metabolized
Keywords: Copper amine oxidase; Bovine plasma amine oxidase;
Mechanism-based inhibitor; Enzyme inactivator; 3-Pyrroline.
*
Corresponding author. Tel.: +1 216 368 3704; fax: +1 216 368
3006; e-mail: LMS3@case.edu
URL: http://www.case.edu/artsci/chem/faculty/sayre/index.html.
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.11.025

Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
1869
O
O
O
H
N
HO
O
HO
HO
HO
N+
N
H
R
R
B
R
OH
N
OH
X
HO
HO
H
N+
R
R
R = H, Ar
Dead Enzyme
Scheme 1. Mechanism of copper amine oxidase inactivation by 3-pyrrolines.
OH
OH
O
R
R
R
R
b, c
a
d
e
N
N
N
N
N
H
2HCl
O
O
OEt
O
O
OEt
2a-c
5a-c
4a-c
3a-c
1
a: R = 2-pyridyl; b: R = 3-pyridyl; c: R = 4-pyridyl
N
N+
N+
I-
I-
f
e
N
N
HCl
N
H
O
O
6b: N-methyl-3-pyridyl
6c: N-methyl-4-pyridyl
4b: 3-pyridyl
4c: 4-pyridyl
7b: N-methyl-3-pyridyl
7c: N-methyl-4-pyridyl
Scheme 2. Synthesis of 3-pyridyl-3-pyrrolines 5a–c and two corresponding N-methylpyridinium derivatives 7b and 7c. Reagents and conditions: (a)
for 2a and 2b: 2- or 3- bromopyridine, n-BuLi, ether, À78 ꢁC, then 1; for 2c: 4-bromopyridine, THF, n-BuLi, À78 ꢁC, then 1; (b) MeOH/KOH/H2O,
reflux; (c) pH 10–12, Ac2O, 25 ꢁC; (d) SOCl2, 25 ꢁC; (e) EtOH/12 N aqueous HCl (1:1 v/v), reflux; (f) excess CH3I, CH2Cl2, 25 ꢁC.
2. Results and discussion
Attempted conversion of 2a–c to the desired 3-pyrro-
lines by deprotection (aqueous methanolic KOH) and
subsequent strong acid-mediated dehydration as de-
scribed for the preparation of simple 3-arylpyrrolines
failed in these cases (a tarry product mixture was ob-
tained). This failure probably reflects the strong elec-
tron-withdrawing property of the protonated pyridine
ring, that destabilizes the E1 carbocationic intermediate.
2.1. Chemistry
Previous preparations of 3-aryl-3-pyrrolines were based
on the reaction of aryl Grignard reagent with ethyl
3-oxopyrrolidine-1-carboxylate (1). However, in the case
of pyridyl reagents, initial experiments revealed that the
corresponding Grignard reagents reacted inefficiently
with 1. By lithium–halogen exchange, the more reactive
pyridyllithium reagents were generated. Thus, reaction
of 2-, 3-, or 4-bromopyridine with n-butyllithium in
THF at À78 ꢁC,11 and reaction of the resulting pyridyl-
lithiums with 1 gave the isomeric ethyl 3-(2-, 3-, or 4-
Resorting to thionyl chloride to achieve dehydration,12
we feared that subsequent base-mediated deprotection
of the carbamate would result in isomerization of the
C@C. We thus chose to convert the N-protecting group
to N-acetyl prior to dehydration, so that we could use an
acid deprotection step following SOCl2-mediated dehy-
dration. Thus, deprotection of 2a–c and reprotection
with Ac2O followed by SOCl2-mediated dehydration
afforded N-acetyl 3-pyrrolines 4a–c along with variable
amounts of 3-chloro-3-pyridyl and 2-pyrroline side
products. Isolated compounds 4a–c were converted to
pyridyl)-3-hydroxypyrrolidine-1-carboxylates
2a–c
(Scheme 2). For synthesis of 2a and 2b, higher yields
(50–70%) were obtained using ether rather than THF
(15–20% yields) as the reaction solvent. For 2c, because
of the low solubility of 4-pyridyllithium in ether, THF
was employed to give the product in moderate yield.

1870
Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
OH
R
O
R
R
c, d
a
b
N
N
N
N
H
nHCl
O
O
OEt
O
OEt
OEt
5d-f
1
4
d-f
2d-f
For 5d, n = 1; for 5e and 5f, n = 2.
d: R = pentafluorophenyl; e: R = 2-quinolyl; f: R = 1-isoquinolyl
F
F
F
F
N
F
N
Scheme 3. Synthesis of 3-aryl-3-pyrrolines 5d-f. Reagents and conditions: (a) arylbromide, n-BuLi, ether, À78 ꢁC, then 1; (b) SOCl2, 25 ꢁC (c) 48%
HBr, AcOH, reflux; (d) 12 N HCl, 25 ꢁC.
2.2. Evaluation of inhibitory potency
100
The time-dependent inhibition of BPAO by the new
3-aryl-3-pyrrolines was determined by pre-incubating
approximately 1 lM BPAO with varying concentrations
of the inhibitors at 30 ꢁC, pH 7.2. The enzyme activity in
aliquots taken over time was determined by the standard
benzylamine assay, monitoring the rate of formation of
Control
benzaldehyde at 250 nm. As with the example shown
(5e, Fig. 1, remaining in Supplementary data), plots of
log activity versus time indicated strictly pseudo first-or-
der loss of activity without any plateau being apparent
even at low concentrations of inhibitor. Thus, IC50 val-
ues are time-dependent, but a sense of the relative poten-
cy of the various 3-aryl-3-pyrroline analogs relative to
the parent 3-phenyl compound and previously report-
ed10 analogs 8 and 9 (Chart 1) can be gained by compar-
ing the concentrations of inhibitors needed to produce
50% inactivation at a fixed time of 20 min (Table 1).
In addition, Kitz and Wilson replots13 of inactivation
half-life data as a function of concentration were
constructed in an effort to obtain kinetic constants kinact
and Ki. For the weaker 4-pyridyl-substituted inhibitor 5c
(activity vs. time and Kitz and Wilson plots are in Sup-
plementary data), kinetic parameters could be deter-
mined. However, for the more potent inhibitors, the
Kitz and Wilson plots passed so close to the 0–0 origin
(Fig. 2 shows an example for 5e) that we could not en-
sure a statistically significant non-zero intercept. This
behavior was seen for the previously described potent
3-aryl-3-pyrroline inhibitors,9,10 where we suggested that
the apparent lack of saturation presumably reflected ra-
pid conversion to inactivated enzyme of the enzyme–
substrate complex, relative to its dissociation. In these
cases it is deemed best not to attempt to extract the indi-
vidual kinetic constants, and the relative potencies of the
analogs are thus instead best compared in terms of the
slopes of the Kitz–Wilson plots, which formally repre-
sent the bimolecular rate constants kinact/Ki. The overall
rank order of potency within the series 5a–5f is
5a > 5f > 5b > 5e > 5d > 5c, although some of the differ-
ences are small. However, since the electron-withdraw-
ing effects of 2- and 4-pyridyl should be similar, the
substantial difference in potency between 5a and 5c
suggests that the electronic effect is not the only factor
governing inhibitory efficiency. Most importantly, com-
10 μM
8 μM
5 μM
3 μM
1 μM
10
0
10
20
30
40
50
60
Incubation Time (min)
Figure 1. Time-dependent inactivation of BPAO (1 lM) by various
concentrations of 3-(2-quinolyl)-3-pyrroline (5e) in pH 7.2 sodium
phosphate buffer at 30 ꢁC.
the dihydrochloride salts of the target 3-pyridyl-3-pyrr-
olines (5a–cÆ2HCl) using 6 N HCl in 50% aqueous etha-
nol. To further enhance the electron-withdrawing
property of the 3-aryl substituent, pyridinium salts were
made by methylation of the N-acetyl-protected pyrrolines
4b, c with MeI prior to N-deprotection, affording pyrro-
lines 7b, c following hydrolysis. However, 4a failed to
undergo methylation, probably due to steric hindrance.
We later found that the N-carbamoyl group could be re-
moved under non-basic conditions using a mixture of
aqueous 48% HBr and acetic acid, which does not affect
the 3-pyrroline C@C. Thus, additional 3-pyrrolines,
bearing the electron-withdrawing pentafluorophenyl,
2-quinolyl, and 1-isoquinolyl as the 3-aryl substituent,
were prepared using the more expedient route shown
in Scheme 3, where SOCl2-mediated dehydration was
carried out directly on alcohols 2d–f. Although the ethyl
carbamate could not be deprotected with 12 N HCl,
following deprotection with HBr, the resulting HBr salts
of 5d–f salts were transformed to the HCl salts of 5d–f
by co-evaporating three times with an excess of aqueous
HCl.

Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
1871
HO
N
OH
R
10a: R = C(CH3)3
10b: R = CH2CH2CC(CH3)3
O2N
CH3O
H
H
N
N
O
10c: R = CH2CH2CH(NHaa)Caa
8
9
O
O2N
CH3O
HO
N
OH
R
11: R = CH2CH2CH(NHaa)Caa
O
aa = adjacent active-site
amino acids
Chart 1.
Table 1. Inactivation of BPAO by 3-substituted-3-pyrrolines (5a–f,
7b–c, 8, 9) at 30 ꢁC
able coulombic interactions associated with substrate
binding or active-site access, since the strong electron-
withdrawing nature of the pyridinium substituent would
otherwise predict high inhibitory potency. A lack of ste-
ric accommodation of the N-methyl group is a less likely
explanation, since annulation of a second ring onto the
2-pyridyl structure 5a to give compounds 5e and 5f does
not reduce the potency of inhibition. It appears that the
pyrroline 3-aryl group is accommodated in a fairly open
region of the active site.
Inhibitors
20 min IC50 (lM) Kinetic data kinact/Ki
(mMÀ1minÀ1
)
3-Phenyl-3-pyrrolinea 160
0.23
0.53
7.0
8b
9b
5a
5b
5c
70
4.4
0.75
1.5
50
65.
24.
1.2 (Ki = 0.076 mM,
kinact = 0.091 minÀ1
)
5d
5e
5f
6.1
2.7
6.0
13.
40.
2.3. Mechanism of enzyme inactivation: mass spectrom-
etry supports cofactor derivatization
0.86
500
7b
7c
We had previously reported that several 3-aryl-3-pyrro-
line analogs are irreversible inactivators of BPAO, and
that inactivation was accompanied by conversion of
the TPQ quinone cofactor into a redox-inactive form.10
In the current study, we found that preparations of
BPAO inactivated by 5c, 5d, and 5f failed to exhibit
any redox cycling activity using the standard sodium
glycinate nitroblue tetrazolium (NBT) assay.14,15 Other
analogs were not evaluated, but our finding that three
new representative analogs all behaved the same way
as the previously investigated 3-aryl-3-pyrrolines10 sug-
gests that cofactor modification is the norm for this fam-
ily of inhibitor–inactivators. In addition to the loss of
cofactor redox cycling activity, we provided spectro-
scopic evidence that inactivation of BPAO by the chro-
mophoric analog 9 results in loss of the typical TPQ
chromophore at 480 nm, with replacement by an
absorption at 370 nm, consistent with the spectrum seen
for the corresponding pyrroloresorcinol structures (e.g.,
10a and 10b) generated using TPQ quinone models.4
TPQ conversion to the 8-derived pyrroloresorcinol was
also supported by a comparison of resonance Raman
spectra on the inactivated enzyme compared to TPQ
model derivatives 10a and 10b.4 However, because the
resonance Raman data comparisons were imperfect
and because the enzyme absorption at 370 nm could
potentially arise from covalent attachment of the nitro-
phenyl-containing inhibitor to any position on the en-
zyme, we sought additional evidence for strict cofactor
pyrrolization by the 3-aryl-3-pyrroline inhibitors. In this
regard, we endeavored to obtain direct mass spectral
evidence for the pertinent modified cofactor-containing
active-site peptide.
>1000
a Data from Ref. 9.
b Data from Ref. 10.
60
50
40
30
20
10
0
0.0
0.2
0.4
0.6
0.8
1.0
1/[Inhibitor], μM-1
Figure 2. Kitz and Wilson replot of data in Figure 1 for inactivation of
BPAO by 3-(2-quinolyl)-3-pyrroline (5e).
pounds 5a, 5b, 5e, and 5f are all more potent than the
most potent analog 9 reported in our previous work.10
The N-methylated pyridinium derivatives 7b and 7c were
much weaker inhibitors, and we were unable to obtain
kinetic constants. Assuming that all the 3-pyrrolines
inactivate through the same mechanism of cofactor
derivatization, the low potency of 7b and 7c suggests
that the permanent cationic charge results in unfavor-
The sequence of BPAO has been known for some time,1
though the protein is glycosylated to a significant extent

1872
Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
778.90
778.93
3.0
1.4
729.05
891.95
0
0
700
900
m/z
700
900
m/z
Figure 3. Negative ion MALDI-TOF spectra for the thermolytic digest of native BPAO (left panel) and BPAO inactivated by 3-(2-naphthyl)-3-
pyrroline (8) (right panel).
on its surface. The crystal structure of the extensively
deglycosylated enzyme was recently published.16 We fo-
cused on the derivatization/inactivation of BPAO by the
two highly chromophoric inhibitors 3-(2-naphthyl)-3-
pyrroline (8) and 3-(4-methoxy-3-nitrophenyl)-3-pyrro-
line (9) (Chart 1), which would allow LC identification
of peptides covalently modified in the inactivation event
using both optical and mass-selective detectors. A num-
ber of attempts were made to effect efficient proteolytic
digestion of the inhibitor-modified samples of purified
BPAO (without deglycosylation) using trypsin, chymo-
trypsin, proteinase K, pepsin, and a combination of
two enzymes. In part these efforts were made with the
hope of establishing reasonable coverage proteolytic
maps for non-active site as well as active-site regions.
However, using both solution and in-gel digestion (with
followup by MALDI-TOF) approaches, results were
mixed, and we resorted to a thermolytic digestion strat-
egy that was successful in first identifying the active-site
TPQ cofactor following phenylhydrazine derivatiza-
tion.1 In this study, it was noted that mass spectrometric
analysis of the digest without derivatization of the cofac-
tor fails to identify it because the quinone tends to con-
dense with released a-amino groups.1 Through use of a
modification of the published approach using thermoly-
sin, we were able to identify the active site peptide pyrr-
olated by 8 by MALDI-TOF, and in the case of 9,
further identify the exact modified residue by ESI-MS/
MS sequencing.
region, where difference peaks near 270 and 310 nm were
observed (along with apparent non-specific absorbance
at 235 nm and below, Supplementary data). The HPLC
elution was monitored at 310 to most readily select for
the chromophore of interest (the lower wavelength band
was deemed to be too near the strong protein absorp-
tion), and showed a major peak eluting at 71 min (Sup-
plementary data). The corresponding mass spectrum
(Supplementary data) showed a singly charged ion at
m/z 919, which is consistent with the mass of the thermo-
lytic active site peptide Leu-Asn-X-Asp-Tyr, where X
corresponds to the 9-TPQ derivative 10c. The tandem
mass spectrum of the singly charged ion at m/z 919
(Fig. 4) shows b and y ions and other internal fragments
that localize to the L-N-X-D-Y sequence.
The negative ion mode MALDI-TOF spectra for the
range of m/z 700–900 of the thermolytic digest of native
BPAO and that following inactivation by 8 are shown in
Figure 3. There are two new prominent peaks that ap-
pear in the spectrum for the modified enzyme at m/z
891.95 and 729.05. These peaks match the calculated
mass for the negative ions of the cofactor-containing
active site peptides Leu-Asn-X-Asp-Tyr and Leu-Asn-
X-Asp, respectively, where X is the 3-(2-naphthyl)pyr-
rol-1-ylresorcinol derivative 11 of the TPQ cofactor.
There is an additional weaker new peak at m/z 876.89,
for which we have no concrete explanation at this time.
Figure 4. Collision-induced dissociation spectrum of the singly
charged ion with m/z 919 for the modified active site thermolytic
peptide eluting at 71 min (Supplementary data) of BPAO inactivated
by 3-(4-methoxy-3-nitrophenyl)-3-pyrroline (9).
O
O
OH
O-
H
O
N
O
N
N
N-
In our previous study focusing on spectroscopic charac-
terization of BPAO inactivated by 9,10 we focused in the
visible region where we could take advantage of reso-
nance Raman. In the current study, for LC–MS detec-
tion of modified peptides, we focused in the UV
OH
OH
Scheme 4. Proposed rate-limiting step in enzymatic metabolism of
3-(2-pyridyl)-3-pyrroline (5a).

Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
1873
3. Conclusions
4. Experimental
4.1. General procedures
We have shown in this study that 3-aryl-3-pyrrolines
containing an exceptionally electron-withdrawing
3-aryl group are some of the most potent mecha-
nism-based inactivators of BPAO reported to date.
Since the IC50 values for some of the analogs ap-
proach the enzyme concentration used in the incuba-
tion, it is possible that the true IC50 values would
be even lower with the use of lower [E]. The high
potency of these analogs likely reflects the enhanced
C–H acidity, through resonance, at the key rate-limit-
ing enzyme step where the active site base abstracts
the proton on the a-C of the substrate Schiff base
(see Scheme 4 for the example of 5a). Previous studies
on the substituent effect on metabolism of substituted
benzylamines by BPAO have found a similar enhance-
ment of substrate activity by acid-strengthening
groups.17
NMR spectra were obtained on Varian Gemini
200 MHz (13C NMR at 50 MHz) or 300 MHz (13C
NMR at 75 MHz) instruments, with chemical shifts ref-
erenced to the residual protonated solvent peak or sodi-
um 2,2-dimethyl-2-silapentane-5-sulfonate in the case of
D2O. Chemical shift ranges without indicated multiplic-
ity are given for resonances reflecting unresolved chem-
ical shift-inequivalent (usually diastereotopic) nuclei. In
the 13C NMR line listings, attached proton test (APT)
designations are given as (+) or (À) following the chem-
ical shifts. High-resolution mass spectra, electron impact
or fast atom bombardment (FAB) were obtained at 20–
40 eV on a Kratos MS-25A instrument. Bromopyridines
and bromopentafluorobenzene were purchased from
commercial sources. 2-Bromoquinoline, 1-bromoiso-
quinoline, and ethyl 3-oxopyrrolidine-1-carboxylate (1)
were prepared according to the indicated literature pro-
cedures. Doubly distilled water was used for all enzyme
experiments. Bovine plasma amine oxidase (BPAO)
(100 U/g of protein) for kinetic studies was purchased
from Sigma or Worthington. Purified BPAO for NBT
assay and mass spectroscopic studies was from our pre-
vious study.10 The purity of final inhibitors evaluated
(all amine salts) was assured to be >95% on the basis
of their displaying a single spot on TLC and the com-
plete absence of extraneous peaks in the 1H NMR
spectra.
It is curious that the 4-pyridyl analog 5c is a much less
efficient inhibitor than are the 2- and 3-pyridyl analogs
5a and 5b. Our interpretation of this finding is that the
pyrroline 3-aryl group resides in a pocket in the en-
zyme active-site that results in favorable interactions
of this group with the protein in the case of good sub-
strates. For the 2- and 3-pyridyl analogs, rotation of
the aryl ring about its single bond attachment to the
pyrroline ring provides the possibility for the ring
nitrogen to sample an arc of potential noncovalent
interactions (e.g., H-bonding) in the active site, or
alternatively, opportunities for avoidance of unfavor-
able noncovalent interactions. In contrast, the same
rotation in the case of the 4-pyridyl analog results in
no change in position of the ring nitrogen. Apparently,
this results in a lack of favorable interaction or in an
unfavorable interaction that cannot be avoided. Over-
all, it seems that further analog development in this
family of inhibitors should focus on other electron-
withdrawing substituents that lack a C2 axis of symme-
try. These compounds could be important leads to
develop pharmaceutically useful inhibitors of certain
copper amine oxidases, since the parent 3-phenyl com-
pound was shown to be a substrate without inhibitory
properties for the flavin-dependent mitochondrial
monoamine oxidase.10
4.2. Preparation of ethyl 3-hydroxy-3-arylpyrrolidine-1-
carboxylates (2a–e)
To a solution of n-butyllithium (2.5 M in hexanes,
1.8 mL, 4.5 mmol) in 25 mL of either anhydrous ether
(for 2a, 2b, 2d and 2e) or THF (for 2c) was added
4 mmol of either 2-bromopyridine, 3-bromopyridine,
pentafluorobromobenzene, or 2-bromoquinoline18 in
anhydrous ether (25 mL) or 4-bromopyridine in THF
(25 mL) under argon at À78 ꢁC. The reaction mixture
was stirred for 2 h at À78 ꢁC to ensure the total conver-
sion to the corresponding aryllithium. To these
solutions, ethyl 3-oxopyrrolidine-1-carboxylate19 (1)
(0.62 g, 4.0 mmol) in ether or THF (25 mL) was intro-
duced dropwise at À78 ꢁC, and the resulting mixture
was stirred for an additional 3 h and then warmed to
room temperature for 1 h. Saturated aqueous NH4Cl
solution (50 mL) was added to quench the reaction.
The organic layer was collected and the aqueous solu-
tion was extracted with ethyl acetate (4· 50 mL). The
combined organic solution was dried over anhydrous
Na2SO4. After evaporation, the residue was purified
by silica chromatography with hexanes–EtOAc as eluent
to give 2a–f.
Supported by model studies, we previously reported
strong evidence that irreversible inactivation of BPAO
by 3-aryl-3-pyrrolines reflects a pyrrolization of the
TPQ cofactor. Confirmatory evidence has now been
provided by mass spectrometric identification of the
modified active-site TPQ-containing peptide in the case
of two potent analogs in this series. Although there is
no guarantee that all analogs inactivate by the same
mechanism, we have no evidence to suggest otherwise.
Unfortunately, it is not possible to make a statement
about the stoichiometry of modification from the mass
spectrometric results because the native peptide is not
seen. However, our previous spectroscopic study on 9
provided data consistent with a stoichiometric derivati-
zation of the TPQ cofactor in the case of totally inacti-
vated enzyme.10
Compound 2a: eluted with EtOAc/hexanes 1:2 (77.6%
1
yield): H NMR (CD3OD) d 1.20–1.40 (br, 3H), 2.10
(m, 1H), 2.52 (m, 1H), 3.55–3.90 (4H), 4.05–4.25 (2H),
7.30 (m, 1H), 7.60–7.90 (2H), 8.53 (d, 1H, J = 4.4 Hz);
13C NMR (APT, CD3OD), there are two amide rota-
mers: d 15.19 (À), 39.52 and 40.27 (+), 46.28 and

1874
Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
46.48 (+), 59.77 and 60.09 (+), 62.50 (+), 81.84 and
82.57 (+), 121.59 (À), 123.82 (À), 138.49 (À), 149.69
(À), 157.24 (+), 163.14 and 163.23 (+).
Compound 3a (88.7% yield): 1H NMR (CD3OD) d 2.07
and 2.12 (2s, 3H total), 2.13 and 2.60 (2m, 2H total),
3.60–4.20 (4H), 7.31 (m, 1H), 7.75–7.90 (2H), 8.55 (m,
1H); 13C NMR (APT, CD3OD) d 21.32 and 21.40 (À),
39.08 and 40.41 (+), 46.13 and 47.91 (+), 59.69 and
61.19 (+), 81.48 and 82.77 (+), 121.64 and 121.67 (À),
123.91 and 123.99 (À), 138.57 and 138.62 (À), 149.74
(À), 163.07 (+), 172.37 and 176.03 (+).
Compound 3b (90.0% yield): 1H NMR (CD3OD) d 1.98
and 2.12 (2s, 3H total), 2.12–2.60 (2H), 3.50–4.00 (4H),
7.53 (m, 1H), 8.15 (m, 1H), 8.48 (m, 1H), 8.76 (m, 1H).
13C NMR (APT, CD3OD) d 21.86 and 22.35 (À), 39.16
and 40.53 (+), 45.93 and 47.65 (+), 59.76 and 61.53 (+),
78.58 and 79.90 (+), 125.09 (À), 135.70 (À), 140.70 and
140.93 (+), 147.66 and 147.71 (À), 149.15 and 149.20
(À), 172.20 and 172.43 (+).
Compound 2b: eluted with EtOAc/MeOH 20:1 (54.7%
yield): 1H NMR (CD3OD) d 1.27 and 1.29 (2t, 3H total,
J = 7.1 Hz for both), 2.15–2.50 (2H), 3.58–3.80 (4H),
4.14 and 4.16 (2q, 2H total, J = 7.1 Hz for both), 7.45
(dd, 1H, J = 4.9, 8.0 Hz), 8.00 (dm, 1H, J = 8.0 Hz),
8.47 (br, 1H), 8.75 (br, 1H); 13C NMR (APT, CD3OD),
there are two amide rotamers: d 15.21 (À), 39.65 and
40.44 (+), 46.07 and 46.27 (+), 60.06 and 60.27 (+),
62.59 (+), 78.94 and 79.68 (+), 125.07 (À), 135.68 (À),
141.05(+), 147.68 (À), 149.08(À), 157.16 and 157.19 (+).
Compound 2c: eluted with EtOAc/hexanes 3:1 (74.7%
yield): 1H NMR (CD3OD) d 1.27 and 1.30 (2t, 3H total,
J = 7.0 Hz for both), 2.09–2.47 (2H), 3.57–3.80 (4H),
4.14 and 4.16 (2q, 2H total, J = 7.0 Hz for both), 7.60
(d, 2H, J = 6.2 Hz), 8.54 (d, 2H, J = 6.2 Hz); 13C
NMR (APT, CD3OD), there are two amide rotamers:
d 15.16 (À), 39.91 and 40.71 (+), 46.16 and 46.38 (+),
60.17 and 60.40 (+), 62.59 (+), 79.61 and 80.34 (+),
122.38 (À), 150.22 (À), 141.05 (+), 155.17 and 155.28
(À), 157.11 and 157.16 (+).
Compound 3c (60.8% yield): 1H NMR (CD3OD) d 2.07
and 2.12 (2s, 3H total), 2.13 and 2.50 (2m, 2H total),
3.60–4.00 (4H), 7.71 (t, 2H, J = 6.6 Hz), 8.64 (dd, 2H,
J = 1.5, 1.8 Hz).
4.4. Preparation of N-acetyl-3-pyridyl-3-pyrrolines (4a–c)
Each compounds 3a–c (0.9 g, 3.8 mmol) was dissolved
in thionyl chloride (5 mL) and stirred at room tempera-
ture for 48 h. After the removal of SOCl2 in vacuo,
aqueous Na2CO3 (5%, 10 mL) was added, and the mix-
ture was extracted with EtOAc/MeOH 20:1 (4· 30 mL).
The combined organic solvent was evaporated under re-
duced pressure, and the residue was purified by silica gel
column chromatography to give 4a–c, NMR character-
ization of which indicated the presence of two amide
rotamers.
Compound 2d: eluted with a gradient of EtOAc/hexanes
(1:2 to 2:1) (77% yield): mp 87–88 ꢁC, there are two
amide rotamers:1H NMR (CDCl3) d 1.27 (t, 3H,
J = 6.9 Hz), 2.30 (m, 1H), 2.70 (m, 1H), 3.18 and 3.45
(2s, 1H total, OH), 3.60–3.68 (3H), 4.08–4.24 (3H);
13C d 28.5, 37.7 (m) and 38.1 (m), 43.4 and 43.7, 57.6–
58.1 (m), 62.0 and 62.1, 79.8, aromatic carbon signals
are complicated.
Compound 2e: eluted with a gradient of EtOAc/hexanes
(1:2 to 3:1) (36.0% yield): oil, there are two amide rota-
mers: 1H NMR (CDCl3) d 1.19–1.30 (3H), 2.14 (m, 1H),
2.40 (m, 1H), 3.66–3.88 (4H), 4.08–4.19 (2H), 5.98 and
6.00 (2s, 1H total, OH), 7.46–7.69 (2H), 7.71 (t, 1H,
J = 7.0 Hz), 7.81 (d, 1H, J = 8.0 Hz), 8.05 (d, 1H,
J = 8.3 Hz), 8.19 (d, 1H, J = 8.5 Hz); 13C NMR (CDCl3)
d 14.9 and 14.8, 46.7 and 39.8, 45.7 and 45.5, 59.7 and
59.4, 61.1, 80.6 and 79.8, 117.1, 126.9, 127.3, 127.6,
128.9, 130.2, 137.8, 146.0, 155.2, 160.19 and 160.16.
Compound 4a (30.0% yield): 1H NMR (CD3OD) d 2.12
and 2.16 (2s, 3H total), 4.40 (m, 1H), 4.60 (m, 2H), 4.80
(m, 1H), 6.66 (m, 1H), 7.31 (m, 1H), 7.65 (m, 1H), 7.82
(m, 1H), 8.60 (m, 1H); 13C NMR (APT, CD3OD) d
21.70 and 21.97 (À), 53.85 and 54.94 (+), 55.26 and
56.32 (+), 121.95 and 122.05 (À), 124.31 (À), 125.02
and 125.27 (À), 138.31 and 138.34 (À), 139.43 and
139.46 (+), 150.25 and 150.36 (À), 152.97 (+), 171.88
(+); HRMS (EI) calcd for C11H13N2O (MH+) m/z
189.1029. Found: 189.1027.
4.3. Preparation of N-acetyl-3-hydroxy-3-pyridylpyrroli-
dines (3a–c)
1
Compound 4b (32.3% yield): H NMR (CDCl3) d 2.05
and 2.12 (2s, 3H total), 4.38 (m, 2H), 4.55 (m, 2H),
6.15–6.30 (1H), 7.27 (m, 1H), 7.68 (m, 1H), 8.50 (m,
1H), 8.60 (m, 1H); 13C NMR (CDCl3) d 21.84 and
22.64 (À), 52.74 and 53.69 (+), 53.98 and 54.96 (+),
120.55 (À), 122.35 and 123.66 (À), 128.73 (+), 132.55
and 132.68 (À), 134.18 and 135.33 (+), 146.61 and
146.88 (À), 149.30 and 149.33 (À), 169.29 and 173.26
(+).
Each compound 2a–c (0.6 g, 2.55 mmol) was dissolved
in methanol (10 mL) and aqueous potassium hydroxide
(5.0 M, 10 mL), and the mixture was heated at reflux for
4 h. TLC indicated complete deprotection of 2. The
methanol was removed by rotary evaporation, and ace-
tonitrile (10 mL) was added. While keeping the reaction
solution pH between 9 and 10 by periodic addition of
2 N NaOH, acetic anhydride (1.5 mL) was added over
30 min, and the reaction mixture was stirred overnight
and then extracted with EtOAc/MeOH 10:1
(3· 50 mL). After removal of the organic solvent, the
residue was purified by column chromatography (eluent
EtOAc/MeOH 3:1) to give 3a–c. NMR characterization
of which indicated the presence of two amide rotamers.
1
Compound 4c (27.0% yield): H NMR (CDCl3) d 2.12
and 2.16 (2s, 3H total), 4.40–4.50 (2H), 4.55–4.70 (2H),
6.41 and 6.50 (2 app p, 1H total, J = 2.1 Hz for both),
7.15–7.40 (2H), 8.40–8.70 (2H); 13C NMR (CD3OD) d
21.84 and 22.16, 52.58 and 53.81, 55.00 and 54.94,
119.95 and 119.98, 125.08 and 125.21, 135.23 and

Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
1875
136.53, 139.87 and 139.93, 150.44 and 150.66, 169.09
and 169.22; HRMS (EI) Calcd for C11H13N2O (MH+)
m/z 189.1029. Found: 189.1027.
4.7. Preparation 3-(N-methylpyridyl)-3-pyrrolines (7b–c)
Compounds 6b–c were heated in a 1:1 mixture of 12 N
HCl and ethanol at reflux for 3 h, followed by complete
removal of solvent to give 7b–c iodide hydrochloride
salts.
4.5. Preparation of 3-pyridyl-3-pyrrolines (5a–c)
Each compounds 4a–c (0.06 g, 0.32 mmol) was dissolved
in a mixture of ethanol (5 mL) and aqueous HCl (12 N,
5 mL), which was heated at reflux for 3 h. Evaporation
to dryness afforded the pure dihydrochloride salts of
5a–c.
1
Compound 7b+IÆHCl (yield 71.9%): mp 196–197 ꢁC; H
NMR (D2O) d 4.41 (s, 3H), 4.43 (d, 2H, J = 2.2 Hz),
4.60 (d, 2H, J = 2.2 Hz), 6.78 (t, 1H, J = 2.2 Hz), 8.06
(dd, 1H, J = 6.2, 8.1 Hz), 8.58 (d, 1H, J = 8.1 Hz), 8.74
(d, 1H, J = 6.1 Hz), 8.97 (s, 1H); 13C NMR (APT,
D2O) d 63.45 (À), 66.32 (+), 68.28 (+), 141.64 (À),
142.82 (À), 145.43 (+), 146.88 (+), 156.78 (À), 157.60
(À), 159.19 (À); HRMS (FAB) calcd for C10H13N2
(M+) m/z 161.1080. Found: 161.1079.
4.5.1. 3-(2-Pyridyl)-3-pyrroline dihydrochloride (5aÆ2HCl).
1
(98.0% yield): mp 223–225 ꢁC; H NMR (D2O) d 4.47
(dd, 2H, J = 2.6, 3.6 Hz), 4.67 (dd, 2H, J = 2.5,
3.7 Hz), 7.05 (app p, 1H, J = 2.2 Hz), 7.96 (dd, 1H,
J = 1.1, 6.0 Hz), 8.03(d, 1H, J = 8.1 Hz), 8.57 (ddd,
1H, J = 1.6, 2.2, 8.0 Hz), 8.70 (dd, 1H, J = 1.6,
6.0 Hz); 13C NMR (APT, D2O) d 65.87 (+), 68.45
(+), 140.24 (À), 141.14 (À), 144.14 (+), 146.85 (À),
156.73 (À), 158.27 (+), 162.08 (À); HRMS (FAB)
Calcd for C9H11N2 (MH+) m/z 147.0922. Found:
147.0927.
1
Compound 7c+IÆHCl (yield 93.9%): mp 193–194 ꢁC; H
NMR (CD3OD) d 4.39 (s, 3H), 4.46 (d, 2H, J = 7.2 Hz),
4.62 (s, 2H), 7.21 (s, 1H), 8.18 (d, 1H, J = 6.0 Hz), 8.90
(d, 1H, J = 6.0 Hz); 13C NMR (CD3OD) d 148.2, 147.0,
134.5, 133.9, 125.6, 54.9, 52.5, 48.0; HRMS (EI) Calcd
for C10H13N2 (M+) m/z 161.1080. Found: 161.1079.
4.5.2. 3-(3-Pyridyl)-3-pyrroline dihydrochloride (5bÆ2HCl).
(68.7% yield): mp 252–254 ꢁC; H NMR (D2O) d 4.37
4.8. Preparation of 3-pentafluorophenyl-3-pyrroline
hydrochloride (5dÆHCl)
1
(dd, 2H, J = 2.4, 3.7 Hz), 4.57 (dd, 2H, J = 2.4,
3.6 Hz), 6.73 (app p, 1H, J = 2.2 Hz), 8.08 (dd, 1H,
J = 5.8, 7.0 Hz), 8.73 (m, 2H), 8.88(d, 1H, J = 1.2 Hz);
13C NMR (APT, D2O) d 66.09 (+), 68.11 (+), 141.18
(À), 142.29 (À), 145.56 (+), 146.33 (+), 153.42 (À),
155.21 (À), 158.48 (À); HRMS (FAB) Calcd for
C9H11N2 (MH+) m/z 147.0922. Found: 147.0923.
Compound 2d was subjected to SOCl2-mediated dehy-
dration according to the above procedure for conversion
of compounds 3–4. Without purification, N-(ethoxycar-
bonyl)-3-pyrroline 4d (0.5 mmol) was dissolved in HBr
(48%, 10 mL) and acetic acid (3 mL), and the mixture
was heated at reflux for 3 h. After removal of all the sol-
vent, the residual solid was coevaporated with HCl
(37.5%, 20 mL) to dryness three times by rotary evapo-
ration. The resulting residue was recrystallized from
EtOH to Et2O to yield 5d hydrochloride (85.0% yield):
mp 203–204 ꢁC; 1H NMR (CD3OD) d 4.35 (s, 2H),
4.52 (s, 2H), 6.56 (s, 1H); 13C (CD3OD) d 53.5, 53.8 (t,
J = 4.2 Hz), other carbon signals are complicated by
19F splitting; HRMS (FAB) calcd for C10H6F5N
(MH+) m/z 236.0499. Found: 236.0497.
4.5.3. 3-(4-Pyridyl)-3-pyrroline dihydrochloride (5cÆ2HCl).
1
(98.7% yield): mp 227–228 ꢁC; H NMR (free base in
CD3OD) d 4.03 (s, 2H), 4.17 (br, 2H), 6.7 (m, 1H),
7.47 (d, 2H, J = 3.1 Hz), 8.73 (d, 2H, J = 3.1 Hz); 13C
NMR (APT, CD3OD) d 53.51 (+), 55.18 (+), 122.04
(À), 129.21 (À), 138.79 (+), 142.97 (+), 150.46 (À),
150.53 (À); HRMS (FAB) calcd for C9H11N2 (MH+)
m/z 147.0922. Found: 147.0927.
4.6. Preparation of N-acetyl-3-(N-methylpyridyl)-3-pyrr-
olines (6b,c)
4.9. Preparation of 3-(2-quinolyl)-3-pyrroline dihydro-
chloride (5eÆ2HCl)
Methyl iodide (2 mL) was added to a solution of 4b or
4c (0.075 g, 0.40 mmol) in CH2Cl2 (5 mL), and the
reaction mixture was stirred for 36 h at room temper-
ature. The slight yellow solid was collected and washed
thoroughly with CH2Cl2 to give methiodides 6bc in
stoichiometric yield, NMR characterization of which
indicated the presence of two amide rotamers.
Compound 2e was subjected to SOCl2-mediated dehy-
dration according to the above procedure for conversion
of compound 3–4. After removal of excess SOCl2, neu-
tralization (Na2CO3), extraction (EtOAc), and silica
gel chromatographic purification (eluent EtOAc/hex-
anes 2:1) yielded N-(ethoxycarbonyl)-3-pyrroline 4e as
an oil in 40.0% yield, NMR characterization of which
indicated the presence of two amide rotamers: 1H
NMR (CD3OD) d 1.26–1.34 (m, 3H), 4.12–4.20 (m,
2H), 4.30 (s, 2H), 4.50–4.57 (m, 2H), 6.61 and 6.64
(2s, 1H total), 7.52 (t, 1H, J = 7.4 Hz), 7.68–7.72 (2H),
7.81 (d, 1H, J = 7.8 Hz), 7.96 (dd, 1H, J = 5.2, 8.3 Hz),
8.15 (d, 1 H, J = 8.7 Hz); 13C NMR (CD3OD) d 15.1,
54.1 and 54.4, 55.2 and 55.5, 62.7 and 62.8, many aryl
peaks. Compound 4e (0.5 mmol) was dissolved in HBr
(48%, 10 mL) and acetic acid (3 mL), and the mixture
was heated at reflux for 3 h. After removal of all the
1
Compound 6b: H NMR (CD3OD) d 2.12 and 2.21 (2s,
3H total), 4.46 (s, 3H), 4.48 (s, 2H), 4.65 (s, 2H), 6.90
(m, 1H), 8.09 (m, 1H), 8.69 (d, 1H, J = 8.6 Hz), 8.84
(d, 1H, J = 3.6 Hz), 9.16(d, 1H, J = 8.0 Hz).
1
Compound 6c: H NMR (DMSO) d 2.02 and 2.06 (2s,
3H total), 4.29 (s, 3H), 4.37 (m, 1H), 4.54 (m, 1H),
4.59 (m, 1H), 4.75 (m, 1H), 7.30 (m, 1H), 8.16–8.20
(2H), 8.92–8.98 (2H).

1876
Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
solvent, the residual solid was coevaporated with HCl
(37.5%, 20 mL) to dryness three times by rotary evapo-
ration. The resulting residue was recrystallized from
EtOH to Et2O to yield 5e dihydrochloride (82.0%): mp
280 ꢁC (dec); 1H NMR (CD3OD) d 4.55 (d, 2H,
J = 2.3 Hz), 4.85 (d, 2H, J = 2.0 Hz), 7.47 (m, 1H),
7.93 (t, 1H, J = 7.5 Hz), 8.12 (dt, 1H, J = 8.4, 1.3 Hz),
8.22 (d, 1H, J = 8.7 Hz), 8.28 (d, 1H, J = 8.4 Hz), 8.46
(d, 1H, J = 8.4 Hz), 9.06 (d, 1H, J = 8.7 Hz); 13C
NMR (CD3OD) d 52.6, 54.9, 121.7, 122.9, 129.7,
130.2, 131.1, 132.9, 136.1, 136.4, 140.8, 147.1, 148.4;
HRMS (FAB) Calcd for C13H12N2 (MH+), 197.1079.
Found: 197.1082.
The samples were then denatured by the addition of sol-
id urea and a 0.5 M aqueous solution of dithiothreitol
(DTT) to reach final concentrations of 6 M and
10 mM, respectively, incubating at 37 ꢁC for 1 h. Iodoa-
cetamide (0.5 M aqueous stock) was then added to a
final concentration of 40 mM, with incubation at room
temperature in the dark for 1 h. Excess DTT solution
was then added (to reach 50 mM) to consume unreacted
iodoacetamide. The sample was then dialyzed against
10 mM NH4HCO3 in a cold room with three changes
in the buffer over 24 h. After dialysis, the protein was
digested by thermolysin at a ratio of BPAO:thermolysin
of 50:1 (w/w) for 24 h at 37 ꢁC. The same amount of
thermolysin was added a second time (bringing the final
ratio to 25:1), with an additional 24-h incubation at
37 ꢁC. The final sample was concentrated to a volume
of 50 lL.
4.10. Preparation of 3-(1-isoquinolyl)-3-pyrroline dihy-
drochloride (5fÆ2HCl)
The lithium reagent from 1-bromoisoquinoline20 was
prepared in diethyl ether at À78 ꢁC and reacted with
ethyl 3-oxopyrrolidine-1-carboxylate as described for
2a–e above. After workup, the resulting N-(ethoxycar-
bonyl)-3-hydroxypyrrolidine 2f was subjected without
purification to SOCl2-mediated dehydration. After
removal of excess SOCl2, neutralization (Na2CO3),
extraction (EtOAc), and silica gel chromatographic
purification (eluent EtOAc/hexanes 2:1) yielded N-(eth-
oxycarbonyl)-3-pyrroline 4f as an oil in 19% overall
yield, NMR characterization of which indicated the
4.13. HPLC–ESI-MS/MS analysis for modification by 9
Aliquots (20 lL) of peptides were separated using a
Michrom Magic 2002 HPLC (Michrom BioResources,
Auburn, CA) equipped with a Varian Vydac C18
1· 250 mm column. The eluate was monitored at 235
and 310 nm, with a binary gradient elution system of
solvent mixtures A (2% aqueous acetonitrile containing
0.05% TFA) and B (90% aqueous acetonitrile containing
0.05% TFA) according to the following 2-h program:
100% A for 8 min, 98% A for 7 min, 95% A for
45 min, 84% A for 20 min, 65% A for 10 min, 0% A
(100% B) for 18 min, 100% A for 12 min, with a flow
rate of 100 lL/min for the first 95 min and 150 lL/min
for the last 25 min. The eluate from the column was
transferred to a Finnigan LCQDuo Mass Spectrometer
(Thermo Electron, San Jose, CA); the ESI was pro-
grammed with the following settings: nitrogen sheath
gas flow rate (arbitrary units) at 50; auxiliary gas flow
rate at 0; capillary temperature at 160 ꢁC; spray voltage
at 4.4 kV. The mass spectrometer was programmed for
selected reaction monitoring and MS/MS, and the nor-
malized collision energy was set to 40%, the tube lens
offset at 30 V. Other parameters were adjusted using a
tune method developed using model compound 10a.
1
presence of two amide rotamers: H NMR (CDCl3)d
1.27–1.36 (m, 3H), 4.18–4.26 (m, 2H), 4.48–4.59 (m,
2H), 4.79–4.85 (m, 2H), 6.20–6.40 (m, 1H), 7.57–7.86
(m, 4H), 8.31–8.60 (m, 2H); 13C NMR (CDCl3) d 14.9
and 15.0, 54.4 and 54.9, 55.2 and 55.8, 61.20 and
61.24, many aryl peaks. Compound 4f (0.5 mmol) was
dissolved in HBr (48%, 10 mL) and acetic acid (3 mL),
and refluxed for 3 h. After removal of all the solvent,
the residue solid was coevaporated with HCl (37.5%,
3· 20 mL) to dryness by rotary evaporator. The result-
ing residue was recrystallized from EtOH to Et2O to
1
yield 5f dihydrochloride (87.0%): mp 208 ꢁC (dec); H
NMR (CD3OD) d 4.61 (s, 2H), 4.83 (s, 2H), 6.92 (s,
1H), 8.12 (t, 1H, J = 7.4 Hz), 8.28 (t, 1H, J = 8.1 Hz),
8.39 (d, 1H, J = 8.1 Hz), 8.52–8.68 (m, 3H); 13C NMR
(CD3OD) d 54.9, 55.0, 127.1, 127.9, 129.6, 129.8,
132.4, 133.0, 137.5, 138.0, 141.0, 151.0, 158.6; HRMS
(FAB) Calcd for C13H12N2 (MH+), 197.1097. Found:
197.1080.
4.14. MALDI-TOF analysis for modification by 8
The protein digest samples were desalted and eluted
using C18 Millipore Ziptips (Millipore, Bedford, MA)
according to the manufacturer’s recommendations.
Briefly, the Ziptip was wet by 50% acetonitrile in dis-
tilled–deionized water and thereafter equilibrated by
0.1% trifluoroacetic acid (TFA). Then the peptides were
extracted with the Ziptip and washed with 0.1% TFA.
Finally the sample was eluted with 2 lL of a solution
of a-cyano-4-hydroxycinnamic acid (20 mg/mL in 70%
acetonitrile and 0.1% TFA) directly onto the stainless
steel target and allowed to dry at room temperature.
The mass spectrometer used was a Bruker BiFlex III
MALDI-TOF instrument (Bruker Daltonics, Bremen,
Germany) equipped with a pulsed nitrogen laser (3 ns
pulse at 337 nm). Spectra were collected in the negative
ion and reflectron mode with an average of 250 laser
shots.
4.11. Enzymology
Time-dependent inactivations of BPAO by candidate
inhibitors, determination of active enzyme concentra-
tion in the various preparations, determinations of irre-
versibility, and the NBT assays for cofactor redox
cycling activity were conducted as previously de-
scribed,15 except that the pH 7.2 sodium phosphate buff-
er concentration was 50 mM rather than 100 mM.
4.12. Enzyme modification for mass spectrometric studies
Purified BPAO10 (2 mg/mL) in 0.1 M potassium phos-
phate buffer, pH 7.4, was incubated with 8 or 9
(1 mM) in a total volume of 0.5 mL at 30 ꢁC for 2 h.

Y. Zhang et al. / Bioorg. Med. Chem. 15 (2007) 1868–1877
1877
6. O’Sullivan, J.; Unzeta, M.; Healy, J.; O’Sullivan, M. I.;
Davey, G.; Tipton, K. F. Neurotoxicology 2004, 25, 303–
315.
7. Boomsma, F.; Hut, H.; Bagghoe, U.; van der Houwen, A.;
van den Meiracker, A. Med. Sci. Monit. 2005, 11, RA122–
RA126.
8. Toninello, A.; Pietrangeli, P.; De Marchi, U.; Salvi, M.;
Mondovi, B. Biochim. Biophys. Acta 2006, 1765,
1–13.
9. Lee, Y.; Huang, H.; Sayre, L. M. J. Am. Chem. Soc. 1996,
118, 7241–7242.
Acknowledgments
We thank Doreen Brown in the laboratory of Professor
David M. Dooley, Montana State University, for the
provision of purified BPAO. We also thank Dr. David
Sell for guidance with the LC–MS work conducted in
the laboratory of Professor Vincent Monnier, and Pro-
fessor Shu Chen for guidance with the MALDI-TOF
work, all in the Department of Pathology, Case Western
Reserve University. Financial support for this work was
provided by NIH Grant GM 48812.
10. Lee, Y.; Ling, K.; Lu, X.; Silverman, R. B.; Shepard, E.
M.; Dooley, D. M.; Sayre, L. M. J. Am. Chem. Soc. 2002,
124, 12135–12138.
11. Wang, X.; Rabbat, P.; O’Shea, P.; Tillyer, R.; Grabowski,
E. J. J.; Reide, P. J. Tetrahedron Lett. 2000, 41, 4335–
4339.
12. Ganellin, C. G.; Jayes, D.; Khalat, Y. S. Collect Czech.
Chem. Commun. 1991, 56, 2448.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmc.
2006.11.025.
13. Kitz, R.; Wilson, I. B. J. Biol. Chem. 1962, 237, 3245–
3249.
14. Paz, M. A.; Fluckiger, R.; Boak, A.; Kagan, H. M.;
Gallop, P. M. J. Biol. Chem. 1991, 266, 689–692.
15. Jeon, H.-B.; Sun, G.; Sayre, L. M. Biochim. Biophys. Acta
2003, 1647, 343–354.
16. Lunelli, M.; Di Paolo, M. L.; Biadene, M.; Calderone, V.;
Battistutta, R.; Scarpa, M.; Rigo, A.; Zanotti, G. J. Mol.
Biol. 2005, 346, 991–1004.
References and notes
1. Janes, S. M.; Mu, D.; Wemmer, D.; Smith, A. J.; Kaur, S.;
Maltby, D.; Burlingame, A. L.; Klinman, J. P. Science
1990, 248, 981–987.
2. Janes, S. M.; Palcic, M. M.; Scaman, C. H.; Smith, A. J.;
Brown, D. E.; Dooley, D. M.; Mure, M.; Klinman, J. P.
Biochemistry 1992, 31, 12147–12154.
17. Hartmann, C.; Klinman, J. P. Biochemistry 1991, 30,
4605–4611.
18. Young, T. E.; Amstutz, E. D. J. Am. Chem. Soc. 1951, 73,
4773–4775.
3. Longu, S.; Mura, A.; Padiglia, A.; Medda, R.; Floris, G.
Phytochemistry 2005, 66, 1751–1758.
19. Wu, Y.-H.; Gould, W. A.; Lobeck, W. G.; Roth, H. R.;
Feldkamp, R. F. J. Med. Pharm. Chem. 1962, 5, 752–
762.
20. Wibaut, J. P.; Haaijman, P. W. Rec. Trav. Chim. 1943, 62,
466–468.
4. Cona, A.; Rea, G.; Angelini, R.; Federico, R.; Tavlado-
raki, P. Trends Plant Sci. 2006, 11, 80–88.
5. Matyus, P.; Dajka-Halasz, B.; Foldi, A.; Haider, N.;
Barlocco, D.; Magyar, K. Curr. Med. Chem. 2004, 11,
1285–1298.