Inhibition of monoamine oxidase B by selected benzimidazole and caffeine analogues

Article abstract of DOI:10.1016/j.bmc.2007.03.046

We have recently reported that a series of (E)-8-styrylcaffeines and (E)-2-styrylbenzimidazoles are moderate to very potent competitive inhibitors of monoamine oxidase B (MAO-B). The most potent member of the series was found to be (E)-8-(3-chlorostyryl)c

Full text of DOI:10.1016/j.bmc.2007.03.046

Bioorganic & Medicinal Chemistry 15 (2007) 3692–3702
Inhibition of monoamine oxidase B by selected benzimidazole
and caffeine analogues
a,
Deidre van den Berg, Kevin R. Zoellner,a, Modupe O. Ogunrombi,a Sarel F. Malan,a
´
Gisella Terre’Blanche,a Neal Castagnoli, Jr.,b Jacobus J. Bergha and Jacobus P. Petzera,*
aPharmaceutical Chemistry, School of Pharmacy, North-West University, Private Bag X6001, Potchefstroom 2520, South Africa
bDepartment of Chemistry, Virginia Tech, Blacksburg, VA 24061, USA
Received 26 January 2007; revised 11 March 2007; accepted 14 March 2007
Available online 18 March 2007
Abstract—We have recently reported that a series of (E)-8-styrylcaffeines and (E)-2-styrylbenzimidazoles are moderate to very
potent competitive inhibitors of monoamine oxidase B (MAO-B). The most potent member of the series was found to be (E)-8-
(3-chlorostyryl)caffeine (CSC) with an enzyme-inhibitor dissociation constant (Ki value) of 128 nM. In the present study, we have
prepared additional caffeine and benzimidazole analogues in an attempt to identify compounds with improved MAO-B inhibition
potency while still acting reversibly. The most potent inhibitor among the caffeine analogues was (E)-8-(3,4-dichlorostyryl)caffeine
with a Ki value of 36 nM, approximately 3.5 times more potent than CSC. The most potent inhibitor among the benzimidazole ana-
logues was (E)-2-(4-trifluoromethylstyryl)-1-methylbenzimidazole with a Ki value of 430 nM. An SAR analysis indicated that the
potency of MAO-B inhibition by (E)-2-styryl-1-methylbenzimidazole analogues depended upon the Taft steric parameter (Es) of
the substituents attached to C-4 of the styryl phenyl ring. Substituents with a large degree of steric hindrance appear to enhance
inhibition potency. The proposal that potent MAO-B inhibition by (E)-8-styrylcaffeines and (E)-2-styrylbenzimidazoles can be
explained by a mode of binding that involves traversing both the entrance and substrate cavities was supported by the finding that
1-methylbenzimidazole only weakly inhibited MAO-B with a Ki value of 2084 lM. Without the styryl side chain, 1-methylbenzim-
idazole is not expected to be able to bind simultaneously to both the entrance and substrate cavities.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
Monoamine oxidase (MAO) is a flavin adenine dinucle-
otide (FAD)-containing enzyme attached to the mito-
chondrial outer membrane of neuronal, glial, and
N
H3C
H
other cells. Its roles include regulation of the levels of
biogenic and xenobiotic amines in the brain and the
peripheral tissues by catalyzing their oxidative deamina-
tion.1 On the basis of their substrate and inhibitor spec-
ificities, two types of MAO (A and B) have been
described. MAO-A preferentially catalyzes the deamina-
tion of serotonin and norepinephrine and is irreversibly
inhibited by low concentrations of clorgyline. MAO-B
preferentially catalyzes the deamination of arylalkylam-
ines such as benzylamine and is irreversibly inhibited by
(R)-deprenyl (1) (Scheme 1).2 Both isoforms utilize
N
(R
)-Deprenyl (1)
Rasagiline (2)
H
O
Cl
N
N
(
(
E
)
)
N
O
N
CSC (3)
N
E
N
6a
Keywords: Monoamine oxidase B; Reversible inhibitors; Structure–
activity relationship; Caffeine; Benzimidazole.
*
Scheme 1. The structures of the mechanism-based MAO-B inactiva-
tors (R)-deprenyl (1) and Rasagiline (2) as well as the reversible
inhibitors CSC (3) and (E)-2-styryl-1-methylbenzimidazole (6a).
Corresponding author. Tel.: +27 18 2992206; fax: +27 18 2994243;
e-mail: jacques.petzer@nwu.ac.za
These authors contributed equally to this work.
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.03.046

D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
3693
dopamine as substrate.1 Due to their role in the metab-
olism of catecholamine neurotransmitters, MAO-A and
MAO-B have long been of considerable pharmacologi-
cal interest, and reversible and irreversible inhibitors of
MAO-A and MAO-B have been used clinically to treat
neurological disorders including depression and Parkin-
son’s disease (PD).1 MAO-B has also been implicated in
neurodegenerative processes resulting from exposure to
xenobiotic amines. For example, the first step of the bio-
activation of the parkinsonian inducing pro-neurotoxin,
1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP),
is catalyzed by MAO-B.3
double bond and 1,3,7-trimethyl substitution of the xan-
thine ring. The literature supports the proposal that a
wide variety of planar, heterocyclic compounds fre-
quently act as competitive inhibitors of MAO-B.
Accordingly, a small series of (E)-2-styrylbenzimidaz-
oles was shown to be moderate competitive inhibitors
of MAO-B.18 In the present study, we have examined
additional caffeine (4a–f and 5a–b) and benzimidazole
(6a–g and 7a–b) analogues (Tables 1–4) in an attempt
Table 1. The Ki values for the inhibition of MAO-B by
(E)-8-styrylcaffeine analogues (4a–f)a
MAO-A and MAO-B are therefore important targets
for the development of new drugs. We are particularly
interested in the therapeutic role of MAO-B inhibitors
in Parkinson’s disease. Since the MAO-B isoform
appears to be predominantly responsible for dopamine
metabolism in the basal ganglia,4,5 inhibition of this
enzyme in the brain may conserve the depleted supply
of dopamine and inhibitors are used in combination
with levodopa as dopamine replacement therapy in
patients diagnosed with early PD.6 MAO-B inhibitors
have been shown to elevate dopamine levels in the
striatum of primates treated with levodopa.7 Further-
more, in the catalytic cycle of MAO, one mole of dopal-
dehyde and H2O2 are produced for each mole of
dopamine oxidized. Both these catabolic products may
be neurotoxic if not rapidly inactivated by centrally
located aldehyde dehydrogenase (ADH)8 and glutathi-
one peroxidase,9, respectively. Thus, inhibitors of
MAO-B may also exert a neuroprotective effect by
stoichiometrically decreasing aldehyde and H2O2
production in the brain. Inhibitors that have been dem-
onstrated to be of clinical value include the mechanism-
based inactivators, (R)-deprenyl10 and Rasagiline (2)11
(Scheme 1), and reversible inhibitors such as lazabe-
mide12 and safinamide.13 From a safety point of view,
reversible inhibitors may be therapeutically more
desirable than inactivators since MAO-B activity can be
regained relatively quickly following withdrawal of the
reversible inhibitor. In contrast, return of enzyme activity
following treatment with inactivators requires de novo
synthesis of the MAO-B protein which may require
several weeks.14 For this reason, several studies are
currently underway to develop reversible inhibitors of
MAO-B.15,16 These inhibitors act in a competitive manner
while retaining selectivity toward MAO-B.
O
N
N
Ar
N
O
N
Compound
Ar
Ki (lM)
Br
4a
0.083
Cl
4b
4c
0.036
0.239
Cl
CF3
CF3
4d
4e
4f
1.58
11.3
39.7
S
S
O
a The enzyme source used was baboon liver mitochondrial MAO-B.
Table 2. The Ki values for the inhibition of MAO-B by the purino[7,8-c]-
quinazoline-8,10(9H,11H)-dione analogues (5a–b)
O
N
We have recently reported that (E)-8-styrylcaffeines act
as moderate to very potent competitive inhibitors of
MAO-B.17–19 In contrast caffeine only weakly inhibited
the enzyme, which indicates that substitution at C-8 en-
hances affinity of caffeine analogues for the active site of
MAO-B. Substitution at C-8 with an electron deficient
styryl functional group produced structures that were
especially potent inhibitors. For example, the most po-
tent member of the series was found to be (E)-8-(3-chlo-
rostyryl)caffeine (CSC) (3) (Scheme 1) with an enzyme-
inhibitor dissociation constant (Ki value) of 128 nM.20
A structure–activity relationship (SAR) study indicated
that the structural features important for MAO-B inhi-
bition are the trans configuration about the styryl
R
N
N
N
O
N
R
Compound
R
Ki valuea (lM)
5a
5b
CH3
C2H5
39.9
No inhibitionb
a The enzyme source used was baboon liver mitochondrial MAO-B.
b No inhibition was observed at a maximum tested concentration of
50 lM. Due to limited solubility in the aqueous incubation solvent
higher concentrations were not tested.

3694
D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
according to a previously reported procedure.20,21 The
key starting material for the procedure, 1,3-dimethyl-
5,6-diaminouracil (8a), was allowed to react with the
appropriate carboxylic acid (9) in the presence of a
carbodiimide activating reagent (1-ethyl-2-[3-(dimeth-
ylamino)propyl]-carbodiimide, EDAC) (Scheme 2).
The resulting amide intermediates underwent ring
closure in refluxing aqueous sodium hydroxide to yield
the corresponding 1,3-dimethyl-8-substituted-7H-xan-
thinyl analogues (10). Without further purification, the
crude product obtained was selectively 7N-methylated
in the presence of an excess of iodomethane and potas-
sium carbonate to yield the target compounds 4a–f.
Following crystallization from a suitable solvent the
structures and purity of all new compounds were
verified by mass spectrometry, 1H NMR, and 13C
NMR. The trans geometry about the styryl p-bonds of
4a–c was confirmed by proton–proton coupling constants
which were in the range of 15.7–15.8 Hz for the olefinic
proton signals. The trans geometry about the ethenyl
p-bonds of 4d–f was confirmed by proton–proton
coupling constants in the range of 15.4–15.6 Hz.
Table 3. The Ki values for the inhibition of MAO-B by (E)-2-styryl-1-
methylbenzimidazole analogues (6a–g)
N
N
R
(E)
Compound
R
Ki valuea (lM)
Esb
pb
0.00
6a
6b
6c
6d
6e
6f
H
Cl
18.7, 17c
2.83
2.54
0.00
ꢀ0.97
ꢀ1.16
ꢀ0.46
ꢀ2.40
ꢀ1.24
ꢀ0.55
0.71
0.86
Br
F
14.95
0.43
3.26
0.14
0.88
CF3
CH3
OCH3
0.56
ꢀ0.02
6g
5.82
The values of the selected physiochemical parameters used in the SAR
study are also listed.27
a The enzyme source used was baboon liver mitochondrial MAO-B.
b Values obtained from Ref. 27.
c Reported Ki value.18
Table 4. The Ki values for the inhibition of baboon liver MAO-B by
methylbenzimidazoles (7a–b)
The purino[7,8-c]quinazoline-8,10(9H,11H)-dione ana-
logues (5a–b) (Table 2) examined in this study were
prepared according to a synthetic pathway reported by
Ceccarelli et al. (Scheme 3).22 The reaction of 1,3-
dimethyl- or 1,3-diethyl-5,6-diaminouracil (8) with
2-nitrobenzaldehyde in the presence of acid gave 6-ami-
no-1,3-dimethyl- or 6-amino-1,3-diethyl-5-(2-nitroben-
zylideneamino)uracil (11) in high yields (83–89%).
Oxidation of these intermediate imines with iodine in
1,2-dimethoxyethane yielded the crystalline 8-(2-nitro-
phenyl)-1,3-dimethyl-7H-xanthines (12). Following
reduction of the nitro group with hydrochloric acid
and tin powder, the resulting amines (13) were treated
with triethyl orthoformate to yield the target com-
pounds (5a–b). The structures and purity of the target
R1
N
R2
N
Compound
R1
R2
Ki value (lM)
7a
7b
CH3
H
H
CH3
2084
1220
to identify compounds with improved MAO-B inhibi-
tion potency. Among the compounds studied was
(E)-8-(3-bromostyryl)caffeine (4a). Applying a multivar-
iate predictive equation constructed in a previous
study,20 this putative inhibitor is predicted to have a
Ki value for the inhibition of MAO-B of 106 nM. Since
electron withdrawing substituents of the styryl ring
appear to enhance MAO-B inhibition potency,20 we
have included in this study additional (E)-8-styrylcaffei-
nyl derivatives (4b–f) with electron deficient styryl rings.
Also studied were seven (E)-2-styryl-1-methylbenzimi-
dazole analogues (6a–g) differing only in substitution
at C-4 of the styryl phenyl ring. The principal aim with
these analogues was to compare the efficacy of inhibi-
tion and possibly the binding mode of the benzimidaz-
oles with that of the caffeine analogues. It was
previously established via a Hansch-type SAR analysis
that potency of MAO-B inhibition by (E)-8-styrylcaffei-
nyl analogues depended upon the descriptors of bulki-
ness (Vw) and lipophilicity (p) of the substituents
attached to C-4 of the styryl phenyl ring.20
O
NH2
Ar
N
+
HO2C
(E)
O
N
NH2
9
8a
O
H
N
i, ii
Ar
N
N
(E)
N
O
O
N
O
10
N
iii
Ar
(E)
2. Results
N
N
4a-f
2.1. Chemistry
Scheme 2. Synthetic pathway to (E)-8-styrylcaffeinyl analogues (4a–f).
Reagents and condition: (i) EDAC, dioxane/H2O; (ii) NaOH (aq),
reflux; (iii) CH3I, K2CO3, DMF.
The (E)-8-styrylcaffeinyl analogues (4a–f) (Table 1)
examined in this study were prepared in high yield

D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
3695
O
O
O2N
NO2
R
O
NH2
NH2
R
O
N
O
H
N
i
N
+
ii
N
R
N
R
NH2
11
8a: R = CH3
8b: R = C2H5
O
O
O
N
H
H
O2N
H2N
R
O
R
O
R
O
N
N
N
N
N
iii
N
N
iv
N
N
N
R
N
R
N
R
13
12
5a: R = CH3
5b: R = C2H5
Scheme 3. Synthetic pathway to purino[7,8-c]quinazoline-8,10(9H,11H)-dione analogues (5a–b). Reagents and conditions: (i) CH3CO2H; (ii) I2,
dimethoxyethane, 50 ꢁC; (iii) Sn, HCl; (iv) HC(OC2H5)3, reflux.
compounds and intermediates were verified by their
mass spectrometric, 1H NMR, and 13C NMR
properties.
NH2
NH2
R
R
+
HO2C
(E)
16
15
A series of (E)-2-styryl-1-methylbenzimidazole ana-
logues (6a–g) (Table 3) also was prepared in an attempt
to compare their efficacy of inhibition and possibly the
binding mode to the active site with those of the caffeine
analogues. Compounds 6a–g were prepared by treating
the corresponding (E)-2-styryl-1H-benzimidazole ana-
logue (14) with one equivalent of iodomethane in the
presence of potassium carbonate. (E)-2-styryl-1H-benz-
imidazole analogues have previously been prepared by
reacting 2-methyl-1H-benzimidazole (7b) with an appro-
priately substituted benzaldehyde at high tempera-
tures.18,23–25 In this study, we have explored a different
synthetic pathway to species 14 (Scheme 4) involving
condensation of o-phenylenediamine (15) with an
appropriately substituted cinnamic acid derivative (16)
in the presence of the carbodiimide activating reagent
(EDAC). 4-(Dimethylamino)pyridine and imidazole
were used as catalysts. The resulting intermediate amide
(17) was cyclized by heating under reflux in the presence
of acid to yield the target (E)-2-styryl-1H-benzimidazole
analogue (14) as the corresponding hydrochloric acid salt.
Attempts to cyclize the intermediate amide under basic
conditions resulted in hydrolysis of the amide. With the
exception of 6e, the synthesis of all the (E)-2-styryl-
1-methylbenzimidazole analogues examined here has
previously been reported. The trans geometry about the
styryl double bonds of 6a–g was confirmed by proton–
proton coupling constants which were in the range of
15.8–15.9 Hz for the olefinic proton signals.
H
N
i
(
E)
O
NH2
17
H
N
R
ii
(E)
N
14
N
N
R
iii
(E)
6a-g
Scheme 4. Synthetic pathway to (E)-2-styryl-1-methylbenzimidazole
analogues (6a–g). Reagents and condition: (i) EDAC, DMF; (ii) HCl,
ethanol, reflux; (iii) CH3I, K2CO3, DMF.
laboratory is based on the MAO-B catalyzed oxidation
of 1-methyl-4-(1-methylpyrrol-2-yl)-1,2,3,6-tetrahydro-
pyridine (MMTP) to the corresponding dihydropyridi-
nium metabolite (MMDP+).26 Since MMDP+ absorbs
light maximally at a wavelength of 420 nm, the enzy-
matic production of MMDP+ may be readily measured
spectrophotometrically. At this wavelength neither the
substrate nor the test compounds absorb light. Because
of these favorable chromophoric characteristics and the
in vitro chemical stability of MMDP+, this assay is
frequently used to evaluate the potencies of potential
inhibitors of MAO-B.18,20 As enzyme source we have
employed the mitochondrial fraction obtained from
baboon liver tissue since it is reported to be devoid of
2.2. Enzymology and inhibition studies
In the present study, we have examined the possibility
that the synthetic caffeine (4a–f, 5a–b) and benzimid-
azole (6a–g, 7a–b) analogues may act as inhibitors of
MAO-B. The measurement of MAO-B activity in our

3696
D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
MAO-A activity while exhibiting a high degree of
MAO-B catalytic activity.26 Therefore, even though
MMTP is a MAO-A/B mixed substrate, its oxidation
by baboon liver mitochondria can be exclusively attrib-
uted to the action of the MAO-B isoform. Also, the
interaction of reversible inhibitors with MAO-B
obtained from baboon liver tissue appears to be similar
to the interaction with the human form of the enzyme
since inhibitors such as CSC are approximately equipo-
tent with both enzyme sources.18
Waals volume (Vw) for a bromine substituted in the
meta position of a phenyl ring are 0.39 and 1.32, respec-
tively.27,28 Substitution of these values into the multivar-
iate equation (logKi = ꢀ2.10rm ꢀ 0.49Vw + 0.49)20
yielded a predicted Ki value of 106 nM. In spite of the
few data points that were used to derive the equation,
the relatively small difference between the experimen-
tally obtained Ki value of 86 nM and the predicted value
reflected the expected trend. Another inhibitor identified
in this study as being exceptionally potent is (E)-8-(3,5-
ditrifluoromethylstyryl)caffeine (4c). This compound
was found to inhibit MAO-B with a Ki value of
239 nM. (E)-8-(2-thienylethenyl)caffeine (4d), (E)-8-(3-
thienylethenyl)caffeine (4e), and (E)-8-(2-furylethe-
nyl)caffeine (4f) were found to be only moderate
inhibitors of MAO-B with Ki values in the low micromo-
lar range. Although only moderately potent, 4d was still
a better inhibitor than (E)-8-styrylcaffeine, which is
reported to inhibit baboon liver MAO-B with a Ki of
2.7 lM.20 The purino[7,8-c]quinazoline-8,10(9H,11H)-
dione analogue 5a was found to be a weak inhibitor of
MAO-B, while analogue 5b exhibited no inhibition at
concentrations up to 50 lM, its limit of solubility in
the aqueous incubation mixtures (Table 2).
With the exception of 5b all of the caffeine (4a–f, 5a) and
benzimidazole (6a–g, 7a–b) analogues tested were found
to be inhibitors of MAO-B. As demonstrated by exam-
ple with (E)-8-(3,5-ditrifluoromethylstyryl)caffeine (4c)
(Fig. 1), the lines of the Lineweaver–Burk plots inter-
sected at the y-axis, indicating the mode of inhibition
to be competitive. The enzyme-inhibitor dissociation
constants (Ki values) for the inhibition of MAO-B by
the test compounds are presented in Tables 1–4. The
most potent inhibitor among the caffeine analogues
was found to be (E)-8-(3,4-dichlorostyryl)caffeine (4b)
with a Ki value of 36 nM (Table 1), approximately 3.5
times more potent than CSC which has a reported Ki va-
lue for the inhibition of baboon liver MAO-B of
128 nM.20 The second most potent inhibitor evaluated
in this study was (E)-8-(3-bromostyryl)caffeine (4a), with
a Ki value of 83 nM, also more potent than the lead
compound CSC.
The most potent inhibitor among the (E)-2-styryl-1-
methylbenzimidazole analogues (6a–g) was (E)-2-(4-tri-
fluoromethylstyryl)-1-methylbenzimidazole (6e) with a
Ki value of 430 nM (Table 3). The least potent of the ser-
ies was the unsubstituted (E)-2-styryl-1-methylbenzimi-
dazole (6a) with a Ki value of 18.7 lM. These results
support evidence that substitution at the styryl phenyl
ring generally leads to enhanced MAO-B inhibition
activity of both (E)-8-styrylcaffeinyl and (E)-2-styryl-
benzimidazolyl analogues. Interestingly, the electronic
characteristics of the styryl substituents do not appear
to correlate with inhibition activity since the analogue
substituted with an electron withdrawing fluorine (6d)
was found to be a weaker inhibitor than analogues bear-
ing electron donating substituent (6f–g). On the other
hand, the analogue bearing a trifluoromethyl substituent
(6e) proved to be the most potent inhibitor in this series.
This result is in agreement with a previous study which
showed that MAO-B inhibition potency of (E)-8-styryl-
caffeine analogues correlated with the size and lipophil-
icity of the substituent at C-4 of the styryl ring, while the
contribution of electronic substituent descriptors was
negligible.20
Applying a multivariate predictive equation constructed
in a previous study,20 we were able to predict the Ki va-
lue for the inhibition of MAO-B by 4a. The reported
electronic Hammett (rm) parameter value and van der
16
12
8
0.5
4
0.4
0.3
0.2
0.1
0.0
0
-0.4 -0.2 0.0 0.2 0.4 0.6
[I], μM
In order to quantify the relationships between MAO-B
inhibitory activity of 6a–g and the physiochemical prop-
erties of the styryl C-4 substituent, a Hansch-type SAR
study was carried out by stepwise multiple linear regres-
sion analysis. Five descriptors were used to define each
substituent. As descriptors of bulkiness, the van der
Waals volume (Vw)28 and Taft steric parameter (Es)27
were employed, while the lipophilicities of the substitu-
ents were described by the Hansch constant (p).27 The
classical Hammett (rp) and Swain–Lupton (F) constants
served as electronic parameters.27 All physiochemical
values of the substituents were obtained from standard
compilations.27,28 Results of the statistical analysis are
presented in Table 5. The only substituent descriptor
-0.01
0.00
0.01
0.02
0.03
0.04
-1
1/[S], μM
Figure 1. Lineweaver–Burk plots of the oxidation of MMTP by
baboon liver MAO-B in the absence (filled circles) and presence of
various concentrations of 4c (open circles, 0.1 lM; filled triangles, 0.2
lM; open triangles, 0.4 lM). The concentration of the baboon liver
mitochondrial preparation was 0.15 mg/mL and the rates are expressed
as nmol mg proteinꢀ1 minꢀ1 of MMDP+ formed. The inset is the
replot of the slopes versus the inhibitor concentrations.

D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
3697
Table 5. Correlations of the MAO-B inhibition constants (logKi) of the (E)-2-styryl-1-methylbenzimidazole analogues (6a–g) with steric, electronic,
and hydrophobic descriptors of the substituents at C-4 of the styryl ringa
Parameter
Slope
y-Intercept
R2
Fb
3.93
Significancec
rp
F
ꢀ1.33 0.67
ꢀ0.90 1.16
ꢀ0.69 0.37
0.70 0.08
0.72 0.18
0.86 0.40
1.24 0.38
1.28 0.09
1.12 0.19
0.44
0.11
0.41
0.94
0.70
0.104
0.47
0.12
0.60
3.52
Vw
Es
p
84.2
11.9
0.0003
0.02
ꢀ1.16 0.34
a The logarithm of the Ki values expressed in lM were used in the linear regression analysis.
b The F test statistic relates the mean squares due to regression to the error variance. Higher F values indicate a better fit and a regression equation
with an F value higher than the critical F value may be judged as significant. Critical F values were calculated as described recently.29
c The significance is the fractional probability that the coefficient of the added variable is zero.
that showed a meaningful correlation with the logarithm
of the Ki values (expressed in lM) was the Taft steric
parameter (Es). Regression analysis of logKi with Es
exhibited a relatively good correlation with a R2 value
of 0.94. The statistical F value was found to be 84.2
for the correlation, a value that is larger than the critical
F value (25.32) for 95% significance (a higher F value
indicates a better correlation).29 The confidence level
of the correlation was greater than 99.9%. All other sin-
gle-parameter correlations with the logKi values exhib-
ited poorer statistical correlations. Therefore, the best
mathematical description of the binding affinity (logKi)
of the C-4 substituted (E)-2-styryl-1-methylbenzimidaz-
ole analogues considered in this study to MAO-B is:
(rp and F) to the regression analysis with Es did not im-
prove the correlation. Although not significant, there
also appears to be a moderate correlation (R2 = 0.70)
between binding affinity (logKi) and the lipophilicity
(p) of the C-4 substituents. Since more lipophilic substit-
uents have increasingly positive Hansch constant (p)
values, the negative correlation (ꢀ1.16 0.34) between
p and the logKi values indicates that enhancement of
the lipophilicity of the C-4 substituents may lead to
better inhibition potency.
3. Discussion
The crystal structures of human recombinant MAO-B in
complex with several pharmacologically important
inhibitors of the enzyme have recently been
reported.30–32 From the surface of the enzyme, the
access channel leading to the active site FAD co-factor
consists of an entrance followed by a substrate cavity.
An inhibitor must traverse an entrance cavity in order
to gain access to the substrate cavity. This is true for
small molecule inhibitors such as isatin that has been
shown to bind within the substrate cavity of the
enzyme.31 A larger inhibitor, such as the reversible
inhibitor 1,4-diphenyl-2-butene (18) (Scheme 5), appears
to exhibit a dual binding mode that involves traversing
both the entrance and substrate cavities.31 Another
inhibitor, trans,trans-farnesol (19), is also reported to
span both the entrance and substrate cavities with the
polar OH moiety in close contact with the flavin located
in the substrate cavity.32 The gate separating the two
cavities is the side chain of Ile-199 which is thought to
exhibit different rotamer conformations that allow for
the fusion of the two cavities in order to accommodate
log Ki ¼ 0:70ðꢁ0:08ÞEs þ 1:28ðꢁ0:09Þ
ðF ¼ 84:2; R2 ¼ 0:94 and n ¼ 7Þ:
ð1Þ
Since larger sterically hindered substituents have
increasingly negative Taft steric parameter (Es) values,
the positive correlation observed with Es (0.70 0.08)
indicates that the MAO-B inhibition potency (logKi)
may be enhanced by substitution at C-4 of the styryl
ring with a sterically bulky substituent (Fig. 2). In
agreement with qualitative inspection of the inhibition
data presented in Table 3, the electronic contribution
of the C-4 substituents to logKi appears to be negligible
since the addition of an electronic substituent parameter
1.6
1.2
0.8
i
K
0.4
0.0
-0.4
-0.8
18
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
OH
Es
Figure 2. Correlation between the logKi values for the inhibition of
MAO-B by (E)-2-styryl-1-methylbenzimidazole analogues (6a–g) and
the Taft steric parameter (Es) of the substituents at C-4 of the styryl
ring. The linear regression line is a graphical representation of Eq. 1.
The correlation coefficient was found to be 0.94.
trans, trans-Farnesol (19)
Scheme 5. The structures of the reversible MAO-B inhibitors
1,4-diphenyl-2-butene (18) and trans,trans-farnesol (19).

3698
D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
these larger inhibitors.32 The potency of MAO-B inhibi-
tion by (E)-8-styrylcaffeines and (E)-2-styrylbenzimidaz-
oles may possibly also be explained by a similar mode of
binding that involves traversing both the entrance and
substrate cavities.33 These inhibitors possibly bind to
MAO-B with the benzimidazole or caffeine ring located
in the substrate cavity of the active site while the styryl
substituent extends into the entrance cavity. In support
of this hypothesis we have found 1-methylbenzimidazole
(7a) to be a very weak inhibitor of MAO-B with a Ki
value of 2084 lM (Table 4). This inhibitor is approxi-
mately 110 times less potent than (E)-2-styryl-1-methyl-
benzimidazole (6a) which was found to have a Ki value
of 18.7 lM (Table 3). This difference in potency is an
indication of the importance of styryl substitution for
a high affinity interaction between the active site of
MAO-B and the (E)-2-styrylbenzimidazolyl type inhibi-
tors. Without the styryl side chain, 1-methylbenzimidaz-
ole is expected to bind to either the substrate or the
entrance cavity leaving the other cavity unoccupied.
Therefore, styryl substitution possibly assists with dual
binding that may lead to increased inhibition potency.
interactions between the substrate cavity and the benz-
imidazolyl analogues are not possible.
Another interesting outcome of this study was that the
(E)-8-styrylcaffeine analogue disubstituted with chlorine
at C-3 and C-4 of the styryl ring (4b) was found to be a
more potent inhibitor (Ki = 36 nM) than CSC (3), which
is monosubstituted with chlorine at C-3 (Ki = 128 nM).
We have previously reported20 that the (E)-8-styrylcaf-
feine analogue that is monosubstituted with chlorine at
C-4 (Ki = 260 nM) also inhibits MAO-B potently. This
synergistic effect of disubstitution on MAO-B inhibition
suggests that there exists, possibly within the entrance
cavity, an aromatic binding pocket with at least two ma-
jor interaction sites, one for substituents at C-3 of the
aromatic ring and another for substituents at C-4. Alter-
natively, the addition of a second substituent may sim-
ply enhance the lipophilicity of the inhibitor and
therefore its affinity for the enzyme. We have previously
shown that the lipophilicity of the C-4 styryl substitu-
ents correlates with MAO-B inhibition potency of
(E)-8-styrylcaffeines.20
The important role played by the styryl side chain is fur-
ther supported by the observation that C-4 substitution
of the phenyl ring has a considerable effect on the MAO-
B inhibition potency of (E)-2-styryl-1-methylbenzimi-
dazole analogues. Since the styryl ring of these inhibitors
is thought to bind within the entrance cavity, specific
interactions between the styryl substituents and amino
acid side chains in the entrance cavity may explain these
observations. It may be predicted that bulky and lipo-
philic styryl substituents may lead to increased binding
affinity since the entrance cavity is lined by the side
chains of hydrophobic amino acids.34 In accordance
with this expectation, the SAR study indicates that
increasing inhibition potency correlates with increasing
bulkiness (Es) and lipophilicity (p) of the C-4 substitu-
ents. We have previously arrived at a similar conclusion
as part of an SAR study with (E)-8-styrylcaffeine ana-
logues bearing different substituents at C-4 of the styryl
ring.20 It was found that the potency of MAO-B inhibi-
tion correlated with both the size (Vw) as well as the
lipophilicity (p) of the C-4 substituents. Enhancement
of the size and lipophilicity of the substituents resulted
in better inhibition. The similarity of the outcomes of
the SAR studies with the (E)-2-styryl-1-methylbenzimi-
dazole analogues and the (E)-8-styrylcaffeine analogues
is an indication that these two types of molecules inter-
act with the active site of the enzyme with a similar
mode of binding—the benzimidazole and caffeine rings
located in the substrate cavity while the styryl side chain
extends into the entrance cavity. The (E)-8-styrylcaffeine
analogues, however, have been found to be considerably
more potent competitive inhibitors of MAO-B than the
corresponding (E)-2-styryl-1-methylbenzimidazole ana-
logues. We believe that this difference may be the result
of binding interactions of the polar functional groups
located on the caffeine ring (e.g., the two carbonyl
oxygens) with the FAD co-factor and/or polar function-
alities of the amino acids located in the substrate
cavity (e.g., Tyr-398 and Tyr-435). Since the benzimid-
azole ring is devoid of polar functional groups, such
4. Experimental
Caution—MMTP is a structural analogue of the nigro-
striatal neurotoxin, 1-methyl-4-phenyl-1,2,3,6-tetrahy-
dropyridine (MPTP), and should be handled using
disposable gloves and protective eyewear. Procedures
for the safe handling of MPTP have been described
previously.35
4.1. Chemicals and instrumentation
All starting materials not described elsewhere were
obtained from Sigma–Aldrich and were used without
purification. 1,3-Dimethyl-5,6-diaminouracil (8a), 1,3-
diethyl-5,6-diaminouracil (8b)36, and the oxalate salt of
MMTP37 were prepared according to previously reported
procedures. Because of chemical instability, compounds
8aand 8b were used within 24 h of preparation. Petroleum
ether used in this study was of a distillation range of
40–60 ꢁC. Proton and carbon NMR spectra were
recorded on a Varian Gemini 300 spectrometer. Proton
(1H) spectra were recorded in CDCl3 or DMSO-d6 at a
frequency of 300 MHz and carbon (13C) spectra at
75 MHz. Chemical shifts are reported in parts per million
(d) downfield from the signal of tetramethylsilane added
to the deuterated solvent. Spin multiplicities are given
as s (singlet), d (doublet), t (triplet), q (quartet), br s
(broad singlet) or m (multiplet) and the coupling
constants (J) are given in hertz (Hz). Direct insertion elec-
tron impact ionization (EIMS) and high resolution mass
spectra (HRMS) were obtained on a VG 7070E mass
spectrometer. Melting points (mp) were determined on a
Gallenkamp melting point apparatus or by differential
scanning calorimetry (DSC) on a Shimadzu DSC-50
instrument. All the melting points are uncorrected.
UV–vis spectra were recorded on a Milton-Roy Spectron-
ic 1201 spectrophotometer. Thin layer chromatography
(TLC) was carried out using silica gel 60 (Macherey-
Nagel) containing UV254 fluorescent indicator.

D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
3699
4.2. General procedure for the synthesis of (E)-8-styryl-
caffeine analogues (4a–f)
(CDCl3) d 3.39 (s, 3H), 3.60 (s, 3H), 4.10 (s, 3H), 7.03
(d, 1H, J = 15.8 Hz), 7.82 (br s, 1H), 7.84 (d, 1H,
J = 15.8 Hz), 7.96 (br s, 2H); 13C NMR (CDCl3) d
27.96, 29.71, 31.70, 108.49, 114.84, 122.42, 124.89,
126.90, 132.48 (q), 134.53, 137.62, 148.44, 148.47,
151.60, 155.25; EIMS m/z 432 (MÅ+); HRMS calcd
432.1021, found 432.1001.
(E)-8-Styrylcaffeine analogues (4a–f) examined in this
study were prepared according to the procedure described
by Suzuki and et al.21 To a solution of 1,3-dimethyl-5,6-
diaminouracil (8a, 3.50 mmol) and 1-ethyl-2-[3-(dimeth-
ylamino)propyl]-carbodiimide hydrochloride (EDAC;
5.11 mmol) in 40 mL dioxane/H2O (1:1) was added the
appropriate commercially available carboxylic acid (9,
3.81 mmol). The pH of the suspension was adjusted to 5
with 2 M aqueous hydrochloric acid and stirring was con-
tinued for an additional 2 h. The reaction mixture was
neutralized with 1 M aqueous sodium hydroxide, cooled
to 0 ꢁC, and the resulting precipitate was collected by fil-
tration. The crude product was dissolved in 40 mL aque-
ous sodium hydroxide (1 M)/dioxane (1:1) and heated for
2 h under reflux. The reaction mixture was cooled to 0 ꢁC,
acidified to a pH of 4 with 4 M aqueous hydrochloric acid,
and the precipitate was collected by filtration. The result-
ing 1,3-methyl-(E)-8-styryl-7H-xanthinyl analogues (10)
were used in the subsequent reaction without further puri-
fication. Iodomethane (0.40 mmol) was added to a stirred
suspension of 10 (0.20 mmol) and potassium carbonate
(0.50 mmol) in 5 mL DMF. Stirring was continued at
60 ꢁC for 60 min, the insoluble materials were removed
by filtration, and sufficient water was added to the filtrate
to precipitate the product (4) that was collected by
filtration. Following crystallization from a mixture of
methanol/ethyl acetate (9:1) analytically pure samples of
4a–f were obtained. For 4e we found the melting point
to be 218 ꢁC, while the reported melting point is
216–218 ꢁC.38 NMR and MS data also correlated to the
corresponding published data.
(E)-8-(2-Thienylethenyl)caffeine (4d) was prepared
from 1,3-dimethyl-5,6-diaminouracil (8a) and 3-(2-thie-
nyl)acrylic acid in a yield of 32%: mp 204 ꢁC (capillary
method); 1H NMR (CDCl3) d 3.36 (s, 3H), 3.57 (s,
3H), 3.99 (s, 3H), 6.63 (d, 1H, J = 15.4 Hz), 7.03 (dd,
1H, J = 5.1, 5.1 Hz), 7.21 (d, 1H, J = 4.8 Hz), 7.30 (d,
1H, J = 5.1 Hz), 7.86 (d, 1H, J = 15.4 Hz); 13C NMR
(CDCl3) d 27.86, 29.67, 31.44, 107.82, 110.21, 127.02,
128.14, 129.46, 130.89, 140.84, 148.53, 149.65, 151.62,
155.09; EIMS m/z 302 (MÅ+); HRMS calcd 302.0837,
found 302.0820.
(E)-8-(2-Furylethenyl)caffeine (4f) was prepared from
1,3-dimethyl-5,6-diaminouracil (8a) and 3-(2-furyl)-
acrylic acid in a yield of 35%: mp >240 ꢁC (capillary
method); 1H NMR (CDCl3) d 3.35 (s, 3H), 3.56 (s,
3H), 3.98 (s, 3H), 6.44 (m, 1H), 6.52 (d, 1H,
J = 3.92 Hz), 6.75 (d, 1H, J = 15.5 Hz), 7.44 (d, 1H,
J = 2.3 Hz), 7.72 (d, 1H, J = 6.32 Hz), 7.50 (d, 1H, J =
15.5 Hz); 13C NMR (CDCl3) d 27.84, 29.63, 31.37,
107.86, 109.11, 112.34, 113.12, 124.63, 143.87, 148.55 ,
149.86, 151.63, 151.79, 155.08; EIMS m/z 286 (MÅ+);
HRMS calcd 286.1066, found 286.1064.
4.3. Synthesis of purino[7,8-c]quinazoline-8,10(9H,11H)-
dione analogues (5a–b)
(E)-8-(3-Bromostyryl)caffeine (4a) was prepared from
1,3-dimethyl-5,6-diaminouracil (8a) and trans-3-bromo-
cinnamic acid in a yield of 37%: mp 229–231 ꢁC (capil-
The purino[7,8-c]quinazoline-8,10(9H,11H)-dione ana-
logues (5a–b) examined here were prepared from the
corresponding 1,3-dialkyl-5,6-diaminouracil (8a–b) and
2-nitrobenzaldehyde according to the synthetic pathway
reported by Ceccarelli et al. (Scheme 3).22 For 5a the
melting point was found to be 258.3 ꢁC (DSC) which
is lower than the reported value of 276.1 ꢁC.22 The
NMR and MS data however were fully consistent with
the structure.
1
lary method); H NMR (CDCl3) d 3.38 (s, 3H), 3.59
(s, 3H), 4.05 (s, 3H), 6.88 (d, 1H, J = 15.8 Hz), 7.25
(m, 1H), 7.45 (dd, 2H, J = 1.7, 7.9 Hz), 7.70 (d, 1H,
J = 15.7 Hz), 7.71 (t, 1H, J = 1.8 Hz); 13C NMR
(CDCl3) d 27.92, 29.72, 31.55, 108.12, 112.59, 123.12,
126.16, 129.81, 130.42, 132.21 , 136.48, 137.64, 148.52,
149.30, 151.65, 155.24; EIMS m/z 374 and 376 (MÅ+);
HRMS calcd 374.0378, found 374.0371.
Purino[7,8-c]quinazoline-8,10(9H,11H)-dione analogue
5b was prepared from 8-(2-aminophenyl)-1,3-diethyl-
7H-xanthine (13b) in a yield of 82.9%: mp 238.2 ꢁC
(chloroform); 1H NMR (DMSO-d6) d 1.21 (t, 3H,
J = 7.0 Hz), 1.35 (t, 3H, J = 7.1 Hz), 4.02 (q, 2H,
J = 6.9 Hz), 4.23 (q, 2H, J = 7.0 Hz), 7.84 (t, 1H,
J = 6.8, 7.9 Hz), 7.96 (t, 1H, J = 6.7, 7.0 Hz), 8.07 (d,
1H, J = 8.1 Hz), 8.53 (d, 1H, J = 8.2 Hz), 9.51 (s, 1H);
13C NMR (DMSO-d6) d 13.13, 13.25, 35.94, 38.63,
101.68, 117.48, 123.56, 128.52, 129.25, 132.51, 136.45,
142.13, 145.78, 149.82, 150.22, 153.89; HRMS calcd
309.1226, found 309.1209.
(E)-8-(3,4-Dichlorostyryl)caffeine (4b) was prepared
from 1,3-dimethyl-5,6-diaminouracil (8a) and trans-
3,4-dichlorocinnamic acid in a yield of 42%: mp
1
>240 ꢁC (capillary method); H NMR (CDCl3) d 3.39
(s, 3H), 3.60 (s, 3H), 4.06 (s, 3H), 6.87 (d, 1H,
J = 15.7 Hz), 7.37 (m, 1H), 7.45 (d, 1H, J = 8.4 Hz),
7.64 (d, 1H, J = 2.1 Hz), 7.68 (d, 1H, J = 15.7 Hz); 13C
NMR (CDCl3) d 27.93, 29.72, 31.56, 108.18, 112.91,
126.46, 128.68, 130.89, 133.24, 133.28, 135.40, 135.57,
148.49, 149.07, 151.62, 155.22; EIMS m/z 365 (MÅ+);
HRMS calcd 364.0494, found 364.0513.
4.4. General procedure for the synthesis of (E)-2-styryl-1-
methylbenzimidazole analogues (6a–g)
(E)-8-(3,5-Ditrifluoromethylstyryl)caffeine (4c) was pre-
pared from 1,3-dimethyl-5,6-diaminouracil (8a) and
trans-3,5-ditrifluoromethylcinnamic acid in a yield of
38%: mp 267–268 ꢁC (capillary method); 1H NMR
To a solution of o-phenylenediamine (15) (10 mmol) and
EDAC (15 mmol) in DMF (10–15 mL) was added the

3700
D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
appropriately substituted cinnamic acid derivative (16)
(10 mmol). 4-(Dimethylamino)pyridine (0.05 equiv)
and imidazole (0.05 equiv) were added as catalysts and
the reaction was stirred at room temperature for 4 h.
Water (150 mL) then was added to the reaction mixture
and the resulting precipitate was collected by filtration.
This intermediate amide (17) was suspended in 20 mL
of 6 M aqueous hydrochloric acid, heated to 90–
100 ꢁC under reflux, and a sufficient amount of ethanol
was added carefully to dissolve the amide. Heating was
continued until TLC (ethyl acetate/petroleum ether, 1:1)
indicated completion of the reaction. The reaction mix-
ture was cooled to 0 ꢁC and the precipitate that formed
was collected by filtration and washed with water. These
(E)-2-styryl-1H-benzimidazole analogues (14) were crys-
tallized from ethanol as the corresponding hydrochloric
acid salts. The yields of 14 were recorded as 24–82%.
The salts (4 mmol) were dissolved in a minimal amount
of DMF (5–10 mL) followed by the addition of potas-
sium carbonate (10 mmol). The suspension was cooled
to 0 ꢁC and iodomethane (4 mmol) was added. The reac-
tion mixture was allowed to return to room temperature
and stirring was continued until TLC (ethyl acetate/
dichloromethane, 1:1) indicated completion of the reac-
tion. Except for compounds 6a and 6f–g, the final prod-
uct was isolated by the addition of water (100 mL). The
precipitate that formed was collected via filtration and
crystallized from an appropriate solvent. In the cases
of 6a and 6f–g, the remaining DMF solvent was
removed via vacuum distillation. The residue was either
extracted to chloroform (3· 30 mL) or directly crystal-
lized using an appropriate solvent. The melting points
of compounds reported previously were as follows: 6a
mp 121–122 ꢁC (from toluene, capillary method), lit.
mp 119–121 ꢁC;18 6b mp 155 ꢁC (from ethyl acetate/
dichloromethane, 1:1), lit. mp 145–146 ꢁC;23 6c mp
161 ꢁC (from ethyl acetate/dichloromethane, 1:1), lit.
mp 152–153 ꢁC;23 6d mp 113 ꢁC (from ethyl acetate/
dichloromethane, 1:1), lit. mp 116–118 ꢁC;23 6f mp
108 ꢁC (from acetonitrile), lit. mp 130–131 ꢁC;23 6g mp
133 ꢁC (from acetonitrile), lit. mp 125–126 ꢁC.23 The
characterization of previously unreported 6e is summa-
rized below.
equipped with a 75 mm silica Versapak cartridge using
dichloromethane/ethylacetate/methanol (10:10:2) as mo-
bile phase. The melting point was found to be 54–57 ꢁC
(capillary method) which corresponded with the
reported value of 57–60 ꢁC.40 1H NMR (DMSO-d6) d
3.81 (s, 3H), 7.23 (m, 2H), 7.58 (m, 1H), 7.66 (m, 1H),
8.17 (s, 1H); 13C NMR (DMSO-d6) d 30.54, 110.05,
119.23, 121.34, 122.15, 134.54, 143.31, 144.46; HRMS
calcd 132.0687, found 132.0691.
2-Methyl-1H-benzimidazole (7b) was synthesized
according to Phillips41 by condensing o-phenylenedi-
amine with acetic acid. The melting point was found
to be 175–176 ꢁC (capillary method) which corre-
sponded with the reported value of 176 ꢁC.41 1H NMR
(DMSO-d6) d 2.50 (s, 3H), 7.11 (m, 2H), 7.47 (m, 2H),
10.92 (br s, 1H); 13C NMR (DMSO-d6) d 14.57,
114.16, 120.97, 121.01, 138.93, 151.20; HRMS calcd
132.0687, found 132.0681.
4.6. MAO-B inhibition studies
Mitochondria were isolated from baboon liver tissue as
described previously42 and stored at ꢀ70 ꢁC in 300 lL
aliquots. Following addition of an equal volume of
sodium phosphate buffer (100 mM, pH 7.4) containing
glycerol (50%, w/v) to the aliquots, the protein concen-
tration was determined by the method of Bradford using
bovine serum albumin as reference standard.43 Since the
mitochondrial fraction obtained from baboon liver
tissue is reported to be devoid of MAO-A activity,26
inactivation of this enzyme was unnecessary. The
MAO-A and -B mixed substrate MMTP (Km = 60.9 lM
for baboon liver MAO-B)26 served as substrate for the
inhibition studies. Incubations were carried out in
sodium phosphate buffer (100 mM, pH 7.4) and con-
tained MMTP (30–120 lM), the mitochondrial isolate
(0.15 mg protein/mL), and various concentrations of
the test inhibitors. The final volume of the incubations
was 500 lL. The stock solutions of the inhibitors were
prepared in DMSO and were added to the incubation
mixtures to yield a final DMSO concentration of 4%
(v/v). DMSO concentrations higher than 4% are
reported to inhibit MAO-B.15 Following incubation at
37 ꢁC for 15 min, the enzyme reactions were terminated
by the addition of 10 lL perchloric acid (70%) and the
samples were centrifuged at 16,000g for 10 min. The
MAO-B catalyzed production of MMDP+ is reported
to be linear for the first 15 min of incubation under these
conditions.26 The supernatant fractions were removed
and the concentrations of the MAO-B generated
product, MMDP+, were measured spectrophotometri-
cally at 420 nm (e = 25,000 Mꢀ1).26 The initial rates of
oxidation at four different substrate concentrations
(30–120 lM) in the absence and presence of three differ-
ent concentrations of the inhibitors were calculated and
Lineweaver–Burk plots were constructed. The slopes of
the Lineweaver–Burk plots were plotted versus the
inhibitor concentration and the Ki value was determined
from the x-axis intercept (intercept = ꢀKi). Linear
regression analysis was performed using the SigmaPlot
software package (Systat Software Inc.). Each Ki value
reported here is representative of a single determination
(E)-2-(4-Trifluoromethylstyryl)-1-methylbenzimidazole
(6e). Following crystallization from ethyl acetate/dichlo-
romethane (1:1) light yellow crystals were obtained in a
yield of 42.8%; mp 127 ꢁC (capillary method); 1H NMR
(DMSO-d6) d 3.94 (s, 3H), 7.22 (m, 2H), 7.52 (m, 1H),
7.62 (m, 1H), 7.66 (d, 1H, J = 15.8 Hz), 7.76 (d, 2H,
J = 8.2 Hz), 7.89 (d, 1H, J = 15.8 Hz), 8.02 (d, 2H,
J = 8.2 Hz); 13C NMR (d6-DMSO) d 29.59, 110.22,
117.27, 118.61, 122.17, 122.25, 125.53 (q), 127.93,
128.30, 128.72, 133.91, 136.06, 140.00, 142.73, 150.31;
HRMS calcd 302.1031, found 302.1043.
4.5. Synthesis of 1-methylbenzimidazole (7a) and
2-methyl-1H-benzimidazole (7b)
1-Methylbenzimidazole (7a) was synthesized from benz-
imidazole according to the literature procedure reported
for the synthesis of 1,2-dimethylbenzimidazole.39 The
crude product was purified with a VersaFlash station

D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
3701
where the correlation coefficient (R2 value) of the replot
of the slopes versus the inhibitor concentrations was at
least 0.98.
14. Thebault, J. J.; Guillaume, M.; Levy, R. Pharmacotherapy
2004, 24, 1295.
15. Gnerre, C.; Catto, M.; Leonetti, F.; Weber, P.; Carrupt, P.
-A.; Altomare, C.; Carotti, A.; Testa, B. J. Med. Chem.
2000, 43, 4747.
16. Chimenti, F.; Maccioni, E.; Secci, D.; Bolasco, A.;
Chimenti, P.; Granese, A.; Befani, O.; Turini, P.; Alcaro,
S.; Ortuso, F.; Cirilli, R.; La Torre, F.; Cardia, M. C.;
Distinto, S. J. Med. Chem. 2005, 48, 7113.
17. Chen, J. F.; Xu, K.; Petzer, J. P.; Staal, R.; Xu, Y. H.;
Beilstein, M.; Sonsalla, P. K.; Castagnoli, K.; Castagnoli,
N., Jr.; Schwarzschild, M. A. J. Neurosci. 2001, 21,
RC143.
18. Petzer, J. P.; Steyn, S.; Castagnoli, K. P.; Chen, J. F.;
Schwarzschild, M. A.; Van der Schyf, C. J.; Castagnoli,
N., Jr. Bioorg. Med. Chem. 2003, 11, 1299.
19. Castagnoli, N., Jr.; Petzer, J. P.; Steyn, S.; Castagnoli, K.;
Chen, J. F.; Schwarzschild, M. A.; Van der Schyf, C. J.
Neurology 2003, 61(Suppl. 6), S62.
20. Vlok, N.; Malan, S. F.; Castagnoli, N., Jr.; Bergh, J. J.;
Petzer, J. P. Bioorg. Med. Chem. 2006, 14, 3512.
21. Suzuki, F.; Shimada, J.; Shiozaki, S.; Ichikawa, S.; Ishii,
A.; Nakamura, J.; Nonaka, H.; Kobayashi, H.; Fuse, E.
J. Med. Chem. 1993, 36, 2508.
4.7. SAR study
The values of the substituent descriptors rp, F, p, and Es
were obtained from Hansch and Leo27, while those for
the van der Waals volume (Vw) were obtained from
compilations by Van de Waterbeemb and Testa.28 Lin-
ear regression analysis of the logKi values as a function
of the substituent descriptor values was carried out with
Statistica software package (StatSoft Inc.). In order to
estimate the significance of the regression equations
the F statistic was employed. An F value higher than
the critical F value was judged to be significant and crit-
ical F values were calculated as described recently.29 The
critical F value (Fmax) for 95% significance for models
constructed from seven logKi values (Table 3) and which
contains one parameter (out of a possible five: Vw, Es, p,
rp, F) was calculated to be 25.32.
22. Ceccarelli, S.; D’Alessandro, A.; Magnante, F.; Scuri, R.;
Zanarella, S. Farmaco 1993, 48, 1301.
Acknowledgments
23. Pushkina, L. N.; Mazalov, S. A.; Postovskii, I. Y. J. Gen.
Chem. USSR (Engl. Transl.) 1962, 32, 2585.
24. Dubey, P. K.; Kumar, R.; Chameron, T. S.; Grosssert,
J. S.; Hooper, D. L. Indian J. Chem., Sect B 1999, 38,
914.
We are grateful to Jan du Preez and the staff of the
Analytical Technology Laboratory, North-West
University, for their support. The NMR and MS spectra
´
were recorded by Andre Joubert, Johan Jordaan, and
25. Ramaiah, K.; Dubey, P. K.; Eswara, R. D.; Ramanatham,
J.; Ramesh, K.; Grossert, J. S.; Hooper, D. L. Indian
J. Chem., Sect B 2000, 39, 904.
26. Inoue, H.; Castagnoli, K.; Van der Schyff, C. J.; Mabic, S.;
Igarashi, K.; Castagnoli, N., Jr. J. Pharmacol. Exp. Ther.
1999, 291, 856.
Louis Fourie of the SASOL Centre for Chemistry,
North-West University. This work was supported by
grants from the National Research Foundation and
the Medical Research Council, South Africa.
27. Hansch, C.; Leo, A. Exploring QSAR. Fundamentals and
Applications in Chemistry and Biology; American Chemical
Society: Washington, DC, 1995, pp 1–124.
References and notes
28. Van de Waterbeemb, H.; Testa, B. In Advances in Drug
Research; Testa, B., Ed.; Academic Press: London, 1987;
pp 85–225.
1. Youdim, M. B. H.; Bakhle, Y. S. Br. J. Pharmacol. 2006,
147, S287.
2. Waldmeier, P. C. J. Neural Transm. Suppl. 1987, 23, 55.
3. Chiba, K.; Trevor, A. J.; Castagnoli, N., Jr. Biochem.
Biophys. Res. Commun 1984, 120, 574.
4. Collins, G. G. S.; Sandler, M.; Williams, E. D.; Youdim,
M. B. H. Nature 1970, 225, 817.
5. Youdim, M. B. H.; Collins, G. G. S.; Sandler, M.; Bevan-
Jones, A. B.; Pare, C. M.; Nicholson, W. J. Nature 1972,
236, 225.
29. Livingstone, D. J.; Salt, D. W. J. Med. Chem. 2005, 48,
661.
´
30. Binda, C.; Hubalek, F.; Li, M.; Herzig, Y.; Sterling, J.;
Edmondson, D. E.; Mattevi, A. J. Med. Chem. 2004, 47,
1767.
31. Binda, C.; Hubalek, F.; Restelli, N.; Edmondson, D. E.;
´
Mattevi, A. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 9750.
´
32. Hubalek, F.; Binda, C.; Khalil, A.; Li, M.; Mattevi, A.;
6. Birkmayer, W.; Riederer, P.; Youdim, M. B. H.; Linauer,
W. J. Neural. Transm 1975, 36, 303.
Castagnoli, N., Jr.; Edmondson, D. E. J. Biol. Chem.
2005, 280, 15761.
7. Finberg, J. P.; Wang, J.; Bankiewich, K.; Harvey-White,
J.; Kopin, I. J.; Goldstein, D. S. J. Neural Transm. Suppl.
1998, 52, 279.
8. Gesi, M.; Santinami, A.; Ruffoli, R.; Conti, G.; Fornai, F.
Pharmacol. Toxicol. 2001, 89, 217.
9. Gotz, M. E.; Freyberger, A.; Riederer, P. J. Neural
Transm. Suppl. 1990, 29, 241.
10. Ebadi, M.; Brown-Borg, H.; Ren, J.; Sharma, S.; Shavali,
S.; El ReFaey, H.; Carlson, E. C. Curr. Drug Targets 2006,
7, 1513.
11. Rabey, J. M.; Sagi, L.; Huberman, M.; Melamed, E.;
Korczyn, A.; Giladi, M.; Inzelberg, R.; Djaldetti, R.;
Klein, C.; Berecz, G. Clin. Neuropharmacol. 2000, 23, 324.
12. Dingemanse, J.; Wood, N.; Jorga, K.; Kettler, R. Br.
J. Clin. Pharmacol. 1997, 43, 41.
´
33. Binda, C.; Hubalek, F.; Li, M.; Castagnoli, N.; Edmond-
son, D. E.; Mattevi, A. Neurology 2006, 67, S5.
´
34. Binda, C.; Newton-Vinson, P.; Hubalek, F.; Edmondson,
D. E.; Mattevi, A. Nat. Struct. Biol. 2002, 9, 22.
35. Pitts, S. M.; Markey, S. P.; Murphy, D. L.; Weisz, A.
In MPTP:
A Neurotoxin Producing a Parkinsonian
Syndrome; Markey, S. P., Castagnoli, N., Jr., Trevor,
A. J., Kopin, I. J., Eds.; Academic Press: New York,
1986; pp 703–716.
36. Blicke, F. F.; Godt, H. C., Jr. J. Am. Chem. Soc. 1954, 76,
2798.
37. Bissel, P.; Bigley, M. C.; Castagnoli, K.; Castagnoli, N.,
Jr. Bioorg. Med. Chem. 2002, 10, 3031.
38. Del Giudice, M. R.; Borioni, A.; Mustazza, C.; Gatta, F.;
Dionisotti, S.; Zocchi, C.; Ongini, E. Eur. J. Med. Chem.
1996, 31, 59.
13. Chazot, P. L. Curr. Opin. Investig. Drugs 2001, 2, 809.

3702
D. van den Berg et al. / Bioorg. Med. Chem. 15 (2007) 3692–3702
39. Gazit, A.; Yee, K.; Uecker, A.; Bo¨hmer, F. D.; Sjo¨blom,
T.; Ostman, A.; Waltenberger, J.; Golomb, G.; Banai, S.;
Heinrich, M. C.; Levitzki, A. Bioorg. Med. Chem. 2003,
11, 2007.
40. Bergman, J.; Sand, P. Tetrahedron Lett. 1984, 25, 1957.
41. Phillips, M. A. J. Chem. Soc. 1928, 2, 2393.
42. Salach, J. I.; Weyler, W. Methods Enzymol. 1987, 142, 627.
43. Bradford, M. M. Anal. Biochem. 1976, 72, 248.