Inhibition of monoamine oxidases by functionalized coumarin derivatives: Biological activities, QSARs, and 3D-QSARs

Article abstract of DOI:10.1021/jm001028o

A large series of coumarin derivatives (71 compounds) were tested for their monoamine oxidase A and B (MAO-A and MAO-B) inhibitory activity. Most of the compounds acted preferentially on MAO-B with IC₅₀ values in the micromolar to low-nanomolar range; high inhibitory activities toward MAO-A were also measured for sulfonic acid esters. The most active compound was 7-[(3,4-difluorobenzyl)oxy]-3,4-dimethylcoumarin, with an IC₅₀ value toward MAO-B of 1.14 nM. A QSAR study of 7-X-benzyloxy meta-substituted 3,4-dimethylcoumarin derivatives acting on MAO-B yielded good statistical results (q² = 0.72, r² = 0.86), revealing the importance of lipophilic interactions in modulating the inhibition and excluding any dependence on electronic properties. CoMFA was performed on two data sets of MAO-A and MAO-B inhibitors. The GOLPE procedure, with variable selection criteria, was applied to improve the predictivity of the models and to facilitate the graphical interpretation of results.

Full text of DOI:10.1021/jm001028o

J . Med. Chem. 2000, 43, 4747-4758
4747
In h ibition of Mon oa m in e Oxid a ses by F u n ction a lized Cou m a r in Der iva tives:
Biologica l Activities, QSARs, a n d 3D-QSARs
Carmela Gnerre,† Marco Catto,‡ Francesco Leonetti,‡ Peter Weber,† Pierre-Alain Carrupt,† Cosimo Altomare,‡
Angelo Carotti,‡ and Bernard Testa*,†
Institut de Chimie The´rapeutique, Universite´ de Lausanne, BEP, CH-1015 Lausanne, Switzerland,
and Dipartimento Farmaco-Chimico, University of Bari, I-70125 Bari, Italy
Received J uly 17, 2000
A large series of coumarin derivatives (71 compounds) were tested for their monoamine oxidase
A and B (MAO-A and MAO-B) inhibitory activity. Most of the compounds acted preferentially
on MAO-B with IC50 values in the micromolar to low-nanomolar range; high inhibitory activities
toward MAO-A were also measured for sulfonic acid esters. The most active compound was
7-[(3,4-difluorobenzyl)oxy]-3,4-dimethylcoumarin, with an IC50 value toward MAO-B of 1.14
nM. A QSAR study of 7-X-benzyloxy meta-substituted 3,4-dimethylcoumarin derivatives acting
on MAO-B yielded good statistical results (q2 ) 0.72, r2 ) 0.86), revealing the importance of
lipophilic interactions in modulating the inhibition and excluding any dependence on electronic
properties. CoMFA was performed on two data sets of MAO-A and MAO-B inhibitors. The
GOLPE procedure, with variable selection criteria, was applied to improve the predictivity of
the models and to facilitate the graphical interpretation of results.
In tr od u ction
MAO-A inhibitors, such as clorgyline, also seem to
reduce neuronal cell death.7 Various observations stress
the need for new, possibly reversible and safer MAO-B
inhibitors in the therapy of PD and call for further
investigations of the possible role of MAO-A inhibitors
(already in use as antidepressants and antianxiety
drugs8) as antiapoptotic agents in neurodegenerative
diseases.
Unfortunately, the rational design of new MAO
inhibitors (MAO-I) is strongly hampered by a lack of
reliable three-dimensional structural information on the
two active sites of the two isoenzymes. For this reason,
several indirect modeling studies9-13 have been carried
out on numerous and structurally diverse classes of
MAO-I, but the molecular requirements responsible for
selective MAO-A or MAO-B inhibition are only partly
understood.
Coumarins are members of the class of compounds
called benzopyrones and display a variety of pharma-
cological properties14 depending on their substitution
pattern. Some coumarin derivatives of natural15 and
synthetic origin16 have been characterized as MAO-I.
To obtain a better understanding of the structure-
activity relationships (SARs) of MAO-I and to improve
their inhibitory activity, we have designed, prepared,
and tested in vitro several 3-, 4-, 6-, and 7-substituted
coumarins, as well as chromone derivatives and bicyclic
compounds. Quantitative structure-activity relation-
ships (QSARs) and 3D-QSAR models using the com-
parative molecular field analysis (CoMFA)/GOLPE ap-
proach are described in this paper. The results obtained
reveal interesting features about the topography of the
active site of MAO-A and MAO-B.
Monoamine oxidase (MAO) is an FAD-containing
enzyme tightly bound to the mitochondrial outer mem-
brane of neuronal, glial, and other cells. Its roles include
regulation of the levels of biogenic amines in the brain
and the peripheral tissues by catalyzing their oxidative
deamination.1 On the basis of their substrate and
inhibitor specificities, two types of MAO (A and B) have
been described: MAO-A preferentially deaminates se-
rotonin, norepinephrine, and epinephrine and is ir-
reversibly inhibited by low concentrations of clorgyline;
MAO-B preferentially deaminates â-phenylethylamine
and benzylamine and is irreversibly inhibited by de-
prenyl.2
Due to their role in the metabolism of monoamine
neurotransmitters, MAO-A and MAO-B present a con-
siderable pharmacological interest. In Parkinson’s dis-
ease (PD), dopamine (DA) replacement therapy with
L-DOPA is the first-line treatment of this severe neu-
rodegenerative disorder.3 Unfortunately, the rapid meta-
bolic transformation of L-DOPA, at both the peripheral
and central levels, strongly hampers its high therapeutic
potential. For this reason, L-DOPA is typically combined
with inhibitors of aromatic amino acid decarboxylase
(AADC), such as benserazide and carbidopa, and/or with
reversible and selective catechol O-methyltransferase
(COMT) inhibitors, such as entacapone and tolcapone,4
and sometimes also with selective and irreversible
MAO-B inhibitors, such as deprenyl.5
Paradoxically, recent findings indicate that both
deprenyl and L-DOPA at high doses may induce neu-
ronal apoptosis.6 In contrast, deprenyl at low doses
exerts a significant neuroprotective effect by blocking
apoptotic cell death.6 Interestingly, selective irreversible
Resu lts a n d Discu ssion
* To whom correspondence should be addressed. Phone: +4121 692
4521. Fax: +4121 692 4525. E-mail: Bernard.Testa@ict.unil.ch.
† Universite´ de Lausanne.
Ch em istr y. The structures of the compounds pre-
pared in this study, mostly coumarin derivatives, are
listed in Tables 1-4. Some of these compounds have
‡ University of Bari.
10.1021/jm001028o CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/11/2000

4748 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25
Gnerre et al.
Sch em e 1
Sch em e 3a
a
(i) N-Bromosuccinimide, CCl4, reflux; (ii) PhOH or PhNH2,
K2CO3, abs. EtOH, reflux.
starting material for the synthesis of compound 14 (i.e.,
4,7-dihydroxycoumarin)24 was easily prepared by the
method of Dickenson25 using 2,4-dihydroxyacetophenone
and diethyl carbonate. Compound 21 was obtained by
oxidizing 20 with DDQ.
7-Phenoxymethyl (3) and 7-anilinomethyl (4) cou-
marins were prepared by bromination of the methyl
group of the commercial 7-methylcoumarin with N-
bromosuccinimide (NBS) and subsequent nucleophilic
displacement of bromide with phenol and aniline,
respectively (Scheme 3).
Sch em e 2a
Compounds 30-32 were obtained through benzyla-
tion of 6,7-dihydroxy-3,4-dimethylcoumarin, the latter
being prepared according to known procedures,26 namely
acetylation of p-benzoquinone to give 1,2,4-triacetoxy-
benzene and cyclization of this intermediate with ethyl
2-methylacetoacetate in 75% sulfuric acid.
The newly synthesized 7-X-substituted benzyloxy-3,4-
dimethylcoumarins (i.e., 52, 54-56, 69-71, see Table
4) were obtained by a classical alkylation reaction of
7-hydroxy-3,4-dimethylcoumarin with the appropriate
benzyl bromide. The preparation of compounds 52 and
69 was carried out with the respective benzyl bromides,
bearing the 3-OH groups protected as benzoyl esters.
Reduction of the nitro group in compound 61 provided
the amino congener 55, and acetylation of 55 gave
compound 56.
a
(i) Benzyl triphenylphosphonium chloride, Na, EtOH; (ii) ethyl
2-methylacetoacetate, H2SO4, 120 °C.
been patented17,18 and are already reported in the
literature as reversible and selective MAO-I.16 The
reaction yields and physicochemical and spectral data
of the newly synthesized compounds are given in the
Experimental Section and in Table S1 (Supporting
Information).
According to reported procedures, the parent chromone
derivatives (Table 2) were prepared from dihydroxyac-
etophenones by treatment with ethyl orthoformate and
perchloric acid (35 and 36)27 or alternatively by acyla-
tion, cyclization of the acyl intermediate in refluxing
toluene in the presence of Na, and finally alkaline
hydrolysis of the acyloxychromone derivative (37-39).28
Compounds 36 and 37 are novel, and their physico-
chemical data are reported in Table S1 (Supporting
Information).
The hetero-benzofused precursors of the compounds
in Table 3 were prepared using previously reported
procedures,29,30 whereas 5-benzyloxyphthalimide was
synthesized starting from the 5-amino congener in two
steps: i.e., a diazotation/hydroxylation achieved via
Sandmeyer reaction, by using Cu(NO3)2 and Cu2O,31 and
a subsequent classical alkylation of the OH group at the
5-position with benzyl bromide (Scheme 4). All other
benzyloxy derivatives reported in Table 3 were prepared
by a classical alkylation reaction of the hydroxy-
substituted benzofused compounds.
The parent 7-hydroxycoumarins, if not commercially
available, were prepared by the method of von Pech-
mann19,20 from m-resorcinol and the ethyl esters of
appropriate â-keto acids, in the presence of catalytic
amounts of sulfuric acid at 120 °C. The synthesis of
7-amino-3,4-dimethylcoumarin, the parent compound of
16, 23, and 28 (Table 1), was performed using ZnCl2 in
order to minimize the formation of the 7-hydroxyquino-
lin-2-one derivative. The latter was used as the starting
material for the synthesis of compound 48 (Table 3).
Hydroxy- or aminocoumarins were reacted with benzyl,
benzoyl, or benzenesulfonyl halides under basic condi-
tions to give the corresponding benzyl, benzoyl, or
benzenesulfonyl derivatives. Scheme 1 illustrates the
general pathway adopted for preparing most of the
compounds examined in this study.
The 7-trans-styryl derivative 29 was prepared by
reacting ethyl 2-methylacetoacetate with 3-trans-
styrylphenol, the latter being prepared by a Wittig
reaction using 3-hydroxybenzaldehyde and benzyltriph-
enylphosphonium chloride (Scheme 2).21,22
MAO-A a n d MAO-B In h ibition . Seventy-one com-
pounds, most of which are derivatives of 7-hydroxycou-
marin, were tested for their inhibitory effect on MAO-A
and MAO-B. To allow a separate analysis of their SARs,
the compounds were divided into four series: namely a
Compound 9 was obtained by benzylation of 7-hy-
droxy-3,4-dihydrocoumarin, prepared by treating resor-
cinol and acrylic acid with Amberlist-15 in refluxing
toluene, according to a known procedure,23 whereas the

Inhibition of MAO by Functionalized Coumarins
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25 4749
Sch em e 4a
act preferentially as MAO-B inhibitors, with IC50 values
in the micromolar to low-nanomolar range. From a SAR
viewpoint, it appears that substitution at the 7-position
of the coumarin nucleus is essential for good inhibitory
activity.
The most active MAO-B inhibitor was 7-[(3,4-difluo-
robenzyl)oxy]-3,4-dimethylcoumarin (67; Table 4) with
an IC50 value of 1.14 nM, whereas the most active
MAO-A inhibitor was 7-(4-nitrophenylsulfonyloxy)-3,4-
dimethylcoumarin (27; Table 1) with an IC50 value of
10.7 nM. These experimental observations agree with
previous literature data16 demonstrating a key role of
the 7-substituent, the ether derivatives being MAO-B
selective but the sulfonic acid esters displaying MAO-A
selectivity. It is worth noting that for compounds 15 and
62-64, we observed the same activity rank order as
reported in the literature (i.e., 62 < 15 < 63 ≈ 64).16
Typically, our data, determined by a different method
(spectrophotometric versus radiochemical assay), showed
pIC50 values lower by up to 1 order of magnitude
compared to those reported by Rendenbach-Mu¨ller et
al.16
a
(i) NaNO2, H2O, H+; (ii) Cu(NO3)2, Cu2O, H2O; (iii) PhCH2Br,
K2CO3, abs. EtOH, reflux.
Qu a lita tive SARs. For MAO-B inhibition, a com-
parison between the biological data of the coumarin
derivatives in Table 1 and the compounds in Tables 2
and 3 unequivocally proved the essential role of the
coumarin nucleus. In fact, compared with the MAO-B
inhibitory activity of the coumarins (Tables 1 and 4),
the chromone derivatives 35-39, the 2-quinolinone
derivative 48, and all other bicyclic compounds contain-
ing a benzofused 5-membered heterocycle showed lower
MAO-B inhibition, the most active inhibitor 37 having
a pIC50 value of 6.90.
Within the coumarin series (Table 1), 7-benzyloxy-
coumarin (2) was quite active and selective toward
MAO-B. Hydrogenation of the 3,4-double bond (9), as
well as a substituent in the 6-position regardless of size
and lipophilicity (5-8), abolished or significantly de-
creased activity. Substitution at positions 3 and/or 4 of
the coumarin nucleus modulated MAO-B inhibitory
activity and A/B selectivity. Indeed, monosubstitution
in position 4 with a CH3 led to more active compounds
(10 and 11), whereas phenyl (12), trifluoromethyl (13),
and hydroxy (14) groups at the same position decreased
activity. Simultaneous substitution with methyl groups
at positions 3 and 4 was even more favorable for MAO-B
inhibition activity (compare 15 with 2 and 10). 3,4-
Annelation with 5- and 6-membered rings was favorable
only for the cycloaliphatic and smaller cyclopentenyl
rings (compare 19 with 15, 20 and 21). The coumarin
derivative 22 bearing phenyl and methyl groups at
positions 3 and 4, respectively, lacked both MAO-A and
MAO-B inhibitory activity. From such a rank order of
activity, we can infer that the MAO-B binding site
should accept lipophilic substituents of limited size at
positions 3 and 4 of the coumarin nucleus, whereas
hydrophilic (e.g., OH) or larger hydrophobic (e.g., CF3,
phenyl) are not well tolerated. Other factors of electronic
nature may interfere in the case of CF3 substituent.
F igu r e 1. Lineweaver-Burk plots for compound 2 toward
MAO-A (a) and MAO-B (b).
coumarin series (34 compounds, Table 1) with variations
mainly in positions 3, 4, 6, and 7 of the coumarin
nucleus, a second series (5 compounds, Table 2) group-
ing chromone derivatives, a third series (13 compounds,
Table 3) with bicyclic structures containing a benzo
moiety fused with various 5(or 6)-membered hetero-
cycles, and a fourth series of 24 7-benzyloxy-substituted
congeners of 3,4-dimethylcoumarin (Table 4) designed
for QSAR purposes.
Lineweaver-Burk representations for compound 2 on
MAO-A and MAO-B (Figure 1) demonstrate that its
mechanism of inhibition was competitive in both cases.
To confirm this mechanism for the other compounds,
incubations were performed for all inhibitors at their
IC50, at a substrate concentration of 500 µM, resulting
in a marked reduction in the calculated degree of
inhibition. Preincubating the compounds for 15 instead
of 5 min at their IC50 and at the same substrate
concentration did not affect the degree of inhibition.
MAO inhibition data expressed as pIC50 values in
Tables 1-4 show that the majority of the compounds
As for the structural variations of substituents at the
7-position, it appears that the benzyloxy substituent
elicits ideal interactions within the MAO-B binding
pocket. In fact, a simple inversion of the direction of the
OCH2 functionality (2 versus 3) leads to a slight

4750 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25
Gnerre et al.
Ta ble 1. Chemical Structures and MAO Inhibition Data of Coumarin Derivatives
a
pIC50
compd
R3
R4
R6
R7
MAO-A
MAO-B
1
2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
4.39
4.92
OCH2C6H5
CH2OC6H5
CH2NHC6H5
OCH2C6H5
OH
5.17
7.26
3
6.41
4.38
18% (50 µM)
4.63
7.07
4
5.67
5
OCH3
5.17
6
OCH2C6H5
5.69
7b
8
glucosyl
OH
9% (100 µM)
15% (100 µM)
4 µMd
5.71
7% (100 µM)
11% (100 µM)
20% (4 µM)
7.74
glucosyl
OCH2C6H5
OCH2C6H5
OCH2C6H5
OCH2C6H4-3′-NO2
OCH2C6H5
OCH2C6H5
OCH2C6H5
OCH2C6H5
NHCH2C6H5
O(CH2)2C6H5
OCH(CH3)C6H5
OCH2C6H5
OCH2C6H5
OCH2C6H5
OCH2C6H5
NHCOC6H5
OSO2C6H5
9c
H2
H
H2
H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33e
34f
CH3
CH3
C6H5
CF3
OH
CH3
CH3
CH3
CH3
H
H
H
6.90
7.88
H
H
4 µMd
5 µMd
26% (15 µM)
6.16
4 µMd
5.86
H
H
H
H
5.80
CH3
CH3
CH3
CH3
H
8.36
H
5.80
6.79
H
6.00
8.25
H
5.45
6.49
(CH2)3
(CH2)4
H
5.80
8.46
H
5.80
8.46
(-CHdCH-)2
H
7% (1 µM)
4% (3 µM)
5.86
7.30
C6H5
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
8% (3 µM)
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
6.72
H
7.12
5.28
H
OSO2C6H4-4′-CH3
OSO2C6H4-4′-OCH3
OSO2C6H4-4′-NO2
NHSO2C6H4-4′-CH3
trans-CHdCHC6H5
OCH2C6H5
OH
OCH2C6H5
OCH2C6H5
OH
7.33
17% (15 µM)
H
7.15
4.77
H
7.90
20% (2 µM)
29% (40 µM)
7.55
H
4.17
H
6.39
OH
5.03
7.55
OCH2C6H5
3.95
5.51
OCH2C6H5
0.4 µMd
6.25
0.4 µMd
5.48
H
H
35% (30 µM)
18% (30 µM)
a
b
Or percent inhibition at maximum solubility in parentheses; SD on IC50 values < 10%. Esculin (natural product). c 3,4-
Dihydrocoumarin derivative. No inhibition at maximum solubility. e 8-Methyl derivative. 5-OCH2C6H5 derivative.
d
f
Ta ble 2. Chemical Structures and MAO Inhibition Data of
Chromone Derivatives
nitude (15 versus 30). The CH2NH bridge at the
7-position (16) decreased both MAO-A and MAO-B
inhibition, the latter much more than the former. The
inverted NHCH2 bridge (compound 4) produced similar
effects. MAO inhibition effects were maintained from a
NHCH2 (16) to a NHCO (23) bridge.
a
Due to its good MAO-B activity and selectivity,
derivative 15 was taken as a lead compound to explore
SARs by various substitutions of the phenyl ring of the
7-benzyloxy group in the 2-, 3-, and 4- positions (Table
4). Derivatives bearing halogen substituents (57, 58, 63,
64, 67, 68), are more active than compound 15 toward
MAO-B. In two cases when a comparison between
positional isomers was possible (CH3 and CN conge-
ners), ortho-substitution appeared unfavorable. Among
the para-substituents, halogens (63 and 64) elicited the
highest activities. The compounds in Table 4 were also
active toward MAO-A, their degree of B/A selectivity
ranging from about 3 (51) to about 1 (69) log units.
However, the congeners showing MAO-A inhibitory
activity higher than (or close to) that of the lead 15 are
the ones bearing strong electron-withdrawing substit-
uents (NO2, CN) irrespective of their position (50, 60,
61, 65, 66, 70) or a halogen in the para-position (63,
64, 67, 69). The role of steric hindrance in influencing
MAO-A inhibition is suggested by the strong decrease
in activity of the congeners bearing bulkier substituents
pIC50
MAO-A
OCH2C6H5 4.79
compd
R2
R6
R7
MAO-B
35
36
37
38
39
H
H
5.98
6.25
6.90
H
OCH2C6H5
CH3 OCH2C6H5
C6H5
C6H5 OCH2C6H5
H
H
5.17
4.66
H
OCH2C6H5 6% (1 µM) 6% (1 µM)
H
3 µMb
38% (3 µM)
a
Or percent inhibition at maximum solubility in parentheses;
SD on IC50 values < 10%. No inhibition at maximum solubility:
b
3 µM.
decrease in MAO-B and a significant increase in MAO-A
inhibitory activity. As a result, a strong drop of B/A
selectivity was observed. Increasing the length of the
ether bridge with a further methylene group did not
significantly change activity (15 versus 17). The com-
parison of the activity values of compounds 30-32, as
well as the activity of derivatives 5-7 and the natural
product esculin (8), clearly reveals the effect of steric
hindrance at position 6: only small substituents such
as OH are tolerated, even though both MAO-A and
MAO-B activities decreased by about 1 order of mag-

Inhibition of MAO by Functionalized Coumarins
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25 4751
Ta ble 3. Chemical Structures and MAO Inhibition Data of
Hetero-Benzofused Derivatives Bearing a Benzyloxy
Substituent
Ta ble 4. MAO Inhibition Data and Physicochemical
Parameters Used for the QSAR Study of 7-X-Substituted
Benzyloxy-3,4-dimethylcoumarins
a
pIC50
compd
X
MAO-A MAO-B
π
Vw
F
15
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
H
2-CH3
2-CN
3-CH3
3-OH
3-OCH3
3-OCF3
3-NH2
3-NHCOCH3
3-F
3-Cl
3-CF3
3-CN
3-NO2
4-CH3
4-F
4-Cl
4-CN
6.16
5.64
6.38
5.48
6.38
5.82
5.23
6.00
5.15
6.24
5.95
5.72
6.66
7.12
5.43
6.91
6.91
7.00
6.08
6.91
6.17
6.94
6.74
6.16
8.36
8.06
7.64
8.36
8.01
8.44
7.94
7.46
6.60
8.55
8.48
8.24
7.97
8.59
8.21
8.52
8.59
8.43
8.07
8.94
8.52
8.13
8.19
8.23
0
0.08
0
0.44 1.01 -0.05
-0.43 1.14 0.64
0.49 1.01 -0.04
-0.67 0.49
-0.09 1.49
1.02 1.60
-1.23 0.67
-0.99 2.70
0.14 0.36
0.71 1.07
0.88 1.11
-0.57 1.14
-0.26 1.20
0.28
0.25
0.37
0.02
0.27
0.42
0.40
0.37
0.50
0.66
0.49 1.01 -0.04
0.14 0.36
0.71 1.07
-0.57 1.14
-0.26 1.20
0.21 0.72
0.28 0.72
-0.43 0.85
-0.52 2.40
0.57 1.80
0.43
0.41
0.51
0.67
0.85
0.84
0.71
1.32
2.35
4-NO2
3,4-F2
3,5-F2
3-OH, 4-F
3,5-(NO2)2
pentafluoro
a
SD on IC50 values < 10%.
a
Or percent inhibition at maximum solubility in parentheses;
coumarin nucleus. In contrast, changing the benzyloxy
substituent at position 7 of compound 2 (∆pIC50 ) 2.09)
with other groups generally resulted in a significant
decline in MAO-B selectivity (see, for example, com-
pounds 3 and 4), the higher effect being observed for
the 7-phenoxymethyl derivative 3.
SD on IC50 values < 10%.
(e.g., 54 and 56). These preliminary considerations bring
useful information on the main physicochemical proper-
ties modulating selectivity toward the two isoforms of
MAO.
QSAR Stu d y of 7-X-Ben zyloxy-Su bstitu ted 3,4-
Dim eth ylcou m a r in Con gen er s. To understand the
relationship between their MAO inhibitory activity and
physicochemical properties, a Hansch-type QSAR study
of the 24 substituted congeners of 7-X-benzyloxy-3,4-
dimethylcoumarin (Table 4) was carried out by stepwise
multiple linear regression (MLR) with cross-validation.
Each substituent was described by six parameters. The
Hansch constant (π), calculated from CLOGP values,32
was used as a descriptor of lipophilicity, bulkiness was
parametrized by the van der Waals volume (Vw) and
molar refractivity (MR), taken from standard compila-
tions,33-35 whereas classical Hammett (σm and σp) and
Swain-Lupton (F and R) constants were used as
electronic parameters. The F and R parameters for the
substituents at the meta- and ortho-positions were
corrected according to Williams and Norrington.36
The squared correlation matrix of variables used in
MLR is reported in Table 5. MAO-A and MAO-B
inhibitory activity values (pIC50) are poorly correlated,
indicating that the same physicochemical properties
modulate differently the inhibition of the two MAO
isoforms.
It had previously been reported16 that the selectivity
of coumarin inhibitors was reversed by an O-SO2
function. Our data confirm this observation (compare
compounds 24-27 with 15 and its congeners). Interest-
ingly, the introduction of the isosteric NHSO2 group
(compound 28) canceled both MAO inhibitory activities,
likely due the ionizable group (pKa ) 7.05 in 5% DMSO/
water). Unfavorable steric effects resulted from the
introduction of a CH3 group on the oxymethylene bridge,
as already shown.16 Indeed, with compound 18 a marked
drop in MAO-B activity relative to compound 15 was
observed whereas MAO-A inhibition was only slightly
decreased. From a structure-selectivity relationship
viewpoint, another interesting comparison can be made
between the inhibitory activity of compounds 15 and 33,
since the introduction of a methyl group at the 8-posi-
tion in compound 15 dramatically inverted selectivity
toward MAO-A (∆pIC50 ranges from 2.20 in compound
15 to -0.77 in compound 33). This appears due mainly
to a drop in MAO-B inhibition of about 3 orders of
magnitude (from 8.24 to 5.48), the MAO-A inhibitory
activity remaining almost constant. Another structural
factor enhancing MAO-B selectivity appears to be the
introduction in compound 2 of methyl groups (15) or a
fused cyclopentenyl ring (19) at positions 3 and 4 of the
Taking into account our previous QSAR results with
pyridazines as reversible MAO-B inhibitors,37,38 we first
examined the relation between lipophilicity and pIC50

4752 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25
Gnerre et al.
Ta ble 5. Squared Correlation Matrix of Parameters Used in
QSAR (values in Table 4)a
pIC50(A) pIC50(B) ∆pIC50
0.243 0.441
π
Vw
F
pIC50(A)
pIC50(B)
∆pIC50
π
1
0.075
0.081
0.132
(0.230) (0.266) (0.086) (0.263) (0.231)
0.104 0.280 0.250 0.026
(0.255) (0.337) (0.373) (0.045)
0.565 0.014 0.066
(0.732) (0.009) (0.070)
0.015 0.003
(0.005) (0.019)
1
1
1
Vw
1
0.143
(0.060)
1
F
a
In parantheses, squared correlation matrix relative to the
subset of meta-X-substituted congeners.
values. Cross-validated MLR carried out on the whole
set of congeners did not yield statistically significant
equations, although an inspection of the plot (see below)
of MAO-B inhibition data against π values suggested a
nonlinear (likely parabolic) relationship. A quadratic
regression performed on all the 24 congeners afforded
an equation explaining less than 60% of the variance
in the biological data and having a cross-validation
F igu r e 2. Plot of pIC50 (MAO-B) values versus Hansch
hydrophobic constant π. The parabolic curve is a graphical
representation of eq 1. Filled circles represent the data set (n
) 11) generating eq 1, whereas unfilled circles represent the
other congeners of Table 4.
steric hindrance unfavorable to the interaction of com-
pounds 49 and 50 with the binding site. Apparently,
electronic factors should not play an important role,
since incorporating any electronic constant into regres-
sion models did not result in equations better than eq
1.
parameter too low to be considered predictive (q2
)
0.25). In contrast, performing regression analysis on the
subset of the meta-X-monosubstitued derivatives, whose
substituents explore a larger space than the para-
derivatives in terms of lipophilic, steric, and electronic
properties, afforded more quantitative information on
the physicochemical properties governing inhibition.
Indeed, by omitting only the meta-acetamido congener
56 as the strongest outlier, we obtained the following
parabolic equation for the meta-isomers:
Within the limits of the examined property space, it
can be inferred that cooperative lipophilic interactions
of meta-X substituents play the most important role in
modulating MAO-B inhibition, with an optimum value
of π around 0.2. Besides lipophilicity, a proper steric fit
of the meta-substituents may be hypothesized, whereas
QSAR analysis clearly excluded a dependence on elec-
tronic properties.
pIC50(MAO-B) ) 0.19((0.15)π - 0.56((0.21)π2 +
8.46((0.14) (1)
With the lipophilicity parameter alone, no significant
result was obtained for MAO-A inhibition, whereas a
trend between electronic, steric, and lipophilic features
and MAO-A inhibition was detected within the subset
of meta-X-substituted 7-benzyloxy coumarins. A three-
parameter equation having satisfactory fitting (but not
cross-validated) statistics (r2 ) 0.85, q2 ) 0.56) was
obtained, which suggests favorable effects of electron-
withdrawing and less bulky substituents and a minor
and negative effect of lipophilicity.
n ) 11; r2 ) 0.86; q2 ) 0.72; s ) 0.14; F ) 24.13
where n represents the numbers of data points, r2 the
squared correlation coefficient, q2 the leave-one-out
cross-validation coefficient, and s the standard deviation
of the regression equation; 95% confidence intervals of
the regression coefficients are given in parentheses.
Equation 1 explains more than 85% of the variance in
the inhibition data and has a satisfactory predictive
ability. The parabolic eq 1 is graphically illustrated in
Figure 2, where the data of the other positional isomers
have been added for comparison.
Figure 2 suggests that most of the 7-X-benzyloxy-3,4-
dimethylcoumarin congeners follow quite well the para-
bolic relationship described by eq 1. Three compounds,
namely the two ortho-substituted derivatives 49 (X )
2-CH3) and 50 (X ) 2-CN) and the meta-NHCOCH3
congener 56, deviated strongly from the regression
curve, suggesting that other factors, presumably steric
and/or conformational, are not accounted for by the
quadratic equation. We believe that the deviating
behavior of compound 56 could be ascribed to the large
size of its meta-acetamido substituent (whose van der
Waals volume equals 2.70, a value significantly higher
than those of the other substituents, ranging between
0.08 and 1.49), whereas cyano and methyl substituents
at the ortho-position of the phenyl ring may produce
The number of para- and ortho-X-substituted conge-
ners synthesized and tested so far and the property
space examined with them are too limited to allow a
systematic comparison between positional isomers.
Nevertheless, the above QSAR results, although limited
to the meta subset, suggest that the MAO-B binding site
(eq 1) should be different from that of MAO-A in terms
of hydrophobic and polar properties.
As far as selectivity is concerned, the regression
analysis performed on the difference between MAO-B
and MAO-A inhibition (∆pIC50 ) pIC50(MAO-B) - pIC50-
(MAO-A)) gave the following equation for the meta-
congeners:
∆pIC50 ) 0.72((0.21)π - 1.08((0.72)F +
2.44((0.26) (2)
n ) 12; r2 ) 0.88; q2 ) 0.83; s ) 0.22; F ) 33.28

Inhibition of MAO by Functionalized Coumarins
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25 4753
Ta ble 6. Statistical Results for Data Sets A, A1, B, and B1 with Steric (ste), Molecular Electrostatic Potential (mep), and Lipophilic
(lip) Fields
SAMPLS Results (q2)
data set
N
ste
mep
lip
ste and mep
ste and lip
mep and lip
ste, mep, and lip
A
A1
B
2
2
2
2
0.14
0.39
0.52
0.27
0.65
0.17
0.17
0.42
0.44
0.44
0.46
0.43
0.67
0.28
0.05
0.45
0.55
0.43
0.56
0.51
0.66
0.45
0.53
0.52
0.70
<0
0.15
0.30
B1
Final models after variable selection
variables
data set
fields
N
q2
r2
A1
B1
mep
ste, mep, and lip
290
355
2
2
0.58
0.80
0.68
0.88
explaining 88% of the variance in the selectivity data
with a good cross-validation result. Attempts to extend
the regression model to the whole set of congeners
resulted in statistically poor equations, most likely
because of the limited physicochemical property space
explored by congeners different from the meta-substi-
tuted ones. However, within the limits of the examined
structural space, eq 2 shows that 7-X-benzyloxy-3,4-
dimethylcoumarins bearing lipophilic and electron-
donating (field effect) substituents at the meta-position
inhibited MAO-B better than MAO-A. Taken globally,
eqs 1 and 2 indicated lipophilicity as the property that
should influence MAO-B versus MAO-A selectivity,
whereas the presence of an electron-withdrawing sub-
stituent should favor MAO-A inhibition.
These findings agree with our previous QSAR results
obtained with a large series of structurally diverse
reversible MAO-I, i.e., 5H-indeno[1,2-c]pyridazine de-
rivatives,37 where we also demonstrated lipophilic in-
teractions to play a significant role in increasing MAO-B
selectivity. However, most of the pyridazine MAO-I
elicited measurable activity only toward the B form, and
this prevented us from obtaining detailed and quantita-
tive information on other physicochemical properties
modulating selectivity. In contrast, the coumarins,
especially the 7-benzyloxy-3,4-dimethylcoumarin de-
rivatives (Table 4) reported herein, were very good
inhibitors of MAO-B with sometimes appreciable activi-
ties toward the A isoform. This allowed us to accurately
assess A/B selectivity and also to carry out multiple
regression analysis providing helpful information to
design inhibitors. Indeed, lipophilicity proved again to
be a major property favoring the binding of coumarins
to MAO-B, whereas a π-electron-poor phenyl ring in the
7-position appears to improve both MAO-A and MAO-B
inhibition, leading to more active but slightly less
selective inhibitors.
Alignment is considered to be a crucial step in every
CoMFA study. The compounds investigated here present
a relatively flexible substituent in position 7. Their
conformational behavior was calculated with the
MMFF94s40 force field, and the conformation chosen
was a flat one with a maximal energy of 4 kcal/mol
above the minimum. The molecules were then super-
imposed on the coumarin ring, and a regularly spaced
(1.5 Å) 3D grid was used.
First, SAMPLS analysis, with steric field, molecular
electrostatic potential, and molecular lipophilicity po-
tential, was run in SYBYL for sets A and B to determine
which fields were statistically significant. Then, the
CoMFA columns were exported in GOLPE where a PCA
was performed on the X-variables matrix and the first
three components were used to observe the objects
(molecules) in the space of the principal components
(PCs). PLS analysis was then applied and a PLS plot
representing the objects in the space of X-scores (T)
against Y-scores (U) was examined. PCA score plots
together with the PLS plot allowed the identification of
outliers, which were removed in the final analysis.
The X-variables of each set were submitted, in GOLPE,
to an advanced pretreatment and then grouped using
the smart region definition (SRD) algorithm. Finally a
variable selection (fractional factorial design, FFD) was
applied in order to retain only the informative variables.
The details of the parameters used in each step are
given in the Experimental Section.
3D-QSAR of MAO-A In h ibitor s. The results for
SAMPLS analysis on data set A are presented in Table
6. The cross-validation values for this data set are
relatively small and would be considered insignificant
were it not for the possibility to use GOLPE. The only
field that could explain MAO-A activity is the electro-
static one. This field was exported in GOLPE, and a
PCA was applied in order to observe the position of the
objects. The first component revealed clearly that
compounds 1 and 4, which have a relatively weak
activity, are not well discriminated in relation to the
other objects of the set. After a PLS analysis, the
viewing of the PLS plot on the first component con-
firmed that these two compounds are outliers. Com-
pound 1 has a smaller size than all other molecules in
the data set and this can be an explanation. For
compound 4, the conversion of the oxymethylene bridge
to a methyleneamino led to a signifiant loss in activity
which could not be explained by the model. Compounds
1 and 4 were removed from data set A; the cross-
validation results for the SAMPLS analysis applied to
the new data set A1 are also presented in Table 6. The
3D-QSAR Stu d y of MAO In h ibition . CoMFA, a
very useful tool in 3D-QSARs,39 was used to analyze the
numerous biological data obtained for MAO-A and
MAO-B inhibition. The need to validate and better
interpret the results generated with the classical CoM-
FA procedure of SYBYL prompted us to use GOLPE.
Training sets of 41 coumarin derivatives acting as
MAO-A inhibitors (set A) and 44 as MAO-B inhibitors
(set B) were chosen among the numerous compounds
available using the following criteria: only the inhibitors
containing a coumarin ring and substituted in the 3-,
4-, or 7-position were selected, to have a good variability
in substituents. The experimental activities were ex-
pressed as the negative logarithm of IC50 (M) (pIC50).

4754 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25
Gnerre et al.
of the sulfone in the most active compounds (24-27).
The third magenta zone in the para-position of the
phenyl ring indicates the favorable effect of electron-
withdrawing substituents on MAO-A activity.
3D-QSAR of MAO-B In h ibitor s. The results for the
SAMPLS analysis on data set B are presented in Table
6. The three fields were exported in GOLPE in order to
observe the objects after a PCA carried out on each field
separately and then in combination. A PLS was also
applied to observe the PLS plot on the first component.
In this case two outliers were identified, namely com-
pounds 3 and 23. These two compounds, again, present
structural modifications of the benzyloxy in position 7,
leading to differences in activity which cannot be
explained by the model. They were removed from set
B, and the results of the SAMPLS analysis are in Table
6. The cross-validation coefficients are significantly
improved by the removal of these two outliers. A model
with the combination of the three fields was chosen and
submitted to a variable selection (see Experimental
Section for details). The graphical results obtained after
variable selection are presented in Figure 3 (B1).
Sterically favorable green zones are present at positions
3 and in a lesser extent 4 of the coumarin ring, because
compounds substituted in position 4 do not always show
a better MAO-B inhibitory activity (e.g., compounds 13
and 14). The other sterically favorable zones on the
phenyl ring are due to the presence of substituents
increasing activity, such as chloro, cyano, and methoxy.
Concerning the lipophilic field, it seems that the pres-
ence of a favorable yellow zone on the lateral chain is
correlated with the information given by the MEP
field: there is a favorable electron-deficient white zone
in the same position. There is another electron-defi-
ciency zone in the coumarin ring generated by substit-
uents such as a fused cyclopentenyl or cyclohexenyl ring
which enhances activity but also generated by substit-
uents such as OH which decrease it. Two other electron-
rich favorable zones are present: one on the lateral
chain near the oxygen of the oxymethylene bridge in
position 7 and the other between the para- and meta-
positions of the phenyl ring.
F igu r e 3. Statistical fields of 3D-QSAR models for MAO-A
and MAO-B inhibition, obtained after variable selection in
GOLPE: PLS coefficients. The color codes are the following:
sterically favorable and nonfavorable influences, green and red
zones, respectively; favorable influence of lipophilic groups,
yellow zone; favorable influence of high electron density and
deficiency, magenta and white zones, respectively. (B1) Results
of the model for MAO-B inhibitors (actual values of PLS
coefficients: ste, 0.0016/-0.0016; lip, 0.0025; mep, 0.0016/-
0.0017). Substitution at positions 3 and 4 of the coumarin ring
have a favorable steric contribution, whereas substitution in
the lateral chain is sterically unfavorable. The substitution of
positions 3, 4, and 5 in the phenyl ring of the benzyloxy is
also sterically favorable but in a less important manner. The
lipophilic character of the CH2 position in the oxymethylene
bridge is favorable for activity. The magenta zones near the
7-position of the coumarin ring indicate that a high electron
density is favorable for activity. Electron-withdrawing sub-
stituents at the 3- and 4-positions in the benzyloxy ring
produce favorable zones for activity. Electron deficiency in
positions 3 and 4 of the coumarin ring and near the 7-position
on the oxymethylene bridge is favorable for the activity. (A1)
Results of the model for MAO-A inhibitors (actual values: mep,
-0.0015). Two electron-rich zones around the oxymethylene
bridge clearly indicate a favorable influence on activity, as well
as another one in the 4-position on the benzyloxy ring.
Con clu sion
Coumarin derivatives proved to be competitive and
reversible MAO-I, usually with a high selectivity for
MAO-B. The most potent compound (67) has an IC50
)
1.14 nM. However, a very potent MAO-A inhibitor (27)
was also found (IC50 ) 10.7 nM).
Taken together, the results from our QSAR and
CoMFA/GOLPE studies afford coherent information
about the nature and spatial location of the main
interactions underlying the potency and selectivity of
MAO-I. The main features required for selectivity are
the lipophilic character of substituents on the phenyl
ring increasing MAO-B selectivity and the presence of
a sulfonic ester function at position 7 leading to a higher
MAO-A activity and suggesting the presence of a
hydrophilic, polar pocket in the active site of the
isoenzyme.
model with the MEP field built with data set A1 is the
only one able to explain the MAO-A activity. Variable
selection, whose parameters can be found in the Ex-
perimental Section, leads to a slightly improved cross-
validation coefficient, also obtained with a “leave-one-
out” procedure (Table 6), and a clearer representation
of the contour maps.
Indeed, graphical results represented in Figure 3 (A1)
show that there are two electron-rich favorable zones
in the lateral chain. These zones are due to the presence
As observed in several previous studies, the coordi-
nated application of 2D- and 3D-QSAR methods yields
significant and complementary insights into SARs,
offering at the three-dimensional level a clear picture

Inhibition of MAO by Functionalized Coumarins
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25 4755
(d, 1 H, J ) 7.9), 7.33 (d, 1 H, J ) 1.6), 7.29 (dd, 1 H, J ) 7.9,
1.6), 6.41 (d, 1 H, J ) 9.5), 4.50 (s, 2 H); IR (cm-1) 1724. Anal.
(C10H7BrO2) C, H.
of the main forces modulating chemical and biological
processes.41
Our findings are in good agreement with previous
SAR studies on MAO-I. Being based on a rather large
array of compounds properly designed for a QSAR study
and on a homogeneous data set spanning quite regularly
a wide range of inhibitory activity, they can be judi-
ciously used for the design of new potent and selective
MAO-I.
Gen er a l P r oced u r e for th e Alk yla tion of Hyd r oxy- or
Am in ocou m a r in s. Hydroxy or amino derivative of coumarin
(10 mmol) was refluxed under magnetic stirring in 100 mL of
absolute ethyl alcohol with 10 mmol of anhydrous potassium
carbonate and 20 mmol of the suitable bromo derivative, until
the reaction was complete (usually 1-4 h). The warm solution
was filtered and cooled to room temperature to give a precipi-
tate that was purified by crystallization. Through this proce-
dure, compounds 3 and 4 were prepared by reacting 7-(bro-
momethyl)coumarin and phenol and aniline, respectively.
Compound 16 was synthesized starting form 7-amino-3,4-
dimethylcoumarin and benzyl bromide, whereas all the other
derivatives in Tables 1 and 4 were obtained from the reaction
of the appropriate hydroxycoumarin and benzyl or phenethyl
(17 and 18) bromides.
Exp er im en ta l Section
Ch em istr y. Melting points were determined by the capil-
lary method on a Gallenkamp MFB 595 010M electrothermal
apparatus and are uncorrected. Elemental analyses were
performed on a Carlo Erba 1106 analyzer for C, H, and N.
The results agreed within (0.40% of the theoretical values.
IR spectra were recorded using a potassium bromide disk on
a Perkin-Elmer 1600 FT-IR spectrophotometer. Only the most
significant IR absorption bands are listed in Table S1 (Sup-
porting Information). 1H NMR spectra were recorded at 300
3-(Ben zyloxy)-6H -b en zo[c]ch r om en -6-on e (21). 2,3-
Dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) (0.23 g, 1.0 mmol)
was added to a solution of 20 (0.31 g, 1.0 mmol), prepared
according to patent literature,17,18 in anhydrous dioxane (10
mL). The reaction mixture was stirred for 2.5 h at reflux, then
filtered off, concentrated under vacuum, and finally diluted
with ethyl acetate. The organic layer was extracted with 2 N
NaOH, washed with brine, dried over Na2SO4, and evaporated
to dryness. The crude residue was purified by flash chroma-
tography (CH2Cl2/hexane, 8:2 v/v, as the mobile phase), giving
the pure product 21.
MHz on
a Bruker 300 instrument. Chemical shifts are
expressed in δ (ppm) and the coupling constants J in hertz
(Hz). The following abbreviations are used: s, singlet; d,
doublet; t, triplet; dd, double doublet; dq, double quadruplet;
td, triple doublet; tt, triple triplet; m, multiplet; br, broad
signal. Signals due to OH and NH protons were located by
deuterium exchange with D2O.
Chromatographic separations were performed on silica gel
(230-400 mesh, Merck) using the flash methodology. Starting
materials, unless otherwise stated, are commercially available
and were purchased from Sigma-Aldrich (Milano, Italy).
Several compounds were synthesized according to reported
procedures with slight modifications; melting points and
spectral data were in full agreement with those reported in
the literature.
Gen er a l P r oced u r e for t h e Syn t h esis of Cou m a r in
Der iva tives. When not commercially available, 7-hydroxy-
coumarins, i.e., the parent compound of most derivatives listed
in Tables 1 and 4, were synthesized according to classical
procedures. m-Resorcinol (20 mmol) and the ethyl ester of an
appropriate â-keto acid (22 mmol) were mixed and kept at 120
°C for 2 h with a catalytic amount of 96% sulfuric acid. After
cooling to room temperature, the mixture was diluted with
ethyl acetate and the solid filtered and purified by crystal-
lization.
N-(3,4-Dim eth yl-2-oxo-2H-ch r om en -7-yl)ben zam ide (23).
Benzoyl chloride (0.62 mL, 5.3 mmol) was added dropwise to
a solution of 7-amino-3,4-dimethylcoumarin (1.01 g, 5.3 mmol)
in 5 mL of 2 N NaOH, cooled at 0 °C. The reaction mixture
was stirred for 15 min, and the white solid obtained filtered
off and purified by crystallization.
N -(3,4-Dim e t h yl-2-oxo-2H -ch r om e n -7-yl)-4-t olu e n e -
su lfon a m id e (28). An excess of 4-toluenesulfonyl chloride
(0.57 g, 3.0 mmol) was added to a solution of 7-amino-3,4-
dimethylcoumarin (0.19 g, 1.0 mmol) in 10 mL of anhydrous
pyridine, cooled at 0 °C. The reaction mixture was stirred for
1 h, and then poured on ice, giving a precipitate that was
filtered, washed, and purified by crystallization.
3,4-Dim e t h yl-7-[(E )-2-p h e n yle t h yl]-2H -ch r om e n -2-
on e (29). 3-Styrylphenol was prepared via a Wittig reaction
from 3-hydroxybenzaldehyde and benzyl triphenylphospho-
nium chloride in ethanolic solution of sodium ethylate, follow-
ing the method described by Friedrich and Henning;21 both
isomers (trans 60% and cis 40%) were obtained and separated
by column chromatography using hexane/ethyl acetate (77.5:
22.5 v/v) as eluent. Reaction of the trans-isomer22 with ethyl
2-methylacetoacetate, according to the above-reported general
procedure for the synthesis of coumarin derivatives, yielded
compound 29.
5,7-Dih yd r oxy-3,4-d im eth ylcou m a r in . Yield: 43%; mp
294-6 °C dec, from ethyl acetate; 1H NMR (CDCl3 + 10%
DMSO-d6) δ 9.41 (s, 1 H, exch.), 9.28 (br, 1 H, exch.), 6.14 (d,
1 H, J ) 2.6), 6.12 (d, 1 H, J ) 2.6), 2.41 (s, 3 H), 1.92 (s, 3 H);
IR (cm-1) 3257, 1626, 1155. Anal. (C11H10O4) C, H.
6,7-Dih yd r oxy-3,4-d im eth ylcou m a r in was synthesized
according to the procedure described by Tamura et al.42
7-Am in o-3,4-d im eth ylcou m a r in . The parent compound of
derivatives 16, 23 and 28 was prepared by addition of ethyl
2-methylacetoacetate (1.4 g, 10 mmol) to a solution of 3-ami-
nophenol (1.1 g, 10 mmol) in dry ethanol (50 mL). Zinc chloride
(1.5 g) was then added and the reaction mixture heated for 30
min at reflux. Solvent was evaporated under reduced pressure
and the obtained residue diluted in 50 mL of a 1:1 solution of
chloroform and dioxane. The organic layer was extracted with
1 N NaOH, washed with brine, dried over Na2SO4 and
concentrated in vacuum to give the desired compound in about
30% yield (lit.43 46% yield).
7-(Br om om eth yl)cou m a r in . 7-Methylcoumarin (1.0 g, 6.2
mmol) was heated in refluxing carbon tetrachloride (20 mL).
N-Bromosuccinimide (1.3 g, 7.5 mmol) and a catalytic amount
of N,N′-azobis(isobutyronitrile) (20 mg) were then added
followed by 30 min of stirring. To the reaction mixture 20 mL
of chloroform were added, and the organic layer was extracted
with brine, dried over Na2SO4 and then concentrated under
vacuum to give the desired compound in 69% yield: mp 164-5
°C, from EtOH; 1H NMR (CDCl3) δ 7.67 (d, 1 H, J ) 9.5), 7.45
7-(Ben zyloxy)-6-h yd r oxy-3,4-d im eth yl-2H-ch r om en -2-
on e (30), 6-(Ben zyloxy)-7-h yd r oxy-3,4-d im eth yl-2H-
ch r om en -2-on e (31), an d 6,7-Bis(ben zyloxy)-3,4-dim eth yl-
2H-ch r om en -2-on e (32). These were prepared following the
general procedure for the alkylation of coumarins bearing
hydroxy groups, by reacting 6,7-dihydroxycoumarin (0.21 g,
1.0 mmol) with benzyl bromide (0.14 mL, 1.2 mmol). After
evaporation of the solvent, the crude residue was washed with
5 mL of hot water, and the precipitate filtered off and
separated by column chromatography using chloroform/hex-
ane/ethyl acetate (90:5:5 v/v) as the mobile phase. The three
compounds were well separated and their structures assessed
by NOESY-NMR experiments.
5-(Ben zyloxy)-7-h yd r oxy-3,4-d im eth yl-2H-ch r om en -2-
on e (34). A solution of benzyl bromide (0.06 mL, 0.5 mmol) in
1 mL of absolute EtOH was added dropwise to a solution of
5,7-dihydroxy-3,4-dimethylcoumarin (0.21 g, 1.0 mmol) in 15
mL of absolute EtOH over a 4-h period, to minimize the
formation of the dibenzyl derivative. After the usual workup,
the residue was separated by column chromatography using

4756 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25
Gnerre et al.
chloroform/ethyl acetate (80:20 v/v) as the mobile phase, and
compound 34, whose structure was assessed by a NOESY-
NMR experiment, was obtained in 17% yield.
Physicochemical and spectral data of the newly synthesized
chromones 36 and 37 and hetero-benzofused derivatives 43,
44, 46, and 48 are reported in Table S1 (Supporting Informa-
tion).
7-[(3-Hyd r oxyben zyl)oxy]-3,4-d im eth yl-2H-ch r om en -2-
on e (52). m-Cresol was benzoylated by a procedure similar to
that described for compound 23. 3-Methylphenyl benzoate,
obtained as a white solid (42% yield), was brominated with
NBS in CCl4, according to the method described above for the
preparation of 7-(bromomethyl)coumarin, to give R-bromo-3-
methylphenyl benzoate (60% yield). This bromo derivative
reacted with 7-hydroxy-3,4-dimethylcoumarin in the usual
alkylation conditions to give 52, which crystallized directly
from the reaction mixture.
7-[(3-Am in ob en zyl)oxy]-3,4-d im et h yl-2H -ch r om en -2-
on e (55). Powdered zinc (0.22 g) was added to a hot solution
of 61 (0.15 g, 0.34 mmol) in 17 mL of ethyl alcohol, under
vigorous magnetic stirring, followed by dropwise addition of
concentrated HCl (0.5 mL). The mixture was refluxed for 30
min, then cooled, filtered off, diluted with water and neutral-
ized with 1 N NaOH. Compound 55 was obtained as a brown
solid, by extraction with ethyl acetate and evaporation of the
organic extracts to dryness.
Biologica l Assa ys. Seventy-one compounds were tested for
their inhibitory activity on MAO-A and MAO-B. Incubations
were carried out at pH 7.4 (Na2HPO4/KH2PO4 made isotonic
with KCl), 37 °C, in rat brain mitochondrial suspensions.51
A
continuous spectrophotometric assay monitoring the rate of
oxidation of the nonselective MAO substrate kynuramine into
4-hydroxyquinoline was used. Clorgyline (250 nM) or selegiline
(250 nM) was preincubated with the suspension for 5 min to
measure MAO-B or MAO-A activity, respectively. The forma-
tion of 4-hydroxyquinoline was followed at 314 nm using a
Kontron UVIKON 941 spectrophotometer. The limited solubil-
ity of the tested compounds in aqueous solution required the
use of 5% DMSO as cosolvent, a concentration that did not
affect MAO activity. The substrate concentration used for the
assays was 60 µM for MAO-B and 90 µM for MAO-A,
corresponding to the respective KM of kynuramine for the two
isoenzymes. IC50 values were determined and calculated from
a hyperbolic equation as already described.52 Lineweaver-
Burk plots on MAO-A and MAO-B were obtained from incuba-
tions at five substrate concentrations, without or with two
different inhibitor concentrations. The inverse values of the
reaction velocities were then represented as a function of the
inverse value of substrate concentration.
N -(3-{[(3,4-D im e t h y l-2-o x o -2H -c h r o m e n -7-y l)o x y ]-
m eth yl}p h en yl)a ceta m id e (56). By treating 55 (0.65 g, 0.22
mmol) with acetic anhydride (1.0 mL) for 1 h at 0 °C, compound
56 was obtained after evaporation of the reaction mixture to
dryness.
p Ka Deter m in a tion . A spectrophotometric method, as
previously described,53 was used for the determination of the
pKa of compound 28. The spectra of the ionized and un-ionized
forms were recorded in 0.1 M HCl and 0.1 M NaOH with 5%
DMSO. A wavelength was chosen where the difference in
absorbance between the un-ionized and ionized forms was
maximal (λ ) 347 nm). The absorbance values at 347 nm and
different pH (7.4, 7.6, 7.0 and 6.8) were measured in a
phosphate buffer (Na2HPO4/KH2PO4, 0.15 M) with 5% DMSO.
The mean value was then determined and the standard
deviation calculated (n ) 4). The spectra and absorbance
values were recorded on a Kontron UVIKON 941 spectropho-
tometer.
7-[(4-F lu or o-3-h yd r oxyb en zyl)oxy]-3,4-d im et h yl-2H -
ch r om en -2-on e (69). 2-Fluoro-5-methylphenol was treated
with benzoyl chloride in the same conditions as for compound
23 (see above). The resulting 3-benzoyloxy-4-fluorotoluene
(nearly quantitative yield) was brominated with NBS in CCl4
(see above) to afford R-bromo-3-benzoyloxy-4-fluorotoluene
(56% yield), which was used without further purification for
the reaction with 7-hydroxy-3,4-dimethylcoumarin. The two
compounds (1 mmol each) reacted for 8 h in 10 mL of refluxing
anhydrous acetone, in the presence of 0.5 mmol of anhydrous
K2CO3, to give 7-(3-benzoyloxy-4-fluorobenzyloxy)-3,4-dim-
ethyl-2H-chromen-2-one (39% yield) which was finally hydro-
lyzed with 2 equiv of NaOH in 10 mL of a mixture of EtOH/
water (1:1, v/v) kept at 60 °C in an oil bath. The clear solution
obtained was then concentrated under vacuum, acidified and
extracted with ethyl acetate; the organic layers, dried over Na2-
SO4, were evaporated giving compound 69 as a solid.
The other newly synthesized coumarin derivatives (i.e., 3-6,
8, 13, 54, 70, 71 in Tables 1 and 4) were prepared according
to the literature as shown in general Scheme 1; their physi-
cochemical and spectral data are listed in Table S1 (Supporting
Information).
6-Benzyloxy-4H-1-benzopyran-4-one (36) and its 2-methyl
congener (37) were prepared by benzylation under classical
conditions (see above for the general procedure) of the corre-
sponding hydroxy derivatives, the latter being synthesized
according to the methods described by J aen et al.27 and Pandit
et al.,28 respectively. Compounds 35,44 3845 and 39,46 already
known, were prepared following reported procedures.
The parent hydroxy compounds of the derivatives reported
in Table 3, when not commercially available, were prepared
using approaches already reported in the literature. Thus,
6-benzyloxybenzofuran 42 was prepared according to a pro-
cedure described by Davies et al.,47 whereas compound 41 was
obtained by catalytic hydrogenation at 60 atm of 42. Com-
pounds 4348 and 4549 were prepared through benzylation of
the hydroxy intermediates, synthesized according to known
methods. 5-Hydroxyphthalimide,30 the precursor of compound
46, was prepared by reduction of m-nitrophthalimide with
SnCl2 in aqueous HCl, whereas diazotation/hydroxylation was
achieved via a Sandmeyer reaction (35% yield).31
The starting material for the preparation of compound 48
was obtained from the reaction of 3-aminophenol with ethyl
2-methylacetoacetate under von Pechmann conditions (see
above), whereas compound 47 was prepared using the ap-
proach of Teshima et al.50
CoMF A (3D-QSAR) Stu d y. All calculations were run on
a workstation Silicon Graphics Indy R4400 175 MHz or Origin
2000 R10000 195 MHz using the SYBYL 6.6 molecular
modeling package (Tripos Associates, St. Louis, MO). Minimal
energy conformations were obtained using the MMFF94s force
field.54 Molecular electrostatic potential (MEP) was calculated
with Gaussian 98 software55 using the 3-21G* basis set. As a
third field, the molecular lipophilicity potential (MLP) was
used.56 Statistics were performed with the QSAR module of
SYBYL. Default settings were used except for the option “drop
electrostatic”, which was set to “No”, meaning that the
electrostatic field was calculated at grid points with high steric
interactions (∆30 kcal/mol). The GOLPE 4.5 software57 (Mul-
tivariate Infometric Analysis, Perugia, Italy) was used to
examine CoMFA statistical treatments and to validate and
interpret the models. An advanced pretreatment was per-
formed on the two data sets (A1 for MAO-A and B1 for MAO-
B) for each field: steric (ste), electrostatic (mep), and lipophilic
(lip). In both cases, “zeroing” positive and negative values was
applied (ste: +0.5/-0.1; mep: +0.5/-0.1; lip: +0.2/-0.1 kcal/
mol), as well as a standard deviation cutoff (ste: 0.1; mep: 1.0;
lip: 0.1 kcal/mol). The 2-, 3- and 4-level variables were
removed; for MAO-B inhibitors, the weights were set in order
to give the same importance to each field. A grouping of
variables was then applied using the smart region definition
(SRD) algorithm: 163 seeds selected on PLS weights space,
critical distance cutoff of 1.0 Å, and collapsing distance cutoff
of 2.0 Å. The regions defined were then used in the fractional
factorial design (FFD), which was finally applied to select the
variables: a 5:1 ratio of true/dummy variables and a 2:1 ratio
of combinations/variables were adopted. The cross-validation
mode chosen was “leave-one-out” and the weights were recal-
culated after object exclusion. Noise variables were excluded
and the uncertain ones were retained.

Inhibition of MAO by Functionalized Coumarins
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25 4757
(21) Friedrich, K.; Henning, H. G. Darstellung stilbenartiger Verbin-
dungen nach der Methode von G. Wittig. Chem. Ber. 1959, 92,
2944-2949.
(22) Bruce, J . M.; Creed, D.; Dawes, K. Light-induced and related
reactions of quinones. Part VIII. Some (2-hydroxyalkyl)-, phen-
ethyl-, (2-ethoxycarbonylethyl)-, and styryl-1,4-benzoquinones.
J . Chem. Soc. C 1971, 3749-3753.
Ack n ow led gm en t. B.T. and P.A.C. are grateful to
the Swiss National Science Foundation for financial
support. A.C. and C.A. gratefully acknowledge the
financial support of the National Council of Research
(CNR-Rome, Italy).
(23) Hoefnagel, A. J .; Gunnewegh, E. A.; Downing, R. S.; van
Bekkum, H. Synthesis of 7-hydrocoumarins catalysed by solid
acid catalysts. J . Chem. Soc., Chem. Commun. 1995, 225-226.
(24) Hermodson, M. A.; Barker, W. M.; Link, K. P. Studies on the
4-hydroxycoumarins. Synthesis, of the metabolites and some
other derivatives of warfarin. J . Med. Chem. 1971, 14, 167-
169.
Su p p or tin g In for m a tion Ava ila ble: Table S1 giving all
further information concerning the spectroscopic and physi-
cochemical data of the newly synthesized compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(25) Dickenson, H. G. Cyclic lactones. U.S. Patent 2, 449, 162, Sept.
14, 1948; Chem. Abstr. 1949, 43, 694.
Refer en ces
(26) Tamura, S.; Okuma, K.; Hayashi, T. The antioxygenic property
of hydroxycoumarins. J . Agric. Chem. Soc. J pn. 1952, 26, 413-
414.
(27) J aen, J . C.; Wise, L. D.; Heffner, T. G.; Pugsley, T. A.; Meltzer,
L. T. Dopamine autoreceptor agonists as potential antipsychot-
ics. 2. (Aminoalkoxy)-4H-1-benzopyran-4-ones. J . Med. Chem.
1991, 34, 248-256.
(28) Pandit, S. S.; Sethna, S. Molecular rearrangement of some
o-acyloxyacylactones in the presence of metallic sodium. J .
Indian Chem. Soc. 1950, 27, 1-4.
(29) Shigematsu, N. Synthetic analgesics. XVI. Synthesis of 1-(2-
tertiary aminoalkyl)-3,4-dihydrocarbostyryls. Chem. Pharm.
Bull. 1961, 9, 970-975.
(30) Wyrick, S. D.; Smith, F. T.; Kemp, W. E.; Grippo, A. A. Effects
of [(N-alkyl-1,3-dioxo-1H,3H-isoindolin-5-yl)oxy]alkanoic acids,
[(N-alkyl-1-oxo-1H,3H-isoindolin-5-yl)oxy]butanoic acids, and
related derivatives on chloride influx in primary astrogial
cultures. J . Med. Chem. 1987, 30, 1798-1806.
(31) Cohen, T.; Dietz, A. G., J r.; Miser, J . R. A simple preparation of
phenols from diazonium ions via the generation and oxidation
of aryl radicals by copper salts. J . Org. Chem. 1977, 42, 2053-
2058.
(1) Bach, A. W. J .; Lan, N. C.; J ohnson, D. L.; Abell, C. W.;
Bembenek, M. E.; Kwan, S. W.; Seeburg, P. H.; Shih, J . C. cDNA
cloning of human liver monoamine oxidase A and B: molecular
basis of differences in enzymatic properties. Proc. Natl. Acad.
Sci. U.S.A. 1988, 85, 4934-4938.
(2) Grimsby, J .; Lan, N. C.; Neve, R.; Chen, K.; Shih, J . C. Tissue
distribution of human monoamine oxidase A and B mRNA. J .
Neurochem. 1990, 55, 1166-1169.
(3) Agid, Y.; Ahlskog, E.; Albanese, A.; Calne, D.; Chase, T.; De
Yebenes, J .; Factor, S.; Fahn, S.; Gershanik, O.; Goetz, C.; Koller,
W.; Kurth, M.; Lang, A.; Lees, A.; Lewitt, P.; Marsden, C. D.;
Melamed, E.; Michel, P. P.; Mizuno, Y.; Obeso, J .; Oertel, W.;
Olanow, W.; Poewe, W.; Pollak, K.; Tolosa, E. L-Dopa in the
treatment of Parkinson’s disease: a consensus meeting. Move-
ment Disorders 1999, 14, 911-913.
(4) LeWitt, P. A. New drugs for the treatment of Parkinson’s disease.
Pharmacotherapy 2000, 20, 26S-32S.
(5) Drucharch, B.; van Muiswinkel, F. L. Drug treatment of Par-
kinson’s disease. Biochem. Pharmacol. 2000, 59, 1023-1031.
(6) Walkinshaw, G.; Waters, C. M. Induction of apoptosis in cat-
echolaminergic PC12 cells by L-dopa. Implications for the
treatment of Parkinson’s disease. J . Clin. Invest. 1995, 95, 2458-
2464.
(32) MacLogP2.0; Biobyte Corp.: Claremont, CA, 1998.
(33) Hansch, C.; Leo, A. Substituent Constants for Correlation
Analysis in Chemistry and Biology; J ohn Wiley and Sons: New
York, 1979; pp 352-352.
(34) Hansch, C.; Leo, A. Exploring QSAR. Fundamentals and Ap-
plications in Chemistry and Biology; ACS Professional Reference
Book; American Chemical Society: Washington, DC, 1995; pp
1-557.
(35) van de Waterbeemd, H.; Testa, B. The parametrization of
lipophilicity and other structural properties in drug design. In
Advances in Drug Research; Testa, B., Ed.; Academic Press:
London, 1987; Vol. 16, pp 87-227.
(36) Williams, S. G.; Norrington, F. E. Determination of positional
weighting factors for the Swain and Lupton substituent con-
stants F and R. J . Am. Chem. Soc. 1976, 98, 508-516.
(37) Kneubu¨hler, S.; Thull, U.; Altomare, C.; Carta, V.; Gaillard, P.;
Carrupt, P. A.; Carotti, A.; Testa, B. Inhibition of monoamine
oxidase-B by 5H-indeno[1,2-c]pyridazine derivatives: biological
activities, quantitative structure-activity relationships (QSARs)
and 3D-QSARs. J . Med. Chem. 1995, 38, 3874-3883.
(38) Altomare, C.; Cellamare, S.; Summo, L.; Catto, M.; Carotti, A.;
Thull, U.; Carrupt, P. A.; Testa, B.; Stoeckli-Evans, H. Inhibition
of monoamine oxidase-B by condensed pyridazines and pyrim-
idines: effects of lipophilicity and structure-activity relation-
ships. J . Med. Chem. 1998, 41, 3812-3820.
(7) Malorni, W.; Giammorioli, A. M.; Matarrese, P.; Pietrangeli, P.;
Agostinelli, E.; Ciaccio, A.; Grassili, E.; Mondovi, B. Protection
against aptotosis by monoamine oxidase A inhibitors. FEBS Lett.
1998, 426, 155-159.
(8) Harfenist, M.; Heuser, D. J .; J oyner, C. T.; Batchelor, J . F.;
White, H. L. Selective inhibitors of monoamine oxidase. 3.
Structure-activity relationship of tricyclic bearing imidazoline,
oxadiazole, or tetrazole groups. J . Med. Chem. 1996, 39, 1857-
1863.
(9) Forna, L. S. Neuropathology of Parkinson’s disease. J . Neuro-
pathol. Exp. Neutol. 1996, 55, 259-272.
(10) Wouters, J . Structural aspects of monoamine oxidase and its
reversible inhibition. Curr. Med. Chem. 1998, 5, 137-162.
(11) Kalgutkar, A. S.; Castagnoli, N., J r.; Testa, B. Selective inhibitors
of monoamine oxidase (MAO-A and MAO-B) as probes of its
catalytic site and mechanism. Med. Res. Rev. 1995, 15, 325-
388.
(12) Efange, S. M. N.; Boudreau, R. J . Molecular determinants in
the bioactivation of the dopaminergic neurotoxin N-methyl-4-
phenyl-1,2,3,6-tetrahydropyridine (MPTP). J . Comput.-Aided
Mol. Des. 1991, 5, 405-417.
(13) Mabic, S.; Castagnoli, N., J r. Assessment of structural require-
ments for the monoamine oxidase-B-catalyzed oxidation of 1,4-
disubstituted-1,2,3,6-tetrahydropyridine derivatives related to
the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahyropyridine. J .
Med. Chem. 1996, 39, 3694-3700.
(14) Hoult, J . R. S.; Payd, M. Pharmacological and biochemical
actions of simple coumarins: natural products with therapeutic
potential. Gen. Pharmacol. 1996, 27, 713-722.
(15) Hossain, C. F.; Okuyama, E.; Yamazaki, M. A new series of
coumarin derivatives having monoamine oxidase inhibitory
activity from Monascus anka. Chem. Pharm. Bull. 1996, 44,
1535-1539.
(16) Rendenbach-Mu¨ller, B.; Schlecker, R.; Traut, M.; Weifenbach,
H. Synthesis of coumarins as subtype-selective inhibitors of
monoamine oxidase. Bioorg. Med. Chem. Lett. 1994, 4, 1195-
1198.
(17) Rendenbach, B.; Weifenbach, H.; Teschendorf, H. J . Arylalkoxy-
cumarine. Verfahren zu ihre Herstellung und diese enthalende
therapeutische Mittel. Patent DE 3834861 A1, 1990 (BASF AG).
(18) Schlecker, R.; Schmidt, P.; Thieme, P. C.; Lenke, D.; Teschendorf,
H. J .; Traut, M.; Mueller, C. D.; Hofmann, J . P.; Kreiskott, H.
Neue Sulfonsa¨ureester von Hydroxycumarinen, ihre Herstellung
und sie enthaltende Arzneimittel. Patent DE 3243158 A1, 1984
(BASF AG).
(39) Cramer, R.D., III; Patterson, D. E.; Bunce, J . D. Comparative
molecular field analysis (CoMFA). 1. Effect of shape on binding
of steroids to carrier proteins. J . Am. Chem. Soc. 1988, 110,
5959-5967.
(40) Kaliszan, R.; Kaliszan, A.; Wainer, I. W. Deactivated hydrocar-
bonaceous silica and immobilized artificial membrane stationary
phases in high-performance liquid chromatographic determina-
tion of hydrophobicities of organic bases: relationship to log P
and CLOGP. J . Pharm. Biomed. Anal. 1993, 11, 505-511.
(41) Kim, K. H. Comparison of classical and 3D QSAR. In 3D QSAR
in Drug Design. Theory Methods and Applications; Kubinyi, H.,
Ed.; ESCOM Science Publishers: Leiden, 1993; pp 619-642.
(42) Pace, C. N.; Heinemann, U.; Hahn, U.; Saenger, W. Ribonuclease
T1: structure, function, and stability. Angew. Chem., Int. Ed.
Engl. 1991, 30, 343-360.
(43) Atkins, R. L.; Bliss, D. E. Substituted coumarins and azacou-
marins. Synthesis and fluorescent properties. J . Org. Chem.
1978, 43, 1975-1980.
(44) Antus, S.; Gottsegen, A.; Nogradi, M. Unusual regioselectivity
in the reduction of R,â-unsaturated carbonyl compounds with
diisobutylaluminium hydride (DIBAH): direct conversion of
isoflavones to iso-flavan-4-ones. Synthesis 1981, 7, 574-576.
(19) van Pechmann, H. Duisberg, C. U¨ ber die verbindungen der
Phenole mit Acetessiga¨ther. Chem. Ber. 1883, 16, 2119-2128.
(20) Sethna, S.; Phadke, R. The Pechmann reaction. Org. React. 1953,
7, 1-58.

4758 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 25
Gnerre et al.
(45) Mahal, H. S.; Rai, H. S.; Venkataraman, K. Synthetical experi-
ments in the chromone group. Part XVI. Chalcones and fla-
vanones and their oxidation to flavones by means of selenium
dioxide. J . Chem. Soc. 1935, 866-868.
(46) Bhalla, D. C.; Mahal, H. S. Synthetical experiments in the
chromone group. Part XVII. Further observations on the action
of sodamide on o-acyloxyacetophenones. J . Chem. Soc. 1935,
868-870.
(53) Albert, A.; Serjeant, E. P. Determination of ionization constants
by spectrophotometry. The Determination of Ionization Con-
stants. A Laboratory Manual; Chapman and Hall: London, 1984;
pp 70-95.
(54) Halgren, T. A. Merck molecular force field. II. MMFF94 van der
Waals and electrostatic parameters for intermolecular interac-
tions. J . Comput. Chem. 1996, 17, 520-552.
(55) Frish, M. J .; Trucks, G. W.; Schelgel, H. B.; Scuseria, M. A.; Robb,
M. A.; Cheeseman, J . R.; Zahrzewski, V. G.; Montgomery, J . A.;
Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi,
J .; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.;
Adamo, C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala,
P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J .; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe,
M.; Gill, P. M. W.; J ohnson, B. G.; Chen, W.; Wong, M. W.;
Andres, J . L.; Head-Gordon, M.; Replogle, E. S.; Pople, J . A.
GAUSSIAN 98 software, revision A.1; Gaussian, Inc.: Pitts-
burgh, PA, 1998.
(47) Davies, J . S. H.; McCrea, P. A.; Norris, W. L.; Ramage, G. L.
Furanochromones. Part IV. Synthesis of 2-(methylfurano)-3′,2′-
(6,7)-chromone and derivatives thereof. J . Chem. Soc. 1950,
3206-3213.
(48) Vaughan, W. R.; Baird, S. L., J r. The preparation of some
phthalazines and related substances. J . Am. Chem. Soc. 1946,
68, 1314-1316.
(49) Beroza, M. 3,4-Methylenedioxyphenoxy compounds as synergists
for natural and synthetic pyrethrins. J . Agric. Food Chem. 1956,
4, 49-53.
(50) Teshima, T.; Nakaji, A.; Shiba, T.; Tsai, L.; Yamazaki, S.
Elucidation of stereospecificity of a selenium-containing hydro-
genase from Metanococcus Vannielii. Synthesis of (R)- and (S)-
[4-2H1]-3,4-dihydro-7-hydroxy-1-hydroxyethylquinolinone. Tet-
rahedron Lett. 1998, 26, 351-354.
(51) Thull, U.; Kneubu¨hler, S.; Gaillard, P.; Carrupt, P. A.; Testa,
B.; Altomare, C.; Carotti, A.; J enner, P.; McNaught, K. S. P.
Inhibition of monoamine oxidase by isoquinoline derivatives:
qualitative and 3D-quantitative structure-activity relationships.
Biochem. Pharmacol. 1995, 50, 869-877.
(52) Thull, U.; Kneubu¨hler, S.; Testa, B.; Borges, M. F. M.; Pinto,
M. M. M. Substituted xanthones as selective and reversible
monoamine oxidase A (MAO-A) inhibitors. Pharm. Res. 1993,
10, 1187-1190.
(56) Gaillard, P.; Carrupt, P. A.; Testa, B.; Boudon, A. Molecular
lipophilicity potential, a tool in 3D-QSAR. Method and applica-
tions. J . Comput.-Aided Mol. Des. 1994, 8, 83-96.
(57) GOLPE 4.5; Multivariate Infometric Analysis: Perugia, Italy,
1999.
J M001028O