Selective inhibitors of monoamine oxidase. 4. SAR of tricyclic N- methylcarboxamides and congeners binding at the tricyclics' hydrophilic binding site

Article abstract of DOI:10.1021/jm9608063

Linear [6.6.6] tricyclic moieties whose center ring is made of two atoms of differing size (here primarily thioxanth-9-ones and phenoxathiins) monosubstituted meta to the sulfur by C(O)NHMe include potent and selective inhibitors of monoamine oxidase A. Similarities with effects on SAR of acylamide and of diazapentacyclic substitution on such rings, including positional variables, the requirement for monomethylation (primary and dialkylated amides are inactive and higher monoalkylated amides show little or no potency), and that sulfur is optimally in sulfone form, suggest that binding to the enzyme occurs similarly in each series. No significantly greater rise in blood pressure was found in rats given sufficient 8 to inhibit most brain and liver MAO A and then followed by oral tyramine than was found on administration of tyramine to controls. This is in contrast to a large blood pressure rise in rats pretreated with phenelzine followed by tyramine, and in accord with the belief that an inhibitor selective for MAO A which is reversibly bound to the enzyme and therefore displaced by any ingested tyramine will not lead to the 'cheese effect' (hypertension during treatment with MAO inhibitors usually caused by ingestion of foods containing tyramine).

Full text of DOI:10.1021/jm9608063

2466
J . Med. Chem. 1997, 40, 2466-2473
Articles
Selective In h ibitor s of Mon oa m in e Oxid a se. 4.1 SAR of Tr icyclic
N-Meth ylca r boxa m id es a n d Con gen er s Bin d in g a t th e Tr icyclics’ Hyd r op h ilic
Bin d in g Site
Morton Harfenist,*,†,‡ Diane M. J oseph,† Sharon C. Spence,† Daniel P. C. Mcgee,† Mark D. Reeves,† and
Helen L. White§
Departments of Organic Chemistry and Pharmacology, The Wellcome Research Laboratories,
Research Triangle Park, North Carolina 27709
Received November 22, 1996X
Linear [6.6.6] tricyclic moieties whose center ring is made of two atoms of differing size (here
primarily thioxanth-9-ones and phenoxathiins) monosubstituted meta to the sulfur by
C(O)NHMe include potent and selective inhibitors of monoamine oxidase A. Similarities with
effects on SAR of acylamide and of diazapentacyclic substitution on such rings, including
positional variables, the requirement for monomethylation (primary and dialkylated amides
are inactive and higher monoalkylated amides show little or no potency), and that sulfur is
optimally in sulfone form, suggest that binding to the enzyme occurs similarly in each series.
No significantly greater rise in blood pressure was found in rats given sufficient 8 to inhibit
most brain and liver MAO A and then followed by oral tyramine than was found on
administration of tyramine to controls. This is in contrast to a large blood pressure rise in
rats pretreated with phenelzine followed by tyramine, and in accord with the belief that an
inhibitor selective for MAO A which is reversibly bound to the enzyme and therefore displaced
by any ingested tyramine will not lead to the “cheese effect” (hypertension during treatment
with MAO inhibitors usually caused by ingestion of foods containing tyramine).
Monoamine oxidase (MAO) (EC 1.4.3.4 amine oxidase,
flavin containing) consists of two forms distinguishable
by their substrate specificities2 and their amino acid
sequences.3 Before this difference was recognized,
inhibition of “MAO” was found to lead to antidepressant
activity, and nonspecific inhibitors such as phenelzine,
which are irreversibly bound to both forms of MAO,
were clinically used as antidepressants. However, some
patients developed hypertension (“cheese effect”), so
clinical use of such MAO inhibitors as antidepressants
was essentially abandoned. Subsequent research4 in-
dicated that the cheese effect was usually due to
ingestion of tyramine, found especially in fermented
foods such as cheeses where it is produced by the action
of microorganisms on milk tyrosine, which was not
destroyed by the inhibited MAO. Tyramine is a sub-
strate of both forms of MAO. Therefore, when it was
realized that inhibition of MAO A was responsible for
the antidepressant effect,5 it seemed logical to look for
an inhibitor selective for MAO A, retaining the MAO B
activity to destroy any ingested tyramine. As a further
precaution, it would be desirable that the inhibitor be
bound to the MAO A reversibly and competitively with
tyramine, so that ingestion of a large amount of tyramine
would lead to displacement of inhibitor from the MAO
A-inhibitor complex, reactivating the MAO A to oxidize
tyramine. Such a selective, reversibly bound, and
tyramine-displaceable inhibitor has been our goal.
Previous publications from our laboratory1,6 have shown
that the set of tricyclic compounds with the two outer
rings aromatic and the center ring made up of two
functions differing in size and with a hydrophilic group
meta to the larger central function (termed “the hydro-
philic binding site”) contains many potent and selective
MAO A inhibitors. The hydrophilic group reported
initially was the N-acylamino function.6 While this
allowed for the ready synthesis of a variety of positional
isomers on many tricyclic systems, the compounds 1
could not be exploited clinically because of their rela-
tionship to known N-arylamide carcinogens.6 Studies
were therefore instituted to learn whether other, me-
dicinally more acceptable hydrophilic groups could
replace the acylamino function and, if so, whether the
MAO inhibitory activity of the resulting compounds
followed the same structure to potency relationships
(SAR) found for the N-arylacylamides 1. A variety of
five-membered rings containing two or more hetero-
cycles (2a -d ) were the first potential hydrophilic re-
placements for acylamino groups that were studied.1
The SAR for potency and selectivity in inhibiting MAO
A as a function of tricyclic moiety, position of substitu-
tion with respect to the middle ring larger group, and
additional effect of other groups in the outer ring not
substituted by the hydrophilic function were similar to
the effects found earlier for the compounds1. Unfortu-
nately, the seemingly best compounds of the series 2
either showed unsatisfactory pharmacokinetic proper-
ties or toxicity on chronic administration to rats. How-
ever, the similarity of SAR between these two series,
and the appreciable number of compounds showing
nanomolar potencies and high selectivity, made it seem
worth while to search for an alternative hydrophilic
† Division of Organic Chemistry.
‡ Current address: 6406C Farrington Rd, Chapel Hill, NC 27514.
§ Division of Pharmacology. Deceased March, 17 1997.
X Abstract published in Advance ACS Abstracts, J uly 1, 1997.
S0022-2623(96)00806-0 CCC: $14.00 © 1997 American Chemical Society

Selective Inhibitors of Monoamine Oxidase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16 2467
Of greater interest is the major thrust of this report,
the MAO inhibition by tricyclics substituted by the
hydrophilic carboxamide group. A qualitative cor-
respondence between the results reported in this paper
and those reported for the tricyclic N-arylacylamides6
and for the tricyclics substituted by imidazolines and
their congeners1 earlier is clear. This includes the
effects on potency both of the position of the substituent
on the tricyclic system with respect to the central
groups, i.e. the greatest potency is conferred by a
hydrophilic substituent on an outer ring meta to the
larger central ring group (most usually here SO2) and
para to the smaller, and of similarity of the effects of
generalized structural details. Thus the thioxanthone
dioxide 8, which has its CONHMe meta to the large
central SO2 group, has an IC50 of 60 nM and is selective
for MAO A, while 6, which has its CONHMe para to
the SO2, shows only 28% inhibition of MAO A at 1000
nM. The loss of specificity with compounds of low
potency which is shown here for 6 has been mentioned
in our earlier papers. A similar effect of substituent
position with respect to the central SO2 group is seen
when potencies of the corresponding phenoxathiin di-
oxides are compared. Thus 59 (N-methylcarboxamide
function meta to sulfone, IC50 ) 30 nM) is far more
potent than 53 (N-methylcarboxamide ortho to sulfone,
inactive) or with 62 (N-methylcarboxamide para to
sulfone, IC50 ) 600 nM). Incidentally, while the in-
creased potency in going from the sulfide 7 to the sulfone
8 is evident here as expected, 58 and 59 appear to invert
this order. We can give no explanation of this.
group to substitute into the same tricyclic systems.
Derivatives likely to be metabolized into aromatic
carboxylic acids were selected for study, since such acids
tend to be nontoxic.
Resu lts a n d Discu ssion
Some confusion has been induced in comparisons of
SAR for substitution in differing tricyclics caused by
historic nomenclature requiring differing locants for
identical positions of substitution with respect to the
central ring components. The numbering used for the
xanthone and the thioxanthone and thioxanthene sys-
tems is the historical system shown at the heading to
Tables 1 and 4. Phenothiazines, as shown in the
heading of Table 3, also have their own numbering
system. The other tricyclics are numbered from the
outer ring carbon ortho to the central element of higher
atomic number, as shown for phenoxathiin in the
heading for Table 2.
The tables show inhibition of rat brain MAO A and
B obtained with tricyclic systems substituted by car-
boxylic acid functions or simple derivatives of these. In
addition, a recent report7 that some unsubstituted
tricyclics had IC50 values in the micromolar range for,
and were specific inhibitors of, MAO A led us to tabulate
some of our results with unsubstituted and with simple
halogenated derivatives of the tricyclics whose hydro-
philic substituted derivatives are the major subject of
this report. Subject to the uncertainties possibly due
to the extremely low solubilities in aqueous systems of
these compounds with no polar substituents, the results
seem to follow one generality reported previously:
where one component of the central ring is sulfur, ring
sulfur in the sulfone form markedly increases potency
compared to unoxidized or sulfoxide sulfur. Thus while
only a little difference in the low potencies reported can
be seen between the unsubstituted thioxanthone 3,
which has the sulfur in the unoxidized form, and the
sulfone 4, the unsubstituted phenoxathiins show dif-
ferences analogous to those reported in our earlier work
for the three degrees of oxidation of heterocyclic sulfur,
with potencies increasing from the sulfoxide 49 through
the sulfide 48 to the sulfone 50. This last showed the
quite respectable IC50 value of 50 nM. Results compat-
ible with this progression can be seen in the 2-chloro-
thioxanthenes of Table 4, where the sulfone 84 has
appreciable though not useful potency and selectivity,
while the corresponding sulfide 81 and sulfoxide 83 have
no detectible activity.
Another point of close similarity between the present
series and the N-arylacylamide series6 is the sharp
maximum found in this series with the N- monomethyl-
ated amides. The C(O)NHMe substituent which confers
maximum potency in the results reported here corre-
sponds to the NHC(O)Me reported previously6 to lead
to maximum potency in the N-arylamide series. In the
N-arylamides, maximum potency was found for N-
acetylamino-substituted tricyclics, with the lower (form-
amido) and higher homologs less active or inactive.
Similarly, as shown in this present report, the N-
methylcarboxamides are far more potent than the
otherwise identical primary amides ArC(O)NH2 (com-
pare 30 to 32 or 34 to 35 in the thioxanthone dioxide
series and 57 to 59 in the phenoxathiin dioxide series.
Also, as in the earlier N-arylacylamide paper, replacing
the pendant methyl in this work with larger R groups
to give ArC(O)NHR where R is Et or higher to give, for
example, 10, 11, 12 and higher thioxanthone dioxides
of Table 1 compared to 8, led to partial or complete loss
of activity at the concentrations of interest. Similar
results were found in the phenoxathiin dioxide series,
where not even the relatively small cyclopropyl function
found in 66 could satisfactorily replace the methyl of
62.
Other substitutions eliminating appreciable activity
include dialkylation of the amide N not only by simple
alkyl groups, as found for 20, 69, and 70 among others,
but also by the N-methyl (16) and O-methyl (17)
hydroxylamine amides. It is interesting that the N-
methylamidine 46, which is a nitrogen analog of a
methylamide, has substantial potency, whereas the
N,N′-dimethylamidine 47 has little, and that the N-meth-
yl-3-thioamide 45 has little activity, in contrast to its

2468 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16
Harfenist et al.
amide analog 8. No further studies of amidines were
carried out because amidines are known generally to
be poorly absorbed in vivo.
are considered markers of possible liver damage, so
further work with this compound was discontinued.
Parenthetically, the reduction product of 8, the alcohol
88, was found to be to be a major metabolite of 8. As
anticipated, the reduction of 8 to 88 was reversible and
was a function of pH in vitro, favoring 8 under alkaline
conditions.
Comparison of the effect of methyl groups on potency
in this present methylamide series with results reported
for the tricyclics whose hydrophilic function was a five-
membered ring with two or more heteroatoms1 turns
out not to be a simple matter. The effect for alkyl
groups greater than methyl is the same in all three
series, N-arylacylamino, five-membered ring, and N-
alkylarylcarboxamide, causing diminution or loss of
activity for all. However, methylation of either the
carbon or a nitrogen of the 2-arylimidazolines destroyed
activity, unlike the effect of methyl substitution in the
present and earlier series, while carbon methylation in
the 1,2,4-oxadiazole and in the 1,3,4-oxadiazole series
led to very potent and specific inhibitors. The 5-aryl-
tetrazoles in the previous paper are an even more
difficult case for comparison, since with these 2-methyl-
ation led to potent inhibitors while 1-methylation gave
far less potent compounds. The simplest conclusion that
can be suggested is that the steric requirement for
binding to the enzyme at the position meta to the larger
central group (in all these examples the SO2 function)
is very sensitive to small changes in the relatively rigid
5-membered rings.
Exp er im en ta l Section
Ch em istr y. Melting points below 305 °C were determined
using an electrically heated oil bath, and those above that
temperature by using a block (MEL-TEMP Laboratory De-
vices). All melting points are uncorrected. Unless otherwise
is stated, all compounds showed
a single spot on UV-
fluorescent silica plates (MK6F, Whatman International, Ltd.).
The solvents used were EtOAc-hexanes or CH2Cl2-hexanes
unless otherwise specified. Reactions in those few cases in
which products could not be distinguished by TLC from
starting materials or from simultaneously produced isomers
were followed by NMR. Products were identified by elemental
analysis and NMR. Preparative conditions were not opti-
mized.
Preparation of the 9-oxothioxanthene-3-carboxylic acid amides
of Table 1 proceeded from the corresponding acid chloride
which was made from 2-nitroterephthalonitrile and the re-
quired aryl thiol by the procedure outlined in our previous
paper1 and the reference cited there. Amides of lower amines
such as methylamine were made by cautiously adding the neat
acid chloride to a stirred large excess of cooled concentrated
preferably aqueous solution of amine. The resulting slurry
could be filtered and the solid water washed free of a few
percent of alkylammonium salt of the acid and was then
essentially pure after drying. Especially the methylamides
made in this way frequently were appreciably soluble in excess
aqueous methylamine, so it was often expedient to concentrate
either the unfiltered reaction or the first filtrate at reduced
pressure (water aspirator) to remove most of the amine and
then filter, or refilter, the aqueous slurry to recover the
resulting solid amide. A small amount of starting acid could
be recovered by acidification of the filtrate. Methylamides
could also be made from the methyl esters and dry methyl-
amine by methoxide-catalyzed reaction in dry methanol.
Amides of higher amines were best made from the acid
chlorides and 2 equiv of amine for readily available amines,
or using a small excess over 1 equiv of the reacting amine and
1 equiv of a tertiary amine such as triethylamine to scavenge
the HCl. The preferred solvent was dry ethyl ether, and the
amine hydrochloride almost always precipitated in filterable
form. Alternatively the amine hydrochloride and any excess
amine were extracted with 0.1 N or more dilute aqueous HCl
and the amide recovered from its ethereal solution generally
by evaporation of the ether but occasionally by cooling the
solution to cause the amide to crystallize. An example of the
preparation of 8 modified from ref 9 is given below. The
N-methylthioamide (45) was made uneventfully from the
corresponding amide (8) without obvious involvement of the
9-keto group, using Lawesson’s reagent10 in xylene. The same
method was also used to prepare the nonmethylated thioamide
corresponding to 8 (not tabulated) in poorer yield. Nitrile, the
obvious side product, was not sought. Both 8 and 45 were
converted by (EtO)3BF4 in toluene to O-ethyl and S-ethyl salts,
respectively, and each of those added to NH3 gave 46 or, added
to MeNH2, gave 47.
The similarities in the effects of substituent position,
size, and hydrophilicity on activity support the idea that
the differing hydrophilic groups cause interaction be-
tween the inhibitors and the enzyme to occur with
essentially identical geometries. This raises the pos-
sibility of using inhibitors based on those reported here
but modified with a reactive group which can bind
covalently to nearby amino acids of the enzyme. Such
reactive groups might be placed in each in turn of
several positions on the tricyclic system to explore the
composition of the enzyme docking site of these inhibi-
tors. Together with similar use of other specific binding
sites on tricyclics which we hope to report on in future
publications, a fairly broad range of probing might be
available.
Compound 8 was selected for further study8 aimed
at its development as a potential antidepressant/anxio-
lytic. The dialysis of MAO A in rat brain homogenates
which had been inhibited by 8 led to complete return of
enzyme activity. As anticipated from this and the
previous discussion, when conscious, unrestrained rats
pretreated with 8 were given oral tyramine, the rise in
blood pressure was within significance of control values
up to 90 mg/kg of tyramine, while phenelzine used as
positive control at equipotent MAO A inhibition in rat
brain (phenelzine is approximately twice as potent as
8 in our tests) gave blood pressure increases 3-10-fold
greater. Activity was shown by 8 in standard anti-
depressant models in rats, but no activities in a large
variety of other pharmacological tests and models at
high multiples of the dose of 50 mg/kg in Sprague
Dawley male rats which gives 80% inhibition of rat
brain MAO A and ca. 90% inhibition of rat liver MAO
A. Details of methods and results have been published.8
Unfortunately, when humans were fed large multiples
of the anticipated therapeutic dose of 8 in early clinical
trials, a small proportion showed an increase above
pretreatment levels in their serum aminotransferases
(“liver enzymes”). Increased levels of these in serum
The phenoxathiins of Table 2 were generally made (Scheme
1) by condensation of an o-hydroxythiophenol with an o-
nitrohalobenzene, usually in a one-pot reaction without isola-
tion of intermediate diaryl sulfide. Normally, anhydrous DMF
was used as solvent with a base of low nucleophilicity such as
KO-t-Bu or KH. Smiles rearrangement was never noted. A
preparation of 3-bromophenoxathiin by this route, oxidation
of that to its 10,10-dioxide, and conversion of that by the
modified von Braun reaction (CuCN in DMF) to nitrile 60 is
given below.
Preparation of phenoxathiins by sulfur fusion of diphenyl
ethers with Friedel-Craft catalysts, while the method of choice

Selective Inhibitors of Monoamine Oxidase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16 2469
Ta ble 1. MAO Inhibition by Thioxanthones
IC50 (µM) or % I at µM
no.
substituent
n
MAO A
MAO B
formula
anal.
mp (°C)
recrysta
3b
4c
0
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
Ca 0.6
Ca 1
1
28% at 1
0.5
0.06
27% at 10
12% at 10
5% at 1
32% at 1
48% at 1
4% at 1
0.6
9% at 1
6% at 1
34% at 1
0.7% at 1
41% at 1
8% at 1
0% at 0.1
5d 2-Br
6
7
2-CONHMe
3-CONHMe
C
C
15H11NO4S
15H11NO2S
C,H,Nb
C,H,N,S 235-236
242-243
A-W
A-W
8e 3-CONHMe
9
10
11
12
13
14
15
16
17
18
2,6-(CONHMe)2
3-CONHEt
0.05
C17H14O5N2S
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
321-322
186-186.4
175
196.5
217
191-192
199
159-161
218-220
170
HOAc
IPO
HOAc-W
HOAc
A
T-SOCl2
A
A-H
A-D
A
Ca 0.5
11% at 1
18% at 1
0% at 1
-6% at 1
18% at 1
0% at 0.1
C
C
C
16H13NO4S
18H17NO4S
18H17NO4S
3-CONHCHMeEt
3-CONHt-Bu
3-CONHC2H4OH
3-CONHCMe2CH2Cl
3-CONCMe2CH2OH
3-CONOHMe
C16H13NO5S
C18H16ClNO4S
C18H17NO5S
C
C
15H11NO5S
15H11NO5S
3-CONHOMe
3-CONHC3H6NMe2
15% at 0.1 0% at 0.1
2% at <0.5 4% at <0.5 C19H20N2O4S‚HCl
0.5 H2O
19
3-CONHCH2C6H2-
3,4,5(OMe)3
2
0.5 at 1
9% at 1
C23H19NO7S
C,H,N
235
A
20
21
3-CONEt2
2
2
17% at 1
0 at 1
0 at 1
0 at 1
C16H13NO4S
C21H21NO5S
C,H,N
C,H,N
189
186
A
HOAc
3–CON
C2H4OH
3–CON
22
23
24
25
2
2
2
0
72% at 1
7% at 0.1
2% at 1
51% at 1
C18H15NO4S
C,H,N
C,H,N
C,H,N
169
A
17% at 0.1 C20H21N2O4SI
322
W
+
3–CON
3–CON
3–CON
NMe2I –
12% at 1
C19H18N2O4S
198-202.5
345
A-W or EA
A-W
N–Me
7% at 10
10% at 10
C20H20N2O2S‚HCl‚0.5H2O C,H,N
NH•HCl
26
27
28
29
30
31
5-Me-3-CONHMe
5-Et-3-CONHMe
7-Me-3-CONHMe
7-i-Pr-3-CN
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0.05
44% at 1
0.008
60% at 0.3 5% at 0.3
0.7 14% at 1
59% at 0.1 14% at 0.1
0.006
64% at 1
0.45
0.002
29% at 1
0.01
0.13
0.3
7% at 1
5% at 1
18% at 1
C
C
C
C
16H13NO4S
17H15NO4S
16H13NO4S
17H13NO3S
C,N,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
232-233.5
169.5
251.5-254
177
A-W
A-W
HOAc
HOAc
HOAc
HOAc
HOAc-W
A
7-i-Pr-3-CONH2
7-i-Pr-3-CONHMe
C17H15NO4S
18H17NO2S
15% at 0.1 C18H17NO4S
19% at 1
11% at 1
4% at 0.03
3% at 1
0at 0.1
27% at 1
5% at 1
235
C
185-186
203-205
320-322.5
240-241
193
32e 7-IsoPr-3-CONHMe
33
34
35
36
37
38
39
40
41
42
43
44
45
7-OH-3-CONHMe
7-PrO-3-CONH2
C
15H11NO5S
C17H15NO5S
18H17NO5S
D-A-W
A-D-W
M-Ac-W
M-W
7-PrO-3-CONHMe
C
7-MeCHOAc-3-CONH2‚1/2H2O
7-MeCHOH-3-CONHMe
7-OAc-3-CONHMe
C18H15NO6S‚0.5H2O
C
C
C18H15NO7S
C20H19NO8S
C17H16N2O4S
C17H15NO4S
C16H13NO2S
C18H17NO4S
C,H,N,S 205-207
C,H,N,S 193-196
17H15NO5S
17H13NO6S‚H2O
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
230-231
140.2-140.7 A-W
140-141
213-215
239-241
292.5
A-W
7-OCH2COOMe-3-CONHMe
7-OCH2COOC2H4OMe
7-NMe2-3-CONHMe
5,7-Me2-3-CONHMe
2,4-Me2-3-CONH2
Ca. 1
0.03
0.02
8% at 0.1
7% at 0.1
3% at 0.1
0 at 0.1
MOX-W
A-D-W
A-W
A
0 at 0.1
0 at 0.1
38% at 0.1 0 at 0.1
0.06
21% at 1
2,4-Me2-3-CONMe2
3-C(S)NHMe
0 at 0.1
189-191
EA-H
E-W
C
15H11NO3S2
C15H12N2O3S‚HCl
16H14N2O3S
C,H,N,S 221.2
C,H,N,S 315.3-317.3 A, EA, E
C,H,N
46f 3-C(dNH)NHMe‚HCl
47
12% at 1
4% at 1
3-C(dNMe)NHMe
C
221-223
A
a
Recrystallization solvents: A ) ethanol; Ac ) acetone; C ) chloroform; D ) DMF; E ) ether; EA ) ethyl acetate; H ) hexanes; HOAc
b
) acetic acid; IPO ) 2-propanol; M ) methanol; MOX ) 2-methoxyethanol; P ) pentane; T ) toluene; W ) water. Aldrich Chemical Co.
c Beaulieu, F.; Snieckus, V. Directed Metallation of Diarylsulfone 2-Amides and 2-O-Carbamates. Regiospecific General Route to
Thioxanthen-9-one 10,10-Dioxides via Anionic Friedel-Crafts and Remote Fries Rearrangement Equivalents. J . Org. Chem. 1994, 59,
d
6508-6509. J ayamma, V.; Badiger, V. V.; Nargund, K. S. Substituted Thioxanthones. J . Karnatak Univ. 1967, 12, 49; Chem. Abstr.
1968, 69, 106456t. e Harfenist, M.; J oyner, C. T.; Heuser, D. J . Tricyclic Compounds, Compositions Containing Them, and Their Use in
Medicine. Eur. Pat. Appl. EP 150, 891. f Compound 47‚HCl is as tabulated. 47‚HBF4 : C,H,N. Mp: 228.7 °C.
for phenoxathiin itself, was not applicable to the congeners
that our work required. This presumably is because of
nonreactivity or differing reactivity of the substituted rings.
Although sulfur fusion of 4,4′-dibromodiphenyl ether was
ineffective, as reported in the literature, the related reaction
with molar amounts of AlCl3 and SOCl2 in CH2Cl2 gave an
appreciable yield of a mixture of the expected 2,8-dibromo-
phenoxathiin 10-oxide and the corresponding phenoxathiin.

2470 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16
Ta ble 2. MAO Inhibition by Phenoxathiins
Harfenist et al.
IC50 (µM) or % I at µM
no.
substituent
n
MAO A
MAO B
formula
anal.
mp (°C)
recrysta
HOAc
48b
49c
50d
51e
52
0
1
2
2
2
2
0
2
2
2
0
2
2
2
2
0
0
2
2
2
2
2
2
0
0
2
2
2
2
0.3
24% at 1
0.05
13% at 1
8% at 1
4% at 1
0.4
-2% at 0.3
7% at 1
3% at 1
5% at 1
15% at 1
7% at 1
14% at 1
0.7% at 0.3
11% at 1
11% at 1
15% at 1
-14% at 0.01
2.5
11% at 1
3.0
8% at 10
8% at 10
4% at 1
10% at 0.1
8% at 0.1
11% at 0.1
10% at 0.1
11% at 0.1
8% at 10
0 at 0.1
18% at 1
0 at 0.1
5% at 1
1-COOH
1-CONH2
1-CONHMe
2-CN
C13H8O5S‚0.15H2O
C13H9NO4S
14H11NO4S
C,H,N
C,H,N
C,H,N
205-207
207.5-208.5
244-247
A-D-W
A
53
C
54f
55
2-CN
0.2
C
13H7NO3S
C,H,N
253-254.8
A-W
56g
57h
58
2-COOH
2-CONH2
2-CONHMe
2-CONHMe
3-CN
3-COOH
3-CONHMe
4-COOH
4-CONHMe
1-aza-7-CONHMe
3-CONH-cyclo-Pr‚0.2H2O
3-CONHBu
3-C(O)NHC2H4NHAc
3-CONMe2
3-CONEt2
4-COOH
9% at 1
45% at 1
0.02
0.03
0.06
C13H9NO4S
14H11NO2S‚0.5H2O
14H11NO4S
13H7NO3S
13H8O5S‚0.3H2O
14H11NO4S
C13H8O3S
14H11NO2S
13H10N2O4S
C16H13NO4S
17H17NO4S
C17H16N2O5S
C15H13NO4S
C17H17NO4S
13H8O3S
C17H17NO2S
14H11NO5S
14H10O4S
C14H10O4S
16H12O5S
C,H,N
C,H,N,S
C,H,N
C,H,N,S
C,H,N
C,H,N,S
C,H,N,S
C,H,N
290-292
102-103
253-257
225-227
261-265
206-207
165-168
122-124
267-270
185-187
135-137
235
173-175
105-108
165-168
85-87
162-164
142-144
169
A-D-W
MeOH
A-D-W
EA-H
IPO-W
EA-H
HOAc
EA
C
C
C
C
C
59
60
61
9% at 1
0.6
62
63i
64
32% at 10
13% at 10
2% at 1
NS at 0.1
14% at 0.1
0.3
0 at 0.1
5% at 0.1
32% at 10
0 at 0.1
0.2
C
C
65
66
67
68
69
70
71
72
C,H,N
A-D-W
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N,S
C,H,N
C,H,N
C,H
C,H
C,H
A-W
C
EA-H
EA
EA-H
EA-H
HOAc
EA-P
A
EA-P
EA-H
A
C
4-CONEt2
2-OC(O)NHMe
1-Ac
73
C
C
74
85% at 0.1
52% at 1
0.5
75j
76
2-Ac
2,7-DiAc
2
C
223-225
a
See footnote a of Table 1. b Parish Chemical Co. c Folsom, H. E.; Castrillo´n, J . The Controlled Oxidation of Organic Sulfides to Sulfoxides
d
with the Help of o-Iodosylbenzoic Acid. Synth. Commun. 1992, 22, 1799-1806. Kemp, D. S.; Buckler, D. R. Synthesis of 1-Methoxy-9-
Mercaptopheroxathiin and the Resolved 1-((N-Benzyloxycarbonyl)-l-Alanyl)oxy-9-(methoxycarbonyl)dithio)phenoxathiin 10-oxide Dia-
stereomers. Comments on Improved Methods for Sulfone Reduction. J . Org. Chem. 1989, 54, 3647-3651. e Shirley, D. A.; Lehto, E. A.
The Metallation of Phenoxathiin 10-oxide and 10,10-Dioxide with n-Butyl Lithium. J . Am. Chem. Soc. 1955, 77, 1841-1843. f Batchelor,
g
J . F.; Gorvin, J . H. Sulfur-containing Tricyclic Compounds. Ger. Offen. 2,344,800 1974. Vasiliu, G.; Gloaba, A.; Maior, O. Condensed
Heterocromatic Systems. Effect of Pyrocatechol Dichloromethylene Ether [2,2-Dichloro-1,3-benzodioxole] on Dibenzfuran, Phenoxathiin,
h
Thianthrene and Phenoxatellurin. See: Chem. Abstr. 1969, 71, 3339u. Popa, G.; Danet, A. F.; Avrigeanu, M. Determination of Some
i
Dissociation Constants of some Heteroaromatic Monoximes. See: Chem. Abstr. 1977, 86, 105620f. Gilman, H.; Van Ess, M. W.; Willis,
j
H. B.; Stuckwisch, C. G. The Metallation of Phenoxathiin. J . Am. Chem. Soc. 1940, 62, 2606-2610. Paget, C. J .; Dennis, E. M.; Nelson,
J .; Delong, D. C. Antiviral Phenoxathiins and Their Analogs. Study of the Structure-Activity Relationships for Antiviral Activity and the
Replacement Ability in Poliovirus Type III Dependent Variants. J . Med. Chem. 1970, 13, 620-623.
Sch em e 1a
at or below room temperature for an additional 2 days. The
resulting slurry was filtered, and the solid was washed with
first water and then aqueous NaHCO3, leaving 328 g of off-
white solid. This was recrystallized by dissolution in DMF
and EtOH by addition of water to the hot solution. Two crops
were taken.
Th ioxa n t h en -9-on e 10,10-Dioxid e 3-N-Met h ylt h io-
a m id e (45). A mixture of 9.99 g (33.2 mmol) of 8 and 20 mL
of sieve-dried toluene was degassed, and 10 g (25 mmol) of
Lawesson’s reagent10 was added under argon, followed by 60
mL more of toluene. The resulting suspension was heated on
a steam bath with occasional shaking for 9.5 h, resulting in a
yellow ball under a yellow supernatant. Removal of the
toluene in vacuo (initial foaming) and addition to the residue
of 80 mL of CH2Cl2 gave a yellow solid, which was rinsed with
a
(a) KOtBu in DMF, room temperature; (b) KOtBu in DMF,
CH2Cl2 to give 12.78 g of solid, mp 212 °C. This was
recrystallized from EtOH by addition of water at the boiling
point, and then had mp 220.5-221.8° and a single spot at Rf
0.70 (faint spot at heavy loading at Rf 0.39), using CH2Cl2 on
silica gel. Starting amide 8 remained at the origin under these
reflux.
3-(N-Meth ylca r ba m oyl)th ioxa n th en -9-on e 10,10-Diox-
id e (8). A mixture of 345.2 g of 3-carboxythioxanth-9-one 10,-
10-dioxide and 1.5 kg of SOCl2 was heated under reflux
overnight. An additional 0.5 kg of SOCl2 was then added, and
refluxing continued an additional day. Excess SOCl2 was then
removed on a steam bath at water aspirator pressure, and the
cooled residue was cautiously and slowly added with stirring
to 1500 g of a cooled aqueous 40% NH2Me solution and stirred
conditions. An additional recrystallization gave 7.90 g.
A
small additional amount and some 8 could be separated from
the initial CH2Cl2 solution.
Th ioxa n th en -9-on e 10,10-Dioxid e 3-Th ioca r boxa m id e.
This was prepared from the amide by modification of the

Selective Inhibitors of Monoamine Oxidase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16 2471
Ta ble 3. MAO Inhibition by Phenothiazines
IC50 (µM) or % I at µM
no.
substituent
3-CONHMe
3-CONHMe
10-Me-3-CONHMe
n
MAO A
MAO B
formula
anal.
mp (°C)
recrysta
77
78
79
0
2
2
0.2
54% at 10
0.3
13% at 10
15% at 10
0 at 1
C14H12N2OS
14H12N2O3S
15H14N2O3S
C,H,N
C,H,N
C,H,N
210.8-213.8
361-364
238.5
A-W
C
C
D-W
D-A-W
a
See footnote a of Table 1.
Ta ble 4. MAO Inhibition by Thioxanthenes
substituent
IC50 (µM) or % I at µM
no.
aromatic ring
9
n
MAO A
MAO B
formula
anal.
mp (°C)
recrysta
80b
81c
82d
83e
84
H/H
H/H
H/H
H/H
H/H
H/H
H/H
Bu/H
H/OH
0
0
1
1
2
0
1
0
2
15% at 1
4% at 1
3% at 1
0 at 1
2
13% at 1
3% at 1
8% at 1
5
13% at 1
0.2% at 1
-33% at 1
-26% at 1
0 at 1
9% at 1
7% at 1
16% at 1
1.6
2-Cl
2-Cl
2-Cl
2-Br
2-Br
2-CONHMe
3-CONHMe
C
C
C
13H9ClO2S
13H9BrS
13H9BrOS
C,H,Cl,S
C,H,Br,S
C,H,Br,S
C,H
144
A
A
A
C-H
W
85
99.5-100
142-145
128-130
200-202
86
87f
88g
C19H21NOS
C15H13NO4S
C,H,N
a
b
See footnote a, Table 1. Commercially available. c Muren, J . F.; Bloom, B. M. Thioxanthene Psychopharmacological Agents. I. 9-(3-
d
Aminopropyl)thioxanthene-2-sulfonamides. J . Med. Chem. 1970, 13, 14-16. Hildich, T. P.; Smiles, S. Intramolecular Rearrangements
of Diphenylmethane o-Sulphoxide. J . Chem. Soc. 1911, 99, 145-160. e Andersen, K. K.; Cinquini, M.; Papanikolaou, E. The Synthesis
and Stereochemistry of TriarylsulfoniumSalts. J . Org. Chem. 1970, 35, 706-710. f Chauhan, P. M. S.; Iyer, R. N.; Shankhdhar, V.; Guru,
P. Y.; Amiya, B. Process for Preparation of 2,7-Diamidinoxanthene and -thioxanthene. Chem. Abstr. 1992, 117, 48338c. g Major metabolite
of 8, Table 1. Not preincubated with enzyme to minimize conversion to 8 before test.
Ta ble 5. MAO Inhibition by Thianthrenes
IC50 (µM) or % I at µM
no.
substituent
n
MAO A
MAO B
formula
anal.
mp (°C)
recrysta
89
90
91
92
2-CONHMe
2-CONHMe
3-CONHMe
2-CONHMe
0/0
2/0
2/0
2/2
79% at 0.1
74% at 0.1
0 at 0.1
0 at 0.1
0 at 0.1
0 at 0.1
0 at 0.1
C14H11NOS2
C,H,N
173-175
236-238
202-204
215-217
EA
A
A
C
C
C
14H11NO3S2
14H11NO3S2
14H11NO5S2
C,H,N,S
C,H,N,S
C,H,N,S
0 at 0.1
M
a
See footnote a, Table 1.
method used to prepare 45, notably using much larger EtOH
volumes for recrystallizations. A yield of 22% of theory was
obtained, and much starting amide was found, mp 226-228.3
°C. Anal. (C14H9NO3S2) C, H, N, S.
of solvent from the main fraction was 42.3 g showing ca. 6.7
aromatic H on NMR. This was oxidized by dissolution in 500
mL of acetic acid, addition of 100 mL of 30% H2O2, and
warming slowly first for 1.5 h at 45° C and then for 1 h at 75
°C. A slight exotherm was noted in the early heating. The
cooled reaction was diluted with 2 L of water, and the resulting
solid was removed by filtration, washed with water, and dried,
yielding 45.77 g of solid.
(2) 3-Cya n op h en oxa th iin 10,10-Dioxid e (60). A mixture
of 10.44 g (33.6 mmol) of 3-bromophenoxathiin 10,10-dioxide,
3.61 g (40.3 mmol calculated as monomer) of CuCN, and 25
mL of dried DMF were stirred and heated under reflux (N2)
for 3 days. The cooled product was stirred at 80 °C for 2 h
with 110 mL of 1 N aqueous HCl and 20 g of FeCl3, cooled,
and partitioned between an additional 200 mL of water and 3
× 200 mL portions of CH2Cl2. The organic layers were washed
successively with 400 mL each of 0.1 N HCl, water, and 0.1 N
3-Cya n op h en oxa th iin 10,10-Dioxid e (62). (1) 3-Br om o-
p h en oxa th iin a n d its 10,10-Dioxid e. A solution in 1200
mL of sieve-dried DMF of 30 g (0.238 mol) of 2-hydroxy-
thiophenol was stirred for 30 min with strict exclusion of air
with 54.7 g (0.487 mol) of KO-t-Bu. Addition of 70.1 g (0.25
mol) of 2,5-dibromonitrobenzene to the stirred reaction under
N2 was followed by 20 min of stirring, and the reaction was
then heated under reflux overnight. After removal of most of
the DMF by vacuum distillation, the residue was partitioned
between 2 L of water and three 800 mL portions of CH2Cl2.
Concentration of the nonaqueous phase was followed by crude
chromatography (5% CH2Cl2 in petroleum ether, 12 cm diam-
eter × 26 cm long silica gel column). The residue of removal

2472 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16
Harfenist et al.
Ta ble 6. MAO Inhibition by Miscellanous Aromatic Amides and Related Compounds
a
b
See footnote a, Table 1. Elliott, A. J .; Eisenstein, N.; Iorio, L. C. Central Nervous System Activity of 7-Substituted 1-Azaphenoxathiin
Analogs and Their Oxidation Products. J . Med. Chem. 1980, 23, 333-335. c Elemental analyses indicated H2O‚1/7 hexanes after
chromatography using 1 EA:4 hexanes. Hexane was confirmed by both 1H and 13C NMR. This sample was used for biological testing.
Activity was too low to merit retesting with a recrystallized sample.
NaOH. TLC (CH2Cl2) showed a major spot and three minor
faster moving spots readily separated on a short silica gel
column. Evaporation of solvent left nitrile suitable for sa-
ponification to the acid. A small amount was recrystallized
from hot ethyl acetate-hexanes for elemental analysis and
testing.
additional 2 h, the reaction was warmed in vacuo to remove
volatiles, and the resulting solid was recrystallized from
ethanol.
P h en oth iazin e-3-N-m eth ylcar boxam ide (77). (1) 3-Br o-
m op h en oth ia zin e. A solution of 5.4 g of “98.5%” NaOH
pellets in 278 mL of 95% EtOH was added to a solution of
46.55 g (0.127 mol) of 2-(acetylamino)phenyl 5-bromo-2-nitro-
phenyl sulfide in 570 mL of acetone under N2, and the reaction
mixture was stirred and heated under reflux for 3 h and let
cool. An additional solution of 6 g of NaOH in 305 mL of 95%
EtOH was added, and heating resumed for 45 min. The dark
reaction was concentrated in vacuo to one-third of its volume
and added to about 3 volumes of water. The resulting salmon-
2-P h en oxath iin yl Meth ylcar bam ate 10,10-Dioxide (73).
A solution of 210 mg (0.85 mM) of 2-hydroxyphenoxathiin
10,10-dioxide and 1 g (excess) of diisopropylethylamine was
stirred with 75 mL of 50% by volume of EtOAc-CH2Cl2, and
72.4 mg (1.3 mM) of methyl isocyanate in 20 mL of the same
solvent mixture was added. After having been stirred an

Selective Inhibitors of Monoamine Oxidase
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 16 2473
colored solid was removed by filtration, rinsed with water, and
dried. It weighed 33.9 g, had mp 180-181.6 °C, and was
sufficiently pure for synthetic use, showing no detectible
uncyclized sulfide by TLC.
of the putative inhibitor. The percent inhibition showed SEM
variation within 5% of mean values. IC50 values were obtained
by plotting mean values vs log of inhibitor concentration and
estimating visually from these plots.
(2) P h en oth ia zin e-3-ca r boxylic Acid . A solution of 250
mL of 1.5 M BuLi in hexane was added during 20 min to a
stirred mixture of 36.7 g (0.132 mol) of 3-bromophenothiazine
in 700 mL of dry THF cooled in a dry ice bath under N2 and
maintained below -70 °C during the reaction and for an
additional 45 min. The contents of the flask were then poured
into excess dry ice and 200 mL of dry Et2O and left an
additional 20 min. After mechanical removal of remaining dry
ice, the residue was partitioned between an additional 500 mL
of Et2O and repeated portions of 0.2 M aqueous Na2CO3.
Acidification of the aqueous solutions led to 22.85 g of mostly
yellow plus a little dark solid, with some odor resembling that
of valeric acid. Recrystallization from EtOH-water gave
yellow prisms melting with darkening about 248 °C. Anal
(C13H9NO2S) C, H, N.
(3) P h en oth ia zin e-3-N-m eth ylca r boxa m id e (77). A so-
lution of 3.26 g (1.34 mmol) of the 3-carboxylic acid in 100 mL
of dry, peroxide-free THF was stirred with 0.305 g (1.88 mmol)
of carbonyldiimidazole for 30 min. A 29 mL portion of a 33%
solution of MeNH2 in ethanol was added under N2 in one
portion, and stirring wa continued overnight. The reaction
mixture was poured into 1600 mL of water and stirred for 3
h, and the resulting 1.59 g of orange solid was filtered off,
rinsed with water, and dried. Two recrystallizations from
EtOH-water gave 1.19 g with mp 210.8-213.8 °C. TLC with
Et2O on silica gel visualized by long-wavelength UV showed
a faint spot at the origin and an intensely fluorescent spot at
Rf 0.19.
P h en oth ia zin e-3-ca r boxylic Acid 5,5-Dioxid e. Oxida-
tion of 4.1 g (16.9 mmol) of phenothiazine-3-carboxylic acid in
HOAc by excess 30% H2O2 added in increments during 5 days
at 80-90 °C yielded 2.75 g of orange solid only slightly soluble
in HOAc. It was recrystallized from EtOH-water and dark-
ened but did not melt at 360 °C. Anal. (C13H9NO4S) C, H, N.
P h en oth ia zin e-3-N-m eth ylca r boxa m id e 10,10-Dioxid e
(77). A mixture of 2.39 g of phenothiazine-3-carboxylic acid
10,10-dioxide (8.36 mmol) and 10 mL of SOCl2 was heated
under reflux (CaCl2 tube) for 2.5 h and then left overnight.
Distillation in vacuo left a residue which was scraped into 100
mL (excess) of aqueous 33% MeNH2 with stirring. After an
additional 4 h of stirring, followed by filtration, water wash-
ing, and drying of the resultant off-white solid, 1.6 g was
obtained. This was dissolved in 50 mL of DMSO at 100 °C by
adding 150 mL of hot water and then cooling, giving 1.45 g of
product.
Refer en ces
(1) Previous paper in this series: 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
Tricyclics Bearing Imidazoline, Oxadiazole, or Tetrazole Groups.
J . Med. Chem. 1996, 39, 1857-1863.
(2) J ohnston, J . P. Some Observations Upon a New Inhibitor of
Monoamine Oxidase in Brain Tissue. Biochem. Pharmacol. 1968,
17, 1285-1297. Squires, R. F. Additional Evidence for the
Existence of Several Forms of Mitochondrial Monoamine Oxi-
dase in the Mouse. Biochem. Pharmacol. 1968, 17, 1401-1409.
(3) 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: Molec-
ular Basis of Differences in Enzymatic Properties. Proc. Natl.
Acad. Sci. U.S.A. 1988, 85, 4934-4938.
(4) Marley, E.; Blackwell, B. Interactions of Monoamine Oxidase
Inhibitors,Amines and Foodstuffs. Adv. Pharmacol. Chemother.
1970, 8, 185-239.
(5) Murphy, D. L.; Lipper, S.; Pickar, D.; J imerson, D.; Cohen, R.
M.; Garrick, N. A.; Alterman, I. S.; Campbell, I. C. Selective
Inhibition of Monoamine Oxidase Type A: Clinical and Anti-
depressant Effects and Metabolic Changes in Man. In Mono-
amine Oxidase Inhibitors - The State of the Art; Youdim, M.,
Paykil, E., Eds.; J ohn Wiley and Sons: New York, 1981; pp 189-
205. Cesura, A. M.; Pletscher, A. The New Generation of
Monoamine Oxidase Inhibitors. Prog. Drug Res. 1992, 38, 171-
297.
(6) Harfenist, M.; J oyner, C. T.; Mize, P. D.; White, H. L. Selective
Inhibitors of Monoamine Oxidase. 2. Arylamide SAR. J . Med.
Chem. 1994, 37, 2085-2089. For a more extensive discussion of
arylamide carcinogenicity and lead references, see Harfenist, M.
Prevention of Ames Test Mutagenicity by Chemical Modification
in a Series of Monoamine Oxidase Inhibitors. J . Med. Chem.
1980, 23, 825-827.
(7) Thull, U.; Testa, B. Screening of Unsubstituted Cyclic Com-
pounds as Inhibitors of Monoamine Oxidases. Biochem. Phar-
macol. 1994, 47, 2307-2310.
(8) White, H. L.; Dren, A. T. Some Preclinical and Clinical Phar-
macological Properties of BW-A616U, A Reversible MAO A
Inhibitor. Clin. Neuropharmacol. 1986, 9, Supplement 4, 561-
562. White, H. L.; Harfenist, M.; Beek, O; Soroko, F.: Cooper,
B. R.; Maxwell, R. A. BWA-616U - A Selective Inhibitor of
MAO-A With Negligible Tyramine-Potentiating Activity and
Potential Antidepressant Activity. Abstr., Soc. Neurosci. 1985,
11 (Oct), 442.
(9) Harfenist, M.; J oyner, C. T.; Heuser, D. J . U.S. Patent 4737514,
Apr 12, 1988.
(10) Lawesson reagent is 2,4-bis(4-methoxyphenyl)cyclodiphos-
phathiane 2,4-disulfide. See: Yde, B.; Yousif, N. M.; Pedersen,
U.; Thomsen, I.; Lawesson, S. O. Studies on Organophosphorus
Compounds.XLVII. Preparation of Thiated Synthons of Amides,
Lactams and Imides by Use of Some New Phosphorus- and
Sulfur-containing Reagents. Tetrahedron 1984, 40, 2047-2052.
Biologica l Meth od s. The methods used were given in
detail in the previous paper.1 They involved a radiometric
procedure using [3H]serotonin and [14C]phenethylamine. The
MAO assays were performed in triplicate at each concentration
J M9608063