1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole derivatives: A new series of potent and selective COX-2 inhibitors in which a sulfonyl group is not a structural requisite

Article abstract of DOI:10.1021/jm990965x

Novel tetrahydro-2H-isoindoles have been prepared and evaluated as inhibitors of the COX-2 isoenzyme. A 1,3-diaryl substitution on the central polycyclic ring system and absence of a sulfonyl moiety are the two structural features of this chemical series.

Full text of DOI:10.1021/jm990965x

4582
J . Med. Chem. 2000, 43, 4582-4593
Articles
1,3-Dia r yl-4,5,6,7-tetr a h yd r o-2H-isoin d ole Der iva tives: A New Ser ies of P oten t
a n d Selective COX-2 In h ibitor s in Wh ich a Su lfon yl Gr ou p Is Not a Str u ctu r a l
Requ isite
Bernard Portevin,† Charles Tordjman,§ Philippe Pastoureau,§ J acqueline Bonnet,§ and Guillaume De Nanteuil*,†
Division D of Medicinal Chemistry and Division of Rhumatology, Institut de Recherche Servier, 11 rue des Moulineaux,
92150 Suresnes, France
Received J une 23, 2000
Novel tetrahydro-2H-isoindoles have been prepared and evaluated as inhibitors of the COX-2
isoenzyme. A 1,3-diaryl substitution on the central polycyclic ring system and absence of a
sulfonyl moiety are the two structural features of this chemical series. A short and easy synthetic
pathway produced several derivatives which were shown to be potent and selective COX-2 vs
COX-1 inhibitors (IC50 ) 0.6-100 nM for COX-2, 100->1000 nM for COX-1). Structural
modifications established that a bicyclic ring appended to the pyrrole nucleus and 4,4′-difluoro
substitution on the phenyl rings were optimal for high inhibitory potency. Activity was confirmed
in the human whole blood assay and subsequently in the murine air-pouch model in which in
vivo PGE2 inhibitory activity was evaluated with respect to gastric tolerance (ED50 for inhibition
of exudate PGE2 of 3 mg/kg and gastric PGE2 of 20 mg/kg). Gastric tolerance was further
assessed after administration to mice of high doses (up to 400 mg/kg) of the inhibitors by
measurement of gastric damage. This panel of studies allowed selection of a number of
tetrahydro-2H-isoindoles which were compared in the adjuvant-induced arthritis model.
Compounds 32 and 37 showed the most potent activity with ED50 values for edema inhibition
in the noninjected paw of 0.35 and 0.15 mg/kg/day, respectively, after oral administration. In
addition, this interesting antiinflammatory profile was accompanied by a protective effect
against arthritis-induced osteopenia, the decrease being 50% with a dose of 0.25 mg/kg/day.
In tr od u ction
chondrocytes from osteoarthritic patients as compared
with normal subjects, whereas COX-1 was not detected
in either normal or osteoarthritic chondrocytes.5 PGE2
overinduction in osteoarthritic cartilage has been shown
to coincide with COX-2 upregulation, hence suggesting
that PGE2 may be differentially regulated in normal
and osteoarthritic cartilage.6
It has been almost 10 years since Kujubu and Xie
simultaneously but independently suggested that an
inflammatory or mitogenic stimulus might result in
gene expression responsible for the synthesis of a
second, inducible cyclooxygenase isoenzyme, called COX-
2.1,2 Although very similar in structure (especially at
the active site) to its constitutive counterpart (hence
named COX-1), COX-2 rapidly emerged as an invaluable
target for developing antiinflammatory agents with
reduced gastrointestinal or renal secondary effects.
Nonsteroidal antiinflammatory drugs (NSAIDs) are
widely used to treat pain, fever, and inflammatory
conditions including osteoarthritis. Recently, it has been
postulated that COX-2 inhibitors may effectively relieve
symptoms in diseases, such as osteoarthritis, while
exhibiting a considerably safer toxicity profile than
classical NSAIDs, especially with respect to gastric
complications.3 It has recently been demonstrated that
IL-1-stimulated chondrocytes from normal human car-
tilage expressed the COX-2 gene and produced the
COX-2 protein, followed by PGE2 release.4 Moreover,
COX-2 protein overexpression has been detected in
When we began our work in this field, only nime-
sulide,7 its analogue NS 398,8 and structural flosulide
analogues9,10 had been described as preferential COX-2
inhibitors. It soon became clear that the pyrazole
derivatives reported by Searle were serious candidates
for potent and selective COX-2 inhibition.11 Soon after,
Merck disclosed the structure of several potent and
selective inhibitors, most of which were structurally
related to the class of cyclopentenyl lactones.12 Struc-
tural optimization work in both laboratories has culmi-
nated in the development of celecoxib (1)13 and rofecoxib
(Vioxx, 2),14 which have recently been marketed.
* To whom correspondence should be addressed. Tel: +1 41 18 22
34. Fax: +1 41 18 24 30.
† Division D of Medicinal Chemistry.
§ Division of Rhumatology.
10.1021/jm990965x CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/28/2000

1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole Derivatives
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24 4583
A 1,2-disubstitution by two aryl groups on a central
core, usually a five- or six-membered heterocycle, seemed
to be a common structural feature of the majority of the
inhibitors. However, very recently, several structures
have been reported in which the 1,2-substitution and/
or the sulfonyl moiety on the aromatic ring were
missing.15 Thus, we reasoned that it would be worth
investigating the inhibitory potential of compounds in
which the central core constituted by a bicyclic ring
system (exemplified below by a 2,3-dihydro-1H-pyrroliz-
ine) was diversely substituted by two aryl or heteroaryl
groups.
To our surprise, we found that 5,7-diphenyl-2,3-
dihydro-1H-pyrrolizine (5) produced the same level of
COX-2 inhibition as 5,6-diphenyl-2,3-dihydro-1H-pyr-
rolizine (3) in the mouse monocyte assay (Table 1).16
This prompted us to extend our evaluation to derivatives
6-8: these three compounds proved fairly potent and
selective with at least 100-fold selectivity for COX-2. 1,3-
Diphenyl-2,3-dihydro-1H-isoindole (7) was then evalu-
ated in the adjuvant-induced arthritis assay in the rat
where it gave a modest but significant 21% inhibition
of the edema at the noninjected paw. Thus 7 was
identified as our initial lead compound, suitable for
further structural modifications.
were obtained via route A as depicted in Scheme 1. The
commercial cyclic anhydrides i were reacted with con-
centrated aqueous ammonia to yield the corresponding
2-carbamoylcycloalkylcarboxylic acids ii. Treatment
with tert-butyl alcohol in a phenylsulfonyl chloride-
pyridine mixture gave the 2-cyano tert-butyl esters iii,
which were readily cyclized with 2 equiv of Grignard
reagent in anhydrous ethyl ether to give the desired
corresponding pyrroles.
Depending upon the substitution pattern, the other
inhibitors described in this paper were prepared ac-
cording to Schemes 2-4. In general, our synthetic
strategy relied on preparation of suitably substituted
1,4-diketones which were subsequently condensed with
appropriate amines (Paal-Knorr reaction) to produce
the corresponding pyrroles.
The second synthetic pathway (route B, Scheme 2)
allowed preparation of a broad array of compounds,
including some already obtained via route A. The key
step in this sequence involved a Diels-Alder reaction
between a 2,3-unsaturated 1,4-diaryl diketone vi and
butadiene or a cyclic diene. When not commercially
available, the diketone was easily prepared by a Friedel-
Craft reaction starting with the diacid chloride of
fumaric acid iv and the suitably substituted aryl moiety
v. The diketone vii resulting from the Diels-Alder
reaction17 was hydrogenated to give the saturated
diketone viii, which was further reacted according to
Paal-Knorr conditions to supply the desired inhibitors
(path B1). Alternatively, diketone vii was condensed
with ammonium formate or ammonium acetate in
ethanol to produce the corresponding unsaturated dia-
rylpyrrole derivatives ix (path B2).18 Reduction of the
double bond was efficiently accomplished using am-
monium formate/palladium or cyclohexene/palladium
(when halogen atoms were present on the aromatic
rings) to give the same final compound as above.
The unsaturated diketone vii-a (R ) H) was also
condensed with methylamine to give intermediate ix-
a , which was further hydrogenated to inhibitor 13. vii-a
was also condensed with carbobenzyloxyhydrazine to
give the hydrazino derivative ix-b, which, upon hydro-
genolysis, produced inhibitor 14. This compound was
further reacted with mesyl chloride to yield the monom-
ethylsulfonyl derivative 15. Condensation of vii-a with
glycine tert-butyl ester followed by reduction of the
double bond gave inhibitor 16. Compounds 22 and 23
were obtained from diketone vii-b, whose fluorine atoms
were substituted with 1 or 2 mol equiv of imidazole to
The first modifications performed in order to enhance
the COX-2 inhibition potency of 7 were simple substitu-
tions with halogen atoms on the phenyl rings: curiously,
compounds 9 and 10 were found to exhibit weak activity,
with an IC50 for COX-2 greater than 100 nM. We
consequently decided to modify the structural core of
compound 7 by replacing the 1,2-dihydroisoindole ring
system by a cyclohexa[c]pyrrole moiety, and we observed
that the resulting derivative 11 was a potent and
selective COX-2 inhibitor with an IC50 of 1 nM vs 100-
500 nM for COX-1.
At that time, 11 became our lead compound and was
subjected to further modifications. In the present paper,
we report our efforts to synthesize and evaluate struc-
tural analogues of compound 11 with the goal of
identifying a potential candidate for preclinical develop-
ment.
Ch em istr y
Most of the compounds described herein were pre-
pared simply using two main synthetic routes. Sym-
metrically substituted phenyl ring-containing inhibitors

4584 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24
Sch em e 1. Route Aa
Portevin et al.
reduction with ammonium formate and Pd/C produced
inhibitor 41.
Resu lts a n d Discu ssion
Biological evaluation of the compounds in this series
initially involved measuring inhibitory potencies against
both constitutive and inducible forms of cyclooxygenase,
isolated from mouse resident peritoneal macrophages
according to our previously published protocol.16 The
activity and selectivity of the most effective compounds
were then confirmed using the human whole blood
assay.22 Assessment of in vivo efficacy with respect to
local inflammation and gastric tolerance was performed
by dosing the inhibitors orally in the murine air-pouch
model.23 Toxicity was evaluated by macroscopic exami-
nation of the gastric mucosa. Finally, the inhibitors with
a good overall profile of efficacy and safety were evalu-
ated in the adjuvant-induced arthritis model of chronic
inflammation in the rat.24
En zym e In h ibitor y Activities. As expected, cele-
coxib (1) proved to be a potent and selective inhibitor
in mouse macrophages giving an IC50 value of 4 nM
against COX-2 (confidence limits, 95%: 1-15 nM), while
it inhibited COX-1 with an IC50 of 3750 nM. The
inhibitory activity of our lead compound 11 was within
the same range of potency but was less selective, with
IC50 values of 1.5 nM for COX-2 and 100 nM against
COX-1.
Va r ia tion s on th e P yr r ole Nitr ogen Atom . To
explore the weakly acidic character of the pyrrole
nucleus, the nitrogen atom was methylated or substi-
tuted with an amino group (Tables 2 and 3). N-
Methylpyrrole derivative 13 was found to be completely
devoid of inhibitory activity on COX-2, whereas com-
pound 14 was a potent COX-2 inhibitor (IC50 ) 1.8 nM);
its affinity for COX-1 was much lower with an IC50 of
1000 nM. Another N-methyl derivative, 26, was also
completely devoid of activity on COX-2. Substitution of
the nitrogen by larger groups bearing an acidic char-
acter, such as in 15 and 16, also resulted in an
significant drop in activity.
Va r ia tion s on th e P h en yl Rin gs. In contrast to the
result obtained with 9 (IC50 COX-2 > 100 nM), intro-
duction of a 4,4′-difluoro substitution (17) preserved a
potent and selective activity with an IC50 for COX-2 of
1.7 nM (Tables 2 and 3). Furthermore, 17 was shown
to be selective, with an IC50 for COX-1 of 100 nM.
Increasing the size of the para substituents failed to
improve efficacy: chlorine, methyl, methoxy, and thi-
omethoxy substitutions gave inhibitors 18-21, with
decreasing potencies from 5 to 500 nM. More complex
variations with mono- (22) or disubstitution (23) incor-
porating an imidazole ring partly decreased or abolished
enzyme inhibition. Thus the para position on the phenyl
rings apparently exhibited limited tolerance to the size
of the substituent, with the difluoro substitution being
definitively the most promising one. Two compounds,
39 and 40, structurally related to 32, contain an
acetamido or methylamino group at position 3; their
COX-2 activity was very weak (>100 nM vs 1.1 nM for
32), suggesting that the presence of a large substituent
at the meta position is deleterious for activity. The
preparation of 2,2′- and 3,3′-difluorophenyl-containing
derivatives is currently under investigation to evaluate
a
Reagents: (a) concd NH4OH; (b) tBuOH, pyridine, phenylsul-
fonyl chloride; (c) PhMgBr, Et2O.
obtain the unsaturated diketones vii-f and vii-g, re-
spectively. The monosubstituted intermediate vii-f was
immediately hydrogenated to afford viii-c, while the bis-
imidazole-containing vii-g was cyclized to give the
unsaturated pyrrole intermediate ix-f. Finally, cycliza-
tion of viii-c and hydrogenation of ix-f gave the desired
inhibitors 22 and 23, respectively. Diketone viii-a (R
) F) was used as a convenient starting intermediate
for preparing inhibitors 34 and 35: Mono- or disubsti-
tution of the fluorine atoms on the phenyl rings of viii-a
was performed with 1 or 2.5 mol equiv of sodium
thiomethoxide in DMSO to give viii-d and viii-f,
respectively. Oxidation with oxone (2KHSO5-KHSO4-
K2SO4) in acetonitrile-water smoothly provided the
corresponding mono- and disulfones viii-e and viii-g,
which were cyclized under ammonium formate condi-
tions as described above. viii-a was also used to prepare
39 and 40: viii-h was obtained by nitration of viii-a in
fuming nitric acid and was then reduced using the Fe/
NH4Cl system in dioxane-water to produce viii-i.
Acetylation and final ring closure were achieved in a
one-step process using ammonium acetate in acetic acid
to give 39. When ammonium formate was used instead
of acetate, formylation and cyclization yielded the bis-
formylamino derivative which was further reduced with
LAH to the bis-methylamino-containing inhibitor 40 as
the hydrochloride.
Two compounds were synthesized in which a pyrro-
lidine ring was appended on the pyrrole nucleus, using
Scheme 3 (route C). According to the method developed
by Padwa, the commercially available N-[(trimethylsi-
lyl)methyl]benzylamine x was reacted with formalde-
hyde in methanol in the presence of potassium carbon-
ate to provide the azomethine ylide equivalent N-[(tri-
methylsilyl)methyl]-N-methoxymethylbenzylamine xi;19
lithium fluoride was used as a desilylating agent to
generate the 1,3-dipole, which was subsequently cyclized
with the dipolarophile vi (R ) H) to pyrrolidine xii. The
methodology reported in Scheme 2 was then successfully
applied to the ring closure of xii (ammonium acetate,
acetic acid, ethanol) to give 29. Further debenzylation
(ammonium formate, Pd/C) yielded inhibitor 30, as the
hydrochloride.
Compound 41 was conveniently prepared using the
sequence outlined in Scheme 4 (route D). The bromo-
methyl ketone xiv as the hydrobromide was prepared
from commercially available 4-acetylpyridine xiii, which
was brominated using the reported methodology.20
Phosphorane xv was then obtained by reacting xiv with
triphenylphosphine and pyridine in acetonitrile and
oxidative coupling with peracetic acid in dichloromethane
smoothly afforded diketone vi-a .21 As previously, this
diketone was reacted with cyclopentadiene to give the
unsaturated bicyclic diketone vii-m . Cyclization with
ammonium acetate in acetic acid gave ix-h , and further

1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole Derivatives
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24 4585
Sch em e 2. Route Ba
a
Reagents: (a) AlCl3, CS2; (b) diene, toluene; (c) ammonium formate, ethanol or ammonium acetate, ethanol; (d) cyclohexene, 10%
Pd/C, N-methylpyrrolidone or ammonium formate, 10% Pd/C, ethanol.
Sch em e 3. Route Ca
Sch em e 4. Route Da
a
Reagents: (a) 37% aq formaldehyde, K2CO3, methanol; (b) 1,2-
a
Reagents: (a) Br2, 48% HBr; (b) (i) PPh3, acetonitrile, (ii) Et3N,
dibenzoylethylene, LiF, acetonitrile; (c) ammonium acetate, acetic
acid, ethanol; (d) ammonium formate, 10% Pd/C, ethanol, then aq
HCl.
acetonitrile; (c) peracetic acid, dichloromethane; (d) freshly distilled
cyclopentadiene, toluene; (e) ammonium acetate, acetic acid; (f)
ammonium formate, 10% Pd/C, ethanol-dioxane.
the role of smaller substituents at ortho and meta
positions. Finally, replacement of the two phenyl rings
by a 4-substituted pyridine (41) resulted in loss of all
activity against the COX-2 isoenzyme.
Ta ble 1. IC50 Values for Various 1,2- and
1,3-Diaryl-Substituted Heterocycles
IC50 (nM)
IC50 (nM)
Va r ia tion s on th e Cycloa lk yl Rin g. At this point,
we turned our attention to the nature of the cycloalkyl
ring appended to the pyrrole nucleus (Tables 2 and 3).
Introduction of a nonconjugated double bond at positions
6, 7 in the cyclohexyl ring gave compound 12, which was
slightly less active than its saturated analogue with an
IC50 of 3.3 nM for COX-2. Introduction of two methyl
groups on the cyclohexyl ring (cis/trans ratio: 85:15) in
24 had little effect on enzyme inhibitory activity but
improved selectivity to a considerable extent (IC50
COX-2 ) 3.1 nM and IC50 COX-1 ) 1000 nM). Like its
counterpart 12, the unsaturated derivative 25 was less
potent than 24 (IC50 COX-2 ) 14.5 nM). The corre-
sponding N-methyl derivative 26 was, as previously
observed, completely inactive. Decreasing the size of the
fused cycloalkyl ring produced encouraging results: the
five-membered ring-containing compound 27 and its
compd
COX-2
COX-1
compd
COX-2
32.4
>100
>100
1.5
COX-1
3
4
5
6
7
10
<1000
2500
<1000
1000
8
9
10
11
>2500
nda
28.7
50.0
10.9
10.0
nd
250
>2500
a
nd, not determined.
difluoro analogue 28 were found to be very potent and
selective COX-2 inhibitors, with IC50 values of 0.7 and
2.9 nM, respectively. The selectivity vs COX-1 was in
the same range for both analogues (IC50 ) 250 and 100
nM, respectively) and similar to that of the six-
membered ring-containing inhibitors. Insertion of a
nitrogen atom into the five-membered ring gave the two
pyrrolidinopyrrole derivatives 29 and 30, which were
essentially devoid of COX-2 inhibitory activity. Although
less selective than their five-membered ring counter-

4586 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24
Ta ble 2. Modifications of the Pyrrole and Aryl Substituents
Portevin et al.
a
b
Compounds gave satisfactory analyses ((0.4%) unless otherwise indicated. IC50 values are concentrations of compounds required to
achieve 50% inhibition against COX-1 and COX-2 from mouse resident macrophages. c Double bond present in positions 6, 7. CH,
d
cyclohexane. e EA, ethyl acetate. f nd, not determined. MC, methylene chloride. C: found, 78.65; calcd, 79.25. Cl: found, 20.22; calcd,
g
h
i
j
k
20.72. C: found, 76.76; calcd, 77.29. Cl: found, 15.30; calcd, 14.82.
parts, the [2.2.1] bicyclic derivatives 31 and 32 exhibited
nanomolar inhibitory activity against COX-2 [IC50 ) 2.6
and 1.1 (0.05-17) nM, respectively] as well as good
selectivity (IC50 for COX-1 ) 100 nM). The correspond-
ing unsaturated derivative 33 was almost as active, with
an IC50 of 4.5 nM and comparable selectivity. Using this
central 4-azatricyclo[5.2.1.02,6]deca-2, 5-diene system,
we introduced one (34) or two (35) methylsulfonyl
moieties in position 4 of the phenyl rings. This resulted
in a major loss of activity, suggesting that the mecha-
nism of interaction of our inhibitors with the active site
of COX-2 is probably different from that of the reference
inhibitors celecoxib or rofecoxib which contain a sul-
fonamide and a sulfone moiety, respectively. Results
with the [2.2.2] bicyclic system-containing 36 and 37
paralleled those obtained with the bicycloheptane series,
with COX-2 IC50 of 1.6 and 0.6 (0.1-3.3) nM, respec-
tively. Selectivity of these inhibitors was also very
encouraging: IC50 values against COX-1 were 250 and
230 nM, respectively. As expected, the unsaturated
derivative 38 was 1 order of magnitude less potent than
its saturated analogue 36.
Hu m a n Wh ole Blood Assa y. The potency and
selectivity of celecoxib and of the most effective inhibi-
tors were confirmed using the human whole blood assay
according to Patrignani et al.22 (Table 4). Compound 32
was found to be almost 1 order of magnitude more
potent than 1 in terms of COX-1 and COX-2 inhibition,
but its COX-1/COX-2 ratio was in the same range as
that of celecoxib. More interestingly, compound 37,
which produced similar COX-2 inhibition as 1 and 32,
was less active on COX-1 with an IC50 of 30 µM and a
COX-1/COX-2 ratio 20 times higher than that of the
reference inhibitor.
Air -P ou ch Mod el in th e Mou se. The activity of our
COX- 2 inhibitors given orally was assessed in an acute
inflammation model, the air-pouch model.23 This assay
allowed the evaluation (Table 5) in the same animal of
the PGE2 levels in the inflammatory exudate (mainly
COX-2) and in the gastric mucosa (mainly COX-1).
The most active compounds in vitro were evaluated
in this model. The reference inhibitor celecoxib (1)
showed potent antiinflammatory activity with an ED50
of 10 mg/kg, while a higher dose (ED50 > 40 mg/kg) was

1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole Derivatives
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24 4587
Ta ble 3. Modifications of the Ring Fused to the Pyrrole Nucleus
a
b
Compounds gave satisfactory analyses ((0.4%) unless otherwise indicated. IC50 values are concentrations of compounds required to
achieve 50% inhibition against COX-1 and COX-2 from mouse resident macrophages. c CH, cyclohexane; EA, ethyl acetate. C: found,
d
87.81; calcd, 88.25. e nd, not determined. f ND, not determined but satisfactory results by high-resolution MS analysis were obtained.g S:
h
i
j
found, 7.94; calcd, 8.41. C: found, 61.93; calcd, 62.56. C: found, 88.36; calcd, 88.85. C: found, 68.45; calcd, 68.95.

4588 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24
Portevin et al.
Ta ble 4. Inhibitory Potencies on Human Blood in Vitro
IC50 (µM)
Ta ble 6. Gastric Toxicity Studies
compda
gastric scoreb
compda
gastric scoreb
compd
COX-1
COX-2
COX-1/COX-2
1
28
31
0.7 ( 0.2
0.7 ( 0.4
0.3 ( 0.3
32
37
0.3 ( 0.2
0.3 ( 0.2
2.4 ( 0.4
1
32
37
9.3 (1.6-49.5)a
1 (0.4-2.2)
30
0.7 (0.1-6.6)
0.08 (0.01-1.04)
0.12 (0.04-0.32)
13.2
12.5
250
indomethacin
a
Compounds were given orally at 400 mg/kg, except in-
domethacin at 2.5 mg/kg. bGastric erosion was evaluated using a
visual score ranging from 0 (normal) to 5 (perforation). Values
shown are means ( SEM (8 animals/group); control values ) 0.4
( 0.2.
a
Confidence limits, 95%.
Ta ble 5. PGE2 Inhibitory Activities in the Air-Pouch Model
ED50 (mg/kg)a
Ta ble 7. Antiinflammatory Activity in the Adjuvant-Induced
Arthritis Model: Noninjected Paw at Day 21
compd
inhib of gastric PGE2
inhib of exudate PGE2
1
7
>40
100
50
10
100
20
compd
inhib of edema ED50 (mg/kg/day) poa
11
14
17
19
21
27
28
31
32
36
37
1
17
28
32
37
0.32 (0.18-0.54)b
15%c
ndb
50
inc
20
14%d
nd
>80
10
>80
80
>40
>40
>20
in
20
20
20
0.35 (0.16-0.69)
0.15 (0.05-0.36)
a
b
10 rats/group; edema volume of control ) 1.18 mL. ED50
values were estimated by linear regression. Confidence limits, 95%.
20
c Inhibition at 2 mg/kg/day. Inhibition at 2 mg/kg, twice daily.
d
6.8
>40
3.5
selectivity, while a five-membered ring-containing de-
rivative seems to produce inconsistent results; a one-
or two-carbon bridge added to the six-membered ring
improves activity, especially when a difluoro substitu-
tion is present on the phenyl rings appended to the
pyrrole nucleus, giving the most interesting inhibitors
32 and 37. Whether this is due to improved lipophilicity,
acidic/basic character, or to other factors remains to be
determined.
Ga str ic Toler a n ce Assa y in th e Mou se. Evalua-
tion of the gastrointestinal (GI) toxicity of our inhibitors
was readily performed by assigning a visual individual
score to the lesions observed 5 h after oral administra-
tion of the compounds in a high dosage (400 mg/kg) to
fasted mice (Table 6). Briefly, a low incidence of gastric
erosion was observed for the compounds reported here-
after. The score obtained with celecoxib was comparable
to the value given by the five-membered ring-containing
28. The gastric erosion score was slightly decreased after
administation of the bicyclo [2.2.1]- and [2.2.2]-contain-
ing inhibitors 31, 32, and 37. However, any of them was
found to be significantly different from the control (0.4
( 0.2).
Ad ju va n t -In d u ced Ar t h r it is in t h e R a t . In this
chronic model of inflammation, four compounds were
evaluated and compared with celecoxib (1). As shown
in Table 7, the results confirmed the strong antiinflam-
matory activity of celecoxib, with ED50 values for edema
inhibition of the noninjected paw at day 21 of 0.32 mg/
kg/day (lit. ED50 ) 0.37 mg/kg/day).11
The four inhibitors reported in Table 7 differ only by
the nature of the ring fused to the pyrrole nucleus; a
remarkable finding is that the compounds containing a
bicyclic structural core appended to the pyrrole ring
were the only compounds that produced a potent and
sustained antiinflammatory effect at doses similar to
1. Dose-response curves obtained for 32 and 37 are
given in Figure 1. While the activity of compound 32
was in the same range as that of celecoxib, 37 was 3.8
times more active (comparison of the regression lines).
Moreover, another aspect of the protective effects of our
inhibitors was shown on osteopenia (decrease in bone
mineral content) induced by arthritis as assessed at the
a
ED50 values were evaluated using 3 or 4 dose points (at least
8 mice/group). nd, not determined. c in, inactive.
b
needed to inhibit gastric PGE2. Although many of our
inhibitors were equipotent to 1 in vitro, their potency
did not always translate well in vivo: for example,
dihydro-1H-isoindole 7 was found to be weakly active
with an ED50 of 100 mg/kg for both PGE2 inhibitions.
The initial lead compound 11 was fairly effective on
exudate PGE2 inhibition, with an ED50 of 20 mg/kg, but
this level of dose also resulted in significant gastric
PGE2 inhibition (ED50 ) 50 mg/kg). No improvement
was obtained when the phenyl rings were 4-fluoro-
disubstituted (17). Substitution of the pyrrole nitrogen
(14) as well as introduction of a methyl group on the
phenyl rings (19) produced low or erratic inhibition.
Chlorine substitution (21) proved to be more effective
on exudate PGE2 inhibition, with an ED50 of 20 mg/kg,
and produced an interesting level of selectivity (ED50
for gastric PGE2 inhibition greater than 80 mg/kg).
Five-membered ring-containing 27 gave strong inhibi-
tion of exudate PGE2 but was found to be half as
selective as its six-membered ring counterpart. The
difluoro analogue 28 was only as potent, but more
selective. Similar levels of potency and selectivity were
obtained with the [2.2.1] bicyclic system-containing
inhibitor 31, whereas remarkable improvements in
activity and selectivity were obtained when a difluoro
substitution was again introduced into this structure
to yield 32, which had an ED50 of 6.8 mg/kg for exudate
PGE2 inhibition and >40 mg/kg for gastric PGE2
inhibition. The [2.2.2] bicyclic system-containing 36 was
inactive on both PGE2 inhibitions, whereas its difluoro
analogue 37 was the most potent PGE2 inhibitor, with
an ED50 of 3.5 mg/kg for exudate PGE2. Selectivity was
similar to that obtained with the [2.2.1] analogue 32.
In summary, these preliminary in vivo studies suggest
that pharmacokinetics may play a crucial role in the
general activity profile of the compounds described
herein: A six-membered ring fused to the non-N-
substituted pyrrole nucleus appears to be of prime
importance for obtaining significant in vivo potency and

1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole Derivatives
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24 4589
Effect on Ga str ic Mu cosa in th e Mou se. Male Swiss
mice, fasted overnight with free access to water, were dosed
orally with compounds or vehicle 5 h before sacrifice by cervical
dislocation. The stomach was removed, rised and opened for
macroscopic assessment of mucosal damage by an observer
unaware of the treatment according to a rating scale from 0
(normal) to 5 (perforation).
Ad ju va n t-In d u ced Ar th r itis in th e Ra t. Arthritis was
induced and assessed in female Lewis rats according to the
method previously described by Bonnet et al.24
Rou te A (Sch em e 1). 2-Ca r ba m oylcycloh exa n eca r -
boxylic Acid , i-a . To 66 mL of concentrated aqueous ammonia
was added portionwise 46.25 g (0.3 mol) of cis-cyclohexane-
1,2-dicarboxylic anhydride while maintaining the reaction
temperature below 25 °C. The reaction mixture was stirred
overnight, and acidified with 12 N HCl. The resulting precipi-
tate was collected, washed with water and dried to give 51 g
(100%) of the desired monoacid: mp 197 °C; IR (Nujol) 3435,
3217, 2600-2200, 1672, 1643 cm-1; 1H NMR (DMSO-d6) δ 1.2-
1.75 (6H, 3m), 1.9 (2H, m), 2.45-2.7 (2H, 2m), 6.70 and 7.15
(2H, br s), 11.85 (1H, br s).
F igu r e 1. Antiinflammatory activity in the adjuvant-induced
arthritis model (noninjected paw edema at day 21): dose-
response curves for edema inhibition by compounds 32 (- -)
and 37 (-); M ( SEM (8 rats/group); relative potency ) 3.8.
proximal site of the femur. Compounds 32 and 37
induced a 50% decrease of osteopenia from a dose of 0.25
mg/kg/day, an effect closely similar to the result ob-
tained with celecoxib (52%).
6-Ca r ba m oylcycloh ex-3-en eca r boxylic a cid , i-b: mp 151
°C; IR (Nujol) 3600-2300, 1702-1650 cm-1; 1H NMR (DMSO-
d6) δ 2.1-2.55 (4H, m), 2.85 (2H, m), 5.6 (2H, br s), 6.85 (1H,
br s), 7.2 (1H, br s), 11.5-12.5 (1H, br s).
In summary, a 1,3-diaryl substitution on the central
polycyclic ring system and absence of a sulfonyl moiety
on the phenyl rings are the two distinctive structural
features that distinguish the new chemical series de-
scribed herein from the reference inhibitors 1 and 2.
COX-2 inhibition in the low-nanomolar range and COX-
1/COX-2 selectivity ratios as high as 3 orders of mag-
nitude were obtained. In the human whole blood assay,
inhibitor 37 confirmed its interesting selectivity with a
COX-1/COX-2 ratio of 250. Evaluation of GI toxicity (up
to 400 mg/kg to normal mice) followed by the air-pouch
assay allowed us to select a few candidates for evalua-
tion in the adjuvant-induced arthritis model: after oral
administration, inhibitors 32 and 37 showed strong
antiinflammatory potential associated with a tissue-
protective effect. On the basis of these findings, these
two compounds have recently been proposed for further
preclinical evaluation.
2-Ca r ba m oylcyclop en ta n eca r boxylic a cid , i-c: mp 140
1
°C; IR (Nujol) 3450-2300, 1706 cm-1; H NMR (D2O) δ 1.7-
2.3 (6H, m), 3.3 (2H, m).
2-Cya n ocycloh exa n eca r boxylic a cid ter t-bu tyl ester ,
ii-a : bp 100-110 °C/0.01 mmHg; IR (KBr) 2238, 1729 cm-1
;
1H NMR (DMSO-d6) δ 1.2-1.75 (6H, 2m), 1.4 (9H, s), 1.9 (2H,
m), 2.6 (1H,m), 3.35 (1H, q).
6-Cya n ocycloh ex-3-en eca r boxylic a cid ter t-bu tyl es-
ter , ii-b: IR (KBr) 2240, 1728, 1657-1620 cm-1 1H NMR
;
(CDCl3) δ 1.5 (9H, s), 2.45 (4H, m), 2.65 (1H, m), 3.35 (1H, m),
5.65 (1H, m), 5.85 (1H, m).
2-Cya n ocyclop en ta n eca r boxylic a cid ter t-bu tyl ester ,
ii-c: bp 80-82 °C/0.02 mmHg; IR (KBr) 2978-2879, 2241,
1728 cm-1; 1H NMR (DMSO-d6) δ 1.48 (9H, s), 1.58-2.22 (6H,
m), 2.91-3.37 (2H, m).
1,3-Dip h en yl-4,5,6,7-tetr a h yd r o-2H-isoin d ole, 11: mp
146 °C; IR (Nujol) 3437, 1603 cm-1 1H NMR (CDCl3) δ 1.8
;
(4H, m), 2.8 (4H, m), 7.1-7.6 (10H, m), 8.25 (1H, br s).
1,3-Dip h en yl-4,5-d ih yd r o-2H-isoin d ole, 12: mp 138 °C;
1
IR (Nujol) 3437,1604 cm-1; H NMR (DMSO-d6) δ 3.40 (4H,
Exp er im en ta l Section
s), 5.85 (2H, s), 7.20 (2H, t), 7.40 (4H, t), 7.65 (4H, d), 11.0
(1H, br s).
Ch em istr y. Melting points were determined on a Tottoli
apparatus and are not corrected. Elemental analyses (C, H,
N, S, Cl) were carried out by the analytical department of the
Institut de Recherches SERVIER; results obtained for specified
elements are within (0.4% of the theoretical values. Infrared
spectra were recorded on a Bruker IFS 28 spectrophotometer
in solution, in Nujol, or using KBr pellets; frequencies are
1,3-Di(4-flu or op h e n yl)-4,5,6,7-t e t r a h yd r o-2H -isoin -
1
d ole, 17: mp 139 °C IR (Nujol) 3469 cm-1; H NMR (DMSO-
d6) δ 1.70 (4H, m), 2.70 (4H, m), 7.25 (4H, m), 7.60 (4H, m),
10.80 (1H, br s).
1,3-Di(4-m eth ylth ioph en yl)-4,5,6,7-tetr ah ydr o-2H-isoin -
d ole, 18: mp 174 °C; IR (Nujol) 3400, 1590, 823 cm-1; 1H NMR
(DMSO-d6) δ 1.74 (4H, m), 2.47 (6H, s), 2.71 (4H, m), 7.30 (4H,
d), 7.54 (4H, d), 10.68 (1H, s).
1,3-Dip h en yl-2,4,5,6-tetr a h yd r ocyclop en ta [c]p yr r ole,
27: mp 190 °C; IR (Nujol) 3437, 3075-3056, 1610, 1591, 1490,
1163-1077 cm-1; 1H NMR (CDCl3) δ 2.50 (6H, m), 7.28-7.70
(10H, m), 8.32 (1H, br s).
expressed in cm-1 1H NMR spectra were recorded in CDCl3
.
or DMSO-d6 solution (S.d.S. isotopes) in 5-mm tubes (Wilmad)
at 27 °C and were collected on a Bruker AC 200 or AM 300
spectrometer. Chemical shifts (δ) are given in ppm with
tetramethylsilane (TMS) as the internal standard. Mass
spectra were obtained on either a Nermag R 1010, a Finnigan
MAT 95-Q, or a Finnigan MAT TSQ-7000.
1,3-Di(4-flu or op h en yl)-2,4,5,6-tetr a h yd r ocyclop en ta [c]-
Biology. In vitro cyclooxygenases assays were performed
as reported in refs 15 and 21.
p yr r ole, 28: mp 170 °C; IR (Nujol) 3450 cm-1 1H NMR
;
(DMSO-d6) δ 2.40 (2H, m), 2.80 (4H, m), 7.25 (4H, m), 7.70
(4H, m), 10.70 (1H, br s).
Air -P ou ch in th e Mou se. Air-pouches were induced in
mice according to the method previously described.23 After 18
h, compounds or vehicle were given orally, and 1 h later, 500
µL of a sterile solution of 1% carrageenan were injected into
the air-pouch. Mice were sacrified by cervical dislocation 5 h
after carrageenan injection. Pouches were lavaged with 2 mL
of heparinized HBSS. The lavage fluid was immediatly cooled
on ice, and PGE2 level was determined by EIA (Enzymo
Immunoassay) after solid-phase extraction. In parrallel, the
stomachs were excised, opened, rinsed in PBS, weighed and
incubated in 5 mL of PBS for 15 min. After centrifugation,
the supernatants were assayed for PGE2 by EIA.
Rou te B (Sch em e 2). (6-Ben zoylcycloh ex-3-en yl)(p h e-
n yl)m eth a n on e, vii-a . The preparation of this compound is
reported in ref 16: mp 114 °C; IR (Nujol) 3037, 1681 cm-1; 1H
NMR (CDCl3) δ 2.13-2.56 (4H, m), 4.15 (2H, m), 5.80 (2H,
m), 7.50 (6H, m), 8.03 (4H, d).
[6-(4-F lu or oben zoyl)cycloh ex-3-en yl](4-flu or op h en yl)-
m eth a n on e, vii-b: mp 125 °C; IR (Nujol) 3100, 3040, 1672,
1
1598, 1237, 852 cm-1; H NMR (CDCl3) δ 2.20 (2H, m), 2.55
(2H, m), 4.10 (2H, m), 5.80 (2H, m), 7.15 (4H, t), 8.05 (4H,
dd).

4590 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24
Portevin et al.
[6-(4-Met h ylb en zoyl)cycloh ex-3-en yl](p -t olyl)m et h a -
n on e, vii-c: mp 124 °C; IR (Nujol) 3466, 1610 cm-1; 1H NMR
(DMSO-d6) δ 2.35 (6H, s), 3.35 (4H, s), 5.90 (2H, s), 7.20 (4H,
d), 7.55 (4H, d), 10.8 (1H, br s).
decanted and washed with water, then dried over calcium
sulfate. Filtration and evaporation gave a crude solid that was
chromatographed (silica gel, cyclohexanes-ethyl acetate 95:
5) to give 0.3 g (11%) of the desired compound: mp 107 °C; IR
1
(Nujol) 1665 cm-1; H NMR (DMSO-d6) δ 0.95-1.75 (6H, m),
[6-(4-Meth oxyben zoyl)cycloh ex-3-en yl](4-m eth oxyp h e-
n yl)m eth a n on e, vii-d : mp 130 °C; IR (Nujol) 1661, 1539 cm-1
;
2.5 (3H, s), 2.5 (1H, m), 2.7 (1H, m), 3.95 (1H, m), 4.3 (1H, m),
7.35 (4H, m), 7.85-8.2 (4H, m).
1H NMR (CDCl3) δ 2.18 (2H, m), 2.5 (2H, m), 3.85 (6H, s), 4.07
(2H, m), 5.8 (2H, m), 6.93 (4H, d), 8.00 (4H, d).
[3-(4-F lu or oben zoyl)bicyclo[2.2.1]h ep t-2-yl](4-m eth yl-
su lfon ylp h en yl)m eth a n on e, viii-e. The methylmercaptan
derivative viii-d (1.1 g, 0.003 mol) was dissolved in 60 mL of
a 1:1 mixture of acetonitrile and dioxane. 30 mL of water was
added, followed by addition of 5.5 g of oxone (potassium
peroxomonosulfate). The reaction mixture was stirred at room
temperature for 24 h, filtered and evaporated. The residue was
taken up with methylene chloride, washed with water, and
dried over calcium sulfate. Evaporation of the solvent afforded
a solid which was washed with pentane to give 1 g (84%) of
the desired sulfone: mp 142 °C; IR (Nujol) 1669, 1740, 1330,
1151 cm-1; 1H NMR (DMSO-d6) δ 1.0-1.8 (6H, 3m), 2.75 (2H,
m), 3.35 (3H, s), 4.0 (1H, m), 4.4 (1H, m), 7.35 (2H, m), 8.1-
8.3 (6H, m).
[3-(4-Meth ylth ioben zoyl)bicyclo[2.2.1]h ep t-2-yl](4-m e-
th ylth iop h en yl)m eth a n on e, viii-f. 7.3 g (0.0215 mol) of
difluoro derivative viii-a was reacted in 120 mL of DMSO with
sodium thiomethoxide (2.74 g, 0.043 mol). The reaction was
heated at 80 °C for 7 h, then at 50 °C for 20 h. After
evaporation of the solvent, the crude residue was chromato-
graphed (silica gel, cyclohexanes-ethyl acetate 95:5) to give
2.7 g (32%) of the desired compound: mp 112 °C; IR (Nujol)
1661, 1590 cm-1; 1H NMR (DMSO-d6) δ 0.9-1.8 (6H, 3m), 2.55
(6H, s), 2.4 and 2.7 (2H, 2 br s), 3.95 and 4.3 (2H, 2 br s), 7.35
(4H, d), 7.8-8.15 (4H, 2d).
[6-(4-Ch lor oben zoyl)cycloh ex-3-en yl](4-ch lor op h en yl)-
m eth a n on e, vii-e: mp 125 °C; IR (Nujol) 1676, 1600 cm-1
;
1H NMR (DMSO-d6) δ 1.95-2.2 (2H, m), 2.45 (2H, m), 4.05
(2H, m), 5.80 (2H, d), 7.65 (4H, d), 8.05 (4H, d).
[6-(4-F lu or ob e n zoyl)cycloh e x-3-e n yl](4-im id a zol-1-
ylp h en yl)m eth a n on e, vii-f: IR (Nujol) 1671, 1596 cm-1; 1H
NMR (CDCl3) δ 2.2 (2H, m), 2.55 (2H, m), 4.1 (2H, m), 5.8 (2H,
d), 7.15 (2H, t), 7.25 (1H, s), 7.35 (1H, s), 7.55 (2H, d), 7.95
(1H, s), 8.05 (2H, m), 8.15 (2H, d).
{6-[4-(Im id a zol-1-yl)ben zoyl]cycloh ex-3-en yl}(4-im id a -
zol-1-ylp h en yl)m eth a n on e, vii-g: mp 268 °C; IR (Nujol)
1662, 1605-1579 cm-1; 1H NMR (CDCl3) δ 2.25 (2H, m), 2.60
(2H, m), 4.1 (2H, m), 5.8 (2H, d), 7.25 (2H, s), 7.35 (2H, s),
7.55 (4H, d), 7.95 (2H, s), 8.15 (4H, d).
(6-Ben zoyl-3,4-d im eth ylcycloh ex-3-en yl)(p h en yl)m eth -
a n on e, vii-h : mp 113 °C; IR (Nujol) 1675 cm-1 1H NMR
;
(DMSO-d6) δ 1.65 (6H, s), 2.1-2.4 (4H, m), 4.1 (2H, m), 7.6
(6H, m), 8.05 (4H, d).
(3-Ben zoylbicyclo[2.2.1]h ep t-5-en -2-yl)(p h en yl)m eth a -
1
n on e, vii-i: mp 80 °C; IR (Nujol) 1666, 1579-1595 cm-1; H
NMR (CDCl3) δ 1.49 (1H, d), 1.91 (1H, d), 3.16 (1H, s), 3.36
(1H, s), 3.98 (1H, dd), 4.51 (1H, dd), 5.84 (1H, m), 6.43 (1H,
m), 7.50 (6H, m), 8.01 (4H, t).
[3-(4-F lu or oben zoyl)bicyclo[2.2.1]h ep t-5-en -2-yl](4-flu -
or op h en yl)m eth a n on e, vii-j: mp 104 °C; IR (Nujol) 1666,
1620, 1598 cm-1; 1H NMR (DMSO-d6) δ 1.4 and 1.8 (2H, 2dd),
3.1 and 3.3 (2H, 2 br s), 3.8 and 4.4 (2H, 2d), 5.85 and 6.4 (2H,
2dd), 7.3 (4H, m), 8.1 (4H, m).
[3-(4-Meth ylsu lfon ylben zoyl)bicyclo[2.2.1]h ep t-2-yl](4-
m eth ylsu lfon ylp h en yl)m eth a n on e, viii-g: obtained as the
monosulfone viii-e; mp 148-150 °C; IR (Nujol) 1682, 1315-
1
1292, 1154 cm-1; H NMR (DMSO-d6) δ 0.9-1.8 (6H, m), 2.5
and 2.75 (2H, m), 3.3 (6H, 2s), 4.05 and 4.45 (2H, m), 8.1 (4H,
(3-Ben zoylb icyclo[2.2.2]oct -5-en -2-yl)(p h en yl)m et h a -
n on e, vii-k : mp 131 °C; IR (Nujol) 3100, 1673, 1596-1579,
d), 8.15-8.3 (4H, m).
1
778-688 cm-1; H NMR (DMSO-d6) δ 0.9-2.0 (4H, m), 2.90
[3-(4-F lu or o-3-n itr oben zoyl)bicyclo[2.2.1]h ep t-2-yl](4-
flu or o-3-n itr op h en yl)m eth a n on e, viii-h . 40 mL of fuming
nitric acid was cooled at 0 °C, and 15 g (0.044 mol) of the
difluoro derivative viii-a was added portionwise. After 15 min
at 0 °C, the reaction was stirred at room temperature for 1 h,
then poured on 100 g of crushed ice. The gummy solid was
taken up with ethyl acetate. The organic layer was washed
with water and dried over calcium sulfate. After evaporation
of the solvent, the resulting solid was washed with ethanol to
give 8.1 g (43%) of a white solid: mp 161 °C; IR (Nujol) 1688,
(2H, m), 4.10-4.30 (2H, dd), 6.10-6.45 (2H, tt), 7.55 (6H, m),
7.90-7.98 (4H, dd).
[3-(4-F lu or oben zoyl)bicyclo[2.2.2]oct-5-en -2-yl](4-flu o-
r op h en yl)m eth a n on e, vii-l: mp 144 °C; IR (Nujol) 1669,
1597-1506, 1228-1208 cm-1; 1H NMR (CDCl3) δ 1.0-2.0 (4H,
m), 3.0 (2H, m), 4.1-4.4 (2H, 2dd), 6.2-6.5 (2H, 2t), 7.2 (4H,
m), 8.1 (4H, m).
P a th B1 (Sch em e 2). [3-(4-F lu or oben zoyl)bicyclo[2.2.1]-
h ep t-2-yl](4-flu or op h en yl)m eth a n on e, viii-a . The ethyl-
enic intermediate vii-j (3.82 g, 0.011 mol) was heated overnight
at 80 °C in 100 mL of N-methylpyrrolidone in the presence of
20 mL of cyclohexene and 0.6 g of 10% Pd/C. After evaporation
of the solvent, the residue was taken up with ethyl acetate,
washed with water, and dried over calcium sulfate. Filtration
and evaporation gave a solid which was washed with ethanol
and pentane to give 2.7 g (71%) of the desired hydrogenated
1
1613, 1539-1348 cm-1; H NMR (CDCl3) δ 1.0-1.9 (6H, m),
2.55 (1H, br s), 2.8 (1H, br s), 4.05 (1H, d), 4.45 (1H, m), 7.45
(2H, td), 8.3 (2H, m), 8.75 (2H, m).
[3-(4-Flu or o-3-a m in oben zoyl)bicyclo[2.2.1]h ep t-2-yl](4-
flu or o-3-a m in op h en yl)m eth a n on e, viii-i. 8 g (0.019 mol)
of the dinitro intermediate viii-h was reacted with 24 g of Fe
and 4 g of ammonium chloride in a mixture of 200 mL of
dioxane, 200 mL of ethanol and 150 mL of water. After 1.5 h
at reflux, the reaction was filtered and evaporated. The residue
was dissolved in ethyl acetate, the organic solution was washed
with water, and dried over calcium sulfate. Evaporation of the
solvent gave a crude solid which was chromatographed (silica
gel, CH2Cl2-ethanol 95:5) to give the diamino compound (5.6
1
compound: mp 108 °C; IR (Nujol) 3100, 1671, 1597 cm-1; H
NMR (DMSO-d6) δ 0.9-1.8 (6H, m), 2.5 and 2.75 (2H, m), 3.98
and 4.35 (2H, m), 7.4 (4H, t), 8.1 (4H, m).
[3-(4-Flu or oben zoyl)bicyclo[2.2.2]oct-2-yl](4-flu or oph e-
n yl)m eth a n on e, viii-b: from vii-l; mp 128 °C; IR (Nujol)
1
1668, 1600-1508 cm-1; H NMR (CDCl3) δ 1.4-1.7 (6H, m),
1.9-2.1 (4H, m), 4.4 (2H, s), 7.1 (4H, m), 8.05 (4H, m).
[6-(4-F lu or ob e n zoyl)cycloh e x-3-a n yl](4-im id a zol-1-
ylp h en yl)m eth a n on e viii-c: from vii-f; IR (Nujol) 3300-
g, 81%): mp 167 °C; IR (Nujol) 3417, 3347, 1661, 1629 cm-1
;
1H NMR (CDCl3) δ 1.1-1.8 (6H, m), 2.5 (1H, br s), 2.70 (1H,
br s), 3.9 (4H, br s), 4.0 (1H, d), 4.3 (1H, m), 7.0 (2H, td), 7.4
(4H, m).
1
2400, 1620, 1580 cm-1; H NMR (DMSO-d6) δ 1.75-2.7 (8H,
2m), 7.15 (1H, s), 7.25 (2H, m), 7.6-7.85 (7H, m), 8.3 (1H, br
s).
P a th B2 (Sch em e 2). 2-Meth yl-1,3-d ip h en yl-4,7-d ih y-
d r o-2H-isoin d ole, ix-a . To 2.9 g (0.01 mol) of vii-a in 50 mL
of ethanol as added 7.75 mL (0.1 mol) of a 40% aqueous
methylamine solution, followed by 2.9 mL (0.05 mol) of acetic
acid. The reaction mixture was refluxed overnight. A white
solid precipitated on cooling, which was washed with ethanol
and pentane. After drying at 50 °C under 0.01 mmHg
overnight, 2.1 g (74%) of the desired compound was obtained:
[3-(4-F lu or ob en zoyl)b icyclo[2.2.1]h ep t -2-yl](4-m et h -
ylth iop h en yl)m eth a n on e, viii-d . 2.5 g (0.0073 mol) of
difluoro derivative viii-a was reacted in 50 mL of DMSO with
1 mol equiv of sodium thiomethoxyde (0.47 g, 0.0073 mol). The
reaction was heated at 80 °C for 2 h, then at room temperature
overnight. The reaction mixture was poured into 150 mL of
water and extracted with ethyl acetate. The organic layer was
mp 129 °C; IR (Nujol) 3075, 3057, 3025, 1602, 1491, 1075 cm-1
;

1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole Derivatives
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24 4591
1H NMR (CDCl3) δ 3.39 (4H, s), 3.49 (3H, s), 5.91 (2H, s), 7.26-
7.48 (10H, m).
2-Ben zyloxyca r bon yla m in o-1,3-d ip h en yl-4,7-d ih yd r o-
2H-isoin d ole, ix-b: from vii-a ; mp 184 °C; IR (Nujol) 3202,
1708, 1261 cm-1; 1H NMR (CDCl3) δ 3.26 (4H, s), 5.05 (2H, s),
5.90 (2H, s), 7.02-7.45 (15H, m).
reaction was stirred overnight; the solvents were evaporated,
and the residue was taken up with ethyl acetate. The organic
layers were washed with water, 10% citric acid and brine. The
solution was dried over calcium sulfate, and evaporated to give
1.6 g of a compound. Purification on silica gel (CH2Cl2-hexane
2:1) gave 340 mg (20%) of the desired compound: mp 196 °C;
IR (Nujol) 3251, 1326, 1146 cm-1; 1H NMR (CDCl3) δ 1.6-2.0
(4H, m), 2.2 (3H, s), 2.35-3.0 (4H, 2m), 7.2-7.5 (10H, m).
(1,3-Diph en yl-4,5,6,7-tetr ah ydr oisoin dol-2-yl)acetic acid,
16: from vii-a ; mp 162 °C; IR (Nujol) 2700-2300, 1679, 1602
1,3-Di-p-tolyl-4,7-d ih yd r o-2H-isoin d ole, ix-c: from vii-
1
c; mp 142 °C; IR (Nujol) 3466, 1610 cm-1; H NMR (DMSO-
d6) δ 2.35 (6H, s), 3.35 (4H, s), 5.9 (2H, s), 7.2 (4H, d), 7.55
(4H, d), 10.8 (1H, s).
1,3-Bis(4-m eth oxyph en yl)-4,7-dih ydr o-2H-isoin dole, ix-
d : from vii-d ; mp 155 °C; IR (Nujol) 3438, 1600 cm-1; 1H NMR
(DMSO-d6) δ 3.3 (4H, m), 3.75 (6H, s), 5.9 (2H, s), 6.95 (4H,
d), 7.55 (4H, d), 10.7 (1H, br s).
1
cm-1; H NMR (CH3OH-d4) δ 1.75 (4H, m), 2.55 (4H, m), 4.3
(2H, s), 7.2-7.4 (10H, m).
1,3-Di-p-tolyl-4,5,6,7-tetr a h yd r o-2H-isoin d ole, 19: from
ix-c; mp 150 °C; IR (Nujol) 3475, 1611 cm-1; 1H NMR (DMSO-
d6) δ 1.70 (4H, m), 2.32 (6H, s), 2.70 (4H, m), 7.18 (4H, d),
7.48 (4H, d), 10.70 (1H, br s).
1,3-Bis(4-ch lor op h en yl)-4,7-d ih yd r o-2H-isoin d ole, ix-
1
e: from vii-e; IR (Nujol) 3460, 1594 cm-1; H NMR (DMSO-
1,3-Di(4-m eth oxyp h en yl)-4,5,6,7-tetr a h yd r o-2H-isoin -
d ole, 20: from ix-d ; mp 146 °C; IR (Nujol) 3421-3396, 1600-
d6) δ 3.4 (4H, s), 5.9 (2H, s), 7.45 (4H, d), 7.7 (4H, d), 11.1 (1H,
br s).
1
1587 cm-1; H NMR (DMSO-d6) δ 1.70 (4H, m), 2.65 (4H, m),
1,3-Bis[4-(im id a zol-1-yl)p h en yl]-4,7-d ih yd r o-2H-isoin -
d ole, ix-f: from vii-g; mp > 260 °C; IR (Nujol) 1510 cm-1; 1H
NMR (DMSO-d6) δ 3.45 (4H, d), 5.94 (2H, m), 7.12 (2H, s),
7.70 (4H, d), 7.78 (6H, d), 8.29 (2H, s), 11.15 (2H, s).
3,5-D i p h e n y l-4-a z a t r i c y c lo [5.2.1.02 ,6 ]d e c a -2,5,8-
tr ien e, ix-g. This compound is described in ref 17.
1,3-Dip h en yl-5,6-d im et h yl-4,7-d ih yd r o-2H -isoin d ole,
3.8 (6H, s), 6.95 (4H, d), 7.5 (4H, d), 10.55 (1H, br s).
1,3-Di(4-ch lor op h e n yl)-4,5,6,7-t e t r a h yd r o-2H -isoin -
d ole, 21: from ix-e; mp 231 °C; IR (Nujol) 3462, 1593, 830
1
cm-1; H NMR (DMSO-d6) δ 1.75 (4H, m), 2.70 (4H, m), 7.46
(4H, d), 7.62 (4H, d), 10.95 (1H, br s).
1-(4-F lu or op h en yl)-3-[4-(im id a zol-1-yl)p h en yl]-4,5,6,7-
tetr a h yd r o-2H-isoin d ole, 22: from viii-c; mp > 250 °C; IR
25: from vii-a ; mp 212 °C; IR (Nujol) 3459, 1601 cm-1 1H
;
NMR (DMSO-d6) δ 1.8 (6H, s), 3.3 (4H, s), 7.2 (2H, t), 7.45
(4H, m), 7.65 (4H, m), 10.9 (1H, br s).
1
(Nujol) 3300-2400, 1620-1580 cm-1; H NMR (DMSO-d6) δ
1.75-2.7 (8H, 2m), 3.4 (4H, m), 5.95 (2H, s), 7.15 (1H, s), 7.25
(2H, m), 7.6-7.85 (7H, m), 8.3 (1H, s), 11.0 (1H, s).
1,3-Di[4-(im id a zol-1-yl)p h en yl]-4,5,6,7-tetr a h yd r o-2H-
isoin d ole h yd r och lor id e, 23: from ix-f; mp > 260 °C; IR
(Nujol) 3700-2000, 1604 cm-1; 1H NMR (DMSO-d6) δ 1.70 (4H,
m), 2.75 (4H, m), 7.05 (2H, s), 7.4 (4H, d), 7.65 (2H, s), 7.75
(4H, d), 8.15 (2H, s).
1,3-Dip h e n yl-2,5,6-t r im e t h yl-4,7-d ih yd r o-2H -isoin -
d ole, 26. 3.18 g (0.01 mol) of vii-h was dissolved in 25 mL of
THF and 0.6 g of acetic acid was added, followed by 8 mL of a
40% aqueous methylamine solution. The reaction was heated
at 60 °C for 18 h. The solvents were evaporated and the residue
was taken up with ethanol. Filtration of the resulting solid
and washing with ethanol and pentane gave 2.95 g (94%) of
the desired compound: mp 200-201 °C; IR (Nujol) 1601, 1576,
1,3-Dip h en yl-5,6-d im eth yl-4,5,6,7-tetr a h yd r o-2H-isoin -
d ole, 24: from 25; mp 133 °C; IR (Nujol) 3438, 1603 cm-1; 1H
NMR (CDCl3) δ 0.85-1.2 (6H, 2d), 2.0 (2H, m), 2.5 (2H, dd),
2.85 (2H, dd), 7.2 (2H, m), 7.4 (4H, m), 7.65 (4H,d), 10.8 (1H,
br s).
1
1496 cm-1; H NMR (CDCl3) δ 1.74 (6H, s), 3.21 (4H, s), 3.50
(3H, s), 7.25-7.5 (10H, m).
3,5-Bis(4-flu or op h en yl)-4-a za t r icyclo[5.2.1.02,6]d eca -
2,5,8-tr ien e, 33. 28 g (0.08 mol) of intermediate vii-j was
refluxed for 24 h with 25 g of ammonium acetate in 400 mL of
2-propanol. The solvent was evaporated, and the residue was
taken up with ethyl acetate. The organic layers were washed
with a 10% solution of sodium bicarbonate, brine, dried over
calcium sulfate, and evaporated. The crude solid was recrys-
tallized from anhydrous ethanol to give 19.3 g (73%) of the
desired compound: mp 189 °C; IR (Nujol) 3442 cm-1; 1H NMR
(DMSO-d6) δ 2.30-2.40 (2H, 2d), 4.05 (2H, s), 6.75 (2H, s),
7.20 (4H, t), 7.70 (4H, dd), 10.0 (1H, s).
3, 5-Dip h en yl-4-a za tr icyclo[5.2.1.02,6]d eca -2, 5-d ien e,
31: from ix-g; mp 162 °C; IR (Nujol) 3328, 1616-1594 cm-1
;
1H NMR (CDCl3) δ 1.38 (2H, m), 1.72 (1H, d), 2.0 (3H, m),
3.60 (2H, m), 7.20 (1H, t), 7.40 (2H, t), 7.55 (2H,d), 7.90 (1H,
br s).
3,5-Bis(4-flu or op h en yl)-4-a za t r icyclo[5.2.1.02,6]d eca -
2,5-d ien e, 32: from viii-a or 33; mp 146 °C; IR (Nujol) 3436,
1
1223, 833 cm-1; H NMR (DMSO-d6) δ 1.2 (2H, d),1.70 (1H,
d), 1.90 (1H, d), 1.95 (2H, d), 3.55 (2H, s), 7.20 (4H, t), 7.70
(4H, dd), 10.35 (1H, br s).
3,5-D ip h e n y l-4-a za t r ic y c lo [5.2.2.02,6]u n d e c a -2,5,8-
tr ien e, 38: from vii-l; mp 224 °C; IR (Nujol) 3413, 1616, 1593
cm-1; 1H NMR (DMSO-d6) δ 1.6 (4H, m), 4.2 (2H, m), 6.55 (2H,
m), 7.2 (2H, m), 7.45 (4H, t), 7.7 (4H, d), 10.4 (1H, br s).
1,3-Dip h e n yl-2-m e t h yl-4,5,6,7-t e t r a h yd r o-2H -isoin -
d ole, 13. 1.4 g (0.0049 mol) of ethylenic compound ix-a was
hydrogenated in 100 mL of ethanol for 3 h at reflux in the
presence of 0.5 g of 10% Pd/C and 1.85 g of ammonium formate.
After cooling, the reaction mixture was filtered, and the filtrate
was evaporated. The residue was taken up in ethyl acetate,
washed with a 10% aqueous sodium hydrogenocarbonate
solution, and water. Drying over calcium sulfate was followed
by filtration and evaporation to give a solid that was recrystal-
lized from ethanol to give 1 g (71%) of the saturated com-
3-(4-F lu o r o p h e n y l)-5-(4-m e t h y ls u lfo n y lp h e n y l)-4-
a za tr icyclo[5.2.1.02,6]d eca -2,5-d ien e, 34: from viii-d ; mp >
260 °C; IR (Nujol) 3360, 1141 cm-1; 1H NMR (DMSO-d6) δ 1.2
and 1.65-2.05 (6H, m), 3.2 (3H, s), 3.65 (2H, m), 7.3 (2H, m),
7.7-8.0 (6H, m), 10.65 (1H, br s).
3,5-Bis(4-m eth ylsu lfon ylph en yl)-4-azatr icyclo[5.2.1.02,6]-
d eca -2,5-d ien e, 35: from viii-g; mp > 260 °C; IR (Nujol) 3342,
1
1588, 1313-1144 cm-1; H NMR (DMSO-d6) δ 1.2-1.75-2.0
(6H, m), 3.35 (6H, s), 3.75 (2H, m), 7.98 (8H, m), 10.88 (1H, br
s).
3,5-Dip h e n yl-4-a za t r icyclo[5.2.2.02,6]u n d e ca -2,5-d i-
en e, 36: from 38; mp 248 °C; IR (Nujol) 3420, 1616-1595 cm-1
;
1H NMR (DMSO-d6) δ 1.4 (4H, m), 1.8 (4H, m), 3.35 (2H, m),
7.15 (2H, m), 7.4 (4H, m), 7.65 (4H, d), 10.6 (1H, br s).
3,5-Bis(4-flu or op h en yl)-4-a za tr icyclo[5.2.2.02,6]u n d eca -
2,5-d ien e, 37: from viii-b; mp 202 °C; IR (Nujol) 3457, 1508
cm-1; 1H NMR (DMSO-d6) δ 1.4 (4H, m), 1.8 (4H, m), 3.3 (2H,
s), 7.2 (4H, dd), 7.65 (4H, dd), 10.6 (1H, br s).
1
pound: mp 143 °C; IR (Nujol) 1598 cm-1; H NMR (CDCl3) δ
1.75 (4H, m), 2.65 (4H, m), 3.45 (3H, s), 7.2-7.5 (10H, m).
1,3-Dip h e n yl-2-a m in o-4,5,6,7-t e t r a h yd r o-2H -isoin -
d ole, 14: from ix-b; mp 161 °C; IR (Nujol) 3357-3224, 3041,
1599 cm-1; 1H NMR (CDCl3) δ 1.76 (4H, m), 2.63 (4H, m), 4.42
(2H, br s), 7.25-7.49 (10H, m).
1,3-Dip h en yl-4,5,6,7-tetr a h yd r oisoin d ole-2-m eth ylsu l-
fon yla m in e, 15. Compound 14 (1.3 g, 0.0045 mol) was
dissolved in 30 mL pyridine and mesyl chloride (0.35 mL,
0.0045 mol) was added dropwise in 5 mL of dichloromethane.
After 2 h, 0.35 mL of mesyl chloride was added, and the
3,5-Bis(3-a ce t a m id o-4-flu or op h e n yl)-4-a za t r icyclo-
[5.2.1.02,6]d eca -2,5-d ien e, 39: obtained from viii-i by cycliza-
tion and acetylation with ammonium acetate in acetic acid as
described above for compound 33; mp 225 °C; IR (Nujol) 3290,
1673, 1608-1530 cm-1; 1H NMR (DMSO-d6) δ 1.2 (2H, m), 1.65
(1H, d), 1.75-2.0 (3H, m), 2.1 (6H, s), 3.6 (2H, s), 7.2 (2H, dd),
7.45 (2H, m), 8.2 (2H, d), 9.7 (2H, br s), 10.3 (1H, br s).

4592 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24
Portevin et al.
3,5-Bis(3-for m a m id o-4-flu or op h e n yl)-4-a za t r icyclo-
[5.2.1.02,6]d eca -2,5-d ien e: IR (Nujol) 3284, 1691, 1609, 1528
cm-1; 1H NMR (DMSO-d6) δ 1.0-2.0 (4H, m), 3.55 (2H, s), 7.2-
7.4 (4H, m), 8.4 (4H, m), 10.1 (3H, br s).
The residue was washed with diisopropyl ether to give a yellow
solid (5.4 g, 96%): mp 171 °C; IR (Nujol) 3062-3047, 1685
cm-1 1H NMR (CDCl3) presence of endo/exo isomers δ 1.55
;
(m) and 1.85 (d) (2H), 3.15 (m) and 3.40 (m) (2H), 3.9 (d) and
4.48 (t) (2H), 5.94 (m) and 6.45 (m) (2H), 7.80 (4H, 2d), 8.83
(4H, d).
3,5-Bis(3-m eth yla m in o-4-flu or op h en yl)-4-a za tr icyclo-
[5.2.1.02,6]d eca -2,5-d ien e Hyd r och lor id e, 40. 8.14 g (0.02
mol) of the diformylated intermediate obtained above was
dissolved in 150 mL of THF and added dropwise to a suspen-
sion of 1.52 g (0.04 mol) of LAH in 150 mL of THF. The reaction
was refluxed for 1.5 h and after cooling was hydrolyzed with
10 mL of 2-propanol and 6 mL of brine. Filtration and
evaporation of the filtrate gave a crude residue that was
chromatographed (silica gel, CH2Cl2). To the purified solid
dissolved in 10 mL of diethyl ether was added 10 mL of 4 M
HCl in dioxane. The resulting solid was filtered, washed with
diethyl ether and dried to give 2.9 g (37%) of the title
compound: mp 196 °C; IR (Nujol) 3254, 2700-2300, 1619,
3,5-Dip yr id in -4-yl-4-a za t r icyclo[5.2.1.02,6]d eca -2,5,8-
tr ien e, ix-h . 4.8 g (0.016 mol) of compound xvii and 12.2 g
(0.016 mol) of ammonium acetate were refluxed in 50 mL of
acetic acid. After 3 h, the solvent was evaporated, water was
added, and the reaction mixture was alkalinized with potas-
sium carbonate. The resulting solid was filtered and washed
with water. Chromatography (silica gel, CH2Cl2-CH3OH 1:1)
gave a yellow solid (2.4 g, 53%) which was identified as the
desired compound: mp > 250 °C; IR (Nujol) 3136, 1607-1592
1
cm-1; H NMR (DMSO-d6) 2.4 (2H, dd), 4.3 (2H, s), 6.8 (2H,
s), 7.7 (4H, d), 8.5 (4H, d), 10.5 (1H, br s).
1
1581, 1519 cm-1; H NMR (DMSO-d6) δ 1.2 (2H, m), 1.7 (1H,
3,5-Dip yr id in -4-yl-4-a za t r icyclo[5.2.1.02,6]d eca -2,5-d i-
en e Dih yd r och lor id e, 41. This compound was prepared
according to the method used for compound 13. The dihydro-
chloride was obtained after dissolving the pure base in
dioxane-methanol 1:1 and adding dropwise a solution of 4 M
HCl in dioxane: mp 264 °C; IR (Nujol) 3400-2500, 1624,
d), 1.8-1.95 (3H, m), 2.9 (6H, s), 3.6 (2H, s), 7.1-7.4 (6H, m),
7.6 (2H, br s), 10.5 (1H, br s).
R ou t e C (Sch em e 3). (4-Ben zoyl-1-b en zylp yr r olid in -
3-yl)(p h en yl)m eth a n on e, xii. 4.75 g (0.02 mol) of N-benzyl-
N-methoxymethyl-N-[(trimethylsilyl)methyl]amine (xi; Ald-
rich) was reacted with 4.72 g (0.02 mol) of dibenzoylethylene
and 0.65 g (0.025 mol) of lithium fluoride in 50 mL of
acetonitrile. The temperature was maintained at 40 °C for 6
h, then the reaction was stirred at room temperature for 18
h. The reaction mixture was then poured into 200 mL of water.
Extraction with ethyl acetate was followed by washing with
water and brine. Drying over calcium sulfate and evaporation
gave 7.4 g of an oil which was chromatographed (silica gel,
CH2Cl2-AcOEt 1:1) to give 6.2 g (84%) of compound xii as a
1602-1586 cm-1 1H NMR (DMSO-d6) 1.2 (2H, m), 1.8 (1H,
;
m), 2.1 (3H, m), 3.9 (2H, s), 8.5 (4H, d), 8.8 (4H, d), 12.5 (1H,
br s).
Ack n ow led gm en t. Dr. J ean-Paul Volland and his
group are thanked for analytical and spectral studies.
Refer en ces
(1) Kujubu, D. A.; Fletcher, B. S.; Varnum, B. C.; Lim, R. W.;
Herschman, H. R. TIS10, a phorbol ester tumor promoter-
inducible mRNA from Swiss 3T3 cells, encodes a novel prostag-
landin synthase/cyclooxygenase homologue. J . Biol. Chem. 1991,
266, 12866-12872.
(2) Xie, W.; Chipman, J . G.; Robertson, D. L.; Erikson, R. L.;
Simmons, D. L. Expression of a mitogen-response gene encoding
prostaglandin synthase is regulated by mRNA splicing. Proc.
Natl. Acad. Sci. U.S.A. 1992, 89, 3917-3921.
(3) Parnham, M. J . Is there a COX-fight during inflammation?
Inflamm. Res. 1998, 47, 43.
(4) Geng, Y.; Blanco, F. J .; Cornellison, M.; Lotz, M. Regulation of
cyclooxygenase-2 expression in normal human articular chon-
drocytes. J . Immunol. 1995, 155, 796-801.
(5) Chan, C.-C.; DiBattista, J . A.; Zhao, J .; Pelletier, J .-P.; Martel-
Pelletier, J .; Kennedy, B. P.; Brideau, C.; Rodger, I. W. Induction
of cyclooxygenase-2 (COX-2) in human articular chondrocytes
and inhibition of PGE2 synthesis by a selective COX-2 inhibitor.
International Association of Inflammation Society, Brighton,
Sept 17-22, 1995; S275.
(6) Amin, A. R.; Attur, M.; Patel, R. N.; Thakker, G. D.; Marshall,
P. J .; Rediske, J .; Stuchin, S. A.; Patel, I. R.; Abramson, S. B.
Superinduction of cyclooxygenase-2 activity in human osteoar-
thritis-affected cartilage. J . Clin. Invest. 1997, 99, 1231-1237.
(7) Davis, R.; Brogden, R. N. Nimesulide. An update of its phar-
macodynamic and pharmacokinetic properties, and therapeutic
efficacy. Drugs 1994, 48, 431-454.
(8) Futaki, N.; Takahashi, S.; Yokoyama, S.; Arai, I.; Higuchi, S.;
Otomo, S. NS-398, a new antiinflammatory agent, selectively
inhibits prostaglandin G/H synthase/cyclooxygenase (COX-2)
activity in vitro. Prostaglandins 1994, 47, 55-59.
(9) Wiesenberg-Boettcher, I.; Schweizer, A.; Mu¨ller, K. The phar-
macological profile of CGP 28238, a highly potent antiinflam-
matory compound. Agents Actions 1989, 26, 240-242.
(10) Li, C.-S.; Black, W C.; Chan, C.-C.; Ford-Hutchinson, A. W.;
Gauthier, J .-Y.; Gordon, R.; Guay, D.; Kargman, S.; Lau, C. K.;
Mancini, J .; Ouimet, N.; Roy, P.; Vickers, P.; Wong, E.; Young,
R. N.; Zamboni, R.; Prasit, P. Cyclooxygenase-2 inhibitors.
Synthesis and pharmacological activities of 5-methanesulfona-
mido-1-indanone derivatives. J . Med. Chem. 1995, 38, 4897-
4905.
(11) Penning, T. D.; Talley, J . J .; Bertenshaw, S. R.; Carter, J . S.;
Collins, P. W.; Docter, S.; Graneto, M. J .; Lee, L. F.; Malecha, J .
W.; Miyashiro, J . M.; Rogers, R. S.; Rogier, D. J .; Yu, S. S.;
Anderson, G. D.; Burton, E. G.; Cogburn, J . N.; Gregory, S. A.;
Koboldt, C. M.; Perkins, W. E.; Seibert, K.; Veenhuizen, A. W.;
Zhang, Y. Y.; Isakson, P. C. Synthesis and biological evaluation
of the 1,5-diarylpyrazole class of cyclooxygenase-2 inhibitors:
identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-
pyrazol-1-yl] benzenesulfonamide (SC-58635, celecoxib). J . Med.
Chem. 1997, 40, 1347-1365.
colorless oily residue: IR (Nujol) 1680, 1581, 1620-1579 cm-1
;
1H NMR (CDCl3) δ 2.8-3.12 (2H, dd) 3.7 (2H, s), 4.65 (2H,
m), 7.15-7.3 (5H, m), 7.4 (4H, m), 7.5 (2H, m), 7.92 (4H, d).
2-Ben zyl-4,6-d ip h en yl-1,2,3,5-tetr a h yd r op yr r olo[3,4-c]-
p yr r ole, 29: cyclization performed according to the method
used for compound 33; mp 182 °C; IR (Nujol) 3429, 1673,
1610-1595 cm-1; 1H NMR (DMSO-d6) δ 3.9 (4H, s), 4.05 (2H,
s), 7.15 (2H, m), 7.2-7.5 (9H, m), 7.6 (4H, d).
4,6-Dip h en yl-1,2,3,5-tetr a h yd r op yr r olo[3,4-c]p yr r ole,
30: prepared according to the method used for compound 13;
mp >260 °C; IR (Nujol) 3303, 2750-2400, 1595 cm-1; 1H NMR
(DMSO-d6) δ 4.55 (4H, s), 7.25 (2H, m), 7.45 (4H, m), 7.7 (4H,
d), 10.25 (1H, br s), 11.45 (1H, br s).
Rou te D (Sch em e 4). 1-P yr id in -4-yl-2-(tr ip h en yl-λ5-
p h osp h a n ylid en e)eth a n on e, xv. Triphenylphosphine (3.25
g, 0.0125 mol) was dissolved in 40 mL of toluene. 4-Bro-
moacetylpyridine xiv (3.5 g, 0.0125 mol) was then added,
followed by triethylamine (1.73 mL, 0.0125 mol) dropwise.
After stirring for 2 h, the reaction mixture was filtered; the
solid was washed with ethyl acetate and ethanol to give 3.1 g
(65%) of a pale brown solid: mp 224 °C; IR (Nujol) 1577, 1515
cm-1; 1H NMR (DMSO-d6) δ 4.6 (1H, d), 7.5-7.8 (17H, m), 8.6
(2H, d).
1,4-Dip yr id in -4-ylbu t-2-en e-1,4-d ion e, vi-a . 30 g (0.079
mol) of phosphorane xv was dissolved in 200 mL of dichlo-
romethane, and 30 g of 4A molecular sieves was added. A
solution of 37% peracetic acid in acetic acid was diluted with
100 mL of dichloromethane and added dropwise at 5 °C within
45 min. After stirring 15 more min, the solution was filtered
and evaporated. The residue was taken up with dichlo-
romethane and a 10% solution of sodium hydrogenocarbonate.
The organic layer was washed with water, dried over calcium
sulfate and evaporated. The solid was taken up with ethanol,
filtered and washed with pentane to give 4.2 g (45%) of a brown
solid: mp 222 °C; IR (Nujol) 3086, 3048, 1659, 1556 cm-1; 1H
NMR (DMSO-d6) δ 7.9 (2H, s), 7.95 (4H, d), 8.9 (4H, d).
[3-(P yr id in -4-ylca r bon yl)bicyclo[2.2.1]h ep t-5-en -2-yl]-
p yr id in -4-ylm eth a n on e, vii-m . 4.4 g (0.018 mol) of diketone
obtained above was dissolved in 80 mL of toluene and 4 g
(0.061 mol) of freshly distilled cyclopentadiene was added. The
reaction mixture was heated at 120 °C for 2.5 h in a sealed
vessel. After cooling, the reaction was filtered and evaporated.

1,3-Diaryl-4,5,6,7-tetrahydro-2H-isoindole Derivatives
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 24 4593
(12) Black, C.; Grimm, E.; Leger, S.; Prasit, P.; Wang, Z. Diaryl-2-
(5H)-furanones as COX-2 inhibitors. World Patent WO 9619469,
1995.
(13) Wallace, J .; Chin, B. Celecoxib. Curr. Opin. Central Peripheral
Nerv. Syst. 1999, 1, 132-141.
(14) Kamali, F. Rofecoxib. Curr. Opin. Central Peripheral Nerv. Syst.
1999, 1, 126-131.
(18) Kobayashi, T.; Ono, K.; Suda, H.; Yamashita, Y.; Kato, H.
Syntheses and spectral properties of norbornadiene-fused het-
erocycles: 1,3-diphenyl-4,7-dihydro-4,7-methanobenzo[c]thiophene,
-4,7-methano-2-H-isoindole, and -4,7-methanoisobenzofuran. Bull.
Chem. Soc. J pn. 1993, 66, 6, 2707-2713.
(19) Padwa, A.; Dent, W. On the use of N-[(trimethylsilyl)methyl]-
amino ethers as capped azomethine ylide equivalents. J . Org.
Chem. 1987, 52, 235-244.
(20) Taurins, A.; Blaga, A. Synthesis of pyridyl- and quinolyl-
substituted 2-aminothiazoles. J . Heterocycl. Chem. 1970, 7,
1137-1141.
(21) Denney, D. B.; Smith, L. C.; Song, J .; Rossi, C. J .; Hall, C. D.
Reactions of phosphoranes with peracids. J . Org. Chem. 1963,
28, 778-780.
(22) Patrignani, P.; Panara, M. R.; Greco, A.; Fusco, O.; Natoli, C.;
Iacobelli, S.; Cipollone, F.; Ganci, A.; Creminon, C.; Maclouf, J .;
Patrono, C. Biochemical and pharmacological characterization
of the cyclooxygenase activity of human blood prostaglandin
endoperoxyde synthases. J . Pharmacol. Exp. Ther. 1994, 271,
1705-1712.
(23) Edwards, J . C. W.; Sedgewick, A. D.; Willoughby, D. A. The
formation of a structure with the features of synovial lining by
subcutaneous injection of air: an in vivo tissue culture system.
J . Pathol. 1981, 134, 147-153.
(24) Bonnet, J .; Zerath, E.; Picaud, N.; Lesur, C.; Mattio, A.; Tordj-
man, C.; Hott, M.; Marie, P. J . Bone morphometric changes in
adjuvant-induced polyarthritic osteopenia in rats: evidence for
an early bone formation defect. J . Bone Miner. Res. 1993, 8,
659-668.
(15) (a) Barnett, J . W.; Dunn, J . P.; Kertesz, D. J .; Miller, A. B.;
Morgans, D. J .; Ramesha, C. S.; Sigal, C. E.; Sjogren, E. B.;
Smith, D. B.; Talamas, F. X. Pyrrole derivatives. European
Patent EP 0 714 895, 1995. (b) Black, W. C.; Bayly, C.; Belley,
M.; Chan, C.-C.; Charleson, S.; Denis, D.; Gauthier, J . Y.;
Gordon, R.; Guay, D.; Kargman, S.; Lau, C. K.; Leblanc, Y.;
Mancini, J ; Ouellet, M.; Percival, D.; Roy, P.; Skorey, K.; Tagari,
P.; Vickers, P.; Wong, E.; Xu, L.; Prasit, P. From indomethacin
to a selective COX-2 inhibitor: development of indolalkanoic
acids as potent and selective cyclooxygenase-2 inhibitors. Bioorg.
Med. Chem. Lett. 1996, 6, 725-730. (c) Leblanc, Y.; Black, W.
C.; Chan, C.-C.; Charleson, S.; Delorme, D.; Denis, D.; Gauthier,
J . Y.; Grimm, E. L.; Gordon, R.; Guay, D.; Hamel, P.; Kargman,
S.; Lau, C. K.; Mancini, J ; Ouellet, M.; Percival, D.; Roy, P.;
Skorey, K.; Tagari, P.; Vickers, P.; Wong, E.; Xu, L.; Prasit, P.
Synthesis and biological evaluation of both enantiomers of
L-761,000 as inhibitors of cyclooxygenase 1 and 2. Bioorg. Med.
Chem. Lett. 1996, 6, 731-736.
(16) Tordjman, C.; Coge´, F.; Andre´, N.; Rique, H.; Spedding, M.;
Bonnet, J . Characterisation of cyclooxygenase 1 and 2 expression
in mouse resident peritoneal macrophages in vitro; interactions
of non steroidal antiinflammatory drugs with cox 2. Biochim.
Biophys. Acta 1995, 1256, 249-256.
(17) Adams, R.; Gold, M. H. The synthesis of 1,3-diphenyldihy-
drobenzofurans, 1,3-diphenylisobenzofurans and o-dibenzoyl-
benzenes from the diene addition products to dibenzoylethylene.
J . Org. Chem. 1940, 62, 56-61.
J M990965X