New ligands at the?melatonin binding site MT3

Article abstract of DOI:10.1016/j.ejmech.2005.12.002

The third melatonin binding site, MT3 is a non-classical one since it is not a seven transmembrane domains receptor, but an enzyme, quinone reductase 2. A major concern for the study of the physiological role of this site is the lack of specific ligands, permitting to more accurately dissect the pathways linked to the activation of MT3. Indeed, in the course of finding new ligands, we identified a new series of compounds with affinity to the binding site in the nM range, particularly 2,3-dimethoxy 7-hydroxy 10-methyl 5H 10H indeno(1,2-b)indol-10-one (DMHMIO), with a Ki of 190 pM. Based on slightly different and novel synthons compared to most of the compounds used in melatonin pharmacology studies, these compounds offer new perspective for the description of the melatonin pathways, so much more by not having any affinity towards the MT1 and MT2 'classical' melatonin receptors.

Full text of DOI:10.1016/j.ejmech.2005.12.002

European Journal of Medicinal Chemistry 41 (2006) 306–320
http://france.elsevier.com/direct/ejmech
Original article
New ligands at the melatonin binding site MT3
Marie-Françoise Boussard a, Sandrine Truche a, Anne Rousseau-Rojas a, Sylvie Briss a,
Sophie Descamps a, Monique Droual a, Michel Wierzbicki a, Gilles Ferry b, Valérie Audinot b,
Philippe Delagrange c, Jean A. Boutin b,*
a Division de Chimie médicinale E, Institut de Recherches SERVIER, 11, rue des Moulineaux, 92150 Suresnes, France
b Pharmacologie moléculaire et cellulaire, Institut de Recherches SERVIER, 125, chemin de Ronde, 78290 Croissy-sur-Seine, France
c Département des Sciences Expérimentales, Institut de Recherches SERVIER, 11, rue des Moulineaux, 92150 Suresnes, France
Received 10 October 2005; received in revised form 9 December 2005; accepted 12 December 2005
Available online 18 January 2006
Abstract
The third melatonin binding site, MT3 is a non-classical one since it is not a seven transmembrane domains receptor, but an enzyme, quinone
reductase 2. A major concern for the study of the physiological role of this site is the lack of specific ligands, permitting to more accurately
dissect the pathways linked to the activation of MT3. Indeed, in the course of finding new ligands, we identified a new series of compounds with
affinity to the binding site in the nM range, particularly 2,3-dimethoxy 7-hydroxy 10-methyl 5H 10H indeno(1,2-b)indol-10-one (DMHMIO),
with a Ki of 190 pM. Based on slightly different and novel synthons compared to most of the compounds used in melatonin pharmacology
studies, these compounds offer new perspective for the description of the melatonin pathways, so much more by not having any affinity towards
the MT1 and MT2 ‘classical’ melatonin receptors.
© 2006 Elsevier SAS. All rights reserved.
Keywords: Melatonin; MT3 binding site; Ligands; Binding; Quinone reductase 2
1. Introduction
melatonin. In addition, there is evidence for a nanomolar mel-
atonin binding site in Hamster brain [8,9] and kidney [9,10],
called MT3. Its specific ligands over MT1 and MT2 known to
date include mainly 5-methoxycarbonylamino-N-acetyltrypta-
mine (MCA-NAT) [11] and prasozin [9,11]. Besides its origi-
nal pharmacology, MT3 is characterized by fast kinetics of li-
gand association/dissociation [8,10,11], raising difficulties for
affinity measurements. We recently stressed the need for more
specific chemical tools in order to better understand the char-
acteristics and the role of this binding site [3]. We reported
several times on the search of new ligands to MT3 [12,13].
These compounds were essentially bicyclic chemicals that
loosely resembled the indolic melatonin.
Melatonin is an indole-derived neurohormone of long stand-
ing interest which is produced in the pineal gland and is de-
rived from serotonin [1]. Because of its nocturnal synthesis,
melatonin is suspected to relay the circadian rhythm and the
information on the photoperiod to the peripheral organs for
daily and seasonal physiological regulations [2,3]. Several cel-
lular targets of melatonin have been detected since 1986, after
the synthesis of 2-iodomelatonin [4], a very potent melatonin
agonist that was rapidly used as 2-[125I]-iodomelatonin. These
melatonin binding sites MT1 [5], MT2 [6] and a Mel1c (only
present in X. laevis [7]) have been cloned. They are G-protein
coupled seven-transmembrane domain receptors and are char-
acterized by subnanomolar affinities for melatonin and 2-iodo-
In the present work, we report a series of new chemical
entities as potent ligands of MT3. They are indeno- and isoin-
dolo-indol-10-one derivatives. These tetracyclic compounds
are quite far from the previous reference compounds at this
binding site and present affinities for MT3 in the low nanomo-
Abbreviations: DMF, dimethylformamide; DMSO, dimethylsulfoxide.
* Corresponding author. Tel.: +33 1 55 72 27 48; fax: +33 1 55 72 28 10.
E-mail address: jean.boutin@fr.netgrs.com (J.A. Boutin).
0223-5234/$ - see front matter © 2006 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.12.002

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
307
lar range. These tools are now ready for pharmacological and
physiopathological studies on classical melatonin models.
sation with a suitable 5-substituted 2-nitro benzaldehyde 6
gives the benzylidene phtalides 7 as a mixture of Z and E iso-
mers. Compounds 7 can be either catalytically reduced to
amines 8, or rearranged to indan-1,3-diones 11 by action of a
strong base as similarly previously described by Eskola and
Hyrkko [16] and by Estevez et al. [17]. Aminobenzylidene lac-
tones 8 are rearranged to 2-carboxyphenyl indoles 9 under the
action of a strong base and then cyclized to give isoindolo in-
doles 10 by elimination of water and deprotected to com-
pounds 1 by hydrogenolysis when necessary. Indan-1,3-diones
11 are reduced and spontaneously cyclized to give indeno in-
doles 12 in a one step procedure. Compound 12 was finally
deprotected to compound 2 by hydrogenolysis. One easily un-
derstands that compound 11 lead to a single derivative if sub-
stitution of 11 is symmetrical and can lead to a mixture of
compound 12 if the substitution is not symmetrical, due to a
possible attack of the amine (resulting from nitro reduction) on
both carbonyl functions, necessitating a chromatographic se-
paration. In some case, substitution can induce the preferential
2. Chemistry
Indeno (1,2-b) indol-10-one and isoindolo (2,1-a) indol-10-
one have been already described in the literature [14]. Their
syntheses start from 3-(o-iodo benzoyl) indole and 1-(o-iodo
benzoyl) indole, respectively. It is difficult to prepare substi-
tuted iodobenzoyl derivatives as some migration of the benzoyl
moiety leads to non-selective cyclization, making this method
unusable in the present case.
Compounds 1 and 2 have been prepared by mean of a
shared intermediate, following the scheme described in
Fig. 1. Phtalides 3 are brominated in carbon tetrachloride or
dichloromethane using N-bromosuccinimide following a very
classical process [15]. Crude bromophtalides 4 are then substi-
tuted by triphenylphosphine to give Wittig reagents 5. Conden-
Fig. 1. General synthetic pathway for the obtention of compounds 1 and 2.

308
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
formation of one isomer. One could have tried to transform in a
one step procedure, phtalides 3 to 2-phenyl indane-1,3-diones
11 by reaction with the appropriate benzaldehyde under the
action of MeONa, as described by Shapiro et al. [19]. We
needed to isolate the nitrobenzylidenes 7 as key intermediates
to produce either isoindolo-indoles 1 or indeno-indoles 2. In
these conditions of Shapiro et al. [19], intermediates 7, prob-
ably formed during the reaction, cannot be isolated and there-
fore spontaneously evolved to compound 11.
low yield, after N-methylation, to give 18 in an attempt to pre-
pare a radioiodinated ligand for MT3 receptors, by using the so-
called Kirsch and Sluzarek reagent as described by Boussard et
al. [22]. We also have access to reduced indeno-indoles 17
directly from the O-benzylated derivatives 12. This is achieved
by a simple modification of the debenzylation solvent (acetic
acid instead of NaOH, in DMF/methanol mixture). Some of
these transformations are very efficient, some others give very
poor yield, but the products have been isolated and character-
ized to be tested in the pharmacological assay. All these trans-
formations are listed in Fig. 2. A particular case (formation of
compound 14d from a different intermediate) is mentioned be-
low in Fig. 3. Treatment with methanol of the crude Wittig
reaction mixture resulting from 5g and 6x, leads to partial
methanolyis of the lactone as indicated in Fig. 3. The isolations
of compounds 7gx (expected) and 13 were achieved with low
yield, for characterization. Further reduction of 13 with Raney
nickel followed by base treatment led directly to the expected
di-hydroxy isoindoloindole 14d described with others polyphe-
nolic compounds in Table 10.
Methoxy-substituted compound 1 or 2 can be demethylated
to give the polyphenolic derivatives 14 and 15, respectively.
Selective monodemethylation of polymethoxylated derivatives
can be achieved with a low to moderate yield if the methoxy
substituent is ortho or para to a carbonyl group (ketone or lac-
tame) as already described in [20]. Using an excess of de-
methylating reagent allows complete demethylation, but the
yields remain low because of the instability of the compounds
in strong acid reaction mixtures. Most of the time, the reaction
leads to a mixture of methoxy and hydroxy derivatives. When
possible, these products have been separated by chromatogra-
phy for characterization in order to illustrate more precisely
their structure/activity relationships. In some case (compound
15), using BeCl2 as described by Sharghi and Tamaddon [21]
gives a selective reaction with an acceptable yield. Indeno-in-
doles 10-ones 2 are selectively substituted on the indolic nitro-
gen to give compound 16, or selectively iodinated with a very
3. Biology
The activity of the compounds was determined by binding
assays on MT3-rich membrane from hamster brain, using iodo-
melatonin as the radioligand, as classical for the study of this
Fig. 2. Schematic pathways for the transformations of compounds 1 and 2.

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
309
Fig. 3. Particular case: preparation of compound 3,8-dihydroxy-isoindolo(1,2-a)indol-10-one (14d).
binding site [9]. We demonstrated [13], after Molinari et al.
[11], that the experimental binding data obtained with iodo-
melatonin are completely correlated with those obtained with
iodo-MCA-NAT. All these results presented here were con-
firmed with [125I]MCA-NAT [11,13] (data not shown). The
compounds have no affinity towards the melatonin receptors.
MT3 is not a membrane-bound receptor, the binding test cannot
be complemented with a functional one. Indeed, despite sug-
gestion that MT3 might be linked to phosphoinositide produc-
tion in RMPI cell lines [23], in other cell lines where the bind-
ing site can be measured, no IP production could be correlate
to MT3 specific ligands. A more integrated functional assay has
been described for a MT3 ligand in vivo, as it reduces the in-
traocular pressure [24,25]. We were not able to test our com-
pounds in similar assays.
been crystallized [27], and that co-crystallizations were also
reported with resveratrol [28] and CB 1954 [29].
Overall variations around compound 2 gave more potent
compounds than variations around compound 1. Some of these
compounds have remarkable affinities towards MT3, far better
than melatonin itself or MCANAT. For example, the com-
pounds 1ax, 16aj or 16cj have subnanomolar affinities for
MT3 (with Ki of 430, 190 and 790 pM, respectively, see
Tables 1B,2B and 3B). The minute modifications of the core,
tetracyclic structures, led to some interesting observations con-
cerning the SAR of this series of compounds.
In brief, methoxylation of position 2 in compound 1 led to
inactive compounds unless, methoxylation occurred also at po-
sition 3 (Table 1A). Furthermore, the presence of hydroxyl, as
opposed to methoxyl moieties at most of the positions of this
aromatic ring, led to less potent ligands (Table 1B). Since the
presence oh two hydroxy onto the molecules might suggest a
possibility of a quinone-like structure, we determined if the
compounds were able to act as substrates of the QR2 enzyme,
as we previously described for other quinones [30]. Having
rules out this possibility, we turned to the moderate hydropho-
bic nature of methoxy residues, as compared to hydroxy resi-
dues. We concluded that the presence at some position (parti-
cularly, the position 2) of slightly more hydrophobic residue
(such as methoxy) were not favorable for the fixation of the
4. Results and discussion
A survey of the literature rapidly leads to an apparent pau-
city of tools to study the MT3 melatonin binding site (see Bou-
tin et al. [3] for review). Indeed, apart from the initially de-
scribed, poorly active and non-specific prasozin [9,10] or
attempts to synthesize loose derivatives of the melatonin core
[11,12], the compounds able to interact with MT3 are rather
scarce. We recently published [13] a series of compounds that
helped to understand the complex relationship between MT3,
the melatonin binding site and QR2, the enzyme that is actually
this binding site [26]. Exploring further the SAR of molecules
binding the MT3 site is crucial for its understanding since it
will provide new potent molecules that are specific towards the
MT3 binding site (i.e. with no affinity for MT1 and MT2). To
keep the system as simple as possible, we started by trying to
find compounds of new structures with affinity for MT3. Such
compounds were found in a classical screening program. A
posteriori, a loose structural similarity can be depicted between
these compounds and melatonin, as they might be thought of
as two fused indole moieties. Furthermore, it is easy to dock
some of these compounds inside the catalytic site of QR2 (A.
Gohier, G. Ferry, J.A. Boutin, in preparation), since QR2 has
Table 1A
Structure–activity relationship for compound 1 derivatives: influence of
methoxy substitution and of demethylation
An
Ki (nM)
0.43
> 100
2.8
> 100
32
1ax
1bx
1cx
1dx
1fx
2,3-diOMe
2,4-diOMe
1,4-diOMe
2-Ome
4-Ome

310
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
Table 1B
Structure–activity relationship for compound 1 derivatives: influence of methoxy substitution and of demethylation
An
Ki (nM)
0.43
2.8
An
Ki (nM)
15
6.8
1.9
> 100
An
Ki (nM)
36
1ax
1cx
1dx
2,3-diOMe
1,4-diOMe
2-OMe
14k
14j
14g
14k
3-OH 2-OMe
1-OH 4-OMe
2-OH
14i
1,4-diOH
> 100
1-OH
Table 2A
Table 3A
Structure–activity relationship for compound 2: influence of demethylation and
of N-methylation
Structure–activity relationship for compound 2: influence of nitrogen
substituent nature, of ketone reduction and of iodination
An
Ki
(nM)
3
An
Ki (nM)
2
R
H
Ki (nM)
0.19
32
16aj
16ak
16al
2c
1,4-diOMe
15
1-OH 4-OMe
CH2-NMe2
26
Table 2B
Structure–activity relationship for compound 2: influence of demethylation and
of N-methylation
16an
CH2-COOH
100
Table 3B
Structure–activity relationship for compound 2: influence of nitrogen
substituent nature, of ketone reduction and of iodination
An
2,3-diOMe
1,4-diOMe
R
H
H
Ki (nM)
1.3
3
R
Me
Me
Ki (nM)
0.19
0.79
2a
2c
16aj
16cj
de-methylated compound—compare the IC50s of compound
1dx, > 100 nM, and its de-methylated counterpart, the com-
pound 14g, 1.9 nM (see Table 1B). Nevertheless, the hydroxy
residue at position 1, was deleterious to the fixation of the
compound (compound 14k, IC50 > 100 nM, see Table 2A).
We intend to better understand these complicated situations
by using either docking techniques on the available crystals,
or to co-crystallisation approaches.
An
2,4-diOMe
1,4-diOMe
X
CO
CO
Ki (nM)
0.58
3
X
CH2
CH2
Ki (nM)
45
38
2b
2c
17b
17c
atom (i.e. susceptible to be radiolabeled) but this transforma-
tion led to an almost inactive compound (Table 3C).
Finally, as expected from our knowledge of MT and MT2
molecular pharmacology, none of these compounds showed
significant affinity towards these receptors (see Table 4).
For compound 2, the most striking variation was on the ni-
trogen, the methylation of which permitted to obtain 16aj, the
most active compound identified so far (Table 2B). Indeed, any
other variations on that nitrogen diminished the potency of the
new derivatives (see Table 3A). On the same core molecule,
attempts were made to reduce the carbonyl. The replacement
of this carbonyl by methylene considerably limited the potency
of the molecule (Table 2B).
5. Conclusion
The present report describes structures and synthesis details
on new ligands of the MT3 melatonin binding site. Provided
more data will be gathered on their broad specificity (i.e. out-
side the melatoninergic pharmacology), these substances will
permit to launch a series of studies of the physio-pathological
Furthermore, in a failing attempt to obtain an iodinated MT3
radioligand, we tried to synthesize a derivative with an iodine

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
311
Table 3C
6.2.1. Procedure for preparation of 2,3-dimethoxy 7-hydroxy
[10H] isoindolo (1,2-a) indol-10-one (1a)
Structure–activity relationship for compound 2: influence of nitrogen
substituent nature, of ketone reduction and of iodination
6.2.1.1. Triphenyl (5,6-dimethoxy phtalid-3-yl) phosphonium
bromide 5a. An amount of 62 g of 5,6-dimethoxy phtalide 3a
was dissolved in 800 ml of carbon tetrachloride. The mixture
was refluxed by heating with an halogen lamp (500 W). Sixty-
eight grams (1.2 equiv.) of N-bromosuccinimide were then
added by portions. After this addition was completed, the re-
flux was maintained for 2 additional hours. The mixture was
then cooled down to 10 °C. The solid precipitate of succini-
mide was collected by filtration. The solvent of the filtrate
was then distilled under reduced pressure. Eight hundred milli-
liters of anhydrous toluene were added to the residue and this
mixture, filtered again to eliminate remaining succinimide. The
crude dimethoxy bromophtalide 4a could be isolated after the
distillation of toluene, as an oily brown residue. 83.7 g (1 eq)
of triphenylphosphine was added to the toluene solution of 4a
and this mixture, heated to reflux for 16 h. After cooling, the
crystalline precipitate of Wittig reagent 5a was collected by
filtration, washed three times with 100 ml of anhydrous toluene
and then with boiling isopropanol. Finally it was washed with
cold diethyl ether and dried under vacuum. One hundred
twenty-eight grams (75%) of triphenyl (5,6-dimethoxy phta-
lid-3-yl) phosphonium bromide 5a were obtained in this way.
Table 5 summarizes the analytical data concerning all the com-
pounds prepared in this way. Wittig reagent 5a must react with
the suitable nitrobenzaldehyde obtained by substitution of the
phenol [18].
R
H
Ki (nM)
0.19
R
I
Ki (nM)
> 100
16aj
18
Table 4
Lack of affinity of MT3 compounds for the MT1 and MT2 melatonin receptors
Compound
1fx
14k
1dx
14g
MT1 IC50 (μM)
> 10
> 10
> 10
> 10
> 10
> 10
> 10
> 10
MT2 IC50 (μM)
> 10
> 10
> 10
> 10
> 10
> 10
> 10
0.56
1bx
16cj
16aj
16al
16ak
16an
> 10
> 10
> 10
> 10
Compounds, dissolved in DMSO were tested at full range of concentrations
from 0.1 nM up to 10 μM against [125I] 2-idomelatonin binding, as described
previously in [19].
role of MT3, as for instance on intraocular pressure or the nu-
merous properties of MT3/QR2 [31].
6. Experimental protocols
6.2.1.2. 5-Benzyloxy 2-nitro benzaldehyde (6y). A solution of
20 g of 5-hydroxy 2-nitrobenzaldehyde 6x in 60 ml of DMF
was added in 5 min at room temperature to a suspension of
33 g (2 equiv.) of potassium carbonate in 20 ml of DMF and
160 ml of diisopropyl ether. This mixture was then refluxed for
2 hours, and kept at room temperature. 13.8 ml of benzyl
chloride was added and the reaction mixture was refluxed for
15 hours. After cooling, potassium salts were filtered out and
washed with DMF. Filtrate was kept to dryness. The residue
was triturated with diethyl ether and filtrated. The benzyloxy
derivative crystallized slowly from filtrate and was collected by
filtration. 24.5 g (79%) of 5-benzyloxy 2-nitro benzaldehyde
6y were obtained in this way. 5-(2-Dimethylaminoethoxy) 2-
nitro benzaldehyde (6z) was prepared following the same pro-
cedure, replacing benzyl chloride by 2-chloroethyl dimethyl
amine, and diethyl ether by methylene chloride. The pure com-
pound was obtained by washing the methylene chloride solu-
tion with water, drying it with magnesium sulfate and eliminat-
ing the solvent under reduced pressure. Table 6 summarizes the
analytical data concerning all the compounds prepared in the
same way.
6.1. General
All compounds were submitted to standard analytical meth-
ods and gave satisfactory results. H NMR spectra were per-
1
formed on an AC200 (if not specified otherwise) spectrometer
(Brucker, Wissembourg, France). The following abbreviations
are used in description: s: singlet, d: doublet, t: triplet, dd: dou-
ble doublet, sl: singlet large, m: multiplet, ex: signal exchange-
able with D2O. δ are expressed in ppm, using TMS as a refer-
ence. Infrared spectra were performed on an IF25 of IF28
spectrometer (Brucker), using samples dispersed in Nujol. Fre-
quencies are expressed in cm−1. Melting points (instant) were
determined on a Leica-Reichert Kofler apparatus (Leica-Reich-
ert, Rueil-Malmaison, France). Preparative chromatography
was performed using 230–400 mesh Merck silica gel 60
(Merck-Darmstadt, Germany). Kugelrohr apparatus was a
GKR-51 (Büchi, Switzerland).
6.2. Typical procedures for the preparation of 2,3-dimethoxy
7-hydroxy [10H] isoindolo (1,2-a) indol-10-one (1a) and 2,3-
dimethoxy 8-hydroxy 5H indeno (1,2-b) indol-10-one (2a)
6.2.1.3. 5,6-Dimethoxy 3-(2-nitro 5-hydroxy benzylidene) phta-
lide (mixture of E/Z isomers) (7ax). One hundred grams of
Wittig reagent 5a was added by small portions to a stirred so-
lution of 31.2 g (1 equiv.) of 5-hydroxy 2-nitro benzaldehyde
For each step, a table (see Tables 5–16) gives the results for
the different substituted derivatives obtained.

312
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
Table 5
Analytical data
A
M.p.
(°C)
Yield
(%)
75
82
78
IR
NMR (δ ppm)
(Nujol suspension)
5a
5b
5ca
5d
5e
5,6-Dimethoxy
~260
> 260
> 260
> 260
> 260
1780
1772
1780
1785
1788
CDCl3 + DMSO: 9.3(s 1H), 7.8(m 15H), 6.5(s 1H), 7.2(s 1H), 3.9(s 3H), 3.6(s 3H)
DMSO: 8.5(s 1H), 7.9(m 15H), 6.9(dd 2H), 3.8(s 3H), 3.3(s 3H)
DMSO: 8.5(s 1H), 7.7(m 15H), 7.4(d 1H), 7.2(d 1H), 3.8(s 3H), 3.35(s 3H)
DMSO: 8.5(s 1H), 7.9(m 15H), 7.4(m 2H), 6.8(d 1H), 3.8(s 3H)
CDCl3 + DMSO: 9.75(s 1H), 7.85–7.65(m 15H), 7.55(t 1H), 7.0(d 1H), 6.65(dd 1H),
3.9(s 3H)
4,6-Dimethoxy
4,7-Dimethoxy
6-Methoxy
56
36
7-Methoxy
5f
5gb
a
4-Methoxy
6-Hydroxy
247
> 260
38
55
1777
1770
CDCl3 : 9.75(s 1H), 8.1–7(m 18H), 3.4(s 3H)
DMSO: 8.5(s 1H), 8.1–7.7(m 15H), 7.2(dd 1H), 7.1(d 1H), 6.7(d 1H)
Bromination of phtalide 4c has been realized in methylene chloride.
Compound 5g has been isolated from 6-acetoxy phtalide, the hydroxy group is liberated from ester function during treatment with isopropanol.
b
Table 6
Analytical data
R1
M.p.
(°C)
Yield
(%)
IR
NMR (δ ppm)
(Nujol suspension)
6x
6y
6z
H
Benzyl
–(CH2)2–N(Me)2
Commercially available
73
79
50
1705
2824, 2774, 1700,
1517, 1331 (film)
DMSO: 10.5(s 1H), 8.2(d 1H), 7.4(m 6H), 7.2(dd 1H), 5.2(s 2H)
CDCl3: 10.45(s 1H), 8.15(d 1H), 7.35(d 1H), 7.18(dd 1H), 4.2(t 2H),
2.8(t 2H), 2.35(s 6H)
Oil
6x in 1000 ml of DMF and 26 ml (1 equiv.) of triethylamine at
room temperature. The reaction mixture was heated to 90 °C
for 90 min, and then cooled to room temperature and kept to
dryness under reduced pressure. The residue was triturated
with diethyl ether. The solid was filtrated, treated with metha-
nol under reflux, cooled to room temperature and filtered again
to give 44.8 g (70%) of compound 7ax as a mixture (75:25) of
E and Z isomers of 5,6-dimethoxy 3-(2-nitro 5-hydroxy benzy-
lidene) phtalide. Table 7 summarizes the analytical data con-
cerning all the compounds prepared in the same way.
95 ml of ethanol. 81 ml (2 eq) of 1 N NaOH solution were
added, the reaction mixture turned to dark red color. It was
refluxed for 45 min. After cooling on an ice bath, 100 ml of
1 N HCl solution were added while stirring. pH decreased to
~1. The mixture was kept at room temperature for 1 hour, the
precipitate was collected by filtration, washed with diethyl
ether and dried. A 7 g fraction of compound 9ax was obtained.
The aqueous phase of the filtrate was concentrated to 1/3. A
dark gummy mass formed and was dissolved in a neutral aqu-
eous methanol mixture. This solution was extracted with
diethyl ether. The last acidic aqueous phase was also extracted
with diethyl ether. All the ethereal phases were added, treated
with active charcoal, and concentrated to obtain a dark gray
residue. This residue was washed with cold diethyl ether and
dried. An additional 2.6 g of compound was obtained. 9.6 g
(76%) of 5-hydroxy 2-(2-carboxy 4,5-dimethoxy phenyl) in-
dole 9ax were obtained in this way. Table 9 summarizes the
analytical data concerning all the compounds obtained in the
same way.
6.2.1.4. 5,6-Dimethoxy 3-(5-hydroxy 2-amino benzylidene)
phtalide (8ax). Fifteen grams of compound 7ax were dissolved
in 460 ml of DMF and reduced in a Parr apparatus with 36 ml
of Raney nickel suspension under mild hydrogen pressure (3–5
atm) until the reduction was complete. The mixture was then
filtered under nitrogen atmosphere. The nickel residue was
rinsed with DMF, and the filtrate kept to dryness, washed with
diethyl ether and dried under vacuum. 12.7 g (93%) of almost
pure 5,6-dimethoxy 3-(5-hydroxy 2-amino benzylidene) phta-
lide 8ax were obtained. Table 8 summarizes the analytical data
concerning all the compounds prepared in the same way.
6.2.1.6. 2,3-Dimethoxy 7-hydroxy 10H isoindolo (1,2-a) indol-
10-one (1ax). Method 1: Cyclo-dehydration using catalytic
APTS. Two grams of compound 9ax and a catalytic amount
of p-toluene sulfonic acid were added to 80 ml of toluene.
The mixture was refluxed with stirring for 12 hours in a reactor
6.2.1.5. 5-Hydroxy 2-(2-carboxy 4,5-dimethoxy phenyl) indole
(9ax). An amount of 12.7 g of compound 8ax was added to

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
313
Table 7
Analytical data
R1
R
M.p.a
(°C)
253
Yield
(%)
70
IR
NMR (δ ppm)
(Nujol suspension)
3248, 1760, 1660
7ax
7ay
7by
7cx
7cy
7dx
7ez
7fx
H
5,6-Dimethoxy
5,6-Dimethoxy
4,6-Dimethoxy
4,7-Dimethoxy
4,7-Dimethoxy
6-Methoxy
7-Methoxy
4-Methoxy
6-Hydroxy
DMSO: 11.2(m(ex) 1H), 8.2(d 1H), 7.4(s 1H), 7.15(d 1H), 7.1(s
1H), 7.0(dd 1H), 6.7(s 1H), 3.9(s 3H), 3.4(s 3H)
Benzyl
190
192
> 260
168
225
134
> 260
-
75
69
85
77
51
73
57
6
1761, 1660, 1570, 1358 CDCl3: 8.3(d 1H), 7.3(m 7H), 7.2(dd 1H), 7.1(s 1H), 6.4(s 1H), 5.1
(s 2H), 4.0(s 3H), 3.6(s 3H)
1774, 1518, 1507, 1334 CDCl3: 8.05(d 1H), 7.75(s 1H), 7.55–7.3(m 6H), 7.0(S 1H), 6.95
(dd 1H), 6.75(s 1H), 5.2(s 2H), 4.0(s 3H), 3.9(s 3H)
3357, 1746, 1589, 1333 DMF (300 MHz): 11.4(sl(ex) 1H), 8.15(d 1H), 7.65(d 1H), 7.6(d
1H), 7.5(s 1H), 7.35(d 1H), 7.0(dd 1H), 4.15(s 3H), 4.0(s 3H)
1781, 1636
Benzyl
H
Benzyl
CDCl3 : 8.05(d 1H), 7.75(d 1H), 7.6(s 1H), 7.55–7.3(m 5H), 7.15
(d 1H), 6.95(m 2H), 5.75(s 2H), 4(2s 6H)
H
3323, 1740, 1514, 1328 DMSO: 12.1–10.5(sl(ex) 1H), 8.25(d 1H), 7.45(d 1H), 7.3(dd 1H),
7.1(d 1H), 7.05(m 3H), 3.9(s 3H)
(CH2)2NMe2
1779, 1503
CDCl3: 8.25(d 1H), 7.4(t 1H), 7.1(m 3H), 6.9(d 1H), 6.6(d 1H),
4.15(t 2H), 3.95(s 3H), 2.75(t 2H), 2.35(s 6H)
DMSO: 11(m(ex) 1H), 8.1(d 1H), 7.7(m 1H), 7.6(m 2H), 7.5(d
1H), 7.4(s 1H), 6.9(dd 1H), 4.1(s 3H)
H
3251, 1740, 1660
7gxb
H
-
DMSO: 8.25(dd 1H), 7.2–6.95 (m 6H)
a
Melting points are given only as an indication, as the compounds 7 are most of the time obtained as mixtures of E and Z isomers.
7gx has been isolated with a very low yield (see Scheme 3), and is crystallized with 30% DMF.
b
Table 8
Analytical data
R1
A
M.p.a
(°C)
231
Yield
(%)
93
IR
NMR (δ ppm)
(Nujol suspension)
1741
8ax
8by
8cx
8dx
8ez
8fx
a
H
5,6-Dimethoxy
4,6-Dimethoxy
4,7-Dimethoxy
6-Methoxy
7-Methoxy
4-Methoxy
DMSO: 8.65(s(ex) 1H), 7.35(2s 1H), 7.15(s 1H), 6.8(2s 1H), 6.7–
6.5(m 3H), 5.05(s(ex) 1H), 4.05(s(ex) 1H), 4.0(s 3H), 3.7(s 3H)
CDCl3 : 7.6–7.25(m 6H), 6.95(d 1H), 6.85–6.6(m 4H), 5.05(s 2H),
4(s 3H), 3.9(s 3H), 3.6(sl(ex) 2H)
DMSO: 8.6(s(ex) 1H), 7.4(d 1H), 7.3(d 1H), 7.1(d 1H), 6.9(s 1H),
6.6(d 1H), 6.5(dd 1H), 4.7(m(ex) 2H), 3.9(2s 6H)
DMSO: 8.6(s(ex) 1H), 7.5(d 1H), 7.4(d 1H), 7.35(dd 1H), 6.65(m
4H), 4.6(s(ex) 2H), 3.9(s 3H)
CDCl3: 7.4(t 1H), 7–6.7(m 5H), 6.6(sl 1H), 4(s 3H), 4(t 2H), 3.5(m
(ex) 2H), 2.7(t 2H), 2.3(s 6H)
DMSO: 8.6(m(ex) 1H), 7.5(m 3H), 7.3(d 1H), 6.9(s 1H), 6.6(m
2H), 4.7(m(ex) 2H), 4.1(s 3H)
Benzyl
141
84
80
94
75
91
1752
H
> 260
203
1723, 1619
1777
H
(CH2)2NMe2
H
148
1773, 1638
1756, 1666
234
Melting points are given only as an indication, as the compounds 7 are most of the time obtained as mixtures of E and Z isomers.
equipped with a Dean-Stark apparatus. The reaction mixture
was then cooled to room temperature and filtered. The precipi-
tate was washed two times with 5 ml of cold toluene, three
times with 5 ml of cold methanol and then with 5 ml of diethyl
ether. This material was then dissolved in the minimum of
THF, and this solution filtered through a Millipore membrane.
The THF solution was kept to dryness under reduced pressure,
the residue refluxed for a few minutes with 10 ml of methanol,
the mixture cooled on an ice bath, filtered again and the residue
finally washed with diethyl ether and dried under vacuum.

314
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
Table 9
Analytical data
R1
A
M.p.
(°C)
Yield
(%)
76
IR
NMR (δ ppm)
(Nujol suspension)
1666, 1641
9ax
H
4′,5′-Dimethoxy 160
DMSO (300 MHz): 12.6(sl(ex) 1H), 10.9(m(ex) 1H), 8.6(m(ex)
1H), 7.3(s 1H), 7.2(d 1H), 7.1(s 1H), 6.8(sl 1H), 6.6(d 1H), 6.3(sl
1H), 3.8(2s 6H)
9by
Benzyl
H
4′,6′-Dimethoxy 210
3′,6′-Dimethoxy 140
98
91
1688
1700
DMSO: 12.6(s(ex) 1H), 10.9(m(ex) 1H), 7.4(m 5H), 7.3(d 1H), 7.1
(d 1H), 6.8(m 3H), 6.2(sl 1H), 5.1(s 2H), 3.9(2s 6H)
DMSO (300 MHz): 12.8(m(ex) 1H), 10.7(sl(ex) 1H), 8.6(s(ex)
1H), 7.2(d 1H), 7.1(2d 2H), 6.8(d 1H), 6.6(dd 1H), 6.3(d 1H), 3.7
(2s 6H)
9cx
9dx
9ez
9fx
H
4′-Methoxy
3′-Methoxy
6′-Methoxy
210
91
81
75
1711
DMSO: 11(m(ex) 2H), 8.65(sl(ex) 1H), 7.6(d 1H), 7.1(d 1H), 6.8(d
1H), 6.6(dd 1H), 6.3(s 1H), 3.8(s 3H)
(CH2)2NMe2
H
> 260
234
1580
DMSOa: 15–14(m(ex) 2H), 7.5(t 1H), 7.3(m 2H), 7.1(m 2H), 6.8
(dd 1H), 6.6(sl 1H), 4.3(m 2H), 3.9(s 3H), 3.5(m 2H), 2.9(s 6H)
DMSO (300 MHz): 12.6(m(ex) 1H), 10.7(sl(ex) 1H), (8.6(s(ex)
1H), 7.4(t 1H), 7.2(m 3H), 6.8(d 1H), 6.6(dd 1H), 6.1(sl 1H), 3.8(s
3H)
1722, 1627
a
With addition of 2 drops of CF3COOD.
1.2 g (64%) of 2,3-dimethoxy 7-hydroxy 10H isoindolo (1,2-a)
indol-10-one 1ax was obtained this way. In one case, purifica-
tion has been completed by chromatography on silica, devel-
oped with dichloromethane.
again. This new residue was then triturated with 25 ml of a 1:1
mixture of methanol and dichloromethane for 10 hours and
filtered. This final residue was dried under vacuum. 2.75 g
(88%) of 2,4-dimethoxy 7-hydroxy 10H isoindolo (2,1-a) in-
dol-10-one 1bx were obtained. Table 10 summarizes the ana-
lytical data concerning the compounds obtained in the same
way.
Method 2: Cyclo-dehydration using dicyclohexyl-carbo-dii-
mide. Eight grams of compound 9ax were dissolved in 175 ml
of DMF and placed in a reactor. 6.37 g of dicyclohexyl-carbo-
diimide were added under stirring and the mixture, refluxed for
90 min. The mixture, cooled at room temperature, was filtered.
The filtrate was kept to dryness under reduced pressure and the
residue triturated, first with diethyl ether, then with dichloro-
methane. The solid obtained was added to the first precipitate
and this mixture was purified by distillation under vacuum in a
Kugelrohr apparatus to separate it from dicyclohexyl-urea. The
compound obtained was finally crystallized with isopropanol
and washed with petroleum ether to give 5.46 g (73%) of
2,3-dimethoxy 7-hydroxy 10H isoindolo (1,2-a) indol-10-one
1ax. In some cases (compounds 10by and 10ez), the distilla-
tion has been replaced by chromatography on silica, developed
with a mixture of dichloromethane and methanol (95:5). Ta-
ble 10 summarizes the analytical data concerning all the com-
pounds obtained in the same way.
6.2.1.8. 1,7-Dihydroxy 4-methoxy isoindolo (1,2-a) indol-10-
one (14c) and 1,4,7-Trihydroxy isoindolo (1,2-a) indol-10-
one (14c′). Demethylation of methoxy derivatives can be
achieved using different demethylating agents but the reaction
remains incomplete, or produces a large proportion of degrada-
tion compounds if an excess of reagent is used. The following
examples are detailed for illustrating one process.
0.44g of 1,4-dimethoxy derivative 1cx was added to a mix-
ture of 20 ml of acetic acid and 0.65 ml of 48% aqueous solu-
tion of bromhydric acid. The mixture was refluxed for 16 h,
and then dried under reduced pressure. The residue was tritu-
rated with a large excess of diethylether and filtrated. The pre-
cipitate was purified by chromatography on silica, eluating
with a mixture of dichloromethane and methanol (98:2 to
95:5). Forty milligrams of 1,4,7-trihydroxy isoindolo (1,2-a)
indol-10-one 14c′ were isolated. The filtrate was chromatogra-
phied separately and developed with the same solvent to finally
produce 40 mg of 1,7-dihydroxy 4-methoxy isoindolo (1,2-a)
indol-10-one 14c and 60 mg of remaining starting material 1cx.
Table 11 summarizes the analytical data concerning the com-
pounds obtained in the same way.
6.2.1.7. 2,4-Dimethoxy 7-hydroxy 10H isoindolo (2,1-a) indol-
10-one (1bx). 4.1 g of compound 10by (see Table 10) were
dissolved in a mixture of 170 ml of DMF and 55 ml of metha-
nol, and the solution was placed in a Parr apparatus. 410 mg of
palladium catalyst (10% on C) were introduced under nitrogen
flow and the mixture was hydrogenated under a 4–6 atm pres-
sure at 50 °C until the theoretical absorption was complete.
After cooling, the catalyst was eliminated by filtration and the
filtrate was dried under reduced pressure. The residue was tri-
turated with an excess of diethyl ether for 24 hours and filtered
6.2.1.9. 6-Hydroxy 3-(2-nitro 5-hydroxy benzylidene) phtalide
(7gx) and 5-Hydroxy 2-(2-nitro 5-hydroxy phenylacetyl) ben-
zoic acid methyl ester (13). Five grams of phosphonium bro-

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
315
Table 10
Analytical data
R1
A
M.p.
(°C)
> 250
Yield
(%, method)
64 (1)
73 (2)
74 (1)
IR
NMR (δ ppm)
(Nujol suspension)
1726
1ax
10by
1bx
1cx
H
2,3-Dimethoxy
2,4-Dimethoxy
2,4-Dimethoxy
1,4-Dimethoxy
2-Methoxy
DMSO: 9.3(m(ex) 1H), 7.5(d 1H), 7.4(s 1H), 7.3(s 1H), 6.9(sl 1H),
6.7(d 1H), 6.6(s 1H), 4(2s 6H)
DMSO: 7.7(d 1H), 7.4(m 5H), 6.9(m 3H), 6.6(d 1H), 6.4(s 1H),
5.1(s 2H), 3.9(2s 6H)
DMSO: 9.3(s(ex) 1H), 7.5(d 1H), 6.9(m 3H), 6.7(dd 1H), 6.4(s
1H), 3.9(2s 6H)
DMSO (300 MHz): 9.3(s(ex) 1H), 7.5(d 1H), 7.3(d 1H), 7.1(d 1H),
6.9(d 1H), 6.7(dd 1H), 6.6(s 1H), 3.9(2s 6H)
DMSO (300 MHz): 9.35(s(ex) 1H), 7.65(d 1H), 7.5(d 1H), 7.25(d
1H), 7.2(dd 1H), 6.9(d 1H), 6.25(dd 1H), 6.2(s 1H), 3.85(s 3H)
DMSO (300 MHz): 7.8(d 1H), 7.4(m 1H), 7.1(d 1H), 7(d 1H), 6.9
(dd 1H), 6.8(d 1H), 6.5(s 1H), 4.1(t 2H), 4(s 3H), 2.7(t 2H), 2.3(s
6H)
Benzyl
194
1720
H
H
H
> 250
> 250
217
88a
1694
84 (2)
57 (2)
70 (2)
1716, 1733
1696
1dx
10ez
(CH2)2NMe2 1-Methoxy
134
1719
1fx
H
4-Methoxy
~250
72 (2)
1697, 1718
DMSO: 9.3(m(ex) 1H), 7.6(d 1H), 7.4(d 1H), 7.3(m 2H), 6.9(d
1H), 6.7(dd 1H), 6.6(s 1H), 4(s 3H)
a
Yield of debenzylation reaction (from 10by).
Table 11
Analytical data
A
M.p.
(°C)
Yield
(%, reagent)
6 (AlCl3/toluene)
IR
(Nujol suspension)
NMR (δ ppm)
14a
14c
14c'
14d
14e
a
3-Hydroxy 2-methoxy > 260
1693
1695
1666
1692
DMSO (400 MHz): 10.25(s(ex) 1H), 9.3(s(ex) 1H), 7.45(d 1H), 7.22(s
1H), 7.08(s 1H), 6.83(d 1H) 6.7(m 1H), 6.63(s 1H), 3.75(s 1H)
DMSO (300 MHz): 10.1(s(ex) 1H), 9.3(s(ex) 1H), 7.5(d 1H), 7.2(d 1H),
6.9(d 1H), 6.8(d 1H), 6.7(m 1H), 6.5(s 1H), 3.9(s 1H)
DMSO: 9.3(sl(ex) 1H), 8.5(sl(ex) 1H), 7.45(d 1H), 6.85(m 2H), 6.7(dd
1H) 6.45(d 1H), 6.45(s 1H)
DMSO: 10.2(m(ex) 1H), 9.3(m(ex) 1H), 7.5(2d 2H), 7(d + dd 2H), 6.9
(d 1H), 6.7(dd 1H)
DMSO (300 MHz): 10.5(sl(ex) 1H), 9.2(sl(ex) 1H), 7.5(d 1H), 7.3(t
1H), 7.1(d 1H), 6.8(sl 1H), 6.8(d 1H), 6.7(m 2H)
1-Hydroxy 4-methoxy 220
10 (HBr 48% aq
CH3COOH)
10 (HBr 48% aq
CH3COOH)
47a
1,4-Dihydroxy
2-Hydroxy
> 260
> 260
250
1-Hydroxy
34b (BBr3 CH2Cl2) 1703, 1689
Prepared from compound 13 as described above.
b
Prepared from 10ez, aminoalkyl chain is eliminated during demethylation process.
mide 5g were placed in a reactor with 57 ml of DMF and
1.4 ml of triethylamine. 1.7 g of 2-nitro 5-hydroxy benzalde-
hyde 6x were added by small fractions in 5 min under stirring.
The brown reaction mixture obtained was warmed to 90 °C for
40 min, then cooled to room temperature and dried under re-
duced pressure. The residue was triturated with an excess of
diethyl ether. Triethyl ammonium bromine was eliminated by
filtration. The ethereal filtrate was dried and the oily residue
purified by silica chromatography developed with a mixture
of dichloromethane and diethyl ether (95:5 to 90:10) to give
0.57 g (18%) of 6-hydroxy 3-(2-nitro 5-hydroxy benzylidene)
phtalide 7gx (described in Table 7) and 1.68 g (51%) of 5-hy-
droxy 2-(2-nitro 5-hydroxy phenylacetyl) benzoic acid methyl
ester 13 (IR (Nujol suspension): 1690, 1666), NMR (DMSO):
10.9(m(ex) 2H), 8.1(d 1H), 7.9(d 1H), 7(dd 1H), 6.9(m 3H),
4.7(s 2H), 3.7(s 3H).
6.2.1.10. 3,8-Dihydroxy isoindolo (1,2-a) indol-10-one (14d).
5.9 g of compound 13 were reduced at room temperature with
14.8 ml of Raney nickel suspension in 300 ml of DMF under a
6 atm hydrogen pressure until the absorption was complete.
The catalyst was then separated by filtration. The DMF filtrate

316
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
was kept to dryness and this residue triturated in an excess of
dichloromethane. The concentration of the dichloromethane so-
lution under reduced pressure to about 100 ml, liberated a first
precipitate of 2.5 g of the intermediary amine. This filtrate and
the precedent residue were added and treated by chromatogra-
phy on silica, developed with a mixture of dichloromethane
and methanol (98:2 to 95:5). A second sample of 1.6 g of
amine was isolated from the first eluted fractions. 4.1 g of the
crude intermediate amine obtained were dissolved in a mixture
of 10 ml of methanol and 20 ml of 1 N NaOH solution. The
mixture was stirred at room temperature for 3 hours and then
concentrated to about 15 ml at room temperature under reduced
pressure. The precipitate obtained was collected, refluxed a few
minutes in 10 ml of methanol, cooled and filtered again. This
last precipitate was then dissolved in a mixture of acetone and
DMF, the solution obtained was filtrated on a Millipore filter.
The filtrate was dried, and the residue triturated with diethyl
ether and finally dried under vacuum. 2.1 g (47%) of 3,8-dihy-
droxy isoindolo (1,2-a) indol-10-one 14d were obtained. Ta-
ble 11 summarizes the analytical data obtained on all the poly-
phenolic isoindolo (1,2-a) indol-10-one compounds.
ing with the deprotected benzylidene phtalide 7x at room tem-
perature for 16 h. In that case, the crude precipitate obtained
after acidification was washed first with water and then with
petroleum ether to give an almost pure compound 5,6-di-
methoxy 2-(2-nitro 5-hydroxy phenyl) indan-1,3-dione
(11ax). Table 12 summarizes data concerning all the com-
pounds prepared in the same way.
6.2.2.2. 2,3-Dimethoxy 8-benzyloxy 5H indeno (1,2-b) indole
10-one (12a). 19.4 g of compound 11ay were dissolved in
250 ml of DMF and reduced in a Parr apparatus with 12 ml
of Raney nickel suspension under mild hydrogen pressure (3–5
atm) until the reduction was complete. This mixture was then
filtered and the nickel residue was washed with DMF. The fil-
trate was dried under reduced pressure. The residue was then
washed with 50 ml of ethanol, 100 ml of diethyl ether and
dried under vacuum. 12.6 g (73%) of 2,3-dimethoxy 8-benzy-
loxy-5H-indeno (1,2-b) indole 10-one 12a were obtained. Ta-
ble 13 summarizes the analytical data concerning all the com-
pounds prepared in the same way.
6.2.2.3. 2,3-Dimethoxy 8-hydroxy 5H indeno (1,2-b) indol-10-
one (2a). 11.3 g of compound 12a were introduced in a reactor
containing 340 ml of methanol, 150 ml of DMF and 59 ml of 1
N NaOH solution. After flushing it out with nitrogen, 3 g of Pd
(10% on C) were added and the mixture, hydrogenated under
low pressure (<80 mbar) until deprotection was complete. The
catalyst was separated by filtration. Sixty milliliters of a 1 N
HCl solution were added while stirring on an ice bath. After
drying under reduced pressure, the residue was treated with an
excess of isopropanol. The sodium chloride was eliminated by
filtration, and isopropanol was distilled under reduced pressure.
The final residue was washed with a few ml of iced water.
8.3 g (95%) of 2,3-dimethoxy 8-hydroxy 5H indeno (1,2-b)
indol-10-one 2a were obtained as a dark red powder. The same
compound could be obtained (80%) by reduction with Raney
nickel and spontaneous cyclization of compound 11ax, using
conditions described above for 11ay. Table 14 summarizes the
6.2.2. Procedure for preparation of 2,3-dimethoxy 8-hydroxy
5H indeno (1,2-b) indol-10-one (2a)
6.2.2.1. 5,6-Dimethoxy 2-(2-nitro 5-hydroxy phenyl) indan-1,3-
dione (11a). Twenty-six grams of compound 7y were added to
a mixture of 60 ml of 4 N NaOH solution and 92 ml of metha-
nol. The dark red mixture was warmed to 40 °C under vigorous
stirring for 1 h, cooled down to 0 °C on an ice bath and acid-
ified with 75 ml of a 4 N HCl solution. An abundant white
precipitate formed. A vigorous stirring was maintained at room
temperature for 12 h. The precipitate was collected by filtra-
tion, washed with cold water and dried under vacuum. The
purification by silica chromatography, developed with a mix-
ture of dichloromethane and methanol (99:1) gives 19.5 g
(75%) of 5,6-dimethoxy 2-(2-nitro 5-benzyloxy phenyl) in-
dan-1,3-dione 11ay. The same reaction can be performed start-
Table 12
Analytical data
R1
A
M.p.
(°C)
230
Yield
(%)
66
IR
NMR (δ ppm)
(Nujol suspension)
1741, 1622, 1586,
11axa
11ay
11by
11cy
a
H
5,6-Dimethoxy
5,6-Dimethoxy
4,6-Dimethoxy
4,7-Dimethoxy
DMSO (300 MHz): 11.2(sl(ex) 1H), 8.1(d 1H), 7.45(s 2H), 6.98(dd
1704, 1507, 1339, 1223 1H), 6.95(d 1H), 5.29(s(ex) 1H), 4(s 6H)
1741, 1698, 1501, 1342 CDCl3: 8.25(d 1H), 7.4(m+2s 7H), 7.05(dd 1H), 6.95(d 1H), 5.15(s
2H), 4.65(s 1H), 4.05(s 6H)
Benzyl
Benzyl
Benzyl
171
153
213
75
95
66
1748, 1710
CDCl3: 8.2(d 1H), 7.4(m 5H), 7.1–6.8(m 4H), 5.2(s 2H), 4.6(m
1H), 4(2s 6H)
CDCl3 (300 MHz): 8.3(d 1H), 7.4(m 5H), 7.3(s 2H), 7.1(dd 1H), 7
(d 1H), 5.2(s 2H), 4.6(m 1H), 4(s 6H)
1742, 1710
Prepared from 7x.

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
317
Table 13
Analytical data
A
M.p. (°C)
235
Yield (%)
73
IR (Nujol suspension)
1669
NMR (δ ppm)
12a
2,3-Dimethoxy
DMSO: 12.2(m(ex 1H), 7.4(m 6H), 7(m 2H), 6.7(dd 1H), 5.1(s 2H), 3.9
(s 6H)
12ba 2,4-Dimethoxy
12b'a 1,3-Dimethoxy
-
(10a)
(33a)
63
1658
1662
1683
CDCl3 (300 MHz): 8.5(m(ex) 1H), 7.5(d 2H), 7.4(m 3H), 7.3(d 1H), 7.2
(d 1H), 6.8(dd 1H), 6.8(d 1H), 6.3(d 1H), 5.1(s 2H), 3.9(2s 6H)
DMSO (300 MHz): 12.2(s(ex) 1H), 7.4(m 6H), 7.1(d 1H), 6.8(dd 1H),
6.55(d 1H), 6.49(d 1H), 5.1(s 2H), 3.9(2s 6H)
DMSO (300 MHz): 11.95(m(ex) 1H), 7.5–7.3(m 6H), 7.15(d 2H), 7.05
(s 1H), 6.95(d 1H), 6.8(dd 1H), 5.15(s 2H), 3.9(s 3H), 3.8(s 3H)
-
12c
1,4-Dimethoxy
> 260
a
As explained, when substitution of the indene part is not symmetrical (11b), reaction leads to a mixture, compounds have been incompletely separated
(~85% purity) with low yields by chromatography, but signals has been identified in NMR. In the case of 12b, purification has been completed after debenzyla-
tion at the next step.
Table 14
Analytical data
A
M.p. (°C)
> 260
Yield (%)
95 (80a)
IR (Nujol suspension)
1669
NMR (δ ppm)
2a
2b
2b'
2c
2,3-Dimethoxy
DMSO: 12.1(m(ex) 1H), 9(m(ex) 1H), 7.2(d 1H), 7(2s 2H), 6.8(s 1H),
6.6(dd 1H), 3.9(2s 6H)
2,4-Dimethoxy
1,3-Dimethoxy
1,4-Dimethoxy
> 260
> 260
> 260
26b
26c
90
1668
DMSO: 11.9(sl(ex) 1H), 9.05(m(ex) 1H), 7.15(d 1H), 6.8(d 1H), 6.6(s
2H), 6.55 (dd 1H), 3.9(s 3H), 3.85(s 3H)
DMSO: 12.1(sl(ex) 1H), 9.05(m(ex) 1H), 7.2(d 1H), 6.87(d 1H), 6.6(dd
1H), 6.55 (d 1H), 6.38(d 1H), 3.85(2s 6H)
DMSO (300 MHz): 11.8(s(ex) 1H), 9.1(sl(ex) 1H), 7.2(d 1H), 6.95(d
1H), 6.85(s 1H), 6.55(d 1H), 3.9(s 3H), 3.8(s 3H)
1654
1688, 1671
a
b
c
By reduction and cyclization of the non-benzylated phenyl indane dione 11ax.
By chromatographic separation from the remaining isomer 2b′ and crystallization in isopropanol.
By chromatographic separation from the remaining isomer 2b and crystallization in isopropanol.
analytical data concerning all the compounds prepared in the
same way. Indeno indoles 2 were selectively substituted on the
indolic nitrogen by using potassium carbonate in acetone or
DMF.
16am were obtained. The free carboxylic acid was obtained
by treating the ester 16am with 1 equiv. of a 1 N NaOH solu-
tion in the same volume of ethanol at room temperature for
30 min, and then by acidifying it with 1.05 equiv. of a 1 N
HCl solution. The precipitate obtained was filtrated, washed
with iced water and finally with petroleum ether. Seventy per-
cent of 2,3-dimethoxy 8-hydroxy 5-carboxymethyl indeno
(1,2-b) indol-10-one 16an are obtained (described in Table 15
as its sodium salt). The same general procedure could be used
with N-(2-chloro ethyl) morpholine hydrochloride or N-(2-
chloro ethyl) dimethylamine hydrochloride (using 2 equiv. of
potassium carbonate). So this is for methylation using a slight
excess of methyl iodide. Table 15 summarizes the analytical
data concerning all the N-substituted derivatives 16.
6.2.2.4. 2,3-Dimethoxy 8-hydroxy 5-carboxymethyl indeno
(1,2-b) indol-10-one (16an). Two grams of compound 2a were
dissolved in 15 ml of DMF and 5 ml of acetone. 0.94 g (1
equiv.) of potassium carbonate were added. The mixture was
stirred at 47 °C for 12 h. 1.13 g of methyl bromacetate in 5 ml
of acetone were added and the temperature maintained for 4 h.
After cooling at room temperature, the reaction medium was
filtered, the precipitate washed with a large amount of diethy-
lether. The solvent phase was washed with water, dried with
magnesium sulfate and kept to dryness. The residue was
washed first with isopropanol and then with a minimum
amount of diethyl ether. 1.95 g (75%) of 2,3-dimethoxy 8-hy-
droxy 5-methoxycarbonylmethyl indeno (1,2-b) indol-10-one
The 1,4-dimethoxy derivative has been selectively demethy-
lated on position -1, starting from the O-benzyl derivative 12c,
the demethylated product has been identified but not purified
and immediately submitted to classical debenzylation.

318
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
Table 15
Analytical data
A
R2
M.p.
(°C)
> 260
Yield
(%)
49
IR
NMR (δ ppm)
(Nujol suspension)
1659
16aj
16ak
2,3-Dimethoxy
H
DMSO (300 MHz): 9.2(s(ex) 1H), 7.28(d 1H), 7.15(s 1H), 6.98(s
1H), 6.82(d 1H), 6.6(dd 1H), 3.98(s 3H), 3.91(s 3H), 3.8(s 3H)
DMSO (300 MHz): 9.23(m(ex) 1H), 7.32(d 1H), 7.07(s 1H), 7.0(s
1H), 6.82(d 1H), 6.58(dd 1H), 4.45(t 2H), 3.9(s 3H), 3.82(s 3H),
2.65(t 2H), 2.22(s 6H)
DMSO: 9.3(s(ex) 1H), 7.35(d 1H), 7.1(s 1H), 7.0(s 1H), 6.8(d 1H),
6.6(dd 1H), 4.45 (t 2H), 3.85(s 3H), 3.8(s 3H), 3.45(m 4H), 2.7(t
2H), 2.4(m 4H)
2,3-Dimethoxy CH2–NMe2
216
40
1667
1666
16al
2,3-Dimethoxy
228–232 40
16am
16an
16cj
2,3-Dimethoxy COOMe
2,3-Dimethoxy COONa
251
75
1735, 1693
1681, 1625
1666
DMSO (300 MHz): 9.25(s(ex) 1H), 7.25(d 1H), 7.02(s 2H), 6.85(d
1H), 6.6(dd 1H), 5.4(s 2H), 3.9–3.8(3s 9H)
DMSO: 9.25(s(ex) 1H), 7.12(d 1H), 6.95(s 1H), 6.92(s 1H), 6.80(d
1H), 6.52(dd 1H), 4.53(s 2H), 3.8(2s 6H)
DMSO (300 MHz): 9.2(m(ex) 1H), 7.3(d 1H), 7.15(d 1H), 7.0(d
1H), 6.9(s 1H), 6.65(dd 1H), 4.05(s 3H), 3.9(s 3H), 3.8(s 3H)
DMSO (500 MHz): 9.1(m(ex) 1H), 7.35(d 1H), 6.85(s 1H), 6.65(s
+ dd 2H), 6.4(s 1H), 4.4(t 2H), 3.85(2s 6H), 2.65(t 2H), 2.2(s 6H)
> 260
> 260
264
52 (2
steps)
56
1,4-Dimethoxy
H
16b'k
1,3-Dimethoxy CH2–NMe2
41
1676
6.2.2.5. 1-Hydroxy 4-methoxy 8-benzyloxy 5H indeno (1,2-b)
indol-10-one (19). 1.5 g of compound 12c was refluxed for
4 days with 0.9 g of beryllium chloride in 150 ml of anhydrous
toluene. 0.9 g were then added and the reflux maintained for 3
more days. The mixture was cooled at room temperature and
filtered. 1-Hydroxy 4-methoxy 8-benzyloxy 5H indeno (1,2-b)
indol-10-one 19 has not been purified but only characterized
by NMR (400 MHz) CDCl3: 8.7(s(ex) 1H), 8.2(s(ex) 1H),
7.4(m 5H), 7.3(d 1H), 7.2(d 1H), 6.85(dd 1H), 6.8(d 1H),
6.72(d 1H), 5.1(s 2H), 3.9(s 3H).
column. The fractions containing the compound were collected
and concentrated under a 0.02 mm vacuum. The residue was
triturated with water to eliminate some remaining DMSO and
finally with diethyl ether. One hundred and seventy milligrams
(25%) of 2,3-dimethoxy 8-hydroxy 9-iodo 5-methyl indeno
(1,2-b) indol-10-one 18 were obtained. IR (Nujol): 1661,
NMR (300 MHz) DMSO: 9.9(s(ex) 1H), 7.35(d 1H), 7.2(s
1H), 7(s 1H), 6.7(d 1H), 3.9(2s 6H), 3.8(s 3H).
6.2.2.8. 1,4-Dimethoxy 8-hydroxy 10H indeno (1,2-b) indole
(17c). Six grams of compound 12c were dissolved in a mixture
of 200 ml of DMF, 40 ml of methanol and 180 ml of acetic
acid. 0.6 g of Pd (10% on C) were added under a nitrogen
flow. The mixture was hydrogenated under low pressure
(< 80 mbar) at room temperature until the absorption was com-
plete. This mixture was then filtered. After keeping filtrate to
dryness under reduced pressure, the residue was triturated with
100 ml of diethyl ether and filtrated. The precipitate was tritu-
rated with a large excess of isopropanol and filtrated. Isopro-
panol filtrate was kept to dryness and the residue, purified by
silica chromatography, developed with a mixture of dichloro-
methane and methanol (99:1). Some impurities were eliminated
by this purification process and 2.7 g (59%) of 1,4-dimethoxy
8-hydroxy 10H indeno (1,2-b) indole 17c were obtained this
way. Table 16 summarizes the data of the compounds obtained
by the same way.
6.2.2.6. 1,8-Dihydroxy 4-methoxy 5H indeno (1,2-b) indol-10-
one (15). Compound 19 obtained above was debenzylated un-
der conditions described for compound 2a. 1.26 g (88%) of
1,8-Dihydroxy 4-methoxy 5H indeno (1,2-b) indol-10-one 15
were isolated. M.P.: >>260 °C, IR: 1616, 1653, NMR
(300 MHz) DMSO: 11.9(m(ex) 1H), 8.8(m(ex) 2H), 7.2(d
1H), 7(d 1H), 6.85(d 1H), 6.7(d 1H), 6.55(dd 1H), 3.9(s 3H).
6.2.2.7. 2,3-Dimethoxy 8-hydroxy 9-iodo 5-methyl indeno (1,2-
b) indol-10-one (18). Five hundred milligrams of compound
16aj were dissolved at 40 °C in a mixture of n-heptane
(10 ml), anhydrous DMSO (50 ml) acetic acid (15 ml). To this
red solution were added simultaneously: 210 mg of potassium
iodide, 150 mg potassium iodate and 500 mg of iodine. After
6 h at the same temperature, the reaction mixture was cooled to
–5 °C, and 1.5 ml of a 50% solution of sodium bisulfite, added.
After warming up the mixture to room temperature, the volatile
solvents were eliminated under reduced pressure. This mixture
was injected in a preparative HPLC developed with a mixture
of DMSO/H2O (65:35 to 73:27) on a LICHROPREP RP-18
6.3. Membranes
As a first step towards describing a precise and complete
pharmacology of the best compounds synthesized during the

M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
319
Table 16
Analytical data
A
M.p. (°C)
193
Yield (%)
59
IR (Nujol suspension)
Not characteristic
NMR (δ ppm)
17c
17b
1,4-Dimethoxy
DMSO: 10.9(sl(ex) 1H), 8.65(m(ex) 1H), 7.25(d 1H), 6.9(d 1H), 6.8(s
1H), 6.75(d 1H), 6.55(d 1H), 3.9(s 3H), 3.8(s 3H), 3.5(s 2H)
DMSO (300 MHz): 10.8(sl(ex) 1H), 8.6(m(ex) 1H), 7.2(d 1H), 6.8(2d
2H), 6.6(d 1H), 6.5 (dd 1H), 3.9(2s 6H), 3.6(s 2H)
2,4-Dimethoxy
197–198
56
Not characteristic
present work, we chose to characterize their capacity to inter-
fere with the MT3 binding of iodo-melatonin. In brief, male
Syrian hamster brains were obtained from Charles River
Breeding Laboratories. Organs were carefully dissected, inten-
sively washed in ice-cold PBS and snap-frozen in liquid nitro-
gen. They were maintained at –80 °C until further use. Mem-
branes were prepared by the following procedure, all steps
being performed at 4 °C. The tissues were thawed, chopped
using a surgical blade and resuspended in 5 volumes of
50 mM Tris–HCl (pH 7.5) containing 250 mM sucrose, 1 mM
CaCl2 and protease inhibitors as a cocktail commercialized by
Boehringer Mannheim (one tablet of Complete™ in 50 ml).
The nuclei and unbroken material were pelleted twice by
10 min centrifugation at 280 × g. The supernatants were
pooled, fivefold diluted with 20 mM Tris–HCl (pH 7.5) con-
taining 1 mM CaCl2 and protease inhibitors and were subjected
to a final, 60-min, centrifugation at 400,000 × g. The mem-
brane pellets were resuspended with a Dounce as a 2–4 mg
ml–1 protein suspension, as measured by the method of Lowry
et al. [32] adapted for membrane proteins (DC Protein assay,
BioRad) using BSA as a standard. Membrane preparations
were flash-frozen in liquid nitrogen and were stored at
–280 °C until use. Membranes from HEK293 cells expressing
MT1 or MT2 and membranes from Syrian hamster kidney or
brain are referred to as MT1, MT2 and MT3 membranes, re-
spectively, throughout the text.
liquid (Microscint 20, Packard) and counting in a β scintillation
counter (TopCount NXT, Packard).
Binding experiments on MT3 kidney membranes were per-
formed in 20 mM Tris–HCl (pH 7.5) buffer containing 1 mM
CaCl2 (binding buffer) in a final volume of 150 ml. Filtrations
were performed through 96-well glass-fiber supports (GF/B
Unifilter, Packard) presoaked for 2 hours before use in 0.3%
(v:v) polyethyleneimine and rinsed extemporaneously three
times with 200 ml per well of binding buffer. After sample
filtration, the filters were rinsed once with 100 ml per well of
binding buffer. The filtration plates were disposed directly onto
a Multiscreen filtering apparatus (Millipore) connected to a va-
cuum pump, allowing rapid filtration after the samples were
loaded using a 96-well pipetting device (Transtar, Costar).
Radioactivity was measured as described above. Unless other-
wise stated, incubations were performed for 30 min at room
temperature (20–25 °C) using 2-[125I]-I-MCANAT as the spe-
cific radioligand. Results are expressed as the specific binding,
i.e. the total binding corrected for the non-specific binding.
References
[1] J. Arendt, S. Deacon, J. English, L. Morgan, J. Sleep Res. 4 (1995) 74–
79.
[2] P. Delagrange, J. Atkinson, J.A. Boutin, L. Casteilla, D. Lesieur, R. Mis-
slin, S. Pelissier, P. Renard, J. Neuroendocrinol. 15 (2003) 442–448.
[3] J.A. Boutin, V. Audinot, G. Ferry, P. Delagrange, Trends Pharmacol.
Sci. 26 (2005) 412–419.
6.4. Binding
[4] O. Vakkuri, J. Leppaluoto, O. Vuolteenaho, Acta Endocrinol. 106 (1984)
152–157.
[5] S.M. Reppert, D.R. Weaver, T. Ebisawa, Neuron 13 (1994) 1177–1185.
Binding experiments on MT1 and MT2 membranes were
realized as described before [33]. Briefly, samples (10 mg of
proteins) were incubated for 2 h at 37 °C with 25 pM (MT1) or
170 pM (MT2) 2-[125I]-melatonin (2200 Ci mmol–1, NEN,
Boston, MA) in the presence (non-specific binding) or not (to-
tal binding) of 10 mM melatonin and with varying concentra-
tions of test drugs. Incubations were carried out in triplicates in
96-well microplates and were terminated by filtration through
96-well format glass-fiber plates (GF/B Unifilter, Packard)
using a Filtermate (Packard) apparatus. Membranes were then
washed three times with 2 ml of 50 mM Tris–HCl (pH 7.5)
buffer before the addition of 30 ml per well of scintillation
[6] S.M. Reppert, C. Godson, C.D. Mahle, D.R. Weaver, S.A. Slaugenhaupt,
J.F. Gusella, Proc. Natl. Acad. Sci. USA 92 (1995) 8734–8738.
[7] T. Ebisawa, S. Karne, M.R. Lerner, S.M. Reppert, Proc. Natl. Acad. Sci.
USA 91 (1994) 6133–6137.
[8] M.J. Duncan, J.S. Takahashi, M.L. Dubocovich, Endocrinology 122
(1988) 1825–1833.
[9] D.S. Pickering, L.P. Niles, Eur. J. Pharmacol. 175 (1990) 71–77.
[10] P. Paul, C. Lahaye, P. Delagrange, J.P. Nicolas, E. Canet, J.A. Boutin, J.
Pharmacol. Exp. Therap. 290 (1999) 334–340.
[11] E.J. Molinari, P.C. North, M.L. Dubocovich, Eur. J. Pharmacol. 301
(1996) 159–168.
[12] V. Leclerc, S. Yous, P. Delagrange, J.A. Boutin, P. Renard, D. Lesieur,
J. Med. Chem. 45 (2002) 1853–1859.

320
M.-F. Boussard et al. / European Journal of Medicinal Chemistry 41 (2006) 306–320
[13] F. Mailliet, G. Ferry, F. Vella, S. Berger, F. Cogé, P. Chomarat, C. Mal-
let, S.P. Guénin, G. Guillaumet, M.C. Viaud, S. Yous, P. Delagrange,
J.A. Boutin, Biochem. Pharmacol. 71 (2005) 74–88.
[14] W. Carruthers, N. Evans, J. Chem. Soc. (1974) 1523–1525 (Perkin I).
[15] I.A. Koten, R.J. Sauer, Org. Synth. Coll. V (1973) 145–147.
[16] S. Eskola, S. Hyrkko, Suom. Kemistil. 21 (1948) 33.
[17] J. Estevez, R. Estevez, L. Castedo, Tetahedron Lett. 34 (1993) 6479–
6480.
[25] J.B. Serle, R.F. Wang, W.M. Peterson, R. Plourde, B.R. Yerxa, J. Glau-
coma 13 (2004) 385–388.
[26] F. Mailliet, G. Ferry, F. Vella, K. Thiam, P. Delagrange, J.A. Boutin,
FEBS Lett. 578 (2004) 116–120.
[27] C.E. Foster, M.A. Bianchet, P. Talalay, Q. Zhao, L.M. Amzel, Biochem-
istry 38 (1999) 9881–9886.
[28] L. Buryanovskyy, Y. Fu, M. Boyd, Y. Ma, T.C. Hsieh, J.M. Wu, Z.
Zhang, Biochemistry 43 (2004) 11417–11426.
[18] C.D. Jones, M.G. Jevnikar, A.J. Pike, M.K. Peters, L.J. Black, A.R.
Thompson, J.F. Falcone, J.A. Clemens, J. Med. Chem. 27 (1984) 1057–
1066.
[19] S.L. Shapiro, K. Geiger, J. Youlus, L. Freedman, J. Org. Chem. 25
(1960) 1860.
[20] E.D. Amstutz, J. Am. Chem. Soc. 72 (1950) 3420–3423.
[21] H. Sharghi, F. Tamaddon, Tetrahedron 52 (1996) 13623–13640.
[22] M.F. Boussard, J.P. Guette, M. Wierzbicki, P. Beal, J. Fournier, M. Bou-
langer, O. Della-Zuana, J. Duhault, Arzneim.-Forsch./Drug Res. 50
(2000) 1084–1092.
[23] A.S. Eison, U.L. Mullins, Life Sci. 53 (1993) 393–398 (1993).
[24] J. Pintor, T. Pelaez, C.H.V. Hoyle, A. Peral, Br. J. Pharmacol. 138
(2003) 831–836.
[29] Y. Fu, L. Buryanovskyy, Z. Zhang, Biochem. Biophys. Res. Commun.
336 (2005) 332–338.
[30] J.A. Boutin, F. Chatelain-Egger, F. Vella, P. Delagrange, G. Ferry,
Chem. Biol. Inter. 151 (2005) 213–228.
[31] F. Vella, G. Ferry, P. Delagrange, J.A. Boutin, Biochem. Pharmacol. 71
(2005) 1–12.
[32] O.H. Lowry, N.J. Rosebrouhg, A.L. Farr, R.J. Randall, J. Biol. Chem.
193 (1951) 265–275.
[33] V. Audinot, F. Mailliet, C. Lahaye-Brasseur, A. Bonnaud, A. Le Gall, C.
Amossé, S. Dromaint, M. Rodriguez, N. Nagel, J.P. Galizzi, B. Mal-
paux, G. Guillaumet, D. Lesieur, F. Lefoulon, P. Renard, P. Delagrange,
J.A. Boutin, Naunyn Schmiedebergs Arch. Pharmacol. 367 (2003) 553–
561.