Substituted benzopyranobenzothiazinones. Synthesis and estrogenic activity on MCF-7 breast carcinoma cells

Article abstract of DOI:10.1016/S0223-5234(00)01207-1

In the search for new agents with estrogenic activity mediated by estrogen receptors (ER), six 6,12-dihydro-1-benzopyrano[3,4-b][1,4]benzothiazin-6-ones 3a-f were synthesized. These compounds were readily prepared by the addition of 2-aminothiophenol 2 to substituted 4-hydroxycoumarin derivatives 1a-e. The estrogenic effect has been evaluated on the proliferation of MCF-7 breast adenocarcinoma cells and the specificity of described compounds was evaluated by the inhibition of their effect by ICI 182,780, an antiestrogenic compound. Among the compounds tested, 6,12-dihydro-3-methoxy-1-benzopyrano[3,4-b][1,4]benzothiazin-6-one 3e and 6,12-dihydro-3-hydroxy-1-benzopyrano[3,4-b][1,4]benzothiazin-6-one 3f exhibited an ER-dependent proliferation and a high binding affinity to ER, but a moderate capacity to activate the transcription of a reporter gene. Their pharmacological profiles are defined by their binding properties and their mechanism of action by computational modelling studies.

Full text of DOI:10.1016/S0223-5234(00)01207-1

Eur. J. Med. Chem. 36 (2001) 127–136
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(00)01207-1/FLA
Original article
Substituted benzopyranobenzothiazinones. Synthesis and estrogenic activity on
MCF-7 breast carcinoma cells
Yves Jacquota, Laurent Bermontb,c, Herve´ Giorgia, Bernard Refouveleta*, Gerard. L. Adessib,
Edwige Daubrossec, Alain Xiclunaa
aEquipe de Chimie Therapeutique, Faculte´ de Me´decine et de Pharmacie, Place Saint-Jacques, F-25030 Besanc¸on, France
bService d’Oncologie et d’Endocrinologie Mole´culaires and CRI 96.01, 240, route de Dole, F-25000 Besanc¸on, France
cEquipe de Biochimie Pharmaceutique, Faculte´ de Me´decine et de Pharmacie, Place Saint-Jacques, F-25030 Besanc¸on, France
Received 5 June 2000; revised 11 November 2000; accepted 1 December 2000
Abstract – In the search for new agents with estrogenic activity mediated by estrogen receptors (ER), six 6,12-dihydro-1-benzopy-
rano[3,4-b][1,4]benzothiazin-6-ones 3a–f were synthesized. These compounds were readily prepared by the addition of 2-aminothio-
phenol 2 to substituted 4-hydroxycoumarin derivatives 1a–e. The estrogenic effect has been evaluated on the proliferation of MCF-7
breast adenocarcinoma cells and the specificity of described compounds was evaluated by the inhibition of their effect by ICI
182,780, an antiestrogenic compound. Among the compounds tested, 6,12-dihydro-3-methoxy-1-benzopyrano[3,4-
b][1,4]benzothiazin-6-one 3e and 6,12-dihydro-3-hydroxy-1-benzopyrano[3,4-b][1,4]benzothiazin-6-one 3f exhibited an ER-dependent
proliferation and a high binding affinity to ER, but a moderate capacity to activate the transcription of a reporter gene. Their
pharmacological profiles are defined by their binding properties and their mechanism of action by computational modelling studies.
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS
benzopyranobenzothiazinones / 2D-NMR / binding assay / estrogenic activity / SAR
1. Introduction
and the steroidal synthetic compound ethynylestra-
diol [5] (figure 1). Nevertheless, what distinguishes ER
within the steroid receptor superfamily is its accom-
modation of a large variety of non-steroidal ligands.
Initially isolated as natural products [6], coumarins
constitute a very relevant family of pharmacological
active compounds. Actually, such derivatives have
been developed as anticoagulant drugs [7], photosen-
sitive drugs [8], potent and selective human dopamine
Even if estrogens have a well-established role in the
growth of hormone-dependant tumours by an estro-
gen receptor (ER)-dependant mitogenic effect in cells
containing ER [1], a member of the nuclear receptor
superfamily [2], they exert numerous favourable activ-
ities in women. From a therapeutical point of view,
17b-estradiol (E2) and derivatives are well known not
only as oral contraceptives, but also in hormone
replacement therapy required in bone loss prevention
or in the control of cardiovascular diseases (particu-
lary atherosclerosis) in postmenopausal women [3].
The most widely used estrogens are 17b-estradiol [4]
* Correspondence and reprints.
E-mail address: Bernard.Refouvelet@univ-fcomte.fr
(B. Refouvelet).
Figure 1. Structure of 17b-estradiol and ethynylestradiol.

128
Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
compounds 3e and 3f, their binding affinity, obtained
by Scatchard analysis using a competitive binding
assay with [3H]-17b-estradiol as tracer. Finally, the
ER-dependant proliferative activity of compounds 3e
and 3f was determined, in transient transfection ex-
periment, through the expression of a luciferase re-
porter gene driven by a consensus estrogen response
element (ERE).
Subsequently,
a structure–activity relationship
study (SAR) was undertaken on described and evalu-
ated compounds to rationalise the activity of this new
class of non-steroidal estrogenic derivatives.
Figure 2. Coumarinic estrogen coumestrol.
2. Chemistry
The synthesis of the quadricyclic 6,12-dihydro-1-
benzopyrano[3,4-b][1,4]benzothiazin-6-one
system
3a–e was carried out according to the method of
Tabacovic and co-workers [15].
Figure 3. Structure of ICI 182,780.
The addition of 2-aminothiophenol 2 on the 2,4-
chromandione anionic tautomeric form [16] of substi-
tuted 4-hydroxycoumarins 1a–e was made in the
oxidizing solvent DMSO [17]. This addition was ac-
companied by the elimination of one molecule of
water and conducted to a non-isolable enaminone
intermediate, bis(o-aminophenyl)disulfide (DAPDS)
[18]. Subject to a nucleophilic attack at the 3-position
of the coumarin, this enaminone led to the desired
products 3a–e by intracyclization. This heterocycliza-
tion was managed by the sulfurꢀsulfur bond scission,
because of the very reactive 3-position of the substi-
tuted coumarins [19] (figure 4).
D4 antagonists [9], non-peptidic HIV protease in-
hibitors [10] or as antibiotic agents, such as novo-
biocin [11].
Among all the potential therapeutical applications
revealed by coumarins, some estrogenic-like activities
have been discovered, leading to the development of
the natural potent non-steroidal estrogenic compound
coumestrol [12] (figure 2).
The aim of our study was to synthesize a series of 2
or 3-substituted benzopyranobenzothiazin-6-ones and
to test their estrogenic activity in vitro.
We describe the synthesis, the pharmacological
profile and the design study of 6,12-dihydro-1-ben-
zopyrano[3,4-b][1,4]benzothiazin-6-ones 3a–f. Their
The expected compounds crystallised with DMSO
when cooling to room temperature, and then were
isolated by filtration under vacuum. Their structure
1
1
was assigned from UV, FTIR, H-NMR spectra, 13C-
structure was elucidated by UV, FTIR, H-NMR,
13C-NMR, 2D-NMR (J, l spectroscopy, correlation
spectroscopy (COSY), heteronuclear multiple bond
correlation (HMBC) and heteronuclear multiple
quantum coherence (HMQC)) experiments and by
EI-mass spectral data.
NMR spectra, 2D-NMR experiments (J, l experi-
ment, COSY, HMQC, HMBC), EI-MS spectra and
elemental analysis.
The synthesis of the hydroxylated derivative was
unsuccessful with the classic methods using Lewis
acid BBr3 or the hard acid and soft nucleophile sys-
tem EtSH–AlCl3 [20]. Nevertheless, the use of the
strong acid HI in acetic acid in the presence of acetic
anhydride [21] successfully led to the desired product
3f by deprotection of the methoxy group of com-
pound 3e (figure 5).
The proliferative activity of these compounds was
tested on MCF-7 breast cancer cells and the involve-
ment of ER was tested through the capacity of ICI
182,780 (figure 3), or faslodex, a pure antiestrogenic
derivative [13, 14], to displace the tested molecules
and to reverse their effect. We reported their ability to
stimulate the proliferation of MCF-7 estrogen-sensi-
tive mammary tumour cells, and for the most potent
The UV spectra of the synthesized compounds ex-
hibited three absorption maxima: a coumarinic chro-

Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
129
mophore at 451 nm, a cinnamoyl chromophore at 315
nm and an aminophenyl chromophore at 290 nm [18].
1
The H–1H shift COSY, J, l spectroscopy, HMQC
and HMBC were successfully performed to establish
the interfragment relationship, and to assign the pro-
ton and carbon signals of compound 3e as shown in
figure 6.
In the HMBC spectrum of 3e, H-1 at l 8.10 corre-
lates with four aromatic carbon signals C-2 (l 129),
C-3 (l 163), C-4a (l 154) and C-1a (l 119). H-2 at l
7.02 is correlated with three aromatic carbon signals
C-3 (l 163), C-4 (l 102) and C-4a (l 154). Finally,
H-4 at l 6.98 is correlated with four aromatic carbon
signals, precisely C-2 (l 129), C-3 (l 163), C-4a (l
154) and C-1a (l 119) of the coumarin ring. More-
over, H-12 at l 9.00, the mobile proton of the amino
group, and the four carbon signals C-11 (l 118), C-1a
(l 119), C-7a (l 107) and C-6a (l 88) of the benzoth-
iazine moiety, were correlated. Furthermore, H-11 at
l 6.98 and carbon signals C-11a (l 138), C-10 (l 126)
were also correlated.
1
Figure 4. Synthesis of substituted benzopyranobenzothiazi-
nones 3a–e.
These results were supported by a J correlation of
H-1, H-2 and H-4 to, respectively, C-1 (l 124), C-2 (l
129) and C-4 (l 102) in the HMQC spectrum.
3. Biological results and discussion
The biological profiles of these new non-steroidal
molecules were evaluated by their capacity to induce
the proliferation of MCF-7 breast carcinoma cells [22,
23], an ER-positive breast cancer cell line. The prolif-
erative effect of these compounds was compared to
the one observed with 17b-estradiol (table I) and the
ER implication in cell proliferation was assessed with
the pure and highly selective antiestrogen ICI 182,780
[24, 25].
Figure 5. Synthesis of compound 3f.
The stimulation of MCF-7 cells with various con-
centrations of 3a–f compounds was performed for 4
days (see Section 5) [26]. At the end of this period, an
estimation of cell proliferation was performed using
tetrazolium salt WST-1 (figure 7). The implication of
estrogen receptors in the action of these compounds
was appraised using ICI 182,780 at 1 mM. The results
for compounds 3a–f (table I) showed that 3a induced
a moderate (<20%) but significant increase (P=8.7×
10−5) of cell proliferation for the highest concentra-
tion tested (1 mM).
On the other hand, MCF-7 cells stimulation by 1
mM of compound 3e induced a significant growth
Figure 6. HMBC spectrum (H“C) of compound 7.

130
Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
Table I. Cell proliferation results for compounds 3a–f.
Cell proliferation (% of control)
49.2
a
95% Confidence limit (%)
P
17b-Estradiol
ICI 182,780
E2+ICI
1 nM
20.2
6.7
B0.01
B0.01
B0.01
1 mM
−11.8
−8.9
1 nM+1 mM
1.1
Compound 3a
Compound 3b
Compound 3c
Compound 3d
Compound 3e
Compound 3f
0.1 mM
−2.5
12.6
−3.2
1.5
4.7
2.3
6.8
6.7
NS b
B0.01
NS
1 mM
0.1 mM+ICI c
1 mM+ICI
NS
1 mM
10 mM
1 mM+ICI
10 mM+ICI
−1.6
2
−7.8
−6.8
1
B0.01
NS
B0.01
B0.01
2.6
4.1
3.2
1 mM
10 mM
1 mM+ICI
10 mM+ICI
2.4
1.5
2.4
1
B0.01
NS
NS
3.4
5.3
3.3
−1.1
NS
1 mM
10 mM
1 mM+ICI
10 mM+ICI
−13.9
−2.7
−25.9
−15.2
4.8
9.4
4.3
4.5
B0.01
NS
B0.01
B0.01
0.1 mM
1 mM
0.1 mM+ICI
1 mM+ICI
40.3
58.2
16.3
9.6
9.9
14
9.7
B0.01
B0.01
B0.01
NS
14.4
1 mM
10 mM
1 mM+ICI
10 mM+ICI
−5.6
52.6
−8.7
−9.1
5.1
10.2
4.1
NS
B0.01
B0.01
B0.01
1.1
a P value of t-test used to compare cell proliferation induced by the compounds with that induced in control cells.
b NS: not significant, i.e. P\0.05.
c ICI (ICI 182,780) concentration used is 1 mM.
response on MCF-7 cells of about 58.2 14% com-
pared to the unstimulated cells (P=2.0×10−8) after 4
days of stimulation. The magnitude of the maximum
effect induced by compound 3e was found to be not
different from the one observed in MCF-7 cells
treated with 1 mM E2 (49.2 20%). The proliferative
activity of compound 3e was, as for the other com-
pounds, almost completely inhibited with 1 mM ICI
182,780, demonstrating that such an effect was ER
dependent.
Therefore, we focused on compound 3e and we
determined the concentration which induced the half-
maximum effect, EC50 (figure 8). The stimulation of
MCF-7 cells with compound 3e from 10−11 to 10−5
M
allowed one to find that the EC50 was obtained with a
Figure 7. Effect of compounds 3a and 3e on MCF-7 cell
proliferation.
concentration of 2.4 1.6×10−8 M (n=4).

Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
131
in the same conditions. Stimulation on MCF-7 cells
by 10 mM of compound 3f induced a significant
proliferative response (52.6 10.2%) compared to the
unstimulated cells (P=5.0×10−8) after 4 days of
stimulation which is not different from 3e. Activity of
compound 3f was, as for 3e, almost completely inhib-
ited with 1 mM ICI 182,780, demonstrating that such
an effect was ER-dependent. The EC50 of compound
3f was obtained with a concentration of 5.9 0.4.10−6
M (n=4) (data not shown). Scatchard analyses
showed that compound 3f was bound to only one
class of estrogen receptors with high affinity Kd=
1.60 0.30 nM, and with a binding capacity of
39 677 14 537 sites per cell. These results show that
3f has the same behaviour as 3e. The evaluation of
the ER–3f complex capacity to bind an ERE se-
quence and to activate the luciferase reporter gene
transcription was then performed and showed a sig-
nificant capacity of 10−5 M 3f (2.64-fold, P<0.01) to
increase the luciferase activity compared to unstimu-
lated cells.
Figure 8. Concentration dependence of compound 3e on
MCF-7 cell proliferation.
Since numerous ERa and ERb variants are ex-
pressed in MCF-7 cells [27], we assume that 3e and 3f
could bind one class of ER implicated in cell prolifer-
ation, but we cannot exclude the implication of an-
other type of receptor in the observed effects of 3e
and 3f. The different sensibility of MCF-7 cells to the
proliferative effect of 3e and 3f derivatives could
partly rely upon a slower degradation of 3e in MCF-7
cells.
The estrogenic activity of compound 3e mediated
by estrogen receptors could be explained by a molecu-
lar modelling approach compared to E2. The
molecules constructed into 3D molecular structures
were minimised and the calculation performed until
Figure 9. Molecular dynamic of compound 3e at 300 K during
100 ps.
In an attempt to obtain further information about
the relationships between 3e and ER, its affinity for
estrogen receptors as well as the activation of an
ERE-containing promoter driving the luciferase gene
were determined. Scatchard analyses showed that the
number of estrogen receptors, evidenced with unla-
belled diethylstilbestrol (DES), was 84 317 33 687
sites per cell and suggest that the compound 3e was
bound to only one class of estrogen receptor with a
high affinity Kd =2.54 0.91 nM, since the number of
sites per cell was 25 264 13 655. The evaluation of
the ER–3e complex capacity to bind an ERE se-
quence and to activate the luciferase reporter gene
transcription was then performed and showed a mod-
erate but significant capacity of 10−6 M of 3e (1.34-
fold, P<0.01) to activate the luciferase transcription
the maximum RMS derivative was less than 0.001
−1
,
kcal A . The molecular dynamic study, carried out
at 300 K during 100 ps, showed a very planar struc-
ture 3e, with a maximum energy variation of 3
kcal mol−1 (figure 9). The planarity of this new substi-
tuted aromatic system does not appear to be an
obstacle to a significant pharmacological activity,
even if it is agreed that good ligands for the estrogen
receptors need some degree of thickness in the central
hydrophobic portion of the ligand binding domain
(LBD).
compared with unstimulated cells treated with 10−8
M
E2 (3.52-fold, P<0.01) as positive control.
Moreover, the volume of 3e and 3f occupies less
3
,
Moreover, in order to explore the implication of
the 3-methoxy group of compound 3e in this estro-
genic activity, we tested the free phenolic derivative 3f
total space than E2 (208.6 and 193.10 A , respectively,
3
3
,
,
versus 232.4 A for E2) in the 450 A binding pocket
of the ER. Compared to E2, the excedentary volume

132
Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
3
,
with a binding of these derivatives in the pocket of
estradiol in the ER. According to the studies of
Duax et al. [28], the observation that the excedentary
volume is principally localised in the D ring region is
not an obstacle for the activity, since ER-LBD is
flexible in this D-ring region [29, 30]. Moreover, the
planarity induced by the keto group should have a
better receptor match than the non-planar B ring
[31].
of 3e, 52.06 A , concerns the methyl group of the
methoxy, the keto group and the angular aromatic D
ring, whereas the excedentary volume of 3f, 36.83
A , concerns only the keto group and the angular
aromatic D ring. Equally, when 3e is compared to
3
,
3
,
E2, the excedentary volume of E2 is 76.33 A , and
concerns the C ring and the cyclopentyl D ring of E2
(figure 10). These results concerning volumes of 3e
and 3f compared to the volume of E2 are compatible
Furthermore, the charge density study of the ben-
zopyranobenzothiazinone core clearly shows some
analogies with E2. By molecular computational
methods, we determined that the A ring of com-
pound 3e, compared to E2, had a similar charge
density, even if the methoxy group, which might
mimic the 3-hydroxyl group of 3f, contributes to
reducing the polarity of the 3-substituent. The calcu-
lation of the net atomic charge of the phenolic hy-
droxylic oxygen of 3f is −0.38, suggesting that the
phenol is a poor H-bond acceptor and is more H-
bond donor as observed with the phenolic group of
E2. Actually, this result has been already observed
with E2; even if the phenol of E2 might be an H-
bond donor and H-bond acceptor, there is evidence
that the donor function is more important, with a
net atomic charge of −0.253 [29].
Nevertheless, as we showed in the present study,
the presence of a methoxy group is also implicated
in the estrogenic activity as it has been already
demonstrated with some non-steroidal estrogens such
as TACE (chlorotrianisene), a triarylethylenic estro-
gen, or centchroman, a chroman derivative. More-
over, it is now admitted that the methyl group of the
3-methoxy substituent is removed by metabolism, re-
sulting in free phenol. However, the polar sub-
stituent at the 3-position is indispensable since the
loss of this substituent showed a significant loss from
20 to 120% of activity [32–34]. Actually, it is known
that the apolar surface of the ligand interacts hydro-
phobically with the LBD of the estrogen receptor a,
whereas the methoxy or the hydroxy group interacts
with the His 524 (helix H11) and the Glu 419 (loop
6–7), a very conserved residue which plays a special
role in polar ligand binding contacts, or with the
Glu 353 (helix H3) of the LBD. Unfortunately, since
the ligand orientation in the ligand binding cavity,
limited to ERa, is still uncertain, it is difficult to
come to a specific conclusion about the binding
mode of the compound 3e to the ER receptor [34].
Figure 10. Excedentary volume of: (a) 3e compared to E2; and
(b) E2 compared to 3e.

Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
133
4. Conclusions
triplet (t), multiplet (m). EI-MS was done on a Nermag
R10-10H apparatus with a Coniphot detector and an
ionizing voltage of 70 eV in the direct-inlet mode. UV
spectra were recorded to a Shimadzu UV-160A spec-
trophotometer. Thin-layer chromatography (TLC) was
carried out on an Alugram Sil G/UV254 plate with
appropriate solvents. Microanalysis were carried out by
the Service Central d’Analyses, Centre National de la
Recherche Scientifique, Vernaison (France). Elemental
analyses indicated by the symbols of the elements were
within 0.4% of the theoretical values for C, H and N.
Yields were not optimised.
In conclusion, our study provides evidence that
non-steroidal compounds 3e and 3f act positively on
MCF-7 cell proliferation and that this effect is medi-
ated by estrogen receptors as evidenced by the inhibi-
tion with ICI 182,780. The binding studies of these
compounds to ER as well as their capacity to induce
the transcription of a reporter gene, driven by a
consensus ERE sequence, allowed one to characterise
the relationship between 3e, 3f and ER. Moreover,
the volumes of 3e and 3f are totally compatible with
the fact that such compounds may fit into the LBD of
ER. This activity is easily assimilated to the methoxy
group at the 3-position of 3e, and to the free phenolic
group of 3f, which could mimic the phenolic group of
E2, playing a fundamental role in the formation of
the ligand–receptor complex. On the other hand, the
non-substituted compounds and the two non-polar
substituted compounds present a faint proliferative
activity or none at all. Therefore, the effects may be
useful in the treatment of pathologies in which estro-
gen-like properties play a fundamental role. Thus,
compounds 3e and 3f could be new candidates that
could contribute to the development of a large chemi-
cal library of related compounds by a combinatorial
synthesis approach. Further studies are needed to
elucidate with precision the type of receptor involved
in the activity of 3e and 3f and their mechanism of
action (binding mode).
5.1.1. General procedure for the condensation of
4-hydroxycoumarin derivatives 1a–e with
2-aminothiophenol 2
4-hydroxycoumarin derivatives 1a–e (61.7 mmol) and
2-aminothiophenol 2 (11.56 g, 92.5 mmol) were added
to 20 mL of DMSO and heated to 150°C for 10 h. On
cooling at room temperature, the products crystallised
and are filtered under vacuum.
5.1.1.1. 6,12-Dihydro-1-benzopyrano[3,4-b][1,4]-
benzothiazin-6-one (3a)
Prepared by coupling 4-hydroxycoumarin 1a with 2-
aminothiophenol 2. Red amorphous powder. (60%),
mp>300°C. IR (KBr) (cm−1): 3335, 3010, 1668, 1618,
1
1015, 737; H-NMR (DMSO-d6) l: 6.90–7.10 (4H, m,
4
Ar-H), 7.50 (2H, m, Ar-H), 7.65 (1H, dd, J=6.5 Hz,
3
3J=7.5 Hz, H-2), 8.20 (1H, m, J=7.5 Hz, H-1), 9.00
(1H, s, NH). MS m/z: 267 [M+]; Anal. C15H9NO2S
(C,H,N).
5. Experimental protocols
5.1.1.2. 6,12-Dihydro-2-methyl-1-benzopyrano-
[3,4-b][1,4]benzothiazin-6-one (3b)
Prepared by coupling 4-hydroxy-2-methylcoumarin 1b
with 2-aminothiophenol 2. Pale red amorphous powder.
(40%), mp>300°C. IR (KBr) (cm−1): 3348, 3010, 2950,
5.1. Chemistry
All melting points (mp) were obtained with a Kofler
Heizbank Reichert 18.43.21 and were uncorrected. In-
frared spectra were recorded on a Shimadzu FTIR-
8201PC spectrometer in potassium bromide pellets (w in
cm−1). 1H-NMR spectra, 13C-NMR spectra and 2D-
NMR experiments (J,l experiment, COSY, HMQC,
HMBC) were recorded on a Bruker AC 200 spectrome-
ter. The samples were dissolved in DMSO-d6 relative to
tetramethylsilane as internal standard. All measure-
ments were performed at 293 K. The chemical shift
values are reported in parts per million (ppm, l units)
and spin–spin coupling J were exposed in Hz. The
following abbreviations are used: singlet (s), doublet (d),
1
1655, 1038, 739; H-NMR (DMSO-d6) l: 2.20 (3H, s,
3
CH3), 6.80–7.10 (4H, m, Ar-H), 7.25 (1H, d, J=8 Hz,
H-4), 7.35 (1H, d, 3J=8 Hz, H-3), 7.90 (1H, s, H-1), 8.9
(1H, s, NH). MS m/z: 281 [M+]; Anal. C16H11NO2S
(C,H,N).
5.1.1.3. 2-Chloro-6,12-dihydro-1-benzopyrano-
[3,4-b][1,4]benzothiazin-6-one (3c)
Prepared by coupling 6-chloro-4-hydroxycoumarin 1c
with 2-aminothiophenol 2. Orange amorphous powder.
(50%), mp=230°C. IR (KBr) (cm−1): 3323, 3065, 1672,
1616, 1030, 745; 1H-NMR (DMSO-d6) l: 6.80–7.00

134
Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
3
(4H, m, Ar-H), 7.36 (1H, s, J=8.0 Hz, H-4), 7.65 (1H,
containing phenol red (Sigma, Saint Quentin, Fallavier,
France) and supplemented with 15 mM HEPES, 2 mM
glutamine, 1% antibiotic cocktail (streptomycin 10
mg mL−1, penicillin 10 000 U mL−1, amphotericin B 25
mg mL−1) and 10% fetal bovine serum (FBS). MCF-7
cells were tested for the absence of mycoplasma before
starting the experiments. For the experiments, cells were
trypsinised and seeded in 96-well plates (Falcon, Elvetec,
Venissieux, France) with 5000 cells per well for 4 days in
a phenol red-free DMEM medium supplemented with
15 mM HEPES, 2 mM glutamine, 1% antibiotic cocktail
and 10% desteroided-FBS (dextran-charcoal-treated
FBS).
All compounds (3a–f) were made soluble in DMSO.
For experiments, the DMSO solution was diluted by a
factor 1000 into the stimulation medium [35]. Stimula-
tions were performed with the same medium supple-
mented with various concentrations of compounds in
eight wells for each condition. Positive controls were
performed with 10−9 M 17b-estradiol and results of
proliferation were compared to those obtained with cells
incubated with the medium alone. The implication of
estrogen receptors in the action of these compounds was
tested by the inhibition of their effect on cell prolifera-
tion with a pure antiestrogenic compound, ICI 182,780,
at 10−6 M.
4
3
4
dd, J=4.0 Hz, J=8.0 Hz, H-3), 8.25 (1H, d, J=8.0
Hz, H-1), 8.95 (1H, s, NH). MS m/z=301 [M+]; Anal.
C15H8NO2SCl (C,H,N).
5.1.1.4. 2-Bromo-6,12-dihydro-1-benzopyrano-
[3,4-b][1,4]benzothiazin-6-one (3d)
Prepared by coupling 6-bromo-4-hydroxycoumarin 1d
with 2-aminothiophenol 2. White amorphous powder.
(36%), mp=290°C. IR (KBr) (cm−1): 3323, 3006, 1672,
1618, 1028, 752; 1H-NMR (DMSO-d6) l: 6.50–6.80
3
(4H, m, Ar-H), 7.00 (1H, d, J=10 Hz, H-4), 7.20 (1H,
d, 3J=10 Hz, H-3), 7.95 (1H, d, H-1), 10.8 (1H, s, NH).
MS m/z=347 [M++1]; Anal. C15H8NO2SBr (C,H,N).
5.1.1.5. 6,12-Dihydro-3-methoxy-1-benzopyrano-
[3,4-b][1,4]benzothiazin-6-one (3e)
Prepared by coupling 4-hydroxy-7-methoxycoumarin
1e with 2-aminothiophenol 2. Pink amorphous powder.
(60%), mp=230. IR (KBr) (cm−1): 3335, 3010, 2934,
1670, 1616, 1022, 750; 1H-NMR (DMSO-d6) l: 3.85
(3H, s, OCH3), 6.85 (2H, m, Ar-H), 6.98 (1H, d, J=2.5
Hz, H-4), 6.95 (2H, m, Ar-H), 7.02 (1H, dd, 4J=2.5 Hz,
3J=9.1 Hz, H-2), 8.10 (1H, d, J=9.1 Hz, H-1), 9.01
3
(1H, s, NH). 13C-NMR (DMSO-d6): l 56.8 (OCH3),
88.3 (C-6a), 101.9 (C-4), 106.5 (C-7a), 113.0 (C-9), 117.7
(C-11), 118.7 (C-1a), 124.0 (C-1), 126.0 (C-10), 127.4
(C-8), 128.5 (C-2), 137.6 (C-11a), 147.5 (C-12a), 154.1
(C-4a), 158.0 (C-6), 163.1 (C-3). MS m/z=297 [M+];
Anal. C16H11NO3S (C,H,N).
5.2.2. Cell proliferation
The measurement of cell proliferation was then per-
formed by a colorimetric assay based on the cleavage of
the tetrazolium salt WST-1 (4-[3-(4-iodophenyl)-2-(4-ni-
trophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate) to
formazan by mitochondrial dehydrogenase in viable
cells after 4 days of stimulation. Assays were performed
according to the manufacturer’s protocol (cell prolifera-
tion reagent WST-1, Roche, Meylan, France).
5.1.1.6. 6,12-Dihydro-3-hydroxy-1-benzopyrano-
[3,4-b][1,4]benzothiazin-6-one (3f)
Prepared by refluxing during 1 h 1 mmol of 3e with 10
mL of HI in a mixture of 5 mL of acetic anhydride and
5 mL of acetic acid. On cooling at room temperature,
the product precipitates and is then filtered under vac-
uum. Ochre amorphous powder. (35%), mp>300°C. IR
(KBr) (cm−1): 3415, 3246, 3010, 1664, 1616, 1050, 763;
1H-NMR (DMSO-d6) l: 5.10 (1H, s, NH), 6.60 (3H, m,
Ar-H), 6.90 (1H, d, J=7.2 Hz, H-4), 7.30 (2H, m,
Ar-H), 7.70 (1H, d, 3J=8.6 Hz, H-1), 10.70 (1H, s,
OH). MS m/z=282 [M+−1]; Anal. C15H9NO3S
(C,H,N).
5.2.3. Binding studies in MCF-7 cells
The determination of the 3e and 3f compounds bind-
ing affinity to the estrogen receptor was performed
according to the method described by Taylor et al. [36].
MCF-7 cells (5×104 cells) were grown for 5 days in
24-well plates in phenol red-free DMEM supplemented
with 10% charcoal-treated FBS, 2 mM glutamine and 15
mM HEPES. Thereafter, as confluence was reached,
cells were incubated with 6 nM [3H]-estradiol (specific
activity: 84.1 Ci mmol−1, NEN) and various concentra-
tions of 3e and 3f compounds (0–10−6 M) for 1 h at
37°C in phenol red-free DMEM supplemented with
0.1% bovine serum albumin (BSA). Medium was then
5.2. Pharmacological methods
5.2.1. Cell line and culture
MCF-7 cells were purchased from ATCC (Biovalley,
Conches, France) and maintained in DMEM medium

Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
135
discarded and cells incubated for 30 min at room tem-
perature (RT) with a phosphate buffer pH 7.4 (sodium
phosphate 5 mM, sucrose 0.25 M, glycerol 10%, BSA
0.5%). MCF-7 cells were then rinsed twice with the same
cold phosphate buffer and incubated with 1 mL abso-
lute ethanol for 30 min at RT. Ethanol fractions, con-
taining the bound fraction of estradiol, were then
transferred to counting vials and placed in a b-counter
(Beckman LS 6000 IC). Scatchard analysis was then
performed to determine the dissociation constants (Kd)
and the number of binding sites (N), using a non-linear
adjustment according to the Marquart’s method pro-
cessed on SAS software (SAS Institute, Cary, NC).
Binding studies were performed in triplicate in three
independent experiments.
Kit (Promega). Results are expressed as ratio of the
luciferase activity (normalised with b-galactosidase ac-
tivity) between stimulated over unstimulated cells. Tran-
sient transfection studies were performed in qua-
druplicate in two independent experiments.
5.3. Computational methods
5.3.1. Molecular modelling
Modelling studies were performed on a Silicon
Graphics Iris Indigo 2 running the INSIGHT II soft-
ware package (Biosym/MSI., version 95.0). The struc-
ture of estradiol and the compound 3e were constructed
from standard bond lengths and bond angles using the
SKETCH option from the Builder module. The 2D
sketches of estradiol and compound 3e were fully opti-
mised in vacuo and converted into 3D molecular struc-
tures. After construction, molecules were minimised
using the Steepest Descents method and CFF91
forcefield supported by the Discover program. Calcula-
tions were performed until the maximum RMS deriva-
5.2.4. Transient transfection experiments
Transient transfection of the ERE-wt-tk-Luc and luci-
ferase activity quantification were performed to evaluate
the implication of the estrogen receptors in the action of
compounds 3e and 3f. MCF-7 cells were transiently
transfected with a plasmid containing a consensus ERE
triggering the luciferase gene expression. Transfections
were achieved with the following plasmids: pUC 18
vector containing a consensus ERE linked to the firefly
luciferase reporter gene and the eukaryotic expression
vector pSGD2 containing NLS LacZ from pMMuLV
NLS LacZ [37] which was used as a positive control for
monitoring transfection efficiency. MCF-7 cells, grown
to a 30–40% confluence, were transiently transfected
using Lipofectine® Reagent (Life Technologies, Cergy-
Pontoise, France) for 12 h with 1 mg of each plasmid
and 4 mg Lipofectine®, in 1 mL phenol red-free DMEM,
supplemented with 0.1% charcoal-treated FBS and 2
mM glutamine. Cells were washed twice with 2 mL
phenol red-free DMEM and were subsequently stimu-
lated for 5 h with various concentrations of 3e and 3f,
and with 10−8 M 17b-estradiol as positive control, in
phenol red-free DMEM supplemented with 15 mM
Hepes, 2 mM glutamine, 10% charcoal-treated FBS and
1% antibiotic cocktail. Transfected MCF-7 cells were
then lysed with Reporter lysis buffer (Promega, Char-
bonnie`res, France) and luciferase and b-galactosidase
activities were measured. b-Galactosidase activity was
assayed in 40 mg cellular extracts, in 1 mL buffer Z (60
mM Na2HPO4, 40 mM NaH2PO4, 1 mM MgSO4, 10
mM KCl, 50 mM b-mercaptoethanol) and 200 mL
ONPG (4 mg mL−1 in buffer Z). Samples were incu-
bated at 37°C until a yellow colouring appeared. Luci-
ferase activity was determined using luciferase detection
tive was less than 0.001 kcal A−1. The molecular
,
dynamics at 300 K during 100 ps was done with the
Discover program and the volumes were calculated us-
ing the Search/Compare module.
6. Statistics
Results were expressed as mean 95% confidence
limits. EC50 values were determined from the linear
portion of concentration/proliferation curves by re-
gression analysis, and expressed as a mean SE.
Statistical analysis was performed using Student’s test
(t-test) and probability values (P) below 0.05 were
considered significant. Before performing the t-test,
the data were tested using an F test, for their variance
homogeneities.
Acknowledgements
This work was supported in part by grants from La
Ligue Nationale contre le Cancer and INSERM.
References
[1] Dickinson R.B., McManaway M.E., Lippman M.E., Science
232 (1986) 1540–1543.
[2] Tsai M.J., O’Malley B.W., Annu. Rev. Biochem. 63 (1994)
451–486.

136
Y. Jacquot et al. / European Journal of Medicinal Chemistry 36 (2001) 127–136
[20] Node M., Nishide K., Fuji K., Fujita E., J. Org. Chem. 45
[3] Sun J., Meyers M.J., Fink B.E., Rajendran R., Katzenellenbo-
(1980) 4275–4277.
gen J.A., Katzenellenbogen B.S., Endocrinology 140 (1999)
800–804.
[21] J. Boyd, A. Robertson, J. Chem. Soc. (1948) 174–176.
[4] Wiese T.E., Polin L.A., Palomino E., Brooks S.C., J. Med.
Chem. 40 (1997) 3659–3669.
[22] Wilson S., Ruenitz P.C., Ruzicka J.A., J. Steroid Biochem.
Mol. Biol. 42 (1992) 613–616.
[5] Li J.J., Hou X., Bentel J., Yazlovitskaya E.M., Li S.A., Car-
cinogenesis 19 (1998) 471–477.
[23] Borra´s M., Laios I., El Khissiin A., Seo H.S., Lempereur F.,
Legros N., Leclercq G., J. Steroid Biochem. Mol. Biol. 57
(1996) 203–213.
[24] Lobaccaro C., Pons J.F., Duchesne M.J., Auzou G., Pons M.,
Nique F., Teutsch G., Borgna J.L., J. Med. Chem. 40 (1997)
2217–2227.
[6] A. Este´vez-Braun, A.G. Gonza´lez, Natural Prod. Rep. (1997)
465–475.
[7] Hermodson M.A., Barker W.M., Link K.P., J. Med. Chem. 14
(1971) 167–169.
[25] Nawaz Z., Stancel G.M., Hyder S.M., Cancer Res. 59 (1999)
372–376.
[8] Wulf H., Rauer H., Du¨ring T., Hanselmann C., Ruff K., Wrish
A., Grissmer S., Ha¨nsel W., J. Med. Chem. 41 (1998) 4542–
4549.
[9] Kesten S.R., Heffner T.G., Johnson S.J., Pugsley T.A., Wright
J.L., Wise L.D., J. Med. Chem. 42 (1999) 3718–3725.
[26] Tan A.S., Berridge M.V., J. Immunol. Methods 238 (2000)
59–68.
[27] Pfeffer U., Fecarotta E., Arena G., Forlani A., J. Steroid
Biochem. Mol. Biol. 56 (1996) 99–105.
[10] Skulnik H.I., Johnson P.D., Aristoff P.A., Morris J.K., Lovasz
K.D., J. Med. Chem. 40 (1997) 1149–1164.
[28] Duax W.L., Griffin J.F., Rohrer D.C., Swenson D.C., Weeks
C.M., J. Steroid Biochem. 15 (1981) 41–47.
[11] Crow F.W., Duholke W.K., Farley K.A., Hadden C.E., Hahn
D.A., Kaluzny B.D., Mallory C.S., Martin G.E., Smith R.F.,
Thamann T.J., J. Heterocyclic Chem. 36 (1999) 365–370.
[29] Anstead G.M., Carlson K.E., Katzenellenbogen J.A., Steroids
62 (1997) 268–302.
[30] Fink B.E., Mortensen D.S., Stauffer S.R., Aron Z.D.,
Katzenellenbogen J.A., Chem. Biol. 6 (1999) 205–218.
[12] Micheli R.A., Booth A.N., Livingston A.L., Bickoff E.M., J.
Med. Chem. 5 (1962) 321–335.
[31] Durani S., Anand N., Int. J. Quantum Chem. 20 (1981) 71–83.
[13] Kendra K.L., Katzenellenbogen B.S., J. Steroid Biochem. 28
(1987) 123–128.
[32] El Garrouj D., Aumelas A., Borgna J.L., J. Med. Chem. 36
(1993) 2973–2983.
[14] Wakeling A.E., Bowler J., J. Steroid Biochem. Mol. Biol. 43
(1992) 173–177.
[33] Brzozowski A.M., Pike A.C.W., Dauter Z., Hubbard R.E.,
8
Bonn T., Engstro¨m O., Ohman L., Greene G.L., Gustafsson
J.A, Carlquist M., Nature 389 (1997) 753–757.
[15] Tabakovic K., Tabakovic I., Trkovnik M., Juric A., Trinajstic
N., J. Heterocyclic Chem. 17 (1980) 801–803.
,
[34] Wurtz J.M., Egner U., Heinrich N., Moras D., Mueller-
Fahrnow A., J. Med. Chem. 41 (1998) 1803–1814.
[16] Vanhaelen M., Vanhaelen-Fastre´ R., Pharm. Acta Helv. 51
(1976) 307–312.
[35] Devraj R., Barrett J.F., Fernandez J.A., Katzenellenbogen J.A.,
Cushman M., J. Med. Chem. 39 (1996) 3367–3374.
[17] Yiannios C.N., Karabinos J.V., J. Org. Chem. 28 (1963) 3246–
3248.
[36] Taylor C.M., Blanchard B., Zava D.T., J. Steroid Biochem. 20
(1984) 1083–1088.
[18] Reddy B.S., Darbarwar M., J. Indian Chem. Soc. 62 (1985)
377–379.
[37] Ambrosino C., Cicatiello L., Cobellis G., Addeo R., Sica V.,
Bresciani F., Weisz A., Mol. Endocrinol. 7 (1993) 1472–1483.
[19] N.P. Buu-Ho¨ı, M. Mangane, P. Jacquignon, J. Chem. Soc. (C)
(1966) 50–52.
.