Lead optimization providing a series of flavone derivatives as potent nonsteroidal inhibitors of the cytochrome P450 aromatase enzyme

Article abstract of DOI:10.1021/jm060186y

Following our SAR studies on aromatase inhibitors, new compounds were designed by appropriately modifying the structure of flavone 1 using our previously reported CoMFA model. While the introduction of substituents on the 2-phenyl ring alone did not cause

Full text of DOI:10.1021/jm060186y

J. Med. Chem. 2006, 49, 4777-4780
4777
Brief Articles
Lead Optimization Providing a Series of Flavone Derivatives as Potent Nonsteroidal Inhibitors
of the Cytochrome P450 Aromatase Enzyme
Silvia Gobbi,*,† Andrea Cavalli,† Angela Rampa,† Federica Belluti,† Lorna Piazzi,† Anja Paluszcak,‡ Rolf W. Hartmann,‡
Maurizio Recanatini,† and Alessandra Bisi†
Department of Pharmaceutical Sciences, UniVersity of Bologna, Via Belmeloro, 6, I-40126 Bologna, Italy, and Pharmaceutical and Medicinal
Chemistry, Saarland UniVersity, P.O. Box 151150, D-66041 Saarbru¨cken, Germany
ReceiVed February 17, 2006
Following our SAR studies on aromatase inhibitors, new compounds were designed by appropriately
modifying the structure of flavone 1 using our previously reported CoMFA model. While the introduction
of substituents on the 2-phenyl ring alone did not cause improvement in potency, these modifications and
the removal of the 7-methoxy group led to compounds showing inhibitory activity in the nanomolar range,
comparable to the marketed drug fadrozole.
Introduction
the previously described alignment hypothesis.10 Briefly, the
phenyl group was aligned onto the p-cyanophenyl of S-fadrozole
(Figure 1A), and as a consequence, the chromone nucleus was
found to be fairly well aligned onto the p-cyanophenyl of
R-fadrozole (Figure 1B). In Figure 1C, 1 is shown surrounded
by the CoMFA contour maps of the model of ref 14. In light of
this picture, we thought the activity of 1 (IC50 ) 0.55 µM) could
be improved in two ways: (i) by inserting into the 2-phenyl
ring of the flavone a suitable substituent mimicking the cyano
group of S-fadrozole, eventually enabling the establishment of
an H-bond interaction with the Ser478 of the biological
counterpart15 (2a-c, Table 1); (ii) by removing the 7-methoxy
group of 1 (3a-e, Table 1). The biological activities of the
resulting compounds were then predicted using our CoMFA
model,14 and it turned out that the designed structural modifica-
tions were predicted to improve the aromatase inhibitory potency
with respect to the parent compound 1, whose calculated pIC50
was 6.11 (Table 1).
Compounds 2a-c (Scheme 1) were synthesized starting from
2,4-dihydroxypropiophenone, which was heated with the ap-
propriately substituted benzoyl chloride and the corresponding
sodium salt to give the 7-hydroxy-3-methylflavones 4a,b after
hydrolysis with sulfuric acid. The hydroxy group in position 7
was methylated with dimethylsulfate to give the intermediates
5a,b, which were then brominated with N-bromosuccinimide
to give 6a,b, and the reaction with imidazole led to the desired
compounds (2a,b). For the synthesis of 2c a different route was
used because of the instability of the nitrile in the presence of
sulfuric acid. 2,4-Dihydroxypropiophenone was first methylated
and then reacted with 4-cyanobenzoyl chloride and sodium
4-cyanobenzoate to give the corresponding 7-methoxyflavone
5c, which was then treated as above to obtain the final
compound. For the synthesis of 3a-e (Scheme 1), starting from
2-hydroxypropiophenone and the appropriately substituted ben-
zoyl chlorides and corresponding sodium salts, the flavones 8a,e
were obtained. These were then brominated (9a,e) and reacted
with imidazole to give the final compounds.
One of the main approaches to control postmenopausal
hormone-dependent breast cancer involves the reduction of
plasma and tissue levels of estrogens via the inhibition of a
key enzyme of their biosynthetic process, and research was soon
focused on aromatase.1 This multienzymatic complex, formed
by cytochrome P450 XIX (CYP19) and NADPH-cytochrome
P450 reductase, catalyzes the conversion of androgens to
estrogens through the aromatization of the A ring of androgen
substrates and has been considered a particularly attractive target
for the treatment of hormone-dependent breast cancer. Since
the early discovery of aromatase inhibitory activity in the known
drug aminoglutethimide,2 the extensive SAR studies performed
in the past 30 years to search for potent, selective, and less toxic
compounds have led to the development of a second and a third
generation of nonsteroidal inhibitors, which have recently been
reviewed.3 Among different classes, the azoles still hold a
prominent position as demonstrated by the recently marketed
drugs, which are now considered a valid alternative to tamox-
ifene as a first line treatment for advanced breast cancer,4 and
recently new series of nonsteroidal aromatase inhibitors have
been synthesized by groups working in the field.5-8
We have been interested for some years in the study of
aromatase inhibitors, and on the basis of a CoMFA analysis,9,10
new aromatase inhibitors were synthesized11 in which the key
features of the azole family of aromatase inhibitors (nitrogen-
containing heterocycle and H-bond accepting group) were
inserted into different scaffolds, such as chromone, xanthone,
and flavone, which are known in the literature to bear aromatese
inhibition properties.12,13 Starting from these results, we per-
formed additional SAR studies, obtaining several series of new
aromatase inhibitors. In this paper we report some modifications
performed on the earlier described flavone 111 to improve the
biological activity of this lead compound. 1 was aligned into
the recently updated enantioselective CoMFA model14 following
* To whom correspondence should be addressed. Phone/fax: +39 051
2099732/34. E-mail: silvia.gobbi@unibo.it.
† University of Bologna.
The new compounds were tested for inhibition of aromatase,
and to assess their selectivity toward related enzymes, their
ability to inhibit another cytochrome P450 enzyme was also
‡ Saarland University.
10.1021/jm060186y CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/30/2006

4778 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15
Brief Articles
Table 1. Structures and Biological Activities of the Target Compounds
against CYP19 and CYP17
CYP17b
CYP19a % inhibition CYP19 CYP19
R
R′
IC50, µMc at 2.5 µM pIC50obsd pIC50pred
∆
1
2a
2b
2c
3a
3b
3c
3d
OCH3
NO2 OCH3
0.55
0.47
4.1
44
9
29
12
6.26
6.33
5.39
5.74
7.15
7.35
6.36
7.16
7.10
6.11
6.78
6.44
6.56
6.22
7.20
6.62
6.84
6.90
0.15
-0.45
-1.05
-0.82
0.93
0.15
-0.26
0.32
Br
CN
H
NO2
Br
CN
OCH3
OCH3
OCH3
H
H
H
1.8
0.071
0.045
0.44
0.069
0.080
0.052
27
26
H
H
3e
0.20
fadrozole
a Substrate is 500 nM 1â-[3H]androstenedione. b Substrate is 25 µM
progesterone. c The given values are mean values of at least three experi-
ments. The deviations were within (5%.
Scheme 1a
Figure 1. (A) Superimposition of 1 onto S-fadrozole following the
previously reported alignment strategy. As a consequence, (B) the
chromone nucleus is fairly well aligned onto the p-cyanophenyl of
R-fadrozole. (C) The CoMFA contour maps are plotted in the 3D space
surrounding the inhibitor 1. The CoMFA contour map levels are those
reported in Figure 1.
evaluated, namely, 17R-hydroxylase/C17,20-lyase (CYP17),
catalyzing the key step of androgen biosynthesis.
a Reagents and conditions: (i) K2CO3, (CH3)2SO4, reflux 7 h; (ii) (1)
180-190 °C, 7 h; (2) H2SO4, reflux 15 min; (iii) NBS, CCl4, reflux 5 h;
(iv) imidazole, CH3CN, reflux 6 h.
Results and Discussion
be a fundamental physicochemical requirement for good CYP19
inhibition.11 Since our CoMFA model does not include log P
or other descriptors to account for hydrophobic features, 2a,
which carried a strong H-bond acceptor group, was predicted
to be more potent than it actually was (pIC50pred ) 6.78 vs
pIC50obsd ) 6.33). In agreement with this observation was also
the activity of 2c. In previously reported series, the CN group
was shown to be a worse H-bond acceptor compared to the NO2
substituent, and this was confirmed here by the pIC50pred of 2c,
which was lower than the pIC50pred of 2a (6.56 and 6.78,
respectively). Still, 2c was observed to be significantly less
potent with an experimental pIC50obsd of 5.74. A slightly different
scenario could be observed for the highly hydrophobic Br
All the new compounds showed inhibition of aromatase, and
some of them proved to be active at the nanomolar level (Table
1). In particular, 3b was slightly more active than the marketed
drug fadrozole. Moreover, all the new compounds proved to
be highly selective for aromatase with respect to CYP17 because
no IC50 for the inhibition of this enzyme could be calculated.
Considering the results collected in Table 1, the introduction
of a NO2 group on the phenyl ring in the 7-methoxy series (2a)
seemed to have no significant effect on CYP19 inhibition. This
could be due to an overall increase of the hydrophilic character,
which counterbalanced the positive effect of a possible H-bond
interaction between NO2 and Ser478 of CYP19. Actually, high
hydrophobicity of CYP19 inhibitors has been demonstrated to

Brief Articles
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15 4779
sulfate (0.7 g, 5.4 mmol) in acetone was refluxed for 7 h. The
mixture was hot-filtered and evaporated to dryness. Purification
by column chromatography (petroleum ether/ethyl acetate 4:1) gave
derivative 2b, which is not able to establish an H-bond with
the biological counterpart, thus likely missing an important
interaction with the protein target. This may be responsible for
both its low potency (pIC50obsd ) 5.39) and its CoMFA-based
overestimated activity (pIC50pred ) 6.44).
1
1.10 g (65%) of 5a, mp 202-204 °C. H NMR: δ 2.20 (s, 3H,
CH3), 3.90 (s, 3H, OCH3), 6.85 (d, J ) 1.8 Hz, 1H, arom), 7.00
(dd, J ) 1.8 and 8.4 Hz, 1H, arom), 7.85 (d, J ) 8.4 Hz, 2H,
arom), 8.15 (d, J ) 8.4 Hz, 1H, arom), 8.40 (d, J ) 8.4 Hz, 2H,
arom).
More promising results were obtained with the removal of
the methoxy group because 3a showed activity in the nanomolar
range, about an 8-fold increase with respect to the parent
compound 1. This seemed to confirm the design hypothesis
based on our CoMFA model, the pIC50pred of 3a (6.22) however
being much lower than its pIC50obsd (7.15). This indeed proved
that a decrease of steric hindrance in a sterically disallowed
region increased the potency of inhibitors. Likely, our CoMFA
model underestimated 3a potency because the compound lacks
the H-bond acceptor group, which is a fundamental feature for
the statistical model.14 This definitively arose by the analysis
of the predicted activity of the most potent compound (3b),
which underwent both structural modifications invoked by our
design hypothesis, namely, the removal of the bulky methoxy
substituent and the introduction of an excellent H-bond acceptor
(NO2) on the phenyl ring. The pIC50obsd of 3b was 7.35, in very
good agreement with the pIC50pred (7.20). Fairly good predictions
were also obtained for the other compounds (3c-e, Table 1),
the Br derivative being the least potent of the present series of
molecules. To evaluate the overall predictive capability of our
3-(Bromomethyl)-7-methoxy-4′-nitroflavone (6a). To a solution
of 5a (3 g, 9.8 mmol) in CCl4 (150 mL) were added N-
bromosuccinimide (1.85 g, 9.8 mmol) and a catalytic amount of
benzoyl peroxide, and the mixture was refluxed for 5 h. The mixture
was hot-filtered and evaporated to dryness to give 1.2 g of 6a that
1
was used without further purification. H NMR: δ 3.90 (s, 3H,
OCH3), 4.40 (s, CH2), 7.00-7.40 (m, 2H, arom), 7.50-8.45 (m,
5H, arom).
3-(Imidazolylmethyl)-7-methoxy-4′-nitroflavone (2a). A mix-
ture of 6a (1.2 g, 3 mmol) and imidazole (0.6 g, 9 mmol) in dry
acetonitrile (40 mL) was refluxed under N2 for 6 h and evaporated
to dryness. The residue was purified by flash cromathography (ethyl
1
acetate) to give 0.52 g (46%) of 2a, mp 195 °C (dec). H NMR:
δ 3.90 (s, 3H, OCH3), 5.00 (s, 2H, CH2), 6.85-7.10 (m, 4H, arom),
7.40 (s, 1H, arom), 7.65 (d, J ) 8.4 Hz, 2H, arom), 8.20 (d, J )
8.4 Hz, 1H, arom), 8.40 (d, J ) 8.4 Hz, 2H, arom). MS: m/z 378
(M + 1). Anal. (C20H15N3O5) C, H, N.
Biology. All the compounds were tested according to previously
reported methods.11
CoMFA model toward flavone-based molecules, we calculated
Acknowledgment. This work was supported by grants from
MIUR-COFIN2004 (Rome, Italy).
16
the r2
to be 0.700 for the eight new compounds. Indeed,
pred
this value may be considered a satisfactory predictive value for
a series of newly designed molecules.
Supporting Information Available: Experimental details for
intermediates 4b-c, 5b-c, 6b-c, 7, 8a-e, 9a-e and final
compounds 2b-c and 3a-e; computational chemistry methods;
elemental analysis results of target compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
In conclusion, the introduction of substituents in position 4′
on the phenyl ring of the flavone (2a-c) did not cause any
improvement in potency. On the contrary, the removal of the
methoxy group in position 7, together with the introduction of
the same substituents on the phenyl ring, led to a series of
compounds (3a-e) showing inhibitory activity in the nanomolar
range, comparable to the marketed drug fadrozole. Moreover,
the activities of the new compounds were predicted by the
CoMFA model, and we can conclude that the overall predictive
capability of the model toward flavone-based molecules, as
References
(1) Brodie, A. M.; Schwarzel, W. C.; Shaikh, A. A.; Brodie, H. J. The
effect of an aromatase inhibitor, 4-hydroxy-4-androstene-3,17-dione,
on estrogen-dependent processes in reproduction and breast cancer.
Endocrinology 1977, 100, 1684-1695.
(2) Chakraborty, J.; Hopkins, R.; Parke, D. V. Inhibition studies on the
aromatization of androst-4-ene-3,17-dione by human placental mi-
crosomal preparations. Biochem. J. 1972, 130, 19P-20P.
(3) Recanatini, M.; Cavalli, A.; Valenti, P. Nonsteroidal aromatase
inhibitors: recent advances. Med. Res. ReV. 2002, 22, 282-304.
(4) Mouridsen, H.; Gershanovich, M.; Sun, Y.; Perez-Carrion, R.; Boni,
C.; Monnier, A.; Apffelstaedt, J.; Smith, R.; Sleeboom, H. P.;
Jaenicke, F.; Pluzanska, A.; Dank, M.; Becquart, D.; Bapsy, P. P.;
Salminen, E.; Snyder, R.; Chaudri-Ross, H.; Lang, R.; Wyld, P.;
Bhatnagar, A. Phase III study of letrozole versus tamoxifen as first-
line therapy of advanced breast cancer in postmenopausal women:
analysis of survival and update of efficacy from the International
Letrozole Breast Cancer Group. J. Clin. Oncol. 2003, 21, 2101-
2109.
(5) Kim, Y. W.; Hackett, J. C.; Brueggemeier, R. W. Synthesis and
aromatase inhibitory activity of novel pyridine-containing isoflavones.
J. Med. Chem. 2004, 47, 4032-4040.
(6) Leonetti, F.; Favia, A.; Rao, A.; Aliano, R.; Paluszcak, A.; Hartmann,
R. W.; Carotti, A. Design, synthesis, and 3D QSAR of novel potent
and selective aromatase inhibitors. J. Med. Chem. 2004, 47, 6792-
6803.
(7) Su, B.; Diaz-Cruz, E. S.; Landini, S.; Brueggemeier, R. W. Novel
sulfonanilide analogues suppress aromatase expression and activity
in breast cancer cells independent of COX-2 inhibition. J. Med. Chem.
2006, 49, 1413-1419.
(8) Saberi, M. R.; Vinh, T. K.; Yee, S. W.; Griffiths, B. J.; Evans, P. J.;
Simons, C. Potent CYP19 (aromatase) 1-[(benzofuran-2-yl)(phenyl-
methyl)pyridine, -imidazole, and -triazole inhibitors: synthesis and
biological evaluation. J. Med. Chem. 2006, 49, 1016-1022.
(9) Recanatini, M. Comparative molecular field analysis of non-steroidal
aromatase inhibitors related to fadrozole. J. Comput.-Aided Mol. Des.
1996, 10, 74-82.
(10) Recanatini, M.; Cavalli, A. Comparative molecular field analysis of
non-steroidal aromatase inhibitors: an extended model for two
different structural classes. Bioorg. Med. Chem. 1998, 6, 377-388.
quantitatively indicated by the r2
value, is fairly good.
pred
Experimental Section
Chemistry. General Methods. All melting points were deter-
mined in open glass capillaries using a Bu¨chi apparatus and are
uncorrected. 1H NMR spectra were recorded in CDCl3 (unless
otherwise indicated) on a Varian Gemini 300 spectrometer with
Me4Si as the internal standard. Mass spectra were recorded on a
Waters ZQ 4000 spectrometer operating in the electrospray (ES)
mode. Silica gel (Merck, 230-400 mesh) was used for purification
with flash chromatography. Elemental analyses were within 0.4%
of theoretical values. Compounds were named following IUPAC
rules as applied by AUTONOM, PC software for nomenclature in
organic chemistry from Beilstein-Institut and Springer.
7-Hydroxy-3-methyl-4′-nitroflavone (4a). A mixture of 2,4-
dihydroxypropiophenone (1.8 g, 11 mmol), 4-nitrobenzoyl chloride
(4.1 g, 22 mmol), and sodium 4-nitrobenzoate (6.3 g, 34.3 mmol)
was heated to 180-190 °C for 7 h. Water was added, and the solid
was filtered and refluxed for 15 min in 50% H2SO4, then poured
into ice. The precipitate was filtered and resuspended in 2 N NaOH,
filtered, and purified by column chromatography (petroleum ether/
1
ethyl acetate 4:1) to give 2.3 g (70%) of 4a, mp 278-280 °C. H
NMR (DMSO-d6): δ 2.00 (s, 3H, CH3), 6.80 (d, J ) 1.8 Hz, 1H,
arom), 6.90 (dd, J ) 1.8 and 8.4 Hz, 1H, arom), 8.00 (d, J ) 8.4
Hz, 1H, arom), 8.05 (d, J ) 8.4 Hz, 2H, arom), 8.40 (d, J ) 8.4
Hz, 2H, arom), 10.80 (s, 1H, OH).
7-Methoxy-3-methyl-4′-nitroflavone (5a). A mixture of 4a
(1.6 g, 5.4 mmol), dry K2CO3 (0.75 g, 5.4 mmol), and dimethyl

4780 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15
Brief Articles
(11) Recanatini, M.; Bisi, A.; Cavalli, A.; Belluti, F.; Gobbi, S.; Rampa,
A.; Valenti, P.; Palzer, M.; Palusczak, A.; Hartmann, R. W. A new
class of nonsteroidal aromatase inhibitors: design and synthesis of
chromone and xanthone derivatives and inhibition of the P450
enzymes aromatase and 17R-hydroxylase/C17,20-lyase. J. Med.
Chem. 2001, 44, 672-680.
(12) Sanderson, J. T.; Hordijk, J.; Denison, M. S.; Springsteel, M. F.;
Nantz, M. H.; van den Berg, M. Induction and inhibition of aromatase
(CYP19) activity by natural and synthetic flavonoid compounds in
H295R human adrenocortical carcinoma cells. Toxicol. Sci. 2004,
82, 70-79.
(13) Vinh, T. K.; Nicholls, P. J.; Kirby, A. J.; Simons, C. Evaluation of
7-hydroxy-flavones as inhibitors of oestrone and oestradiol biosyn-
thesis. J. Enzyme Inhib. 2001, 16, 417-424.
(14) Cavalli, A.; Bisi, A.; Bertucci, C.; Rosini, C.; Paluszcak, A.; Gobbi,
S.; Giorgio, E.; Rampa, A.; Belluti, F.; Piazzi, L.; Valenti, P.;
Hartmann, R. W.; Recanatini, M. Enantioselective nonsteroidal
aromatase inhibitors identified through a multidisciplinary medicinal
chemistry approach. J. Med. Chem. 2005, 48, 7282-7289.
(15) Favia, A. D.; Cavalli, A.; Masetti, M.; Carotti, A.; Recanatini, M.
Three-dimensional model of the human aromatase enzyme and
density functional parameterization of the iron-containing protopor-
phyrin IX for molecular dynamics study of heme-cysteinato cyto-
chromes. Proteins, in press.
(16) r2
was calculated with the following equation: r2
) (SD -
pred
pred
PRESS)/SD, where SD is the sum of the squared deviations of each
observed activity value for each molecule of the test set (pIC50obsd
values of Table 1) from the mean of the observed activity values of
the training set (pIC50obsd reported in ref 12) and PRESS is the sum
of the squared deviations between predicted and observed values for
the molecules of the test set (pIC50pred and pIC50obsd of Table 1,
respectively).
JM060186Y