Synthesis and biological evaluation of 4-imidazolylflavans as nonsteroidal aromatase inhibitors

Article abstract of DOI:10.1016/j.bioorg.2004.06.008

A series of 4-imidazolylflavans having a variety of substituents on the 2-phenyl ring was synthesized and investigated for their inhibitory effect against aromatase. Structure-activity relationships of these compounds were determined. An additional chlorine atom or a cyano group at the 4′ position did not enhance aromatase inhibition as well as a 3′-hydroxyl group. The influence of an additional 4′-hydroxyl group depends on the substitution pattern of A ring. Among these molecules, 4′-hydroxy-4- imidazolyl-7-methoxyflavan is only 2.2-fold less active than the letrozole (as assessed by IC₅₀ values). Letrozole is used as the first-line therapy for metastatic breast cancer.

Full text of DOI:10.1016/j.bioorg.2004.06.008

Bioorganic Chemistry 32 (2004) 494–503
www.elsevier.com/locate/bioorg
Synthesis and biological evaluation
of 4-imidazolylflavans as nonsteroidal
aromatase inhibitors
Christelle Pouget, Samir Yahiaoui, Catherine Fagnere,
Gerard Habrioux, Albert Jose Chulia*
´ ´
UPRES EA 1085, Biomolecules et Cibles Cellulaires Tumorales, Faculte de Pharmacie,
2 rue du Docteur Marcland, 87025 LIMOGES cedex, France
Received 17 December 2003
Available online 6 August 2004
Abstract
A series of 4-imidazolylflavans having a variety of substituents on the 2-phenyl ring was
synthesized and investigated for their inhibitory effect against aromatase. Structure–activity
relationships of these compounds were determined. An additional chlorine atom or a cyano
group at the 40 position did not enhance aromatase inhibition as well as a 30-hydroxyl group.
The influence of an additional 40-hydroxyl group depends on the substitution pattern of A
ring. Among these molecules, 40-hydroxy-4-imidazolyl-7-methoxyflavan is only 2.2-fold less
active than the letrozole (as assessed by IC50 values). Letrozole is used as the first-line therapy
for metastatic breast cancer.
ꢀ 2004 Elsevier Inc. All rights reserved.
Keywords: 4-Imidazolylflavans; Breast cancer; Anti-aromatase effect; Structure–activity relationships
*
Corresponding author. Fax: +33-555-435-910.
E-mail address: jose.chulia@unilim.fr (A.J. Chulia).
0045-2068/$ - see front matter ꢀ 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.bioorg.2004.06.008

C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
495
1. Introduction
Estrogens are known to be important in the growth of breast cancer in both pre-
and post-menopausal women. Therefore, two effective strategies have been devised
to antagonize the action of these hormones and to treat endocrine-related breast can-
cers. One pharmacological approach is the inhibition of estrogen action by anti-es-
trogens (e.g., tamoxifen) that interact with estrogen receptors [1]. The second
approach is to block estrogen synthesis by inhibiting a key enzyme of the process.
Aromatase, which catalyses the final step in the biosynthesis of estrogens, has been
the target for the design of inhibitors which have become the established second-line
treatment for metastatic breast cancer in post-menopausal women [2].
Aromatase inhibitors can be both steroidal and nonsteroidal compounds. The
first nonsteroidal inhibitor was aminoglutethimide. This compound has now been re-
placed by other more selective and potent nonsteroidal inhibitors like fadrozole,
letrozole, or anastrozole. These compounds are currently being compared with
tamoxifen in first-line metastatic, adjuvant, and neoadjuvant settings. Thus, based
primarily on a superior side effect profile, letrozole has recently been approved as
a first-line therapy of metastatic breast cancer in several countries [3]. The structure
of these nonsteroidal aromatase inhibitors can be considered to consist of two parts.
One part is the azole part with a nitrogen atom which interacts with the heme iron
atom of the cytochrome P450 of aromatase. The second part is the bulky aryl part,
which mimics the steroid ring of the substrate.
Our previous studies focused mainly on the flavonoids, which are natural com-
pounds inhibiting aromatase with about the same activity as aminoglutethimide
[4,5]. Structure–activity relationship studies enabled us to determine the binding
characteristics and the structural requirements for flavonoids to inhibit aromatase.
First, these compounds bind to the active site of aromatase in an orientation in
which A and C rings mimic rings D and C of the androgen substrate, respectively
[6]. It also appears that the C-4 keto group of flavonoids is essential for aromatase
inhibition because it interacts with the heme of the enzyme [6]. With respect to sub-
stitutions in the A ring, a 7-hydroxyl group is essential for enhanced aromatase
inhibitory activity while a 7-methoxy group is also responsible for an anti-aromatase
effect [4]. Finally, we demonstrated that additional hydroxyl groups at position 30
and/or 40 on the 7-methoxyflavanone skeleton led to an increase in aromatase inhi-
bition [5].
In the search of new lead compounds that inhibit aromatase, we have recently
described the synthesis and biological evaluation of two new 4-imidazolylflavans,
the 2,4-trans-7-hydroxy-4-imidazolylflavan A1 and the 2,4-trans-4-imidazolyl-7-
methoxyflavan A2 (Scheme 1) [7]. Replacement of the carbonyl function of flav-
anones by an imidazolyl group led to a marked increase of inhibitory potency
as these two compounds were found to have IC50 values of 0.041 and
0.091lM, respectively, while the IC50 value of aminoglutethimide was 5.2lM.
These encouraging results prompted us to modify these lead compounds. In par-
ticular, we examined the effects of substitutions on the 2-phenyl group (B ring in
A1 and A2).

496
C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
Scheme 1. Structure of 4-imidazolylflavans A1 and A2.
In the present paper, we report the synthesis of the 4-imidazolylflavans 3a–3e and
their biological evaluation against aromatase. Our previous studies concerning the B
ring structural requirements for enhanced aromatase inhibitory effect of flavonoids
allowed us to identify some active flavanones as molecular scaffolds [5]. Thus, 30
and 40-hydroxy groups, which were found to increase aromatase inhibition, were
introduced onto the 2-phenyl of the 4-imidazolylflavan skeleton. A chlorine atom
and a cyano group were also introduced at the 40 position of the B ring since it is
known that in the azole family of aromatase inhibitors, the presence of such a
para-substituted phenyl group reinforces the interaction between the azole nitrogen
atom of these inhibitors and the heme iron atom of the enzyme [8,9]. The results show
that an additional chlorine atom or a cyano group at the 40 position did not enhance
aromatase inhibition as well as a 30-hydroxyl group. The influence of an additional 40-
hydroxyl group was found to depend on the substitution pattern of A ring.
2. Materials and methods
2.1. Chemistry
2.1.1. General experimental procedures
NMR spectra were recorded on a Bruker 400MHz spectrometer with Me4Si as
the internal standard. ESI-MS spectra were performed on a Waters Alliance system
equipped with an API–MS interface. Compounds were purified by preparative thin
layer chromatography on Macherey–Nagel silica gel. Preparative HPLC was per-
formed on a Merck prep Sep Tech system.
2.1.2. Synthesis and characterization of the 4-imidazolylflavans
2.1.2.1. General procedure. To a solution of flavanone in ethanol was added sodium
borohydride (5 eq). The mixture was kept under a nitrogen atmosphere at a temper-

C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
497
ature of 6ꢁC for 24–72h (except flavanone 1a for which the reaction occurred at
room temperature for 4h). The mixture was evaporated under reduced pressure
and water was added. The pH of the mixture was adjusted to 6 using 1M HCl
and the resulting solution was extracted with chloroform. The organic layers were
washed, dried over Na2SO4 and evaporated. The resulting flavan-4-ols were purified
via preparative TLC on silica gel using toluene/Et2O (75/25) as eluent.
To a solution of flavan-4-ol in anhydrous THF was added 1,10-carbonyldiimidaz-
ole (4 eq). The mixture was kept under a nitrogen atmosphere at room temperature
for 4h. Subsequently, it was evaporated under reduced pressure. The residue was dis-
solved into chloroform and the organic layer was washed, dried over Na2SO4 and
evaporated. Preparative TLC on silica gel using CHCl3/MeOH (97/3) as eluent affor-
ded pure 4-imidazolylflavans.
2,4-cis-30-hydroxy-7-methoxyflavan-4-ol (2a) was obtained in 70% yield. 1H NMR
(400MHz; CDCl3): d 2.11(1H, ddd, J = 10.0, 11.2, and 13.2Hz, H-3ax), 2.50 (1H,
ddd, J = 2.0, 6.4, and 13.2Hz, H-3eq), 3.78 (3H, s, OCH3), 5.04 (1H, br s, H-4),
5.13 (1H, dd, J = 2.0 and 11.2Hz, H-2), 6.45 (1H, d, J = 2.4Hz, H-8), 6.58 (1H,
dd, J = 2.4 and 8.8Hz, H-6), 6.80 (1H, br dd, J = 2.0 and 8.0Hz, H-40), 6.94 (1H,
t, J = 2.0Hz, H-20), 6.99 (1H, br d, J = 8.0Hz, H-60), 7.26 (1H, t, J = 8.0Hz, H-
50), 7.40 (1H, d, J = 8.8Hz, H-5); 13C NMR (100MHz; CDCl3): d 39.9 (C-3), 55.4
(OCH3), 65.5 (C-4), 76.8 (C-2), 101.3 (C-8), 108.2 (C-6), 112.9 (C-20), 115.4 (C-40),
118.0 (C-4a), 118.2 (C-60), 128.1 (C-5), 130.0 (C-50), 142.1 (C-10), 155.4 (C-30),
156.2 (C-8a), and 160.3 (C-7); ESP-MS (ꢀ40V): m/z 271 [MꢀH]ꢀ.
2,4-trans-30-hydroxy-4-imidazolyl-7-methoxyflavan (3a) was obtained in 40% yield.
1H NMR (400MHz; CDCl3): d 2.31(1H, br dt, J = 2.2 and 14.5Hz, H-3eq), 2.46 (1H,
ddd, J = 4.4, 11.2, and 14.5Hz, H-3ax), 3.79 (3H, s, OCH3), 4.75 (1H, dd, J = 1.4 and
11.2Hz, H-2), 5.33 (1H, br t, J = 3.2Hz, H-4), 6.55 (1H, br s, H-8), 6.56 (1H, dd,
J = 2.5 and 8.2Hz, H-6), 6.75 (1H, br s, H-20), 6.85 (1H, br d, J = 8.1Hz, H-40),
6.92 (1H, br d, J = 7.7Hz, H-60), 6.97 (1H, s, H-400), 7.01 (1H, d, J = 8.2Hz, H-5),
7.11 (1H, s, H-500), 7.26 (1H, t, J = 7.9Hz, H-50), 7.45 (1H, s, H-200); 13C NMR
(100MHz; CDCl3): d 38.1 (C-3), 51.5 (C-4), 55.4 (OCH3), 73.3 (C-2), 101.8 (C-8),
109.0 (C-4a), 109.3 (C-6), 118.6 (C-400), 113.6 (C-20), 116.2 (C-40), 116.4 (C-60),
128.6 (C-500), 130.2 (C-50), 131.2 (C-5), 136.4 (C-200), 140.6 (C-10), 156.6 (C-30),
158.0 (C-8a), and 161.6 (C-7); ESP-MS (+40V): m/z 323 [M + H]+.
1
2,4-cis-40-chloro-7-methoxyflavan-4-ol (2b) was obtained in 61% yield. H NMR
(400MHz; CDCl3): d 2.06 (1H, ddd, J = 10.3, 11.4, and 13.1Hz, H-3ax), 2.48 (1H,
ddd, J = 1.9, 6.2, and 13.2Hz, H-3eq), 3.77 (3H, s, OCH3), 5.04 (1H, dd, J = 6.3
and 10.1Hz, H-4), 5.13 (1H, dd, J = 1.9 and 11.5Hz, H-2), 6.43 (1H, d,
J = 2.5Hz, H-8), 6.58 (1H, dd, J = 2.5 and 8.6Hz, H-6), 7.38 (4H, m, H-20,H-
30,H-50, and H-60), 7.39 (1H, d, J = 8.6Hz, H-5); 13C NMR (100MHz; CDCl3): d
40.3 (C-3), 55.4 (OCH3), 65.4 (C-4), 76.4 (C-2), 101.2 (C-8), 108.3 (C-6), 118.0 (C-
4a), 127.5 (C-30 and C-50), 127.9 (C-5), 128.9 (C-20 and C-60), 134.0 (C-10), 139.0
(C-40), 155.3 (C-8a), and 160.5 (C-7); ESP-MS (+40V): m/z 273 and m/z 275
[M + H]+.
2,4-trans-40-chloro-4-imidazolyl-7-methoxyflavan (3b) was obtained in 54% yield.
1H NMR (400MHz; CDCl3): d 2.34 (1H, br dt, J = 2.8 and 14.4Hz, H-3eq), 2.43

498
C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
(1H, ddd, J = 4.4, 11.1, and 14.4Hz, H-3ax), 3.82 (3H, s, OCH3), 4.94 (1H, dd,
J = 2.2 and 11.1Hz, H-2), 5.34 (1H, br t, J = 3.5Hz, H-4), 6.56 (1H, d, J = 2.4Hz,
H-8), 6.59 (1H, dd, J = 2.4 and 8.5Hz, H-6), 6.93 (1H, br s, H-400), 7.04 (1H, d,
J = 8.5Hz, H-5), 7.12 (1H, br s, H-500), 7.29 (2H, d, J = 8.5Hz, H-30 and H-50),
7.36 (2H, d, J = 8.5Hz, H-20 and H-60), 7.43 (1H, s, H-200); 13C NMR (100MHz;
CDCl3): d 38.4 (C-3), 50.9 (C-4), 55.4 (OCH3), 72.5 (C-2), 101.7 (C-8), 109.4 (C-
6), 109.5 (C-4a), 118.3 (C-400), 127.5 (C-30 and C-50), 128.9 (C-20 and C-60), 129.8
(C-500), 131.2 (C-5), 134.2 (C-10), 136.8 (C-200), 138.2 (C-40), 156.2 (C-8a), and
161.5 (C-7); ESP-MS (+40V): m/z 341 [M + H]+.
1
2,4-cis-40-cyano-7-methoxyflavan-4-ol (2c) was obtained in 54% yield. H NMR
(400MHz; CDCl3): d 2.05 (1H, ddd, J = 10.2, 11.4, and 13.2Hz, H-3ax), 2.51 (1H,
ddd, J = 2.1, 6.2, and 13.2Hz, H-3eq), 3.79 (3H, s, OCH3), 5.07 (1H, m, H-4), 5.23
(1H, dd, J = 1.8 and 11.4Hz, H-2), 6.45 (1H, d, J = 2.5Hz, H-8), 6.61 (1H, dd,
J = 2.5 and 8.6Hz, H-6), 7.40 (1H, d, J = 8.5Hz, H-5), 7.57 (2H, d, J = 8.2Hz, H-
20 and H-60), 7.70 (2H, d, J = 8.3Hz, H-30 and H-50); 13C NMR (100MHz; CDCl3):
d 40.2 (C-3), 55.4 (OCH3), 65.2 (C-4), 76.1(C-2), 101.3 (C-8), 108.5 (C-6), 112.0 (C-
40), 117.9 (C-4a), 118.6 (CN), 126.6 (C-20 and C-60), 128.0 (C-5), 132.5 (C-30 and C-
50), 145.9 (C-10), 154.9 (C-8a), and 160.6 (C-7); ESP-MS (ꢀ40V): m/z 280 [MꢀH]ꢀ.
2,4-trans-40-cyano-4-imidazolyl-7-methoxyflavan (3c) was obtained in 42% yield.
1H NMR (400MHz; CDCl3): d 2.37 (1H, br dt, J = 3.0 and 14.2Hz, H-3eq), 2.41
(1H, ddd, J = 4.3, 9.8, and 14.2Hz, H-3ax), 3.83 (3H, s, OCH3), 5.01 (1H, dd,
J = 4.2 and 9.6Hz, H-2), 5.36 (1H, br t, J = 3.5Hz, H-4), 6.58 (1H, d, J = 2.5Hz,
H-8), 6.62 (1H, dd, J = 2.5 and 8.5Hz, H-6), 6.94 (1H, br s, H-400), 7.06 (1H, d,
J = 8.5Hz, H-5), 7.14 (1H, br s, H-500), 7.47 (1H, s, H-200), 7.48 (2H, d, J = 8.2Hz,
H-20 and H-60), 7.69 (2H, d, J = 8.2Hz, H-30 and H-50); 13C NMR (100MHz;
CDCl3): d 38.6 (C-3), 50.7 (C-4), 55.5 (OCH3), 72.3 (C-2), 101.7 (C-8), 109.3 (C-
4a), 109.6 (C-6), 112.3 (C-40), 118.3 (C-400), 118.5 (CN), 126.7 (C-20 and C-60),
129.9 (C-500), 131.3 (C-5), 132.6 (C-30 and C-50), 136.8 (C-200), 144.9 (C-10), 155.8
(C-8a), and 161.6 (C-7); ESP-MS (+40V): m/z 332 [M + H]+.
2,4-cis-40-hydroxy-7-methoxyflavan-4-ol (2d) was obtained in 14% yield. 1H NMR
(400MHz; CD3OD): d 2.02 (1H, ddd, J = 10.9, 12.0, and 12.9Hz, H-3ax), 2.33 (1H,
ddd, J = 1.6, 6.2, and 12.9Hz, H-3eq), 3.73 (3H, s, OCH3), 4.99 (1H, dd, J = 6.2 and
10.7Hz, H-4), 5.03 (1H, dd, J = 1.0 and 12.0Hz, H-2), 6.34 (1H, d, J = 2.5Hz, H-8),
6.52 (1H, dd, J = 2.5 and 8.6Hz, H-6), 6.80 (2H, d, J = 8.6Hz, H-30 and H-50), 7.27
(2H, d, J = 8.6Hz, H-20 and H-60), 7.36 (1H, d, J = 8.6Hz, H-5); 13C NMR
(100MHz; CD3OD): d 41.2 (C-3), 55.7 (OCH3), 66.3 (C-4), 78.6 (C-2), 102.0 (C-
8), 108.4 (C-6), 116.2 (C-30 and C-50), 120.0 (C-4a), 128.7 (C-20 and C-60), 129.1
(C-5), 133.3 (C-10), 157.2 (C-40), 158.5 (C-8a), and 161.6 (C-7); ESP-MS (ꢀ40V):
m/z 271 [MꢀH]ꢀ.
2,4-trans-40-hydroxy-7-methoxyflavan-4-ol (2d0) was obtained in 17% yield. 1H
NMR (400MHz; CD3OD): d 2.07 (1H, br dt, J = 3.3 and 14.2Hz, H-3eq), 2.11
(1H, ddd, J = 2.5, 9.6, and 14.2Hz, H-3ax), 3.75 (3H, s, OCH3), 4.71 (1H, br t,
J = 2.9Hz, H-4), 5.12 (1H, dd, J = 4.5 and 9.6Hz, H-2), 6.40 (1H, d, J = 2.5Hz,
H-8), 6.52 (1H, dd, J = 2.5 and 8.5Hz, H-6), 6.80 (2H, d, J = 8.6Hz, H-30 and H-
50), 7.22 (1H, d, J = 8.5Hz, H-5), 7.27 (2H, d, J = 8.6Hz, H-20 and H-60); 13C

C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
499
NMR (100MHz; CD3OD): d 40.0 (C-3), 55.7 (OCH3), 64.3 (C-4), 74.3 (C-2), 102.2
(C-8), 108.8 (C-6), 116.2 (C-30 and C-50), 117.5 (C-4a), 128.8 (C-20 and C-60), 132.5
(C-5), 133.5 (C-10), 157.5 (C-40), 158.4 (C-8a), 162.3 (C-7); ESP-MS (ꢀ40V): m/z 271
[MꢀH]ꢀ.
2,4-trans-40-hydroxy-4-imidazolyl-7-methoxyflavan (3d) was obtained in 35% yield.
1H NMR (400MHz; CD3OD): d 2.34 (1H, br dt, J = 2.5 and 14.6Hz, H-3eq),
2.49(1H, ddd, J = 4.5, 11.5, and 14.6Hz, H-3ax), 3.79 (3H, s, OCH3), 4.86 (1H,
dd, J = 2.0 and 11.5Hz, H-2), 5.49 (1H, br t, J = 3.5Hz, H-4), 6.54 (1H, d,
J = 2.5Hz, H-8), 6.58(1H, dd, J = 2.5 and 8.5Hz, H-6), 6.78 (2H, d, J = 8.6Hz, H-
30 and H-50), 7.03 (1H, br s, H-400), 7.05 (1H, d, J = 8.5Hz, H-5), 7.13 (1H, br s,
H-500), 7.18 (2H, d, J = 8.6Hz, H-20 and H-60), 7.59 (1H, s, H-200); 13C NMR
(100MHz; CD3OD): d 39.1 (C-3), 52.7 (C-4), 55.9 (OCH3), 74.4 (C-2), 102.7 (C-
8), 109.8 (C-6), 111.4 (C-4a), 116.3 (C-30 and C-50), 120.3 (C-400), 128.7 (C-20 and
C-60), 128.9 (C-500), 132.2 (C-10), 132.3 (C-5), 138.0 (C-200), 158.2 (C-40), 158.7 (C-
8a), and 163.0 (C-7); ESP-MS (+40V): m/z 323 [M + H]+.
2,4-cis-40,7-dihydroxyflavan-4-ol (2e) was obtained in 10% yield. 1H NMR
(400MHz; CD3OD): d 2.00 (1H, ddd, J = 10.8, 11.9, and 12.9Hz, H-3ax), 2.31
(1H, ddd, J = 1.6, 6.4, and 12.9Hz, H-3eq), 4.96 (1H, dd, J = 6.4 and 10.7Hz,
H-4), 5.00 (1H, dd, J = 1.4 and 11.9Hz, H-2), 6.21 (1H, d, J = 2.4Hz, H-8), 6.40
(1H, dd, J = 2.4 and 8.4Hz, H-6), 6.80 (2H, d, J = 8.5Hz, H-30 and H-50), 7.26
(2H, d, J = 8.5Hz, H-20 and H-60), 7.26 (1H, d, J = 8.4Hz, H-5); 13C NMR
(100MHz; CD3OD): d 41.3 (C-3), 66.3 (C-4), 78.5 (C-2), 103.6 (C-8), 109.5 (C-
6), 116.2 (C-30 and C-50), 119.0 (C-4a), 128.7 (C-20 and C-60), 129.1 (C-5), 133.4
(C-10), 157.2 (C-40), 158.5 (C-8a), and 158.9 (C-7); ESP-MS (ꢀ40V): m/z 257
[MꢀH]ꢀ.
2,4-trans-40,7-dihydroxyflavan-4-ol (2e0) was obtained in 30% yield. 1H NMR
(400MHz; CD3OD): d 2.04 (1H, dt, J = 3.4 and 14.3Hz, H-3eq), 2.08 (1H, ddd,
J = 2.6, 9.4, and 14.3Hz, H-3ax), 4.68 (1H, t, J = 2.8Hz, H-4), 5.10 (1H, dd,
J = 4.7 and 9.4Hz, H-2), 6.28 (1H, d, J = 2.4Hz, H-8), 6.40 (1H, dd, J = 2.4 and
8.4Hz, H-6), 6.80 (2H, d, J = 8.5Hz, H-30 and H-50), 7.14 (1H, d, J = 8.4Hz, H-
5), 7.26 (2H, d, J = 8.5Hz, H-20 and H-60); 13C NMR (100MHz; CD3OD): d 40.1
(C-3), 64.4 (C-4), 74.2 (C-2), 103.8 (C-8), 109.7 (C-6), 116.2 (C-30 and C-50), 116.5
(C-4a), 128.8 (C-20 and C-60), 132.6 (C-5), 133.6 (C-10), 157.4 (C-40), 158.4 (C-8a),
and 159.7 (C-7); ESP-MS (ꢀ40V): m/z 257 [MꢀH]ꢀ.
2,4-trans-40,7-dihydroxy-4-imidazolylflavan (3e) was obtained in 16% yield. 1H
NMR (400MHz; CD3OD): d 2.32 (1H, dt, J = 2.3 and 14.5Hz, H-3eq), 2.47 (1H,
ddd, J = 4.5, 11.6, and 14.5Hz, H-3ax), 4.82 (1H, dd, J = 1.5 and 11.6Hz, H-2),
5.46 (1H, br t, J = 3.3Hz, H-4), 6.40 (1H, d, J = 2.4Hz, H-8), 6.46 (1H, dd,
J = 2.4 and 8.4Hz, H-6), 6.77 (2H, d, J = 8.6Hz, H-30 and H-50), 6.97 (1H, d,
J = 8.4Hz, H-5), 7.03 (1H, br s, H-400), 7.09 (1H, br s, H-500), 7.17 (2H, d,
J = 8.6Hz, H-20 and H-60), 7.59 (1H, br s, H-200); 13C NMR (100MHz; CD3OD):
d 39.2 (C-3), 52.8 (C-4), 74.3 (C-2), 104.3 (C-8), 110.3 (C-4a), 110.6 (C-6), 116.3
(C-30/50), 120.2 (C-400), 128.7 (C-20/60), 129.2 (C-500), 132.3 (C-10), 132.4 (C-5),
138.1 (C-200), 158.1 (C-40), 158.7 (C-8a), and 160.7 (C-7); ESP-MS (+40V): m/z
309 [M + H]+.

500
C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
2.2. Biological tests
The inhibitory activities of the compounds towards aromatase were determined in
vitro using human placental microsomes and [1,2,6,7-3H]-androstenedione as previ-
ously described [4]. The IC50 values were determined graphically by testing com-
pounds in six appropriate concentrations (0.02, 0.05, 0.07, 0.10, 0.50, and 1.00lM)
with each experiment performed in duplicate. The deviations were within 5%.
3. Results and discussion
The route for the synthesis of the target compounds is outlined in Scheme 2.
Flavanones 1a–1d were prepared by cyclization of corresponding 20-hydroxychal-
cones previously obtained by Claisen–Schmidt condensation between 2-hydroxy-
4-methoxyacetophenone and appropriately substituted benzaldehydes [5]. For
flavanone 1d, condensation was carried out with 4-hydroxybenzaldehyde protected
as tetrahydropyranyl ether. Flavanone 1e was synthesized by cyclization of the
20,4,40-trihydroxychalcone, obtained by the condensation of 4-hydroxybenzalde-
hyde and 2,4-dihydroxyacetophenone. Both compounds were protected as their
tetrahydropyranyl ethers.
Reduction of the 40-substituted flavanones was carried out at a temperature of
6ꢁC because at room temperature, the alkaline medium resulting from the presence
of NaBH4 was responsible for the isomerization into the corresponding 20-hydroxy-
chalcones. Reduction by NaBH4 of the flavanones 1a–1c gave the 2,4-cis-flavan-4-ols
Scheme 2. Synthesis of 4-imidazolylflavans 3a–3e.

C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
501
(2a–2c). The NaBH4 reduction of the 40-hydroxy-7-methoxyflavanone 1d afforded
the cis and trans isomers (2d and 2d0, respectively) of 40-hydroxy-7-methoxyflavan-
4-ol, in a 2:1 ratio, which were separated by preparative HPLC using a diol column
Merck LichroCART, Lichrospher 10l (250 · 10mm) with a mixture of hexane
containing 5% isopropanol/methanol (97/3) as eluent. Reduction of the 40,7-dihydr-
oxyflavanone 1e also gave the cis and trans isomers (2e and 2e0, respectively) of 40,7-
dihydroxyflavan-4-ol, in a 1:3 ratio, which could be separated by HPLC on a reverse
phase C18 column Waters Nova-Pak 4l (100 · 8mm) with H2O/methanol (85/15).
However, preparative chromatography was not performed and the condensation
with 1,10-carbonyldiimidazole was carried out on the mixture of the two isomers.
Treatment of the compounds 2a–2c with 1,10-carbonyldiimidazole in THF [10] led
to the 2,4-trans-4-imidazolylflavans 3a–3c while both the two isomers of 40-hydroxy-
7-methoxyflavan-4-ol, after condensation with 1,10-carbonyldiimidazole, gave the
2,4-trans-4-imidazolylflavan 3d. In the same way, the mixture of the two isomers
of the 40,7-dihydroxyflavan-4-ol led to the only 2,4-trans-4-imidazolylflavan 3e.
1
The 2,4-trans configuration for these compounds was determined from the H
NMR vicinal coupling constants as previously described [7]. These constants are
consistent with either the half chair (a) or sofa (b) conformation of the heterocyclic
ring in which the H-2 is axial (Scheme 3). The coupling constant for the H-3ax and H-
2 is approximately 11Hz, which corresponds to a trans-diaxial relationship between
the protons. The coupling constant for the H-4 and H-3ax in compounds 3a–3e is
4.4Hz, which corresponds to a quasi-equatorial position for the proton H-4 and al-
lowed us to determine a 2,4-trans configuration.
It was observed that steric hindrance due to the 2-phenyl group influenced the
type of substitution and was responsible for the 2,4-trans-stereochemistry of the 4-
imidazolylflavans. From the 2,4-cis-flavan-4-ols, the inversion of configuration at
the carbon C-4 was consistent with the occurrence of a SN2 pathway while for the
2,4-trans isomers 2d0 and 2e0, the reaction likely involved the formation of a carbo-
cation followed by a nucleophilic attack by the azole on the only opposite side from
the 2-phenyl group.
The IC50 values and inhibitory potencies of the compounds 3a–3e relative to ami-
noglutethimide are reported in Table 1. The derivatives show high inhibitory activity
against aromatase with IC50 values from 0.04 to 0.20lM and are more potent than
Scheme 3. Conformation of the heterocyclic ring of 2,4-trans-4-imidazolylflavans.

502
C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
Table 1
Aromatase inhibitory activity of compounds 3a–3e
Compound
3a
3b
3c
3d
3e
Letrozole
IC50 (lM)
0.130
40
0.200
26
0.133
39
0.040
130
0.077
68
0.018
289
RPa/aminoglutethimide
a
The relative potency (RP) was calculated from the ratio of IC50 values for compounds 3a–3e to that
measured for aminoglutethimide.
aminoglutethimide (IC50 = 5.2lM), exhibiting relative potencies (RP) from 26 to
130.
The influence of the B ring substitution pattern on the aromatase inhibitory effect
was investigated by comparing activities of compounds 3a–3d with that of the 2,4-
trans-4-imidazolyl-7-methoxyflavan A2 (IC50 = 0.091lM). It appears that a cyano
group or a chlorine atom at position 40 did not enhance the aromatase inhibition
as previously described for the 7-methoxyflavanone skeleton [5]. These results sug-
gest that the 4-imidazolylflavans and known azole compounds bind to the active site
of aromatase in a different way. For the azole compounds, para-substitution of a
phenyl group by halogen atoms or a cyano group led to a significant enhancement
in inhibitory activity [8,9]. The presence of a 30-hydroxy group did not increase
the anti-aromatase effect of these imidazolylflavans whereas this group was found
to be essential for enhanced aromatase inhibitory activity of flavanones [5]. In con-
trast, the 40-hydroxy substitution was shown to be responsible for an increase in aro-
matase inhibition. Thus, the compound 3d is only 2.2-fold less active than letrozole
(IC50 = 0.018lM), which is used as the first-line therapy for metastatic breast cancer.
The 7-hydroxy-4-imidazolylflavan A1 (IC50 = 0.041lM) was more potent than the
7-methoxy counterpart A2. For this reason, we synthesized the 40,7-dihydroxy-4-imi-
dazolylflavan 3e. Unfortunately, this compound was found to be less active than
expected (IC50 = 0.077lM) even though inhibitory potency relative to aminogluteth-
imide was equal to 68. In contrast to what was observed for the 7-methoxyflavan
skeleton, the presence of an additional 40-hydroxyl group led to a decrease in aroma-
tase inhibition. This result is in agreement with those previously established for the
7-hydroxyflavanone scaffold since we demonstrated that 40,7-dihydroxyflavanone
was less potent than 7-hydroxyflavanone (unpublished results). These findings may
suggest a different mode of binding to the active site of aromatase between 4-imid-
azolyl-7-methoxyflavans and the 7-hydroxy analogues.
4. Conclusion
The synthesis of five new 4-imidazolylflavans, which were found to be highly po-
tent as aromatase inhibitors, has been described. This work allowed us to obtain a
preliminary biological profile of the series indicating which compounds might be
suitable for further investigation and in which direction optimization efforts should
be pursued. The selectivity towards other P450 enzymes and the in vivo activity must
be investigated to answer the question whether these compounds can be considered

C. Pouget et al. / Bioorganic Chemistry 32 (2004) 494–503
503
as novel drug candidates for the treatment of hormone-dependent breast cancer. The
next step in this work is to separate the most active racemic compounds by chiral
HPLC, so as to enhance aromatase inhibitory activity of these 4-imidazolylflavans
by testing enantiomerically pure compounds.
Acknowledgments
The authors are grateful to the Region Limousin for the grants awarded to C.
Pouget and S. Yahiaoui and to Y. Champavier for acquiring the NMR spectra.
References
[1] M. Clemons, S. Danson, A. Howell, Cancer Treat. Rev. 28 (2002) 165–180.
[2] A.M.H. Brodie, V.C.O. Njar, Steroids 65 (2000) 171–179.
[3] P.E. Goss, K. Strasser, J. Clin. Oncol. 19 (2001) 881–894.
[4] J.C. Le Bail, T. Laroche, F. Marre-Fournier, G. Habrioux, Cancer Lett. 133 (1998) 101–106.
[5] C. Pouget, C. Fagnere, J.P. Basly, A.E. Besson, Y. Champavier, G. Habrioux, A.J. Chulia, Pharm.
Res. 19 (2002) 286–291.
[6] Y.C. Kao, C. Zhou, M. Sherman, C.A. Laughton, S. Chen, Environ. Health Persp. 106 (1998) 85–92.
[7] C. Pouget, C. Fagnere, J.P. Basly, G. Habrioux, A.J. Chulia, Bioorg. Med. Chem. Lett. 12 (2002)
2859–2861.
[8] M. Lang, C. Batzl, P. Furet, R. Bowman, A. Ha¨usler, S. Bhatnagar, J. Steroid Biochem. Mol. Biol.
44 (1993) 421–428.
[9] M. Okada, T. Yoden, E. Kawaminami, Y. Shimada, M. Kudoh, Y. Isomura, H. Shikama, T.
Fujikura, Chem. Pharm. Bull. 44 (1996) 1871–1879.
[10] M. Le Borgne, P. Marchand, B. Delevoye-Seiller, J.M. Robert, G. Le Baut, R.W. Hartmann, M.
Palzer, Bioorg. Med. Chem. Lett. 9 (1999) 333–336.