Synthesis and biological evaluation of phenolic 4,5-dihydroisoxazoles and 3-hydroxy ketones as estrogen receptor α and β agonists

Article abstract of DOI:10.1016/j.bmc.2010.04.007

In this work, 52 diphenyl-4,5-dihydroisoxazoles and -3-hydroxy ketones were prepared and their estrogen receptor α (ERα) and estrogen receptor β (ERβ) activities were explored in order to systematize and maximize their biological activity. The biological

Full text of DOI:10.1016/j.bmc.2010.04.007

Bioorganic & Medicinal Chemistry 18 (2010) 3437–3447
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and biological evaluation of phenolic 4,5-dihydroisoxazoles
and 3-hydroxy ketones as estrogen receptor
a and b agonists
Pekka K. Poutiainen a, , Tuomas A. Venäläinen b, , Mikael Peräkylä b, Juha M. Matilainen b, Sami Väisänen b,
Paavo Honkakoski c, Reino Laatikainen a, Juha T. Pulkkinen a,
*
a Laboratory of Chemistry, Department of Biosciences, University of Eastern Finland, PO Box 1627, FIN-70211 Kuopio, Finland
b Laboratory of Biochemistry, Department of Biosciences, University of Eastern Finland, PO Box 1627, FIN-70211 Kuopio, Finland
c Department of Pharmaceutics and Biocenter Kuopio, University of Eastern Finland, PO Box 1627, FIN-70211 Kuopio, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work, 52 diphenyl-4,5-dihydroisoxazoles and -3-hydroxy ketones were prepared and their estro-
Received 18 December 2009
Revised 29 March 2010
Accepted 1 April 2010
Available online 07 April 2010
gen receptor
a (ERa) and estrogen receptor b (ERb) activities were explored in order to systematize and
maximize their biological activity. The biological activity was firstly screened by using ERE reporter assay
to find out how aromatic hydroxylation and methylation of the chiral centers of the compounds affect the
ability of ER to mediate biological responses. For selected 19 compounds, the relative binding affinities
(RBA, relative to 3,17b-estradiol) and ability to induce transcription of primary E2 target gene pS2 in
human MCF-7 breast cancer cells were determined. In the reporter assay, many compounds showed even
stronger activity than E2 and some of them showed RBA larger than 1%. The highest RBAs were deter-
mined for the enantiomers of 1-hydroxy-6-(4-hydroxy-phenyl)-1-phenyl-hexan-3-one (50a and 50b).
Keywords:
Estrogen receptor
ER agonist
Estradiol
4,5-Dihydroisoxazoles
3-Hydroxyketones
Isomer 50a showed high binding affinity both to ER
the RBAs of 50b were ca. 40% of those. Some of the other compounds (with RBA ꢀ1–16%) showed also
notable ER binding selectivity. When four most promising ligands (50a, 50b, 45a, and 45b) were studied
a
(with RBA ꢀ200%) and ERb (with RBA ꢀ60%), while
a
with respect to their ability to induce the transcription of primary E2 target gene pS2, the compounds
acted as agonists or partial agonists. Computer modeling was used to predict receptor binding conforma-
tions and to rationalize the RBA differences of the compounds.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
used for the prevention and treatment of osteoporosis, is a rela-
tively ERb-selective SERM.22 However, not many of the ER agonists
Naturally occurring and synthetic estrogens have been widely
utilized in the treatment of broad variety of indications.1–5 Because
of the high therapeutic value, there is a continuous need for novel
compounds which display estrogenic behavior with selective ef-
fects on different estrogen-responsive tissues (i.e., selective estro-
gen receptor modulators = SERMs).6 There are two common
subtypes of estrogen receptors, ERa7 and ERb,8 both of which are
activated by their endogenous ligand 3,17b-estradiol (E2).
Although most estrogen target tissues comprise cells expressing
or antagonists currently in clinical use are subtype-specific. Due to
the adverse effects of ER-targeted drugs, wide variation in ER and
ERb expression in various tissues, and cell- and promoter-specific
effects of the ER subtypes, novel chemical scaffolds are being
sought in order to develop both subtype-specific and tissue-selec-
tive ligands.
The vast majority of the synthetic estrogenic compounds have a
general structure of ARYL1–CORE STRUCTURE–ARYL219 in which
the core may have a third aromatic ring and an additional substi-
tuent or side chain. With a few exceptions,20,21 the structures can
be considered as more or less rigid. Recently, we reported novel
estrogen agonists, which have a core structure consisting of a
spacer and flexible linkers between the spacer and the aryl
groups.23 The spacers have chiral sp3-centres and varying lengths
of the CH2-linker provide a distance of 4–8 carbons between the
aromatic rings. The in vitro studies showed that the best estrogen
agonists had a 4,5-dihydroisoxazole or 3-hydroxyketone moiety as
the spacer. As a follow-up of the study, we report here the estrogen
agonism of several novel phenolic diaryl 4,5-dihydroisoxazoles and
3-hydroxyketones and show how the aromatic hydroxylation and
a
both ER isoforms, ERa is predominantly found in reproductive tis-
sues, kidney, liver as well as central nervous system,9,10 and it is
the predominant subtype expressed in breast cancer.11 On the
other hand, ERb is present in many tissues including colon, pros-
tate, testis, ovary, vascular endothelium, and bone.9,11 There are
plenty of examples of subtype-specific compounds2,4,9–21 and re-
cently it has been shown that, for example, raloxifene which is
* Corresponding author. Tel.: +358 40 3553247; fax: +358 17 163259.
E-mail address: juha.t.pulkkinen@uef.fi (J.T. Pulkkinen).
Authors with equal contributions.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.04.007

3438
P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
methylation of the spacers’ chiral centres affect the activity. The fit
of the agonists into the ER and ERb binding pockets is explored
whose affinities are comparable to those of the endogenous ERa/
ERb ligand E2.
a
and rationalized by molecular modeling.
The most promising ligands 45a, 45b, 50a, and 50b that had
high or medium high affinities to ER and ERb were further studied
a
with respect to their ability to induce transcription of primary E2
target gene pS2 in human MCF-7 breast cancer cells. Our data sug-
gest that with close to physiological hormone concentration of
10 nM the pS2 mRNA expression was induced by 2.2-fold in re-
sponse to 24 h E2 treatment (Fig. 1A). Enantiomers 45b and 50a
did not show statistically significant ligand effects with 10 nM
(Fig. 1C and D) or with 1000-fold higher ligand concentrations
(data not shown), whereas 50b showed identical pS2 mRNA
expression profile to that of E2 (Fig. 1E). Interestingly, the com-
pound 45a showed a significant 1.8-fold induction already after
6 h ligand treatment but after 24 h treatment only basal expression
could be observed. The above four compounds were also tested by
competing 10 nM E2 with 100-fold higher concentrations of the
different compounds prior to pS2 mRNA expression assays (Fig
1F). According to our results, competition with ligands 45a, 45b,
50a or 50b decreased the E2 induced pS2 activity by 32%, 27%,
23% or 29%, respectively. Thus, the pS2 activity after competition
was close to that observed in respective expression assays, sug-
gesting that the synthetic compounds replaced the natural E2.
2. Results
2.1. Syntheses
All designed compounds were synthesized starting from appro-
priate aromatic aldoximes A and alkenes B, as shown in Scheme 1,
using similar methods and experimental conditions described pre-
viously in detail by us.23,24 The dipolar cycloaddition reaction leads
to 4,5-dihydroisoxazoles C, the aromatic methoxy groups of which
can be demethylated by BBr3 to produce the phenol function of the
compounds D. Finally, the phenolic isoxazoles D were converted to
corresponding hydroxy ketones E by catalytic hydrogenation and
hydrolysation with Raney Ni in the presence of H2O and AcOH.
After preparation, all compounds were run through semi-prepara-
tive HPLC system23 and, with one exception, we were able to sep-
arate all the enantiomers of compounds D (Tables 1 and 2). Neither
specific rotation nor absolute configuration was resolved due to
the small amounts of the purified enantiomers. For each separa-
tion, the enantiomer with a shorter retention time in the chiral
HPLC column is marked in the tables as enantiomer a and the other
form as b.
2.3. Modeling
To characterize binding conformations and to rationalize the
ERE reporter activities and RBAs, compounds 12–51 (total of 24
2.2. Biological evaluation
compounds) of Table 3 were docked to the ER
and MD simulations were done for the most potent compounds
50a/50b to relax their ER and ERb complexes. Based on modeling,
a and ERb receptors,
The estrogen activities (Tables 1 and 2) were evaluated in
HEK293 cells that lack endogenous ER expression25 by using tran-
a
sient transfection of ER
ven luciferase reporter as described.23,26 The reporter activities of
the test compounds (10 M) were normalized to the activation ob-
a or ERb expression vectors and an ERE-dri-
the Aryl 1 (Aryl 2 in some of the molecules) phenol group mimics,
as expected, the A ring of the natural agonist E2 by being involved
in a hydrogen bond network with Arg394/Arg346 (ERa/ERb), Glu353/
Glu305 and a structural water molecule (Fig. 2). It also forms an
edge-to-face interaction with Phe404/Phe356. Ligand docking also
indicated that the increase in affinity could be largely explained
by formation of hydrogen bonds between the compounds and
the binding site residues. Molecules 21ab, 45a, 45b, 50a, and
l
served for 10 nM E2 (= 100), and the specificity activation was
ascertained by transfections lacking ERs and inclusion of the ER
antagonist ICI-182,780.23 The relative ER binding affinities (RBA,
Table 3) were determined by
a competitive assay against
[6,7-3H(N)]estradiol in COS-1 cells using a wide range of concen-
trations to determine each chemical’s EC50 and relative binding
affinity (RBA) values as compared to E2 (= 100%).27 Overall, the
phenolic hydroxyl group as R1 substituent (Aryl 1, see Tables 1
and 2 for the structures) is a structural feature important for the
affinity of the ligands and in comparison to unsubstituted analogs,
compounds with a phenolic Aryl 1 have both higher receptor activ-
50b, which are good ER
a and ERb activators and had the largest
RBAs of the molecules in this work, were the only ones predicted
to form three hydrogen bonds of favorable geometry with the
receptors. The other good receptor binders form similar van der
Waals contacts with the receptor, but lack the hydrogen bond to
His540/His475.
ities and RBAs. Although most phenolic compounds are ERa selec-
The MM-PBSA calculations proposed also the absolute configu-
rations of 50a and 50b, which are not available within reasonable
efforts because the compounds form oily products and their X-ray
analysis would demand isolation of the compounds in larger
amounts than done so far and then also derivatization. The calcu-
lations for (R)-50 resulted more negative binding free energy than
tive, in several cases (compounds 3, 10, 21, 24, 30, 42, 45, and 51)
high ERb activities emerge by adding the hydroxyl especially to the
para-position of Aryl 1. On the other hand, para-substitution of Aryl
2 (5b, 12, and 25) also enhances the activities to some extent.
Based on the RBAs phenolic 4,5-dihydroxyisoxazole derivative 45
as well as 3-hydroxyketones 21 and 50 bind strongly to ER
a
the (S)-50 isomer to both ER
a
(À34.2 kcal/mol vs À32.9 kcal/mol)
(RBA = 11–218%). Ligands 21ab, 45a, and 45b show 40–150-fold
and ERb (À30.8 kcal/mol vs À29.1 kcal/mol). In line with this, the
ERa
selectivity over ERb. The most potent compound is 50a/50b,
R
MeO
HO
OMe
OH
i
MeO
+
MeO
N
O
R
NOH
m
n
n
m
A
B
C
HO
OH
O
OH
R
iii
ii
N
O
R
C
m
n
m
n
E
D
Scheme 1. Synthesis of the hydroxyl substituted diphenyl 4,5-dihydroisoxazoles and 3-hydroxy ketones; m = 2 or 3, n = 0–2, R = H or Me. Reagents and conditions: (i) NaOCl,
pyridine, CH2Cl2, 0 °C, 2 h; (ii) BBr3, CH2Cl2, rt, overnight; (iii) H2, Raney Ni, H2O, AcOH, MeOH–THF, rt, overnight.

P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
3439
Table 1
In vitro estrogen receptor activation of ligands 1–34
R2
O
OH
R2
R1
N
O
Substituents
R1
R1
R2
No.
ERaa
ERbb
No.
ER
a
ERb
No.
ER
a
ERb
Aryl 1
Aryl 2
H
H
H
H
4-F
4-OH
4-F
4-F
1ac
2a
3a
4a
5a
6a
7a
68.4
111.3
171.1
99.2
34.2
82.0
1.1
1bd
2b
3b
4b
5b
6b
7b
38.2
39.7
24.3
28.9
105.2
11.1
5.7
1.3
8abe
9ab
10ab
11ab
12ab
13ab
14ab
12.8
45.2
55.0
30.1
51.4
21.6
28.6
0.4
2-OH
4-OH
H
H
2-OH
4-OH
18.4
143.1
11.2
9.1
3.4
28.4
30.5
41.5
7.4
42.4
À2.4
4.4
5.1
106.7
5.7
6.1
11.1
21.2
77.8
N O
O
OH
R
R
R
No.
ER
a
ERb
No.
ER
a
ERb
No.
ER
a
ERb
Aryl 1
H
15a
16a
17a
18a
3.4
4.5
8.3
0.2
4.2
2.6
15b
16b
17b
18b
43.9
21.8
3.0
À0.4
2.6
3.6
19ab
20ab
21ab
22ab
16.2
4.0
87.3
46.8
1.0
1.1
48.1
10.1
2-OH
3-OH
4-OH
62.8
15.4
15.8
0.8
R1
R1
O
OH
R2
R2
N O
R1
R2
No.
ER
a
ERb
No.
ER
a
ERb
No.
ER
a
ERb
Aryl 1
Aryl 2
H
4-OH
H
H
H
4-OH
4-OH
23a
24a
25a
26a
31.3
153.7
67.1
13.6
78.2
4.7
À0.9
23b
24b
25b
26b
28.1
75.1
97.2
13.6
5.1
19.8
17.0
10.6
29ab
30ab
31ab
32a
32b
33ab
34ab
24.7
105.2
16.0
15.4
2.8
9.1
82.3
2.4
4-OH
14.4
1.8
À1.6
À0.8
À0.6
3,4-(OH)2
3,4-(OH)2
H
4-OH
27a
28a
0.3
1.7
1.3
À1.1
27b
28b
2.1
1.4
À1.3
À0.7
À0.5
0.5
a,b
Receptor activation relative to 10 nM E2 corresponding 100 (a mean of at least three independent transfections, SEM typically <15%), sample concentration 10
lM. The
activities of 1–8, 11, 15, 19, 23 and 29 have been reported previously.23
c
a = enantiomer with a shorter retention time in the chiral separation.
b = enantiomer with a longer retention time in the chiral separation.
Enantiomers not separated, racemic mixture tested.
d
e
Table 2
In vitro estrogen receptor activation of ligands 35–52
R2
N O
O
OH
R2
R1
Substituents
R1
R1
R2
No.
ERaa
ERbb
No.
ER
a
ERb
No.
ER
a
ERb
No.
ER
a
ERb
Aryl 1
H
H
2-OH
3-OH
4-OH
H
35ac
36a
37a
38a
39a
16.4
8.1
11.8
21.6
39.7
9.8
35bd
36b
37b
38b
39b
11.2
10.9
21.7
7.1
8.1
40a
78.8
52.1
20.5
38.4
40b
25.3
2.8
CH3
CH3
CH3
CH3
À20.0
À0.4
0.5
À7.3
30.0
À2.9
37.4
41abe
18.0
18.7
42ab
80.3
116.3
R1
O
OH
R1
R3
R3
O
N
R2
R2
R1
R2
R3
No.
ER
a
ERb
No.
ER
a
ERb
No.
ER
a
ERb
No.
ER
a
ERb
Aryl 1
Aryl 2
H
H
4-OH
4-OH
4-OH
H
H
H
H
H
CH3
H
CH3
H
43a
44ab
45a
46a
47a
44.3
69.5
103.6
14.4
À3.1
15.5
44.7
À0.9
12.0
43b
11.5
17.2
48a
49ab
50a
51ab
52a
91.4
83.3
139.2
90.9
147.3
15.1
170.2
103.1
107.8
48b
50b
52b
59.3
140.0
49.3
28.2
45b
46b
47b
65.0
13.6
27.6
77.1
10.6
7.2
155.3
43.5
3-OH
32.9
87.4
a–e
See Table 1. The activities of 35, 40, 43, and 48 have been reported previously.23

3440
P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
Table 3
structure) and 1.68 kcal/mol (ERa–(S)-50 starting structure) more
Relative binding affinities (RBA) of some of the estrogen active compounds in
comparison to E2 and tamoxifen for ER and ERb
negative binding free energy than the (S)-50 enantiomer. This sug-
gests that 50a having the higher experimental affinity has R
configuration.
a
No.
ER
a
RBA (%)
ERb RBA (%)
a/b ratio
Docking of compound 45 suggests that the main difference be-
tween the R and S enantiomers is in the rotation of the isoxazole
ring (Fig. 3). The 4-OH of the Aryl 1 mimics the hydroxyl group
of the natural ligand E2 and participates in the hydrogen bonding
network between the residues Arg394/Arg346 (ERa/ERb), Glu353/
Glu305 and a structural water molecule. The isoxazole ring of the
E2
100.00
1.51
0.91
0.09
0.03
10.60
0.05
1.90
0.10
0.04
100.00
2.18
0.16
0.18
0.16
0.10
0.20
0.18
N.D.a
0.04
0.02
0.10
0.16
0.06
0.01
0.18
0.02
0.10
57.65
18.67
0.04
1
0.7
6
0.5
0.2
106
0.3
11
Tamoxifen
3a
3b
10ab
21ab
22ab
30ab
43a
43b
44ab
45a
45b
46a
R isomer was found to be in close contact with Met336 of ERb. How-
ever, in the ERa the corresponding residue Leu384 is smaller and
1
leaves more space for the isoxazole ring of the R isomer. In this
group of compounds the a isomer showed higher affinities and
activities in ERa, whereas in ERb the b isomer displayed higher
affinity and activity than the a isomer (Tables 2 and 3). This differ-
ence suggests that the a isomer of this molecule series is the R
enantiomer, and the change of the favored form results from the
steric clash of the Met336 with the isoxazole ring reducing the affin-
ity of the R isomer to ERb.
0.03
1.5
157
41
13
1
0.7
5
2
15.68
6.53
0.78
0.01
0.12
0.09
0.24
217.80
92.02
0.32
46b
48a
48b
49ab
50a
4
5
8
50b
51ab
3. Discussion
a
Not detected.
Tables 1 and 2 show the ERE reporter activation indexes (in per-
cents). There are several compounds for which the index is >100,
which means that the compounds at 10 lM concentration are able
thermodynamic integration (TI) free energy calculations predicted
that the (R)-50 enantiomer has 1.64 kcal/mol (ER –(R)-50 starting
a
Figure 1. Expression profiling of the human pS2 gene. The MCF-7 human breast cancer cells were treated with 10 nM of E2 (A) or compounds 45a (B), 45b (C), 50a (D) or 50b
(E) for 6 and 24 h prior to RNA extraction. In competition assays (F), the cells were treated for 24 h with 10 nM E2 and 1 M of above compounds prior to RNA extraction. The
remaining pS2 activities are indicated as % of E2 activity above bars. The results are calculated as fold changes to solvent (EtOH, 0.1% final concentration), and they are
l
*
**
***
p <0.001). In each
normalized to RPLP0. Two-tailed Student’s t-tests were performed to calculate p-values in reference to EtOH treatment ( p = 0.01–0.05, p = 0.001–0.01,
panel, n is at least 3. Error bars indicate S.D.

P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
3441
Figure 2. Average structures of the MD simulations of (R)-50 and (S)-50 in ERa (A) and ERb (B). The positions of the residues shown are from the simulations with the R
isomer (light gray, the S isomer is shown in dark gray). According to docking, the ligand-binding modes differ between the ER subtypes: the phenolic (Aryl 1) and skeletal
hydroxyl groups are located at the same position and form hydrogen bonds in both of subtypes, but the unsubstituted Aryl 2 ring adopts two different orientations. The
favored binding mode results from the amino acid differences Leu384/Met336 and Met421/Ile373 (ER
a
/ERb): the unsubstituted phenyl ring binds so that it is close to Leu384 (ER
a
or Ile373 (ERb). In ERb, compound (R)-50 formed hydrogen bond with the backbone carbonyl oxygen of Gly472
.
ally one of the strongest ER
a
activators in the present paper but it
shows only modest binding affinity to ER
hydroxyketone 10ab, both ER and ERb activities are enhanced,
showing also a rare ERb selectivity ( /b of 0.5) in our collection
of compounds. As to the Aryl 2, para-substitution with fluorine
and OH has only minor or negative effects on ER activity of the
a. For the corresponding
a
a
a
a isomers. Para-OH group in 5b enhances both the activities, sug-
gesting that the binding mode is reverse. The trends with hydroxy-
ketones are more or less similar.
In comparison to 1–14, compounds 15–22 have one carbon
longer core structure (m = 2, n = 2), which seems to affect nega-
tively to the receptor activation. The unsubstituted isoxazoline
enantiomer 15b activates ER
substitution of Aryl 1 clearly decreases the activation while a clear
ER activation emerges by para-substitution (18a). Unexpectedly,
in the case of the hydroxyketones meta-substitution results in the
a to some extent. Ortho- or meta-OH
Figure 3. Binding of (R)-45 (light gray) and (S)-45 (dark gray). Met336 hinders
sterically especially the R isomer of compound 45 in ERb reducing its affinity. The
a
corresponding residue Leu384 of ER
a is smaller which, in comparison to ERb, leads
highest ER
22ab is ER
to ER
/b ratio of 106. On the basis of ligand docking, 21ab is capable of
forming three hydrogen bonds with ER (Fig. 4).
a
and ERb activities (21ab) and also para-substituted
to generally higher affinities of compounds 43–46 and favors the R form.
a
active. The former shows relatively high RBA of 10.6%
a
(Table 3) and its binding is strongly ERa selective with an
a
to activate the ERa and ERb receptors in HEK293 cells better than
a
E2 at 10 nM. On the other hand, the RBAs (Table 3) determined
with competitive assay against [6,7-3H(N)]estradiol in COS-1 cells
for a few selected structures show that there are compounds,
which own good or even higher binding affinity to the receptors
The core structure of ligands 23–34 (m = 3, n = 1) has the same
length as compounds 15–22, but the spacer is shifted by one car-
bon. The change in the spacer position mostly leads to higher
receptor activities and, in comparison to the unsubstituted ligands
than E2 (RBA = 100). At the 10 lM concentration used in the repor-
ter activity screening, the receptors can be assumed to be fully sat-
urated by these ligands. In other examples the large activation
indexes, in spite of their low RBA, also suggest that corresponding
binding affinity must be so high that the receptor is largely occu-
pied by them at 10 lM. In the following discussion we try to char-
acterize the key features of good binders and activators.
Ligands 1–14 have a core structure (m = 2, n = 1, Scheme 1) of
length of six carbons (Table 1). Several of these compounds show
high estrogen activity but their RBAs are low and, therefore, we
conclude that a high RBA may not be obtainable with these spacer
constructs. In this construct the compounds with 4,5-dihydroisox-
azole spacer are in all cases more active than the corresponding
hydroxyketones. The unsubstituted compounds 1 and 8 are both
clearly ER
selective one. Substitution of Aryl 1 with an ortho-hydroxyl group
(compound 2) enhances the ER activity of the isoxazole isomer
a selective, enantiomer 1a being the most active and
a
2a, and retains its receptor selectivity whereas for 2b also the
ERb activity is increased and the receptor selectivity lost. On the
other hand, para-OH substitution of Aryl 1, which is supposed to
mimic the E2 phenolic A-ring,23 leads to dramatic enhancement
of both receptor activities of the isoxazole isomer 3a, which is actu-
Figure 4. Binding of (R)-21 to ERa. In comparison to 45 and 50, the one carbon
longer linker leads to different binding orientation of the phenol ring. Consequently,
the 3-OH substitution of 21 leads to higher affinity than the 4-OH substitution of
compound 22.

3442
P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
23a/23b and 29ab, para-OH substitution of Aryl 1 strongly en-
hances ER and ERb activities of both isoxazoline 24a/24b and
hydroxyketone 30ab, which also shows notable ER binding affin-
ity (1.9%) and ER selectivity of 11/1. On the other hand, in the case
of isoxazoline 25a/25b para-substitution of Aryl 2 increases only
ER activation and for the corresponding hydroxyketone 31ab
high concentrations. Still, 45b had medium high affinity to ERa and
a
50a even twofold higher than that of E2 and both compounds were
able to compete with E2. The remaining pS2 activities after compe-
tition suggest that these two compounds act as partial agonists in
breast cancer cells. Isomer 45a appears to be especially interesting
a
a
a
with respect to its biological properties. It is highly ER
as it has a relatively good binding affinity to ER but poor to
ERb. The different affinities likely reflect to its ability to activate
ER receptor twofold better than ERb in reporter gene assays. Even
a selective,
the effect is negative for both receptors. Nevertheless, we conclude
that neither core structure is optimal for ER binding.
a
Ligands 35–42 own the shortest skeleton (m = 2, n = 0) of all the
compounds (Table 2). The low receptor activities of the isoxazoline
35a/35b are not affected by the addition of a methyl group to the
chiral center but the OH substitution of Aryl 1 enhances the recep-
tor activities to some extent. Especially both ortho- (37b) and para-
substitution (39b) increase ERb activation of enantiomer b. The
a
more interesting is the finding that 45a activates primary E2 target
gene pS2 very rapidly but temporally in MCF-7 cells. These data
suggest that 45a is a potent activator of E2 target genes, but also
that it is metabolized quite differently and significantly faster than
E2 in breast cancer cells. This could make 45a potentially safer
compound in hormone replacement therapy for women suffering
menopausal symptoms, since fast metabolism rate may decrease
the risk of estrogen dependent carcinogenesis.
hydroxyketone isomer 40a has the second highest ER
a but low
ERb activities among the unsubstituted compounds, and a substitu-
tion of the chiral carbon retains the former but increases the latter
to some extent (racemate 41ab). Para-OH substitution of Aryl 1 in
42ab, in comparison to 41ab, enhances especially the ERb activa-
tion. In general, it seems that methyl in the hydroxyketone spacer
of this set of ligands does not seriously disturb the activation. Be-
cause of the relatively low activities, RBAs were not determined.
Compounds 43–52 have similar structures with 35–42 but the
linker (m = 3, n = 0) between the spacer and Aryl 1 is longer by
one carbon. All the results indicate that this construct fits best to
the receptors and, interestingly, that the flexible hydroxyketones
do so better than the isoxazolines. This set of compounds contains
many good activators and even the unsubstituted hydroxyketone
4. Conclusions
Our analyses on the present molecular constructs shows that the
key features for good binding and receptor activation are, expect-
edly, para-OH at Aryl 1 and hydrogen bonding. It also seems that
the m = 3, n = 0 spacer construct fits best to the receptor, although
good activation is obtained with many other systems. The struc-
tures 45a/45b and 50a/50b happen just to own nearly perfect fit
to the receptors. The explanation to the superior RBAs of 50a and
50b is the hydrogen bond formed by the skeletal OH-group of the
48 (especially enantiomer 48a) is a strong activator of both ER
a
hydroxyketone. The remarkable selectivity (up to
a/b ratio of 157
and ERb. However, the RBAs are poor with two exceptions: the isox-
azoline 45a/45b and its hydroxyketone derivative 50a/50b with
para-OH at Aryl 1 show both very high activation and RBAs to
with 45a) to ER can probably be accounted for the larger size of
a
our ligand in comparison to E2, while the reason for the very high
binding affinities of the best ligands (50a and 50b) and obviously
also their high ERE agonism, is that the low rotational energies al-
low formation of favorable hydrogen bonds and numerous van
der Waals contacts with the receptor. The extensive contacts are
able to compensate the entropy loss due to rigidifying of the flexible
ligand upon receptor binding.
ER
ER
a
a
. The isoxazoline 45a shows also very high RBA selectivity to
(157:1 over ERb) while the ratio for hydroxyketones is smaller.
Methylation of the chiral center enhances both ER
a and ERb
activities of the isoxazoline ligand 43a/43b, but surprisingly
para-OH substitution of Aryl 1 of the methylated compound 44ab
deactivates the compound almost completely (46a/46b). When
also the very low RBAs (Table 3) of the methylated analog (51ab)
of the best binders 50a and 50b are considered, it seems improba-
ble that the replacement of hydrogen of the chiral center with a
larger group would improve the biological activity.
Both the MM-PBSA and TI calculations predict that 50a with the
higher affinity has R configuration. It must be noted that the affin-
ity difference between 50a and 50b is rather small and therefore
challenging to calculate using computer simulations. The high po-
tency of 50a and 50b can be attributed to the correct distance be-
tween Aryl 1 and skeletal OH-groups, which both are capable of
In general, our results show that very flexible molecules are able
to bind and to activate the ER
rigid estradioles. The in vitro ERE reporter assay revealed several
compounds having a higher agonism at high (10 M) concentration
a/ERb receptor even better than the
l
than E2 has at 10 nM. However, there were only a few compounds
(Table 3) showing significant (>1%) RBA in COS-1 cells. These good
binders were also good ERE activators, without exception. Interest-
ing is that there were many other compounds which were good re-
porter activators but have or, on the basis of structural features, can
be assumed to have low RBA. One could say that when only the ERE
reporter activation is considered, the ER receptors are not finical at
all while the RBAs are very sensitive to geometry and, are thus
much better measures of the real biological activation.
In the most critical test, the best binders acted as partial ago-
nists also in pS2 gene expression assay in breast cancer cells. Nota-
ble is also that, in these experiments (Fig. 1), the compound 45a
showed unusual kinetics with fast but temporary activation, while
50b behaved more like E2. This compound can be considered as
convergent of the agonistic properties of the entire compound
group and that it owns the best potency for further developments
and applications.
forming hydrogen bonds of favorable geometry with ER
a and
ERb. Since some of the compounds studied show high affinities
and activation without being able to form three hydrogen bonds
like 50a and 50b, receptor–ligand interactions other than hydrogen
bonds play also role in determining compounds’ potency. For
example, the isoxazole ring of 45a/45b is predicted to bind roughly
at the same location as the hydrogen bond-forming hydroxyketone
part of 50a and 50b, but stabilizing the receptor complex by polar
interactions instead of hydrogen bonding. In addition, the best bin-
der molecules are slightly larger than E2 and therefore more effi-
ciently fill the ligand-binding cavity. Similar explanation has
been given for the superagonism of several vitamin D receptor
(VDR) ligands.28–30 The larger molecule size also explains the ER
selectivity, since the volume of the ligand-binding cavity of ER
(450 Å3) is slightly larger than that of ERb (390 Å3).31
a
a
5. Experimental
5.1. Chemistry
According to the expression assays compound 50b behaved
highly similarly to E2. Instead, compounds 45b and 50a did not
show statistically significant effects to pS2 expression, even at very
All moisture sensitive reactions were performed under a posi-
tive pressure of argon in oven dried glassware equipped with

P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
3443
rubber septa; solvents and liquid reagents were dried, distilled,
stored under argon, and transferred using a syringe flushed with
argon. Column chromatography was performed with silica gel 60
(SiO2, 70–230 mesh). Purity of >95% was confirmed via elemental
analyses (CHN) which were carried out with a Thermo Quest CE
Instruments EA 1110 CHNS-O elemental analyzer. 1H and 13C
NMR spectra (TMS/CDCl3) were recorded on a Bruker Avance 500
spectrometer operating at 500.13 MHz and 125.77 MHz, respec-
tively. All J coupling constants are given in Hz. Semi preparative
HPLC-separations of the enantiomers were performed on a Shima-
dzu chromatography system using a Regis Technologies (R,R)-
Whelk-O 1 (25 cm  10 mm id) chiral column in n-hexane/i-
PrOH/AcOH 90:10:0.5 at flow rate 5 ml/min. Compounds were de-
tected by UV absorption at 254 nm and 270 nm.
The methoxy aryl precursors C of the phenolic 4,5-dihydroisox-
azoles D were synthesized according to the procedure in Section
5.1.1. but not characterized. The preparation and characterization
of the following compounds have been described previously:23
5-benzyl-3-phenethyl-4,5-dihydroisoxazole (1), 5-benzyl-3-[2-(2-
hydroxy-phenyl)-ethyl]-4,5-dihydro-isoxazole (2), 5-benzyl-3-[2
-(4-hydroxy-phenyl)-ethyl]-4,5-dihydro-isoxazole (3), 5-(4-fuo-
robenzyl)-3-phenethyl-4,5-dihydroisoxazole (4), 5-(4-hydroxy-
benzyl)-3-phenethyl-4,5-dihydro-isoxazole (5), 5-(4-fluoro-ben-
zyl)-3-(2-hydroxy-phenethyl)-4,5-dihydro-isoxazole (6), 5-(4-flu-
oro-benzyl)-3-(4-hydroxy-phenethyl)-4,5-dihydro-isoxazole (7),
5-hydroxy-1,6-diphenylhexan-3-one (8), 5-hydroxy-6-(4-fuor-
ophenyl)-1-phenyl-hexan-3-one (11), 3,5-di-phenethyl-4,5-dihyd-
roisoxazole (15), 5-hydroxy-1,7-diphenyl-heptan-3-one (19),
5-benzyl-3-(3-phenyl-propyl)-4,5-dihydro-isoxazole (23), 2-hy-
droxy-1,7-diphenyl-heptan-4-one (29), 3-phenethyl-5-phenyl-
4,5-dihydroisoxazole (35), 1-hydroxy-1,5-diphenyl-pentan-3-one
(40), 5-phenyl-3-(3-phenyl-propyl)-4,5-dihydro-isoxazole (43),
1-hydroxy-1,6-diphenyl-hexan-3-one (48).
was added 2.2 ml of 1 M BBr3 in DCM (2.2 mmol, for compounds
having two aromatic methoxyl groups, 4.4 mmol was used) and
the solution was stirred under argon at rt overnight. The organic
layer was evaporated to give the oily product, which was purified
by column chromatography using DCM as an eluent.
5.1.2.1. 3-(2-Hydroxy-phenethyl)-5-phenethyl-4,5-dihydro-isox-
azole (16). Yield 85%, a dark brown viscous oil, 1H NMR d 7.27 (m,
2H), 7.19 (d, 1H, J = 7.3), 7.16 (m, 2H), 7.12 (dd, 1H, J = 8.0, 1.7), 7.09
(d, 1H, J = 7.4), 6.89 (d, 1H, J = 7.7), 6.85 (t, 1H, J = 7.4), 4.52 (m, 1H),
2.97–2.89 (m, 3H), 2.73 (m, 1H), 2.68–2.64 (m, 3H), 2.52 (dd, 1H,
J = 17.0, 8.3), 1.97 (m, 1H), 1.79 (m, 1H); 13C NMR d 159.2, 154.5,
141.1 (3s), 130.4, 128.5 (2d), 128.4 (s), 127.9, 126.1, 120.7, 117.8,
79.7 (5d), 42.8, 36.7, 31.8, 30.0, 25.9 (5t). Elemental analysis data
agreed: for C19H21NO2Á0.2H2O.
5.1.2.2. 3-(3-Hydroxy-phenethyl)-5-phenethyl-4,5-dihydro-isox-
azole (17). Yield 88%, a dark brown viscous oil, 1H NMR d 7.27 (m,
2H), 7.19 (d, 1H, J = 7.3), 7.16 (m, 2H), 7.12 (t, 1H, J = 8.0), 6.73–6.69
(m, 3H), 6.12 (br s, 1H, OH), 4.53 (m, 1H), 2.92 (dd, 1H, J = 17.1,
10.3), 2.84 (t, 2H, J = 7.6), 2.73 (ddd, 1H, J = 14.2, 9.6, 5.6), 2.68–
2.62 (m, 3H), 2.52 (dd, 1H, J = 17.1., 7.9), 1.96 (m, 1H), 1.77 (m,
1H); 13C NMR d 159.1, 156.2, 142.0, 141.1 (4s), 129.8, 128.5,
128.5, 126.1, 120.4, 115.3, 113.6, 79.6 (8d), 42.3, 36.8, 32.4, 31.7,
29.2 (5t). Elemental analysis data agreed: for C19H21NO2Á0.1H2O.
5.1.2.3. 3-(4-Hydroxy-phenethyl)-5-phenethyl-4,5-dihydro-isox-
azole (18). Yield 95%, a dark brown wax, 1H NMR d 7.28 (m, 2H),
7.20–7.17 (m, 3H), 7.03 (m, 2H), 6.74 (m, 2H), 4.50 (m, 1H), 3.02
(br s, 1H, OH), 2.90 (dd, 1H, J = 16.9, 10.2), 2.80 (t, 2H, J = 7.6),
2.73 (ddd, 1H, J = 16.1, 10.2, 5.6), 3.04 (ddd, 1H, J = 16.1, 9.5, 7.9),
2.61 (t, 2H, J = 7.7), 2.50 (dd, 1H, J = 16.9, 7.9), 1.95 (m, 1H), 1.77
(m, 1H). Elemental analysis data agreed: for C19H21NO2Á0.2H2O.
5.1.1. Synthesis of 4,5-dihydroisoxazoles C (compounds 36 and
44)
5.1.2.4. 5-Benzyl-3-[3-(4-hydroxy-phenyl)-propyl]-4,5-dihydro-
isoxazole (24). Yield 99%, a dark brown wax; 1H NMR d 7.28 (m,
2H), 7.21 (m, 3H), 6.97 (m, 2H), 6.76 (m, 2H), 4.78 (m, 1H), 3.00
(dd, 1H, J = 13.9, 6.0), 2.86 (dd, 1H, J = 17.0, 10.1), 2.81 (dd, 1H,
J = 13.9, 6.9), 2.62 (dd, 1H, J = 17.0, 7.6), 2.51 (t, 2H, J = 7.5), 2.28
(t, 2H, J = 7.7) 1.75 (m, 2H); 13C NMR d 159.3, 154.5, 136.9, 132.7
(4s), 129.5, 129.4, 128.6, 126.7, 115.3, 80.5, (6d), 41.5, 40.9, 34.3,
28.1, 27.2 (5t). Elemental analysis data agreed: for
C19H21NO2Á0.2H2O.
To a vigorously stirred solution of oxime A (20 mmol), alkene B
(20 mmol), and pyridine (4 mmol) in 100 ml of DCM at 0 °C was
added drop wise 5% NaOCl solution (60 ml, 40 mmol) keeping the
temperature under 5 °C and the mixture was reacted for 1.5 h.
The organic layer was separated and washed with 2 M HCl, 2 M
NaHCO3, and water, and evaporated to give the crude product,
which was purified by column chromatography using DCM as an
eluent.
5.1.2.5. 5-(4-Hydroxybenzyl)-3-(3-phenyl-propyl)-4,5-dihydro-
isoxazole (25). Yield 65%, a dark brown oil; 1H NMR d 7.24 (m,
2H), 7.15 (m, 1H), 7.11 (m, 2H), 7.00 (m, 2H), 6.80 (m, 2H), 4.71
(m, 1H), 2.87–2.81 (m, 2H), 2.72 (dd, 1H, J = 14.1, 6.3), 2.60–2.54
(m, 3H), 2.28 (t, 2H, J = 7.6), 1.79 (m, 2H); 13C NMR d 159.5,
155.1, 141.2 (3s), 130.4, 128.4, 128.4 (3d), 128.1 (s), 126.0, 115.6,
80.8, (3d), 41.3, 39.9, 35.1, 27.7, 27.2 (5t). Elemental analysis data
agreed: for C19H21NO2Á0.1H2O.
5.1.1.1. 5-Methyl-3-phenethyl-5-phenyl-4,5-dihydroisoxazole
(36). Yield 18%, colorless oil, 1H NMR d 7.39–7.36 (m, 2H), 7.33 (m,
2H), 7.24 (m, 3H), 7.19–7.16 (1H), 7.15–7.12 (2H), 3.01 (d, 1H,
J = 16.8), 2.96 (d, 1H, J = 16.8), 2.87 (t, 2H, J = 8.0), 2.62 (t, 2H,
J = 8.0), 1.65 (s, 3H); 13C NMR d 158.0, 145.8, 140.5 (3 s), 128.5,
128.4, 128.3, 127.2, 126.3, 124.7 (6d), 86.8 (s), 51.0, 32.7, 29.7
(3t), 28.1 (q). Elemental analysis data agreed: for C18H19NO.
5.1.1.2. 5-Methyl-5-phenyl-3-(3-phenyl-propyl)-4,5-dihydro-
isoxazole (44). Yield 34%, yellow viscous oil, 1H NMR d 7.40 (m,
2H), 7.32 (m, 2H), 7.24 (m, 3H), 7.17 (m, 1H), 7.10 (m, 2H), 3.03
(d, 1H, J = 16.7), 2.99 (d, 1H, J = 16.7), 2.58 (m, 2H), 2.31 (t, 2H,
J = 7.5), 1.84 (m, 2H), 1.68 (s, 3H); 13C NMR d 158.5, 145.8, 141.4
(3s), 128.4, 128.4, 128.4, 127.2, 125.9, 124.6 (6d), 86.6 (s), 50.7,
35.1 (2t), 28.1 (q), 27.9, 27.4 (2t). Elemental analysis data agreed:
for C19H21NO.
5.1.2.6. 5-(4-Hydroxy-benzyl)-3-[3-(4-hydroxy-phenyl)-propyl]-
4,5-dihydro-isoxazole (26). Yield 68%, a dark brown wax; 1H NMR
d 7.02 (m, 2H), 6.95 (m, 2H), 6.77 (m, 2H), 6.76 (m, 2H), 4.73 (m,
1H), 2.86 (dd, 1H, J = 14.1, 6.2), 2.85 (dd, 1H, J = 17.2, 10.4), 2.74
(dd, 1H, J = 14.1, 6.6), 2.62 (dd, 1H, J = 17.2, 7.7), 2.50 (t, 2H,
J = 7.5), 2.26 (t, 2H, J = 7.7) 1.73 (m, 2H); 13C NMR d 159.9, 155.5,
154.8, 132.5 (4s), 130.5, 129.5 (2d), 128.3 (s), 115.5, 115.3, 80.9
(3d), 41.3, 39.9, 34.3, 28.1, 27.1 (5t). Elemental analysis data
agreed: for C19H21NO3Á0.33H2O.
5.1.2. Synthesis of phenolic 4,5-dihydroisoxazoles D
(compounds 16–18, 24–28, 37–39, and 45–47)
To a solution of a 4,5-dihydroisoxazole C having an aromatic
methoxyl group (2 mmol, compounds 53–66) in 2 ml of DCM
5.1.2.7. 5-Benzyl-3-[3-(3,4-dihydroxy-phenyl)-propyl]-4,5-dihy-
dro-isoxazole (27). Yield 87%, a dark brown viscous oil, 1H NMR d
7.24 (m, 2H), 7.19 (d, 1H, J = 7.3), 7.17–7.14 (m, 2H), 6.76 (d, 1H,

3444
P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
J = 8.1), 6.70 (d, 1H, J = 1.8), 6.61 (br s, 2H, OH), 6.47 (dd, 1H, J = 8.1,
1.8), 4.75 (m, 1H), 2.94 (dd, 1H, J = 13.9, 6.1), 2.83 (dd, 1H, J = 17.3,
10.2), 2.77 (dd, 1H, J = 13.9, 6.6), 2.58 (dd, 1H, J = 17.3, 7.7), 2.38 (t,
2H, J = 7.4), 2.21 (t, 2H, J = 7.3), 1.68 (m, 2H); 13C NMR d 160.1,
143.9, 142.2, 136.7, 133.7 (5s), 129.4, 128.5, 126.7, 120.5, 115.7,
115.4, 80.6 (7d), 41.4, 40.7, 34.3, 20.7, 26.9 (5t). Elemental analysis
data agreed: for C19H21NO2Á0.5H2O.
d 7.18 (m, 1H), 6.99 (m, 2H), 6.79–6.77 (m, 3H), 6.75 (m, 2H),
5.46 (dd, 1H, J = 10.8, 8.0), 3.33 (dd, 1H, J = 17.2, 10.8), 2.89 (dd,
1H, J = 17.2, 8.0), 2.57 (t, 2H, J = 7.5), 2.37 (t, 2H, J = 7.6), 1.85 (tt,
2H, J = 7.6, 7.5); 13C NMR d 159.6, 157.4, 155.1, 142.8, 132.5 (5s),
130.1, 129.6, 117.0, 115.5, 115.3, 112.7, 81.4 (7d), 45.5,
34.5, 28.4, 27.2 (4t). Elemental analysis data agreed: for
C18H21NO3Á0.5H2O.
5.1.2.8. 3-[3-(3,4-Dihydroxy-phenyl)-propyl]-5-(4-hydroxy-ben-
zyl)-4,5-dihydro-isoxazole (28). Yield 29%, a dark brown viscous
oil, 1H NMR (in CD3OD) d 7.05 (m, 2H), 6.76 (m, 2H), 6.75 (d, 1H,
J = 8.0), 6.60 (d, 1H, J = 2.0), 6.51 (dd, 1H, J = 8.0, 2.0), 4.76 (m,
1H), 2.87–2.82 (m, 2H), 2.79 (dd, 1H, J = 14.1, 6.3), 2.61 (dd, 1H,
J = 17.1, 7.5), 2.43 (m, 2H), 2.32 (m, 2H), 1.69 (m, 2H); 13C NMR
(in CD3OD) d 159.4, 155.3, 144.1, 142.7, 133.4 (5s), 130.7 (d),
127.9 (s), 120.2, 115.5, 115.4, 115.0, 80.6 (5d), 41.0, 39.8, 34.2,
27.8, 26.9 (5t). Elemental analysis data agreed: for C19H21
NO3Á0.2H2O.
5.1.3. Synthesis of hydroxyketones E (compounds 9–14, 20–22,
30–34, 41, 42, and 49–52)
A solution of an isoxazoline D (10 mmol), AcOH (0.1 mol) and
water (1.0 mol) in 100 ml of MeOH/THF (1:1) was stirred under
H2 at rt overnight in the presence of Raney Ni (1.5 g). The catalyst
was removed by filtering through CeliteÒ and the filtrate extracted
by DCM. The extracts were washed with satd NaHCO3 and water,
and evaporated to give the pure product.
5.1.3.1. 5-Hydroxy-1-(2-hydroxy-phenyl)-6-phenyl-hexan-3-
one (9). Yield 68%, a brown viscous oil, 1H NMR d 7.26 (m, 2H),
7.19 (m, 1H), 7.14 (m, 2H), 7.07 (t, 1H, J = 7.8), 6.65–6.62 (m, 2H),
6.59 (m, 1H), 4.29 (m, 1H), 2.80 (dd, 1H, J = 13.6, 7.2), 2.78–2.73
(m, 2H), 2.68 (dd, 1H, J = 13.6, 6.4), 2.66–2.63 (m, 2H), 2.51 (m,
2H); 13C NMR d 211.1, 156.1, 142.4, 13.6 (4s), 129.7, 129.4, 128.6,
126.6, 120.2, 115.3, 113.3, 68.9 (8d), 48.3, 44.7, 42.9, 29.2 (4t). Ele-
mental analysis data agreed: for C18H20O3Á0.2H2O.
5.1.2.9. 3-[2-(2-Hydroxy-phenyl)-ethyl]-5-methyl-5-phenyl-4,5-
dihydro-isoxazole (37). Yield 22%, a brown viscous oil, 1H NMR d
7.39–7.32 (m, 4H), 7.26 (m, 1H), 7.09 (m, 1H), 6.70 (m, 2H), 6.60
(m, 1H), 5.56 (br s, 1H, OH), 3.03 (d, 1H, J = 16.9), 2.99 (d, 1H,
J = 16.9), 2.81 (t, 2H, J = 7.5), 2.62 (t, 2H, J = 7.5), 1.66 (s, 3H); 13C
NMR d 158.5, 156.0, 145.6, 142.0 (4s), 129.7, 128.5, 127.3, 124.7,
120.5, 115.3, 113.5 (7d), 87.0 (s), 51.0, 32.5, 29.4 (3t), 28.0 (q). Ele-
mental analysis data agreed: for C18H19NO2Á0.1H2O.
5.1.3.2. 5-Hydroxy-1-(4-hydroxy-phenyl)-6-phenyl-hexan-3-one
(10). Yield 54%, a brown viscous oil, 1H NMR d 7.27 (m, 2H), 7.21
(m, 1H), 7.15 (m, 2H) 6.94 (m, 2H), 6.70 (m, 2H), 4.67 (br s, 1H,
OH), 4.29 (m, 1H), 2.81 (dd, 1H, J = 13.6, 7.2), 2.77–2.73 (m, 2H),
2.69 (dd, 1H, J = 13.7, 6.5), 2.66–2.62 (m, 2H), 2.51 (m, 2H); 13C
NMR d 211.4, 154.4, 137.6, 132.2 (4s), 129.4, 129.3, 128.6, 126.6,
115.5, 68.9 (6d), 48.3, 45.3, 42.9, 25.1 (4t). Elemental analysis data
agreed: for C18H20O3Á0.33H2O.
5.1.2.10. 3-[2-(3-Hydroxy-phenyl)-ethyl]-5-methyl-5-phenyl-
4,5-dihydro-isoxazole (38). Yield 97%, a brown viscous oil, 1H
NMR d 7.37–7.31 (m, 4H), 7.25 (t, 1H, J = 6.9), 7.04 (t, 1H, J = 7.6),
7.00 (d, 1H, J = 7.5), 6.81 (d, 1H, J = 7.9), 6.75 (t, 1H, J = 7.4), 4.68
(br s, 1H, OH), 3.08 (d, 1H, J = 17.1), 3.06 (d, 1H, J = 17.1), 2.86 (t,
2H, J = 7.5), 2.65 (t, 2H, J = 7.3), 1.64 (s, 3H); 13C NMR d 159.8,
154.9, 145.6, (3s), 130.2, 128.5, 127.7, 127.4 (4d), 127.0 (s),
124.7, 120.0, 115.7 (3d), 87.1 (s), 51.0 (t), 28.3 (q), 28.0, 27.4 (2t).
Elemental analysis data agreed: for C18H19NO2.
5.1.3.3. 5-Hydroxy-6-(4-fuorophenyl)-1-phenyl-hexan-3-one
(11). Yield 66%, a yellow viscous oil; 1H NMR d 7.25 (m, 2H), 7.17
(m, 1H), 7.15–7.10 (m, 4H), 6.96 (m, 2H), 4.23 (m, 1H), 2.96 (s, 1H,
OH), 2.87 (t, 2H, J = 7.5), 2.77–2.70 (m, 3H), 2.65 (dd, 1H, J = 13.8,
5.9), 2.50 (m, 2H); 13C NMR d 210.5, 161.7 (d), 140.7 (s), 133.6
(d), 130.8 (dd), 128.5, 128.2, 126.2 (3d), 115.2 (dd), 68.5 (d), 48.4,
45.0, 42.0, 29.4 (4t). Elemental analysis data agreed: for C18H20FO2.
5.1.2.11. 3-[2-(4-Hydroxy-phenyl)-ethyl]-5-methyl-5-phenyl-4,5
-dihydro-isoxazole (39). Yield 19%, a brown viscous oil, 1H NMR d
7.38–7.32 (m, 4H), 7.26 (m, 1H), 6.96 (m, 2H), 6.69 (m, 2H), 3.03 (d,
1H, J = 16.8), 2.99 (d, 1H, J = 16.8), 2.79 (t, 2H, J = 7.6), 2.61 (t, 2H,
J = 7.6), 1.65 (s, 3H); 13C NMR d 158.4, 154.3, 145.6, 132.1 (4s),
129.4, 128.5, 127.2, 124.7, 115.4 (5d), 86.9 (s), 50.9, 31.8, 29.8
(3t), 28.0 (q). Elemental analysis data agreed: for C18H19NO2.
5.1.3.4. 5-Hydroxy-6-(4-Hydroxy-phenyl)-1-phenyl-hexan-3-one
(12). Yield 99%, a brown viscous oil, 1H NMR d 7.24 (m, 2H), 7.16
(m, 1H), 7.11 (m, 2H), 6.97 (m, 2H), 6.71 (m, 2H), 4.23 (m, 1H),
2.84 (t, 2H, J = 15.0), 2.73–2.68 (m, 3H), 2.61 (dd, 1H, J = 13.0,
5.5), 2.52 (m, 2H); 13C NMR d 211.2, 154.8, 140.6 (3s), 130.5 (d),
129.2 (s), 128.5, 128.2, 126.2, 115.5, 69.1 (5d), 48.3, 45.0, 42.0,
29.4 (4t). Elemental analysis data agreed: for C18H20O3Á0.5H2O.
5.1.2.12. 3-[3-(4-Hydroxy-phenyl)-propyl]-5-phenyl-4,5-dihy-
dro-isoxazole (45). Yield 99%, a brown viscous oil, 1H NMR d
7.36–7.27 (m, 5H), 6.96 (m, 2H), 6.80 (m, 2H), 5.54 (dd, 1H,
J = 10.7, 8.3), 3.35 (dd, 1H, J = 17.4, 10.7), 2.91 (dd, 1H, J = 17.4,
8.3), 2.55 (t, 2H, J = 7.5), 2.40 (t, 2H, J = 7.6), 1.84 (tt, 2H, J = 7.6,
7.5); 13C NMR d 159.9, 154.5, 140.5, 132.6 (4s), 129.5, 128.8,
128.3, 125.8, 115.7, 81.6 (6d), 45.2, 34.3, 28.1, 27.1 (4t). Elemental
analysis data agreed: for C18H21NO2Á0.2H2O.
5.1.3.5. 6-(4-Fluoro-phenyl)-5-hydroxy-1-(2-hydroxy-phenyl)-
hexan-3-one (13). Yield 94%, a brown wax, 1H NMR d 7.09–7.03
(m, 3H), 6.91 (m, 2H), 6.64–6.61 (m, 2H), 6.59 (m, 1H), 4.25 (m,
1H), 2.76–2.71 (m, 3H), 2.66–2.60 (m, 3H), 2.50 (dd, 1H, J = 17.3,
8.3), 2.47 (dd, 1H, J = 17.3, 3.8); 13C NMR d 211.3 (s), 161.7 (d),
156.1, 142.4 (2s), 133.3 (d), 130.8 (dd), 129.7, 120.2, 115.3 (d),
115.2 (dd), 113.3, 68.9 (2d), 48.3, 44.7, 41.9, 29.1 (4t). Elemental
analysis data agreed: for C18H20FO3Á0.1H2O.
5.1.2.13. 3-[3-(4-Hydroxy-phenyl)-propyl]-5-methyl-5-phenyl-
4,5-dihydro-isoxazole (46). Yield 99%, a brown viscous oil, 1H
NMR d 7.40 (m, 2H), 7.33 (m, 2H), 7.24 (m, 2H), 6.92 (m, 2H),
6.75 (m, 2H), 6.54 (br s, 1H, OH), 3.06 (d, 1H, J = 16.9), 3.02 (d,
1H, J = 16.9), 2.49 (m, 2H), 2.31 (t, 2H, J = 7.6), 1.79 (m, 2H), 1.69
(s, 3H); 13C NMR d 159.3, 154.3, 145.5, 132.9 (4s), 129.4, 128.5,
127.3, 124.6, 115.4 (5d), 86.7 (s), 50.7, 34.2, 28.1 (3t), 28.0 (q),
27.3 (t). Elemental analysis data agreed: for C19H23NO2.
5.1.3.6. 6-(4-Fluoro-phenyl)-5-hydroxy-1-(4-hydroxy-phenyl)-
hexan-3-one (14). Yield 99%, a dark brown viscous oil, 1H NMR d
7.08 (m, 2H), 6.95–6.90 (m, 4H), 6.69 (m, 2H), 4.25 (m, 1H),
2.78–2.72 (m, 3H), 2.67–2.61 (m, 3H), 2.50 (dd, 1H, J = 17.3, 7.6),
2.49 (dd, 1H, J = 17.3, 4.0); 13C NMR d 211.5 (s), 161.7 (d), 154.4
(s), 133.3 (d), 132.2 (s), 130.8 (dd), 129.3, 115.5 (2d), 115.3 (dd),
5.1.2.14. 3-[3-(4-Hydroxy-phenyl)-propyl]-5-(3-hydroxy-phenyl)-
4,5-dihydro-isoxazole (47). Yield 99%, a brown viscous oil, 1H NMR

P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
3445
68.9 (d), 48.3, 45.3, 41.9, 28.6 (4t). Elemental analysis data agreed:
5.1.3.14.
2-Hydroxy-7-(3,4-dihydroxy-phenyl)-1-(4-hydroxy-
for C18H20FO3Á0.2H2O.
phenyl)-heptan-4-one (34). Yield 60%, a reddish brown viscous
oil, 1H NMR d 7.02 (m, 2H), 6.98 (m, 1H), 6.76 (d, 1H, J = 1.9),
6.72 (m, 2H), 6.61 (d, 1H, J = 7.2), 6.50 (dd, 1H, J = 7.2, 1.9), 4.21
(m, 1H), 2.78 (dd, 1H, J = 13.5, 7.0), 2.63 (dd, 1H, J = 13.5, 6.1),
2.50–2.38 (m, 2H), 2.37 (t, 2H, J = 7.0), 2.35 (t, 2H, J = 7.0), 1.65
(m, 2H). Elemental analysis data agreed: for C19H22O5ÁH2O.
5.1.3.7. 5-Hydroxy-1-(2-hydroxy-phenyl)-7-phenyl-heptan-3-
one (20). Yield 74%, a brown viscous oil, 1H NMR d 7.24 (m, 2H),
7.20–7.13 (m, 3H), 7.08–7.02 (m, 2H), 6.79 (m, 2H), 4.02 (m, 1H),
2.83 (t, 2H, J = 7.3), 2.78–2.71 (m, 3H), 2.62 (m, 1H), 2.57–2.48
(m, 2H), 1.76 (m, 1H), 1.65 (m, 1H); 13C NMR d 213.1, 154.5,
141.7 (3s), 130.3, 128.9, 128.4, 127.7 (4d), 127.2 (s), 125.9, 120.3,
116.0, 67.1 (4d), 49.2, 44.1, 38.2, 31.7, 24.1 (5t). Elemental analysis
data agreed: for C19H22O3Á0.5H2O.
5.1.3.15. 5-Hydroxy-1,5-diphenyl-hexan-3-one (41). Yield 73%,
pale brown viscous oil, 1H NMR d 7.38 (m, 2H), 7.30 (m, 2H),
7.25–7.19 (m 3H), 7.15 (m, 1H), 7.04 (m, 2H), 3.11 (d, 1H,
J = 16.7), 2.77 (d, 1H, J = 16.7), 2.77–2.63 (m, 3H), 2.55 (m, 1H),
1.49 (s, 3H); 13C NMR d 211.7, 147.2, 140.5 (3s), 128.5, 128.3,
128.2, 126.7, 126.1, 124.3 (6d), 73.3 (s), 53.5, 45.9 (2t), 30.6 (q),
29.1 (t). Elemental analysis data agreed: for C18H20O2Á0.1H2O.
5.1.3.8. 5-Hydroxy-1-(3-hydroxy-phenyl)-7-phenyl-heptan-3-one
(21). Yield 61%, a brown viscous oil, 1H NMR d 7.23 (m, 2H), 7.16–
7.11 (m, 3H), 7.08 (dd, 1H, J = 7.8, 7.8), 6.65 (m, 2H), 6.62 (m, 1H),
4.06 (m, 1H), 2.78 (t, 2H, J = 7.3), 2.75 (m, 1H), 2.64 (t, 2H, J = 7.3),
2.61 (m, 1H), 2.55–2.46 (m, 2H), 1.79 (m, 1H), 1.66 (m, 1H); 13C
NMR d 211.6, 156.1, 142.4, 141.6 (4s), 129.7, 128.4, 128.4, 125.9,
120.6, 115.3, 113.4, 67.3 (8d), 49.2, 44.7, 38.0, 31.7, 29.3 (5t). Ele-
mental analysis data agreed: for C19H22O3Á0.2H2O.
5.1.3.16. 5-Hydroxy-1-(4-hydroxy-phenyl)-5-phenyl-hexan-3-
one (42). Yield 25%, yellow viscous oil, 1H NMR d 7.37 (m, 2H),
7.30 (m, 2H), 7.21 (m, 1H), 6.87 (m, 2H), 6.68 (m, 2H), 6.05 (br s,
1H, OH), 4.74 (br s, 1H, OH), 3.14 (d, 1H, J = 16.8), 2.77 (d, 1H,
J = 16.8), 2.72–2.59 (m, 3H), 2.53 (m, 1H), 1.49 (s, 3H); 13C NMR d
212.4, 154.2, 146.9, 132.2 (4s), 129.3, 128.4, 126.9, 124.3, 115.4
(5d), 73.8 (s), 53.4, 46.2 (2t), 30.6 (q), 28.3 (t). Elemental analysis
data agreed: for C18H20O3Á0.2H2O.
5.1.3.9. 5-Hydroxy-1-(4-hydroxy-phenyl)-7-phenyl-heptan-3-
one (22). Yield 54%, a pale brown viscous oil, 1H NMR d 7.25 (m,
2H), 7.17–7.13 (m, 3H), 6.97 (m, 2H), 6.72 (m, 2H), 4.05 (m, 1H),
2.79–2.60 (m, 6H), 2.52 (m, 2H), 1.79 (m, 1H), 1.65 (m, 1H); 13C
NMR d 211.8, 154.4, 141.6, 132.2 (4s), 129.3, 128.4, 128.4, 125.9,
115.5, 67.2 (6d), 49.1, 45.3, 37.9, 31.7, 28.6 (5t). Elemental analysis
data agreed: for C19H22O3Á0.33H2O.
5.1.3.17. 6-Hydroxy-1,6-diphenyl-heptan-4-one (49). Yield 16%,
yellow viscous oil, 1H NMR d 7.40 (m, 2H), 7.31 (m, 2H), 7.25 (m,
2H), 7.21 (m, 1H), 7.17 (m, 1H), 7.07 (m, 2H), 3.10 (d, 1H,
J = 16.8), 2.75 (d, 1H, J = 16.8), 2.48 (m, 2H), 2.34 (dt, 1H, J = 17.4,
7.3), 2.24 (dt, 1H, J = 17.4, 7.3), 1.77 (m, 2H), 1.50 (s, 3H); 13C
NMR d 212.6, 147.3, 141.3 (3s), 128.4, 128.4, 128.3, 126.7, 126.0,
124.3 (6d), 73.4 (s), 53.3, 46.3, 34.7 (3t), 30.7 (q), 24.5 (t). Elemental
analysis data agreed: for C19H22O2Á0.1H2O.
5.1.3.10 . 2-Hydroxy-7-(4-hydroxy-phenyl)-1-phenyl-heptan-4-
one (30). Yield 49%, a pale brown viscous oil; 1H NMR d 7.27 (m,
2H), 7.21 (m, 1H), 7.16 (m, 2H), 6.94 (m, 2H), 6.72 (m, 2H), 4.29
(m, 1H), 2.82 (dd, 1H, J = 13.6, 8.9), 2.70 (dd, 1H, J = 13.6, 6.4),
2.52 (m, 2H), 2.48 (t, 2H, J = 7.5), 2.35 (t, 2H, J = 7.3), 1.81 (tt, 2H,
J = 7.5, 7.3); 13C NMR d 212.3, 154.1, 137.6, 133.1 (4s), 129.5,
129.4, 128.6, 126.6, 115.3, 69.0, (6d), 48.0, 42.8, 42.7, 34.0, 25.1
(5t). Elemental analysis data agreed: for C19H22O3Á0.25H2O.
5.1.3.18. 1-Hydroxy-6-(4-hydroxy-phenyl)-1-phenyl-hexan-3-
one (50). Yield 14%, a brown viscous oil; 1H NMR d 7.33 (m, 4H),
7.24 (m, 1H), 6.94 (m, 2H), 6.71 (m, 2H), 6.33 (br s, 1H, OH), 5.12
(dd, 1H, J = 9.3, 3.3), 4.00 (br s, 1H, OH), 2.81 (dd, 1H, J = 17.2,
9.3), 2.71 (dd, 1H, J = 17.2, 3.3), 2.48 (t, 2H, J = 7.5), 2.39 (t, 2H,
J = 7.3), 1.83 (tt, 2H, J = 7.5, 7.3); 13C NMR d 211.7, 154.2, 142.6,
133.1 (4s), 129.5, 128.6, 127.7, 125.7, 115.3, 70.1 (6d), 50.9, 42.8,
34.0, 25.1 (4t). Elemental analysis data agreed: for
C18H20O3Á0.25H2O.
5.1.3.11. 2-Hydroxy-1-(4-hydroxy-phenyl)-7-phenyl-heptan-4-
one (31). Yield 99%, a pale brown wax; 1H NMR d 7.26 (m, 2H),
7.17 (m, 1H), 7.13 (m, 2H), 7.00 (m, 2H), 6.72 (m, 2H), 4.23 (m,
1H), 2.72 (dd, 1H, J = 13.7, 7.2), 2.63 (dd, 1H, J = 13.7, 6.2), 2.58 (t,
2H, J = 7.6), 2.51 (m, 2H), 2.38 (t, 2H, J = 7.4), 1.87 (tt, 2H, J = 7.6,
7.4); 13C NMR d 212.0, 154.7, 141.3 (3s), 130.5 (d), 129.3 (s),
128.4, 128.4, 126.0, 115.5, 69.1, (6d), 48.0, 42.8, 41.9, 34.9, 24.8
(5t). Elemental analysis data agreed: for C19H22O3Á0.25H2O.
5.1.3.19. 6-Hydroxy-1-(4-hydroxy-phenyl)-6-phenyl-heptan-4-
one (51). Yield 11%, a brown viscous oil; 1H NMR d 7.39 (m, 2H),
7.31 (m, 2H), 7.21 (m, 1H), 6.90 (m, 2H), 6.72 (m, 2H), 5.90 (br s,
1H, OH), 4.73 (br s, 1H, OH), 3.13 (d, 1H, J = 16.9), 2.76 (d, 1H,
J = 16.9), 2.43–2.29 (m, 4H), 1.72 (m, 2H), 1.51 (s, 3H); 13C NMR d
213.1, 154.1, 147.0, 133.1 (4s), 129.5, 128.3, 126.8, 124.3, 115.3,
73.7 (6d), 53.2, 43.6, 33.8 (3t), 30.6 (q), 24.8 (t). Elemental analysis
data agreed: for C19H22O3Á0.2H2O.
5.1.3.12. 2-Hydroxy-1,7-di-(4-hydroxy-phenyl)-heptan-4-one
(32). Yield 16%, a colorless viscous oil; 1H NMR d 7.06 (m, 2H),
6.99 (m, 2H), 6.76 (m, 2H), 6.73 (m, 2H), 4.22 (m, 1H), 2.95 (br s,
1H, OH), 2.86 (dd, 1H, J = 13.6, 7.1), 2.65 (dd, 1H, J = 13.6, 6.2),
2.55–2.51 (m, 3H), 2.47 (dd, 1H, J = 17.4, 8.5), 2.38 (t, 2H, J = 7.3),
1.85 (tt, 2H, 7.5, 7.3); 13C NMR d 215.9, 154.3, 153.8, 133.6 (4s),
130.6, 129.5 (2d), 125.0 (s), 115.4, 115.3, 68.9 (3d), 48.1, 42.6,
42.0, 29.7, 25.1 (5t). Elemental analysis data agreed: for C19H22O4.
5.1.3.20. 1-Hydroxy-1-(3-hydroxy-phenyl)-6-(4-hydroxy-phe-
nyl)-hexan-3-one (52). Yield 19%, a brown viscous oil; 1H NMR d
8.51 (br s, 1H, OH), 8.21 (br s, 1H, OH), 7.12 (t, 1H, J = 7.8), 6.93 (m,
2H), 6.79–6.76 (m, 2H), 6.73–6.69 (m, 3H), 5.01 (dd, 1H, J = 9.1,
3.5), 4.37 (br s, 1H, OH), 2.77 (dd, 1H, J = 16.8, 9.1), 2.65 (dd, 1H,
J = 16.8, 3.5), 2.47 (t, 2H, J = 7.4), 2.36 (td, 2H, J = 7.4, 2.5), 1.80
(tt, 2H, J = 7.4, 7.4); 13C NMR d 211.8, 156.7, 154.6, 144.6, 132.7
(5s), 129.7, 129.5, 117.3, 115.3, 114.9, 112.6, 69.9 (7d), 51.1, 42.8,
34.0, 25.2 (4t). Elemental analysis data agreed: for
C18H20O4Á0.33H2O.
5.1.3.13. 2-Hydroxy-7-(3,4-dihydroxy-phenyl)-1-phenyl-heptan-
4-one (33). Yield 59%, a reddish brown viscous oil, 1H NMR d 7.26–
7.21 (m, 2H), 7.18 (m, 1H), 7.13–7.10 (m, 2H), 6.70 (d, 1H, J = 7.3),
6.59 (d, 1H, J = 1.8), 6.47 (dd, 1H, J = 7.3, 1.8), 4.28 (m, 1H), 2.78 (dd,
1H, J = 13.5, 7.1), 2.66 (dd, 1H, J = 13.5, 6.2), 2.53-2.40 (m, 2H), 2.35
(t, 2H, J = 7.0), 2.28 (t, 2H, J = 7.0), 1.64 (m, 2H); 13C NMR d 213.1,
144.2, 142.5, 138.0, 134.5 (5s), 129.8, 129.0, 127.1, 121.0, 116.1,
115.9, 69.5 (7d), 48.5, 43.2, 43.0, 34.5, 25.3 (5t). Elemental analysis
data agreed: for C19H22O4Á0.5H2O.

3446
P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
5.2. Biology
(Roche) using TaqMan Gene Expression Assays for RPLP0
(4333761F) and pS2 (4331182) (Applied Biosystems, Warrington,
DDCt)
5.2.1. Agonistic activity assay
UK). Fold changes were calculated using the formula 2À(
,
E2 and ICI-182,780 were from Sigma Chemical Co. (St. Louis,
MO) and Tocris (Avonmouth, UK), respectively. All other reagents
were of reagent grade from Sigma or Fluka. HEK293 cells were
seeded in 48-well plates (70 Â 103 cells per well) in phenol-free
Dulbecco’s modified Eagle medium supplemented with 5% delipi-
dated fetal bovine serum (Sigma) and antibiotics.32 After a 24 h
culture, the medium was changed and the cells were transfected
where DDCt = Ct(stimulus) Ct(solvent) and Ct = Ct(pS2)
Ct(RPLP0). Ct is the cycle at which the threshold line is crossed.
Two-tailed Student’s t-tests were performed to calculate p-values.
D
À
D
,
D
À
5.3. Computational methods
Selected ligands were docked to the ER
a
(PDB entry 1X7R)34
for 4 h with 5 ng ER
a
or ERb expression vector,33 75 ng reporter
and ERb (PDB entry 1X7B)14 ligand-binding pockets using flexible
docking and extra precision mode of the GLIDE program35,36 of
Schrödinger package. Molecular dynamics (MD) simulations of
plasmid pERE2TATA-LUC26 and 20 ng control plasmid pCMVb as
described.32 After transfection, the cells received fresh medium
containing either vehicle (0.1% v/v), test compound (10 lM) or
the R and S enantiomers of 50 bound to ER
a and ERb were done
10 nM E2. When appropriate, the antagonist ICI-182,780
(1000 nM) or vehicle was included in medium. After 24 h, the cells
were washed, lysed and assayed for luciferase and b-galactosidase
activities32 with a Victor2™ reader (PerkinElmer Wallac, Turku,
Finland). After normalization for b-galactosidase activity, luciferase
activities are expressed relative to that of 10 nM E2 by the formula:
Activity = 100% Â [(Test compound) À (Vehicle)/(E2) À (Vehicle)],
where terms in parenthesis indicate the corresponding normalized
luciferase activities. Typically, more than 90-fold activation by
10 nM E2 of luciferase with both ER subtypes was seen. The data
are means SEM of at least three independent transfections.
with the AMBER 9 program37 and the Amber ff99SB force field.38,39
To start the MD simulations, all the hydrogens and waters were
minimized and then heated to 300 K and simulated for 100 ps.
After that the whole system was minimized and heated to 300 K.
After 300 ps of equilibration, 600 ps production runs were per-
formed for each system.
The binding free energies of (R)-50 and (S)-50 to ER
a and ERb
were estimated using the MM-PBSA40,41 approach as implemented
in the AMBER 9. The structures used in the free energy calculations
were extracted from the first 150 ps of the MD trajectories. The
structural water molecule participating in the hydrogen bonding
network of the hydroxyl group of the ligand, Arg394/Arg346 (ERa/
ERb) and Glu353/Glu305 (Fig. 2) was included in the MM-PBSA
5.2.2. Relative binding affinity assay
Relative binding affinities (RBA) were measured by a competi-
tive assay against [6,7-3H(N)]estradiol (PerkinElmer) in transiently
transfected COS-1 cells. One day before transfection, COS-1 cells
were seeded into 2 ml of DMEM (Dulbecco‘s modified Eagle
medium, Gipco) with 10% delipidated fetal bovine serum and
0,25% (vol/vol) Penicillin–Streptomycin (Euroclone) at a density
of 140 Â 103 cells/well. After medium change to DMEM 2.5% FBS,
the cells were transfected for 24 h with 10 ng/well of human
calculations.
Two different thermodynamic integration (TI) calculations42
were done: (i) the starting structure of the ER
was the end structure of 600 ps MD simulation of ER
plex and the atomic point charges of both (R)-50 and (S)-50 were
those of (R)-50, and (ii) the end structure of the 600 ps MD
a–ligand complex
a
–(R)-50 com-
simulation of ERa–(S)-50 complex and the atomic point charges
of (S)-50 were used to set up the TI calculations. Nine-point Gauss-
ian quadratures were used to calculate the total relative free
energy. In the TI simulations of each of the nine different lambda
values 100 ps equilibration simulations were done first, followed
by 450 ps production simulations.
ERa/ERb expression vector pSG5-hERa/b by using the TransIT
method (Micrus Bio TransIT-LT1, Transfection Reagent). After
36 h, the cells were treated with tested compounds using 0.01-,
0.1-, 1-, 10-, 100-, 1000-, and 10,000-fold excess as compared to
labeled E2 (1.96 pM/well). After 2 h of incubation at 37 °C the med-
ium was removed. The cells were removed from the wells to 150 ll
Acknowledgments
of 1Â phosphate buffered saline (PBS), transferred to Eppendorf
tubes and centrifuged at 4 °C using 4000g for 5 min, and then
We would like to thank Mrs. Tarja Ihalainen and Mrs. Kaarina
Pitkänen for expert technical assistance and Mrs. Tiina Koivunen
for performing the elemental analyses. J.P. (Grant 75791), S.V.
(Grant 00897) and P.H. (Grant 51610) were supported in part by
the Academy of Finland. Financial support from the Foundation
for Finnish Innovations (projects 20740 and 20741) is thankfully
acknowledged.
washed twice with 150
ll of PBS. The cell pellets were dissolved
to 50 l of 0.5 M NaOH and incubated for 15 min at 56 °C, after
l
which the samples were transferred to liquid squintillation tubes
and treated with 3 ml of OptiPhase HiSafe 3 solution (PerkinEl-
mer). The results were measured with LKB WALLAC 1214 Racbeta
equipment and non-specific binding was subtracted from the total
[3H]estradiol binding to obtain the specific ER binding.
5.2.3. Gene expression assay
References and notes
MCF-7 human breast cancer cells were maintained in DMEM
containing 5% FBS, 2 mM L-glutamine, 0.1 mg/ml streptomycin
and 100 U/ml penicillin in humidified 95% air/5% CO2 incubator.
Prior to RNA extraction, the cells were grown overnight in phenol
red-free DMEM, supplemented with 5% charcoal-stripped FBS,
1. Manson, J. E.; Martin, K. A. N. Eng. J. Med. 2001, 345, 34.
2. Chiara, B.; Vegeto, E.; Pinna, C.; Maggi, A.; Cignarella, A. Arterioscler. Thromb.
Vasc. Biol. 2006, 26, 2192.
3. Raafat, A. M.; Hofseth, L. J.; Li, S.; Bennett, J. M.; Haslam, S. Z. Endocrinology
1999, 140, 2570.
4. Norman, B. H.; Dodge, J. A.; Richardson, T. I.; Borromeo, P. S.; Lugar, C. W.; Jones,
S. A.; Chen, K.; Wang, Y.; Durst, G. L.; Barr, R. J.; Montrose-Rafizadeh, C.;
Osborne, H. E.; Amos, R. M.; Guo, S.; Boodhoo, A.; Krishnan, V. J. Med. Chem.
2006, 49, 6155.
5. Wallace, O. B.; Lauwers, K. S.; Dodge, J. A.; May, S. A.; Calvin, J. R.; Hinklin, R.;
Bryant, H. U.; Shetler, P. K.; Adrian, M. D.; Geiser, A. G.; Sato, M.; Burris, T. P. J.
Med. Chem. 2006, 49, 843.
6. Grese, T. A.; Dodge, J. Curr. Pharm. Des. 1998, 4, 71.
7. Green, S.; Walter, P.; Kumar, V.; Krust, A.; Bornert, J. M.; Argos, P.; Chambon, P.
Nature 1986, 320, 134.
8. Kuiper, G. G. J. M.; Enmark, E.; Pelto-Huikko, M.; Nilsson, S.; Gustafsson, J. Å.
Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 5925.
L
-glutamine and antibiotics. In expression assays, the cells were
treated with different ligands (final concentration 10 nM) or sol-
vent for 6 h or 24 h prior to extraction of total RNA using High Pure
RNA Isolation Kit (Roche, Mannheim, Germany) and cDNA synthe-
sis was performed using Transcriptor High Fidelity cDNA Synthesis
Kit (Roche) according to the manufacturer’s instructions. In compe-
tition assays, the cells were treated for 24 h with 10 nM E2 and
1
lM of different compounds prior to RNA extraction. Real-time
quantitative PCR was performed with LightCycler 480 apparatus

P. K. Poutiainen et al. / Bioorg. Med. Chem. 18 (2010) 3437–3447
3447
9. Dahlman-Wright, K.; Cavailles, V.; Fuqua, S. A.; Jordan, V. C.; Katzenellenbogen,
J. A.; Korach, K. S.; Maggi, A.; Muramatsu, M.; Parker, M. G.; Gustafsson, J.-Å.
Pharmacol. Rev. 2006, 58, 773.
10. Moon, B. S.; Carlson, K. E.; Katzenellenbogen, J. A.; Choi, T. H.; Chi, D. Y.; Kim, J.
Y.; Cheon, G. J.; Koh, H. Y.; Lee, K. C.; An, G. Bioorg. Med. Chem. 2009, 17, 3479.
11. Barrett, I.; Meegan, M. J.; Hughes, R. B.; Carr, M.; Knox, A. J. S.; Artemenko, N.;
Golfis, G.; Zisterer, D. M.; Lloyd, D. G. Bioorg. Med. Chem. 2008, 16, 9554.
12. Stauffer, S. R.; Sun, J.; Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Bioorg.
Med. Chem. 2000, 8, 1293.
13. Mortensen, D. S.; Rodriguez, A. L.; Sun, J.; Katzenellenbogen, B. S.;
Katzenellenbogen, J. A. Bioorg. Med. Chem. Lett. 2001, 11, 2521.
14. Manas, E. S.; Unwalla, R. J.; Xu, Z. B.; Malamas, M. S.; Miller, C. P.; Harris, H. A.;
Hsiao, C.; Akopian, T.; Hum, W. T.; Malakian, K.; Wolfrom, S.; Bapat, A.; Bhat, R.
A.; Stahl, M. L.; Somers, W. S.; Alvarez, J. C. J. Am. Chem. Soc. 2004, 126, 15106.
15. Schopfer, U.; Schoeffer, P.; Bischoff, S. F.; Nozulak, J.; Feuerbach, D.; Floersheim,
P. J. Med. Chem. 2002, 45, 1399.
25. Kuiper, G. G. J. M.; Lemmen, J. G.; Carlsson, B.; Corton, J. C.; Safe, S. H.; van der
Saag, P. T.; van der Burg, B.; Gustafsson, J. Å. Endocrinology 1998, 139, 4252.
26. Honkakoski, P.; Palvimo, J. J.; Penttilä, L.; Vepsäläinen, J.; Auriola, S. Biochem.
Pharmacol. 2004, 67, 97.
27. Thompson, J.; Saatcioglu, F.; Jänne, O. A.; Palvimo, J. J. Mol. Endocrinol. 2001, 15,
923.
28. Tocchini-Valentini, G.; Rochel, N.; Wurtz, A.; Mitschler, A.; Moras, D. Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 5491.
29. Yamamoto, K.; Inaba, Y.; Yoshimoto, N.; Choi, M.; DeLuca, H. F.; Yamada, S. J.
Med. Chem. 2007, 50, 932.
30. Hourai, S.; Rodrigues, L. C.; Antony, P.; Reina-San-Martin, B.; Ciesielski, F.;
Magnier, B. C.; Schoonjans, K.; Mouriño, A.; Rcchel, N.; Moras, D. Chem. Biol.
2008, 15, 383.
31. Ruff, M.; Gangloff, M.; Wurtz, J. M.; Moras, D. Breast Cancer Res. 2000, 2, 353.
32. Mäkinen, J.; Frank, C.; Jyrkkärinne, J.; Gynther, J.; Carlberg, C.; Honkakoski, P.
Mol. Pharmacol. 2002, 62, 366.
16. Yang, C.; Xu, G.; Li, J.; Wu, X.; Liu, B.; Yan, X.; Wang, M.; Xie, Y. Bioorg. Med.
Chem. Lett. 2005, 15, 1505.
33. Kotaja, N.; Aittomäki, S.; Silvennoinen, O.; Palvimo, J. J.; Jänne, O. A. Mol.
Endocrinol. 2000, 14, 1986.
17. Blizzard, T. A.; Gude, C.; Morgan, J. D.; Chan, W.; Birzin, E. T.; Mojena, M.;
Tudela, C.; Chen, F.; Knecht, K.; Su, Q.; Kraker, B.; Mosley, R. T.; Holmes, M. A.;
Rohrer, S. P.; Hammond, M. L. Bioorg. Med. Chem. Lett. 2007, 17, 6295.
18. Waibel, M.; De Angelis, M.; Stossi, F.; Kieser, K. J.; Carlson, K. E.;
Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Eur. J. Med. Chem. 2009, 44,
3412.
19. Zhou, H.-B.; Carlson, K. E.; Stossi, F.; Katzenellenbogen, B. S.; Katzenellenbogen,
J. A. Bioorg. Med. Chem. Lett. 2009, 19, 108.
20. Smith, H. M.; Knox, A. S.; Zisterer, D. M.; Lloyd, D. G.; Meegan, M. J. Med. Chem.
2007, 3, 135.
34. Manas, E. S.; Xu, Z. B.; Unwalla, R. J.; Somers, W. S. Structure 2004, 12, 2197.
35. Glide, version 4.5, Schrödinger, LLC, New York, NY, 2007.
36. Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.; Greenwood, J. R.;
Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T. J. Med. Chem. 2006, 49, 6177.
37. Case, D. A.; Darden, T. A.; Cheatham, T. E. III; Simmerling, C. L.; Wang, J.; Duke,
R. E.; Luo, R.; Merz, K. M.; Pearlman, D. A.; Crowley, M.; Walker, R. C.; Zhang,
W.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.; Wong, K. F.; Paesani, F.; Wu, X.;
Brozell, S.; Tsui, V.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.; Hornak, V.; Cui, G.;
Beroza, P.; Mathews, H. D.; Schafmeister, C.; Ross, W. S.; Kollman, P.A. AMBER
9, University of California, San Francisco, CA, 2006.
21. Henke, B. R.; Consler, T. G.; Go, N.; Hale, R. L.; Hohman, D. R.; Jones, S. A.; Lu, A.
T.; Moore, L. B.; Moore, J. T.; Orband-Miller, L. A.; Robinett, R. G.; Shearin, J.;
Spearing, P. K.; Stewart, E. L.; Turnbull, P. S.; Weaver, S. L.; Williams, S. P.;
Wisely, G. B.; Lambert, M. H. J. Med. Chem. 2002, 45, 5492.
22. Ball, L. J.; Levy, N.; Zhao, X.; Griffin, C.; Tagliaferri, M.; Cohen, I.; Ricke, W. A.;
Speed, T. P.; Firestone, G. L.; Leitman, D. C. Mol. Cell. Endocrinol. 2009, 299, 204.
23. Pulkkinen, J. T.; Honkakoski, P.; Peräkylä, M.; Berczi, I.; Laatikainen, R. J. Med.
Chem. 2008, 51, 3562.
38. Hornak, V.; Abel, R.; Okur, A.; Strockbine, B.; Roitberg, A.; Simmerling, C.
Proteins 2006, 65, 712.
39. Wang, J.; Cieplak, P.; Kollman, P. A. J. Comput. Chem. 2000, 21, 1049.
40. Kuhn, B.; Kollman, P. A. J. Med. Chem. 2000, 43, 3786.
41. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong, L.; Lee, M.; Lee,
T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D. A.;
Cheatham, T. E. Acc. Chem. Res. 2000, 33, 889.
42. Kollman, P. A. Chem. Rev. 1993, 93, 2395.
24. Pulkkinen, J. T.; Vepsäläinen, J. J. J. Org. Chem. 1996, 61, 8604.