Influence of chlorine or fluorine substitution on the estrogenic properties of 1-alkyl-2,3,5-tris(4-hydroxyphenyl)-1H-pyrroles

Article abstract of DOI:10.1021/jm300860j

In continuation of our previous work, several 1-alkyl-2,3,5-tris(4- hydroxyphenyl)aryl-1H-pyrroles with chlorine or fluorine substituents in the aryl residues were synthesized and tested for estrogen receptor (ER) binding at isolated ERα/ERβ receptors (HAP assay) and in transactivation assays using ERα-positive MCF-7/2a as well as U2-OS/ERα and U2-OS/ERβ cells. In the competition experiment at ERα the compounds displayed very high relative binding affinities of up to 37% (determined for 8m) but with restricted subtype selectivity (e.g., ERα/ERβ (8m) = 9). The highest estrogenic potency in ERα-positive MCF-7/2a cells was determined for 2,3,5-tris(2-fluoro-4-hydroxyphenyl)-1-propyl-1H-pyrrole 8m (EC₅₀ = 23 nM), while in U2-OS/ERα cells 2-(2-fluoro-4-hydroxyphenyl)-3,5-bis(4- hydroxyphenyl)-1-propyl-1H-pyrrole 8b (EC₅₀ = 0.12 nM) was the most potent agonist, only 30-fold less active than estradiol (E2, EC₅₀ = 0.004 nM). In U2-OS/ERβ cells for all pyrroles no transactivation could be observed, which indicates that they are selective ERα agonists in cellular systems.

Full text of DOI:10.1021/jm300860j

Article
pubs.acs.org/jmc
Influence of Chlorine or Fluorine Substitution on the Estrogenic
Properties of 1‑Alkyl-2,3,5-tris(4-hydroxyphenyl)‑1H‑pyrroles
Anja Schafer,† Anja Wellner,† Martin Strauss,† Andreas Schafer,‡ Gerhard Wolber,† and Ronald Gust*,†,§
̈
̈
†Institute of Pharmacy, Freie Universitat Berlin, Konigin-Luise-Strasse 2+4, D-14195 Berlin, Germany
̈
̈
‡Institute of Chemistry and Biochemistry, Freie Universitat Berlin, Takustrasse 3, D-14195 Berlin, Germany
̈
§Department of Pharmaceutical Chemistry, Institute of Pharmacy, Center for Molecular Biosciences Innsbruck, CCBCentrum for
Chemistry and Biomedicine, University of Innsbruck, Innrain 80/82, A-6020 Innsbruck, Austria
S
* Supporting Information
ABSTRACT: In continuation of our previous work, several 1-
alkyl-2,3,5-tris(4-hydroxyphenyl)aryl-1H-pyrroles with chlor-
ine or fluorine substituents in the aryl residues were
synthesized and tested for estrogen receptor (ER) binding at
isolated ERα/ERβ receptors (HAP assay) and in trans-
activation assays using ERα-positive MCF-7/2a as well as
U2-OS/ERα and U2-OS/ERβ cells. In the competition
experiment at ERα the compounds displayed very high
relative binding affinities of up to 37% (determined for 8m)
but with restricted subtype selectivity (e.g., ERα/ERβ (8m) =
9). The highest estrogenic potency in ERα-positive MCF-7/2a
cells was determined for 2,3,5-tris(2-fluoro-4-hydroxyphenyl)-1-propyl-1H-pyrrole 8m (EC50 = 23 nM), while in U2-OS/ERα
cells 2-(2-fluoro-4-hydroxyphenyl)-3,5-bis(4-hydroxyphenyl)-1-propyl-1H-pyrrole 8b (EC50 = 0.12 nM) was the most potent
agonist, only 30-fold less active than estradiol (E2, EC50 = 0.004 nM). In U2-OS/ERβ cells for all pyrroles no transactivation
could be observed, which indicates that they are selective ERα agonists in cellular systems.
Using propylpyrazole triol (PPT; for structural formulas see
Figure 1) as lead structure,10,11 we developed triarylpyr-
roles12−14 as selective ERα agonists. Selected compounds are
depicted in Figure 1. Unfortunately, type A pyrroles showed
high susceptibility to oxidation reactions with only limited half-
life for pharmacological studies (see below).12 In contrast, the
type B pyrroles displayed sufficient stability to study ER
interaction.13 These compounds demonstrated high selectivity
for ERα depending on alkyl chains at the heterocyclic core.
In this part of our structure−activity study, we try to optimize
the hormonal effects by introduction of chlorine or fluorine
substituents in the aromatic rings. All compounds were
evaluated for ER binding in a competition experiment with
radiolabeled estradiol ([3H]E2) as well as in transactivation
experiments.
INTRODUCTION
■
As estrogen receptors (ERα and ERβ) play an important role in
the growth of hormone-dependent breast cancer, selective
estrogen receptor modulators (SERMs) such as tamoxifen
(TAM, active metabolite 4-OHT; see Figure 1) or raloxifene
(RAL), which act at the breast as anti-estrogens, are suitable
drugs for the treatment of this disease. Unfortunately, only a
very limited number of related SERMs are available for
therapeutical use, and new drugs (e.g., lasofoxifene and
bazedoxifene1−3), apart from only being approved for the
treatment of osteoporosis, possess merely minor differences in
their activities and effects compared to TAM or RAL.
Immunohistochemical analyses suggest that ERβ may have a
dominant role in the healthy mammary gland, and its expression
significantly decreases in cancer and metastatic lymph node
tissues.4−7 ERβ appears to suppress the function of ERα through
different mechanisms. Overexpression of ERα is frequently
observed in the early stage of breast cancer8 and is accompanied
by increased cell proliferation and tumor growth.9
This finding opens up the possibility of designing drugs that
distinguish between malignant and nonmalignant breast tissue.
If an ERα selective compound could be further equipped with
an antagonistic hormonal profile, it could be an effective drug for
the treatment of the mammary carcinoma. Therefore, as a first
step we created molecules that bind exclusively to ERα to obtain
a deeper insight into the requirements for ERα selectivity and
estrogenic action.
SYNTHESIS
■
The 2,3,5-triarylpyrroles were obtained following a synthetic
route described earlier.13,15
As it was not commercially available, precursor 3a had to be
synthesized by reacting chloroform with 2-fluoro-4-methoxy-
benzaldehyde in DMF under basic conditions to form the α-
trichloromethylcarbinol 1 (Scheme 1).16
Received: June 19, 2012
Published: October 8, 2012
© 2012 American Chemical Society
9607
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
Figure 1. Ligands of the ER: 4-hydroxytamoxifen (4-OHT), raloxifene (RAL), propylpyrazole triol (PPT),11 and type A (1,2,4-triaryl-3-alkylpyrroles)
and type B (2,3,5-triaryl-1-alkylpyrroles) pyrroles.12,13
a
Scheme 1. Synthesis of Precursor 3a
a
Conditions: (a) −9 °C, dry DMF, KOH, dry methanol, 3 h, toluene, 1 N HCl; (b) dry ethanol, diphenyldiselenide, NaBH4, NaOH, 40 °C, 36 h, 1
N HCl; (c) SOCl2, reflux gently, 1 h.
In the subsequent reaction, the phenylseleno(triethyl)borate
complex was generated from diphenyl diselenide and NaBH4 in
ethanol. The phenylseleno moiety then acts as a nucleophile at
the dichloroepoxide prevalent in basic solution to form the α-
phenylselenylacetyl chloride as an intermediate. The isolation of
diphenyldiselenide after acidification points to the participation
of an additional equivalent of PhSeB(OEt)3− on the elimination
of phenylselenide, yielding the phenylacetic acid 2.17 In the final
step 2 was converted to the acid chloride 3a by warming in
thionyl chloride.
In a Friedel−Crafts acylation the acid chlorides 3a,b and the
respective anisoles (Scheme 2, method d) gave the desoxy-
benzoines 4a−e which were then reacted with α-bromoketones
5a,b,d to obtain the 1,2,4-tris(4-methoxyphenyl)butane-1,4-
diones 6a−c,e−h (Scheme 2, method e). Condensation of the
latter with the appropriate primary amine (→7a−m; see
Scheme 2, method f) and ether cleavage with BBr3 (Scheme
2, method g) yielded the pyrroles 8a−m.
RESULTS
■
Stability. The stability of the new compounds in aqueous
solution was studied exemplarily on 8b, 8h, and 8m by HPLC
analysis (RP18 column, mobile phase of methanol/H2O = 7/3,
detection 265 nm). Stock solutions of compounds in methanol
were diluted with PBS (pH 7.3) to a final concentration of 10−5
M and kept at 37 °C (for details see Experimental Section). As
expected from our earlier studies,12,13 the −I effect of
substituents on the phenol rings at C2 and C5 and an alkyl
chain at N1 increased the stability. During 195 h, 8b (t1/2 = 745
h), 8h (t1/2 = 1474 h), and 8m (t1/2 = 1055 h) degraded only
9608
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Fluorine- or Chlorine-Substituted Type B 2,3,5-Triarylpyrroles
a
Conditions: (d) dry DCM, 0 °C, AlCl3; (e) dry THF, −78 °C, KHMDS solution; (f) dry ethanol, p-toluenesulfonic acid, reflux, appropriate amine,
3 h; (g) dry DCM, BBr3, 0 °C.
marginally (see Figure 2). This is a clear improvement
axis, resulting in a distereomeric splitting of the N−CH2 protons
in the NMR spectra of 8b, 8f, 8l, and 8m.
The signals of the N−CH2 protons in the H NMR of 8b
compared to compounds 5c (t1/2 = 256 h)13 and 6d (t1/2
=
169 h)12 (for formulas see Table 1).
1
measured at room temperature are depicted in Figure 3 (J = 7.3
Hz, Δν = 28 Hz). Heating the NMR solution to 409 K fused the
signals to the expected triplet. From the coalescence temper-
ature Tc = 353 5 K the rotational barrier of the aromatic ring
was calculated using the Eyring equation.18 The Gibbs free
energy for rotation of ΔG⧧c = 17.74 0.25 kcal/mol was below
the 25 kcal/mol necessary for separation of rotational isomers at
room temperature.19 The data obtained indicate that 8b would
have to be cooled to approximately −40 °C to obtain
atropisomers which would have a half-life of 6 h, allowing a
chromatographic separation and subsequent NMR character-
ization.
Rotational Isomers. Halides (F or Cl) in ortho-position on
the aromatic rings hinder the free rotation around the sp2−sp2
In order to investigate if the rotational isomers are of
relevance for in vitro studies, the measurements were performed
at 310 K (see Figure S1 in Supporting Information). Under
these conditions the rotational isomers exchange within 0.2 s, so
their existence is of no importance for cell culture experiments.
Relative Binding Affinity (RBA). All compounds were
tested in a competition experiment (hydroxylapatite (HAP)
assay) using radiolabeled estradiol ([3H]E2) and hERα or hERβ
(for details see Experimental Section). 2,3,5-Tris(4-hydrox-
yphenyl)-1-propyl-1H-pyrrole 5c13 and 1,2,4-tris(4-hydroxy-
phenyl)-3-propyl-1H-pyrrole 6d12 displaced E2 from the LBD
of ERα with RBAs of 4.9% and 1.85%, respectively. At ERβ no
Figure 2. Percentage of original peak area over time diagram for
compounds 8b, 8h, and 8m (for HPLC method see Experimental
Section). Values reflect SD of three determinations.
9609
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
The RBA values of the monofluorinated compounds
increased in the series R2 = F (8d, 13.0%) < R3 = F (8b,
22.6%) < R4 = F (8i, 36%). An additional F substituent can only
increase the RBA of 8d but not that of 8b. The pyrrole 8l
(23.3%) showed nearly the same affinity as 8b. Furthermore, the
RBA of 37% calculated for 8m demonstrates that the fluorine at
R4 is of highest significance for the binding to ERα (see Figure
4).
Estrogenic Activity in MCF-7 Cells. In order to evaluate
the consequence of ER binding, transcription activation in ERα-
positive MCF-7 cells stably transfected with the reporter
plasmid EREwtcluc (MCF-7/2a cells)20,21 (Figure 5 and Table
2) was evaluated22 and compared with those of the endogenous
ligand E2.
In MCF-7/2a cells, the pyrroles 8b, 8f, 8l, 8m and also PPT
reached an intrinsic activity (IA) of 100% at 1 μM, while 8i and
8k were partial agonists with IAs of 50% and 78%, respectively.
Compounds 8a, 8c, 8d, and 8e showed IAs between 40% and
70%. Their curves are characterized by a decrease of IA at higher
concentrations. Such effects were also observed for 8b, 8f, and
8l. 8g and 8h were completely inactive. Because partial agonists
may have an antagonistic profile, the compounds were tested for
antagonistic activity. However, none of them could antagonize
the effect of 10−9 M E2 (data not shown).
The calculated EC50 values indicated that higher effects were
achieved with compounds bearing a propyl chain at the
nitrogen. Halides in the C2-standing phenolic ring of the
propyl derivative 5c (→8b, 8f) did not change the hormonal
potency (EC50 = 50−70 nM), while 2-F substituents in the C3-
or C5-aryl ring (8i, 8l, and 8m) increased the hormonal potency
to EC50 = 23−30 nM. These data obtained with the ERα-
positive MCF-7/2a cell line correlated well with the RBA values
determined in the HAP assay using hERα.
A further indication of hormonal activity is the regulation of
the ER level in MCF-7 cells. With the exception of 8i, which is a
partial agonist (see Figure 5), the compounds with profiles of
full agonists and the highest RBA values caused a clear down-
regulation of the estrogen receptor (10 μM; see Figure S2).
Their influence decreased in the series 8m > 8l > 8f > 8b. These
compounds were more active than PPT at 10 μM and E2 at 1
μM. A depletion of ERα comparable to those of the reference
substances E2 and PPT was detected for 8d, 8e, 8h, 8i, and 8k,
while 8a, 8c, and 8g had no influence on the ERα level.
In order to verify the binding of the triarylpyrroles in the
ligand binding domain (LBD) of ERα, compounds 6d, 5c, and
8l (1 μM), respectively, were co-incubated with 4-OHT (0.1
μM) in MCF-7/2a cells. As outlined in Figure 6, 4-OHT
severely decreased the capacity of 6d, 5c, and 8l to induce
luciferase expression, ascertaining that ERα was involved in the
transactivation process. In addition, the compounds were added
to the cells 1 and 3 h before addition of 4-OHT. If the pyrroles
were added to the cells 1 or 3 h before 4-OHT, a slight tendency
of increasing activity could be observed.
Figure 3. Temperature dependent 1H NMR spectra (400 MHz) of 8b
in DMSO-d6: partial NMR spectra at 299, 349, 379, and 409 K: (left)
observed spectra of the protons in the N-CH2- group (see arrows in
formula); (right) computer line shape simulation obtained with the rate
constants (k) indicated.
measurable RBA could be observed (Table 1). An additional N-
atom at position 5 of 6d, resulting in the reference compound
PPT, increased the binding affinity to ERα (10.1%) and ERβ
(0.5%) with an ERα/β ratio of 20.2.
Halogen substitution in the aromatic rings of the type B
pyrroles increased the affinity not only to ERα but also to ERβ.
8m and 8i displayed the highest RBA to ERα (37% and 36%,
respectively) with an α/β-ratio of 9 and 4.7.
From the data listed in Table 1, the structural requirements
for high ER binding can be deduced. All pyrroles bearing a
propyl chain bind better to ERα and ERβ than their respective
methyl chain bearing congeners, indicating stronger hydro-
phobic contacts of the alkyl chain in the LBD.
Substitution of the aromatic rings with chlorine or fluorine
increased the RBA values at both ER. In the N-methyl series F
and Cl containing congeners bind with comparable affinity,
while their N-propyl derivatives showed a higher affinity to ERα
if they are F-substituted. At ERβ the F and Cl substitution
caused comparable effects (compare 8b and 8f). Therefore, we
refrained from the synthesis of further Cl-containing type B
pyrroles.
Transactivation in U2-OS Cells. To demonstrate ER-
subtype selective transactivation, all compounds were tested in
U2-OS/α and U2-OS/β cells transiently transfected with the
plasmid pSG5-ERα or pSG5-ERβ FL (Table 2) and the reporter
plasmid p(ERE)2-luc+.
With U2-OS/α cells a clear concentration−activity relation is
given (no decrease of activation at 10 μM), so IAs and EC50
values could be exactly evaluated: 8a, 8c, 8d, 8e, 8f, 8k, and 8l
were full agonists; the other compounds were partial agonists
with an IA between 46% and 79% (1 μM). Interestingly, 8g and
The position but not the number of F substituents in the
2,3,5-tris(4-hydroxyphenyl)-1-propyl-1H-pyrrole is of high
importance for the binding to ERα. The RBAs at ERβ remain,
with exception of 8i (RBA= 7.6%), nearly unchanged at about
RBA ≈ 4.
9610
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
a
Table 1. Binding Affinity for ERα and ERβ of 8a−m, 5c, 6d, PPT, and E2
HAP assay: rel binding affinity (%)
compd
R1
R2
R3
R4
ERα
ERβ
ERα/ERβ
b
5c
Pr
H
H
H
F
H
F
H
H
H
H
H
H
H
H
H
F
4.9
<0.01
<0.3
3.7
>490
>6.7
6.1
1.8
2.9
3.3
3.7
2.3
3.8
4.7
6.5
8.3
9
8a
8b
8c
8d
8e
8f
Me
Pr
2.0
F
22.6
3.2
Me
Pr
H
H
Cl
Cl
H
H
H
F
1.8
F
13.0
3.9
4.5
Me
Pr
H
H
Cl
Cl
H
F
1.2
11.5
3.9
3.1
8g
8h
8i
Me
Pr
1.7
9.0
2.4
Pr
36.0
9.1
7.6
8k
8l
Me
Pr
H
H
F
1.4
F
F
23.3
37.0
1.85
10.1
100
2.8
8m
Pr
F
F
4.1
b
6d
na
na
na
na
na
na
na
na
na
na
<0.01
0.5
>185
20.2
1
PPT
E2
na
na
100
a
b
HAP assay: mean value of independent determinations in duplicate. Results are reproducible within 20%. na: not applicable. For comparison, see
refs 12 and 13.
8h, which were inactive at MCF-7/2a cells, activated the
receptor in U2-OS/α cells with EC50 of 100 and 47 nM,
respectively. Once again, the N-propyl derivatives were more
active than their methyl-substituted analogues. Introduction of a
2-F substituent in the C2-aryl ring of 5c increased the hormonal
potency from EC50 = 0.4 nM to EC50 = 0.12 nM (8b). The same
structural modification at C5 (8d) decreased the EC50 to 0.67
nM and at C3 (8i) to EC50 = 0.84 nM. The 2-Cl congener of 8d
was less active (8h, EC50 = 47 nM) and indicated the high
importance of fluorine for the induction of transcriptional
activity. Interestingly, it was impossible to strengthen the
hormonal profile of the pyrroles by simultaneous F-substitution
of the phenolic rings. 8b is equipotent to PPT and the most
potent compound in this transactivation assay.
Generally, the compounds were more active in U2-OS/α cells
compared to MCF-7/2a cells (naturally containing ERα). For
instance, 8b displayed an EC50 in U2-OS/α cells (EC50 = 0.12
nM) comparable to that of PPT (0.1 nM). In MCF-7/2a cells, it
was more than 20 times less active than in PPT (EC50 of 70 and
3 nM, respectively).
All compounds demonstrated selectivity for ERα. In the
experiments using U2-OS/β cells maximum IAs of about 15%
were measured (Table 2). Compounds with an RBA higher than
10% at hERα resulted in EC50 values below 1 nM in U2-OS/α
cells.
Inhibition of Cell Growth. The reduction of the IA at the
highest concentration in the luciferase assay with MCF-7/2a
cells induced us to examine if it was due to the inhibition of cell
growth. All compounds were tested in hormone-dependent
MCF-7 and hormone-independent MDA-MB-231 cells. The N-
methyl derivatives showed very low cytotoxicity in MCF-7 cells
(IC50 > 20 μM; see Table S1) comparable to that of PPT (IC50
= 27.2 μM). Elongation of the N-alkyl chain and introduction of
fluorine substituents in the aromatic rings increased the cytoxic
activity. 2,3,5-Tris(2-fluoro-4-hydroxyphenyl)-1-propyl-1H-pyr-
role 8m caused an IC50 of 5.6 μM very similar to that of E2. This
fact and the growth inhibitory effects in the hormone-
independent MDA-MB-231 cell line point to an unspecific
cytotoxicity. Nevertheless the tested type B pyrroles influence
the growth of the MCF-7 cells to a lesser extent than E2 or DES.
For instance, none of the compounds stimulate the proliferation
of the MCF-7 cells in low concentrations (an example for 8m,
8k is given in Figure S4).
The reason for this finding is still unclear. However, it may be
possible that the endogenous ERα signal cascade in MCF-7/2a
cells might respond differently to various ligands. Another
explanation could be that because of the overexpression of the
ER in U2-OS cells, even very weak agonists such as 8g and 8h
elicit a response.23 Interestingly, the lack of activity in MCF-7/
2a cells was also observed for imidazoles and other pyrroles with
an analogue substitution pattern to 8g and 8h.14,24
3D Modeling and Structure−Activity Relationships
(SARs). In order to rationalize the SAR of the series of
compounds presented in this study, CCDCs software package
GOLD25 was used to dock compounds 8a−m and PPT11 into
different crystal structures of ERα and ERβ. The resulting
9611
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
Figure 4. Comparison of the relative binding affinities of 5c and F-substituted 8b, 8d, 8i, 8l, and 8m.
binding poses were prioritized according to the proposed
pharmacophore of PPT within ERα described previously.11,13,33
Our computational studies suggest that the R1 alkyl
substituent should not be too bulky, while the size affects
selectivity: The propyl moiety at R1 perfectly complements the
cone-shaped cavity of ERα (as in the case of PPT),11 while the
narrower binding pocket of ERβ is limited to methyl
substituents. Compounds 8a−m are able to bind in orientations
comparable to that of PPT (see Figures 7 and S5),13 while no
significant differences in orientation were observed. Overall, the
N-propyl substituted compounds show higher activity on both
estrogen receptor subtypes, mainly because of their increased
lipophilicity (see above).
Figure S5 illustrates representative docking poses for 8b, 8f, 8l
and PPT in ERα (PDB code 3ERD28). Our docking
experiments show that binding mode C(5)′ is preferred in
3ERD: The hydroxyl group of the phenyl moiety at position 5
presumably forms hydrogen bonds to Glu353 and Arg394. We
assume that the hydroxyl group in the aryl residue in C2
interacts with Met343 and His524 and the one in C3 with
Gly521 and Met522. The fluorine substituent at R3 may interact
in a hydrogen-bond-like way with the nitrogen of Thr347,
although the geometry for this interaction is not ideal and
therefore not observed in the docking experiments.
An alternative possible orientation following the scheme of
C(5) when docked into 3L0329 is shown in Figure 8 (8l and 8m
as representative poses). The phenol moiety in position 5 shows
hydrogen bonds with Glu353 and Arg394, while the phenol
group at position 2 interacts with His524. In position 3 an
interaction with the backbone at Met421 and Ile424 is formed
corresponding to orientation C(5)′. The alkyl chain at R3
interacts with the hydrophobic cavity consisting of Leu346,
Thr347, Ala350, Trp383, Leu384, Leu387, and Leu540 (not all
residues shown to improve clarity).
Fluorine substituents at R2, R3, and R4 lead to an increase of
activity (see above) at ERα because of increased lipophilic
contact with the inner area of the estrogen receptor.
The binding mode of halogen-substituted pyrroles within
ERα is ambiguous. While the pose presented in Figure S5 shows
the binding pose analogue in Stauffer et al.,11 docking
experiments using different crystal structures (ERα, 2QZO26
and 3L03;29 ERβ, 2NV7,30 3OLL,32 3OLS32) revealed a
different plausible binding mode. 2,3,5-Triaryl-1H-pyrroles
have three hydroxyl groups in nearly equal distances, so six
orientations in the LBD are possible, as shown in Figure 7.
9612
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
Figure 6. Co-incubation of 6d, 5c, or 8l (1 μM) with 4-OHT (0.1 μM)
in MCF-7/2a cells. Shown are mean values of two independent
determinations, each performed in quadruplicate, SD. Control is
vehicle, with 0% relative luciferase expression. Compounds (1 μM)
were added to the cells without or simultaneously with 4-OHT or at 1
and 3 h before administration of 4-OHT.
substituted pyrroles. At isolated receptors, the relative binding
affinity to ERα and ERβ increased if halides are substituted in
the ortho-position of the phenolic rings. Contrary to our earlier
studies with hexestrol derivatives (e.g., 1,2-diamino-1,2-diaryl-
ethanes,34 2,3-diarylpiperazines,35 and 4,5-diaryl-2-imidazo-
lines),36 in the class of type B pyrroles fluorine substituents
caused a higher binding affinity than chlorine substituents. In the
competition experiment with hERα the highest RBA values were
obtained with R4 = F (the influence of F-substitution is
summarized in Figure 4). The ERα/ERβ ratio of all compounds
was in the range 2−9.
Of high importance is the finding that the RBA values do not
reflect the situation in cellular systems. Especially the ER
selectivity is increased in cells. All 2,3,5-triarylpyrroles showed
transactivation exclusively at ERα if the receptor is expressed in
U2-OS cells. In the same experiment with ERβ no gene
activation was determined. An explanation cannot be given at
the moment. But it is well-known that for transcriptional activity
a complex machinery has to be started. After ligand binding and
receptor dimerization, the dimers translocate into the nucleus
and bind to specific DNA sequences, the estrogen response
elements. Then the binding of co-regulators starts the gene
transcription. We assume that the binding of pyrroles to ERβ
either caused a conformation that is not suitable for dimerization
or prevented co-regulator binding. Especially the latter will be
part of a new SAR study.
2,3,5-Triarylpyrroles caused growth inhibitory effects in
tumor cells. The effect is observed in hormone-dependent
MCF-7 as well as in MDA-MB-231 cells, so we assume an
unselective mode of action. All compounds reduced the cell
growth in MCF-7 cells to a lesser extent than DES and E2. Some
of the compounds are even less cytotoxic than PPT. All 2,3,5-
triarylpyrroles did not stimulate the cell growth in low
concentration but influence the ERα content. The hormonally
most active compounds 8m and 8l were also potent ER down-
regulators in MCF-7 cells. In this case the growth stimulating ER
system was destabilized causing reduced cell growth. As a
consequence, 8m and 8l were more cytotoxic in hormone-
dependent MCF 7 cells (IC50 = 5.6 and 8.0 μM, Table S1) than
against hormone-independent MDA-MB-231 cells (IC50 = 21.9
and 20.9 μM, Table S1)
Figure 5. Top: Estrogenic activities of compounds 8a−d and PPT in
MCF-7/2a cells. Middle: Estrogenic activities of compounds 8e−h in
MCF-7/2a cells. Bottom: Estrogenic activities of compounds 8i−m in
MCF-7/2a cells. Diagrams show mean SD of a single representative
experiment performed in quadruplicate, with estradiol graph included
for comparison.
DISCUSSION
■
In continuation of our recently published SAR studies, we tried
to optimize the ER binding of 2,3,5-tris(4-hydroxyphenyl)-
9613
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
Table 2. Estrogenic Properties of 5c, 6d, 8a−m, PPT, and E2 in MCF-7/2a, U2-OS/α, and β Cells
MCF-7/2a cells
U2-OS/α cells
U2-OS/β cells
a
a
a
compd
R1
R2
R3
R4
IA (%)
EC50 (nM)
50
IA (%)
EC50 (nM)
IA (%)
EC50 (nM)
5c
8a
8b
8c
8d
8e
8f
Pr
H
H
H
F
H
F
H
H
H
H
H
H
H
H
H
F
65 12
61 11
55
0.4
16
12
16
15
15
12
10
26
13
0
Me
Pr
101
72
60
F
101
40
1
70
0.12
24
Me
Pr
H
H
Cl
Cl
H
H
H
F
9
1
7
86
F
71
240
600
63
91
0.67
16
Me
Pr
H
H
Cl
Cl
H
F
64
97
92 10
87
0.31
100
47
8g
8h
8i
Me
Pr
0
0
46
79
c
c
Pr
50
78
7
5
26
290
30
23
70
3
105
110
107
100
111
119
100
0.84
15
8k
8l
Me
Pr
H
H
F
5
F
F
88 14
94
99 18
0.27
0.2
12
0
8m
Pr
F
F
2
b
6d
na
na
na
na
na
na
na
na
na
na
1.6
0
PPT
E2
na
97
6
0.1
5
na
100
0.1
0.004
100
0.01
a
EC50 values of compounds: for MCF-7/2a cells, mean values of three independent determinations, each performed in quadruplicate, SD; for U2-
b
OS cells, mean values of three independent determinations, SD < 20%. na: not applicable. IA: intrinsic activity. For comparison, see refs 12 and 13.
For structural formulas, see Table 1. IA: intrinsic activity at 1 μM.
c
8.6, ArH), 7.07 (t, 1H, J = 8.8, ArH), 6.94 (d, 2H, J = 8.6, ArH), 6.84 (d,
2H, J = 8.6, ArH), 6.65 (m, 2H, ArH), 6.57 (d, 2H, J = 8.7, ArH), 6.26
(s, 1H, CH), 3.28 (s, 3H, NCH3). Anal. (C23H18FNO3·1.6H2O) C, H,
N.
EXPERIMENTAL SECTION
■
General. Melting points were determined on a Buchi B-545 and are
̈
uncorrected. NMR spectra were obtained with an Avance/DPX 400
(Bruker, Karlsruhe) 400 MHz spectrometer. Chemical shifts (δ) are
reported in parts per million downfield from tetramethylsilane and
referenced from solvent references. All aromatic d are “apparent”
doublets. Electron impact (EI, 70 eV) mass spectra were obtained on a
CH-7A (Varian MAT, Bremen) mass spectrometer. IR spectroscopy
was performed on an ATI Mattson Genesis (Wigan, GB, KBr disks) or
a Perkin-Elmer Spectrum 100. The elemental compositions of the
compounds 8a−m agreed to within 0.4% of the calculated values
(Vario EL, Elementar, Hanau). HPLC studies confirmed purity of
compounds 8a−m to be between 95% and 100% (Bio-Tek Kontron
Instruments, Neufahrn). Chromatographic separations were performed
on silica gel (Kieselgel 60, 0.063−0.100 mm; VWR, Bruchsal) columns
by flash chromatography. Reactions were followed by thin-layer
chromatography (TLC) on Merck aluminum silica gel (60 F254) sheets
that were visualized under a UV lamp.
Molecular Modeling and Docking Studies. Initial 3D
coordinates of the compounds were generated using the software
CORINA37 by Molecular Networks (Erlangen, Germany) and
subsequently docked using GOLD,25 version 4.11 (CCDC, U.K.).
Ensemble docking of compounds 8a−m, PPT, and E2 was conducted
using PDB complexes 3DT3,27 3ERD,28 3L03,29 2QZO26 (for ERα),
and 2NV7,30 2QTU,31 3OLL,32 3OLS,32 (for ERβ). For pose ranking
the PPT 3D pharmacophore was used in an analogue way as in a
previous study,38 following the already known 3D pharmacophore
pattern of PPT. Pharmacophore creation and visualization were
performed using the program LigandScout 3.02.39−41
2-(2-Fluoro-4-hydroxyphenyl)-3,5-bis(4-hydroxyphenyl)-1-
propyl-1H-pyrrole (8b). 8b was produced from 2-(2-fluoro-4-
methoxyphenyl)-3,5-bis(4-methoxyphenyl)-1-propyl-1H-pyrrole (7b),
0.40 mmol (178 mg), and BBr3, 2.0 mmol (500 mg). Brown solid, mp
106 °C. Yield: 0.31 mmol (126 mg), 71%. 1H NMR (DMSO-d6): δ =
10.14 (s, 1H, OH), 9.59 (s, 1H, OH), 9.17 (s, 1H, OH), 7.31 (d, 2H, J =
8.6, ArH), 7.16 (t, 1H, J = 8.8, ArH), 6.99 (d, 2H, J = 8.6, ArH), 6.89 (d,
2H, J = 8.5, ArH), 6.71 (m, 2H, ArH), 6.62 (d, 2H, J = 8.6, ArH), 6.28
(s, 1H, CH), 3.70 (m, 2H, NCH2CH2CH3), 1.24 (m, 2H,
NCH2CH2CH3), 0.52 (t, 3H, J = 7.4, NCH2CH2CH3). Anal.
(C25H22FNO3·2H2O) C, H, N.
5-(2-Fluoro-4-hydroxyphenyl)-2,3-bis(4-hydroxyphenyl)-1-
methyl-1H-pyrrole (8c). 8c was produced from 5-(2-fluoro-4-
methoxyphenyl)-2,3-bis(4-methoxyphenyl)-1-methyl-1H-pyrrole (7c),
0.72 mmol (300 mg), and BBr3, 3.59 mmol (900 mg). Brown solid, mp
137 °C. Yield: 0.59 mmol (220 mg), 82%. 1H NMR (DMSO-d6): δ =
10.07 (s, 1H, OH), 9.58 (s, 1H, OH), 9.12 (s, 1H, OH), 7.27 (t, 1H, J =
8.9, ArH), 7.09 (d, 2H, J = 8.6, ArH), 6.93 (d, 2H, J = 8.6, ArH), 6.80
(d, 2H, J = 8.6, ArH), 6.70 (m, 2H, ArH), 6.56 (d, 2H, J = 8.7, ArH),
6.22 (s, 1H, CH), 3.20 (s, 3H, NCH3). Anal. (C23H18FNO3·1.3H2O) C,
H, N.
5-(2-Fluoro-4-hydroxyphenyl)-2,3-bis(4-hydroxyphenyl)-1-
propyl-1H-pyrrole (8d). 8d was produced from 5-(2-fluoro-4-
methoxyphenyl)-2,3-bis(4-methoxyphenyl)-1-propyl-1H-pyrrole (7d),
0.44 mmol (194 mg), and BBr3, 2.2 mmol (554 mg). Brown solid, mp
1
100 °C. Yield: 0.22 mmol (90 mg), 50%. H NMR (DMSO-d6): δ =
9.80 (s, 1H, OH), 9.32 (s, 1H, OH), 8.84 (s, 1H, OH), 7.00 (t, 1H, J =
8.7, ArH), 6.84 (d, 2H, J = 8.5, ArH), 6.67 (d, 2H, J = 8.7, ArH), 6.56
(d, 2H, J = 8.6, ArH), 6.44 (m, 2H, ArH), 6.30 (d, 2H, J = 8.7, ArH),
5.94 (s, 1H, CH), 3.34 (t, 2H, J = 7.2, NCH2CH2CH3), 0.89 (sextet,
2H, J = 7.6, NCH2CH2CH3), 0.18 (t, 3H, J = 7.4, NCH2CH2CH3).
Anal. (C25H22FNO3·1.3H2O) C, H, N.
2-(2-Chloro-4-hydroxyphenyl)-3,5-bis(4-hydroxyphenyl)-1-
methyl-1H-pyrrole (8e). 8e was produced from 2-(2-chloro-4-
methoxyphenyl)-3,5-bis(4-methoxyphenyl)-1-methyl-1H-pyrrole (7e),
0.67 mmol (291 mg), and BBr3, 3.4 mmol (840 mg). Orange solid, mp
105 °C. Yield: 0.33 mmol (128 mg), 49%. 1H NMR (DMSO-d6): δ =
10.10 (s, 1H, OH), 9.53 (s, 1H, OH), 9.11 (s, 1H, OH), 7.30 (d, 2H, J =
8.6, ArH), 7.14 (d, 1H, J = 8.4, ArH), 6.99 (d, 1H, J = 2.4, ArH), 6.92
(d, 2H, J = 8.6, ArH), 6.85 (d, 2H, J = 8.6, ArH), 6.79 (dd, 1H, J = 8.4, J
Chemistry. General Method for Ether Cleavage (Method g,
Scheme 2). The respective 1,3,5-tris(4-methoxyphenyl)-1H-pyrrole
(8a−m) was dissolved in 15 mL of dry DCM under inert gas and
cooled to 0 °C. BBr3 dissolved in DCM (5 mL) was added slowly and
the reaction mixture stirred overnight at room temperature. After
evaporation of the solvent, redissolving several times in MeOH,
purification of the residue was performed on silica gel by column
chromatography using DCM/MeOH (9:1).
2-(2-Fluoro-4-hydroxyphenyl)-3,5-bis(4-hydroxyphenyl)-1-
methyl-1H-pyrrole (8a). 8a was produced from 2-(2-fluoro-4-
methoxyphenyl)-3,5-bis(4-methoxyphenyl)-1-methyl-1H-pyrrole (7a),
0.84 mmol (352 mg), and BBr3, 4.2 mmol (1056 mg). Brown solid, mp
158 °C. Yield: 0.28 mmol (106 mg), 34%. 1H NMR (DMSO-d6): δ =
10.05 (s, 1H, OH), 9.55 (s, 1H, OH), 9.14 (s, 1H, OH), 7.30 (d, 2H, J =
9614
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
Figure 7. Potential 2,3,5-triaryl-1H-pyrroles binding modes relative to E2 as 2D projection, oriented to project into the estradiol A-ring binding pocket
(red circle; Glu353, Arg394) while interacting with His524 (blue circle). By rotation of the pyrrole core around the lateral axis (darkened bond),
modes C and C′ are related to each other.11
= 2.5, ArH), 6.57 (d, 2H, J = 8.7, ArH),6.28 (s, 1H, CH), 3.24 (s, 3H,
NCH3). Anal. (C23H18ClNO3·1.3H2O) C, H, N.
ArH), 6.56 (d, 2H, J = 8.6, ArH), 6.19 (s, 1H, CH), 3.13 (s, 3H, NCH3).
Anal. (C23H18ClNO3·1H2O) C, H, N.
5-(2-Chloro-4-hydroxyphenyl)-2,3-bis(4-hydroxyphenyl)-1-
propyl-1H-pyrrole (8h). 8h was produced from 5-(2-chloro-4-
methoxyphenyl)-2,3-bis(4-methoxyphenyl)-1-propyl-1H-pyrrole (7h),
0.65 mmol (300 mg), and BBr3, 3.2 mmol (813 mg). After flash
chromatography, the compound is purified further by MPLC. Orange
2-(2-Chloro-4-hydroxyphenyl)-3,5-bis(4-hydroxyphenyl)-1-
propyl-1H-pyrrole (8f). 8f was produced from 2-(2-chloro-4-
methoxyphenyl)-3,5-bis(4-methoxyphenyl)-1-propyl-1H-pyrrole (7f),
0.44 mmol (220 mg), and BBr3, 2.1 mmol (550 mg). Orange solid,
mp 102 °C. Yield: 0.17 mmol (70 mg), 38%. 1H NMR (DMSO-d6): δ =
10.09 (s, 1H, OH), 9.52 (s, 1H, OH), 9.09 (s, 1H, OH), 7.25 (d, 2H, J =
8.5, ArH), 7.17 (d, 1H, J = 8.4, ArH), 6.95 (d, 1H, J = 2.4, ArH), 6.89
(d, 2H, J = 8.6, ArH), 6.83 (d, 2H, J = 8.6, ArH), 6.80 (dd, 1H, J = 8.4, J
= 2.5, ArH), 6.55 (d, 2H, J = 8.6, ArH), 6.22 (s, 1H, CH), 3.60 (m, 2H,
NCH2CH2CH3), 1.17 (m, 2H, NCH2CH2CH3), 0.46 (t, 3H, J = 7.4,
NCH2CH2CH3). Anal. (C25H22ClNO3·1.1H2O) C, H, N.
5-(2-Chloro-4-hydroxyphenyl)-2,3-bis(4-hydroxyphenyl)-1-
methyl-1H-pyrrole (8g). 8g was produced from 5-(2-chloro-4-
methoxyphenyl)-2,3-bis(4-methoxyphenyl)-1-methyl-1H-pyrrole (7g),
0.45 mmol (195 mg), and BBr3, 2.2 mmol (563 mg). Orange solid, mp
142 °C. Yield: 0.26 mmol (100 mg), 57%. 1H NMR (DMSO-d6): δ =
10.07 (s, 1H, OH), 9.57 (s, 1H, OH), 9.11 (s, 1H, OH), 7.29 (d, 1H, J =
8.4, ArH), 7.09 (d, 2H, J = 8.5, ArH), 6.94 (m, 3H, ArH), 6.81 (m, 3H,
1
solid, mp 123 °C. Yield: 0.05 mmol (20 mg), 7%. H NMR (DMSO-
d6): δ = 10.06 (s, 1H, OH), 9.58 (s, 1H, OH), 9.09 (s, 1H, OH), 7.29
(d, 1H, J = 8.4, ArH), 7.07 (d, 2H, J = 8.5, ArH), 6.94 (d, 1H, J = 2.4,
ArH), 6.92 (d, 2H, J = 8.6, ArH), 6.82 (m, 3H, ArH), 6.54 (d, 2H, J =
8.7, ArH), 6.16 (s, 1H, CH), 3.53 (t, 2H, J = 7.6, NCH2CH2CH3), 1.11
(sextet, 2H, J = 7.4, NCH2CH2CH3), 0.41 (t, 3H, J = 7.4,
NCH2CH2CH3). Anal. (C25H22ClNO3·1.5H2O) C, H, N.
3-(2-Fluoro-4-hydroxyphenyl)-2,5-bis(4-hydroxyphenyl)-1-
propyl-1H-pyrrole (8i). 8i was produced from 3-(2-fluoro-4-
methoxyphenyl)-2,5-bis(4-methoxyphenyl)-1-propyl-1H-pyrrole (7i),
0.22 mmol (100 mg), and BBr3, 1.0 mmol (253 mg). Yellow solid,
mp 201 °C. Yield: 0.13 mmol (53 mg), 60%. 1H NMR (DMSO-d6): δ =
9.58 (s, 1H, OH), 9.51 (s, 1H, OH), 9.50 (s, 1H, OH), 7.26 (d, J = 8.4,
2H, ArH), 7.04 (d, J = 8.4, 2H, ArH), 6.79 (m, 5H, ArH), 6.42 (dd, J =
9615
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
Figure 8. Presumed upside-down poses of 8l and 8m inside the cavity of ERα (PDB code 3L0323). Hydrogen bonds are shown as green and red
arrows, and lipophilic features are shown as yellow spheres. Font size for amino acid labels corresponds to position on the z-axis (depth).
26.4, 2.2, 1H, ArH), 6.39 (dd, J = 22.6, 2.2, 1H, ArH), 6.10 (d, J = 1.9,
1H, CH), 3.78 (t, J = 7.3, 2H, NCH2CH2CH3), 1.15 (m, 2H,
NCH2CH2CH3), 0.42 (t, J = 7.4, 3H, NCH2CH2CH3). Anal.
(C25H22FNO3·1H2O) C, H, N.
min. Compounds were dissolved in methanol (10−2 M) and diluted
with PBS to a final concentration of 10−5 M. Compounds were kept at
37 °C and were measured once directly after dissolving and six more
times during the following 195 h.
2,5-Bis(2-fluoro-4-hydroxyphenyl)-3-(4-hydroxyphenyl)-1-
methyl-1H-pyrrole (8k). 8k was produced from 2,5-bis(2-fluoro-4-
methoxyphenyl)-3-(4-methoxyphenyl)-1-methyl-1H-pyrrole (7k), 1.6
mmol (700 mg), and BBr3, 8.0 mmol (2010 mg). Orange solid, mp 126
°C. Yield: 0.89 mmol (350 mg), 55%. 1H NMR (DMSO-d6): δ = 10.10
(s, 1H, OH), 10.07 (s, 1H, OH), 9.14 (s, 1H, OH) 7.27 (t, 1H, J = 8.8,
ArH), 7.06 (t, 1H, J = 8.8, ArH), 6.94 (d, 2H, J = 8.5, ArH), 6.69 (m,
4H, ArH), 6.58 (d, 2H, J = 8.6, ArH), 6.27 (s, 1H, CH), 3.16 (s, 3H,
NCH3). Anal. (C23H17F2NO3·1.3H2O) C, H, N.
2,5-Bis(2-fluoro-4-hydroxyphenyl)-3-(4-hydroxyphenyl)-1-
propyl-1H-pyrrole (8l). 8l was produced from 2,5-bis(2-fluoro-4-
methoxyphenyl)-3-(4-methoxyphenyl)-1-propyl-1H-pyrrole (7l), 0.86
mmol (400 mg), and BBr3, 4.3 mmol (1080 mg). Orange solid, mp 115
°C. Yield: 0.83 mmol (350 mg), 96%. 1H NMR (DMSO-d6): δ = 10.18
(s, 1H, OH), 10.15 (s, 1H, OH), 9.21 (s, 1H, OH), 7.32 (t, 1H, J = 8.6,
ArH), 7.17 (t, 1H, J = 8.8, ArH), 7.01 (d, 2H, J = 8.5, ArH), 6.76 (m,
4H, ArH), 6.65 (m, 2H, J = 8.6, ArH), 6.33 (s, 1H, CH), 3.61 (m, 2H,
NCH2CH2CH3), 1.23 (sextet, 2H, J = 7,2, NCH2CH2CH3), 0.53 (t,
3H, J = 7.4, NCH2CH2CH3). Anal. (C25H21F2NO3·1.5H2O) C, H, N.
2,3,5-Tris(2-fluoro-4-hydroxyphenyl)-1-propyl-1H-pyrrole
(8m). 8m was produced from 2,3,5-tris(2-fluoro-4-methoxyphenyl)-1-
propyl-1H-pyrrole (7m), 0.15 mmol (72 mg), and BBr3, 0.67 mmol
(169 mg). Yellow solid, mp 105 °C. Yield: 0.06 mmol (26 mg), 40%.
1H NMR (DMSO-d6): δ = 10.10 (s, 1H, OH), 10.07 (s, 1H, OH), 9.66
(s, 1H, OH), 7.24 (d, 1H, J = 8.7, ArH), 7.01 (d, 1H, J = 8.7, ArH), 6.82
(d, 1H, J = 8.8, ArH), 6.66 (m, 4H, ArH), 6.43 (m, 2H, ArH), 6.17 (d,
1H, J = 1.8, CH), 3.59 (m, 2H, NCH2CH2CH3), 1.14 (m, 2H, J = 7.3,
NCH2CH2CH3), 0.43 (t, 3H, J = 7.4, NCH2CH2CH3). Anal.
(C25H20F3NO3·1H2O) C, H, N.
Stability Studies. The following was used: HPLC, Bio-Tek
Kontron Instruments, Germany; C18 reverse phase column, Eurospher
100-5 C18, 250 mm × 4 mm, Knauer GmbH, Germany. HPLC
conditions were the following: methanol/H2O, 7/3; isocratic elution;
detection wavelength, 265 nm; flow rate, 0.8 mL/min; run time, 20
Cell Lines and Growth Conditions. MCF-7/2a cells, MCF-7
cells, and MDA-MB-231 cells were maintained as a monolayer culture
at 37 °C in a humidified atmosphere (95% air, 5% CO2) in T-25 flasks.
U2-OS cells were maintained as a monolayer culture at 37 °C in
humidified atmosphere (92.5% air, 7.5% CO2) in T-25 flasks. Growth
medium for MCF-7/2a cells consisted of phenol red free DMEM high
glucose with sodium pyruvate (110 mg/L), ct-FCS 5%, and Geneticin
solution, 0.5%. Growth medium for U2-OS, MCF-7, and MDA-MB-
231 cells consisted of L-glutamine and sodium pyruvate containing
DMEM high glucose with FCS, 5%.
Estrogenic Activity. Luciferase Assay (MCF-7/2a Cells). Four
days before starting the experiment MCF-7/2a cells were cultivated in
DMEM supplemented with dextran−charcoal-treated FCS (ct-FCS, 50
mL/L). Cells from an 80% confluent monolayer were removed by
trypsinization and suspended to approximately 105 cells/mL in the
growth medium mentioned above. The cell suspension was then
cultivated in 96-well flat-bottomed plates at growing conditions (see
above). After 24 h, the test compounds were added to achieve
concentrations ranging from 10−5 to 10−11 M (estradiol from 10−7 to
10−13 M), and the plates were incubated for 50 h. Subsequently, 50 μL
of cell culture lysis reagent was added to each well. After 30 min of lysis
at room temperature, an amount of 50 μL of the Promega luciferase
assay reagent (Promega, Mannheim, Germany) was added. Lumines-
cence (in relative light unit, RLU) was measured for 10 s using a
luminometer. Estrogenic activity was expressed as percent activation of
a 3 × 10−9 M estradiol control (100%).
Luciferase Assay (U2-OS Cells). U2-OS cells were transferred to
DMEM supplemented with dextran−charcoal-treated FCS 24 h before
the experiment. Cells from an almost confluent monolayer were split
and seeded in 10 cm i.d. Petri dishes at 1 × 106 cells per dish at least 24
h prior to transfection. Transient transfection of the cells with 0.05 μg
of receptor plasmid pSG5-ERα or pSG5-ERβ FL and 5 μg of the
reporter plasmid p(ERE)2-luc+ was carried out after 6 h using 15 μL of
9616
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
FuGENE 6 (Roche Diagnostics, Mannheim, Germany) according to
the manufacturer’s instructions. After 18 h the cells were washed with
PBS and harvested by trypsinization and seeded in white 96-well plates
at 104 cells/well in 100 μL of ct-DMEM. After 3 h, the medium was
replaced by medium containing either E2 or test compounds in final
concentrations ranging from 0.1 pM to 10 nM (E2) or 0.1 nM to 10
mM (test compounds). After 18 h of incubation, the medium was
removed and cells were lyzed and luciferase activity assayed as
described above.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthesis and analytical data of all precursors, additional
spectroscopic data of target compounds 8a−m, results from
cytotoxicity tests, ER down-regulation, and binding pose from
the molecular modeling. This material is available free of charge
via the Internet at http://pubs.acs.org.
Ability of Compounds To Regulate ERα Levels in MCF-7
Cells. MCF-7 cells are plated in six-well plates (500 000 cells per well)
containing ct-DMEM. After 24 h of culture, medium is removed and
cells are exposed to increasing amounts of the investigated compounds
for 24 h in a fresh medium; control cells are maintained in culture
without any compound. Cells are then washed with Tris buffered saline
(50 mM Tris, pH 7.5, 150 mM NaCl) and lysed for 30 min at 4 °C in
radioimmunoprecipitation assay buffer. ERα levels in these lysates were
finally assessed by Western blotting following a procedure described in
ref 42. Briefly, ERα is detected with a mouse monoclonal antibody (F-
10, Santa Cruz Biotechnology, Santa Cruz, CA). Actin, used as a
loading control, is detected with mouse monoclonal antibody
MAB1501R from Millipore. Exposure to primary antibodies is followed
by incubation in the presence of a horseradish peroxidase-conjugated
goat anti-mouse secondary antibody (Pierce, Rockford, IL). Peroxidase
activity is revealed with Supersignal West Pico chemiluminescent
substrate from Pierce. ERα and actin bands are captured with a CCD
camera.
Recombinant ERα/ERβ (HAP Assay). An ethanolic solution (250
fmol/μL) of recombinant human estrogen receptor (hERα or hERβ,
Calbiochem) was diluted with Tris-HCl buffer, pH 8.0 (1:100). These
hERα or hERβ preparations adsorbed on hydroxyl apatite pellets
(prepared in 10 mmol/L of Tris-HCl buffer, pH 8.0, containing 1 mg/
mL BSA) which were labeled with increasing concentrations of the test
compound (10−10−10−5 M) and [3H]E2 (5 nM, Perkin-Elmer) and
were incubated for 3 h at room temperature. Bond [3H]E2 was then
successively extracted with ethanol, and aliquots of 200 μL of the
ethanolic extracts were transferred to scintillation vials containing 3.5
mL of scintillator Ecoscint H (National Diagnostics, Atlanta, GA) for
radioactivity counting. All measurements were performed in duplicate.
The results (relative radioactivity) were expressed as percentage of a
solvent treated control sample. Relative concentrations of compounds
required to reduce the binding of [3H]E2 by 50% gave their IC50 values.
The relative binding affinity was calculated by dividing the IC50 of E2 by
the IC50 of the respective compound. Binding affinities are expressed
relative to estradiol (RBA(E2) = 100%).
Determination of Inhibition of Cell Growth in MCF-7 and
MDA-MB-231 Human Breast Cancer Cells. Cells from an 80%
confluent monolayer were harvested by trypsinization and suspended
to approximately 104 (MCF-7 cells) or 7.5 × 103 cells/mL (MDA-MB-
231 cells). At the beginning of the experiment, the cell suspension was
transferred to 96-well microplates (100 μL/well). After their cultivation
for 3 days at growing conditions, the test compounds dissolved in
DMEM were added at final concentrations ranging from 3.13 to 50 μM.
Control wells (16/plate) contained 0.1% of DMSO, which was used for
the preparation of the stock solutions. The initial cell density was
determined by addition of glutaric dialdehyde to one plate (1% in PBS,
100 μL/well). After incubation for 3 days (MDA-MB-231 cells) or 4
days (MCF-7 cells), the medium was removed and glutaric dialdehyde
(1% in PBS, 100 μL/well) was added for fixation. After 30 min, the
solution of the aldehyde was decanted and an amount of 150 μL of
PBS/well was added. The plates were stored at 4 °C until staining. Cells
were stained by treating them for 30 min with 100 μL of an aqueous
solution of crystal violet (0.02%). After decanting, cells were washed
several times with water to remove the adherent dye. After addition of
180 μL of ethanol (70%), plates were gently shaken for 3−4 h. Optical
density of each well was measured in a microplate autoreader at 590
nm.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: +43 512 507 58200. Fax: +43 512 507 58299. E-mail:
ronald.gust@uibk.ac.at.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The technical assistance of Silke Bergemann and Maxi Wenzel is
acknowledged.
ABBREVIATIONS USED
■
E2, estradiol; (h)ER, (human) estrogen receptor; ERα, estrogen
receptor alpha; ERβ, estrogen receptor β; PPT, propylpyrazole
triol; TAM, tamoxifen; RAL, raloxifene; 4-OHT, 4-hydrox-
ytamoxifen; HAP, hydroxyl apatite; MCF-7, Michigan Cancer
Foundation 7; U2-OS, human osteosarcoma; [3H]E2, tritium
labeled estradiol; RBA, relative binding affinity; IA, intrinsic
activity; MDA-MB-231, M. D. Anderson metastatic breast
cancer; DMEM, Dulbecco’s modified Eagle medium; ct,
dextran-charcoal-treated; FCS, fetal calf serum; ERE, estrogen
response element; DCM, dichloromethane; PBS, phosphate
buffered saline; RLU, relative light unit
REFERENCES
■
(1) Silverman, S. L.; Christiansen, C.; Genant, H. K.; Vukicevic, S.;
Zanchetta, J. R.; de Villiers, T. J.; Constantine, G. D.; Chines, A. A.
Efficacy of bazedoxifene in reducing new vertebral fracture risk in
postmenopausal women with osteoporosis: results from a 3-year,
randomized, placebo-, and active-controlled clinical trial. J. Bone Miner.
Res. 2008, 23, 1923−1934.
(2) Kawate, H.; Takayanagi, R. Efficacy and safety of bazedoxifene for
postmenopausal osteoporosis. Clin. Interventions Aging 2011, 6, 151−
160.
(3) Crabtree, J. S.; Peano, B. J.; Zhang, X.; Komm, B. S.; Winneker, R.
C.; Harris, H. A. Activity of three selective estrogen receptor
modulators on hormone-dependent responses in the mouse uterus
and mammary gland. Mol. Cell. Endocrinol. 2008, 287, 40−46.
(4) Bardin, A.; Boulle, N.; Lazennec, G.; Vignon, F.; Pujol, P. Loss of
ERbeta expression as a common step in estrogen-dependent tumor
progression. Endocr.-Relat. Cancer 2004, 11, 537−551.
(5) Park, B.-W.; Kim, K.-S.; Heo, M.-K.; Ko, S.-S.; Lee, K. S.; Hong, S.
W.; Yang, W.-I.; Kim, J.-H.; Kim, G. E. Expression of estrogen receptor-
β in normal mammary and tumor tissues: Is it protective in breast
carcinogenesis? Breast Cancer Res. Treat. 2003, 80, 79−85.
(6) Gustafsson, J.-Å.; Warner, M. Estrogen receptor β in the breast:
role in estrogen responsiveness and development of breast cancer. J.
Steroid Biochem. Mol. Biol. 2000, 74, 245−248.
(7) Speirs, V.; Skliris, G. P.; Burdall, S. E.; Carder, P. J. Distinct
expression patterns of ERα and ERβ in normal human mammary gland.
J. Clin. Pathol. 2002, 55, 371−374.
(8) Hayashi, S.-I.; Eguchi, H.; Tanimoto, K.; Yoshida, T.; Omoto, Y.;
Inoue, A.; Yoshida, N.; Yamaguchi, Y. The expression and function of
estrogen receptor alpha and beta in human breast cancer and its clinical
application. Endocr.-Relat. Cancer 2003, 10, 193−202.
9617
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618

Journal of Medicinal Chemistry
Article
(9) Clarke, R. B.; Howell, A.; Potten, C. S.; Anderson, E. Dissociation
between steroid receptor expression and cell proliferation in the human
breast. Cancer Res. 1997, 57, 4987−4991.
(10) Fink, B. E.; Mortensen, D. S.; Stauffer, S. R.; Aron, Z. D.;
Katzenellenbogen, J. A. Novel structural templates for estrogen-
receptor ligands and prospects for combinatorial synthesis of estrogens.
Chem. Biol. 1999, 6, 205−219.
(28) Shiau, A. K.; Barstad, D.; Loria, P. M.; Cheng, L.; Kushner, P. J.;
Agard, D. A.; Greene, G. L. The structural basis of estrogen receptor/
coactivator recognition and the antagonism of this interaction by
tamoxifen. Cell 1998, 95, 927−937.
(29) Rajan, S. S.; Kim, Y.; Vanek, K.; Joachimiak, A.; Greene, G. L.
Crystal structure of human estrogen receptor alpha ligand-binding
domain in complex with a glucocorticoid receptor interacting protein 1
nr box II peptide and estra-1,3,5(10)-triene-3,15 alpha,16alpha,17beta-
tetrol. To be published.
(30) Mewshaw, R. E.; Bowen, S. M.; Harris, H. A.; Xu, Z. B.; Manas, E.
S.; Cohn, S. T. ERbeta ligands. Part 5: synthesis and structure−activity
relationships of a series of 4′-hydroxyphenyl-aryl-carbaldehyde oxime
derivatives. Bioorg. Med. Chem. Lett. 2007, 17, 902−906.
(11) Stauffer, S. R.; Coletta, C. J.; Tedesco, R.; Nishiguchi, G.;
Carlson, K.; Sun, J.; Katzenellenbogen, B. S.; Katzenellenbogen, J. A.
Pyrazole ligands: structure−affinity/activity relationships and estrogen
receptor alpha selective agonists. J. Med. Chem. 2000, 43, 4934−4947.
(12) Schafer, A.; Wellner, A.; Gust, R. Synthesis and investigations on
̈
the oxidative degradation of C3/C5-alkyl-1,2,4-triarylpyrroles as
(31) Richardson, T. I.; Dodge, J. A.; Wang, Y.; Durbin, J. D.; Krishnan,
V.; Norman, B. H. Benzopyrans as selective estrogen receptor beta
agonists (SERBAs). Part 5: Combined A- and C-ring structure−activity
relationship studies. Bioorg. Med. Chem. Lett. 2007, 17, 5563−5566.
ligands for the estrogen receptor. ChemMedChem 2011, 6, 794−803.
(13) Schafer, A.; Wellner, A.; Strauss, M.; Wolber, G.; Gust, R.
̈
Development of 2,3,5-triaryl-1H-pyrroles as estrogen receptor alpha
selective ligands. ChemMedChem 2011, 6, 2055−2062.
(14) Hoffmann, R. Synthese und Pharmakologische Untersuchung
Neuer Triaryl-1H-pyrrole mit Estrogenen, Zytotoxischen und COX-
(32) Mocklinghoff, S.; Rose, R.; Carraz, M.; Visser, A.; Ottmann, C.;
̈
Brunsveld, L. Synthesis and crystal structure of a phosphorylated
estrogen receptor ligand binding domain. ChemBioChem 2010, 11,
2251−2254.
Inhibitorischen Eigenschaften. Ph.D. Thesis. Freie Universitat Berlin,
̈
Berlin, Germany, 2007.
(33) Stauffer, S. R.; Huang, Y.; Coletta, C. J.; Tedesco, R.;
Katzenellenbogen, J. A. Estrogen pyrazoles: defining the pyrazole
core structure and the orientation of substituents in the ligand binding
pocket of the estrogen receptor. Bioorg. Med. Chem. 2001, 9, 141−150.
(34) Karl, J.; Gust, R.; Spruss, T.; Schneider, M. R.; Schoenenberger,
H.; Engel, J.; Wrobel, K. H.; Lux, F.; Haeberlin, S. T. Ring-substituted
[1,2-bis(4-hydroxyphenyl)ethylenediamine]dichloroplatinum(II) com-
plexes: compounds with a selective effect on the hormone-dependent
mammary carcinoma. J. Med. Chem. 1988, 31, 72−83.
(15) Mortensen, D. S.; Rodriguez, A. L.; Carlson, K. E.; Sun, J.;
Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Synthesis and
biological evaluation of a novel series of furans: ligands selective for
estrogen receptor alpha. J. Med. Chem. 2001, 44, 3838−3848.
(16) Wyvratt, J. M.; Hazen, G. G.; Weinstock, L. M. A convenient,
general synthesis of alpha-trichloro-methyl carbinols. J. Org. Chem.
1987, 52, 944−945.
(17) Cafiero, L. R.; Snowden, T. S. General and practical conversion
of aldehydes to homologated carboxylic acids. Org. Lett. 2008, 10,
3853−3856.
(35) Gust, R.; Keilitz, R.; Schmidt, K. Synthesis, structural evaluation,
and estrogen receptor interaction of 2,3-diarylpiperazines. J. Med. Chem.
2002, 45, 2325−2337.
(18) Oki, M. Applications of Dynamic NMR Spectroscopy to Organic
Chemistry; VCH Publishers: Deerfield Beach, FL, 1985; Vol. 4, p 423.
(19) Oki, M. The Chemistry of Rotational Isomers; Springer: Berlin,
Germany, 1993; Vol. 30, p 172.
(36) Gust, R.; Keilitz, R.; Schmidt, K.; von Rauch, M. (4R,5S)/
(4S,5R)-4,5-Bis(4-hydroxyphenyl)-2-imidazolines: ligands for the
estrogen receptor with a novel binding mode. J. Med. Chem. 2002,
45, 3356−3365.
(37) Sadowski, J.; Gasteiger, J.; Klebe, G. Comparison of automatic
three-dimensional model builders using 639 X-ray structures. J. Chem.
Inf. Comput. Sci. 1994, 34, 1000−1008.
(38) Goebel, M.; Wolber, G.; Markt, P.; Staels, B.; Unger, T.;
Kintscher, U.; Gust, R. Characterization of new PPAR[gamma]
agonists: benzimidazole derivativesimportance of positions 5 and
6, and computational studies on the binding mode. Bioorg. Med. Chem.
2010, 18, 5885−5895.
(39) Wolber, G.; Langer, T. LigandScout: 3-D pharmacophores
derived from protein-bound ligands and their use as virtual screening
filters. J. Chem. Inf. Model. 2005, 45, 160−169.
(40) Wolber, G.; Dornhofer, A. A.; Langer, T. Efficient overlay of
small organic molecules using 3D pharmacophores. J. Comput.-Aided
Mol. Des. 2006, 20, 773−788.
(41) Wolber, G.; Seidel, T.; Bendix, F.; Langer, T. Molecule-
pharmacophore superpositioning and pattern matching in computa-
tional drug design. Drug Discovery Today 2008, 13, 23−29.
(42) El Khissiin, A.; Journe, F.; Laios, I.; Seo, H. S.; Leclercq, G.
Evidence of an estrogen receptor form devoid of estrogen binding
ability in MCF-7 cells. Steroids 2000, 65, 903−913.
(20) Hafner, F.; Holler, E.; von Angerer, E. Effect of growth factors on
estrogen receptor mediated gene expression. J. Steroid Biochem. Mol.
Biol. 1996, 58, 385−393.
(21) Meyer, T.; Koop, R.; von Angerer, E.; Schonenberger, H.; Holler,
̈
E. A rapid luciferase transfection assay for transcription activation
effects and stability control of estrogenic drugs in cell cultures. J. Cancer
Res. Clin. Oncol. 1994, 120, 359−364.
̈
(22) Laıos, I.; Cleeren, A.; Leclercq, G.; Nonclercq, D.; Laurent, G.;
Schlenk, M.; Wellner, A.; Gust, R. Effects of (R,S)/(S,R)-4,5-bis(2-
chloro-4-hydroxyphenyl)-2-imidazolines and (R,S)/(S,R)-2,3-bis(2-
chloro-4-hydroxyphenyl)piperazines on estrogen receptor alpha level
and transcriptional activity in MCF-7 cells. Biochem. Pharmacol. 2007,
74, 1029−1038.
(23) Watts, C. K. W.; King, R. J. B. Overexpression of estrogen
receptor in HTB 96 human osteosarcoma cells results in estrogen-
induced growth inhibition and receptor cross talk. J. Bone Miner. Res.
1994, 9, 1251−1258.
(24) Wiglenda, T.; Gust, R. Structure−activity relationship study to
understand the estrogen receptor-dependent gene activation of aryl-
and alkyl-substituted 1H-imidazoles. J. Med. Chem. 2007, 50, 1475−
1484.
(25) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible docking.
J. Mol. Biol. 1997, 267, 727−748.
(26) Bruning, J. B.; Parent, A. A.; Gil, G.; Zhao, M.; Nowak, J.; Pace,
M. C.; Smith, C. L.; Afonine, P. V.; Adams, P. D.; Katzenellenbogen, J.
A.; Nettles, K. W. Coupling of receptor conformation and ligand
orientation determine graded activity. Nat. Chem. Biol. 2010, 6, 837−
843.
(27) Fang, J.; Akwabi-Ameyaw, A.; Britton, J. E.; Katamreddy, S. R.;
Navas, F., 3rd; Miller, A. B.; Williams, S. P.; Gray, D. W.; Orband-
Miller, L. A.; Shearin, J.; Heyer, D. Synthesis of 3-alkyl naphthalenes as
novel estrogen receptor ligands. Bioorg. Med. Chem. Lett. 2008, 18,
5075−5077.
9618
dx.doi.org/10.1021/jm300860j | J. Med. Chem. 2012, 55, 9607−9618