Nonsteroidal estrogen receptor isoform-selective biphenyls

Article abstract of DOI:10.1111/cbdd.13126

Estrogen receptor (ER) has been a therapeutic target to treat ER-positive breast cancer, most notably by agents known as selective estrogen receptor modulators (SERMs). However, resistance and severe adverse effects of known drugs gave impetus to the sear

Full text of DOI:10.1111/cbdd.13126

DR B. JAYARAM (Orcid ID : 0000-0002-5495-2213)
Article type : Research Article
Non-steroidal Estrogen Receptor Isoform Selective Biphenyls
Seema Bhatnagara, #,*, Anjali Sonib,#, Swati Kaushika,c, Megha Rikhia, T R Santhosh Kumarc,
B. Jayaramb,d,*
aAmity Institute of Biotechnology, Amity University, Noida-201303, India
bDepartment of Chemistry and Supercomputing Facility for Bioinformatics & Computational Biology,
Indian Institute of Technology, Hauz Khas, New Delhi-110016, India
cCancer Research Programme Lab1, Rajiv Gandhi Centre for Biotechnology, Poojappura,
Thiruvananthapuram, Kerala- 695014, India
dKusuma School of Biological Sciences, Indian Institute of Technology, Hauz Khas, New Delhi-
110016, India
Email: bjayaram@chemistry.iitd.ac.in; sbhatnagar1@amity.edu
*Corresponding authors
#Equal contributions
Abstract
Estrogen receptor (ER) has been a therapeutic target to treat ER positive breast cancer, most notably
by agents known as selective estrogen receptor modulators (SERMs). However, resistance and severe
adverse effects of known drugs gave impetus to the search for newer agents with better therapeutic
profile. ERα and ERβ are two isoforms sharing 56% identity, and having different physiological
functions and expressions in various tissues. Only two residues differ in the active sites of the two
isoforms motivating us to design isoform selective ligands. Guided by computational docking and
molecular dynamics simulations, we have designed, synthesized and tested, substituted biphenyl-2,6-
diethanones and their derivatives as potential agents targeting ERα. Four of the molecules synthesized
exhibited preferential cytotoxicity in ERα+ cell line (MCF-7) compared to ERβ+ cell line (MDA-MB-
231). Molecular dynamics (MD) in combination with molecular mechanics-generalized born surface
area (MM-GBSA) methods could account for binding selectivity. Further co-treatment and E-screen
studies with known ER ligands- estradiol (E2) and tamoxifen (Tam) indicated isoform selective anti-
estrogenicity in ERα+ cell line which might be ER mediated. ERα siRNA silencing experiments
further confirmed the ER selective nature of ligands.
Introduction
Estrogen Receptor (ER) is the mainstay in breast cancer endocrine therapies1. Estrogen plays a critical
role in hormone responsive breast cancer progression by binding to ER leading to DNA synthesis and
cellular proliferation2. The ER isoforms, ERα and ERβ, encoded by separate genes are responsible for
mediating estrogen responsive action3. ERα is expressed in about 75% of the diagnosed breast tumors
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/cbdd.13126
This article is protected by copyright. All rights reserved.

and can be treated with agents targeting estrogen mediated signaling4,5. One of the most prescribed,
first-line adjuvant treatments available for ERα+ breast cancer is tamoxifen, a selective ER
modulator6,7. Tamoxifen acts as an antagonist in breast tissue but as an agonist in bone and uterus.
Contrastingly, its estrogenicity is beneficial in bone to prevent osteoporosis while in uterus, it can lead
to endometrial cancer8. Moreover, it is observed that tamoxifen responsive breast cancer patients
become resistant and long term therapy predisposes them to severe adverse effects such as
development of endometrial cancer9–11. Besides, signaling through protein kinases also leads to
tamoxifen resistance in an estradiol-independent manner12. The limited effectiveness of current
chemotherapeutic drugs urges identification of some novel ER targeting agents with high efficacy and
minimal side effects.
Despite tremendous consistent efforts in this area, ER molecular biology has proven to be
exceedingly complex. Though ERs behave as ligand activated transcription factors, ER ligands cannot
be considered as ‘molecular switches’ that activate or deactivate the receptor post ligand binding. It is
well established that different ER ligands lead to different ER conformations resulting in agonist,
partial agonist, inverse agonist and antagonist behaviors13.
Nonetheless, identification of potent ER ligands remains a hot area of research providing
excellent opportunities for the design of isoform selective agents and is therefore of immediate
interest to the drug design community. The ligand binding domain (LBD) of ER isoforms share 56%
homology14. The dynamic and plastic nature of LBD makes it an attractive target for a wide spectrum
of ligands. However in most cases, potent and efficacious ligands are not ER subtype selective; which
often translates into off-target toxicity15,16. On the contrary, ER subtype selective ligands have
suboptimal in vivo pharmacokinetics. Extensive studies have been undertaken in the last 50 years to
determine critical interacting residues in ER LBD that translate into enhanced potency of ligands. It
has been elucidated from previous reports that subtype selective agents must firstly, position
substituents close to amino acid differences between ERα and ERβ to develop subtype selectivity;
secondly, the ligand must be conformationally rigid; and thirdly, the presence of various substituents
including the phenol (or phenolic bioisosteres) can aid in achieving selectivity for ER subtypes
through molecular recognition17–20
.
In continuation of our efforts to identify ER subtype selective ligands, we focus here on
nonsteroidal scaffolds since nonsteroidal ligands are considered as remarkable high affinity binders
against ER21. Biphenyl scaffold (Fig. 1) is chosen for designing potential ERα hits, as they could be
regarded as surrogates of the steroidal backbone22. The current study on biphenyl-2,6-diethanones was
undertaken to design and determine key structural determinants in biphenyl scaffold that may confer
ERα subtype selectivity.
Methods and Materials
Crystal structures of human ERα and ERβ complexed with 4-hydroxytamoxifen (OHT) and
R,R-5,11-cis-diethyl-5,6,11,12-tetrahydrochrysene-2,8-diol (THC) are extracted from RCSB protein
data bank with codes 3ERT and 1L2J, respectively23–25. Both the complexes have antagonist bound
protein conformations. The missing loop regions in 1L2J are modeled using Modeller26. Biphenyl
compounds 3(a-d) were designed against ERα using structure-based drug design approach27. Docking
is performed on both the ER isoforms using ParDOCK module of Sanjeevini software suite developed
in-house, which contains built-in features to take care of geometry optimization, partial charge
derivation and parameter assignment of the molecules in an automated way27–29
.
To further develop a molecular view of the structure and dynamics of ERα-ligand complexes,
molecular dynamics (MD) simulations are carried out. 3ERT is used as ERα control. MD simulations
are performed using AMBER14 software suite on all the docked complexes obtained from
ParDOCK30. Amber ‘ff99SB’ and ‘GAFF’ force fields are used for the protein and ligand file
preparation. All the simulations are conducted with periodic boundary conditions. The protein-ligand
complexes are solvated in a box of TIP3P water molecules and electroneutrality of the system is
This article is protected by copyright. All rights reserved.

maintained31. An initial minimization is conducted with 5000 steps (2500SD+2500CG). Simulations
are started by slow heating of the solvent to 300 K at constant volume using harmonic restraints of 25
kcal/mol Å-2 on the solute atoms for a period of 200ps. These restraints are slowly relaxed to 1
kcal/mol Å-2 during equilibration of 1ns. The system is further simulated for production run of 100ns
under NPT conditions using Berendsen algorithm with a coupling constant of 5ps32. Long-range
coulombic interactions are treated using particle mesh ewald method (PME) and a cutoff of 12 Å
distance is used for van der Waals (VDW) interactions33. Shake is enabled to constrain the covalent
bonds involving hydrogens34.
A detailed analysis of the interaction energies has been carried out using molecular mechanics
generalized Born surface area (MM-GBSA) methodology to further provide insights into protein-
ligand binding and thus on the inhibitory mechanism of compounds 3(a-d) against ERα and further to
elucidate the basis of specificity of compound 3b for ERα over ERβ35,36. MM-GBSA method
employed to predict the binding free energies (∆Gbind) comprises gas phase electrostatics (ELE) and
VDW energies, polar and non-polar solvation energies and entropic contributions. 200 snapshots are
extracted at equal intervals from the last 10ns stable MD trajectories for binding free energy
estimations. For each snapshot, the binding free energy of protein-ligand complex is estimated as:
∆Gbind = Gcomplex – (Gprotein + Gligand
)
(1)
The ∆Gbind is the sum of the changes in molecular mechanics gas-phase binding free energy
(∆GMM), the solvation free energy (∆Gsolv) and the entropic contributions (-T∆S).
∆Gbind = ∆GMM + ∆Gsolv - T∆S
∆GMM = ∆Gele + ∆GvdW
(2)
(3)
(4)
∆Gsolv = ∆Gpol,solv + ∆Gnonpol,solv
The ELE and VDW (∆Gele and ∆GvdW) energies are derived from sander as used for MD
simulation. The polar contribution to solvation (∆Gpol,solv) is obtained using GBSA module of
AMBER14. The non-polar contribution to solvation (∆Gnonpol,solv) is determined using
∆Gnonpol,solv = γ*SASA, where SASA is the solvent-accessible surface area and γ is the empirical
surface tension and is set to 0.0072 kcal/mol-Å-2. The conformational entropy contributions to free
energies upon ligand binding are calculated using normal-mode analysis by the nmode AMBER
module30. A total of 200 snapshots are chosen from the last 10ns of MD simulations for the molecular
mechanical free energies and solvation energies. However, due to high computational demand for
entropy calculations, only 20 snapshots are extracted from the last 10ns of MD trajectories. In
addition to MM-GBSA binding free energies, per residue binding free energies are also calculated.
Further, to corroborate the computational findings, the initial set of compounds 1(a-i) are
synthesized by adopting known procedures37. The biphenyl derivatives 3(a-d) are synthesized in two
steps from compounds 1(a-b) as depicted in Fig. 2. In the first step, biphenyl-2,6-diethanone
derivatives 1(a-b) are reacted with NBS in methanol which yielded 1,1’-(3,4’-dihydroxy-5-
methylbiphenyl)-2,6-bis-bromodiethanone 2a and 1,1’-(3,2’-dihydroxy-5-methylbiphenyl)-2,6-bis-
bromodiethanone 2b. In the second step, compounds 2a and 2b are further reacted with primary
amines, butylamine and 1,4-diaminobutane in presence of acetonitrile/pyridine, yielding the desired
products 3(a-d) (Fig. 2). The structures of compounds are characterized based on their spectroscopic
data (1H, 13C NMR, IR and Mass spectra) as provided in supporting information.
The MTT assay is performed to evaluate the anti-breast activity of all the synthesized
compounds against ERα+ (MCF-7) and ERβ+ (MDA-MB-231) cell lines. MRC-5 cell line is used as
a control. Followed by, the co-treatment studies are performed with known ER ligands, Tam and E2.
This article is protected by copyright. All rights reserved.

Protocol adopted is reported by earlier workers38. The anti-estrogenicity and antagonist nature of the
most active compound is evaluated by E-screen assay, where tumor cells are exposed to putative anti-
estrogens and cellular proliferation is compared with the tumor cells treated with E239. The E-screen
assay is performed in two steps according to the protocol slightly modified from Soto et al40.
Furthermore, the siRNA silencing experiments are performed to confirm ERα selectivity where we
silenced ER in ER+ MCF-7 cells using specific siRNA against human ERα. The whole cell extract
from silenced and non-silenced cells are prepared to confirm the silencing. The whole cell extract is
separated by electrophoresis and western blotting is carried out as described in the supporting
information. Further details of all experiments are provided in supporting information.
Results and Discussion
The calculated docking energies of all the four compounds against the active sites of both ER
isoforms are shown in Table 1. Reasonably good, preferential docking energies are obtained for most
of the biphenyl compounds against ERα but since both isoforms are highly homologous, differing by
only two amino acid substitutions in the active sites - Leu384 of ERα is substituted by Met336 in ERβ
and Met421 of ERα is substituted by Ile373 in ERβ - the energy differences are small41. Docking is
also carried out using glide software of Schrodinger and similar results are obtained.
The binding cavity of ERα is mainly comprised of following residues- Arg394, Met421,
Glu353, Met343, His524, Ala350, Phe404, Thr347, Leu525, Trp383, Leu384, Ile424 and Gly521. The
binding modes of various compounds 3(a-d) captured from the last stable MD trajectories are shown
in Fig. 3. Compound 3a forms a few nonpolar contacts with His524, Met343, Ala350, Met421 and
Met388 with no hydrogen bond formation. Compound 3b is anchored to the cavity via a hydrogen
bond formation between amine group of one of the butylamine derivatives and main chain atoms of
Leu525 and capturing almost all the VDW and hydrophobic contacts required for the activity of ERα
as depicted in Fig 3. The side chains of compound 3b make extensive hydrophobic contacts with
residues- Met343, Thr347, Ala350, Met388, Phe404, Arg394, Leu391, Met421, Leu346, Leu525 and
Trp383. It fits snugly in the cavity of ERα. Compound 3c forms various hydrogen bonds with Glu353,
Leu387, Glu419 and Thr347 with the active site residues of ERα. Its orientation in the cavity is
different and therefore forms only a few VDW contacts. Compound 3d also forms fewer non-polar
contacts. Based on these MD derived interaction patterns, we identified compound 3b to be the most
active compound of the biphenyl series 3(a-d) as it retains some important contacts throughout the
simulation as formed by tamoxifen23.
To assess the dynamic stability of all the complexes during the entire 100ns production MD
run, their structural and energetic properties are closely monitored. Root mean square deviations
(RMSDs) of compounds 3(a-d) with ERα backbone in relation to their initial structures are plotted in
Fig. S1(a) of supporting information. RMSD tends to converge after 40-50ns indicating the systems
stability. RMSD of compound 3b with ERα and ERβ backbone in comparison to ERα-Tam (3ERT) is
plotted in Fig. S1(b) indicating similar RMSD pattern of ERα-3b with ERα-Tam and their high
stabilities throughout the MD.
The calculated binding free energies (∆Gbind) using MM-GBSA suggest ERα-3b complex to
be the most stable among the series 3(a-d) (Table 2). By comparing the energetic contributions of
ERα and 3(a-d), it is observed that the favorable electrostatics (∆Gele) are compensated by
unfavorable polar solvation energies (∆Gpol,solv) in all the four complexes. The net electrostatic
(∆Gele+∆Gpol,solv) contributions are unfavorable to binding. The favorable sum of gas phase VDW
(∆GvdW) and nonpolar solvation energies (∆Gnonpol,solv) arising from the ligands hydrophobic core
forms the basis of favorable binding energies. The net VDW (∆GvdW+∆Gnonpol,solv) values for ERα-
3a, ERα-3b, ERα-3c and ERα-3d are -64.22, -64.65, -56.75 and -49.92 kcal/mol, respectively.
Enthalpic contributions provide the strength of interactions between the receptor and the ligand and
entropic contributions provide the change in entropy of the solvent due to ligand binding and loss of
conformational degrees of freedom (rotational, vibrational and translational). The unfavorable entropy
This article is protected by copyright. All rights reserved.

contributions in all the systems are opposed by ~23-27 kcal/mol. The net binding free energy
comprising enthalpic and entropic terms is higher for ERα-3b (-23.63 kcal/mol); dominated by shape
complementarity, and thereby resulting in its higher affinity.
To elucidate the selective inhibitory activity of 3b towards ERα over ERβ, we calculated the
MM-GBSA binding free energies for ERβ-3b as well. To unravel the selectivity mechanism of 3b, the
binding free energy components of both the complexes ERα-3b and ERβ-3b are compared. The gas-
phase VDW and nonpolar solvation energies are favorable for binding (-64.65 for ERα-3b and -66.25
and ERβ-3b), but their difference is negligible to account for selectivity. However, the gas-phase
electrostatics is playing a subtle role in selectivity of 3b towards ERα as indicated by the large
difference in electrostatic energy of -8.04 kcal/mol (Table 2). Altogether, the enthalpic interactions
and the net binding free energies are better for ERα-3b explaining its selectivity over ERβ-3b.
The important active site residues governing selectivity of 3b to ERα over (ERβ) are Thr347
(Thr299), Ala350 (Ala302), Trp383 (Trp335), Arg394 (Arg346), Met421 (Ile373), His524 (His475)
and Leu525 (Leu476) as shown in Fig. 4. Larger contribution to selectivity is mainly achieved by
Thr347 (Thr299), Ala350 (Ala302), Met421 (Ile373) and Leu525 (Leu476). Per-residue energy
calculation is also performed on the same 200 snapshots chosen from the last 10ns MD trajectories for
MM-GBSA calculations. The relative positions of ligand 3b in ERα and ERβ are compared with the
crystallographic complexes (3ERT and 1L2J) especially with two key residues: Leu384 (Met336) and
Met421 (Ile373) shown in Fig. 5. The superposition highlights the orientation of two amino acids
which actually differ in ERα (Leu384 and Met421) and ERβ (Met336 and Ile373) resulting in 3b
specificity towards ERα.
Overall, the simulations and the energy analyses strongly point to a preferential binding of 3b
with ERα over ERβ providing a molecular level explanation.
The anti-breast cancer potential of all the synthesized compounds are evaluated against ERα+
(MCF-7) and ERβ+ (MDA-MB-231) cell lines with relevant controls (MRC-5) using MTT assay. The
IC50 values of compounds 3a, 3b, 3c, 3d against ERα+ cell line are 62.5, 0.97, 31.25 and 1.95 µg/ml
whereas against ERβ+ cell line are >125, 62.5, >125 and 62.5 µg/ml (Fig. 6). The MTT assay
measures the cell proliferation rate and conversely, when metabolic events lead to apoptosis or
necrosis, the reduction in the cell viability. The MRC-5 cell line is derived from human lung tissue
and is used here as control for serving the role of an ER negative cell line. Results for 1(a-i) are
provided in Fig. S2 (supporting information). All the 3(a-d) compounds exhibit selective inhibitory
activity against ERα+ (MCF-7) cell line over ERβ+ (MDA-MB-231). Among them, compound 3b is
found to be the most potent with an IC50 value of 0.97 µg/ml (2.27 µM) compared to 3a, 3c and 3d
through MTT assay (Fig. 6). MD simulations combined with MM-GBSA and cell line studies suggest
compound 3b as the best representative compound of the series 3(a-d), building a strong synergism
between computations and experiments. Competitive binding assay could be performed to measure
directly the relative binding affinity of compound(s) to ERα, but because of the regulatory issues
associated with the use and handling of radioactive compound at our current facility, we could not
carry out this work.
In order to determine that the observed anti-proliferative effect of compound 3b is ER
mediated, co-treatment studies are performed to evaluate its synergism with known ER ligands, Tam
and E2. In MCF-7 cell line, our co-treatment results indicate enhanced cytotoxicity when compound
3b is taken along with Tam and similarly, a reversal effect is observed when compound 3b is taken
with E2 (Fig. 7). Co-treatment of compound 3b in MDA-MB-231 and MRC-5, however did not
exhibit enhanced cytotoxicity (Fig. S3). Therefore, the co-treatment results exhibited significant
synergism of compound 3b with Tam in ERα+ (MCF-7) cell lines, whereas no synergism was
observed in ERβ+ (MDA-MB-231) and ER negative (MRC-5) cell lines, which further indicate that
the synergism is directly associated with ERα target and hence affected only ERα+ cell line. Details of
co-treatment studies of all the compounds are provided in supporting information.
This article is protected by copyright. All rights reserved.

Time dependent studies with compound 3b and Tam at IC50 concentration in MCF-7 cells also
exhibited significant decline in the cell viability (Fig. 8). Altogether, these co-treatment results
indicate enhanced cytotoxicity of 3b in dose and time dependent manner against ERα+ (MCF-7) cell
line which plausibly could be ER mediated. These interesting results prompted us to evaluate the anti-
estrogenicity of compound 3b.
Our results from first step (time course studies) of E-Screen assay indicate that when MCF-7
cells are exposed to the lowest concentration of E2 (0.002 µg/ml) along with the range of
concentrations of compound 3b from 24-96 hours, the cytotoxic effects of compound 3b are significant
with a decline in cell viability in a dose and time dependent manner. It is also observed that at
concentration 0.48µg/ml, the cytotoxic effect of compound 3b is abated and a mitogenic response
similar to E2 alone is apparent (Fig. 9(a)). However, as the concentration of compound 3b is increased,
there is a decrease in the cell viability.
Results from second step of E-screen assay, when MCF-7 cells are exposed to a range of
increasing concentration of E2 along with IC50 concentration of 3b (0.97µg/ml), show a noticeable
effect with increase in MCF-7 cellular proliferation, maximal at 0.013µg/ml (50nM) concentration of
E2 throughout the time period (Fig. 9(b)). The E-screen assay results clearly demonstrate that the
inhibition of MCF-7 cells induced by compound 3b can be reversed by the addition of E2. Since 17β-
estradiol has more than 100-fold-greater affinity for ER than Tam, it can compete with Tam or any
putative antagonists and restore the receptor processing and other estrogen associated events38,42
.
The above studies indicate that the compound 3b is selective to ER and its effect on cell
proliferation and cell death are subsequent to its ability to target ER. Further we performed siRNA
silencing experiments to confirm ERα selectivity of compound 3b. The specific antibody against ERα
is used to detect ERα levels and compared with the house keeping protein β actin (Fig. 10(a)). As
shown in the figure, ER has been significantly down regulated in silenced cells compared to parental
MCF-7 cells. The parental MCF-7 and silenced cells were exposed to tamoxifen and compound 3b for
48 hours followed by apoptosis analysis by chromatin condensation as presented in Fig. 10(b) and
Fig. 10(c). Silencing of ERα results in a reduction of cell death induced by tamoxifen and 3b.
Conclusions
In a bid to identify non-steroidal estrogen receptor isoform selective biphenyls, our studies
encompassed the use of computational tools for ligand design followed by synthesis and bioactivity
evaluation. Our studies indicate that the presence of hydroxyl substituents at ortho position of ring B
of the biphenyl scaffold (Fig. 1) resulted in compounds 3b and 3d exhibiting ERα selectivity.
Docking studies indicated that biphenyl compounds containing butylamine groups 3a and 3b act as
good binders compared to diaminobutane derivatives 3c and 3d which act as weak binders.
Introduction of long chain substituents on the biphenyl core of ring A in compounds 3(a-d) resulted in
enhanced docking scores indicating better binding affinities with ERα. Further docking and MD
studies indicated presence of bulky groups on ring A resulted in non-coplanarity of biphenyl ring
which possibly resulted in enhanced docking scores. Analysis of MD results including MM-GBSA
suggested 3b as the best binder with high enthalpic contributions, dominated by van der Waals
interactions. The selectivity of 3b towards ERα over ERβ is dominated by electrostatics and is
contributed mainly by back bone atoms of four amino acids residues namely Thr347 (Thr299),
Ala350 (Ala302), Met421 (Ile373) and Leu525 (Leu476) as indicated by free energy per-residue
decomposition. Cytotoxicity studies of all the compounds belonging to this series indicated higher
cytotoxicity in ERα+ cell line (MCF-7) as compared ERβ+ cell line (MDA-MB-231). This is a new
finding in the light of reports on 4-hydroxybiphenyl compounds which have 20-70 fold ERβ
selectivity43. Compound 3b containing ortho hydroxy group was found to be more potent with IC50 of
0.97 µg/ml as compared to other compounds. Co-treatment of compound 3b with known ER
antagonist Tam also led to enhanced cytotoxicity which is dose dependent (Fig. 7) and time dependent
(Fig. 8). The antagonistic nature of the ligands is confirmed by E-screen assay wherein the test
compound 3b is evaluated at IC50 concentration in the presence of increasing concentrations of known
ER agonist, E2. The relative proliferative potency (RPP) of compound 3b as compared to E2 is
This article is protected by copyright. All rights reserved.

determined to be 0.004. In addition, gene silencing of ER using human siRNA against ER+ MCF-7
cells led to a reduction in cell death induced by compound 3b. Thus, the results on gene silencing
experiments confirmed ERα selectivity of the identified lead 3b (Fig. 10). The results of our studies
could be valuable for future rational design of selective ER inhibitors, providing a better
understanding of the energetics of affinity versus selectivity. Results also suggest compound 3b as a
good candidate lead molecule which could be further optimized to explore biphenyl scaffolds in
search of better, selective and high affinity ligands.
References
1.
2.
3.
4.
5.
6.
Musgrove, E. A & Sutherland, R. L. Biological determinants of endocrine resistance in breast
cancer. Nat. Rev. Cancer 9, 631–43 (2009).
Hall, J. M., Couse, J. F. & Korach, K. S. The Multifaceted Mechanisms of Estradiol and
Estrogen Receptor Signaling. J. Biol. Chem. 276, 36869–36872 (2001).
Tsai, M. & O’Malley, B. W. Molecular mechanisms of action of steroid / thyroid receptor
superfamily members. Annu. Rev. Biochem. 63, 451–486 (1994).
Popolo, A. et al. Antiproliferative activity of brown Cuban propolis extract on human breast
cancer cells. Nat Prod Commun 4, 1711–1716 (2009).
Ali, S. & Coombes, R. C. Endocrine-responsive breast cancer and strategies for combating
resistance. Nat Rev Cancer 2, 101–112 (2002).
Group, E. B. C. T. C. Relevance of breast cancer hormone receptors and other factors to the
efficacy of adjuvant tamoxifen: Patient-level meta-analysis of randomised trials. Lancet 378,
771–784 (2011).
7.
8.
Osborne, C. K., Zhao, H. & Fuqua, S. a. Selective estrogen receptor modulators: structure,
function, and clinical use. J. Clin. Oncol. 18, 3172–86 (2000).
Bernstein, L. et al. Tamoxifen Therapy for Breast Cancer and Endometrial Cancer Risk. JNCI
J Natl Cancer Inst 91, 1654–1662 (1999).
9.
10.
Chang, M. Tamoxifen resistance in breast cancer. Biomol. Ther. (Seoul). 20, 256–67 (2012).
Hu, R., Hilakivi-Clarke, L. & Clarke, R. Molecular mechanisms of tamoxifen-associated
endometrial cancer (Review). Oncol. Lett. 9, 1495–1501 (2015).
11.
12.
Swaby, R. F., Sharma, C. G. N. & Jordan, V. C. SERMs for the treatment and prevention of
breast cancer. Rev. Endocr. Metab. Disord. 8, 229–239 (2007).
McClaine, R. J., Marshall, A. M., Wagh, P. K. & Waltz, S. E. Ron receptor tyrosine kinase
activation confers resistance to tamoxifen in breast cancer cell lines. Neoplasia 12, 650–658
(2010).
13.
14.
15.
16.
17.
Heldring, N. et al. Estrogen receptors: how do they signal and what are their targets. Physiol.
Rev. 87, 905–31 (2007).
Dahlman-Wright, K. et al. International Union of Pharmacology. LXIV. Estrogen receptors.
Pharmacol. Rev. 58, 773–781 (2006).
Martinkovich, S., Shah, D., Planey, S. L. & Arnott, J. Selective estrogen receptor modulators:
tissue specificity and clinical utility. Clin. Interv. Aging 9, 1437–1452 (2014).
Moore, T. W., Mayne, C. G. & Katzenellenbogen, J. A. Minireview: Not Picking Pockets:
Nuclear Receptor Alternate-Site Modulators (NRAMs). Mol. Endocrinol. 24, 683–695 (2010).
Ng, H. W., Perkins, R., Tong, W. & Hong, H. Versatility or promiscuity: The estrogen
receptors, control of ligand selectivity and an update on subtype selective ligands. Int. J.
Environ. Res. Public Health 11, 8709–8742 (2014).
18.
19.
Paternia, I., Granchia, C., Katzenellenbogen, J. A. & Minutoloa, F. Estrogen Receptors Alpha
(ERα) and Beta (ERβ): Subtype Selective Ligands and Clinical Potential. J.Steroids 0, 13–29
(2015).
Anstead, G. M., Carlson, K. E. & Katzenellenbogen, J. A. The estradiol pharmacophore:
Ligand structure-estrogen receptor binding affinity relationships and a model for the receptor
binding site. Steroids 62, 268–303 (1997).
20.
21.
Nilsson, S., Koehler, K. F. & Gustafsson, J.-Å. Development of subtype-selective oestrogen
receptor-based therapeutics. Nat. Rev. Drug Discov. 10, 778–792 (2011).
Gao, H., Katzenellenbogen, J. a, Garg, R. & Hansch, C. Comparative QSAR analysis of
This article is protected by copyright. All rights reserved.

estrogen receptor ligands. Chem. Rev. 99, 723–44 (1999).
22.
Lesuisse, D. et al. Biphenyls as surrogates of the steroidal backbone. Part 1: synthesis and
estrogen receptor affinity of an original series of polysubstituted biphenyls. Bioorg. Med.
Chem. Lett. 11, 1709–1712 (2001).
23.
24.
Shiau, A. K. et al. The structural basis of estrogen receptor/coactivator recognition and the
antagonism of this interaction by tamoxifen. Cell 95, 927–937 (1998).
Shiau, A. K. et al. Structural characterization of a subtype-selective ligand reveals a novel
mode of estrogen receptor antagonism. Nat. Struct. Biol. 9, 359–64 (2002).
Berman, H. M. et al. The Protein Data Bank. Nucleic Acids Res. 28, 235–242 (2000).
Fiser, a et al. Modeling of loops in protein structures. Protein Sci. 9, 1753–73 (2000).
Jayaram, B. et al. Sanjeevini: a freely accessible web-server for target directed lead molecule
discovery. BMC Bioinformatics 13 Suppl 1, S7 (2012).
25.
26.
27.
28.
29.
Jain, T. & Jayaram, B. An all atom energy based computational protocol for predicting binding
affinities of protein-ligand complexes. FEBS Lett. 579, 6659–6666 (2005).
Gupta, a, Gandhimathi, a, Sharma, P. & Jayaram, B. ParDOCK: an all atom energy based
Monte Carlo docking protocol for protein-ligand complexes. Protein Pept. Lett. 14, 632–646
(2007).
30.
31.
32.
33.
34.
Case, D. A. et al. The Amber biomolecular simulation programs. Journal of Computational
Chemistry 26, 1668–1688 (2005).
Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. Comparison
of simple potential functions for simulating liquid water. J. Chem. Phys. 79, 926 (1983).
Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., DiNola, a & Haak, J. R.
Molecular dynamics with coupling to an external bath. J. Chem. Phys. 81, 3684–3690 (1984).
Essmann, U. et al. A smooth particle mesh Ewald method. J Chem Phys 103, 8577–8593
(1995).
Ryckaert, J. P., Ciccotti, G. & Berendsen, H. J. C. Numerical integration of the cartesian
equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput.
Phys. 23, 327–341 (1977).
35.
Gohlke, H., Kiel, C. & Case, D. A. Insights into Protein–Protein Binding by Binding Free
Energy Calculation and Free Energy Decomposition for the Ras–Raf and Ras–RalGDS
Complexes. J. Mol. Biol. 330, 891–913 (2003).
36.
37.
Kalra, P., Reddy, T. V. & Jayaram, B. Free energy component analysis for drug design: A case
study of HIV-1 protease-inhibitor binding. J. Med. Chem. 44, 4325–4338 (2001).
Sharma, A., Pandey, J. & Tripathi, R. P. An efficient regioselective synthesis of functionalized
biphenyls via sequential reactions of aromatic aldehydes and β-keto esters or ketones.
Tetrahedron Lett. 50, 1812–1816 (2009).
38.
39.
40.
Lewis, J. B. et al. 4-Hydroxytamoxifen-induced cytotoxicity and bisphenol A: competition for
estrogen receptors in human breast cancer cell lines. In Vitro Cell. Dev. Biol. Anim. 36, 320–6
(2000).
Soto, A. M.; Lin, T. M.; Justicia, H.; Silvia, R. M.; Sonnenschein, C. in Chemically induced
alterations in sexual and functional development; the wildlife/human connection (ed. Colborn
T, C. C.) 295–309 (NJ: Princeton Scientific Publishing, 1992).
Chow, S. Suppression of cell proliferation and regulation of estrogen receptor alpha signaling
pathway by arsenic trioxide on human breast cancer MCF-7 cells. J. Endocrinol. 182, 325–337
(2004).
41.
42.
Pike, a C. et al. Structure of the ligand-binding domain of oestrogen receptor beta in the
presence of a partial agonist and a full antagonist. EMBO J. 18, 4608–4618 (1999).
Chen, H., Clemens, T. L., Hewison, M. & Adams, J. S. Estradiol and tamoxifen mediate
rescue of the dominant-negative effects of estrogen response element-binding protein in vivo
and in vitro. Endocrinology 150, 2429–2435 (2009).
43.
Edsall, R. J., Harris, H. A., Manas, E. S. & Mewshaw, R. E. ERβ ligands. Part 1: The
discovery of ERβ selective ligands which embrace the 4-hydroxy-biphenyl template.
Bioorganic Med. Chem. 11, 3457–3474 (2003).
This article is protected by copyright. All rights reserved.

Acknowledgements
The authors gratefully acknowledge Amity University for infrastructure and facilities. Author, MR
thanks financial assistance through Amity Science, Technology and Innovation Foundation (ASTIF)
fellowship provided by Amity University. Support to the Supercomputing Facility for Bioinformatics
& Computational Biology (SCFBio), IIT Delhi from the Department of Biotechnology, Govt. of India
is gratefully acknowledged. The authors also acknowledge the support from Cancer Research
Programme Lab 1, Rajiv Gandhi Centre for Biotechnology (RGCB).
Author’s contributions
AS performed computational studies. MR performed the synthesis and biological activities. SK and
TRSK performed ER siRNA silencing experiments. SB, BJ designed the project. All authors analyzed
the results and wrote the manuscript.
Corresponding Author
*E-mail: sbhatnagar1@amity.edu; bjayaram@chemistry.iitd.ac.in
Competing financial interests
The authors declare no competing financial interests.
Supporting information available
The supporting information contains the detail of experimental procedures for organic synthesis,
bioactivity assays, spectroscopic characterization of compounds and anticancer activity of 1(a-i) and
3(a-d).
Figure and Table captions
Figure 1: Compounds of interest
Figure 2: Synthesis of biphenyl-2,6-diethanone derivatives.
Figure 3: A comparative depiction of interactions of biphenyl compounds 3(a-d) with the active site
residues of ERα. The violet color represents various biphenyl compounds. Dashed bonds represent the
hydrogen bonds along with the distances whereas arcs represent VDW and hydrophobic contacts.
Circled residues represent hydrophobic side chain interactions with the corresponding ligands which
are also observed in ERα complexed with tamoxifen. All the snapshots for the above analyses are
taken from the last 10ns of the MD trajectories.
Figure 4: A comparison of per-residue binding free energy for key residues between ERα-3b and
ERβ-3b.
Figure 5: (a) A superposition of complex ERα-3b (cyan) with ERα-Tam (pdb id: 3ERT) (magenta).
(b) A superposition of complex ERβ-3b (cyan) with ERβ-THC (pdb id: 1L2J) (magenta). For clarity
all hydrogens are removed and only a few important amino acids interactions are displayed to
highlight selectivity. Yellow dotted bonds represent hydrogen bonds.
Figure 6: Dose dependent studies of biphenyl-2,6-diethanone derivatives 3(a-d) plotted against optical
density (OD) at 570nm in MCF-7, MDA-MB-231 and MRC-5 cell lines using MTT assay. The data
are means and standard error of the mean (SEM) from three samples of each group. Two-way
ANOVA followed by Bonferroni post test where p-value < 0.05 is significant for compounds.
Figure 7: Results of co-treatment studies at varied concentrations of compound 3b alone and in
combination with Tam against MCF-7 cell line. All the concentrations are in µg/ml.
Figure 8: Time course of the effect of combination of 0.056 µg/ml Tam and 0.97 µg/ml 3b on MCF-7
cells. Mitochondrial dehydrogenase activity is measured via MTT assay for the selected time period.
This article is protected by copyright. All rights reserved.

Cells with media are considered as negative control and with 0.002 µg/ml (10nM) E2 as positive
control.
Figure 9: (a) Time course of the effect of treatment with 0.002µg/ml (10nM) E2 and a range of
concentration of 3b on mitochondrial dehydrogenase activity of the MCF-7 cells; (b): Time course of
the effect of co-treatment with 0.97µg/ml (2.27µM) 3b and a range of concentrations of E2 on
mitochondrial dehydrogenase activity of the MCF-7 cells to establish estrogen rescue phenomena. All
the concentration units are µg/ml.
Figure 10: (a) Gene silencing experiment. Expression of ERα protein by western blot analysis
following transfection with ERα siRNA. The difference in the expression of ERα in silenced as
compared to parent MCF-7 cells is shown. (b) Effect of tamoxifen and 3b on induction of apoptosis in
parental and silenced cells. Parental and ERα silenced MCF-7 cells are assessed for nuclear
condensation following 48 hours treatment with tamoxifen (5.63 µg/ml) and 3b (15, 31.25, 62.5, 125,
250 µg/ml) by Hoechst staining. (c) Data is expressed as Mean±STD of nuclear condensation
expressed as a percentage (n=3) (*P<0.001).
Table 1: Calculated binding free energies (in kcal/mol) for ERα and ERβ with compounds 3(a-d)
using ParDOCK
Table 2: Binding free energies (kcal/mol) calculated using MM-GBSA method for ERα and ERβ
Biphenyl compounds
ERα (3ERT)
-8.00
ERβ (1L2J)
-7.67
3a
3b
3c
3d
-7.50
-7.43
-6.81
-6.48
-6.01
-5.46
This article is protected by copyright. All rights reserved.

Table 2: Binding free energies (kcal/mol) calculated using MM-GBSA method for ERα and ERβ.
ERα-3a
ERα-3b
ERα-3c
ERα-3d
ERβ-3b
Terms
-7.20 (3.68)
-16.44 (4.37) -54.11 (12.59) -14.97 (5.93)
-8.40 (4.20)
-58.39 (2.68)
28.20 (2.33)
-7.86 (0.27)
19.8
∆Gele
-56.15 (3.24) -57.13 (3.47) -50.61 (3.89)
30.91 (3.36) 33.39 (3.31) 62.49 (8.71)
-43.77 (3.09)
36.45 (4.54)
-6.15 (0.39)
21.48
∆GvdW
∆Gpol,solv
∆Gnonpol,solv
∆Gele+∆Gpol,solv
∆GvdW+∆Gnonpol,solv
∆H
-8.07 (0.33)
23.71
-7.52 (0.44)
16.95
-7.64 (0.29)
8.38
-64.22
-64.65
-56.75
-49.92
-66.25
-40.51 (3.21) -47.71 (3.82) -49.87 (5.07)
-23.64 (3.24) -24.08 (4.84) -27.35 (4.53)
-16.87 (4.98) -23.63 (4.72) -22.52 (6.21)
-28.45 (3.64)
-23.36 (4.01)
-5.09 (6.06)
-46.46 (3.41)
-26.05 (5.32)
-20.41 (5.81)
T∆S
∆Gbind
* Standard deviations (STD) are displayed in parentheses.
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.